Discover Where Innovation Can Take Youcanjsurg.ca/wp-content/uploads/2014/10/CJS_56-5.pdf · cma.ca/cjs Laparoscopic versus open surgery for the treatment of colorectal ... Bipolar

© 2013 Olympus America Inc. Trademark or Registered Trademark of Olympus or its affiliate entities in the U.S. and/or other countries of the world. All patents apply.OAIENE0613AD11270

The World’s ONLYArticulating HD 3D Video System

The World’s ONLY

Discover Where Innovation Can Take You

Fully-Integrated Ultrasonic and Advanced Bipolar Technology

For more information please contact your local Olympus Representative or call 1.800.387.0437

A

SPONSORS PARRAINSCanadian Association of General Surgeons

Canadian Society for Vascular SurgeryCanadian Society of Surgical Oncology

Canadian Association of Thoracic Surgeons

Association canadienne des chirurgiens générauxSociété canadienne de chirurgie vasculaireSociété canadienne d’oncologie chirurgicaleAssociation canadienne des chirurgiens thoraciques

Vol. 56, No. 5, October/octobre 2013cma.ca/cjs

Laparoscopic versus open surgery for the treatment of colorectalcancer: a literature review and recommendations

Comparison of the major intraoperative and postoperative complications between unilateral and sequential bilateral TKA

Incremental value and clinical impact of neck sonography for primary hyperparathyroidism: a risk-adjusted analysis

Complications associated with laparoscopic sleeve gastrectomyfor morbid obesity: a surgeon’s guide

Vol. 5

6, N

o. 5

Pages 2

89–360

Octo

ber / o

ctobre 2013

Canadian Jo

urnal o

f Surgery •Jo

urnal ca

nadien de ch

irurgie

covercjs-oct13_Layout 1 13-09-16 8:08 AM Page 1

© 2013 Olympus America Inc. Trademark or Registered Trademark of Olympus or its affiliate entities in the U.S. and/or other countries of the world. All patents apply. OCIAD412-200113

The World’s Only Fully-Integrated Bipolar & Ultrasonic Technology

Groundbreaking Integrated Technology

For more information contact your local Olympus representative or call 888-387-0437 or visit www.olympusamerica.com/thunderbeat.

THUNDERBEAT is the world’s ONLY integration of both advanced bipolar and ultrasonic energies delivered simultaneously from a single, multi-functional instrument. This integration delivers the widely recognized benefits of each individual energy; the ability to rapidly cut tissue with ultrasonic energy, and the ability to create reliable vessel seals with bipolar energy.

THE BENEFITS OF UNPRECEDENTED VERSATILITY� Fastest-in-class cutting speed � Reliable 7 mm vessel sealing � Highest-in-class tip grasping force� Precise dissection with fine jaw design � Minimal thermal spread � Fewer instrument exchanges� Always-available bipolar energy � Reduced mist generation for

for hemostasis without cutting improved visibility

Western University and London Health Sciences Centre/St. Joseph’s Health Care LondonInvite Applications for the Position of

CHAIR/CHIEF, DIVISION OF ORTHOPAEDIC SURGERY, DEPARTMENT OF SURGERYThe Division of Orthopaedic Surgery, Department of Surgery, Schulich School of Medicine & Dentistry at Western University is seeking a Chair/Chief,full-time clinical academic orthopaedic surgeon, at the rank of Associate or Full Professor. The Division of Orthopaedic Surgery has approximately 25 full-time clinical academic surgeons and three scientists located at three Hospital sites. It is the largest Division in the Department of Surgery and consists ofsix clinical teaching units: Adult Reconstruction; Hand and Upper Limb; Paediatric Orthopaedics; Spine; Sport Medicine; and Trauma. The Division has anational and international reputation for teaching, clinical and research excellence.

Candidates must be certified in Orthopaedic Surgery by the Royal College of Physicians and Surgeons of Canada and licensed in the Province of Ontarioand have a strong record of accomplishment in clinical care, research, education, leadership, and administrative experience.

This is a dual position as Chair, Division of Orthopaedic Surgery – Western University; and as Chief of the Division of Orthopaedic Surgery – LondonHealth Sciences Centre & St. Joseph’s Health Care London.

London is home to Western University, with a full-time enrollment of about 32,000 students in a range of academic and professional programs. LondonHealth Sciences Centre and St. Joseph’s Health Care London serve a population base of more than 2 million Ontario residents from approximately 35hospitals within the region. London’s affordable housing, numerous parks, trails, recreational, cultural and sporting opportunities make it an attractive placeto live. London is close to the Great Lakes and is within easy reach of Toronto, Ontario and Detroit, Michigan. Details about the Schulich School of Medicine& Dentistry can be found at www.schulich.uwo.ca; Western University at www.uwo.ca; London Health Sciences Centre at www.lhsc.on.ca; and St.Joseph’s Health Care at www.sjhc.london.on.ca.

Interested candidates should send a letter of interest, curriculum vitae, and the names and addresses of three references to:

John D. Denstedt, MD, FRCSC, FACS, Chair/Chief, Department of SurgerySJHC, 268 Grosvenor Street, Room E3-117, London, Ontario, Canada N6A 4V2

Telephone 519 663-3349 • Fax 519 646-6347

Applications will be accepted until the position is filled. Review of applications will begin after September 1st, 2013.

Positions are subject to budget approval. Applicants should have fluent written and oral communications skills in English. All qualified candidates are encouragedto apply; however, Canadians and permanent residents will be given priority. Western University, London Health Sciences Centre and St. Joseph’s Health CareLondon are committed to employment equity and welcome applications from all qualified women and men, including visible minorities, aboriginal people andpersons with disabilities.

S-746

PAEDIATRIC ORTHOPAEDIC SURGEONIWK Health Centre

The Division of Paediatric Orthopaedic Surgery, Department of Surgery at Dalhousie University and the IWK HealthCentre are seeking a new colleague to join our academic group of Paediatric Orthopaedic Surgeons. The successfulcandidate will have an accredited Fellowship in Paediatric Orthopaedic Surgery and be eligible to practice in Canada.Potential colleagues in all stages of their career are encouraged to apply.

The IWK Health Centre serves a catchment area of nearly two million people in the Maritime Provinces of Canada andis an active participant in the post-graduate training of orthopaedic surgeons. There are ample opportunities for thedevelopment of an academic career in the many paediatric sub-specialty areas.

The candidate would join an established and stable funding environment and participate in the competitive alternatefunding plan which will allow time to pursue research or other academic interests.

Historic Halifax, the business and cultural centre of Atlantic Canada, is home to five Universities and offers an enviablelifestyle opportunity.

All qualified candidates are encouraged to apply; however, Canadians and permanent residents will be given priority.Dalhousie University and the IWK Health Centre are Employment Equity/Affirmative Action employers. The Health Centreand University encourage applications from qualified Aboriginal people, persons with a disability, racially visible personsand women.

Applications accompanied by a curriculum vitae and the names of three referees should be sent to:

Dr. Ron El-Hawary, Head, Division of Paediatric Orthopaedic SurgeryIWK Health Centre, Halifax, Nova Scotia, Canada

5850/5980 University Ave., PO Box 9700, Halifax, NS B3K 6R8Email [email protected]

S-7

47

CAREER/CLASSIFIED ADVERTISING/CARRIÈRES ET ANNONCES CLASSÉES


© 2013 Canadian Medical Association Can J Surg, Vol. 56, No. 5, October 2013 289

EDITORIAL • ÉDITORIAL

293 Resident work conditions under the microscopeE.J. Harvey

294 Les conditions de travail des résidents sous le microscopeE.J. Harvey

COMMENTARY • COMMENTAIRE

295 A new era for resident duty hours in surgerycalls for greater emphasis on resident wellnessK. Imrie, J.R. Frank, N. Ahmed, L. Gorman, K.A. Harris

RESEARCH • RECHERCHE

297 Laparoscopic versus open surgery for the treatment of colorectal cancer: a literaturereview and recommendations from theComité de l’évolution des pratiques enoncologieM. Morneau, J. Boulanger, P. Charlebois, J.-F. Latulippe, R. Lougnarath, C. Thibault, N. Gervais; Comité de l’évolution des pratiques en oncologie

311 Comparison of the major intraoperative andpostoperative complications between unilateral and sequential bilateral total kneearthroplasty in a high-volume communityhospitalE. Spicer, G.R. Thomas, E.J. Rumble

318 Impact of a regional acute care surgery modelon patient access and outcomesS.A. Kreindler, L. Zhang, C.J. Metge, R.W. Nason, B. Wright, W. Rudnick, M.E.K. Moffatt

325 Incremental value and clinical impact of necksonography for primary hyperparathyroidism: a risk-adjusted analysisM.C. Tee, S.K. Chan, V. Nguyen, S.S. Strugnell, J. Yang, S. Jones, P. Tiwari, D.S. Levine, S.M. Wiseman

332 Total hip arthroplasty using a combined anterior and posterior approach via a lateralincision in patients with ankylosed hipsJ. Li, Z. Wang, M. Li, Y. Wu, W. Xu, Z. Wang

341 Clinical outcomes compared between laparoscopic and open appendectomy in pregnant womenJ.C. Chung, G.S. Cho, E.J. Shin, H.C. Kim, O.P. Song

ONLINE RESEARCH • RECHERCHE EN LIGNE

E105 Adding an endovascular aortic surgery program to a rural regional medical centreD. Marelli, E. Woo, M. Watson, P. Fedalen, G. Wang,M. Stallings, R. Fairman, J. Mannion

E114 Development and preliminary validation of a Function IndeX for Trauma (FIX-IT)M. Bhandari, S.M. Wasserman, N. Yurgin,B. Petrisor, S. Sprague, R.E. Dent

E121 Quality of narrative operative reports in pancreatic surgeryM.E. Wiebe, L. Sandhu, J.L. Takata, E.D. Kennedy,N.N. Baxter, A.R. Gagliardi, D.R. Urbach, A.C. Wei

E128 The impact of a massive transfusion protocol(1:1:1) on major hepatic injuries: Does itincrease abdominal wall closure rates?C.G. Ball, C.J. Dente, B. Shaz, A.D. Wyrzykowski,J.M. Nicholas, A.W. Kirkpatrick, D.V. Feliciano


cma.ca/cjs • cma.ca/jcc

contentscjs-oct13_Layout 1 13-09-16 11:09 AM Page 289



The World’s ONLY




A contentscjs-oct13_Layout 1 13-09-16 11:10 AM Page 290


E135 Differences between referred and nonreferredpatients in cancer researchJ. Faulds, C.E. McGahan, P.T. Phang, M.J. Raval, C.J. Brown

REVIEW • REVUE

347 Complications associated with laparoscopicsleeve gastrectomy for morbid obesity: a surgeon’s guideK. Sarkhosh, D.W. Birch, A. Sharma, S. Karmali

CONTINUING MEDICAL EDUCATIONFORMATION MÉDICALE CONTINUE

EVIDENCE-BASED SURGERY

353 Cost-effectiveness of bariatric surgery forseverely obese adults with diabetesH.J. Henteleff, D.W. Birch, P.T. Hallowell; EvidenceBased Reviews in Surgery Group

E103 ONLINE CORRESPONDENCECORRESPONDANCE EN LIGNE

356 INSTRUCTIONS FOR AUTHORS

358 DIRECTIVES AUX AUTEURS

360 SERVICE INFORMATION RENSEIGNEMENTS AUX LECTEURS

IBC CAREER/CLASSIFIED ADVERTISINGANNONCES SUR LES CARRIÈRES ET ANNONCESCLASSÉS

© 2013 Canadian Medical Association. ISSN 0008-428X. For informa-tion on permission to reproduce material from the Can adian Journal ofSurgery (CJS) see Service Information or cma.ca/cjs.

All editorial matter in CJS represents the opinions of the auth ors andnot necessarily those of the Canadian Medical Association (CMA). TheCMA does not assume any responsibility or lia bil ity for damages aris-ing from any error or omission or from the use of any information oradvice contained in CJS, including articles, editorials, reviews, lettersand advertisements. All reproduction rights are reserved.

Printed by Dollco Integrated Print Solutions, Ottawa. Appears in February,April, June, August, October and December.

Return undeliverable Canadian copies to the CMA Member Service Centre, 1870 Alta Vista Dr., Ottawa ON K1G 6R7 (email [email protected]).

© 2013 Association médicale canadienne. ISSN 0008-428X. Pourobtenir des renseignements au sujet des permissions à obtenir afin dereproduire des extraits du Journal canadien de chirurgie (JCC), con-sulter les Renseignements aux lecteurs ou en direct à l’adresse cma.ca /cjs-f.

Tous les articles à caractère éditorial dans le JCC représentent les opi -nions de leurs auteurs, qui ne sont pas nécessairement celles de l’Asso-ciation médicale canadienne (AMC). L’AMC n’assume aucune respon -sabilité pour les dommages résultant de toute erreur ou omission, ou del’utilisation de renseignements ou de conseils contenus dans le JCC, ycompris les articles, éditoriaux, revues, lettres et annon ces. Tous droitsde reproduction réservés.

La revue est imprimée par Dollco Integrated Print Solutions, Ottawa.Elle paraît en février, avril, juin, août, octobre et décembre.

Retournez toutes copies canadiennes non livrées au Centre des ser -vices aux membres, Association médicale canadienne, 1870, prom. AltaVista, Ottawa (Ontario) K1G 6R7 (courriel : cmamsc @cma.ca).

contentscjs-oct13_Layout 1 13-09-16 11:10 AM Page 291

292 J can chir, Vol. 56, No 5, octobre 2013 © 2013 Association médicale canadienne

COEDITORSCORÉDACTEURS

Edward J. Harvey, MD, Montré[email protected]

Major Vivian C. McAlister, MB, [email protected]

ASSOCIATE EDITORSRÉDACTEURS ASSOCIÉS

BASIC SCIENCE AND SURGICAL BIOLOGYIan McGilvray, MD, PhD, TorontoTimothy Daniels, MD, Toronto

BREAST SURGERYMuriel Brackstone, MD, London

CARDIOVASCULAR SURGERYMichel Carrier, MD, MontréalMichael Chu, MD, London

CRITICAL CARERaymond Kao, MD, London

ENDOCRINE SYSTEMSam Wiseman, MD, Vancouver

EVIDENCE-BASED MEDICINEMichelle Ghert, MD, Hamilton

GASTROINTESTINAL AND COLORECTAL SURGERYMarcus Burnstein, MD, Toronto

GLOBAL SURGERYDan Deckelbaum, MD, Montreal

GENERAL SURGERYDaniel Birch, MD, Edmonton

HEPATOBILIARY AND PANCREATIC SURGERYShiva Jayaraman, MD, Toronto

MEDICAL EDUCATIONCarol Hutchison, MD, Calgary

MILITARY MEDICINEHomer Tien, MD, Toronto

ORTHOPEDIC FOOT AND ANKLE SURGERYKarl-André Lalonde, MD, Ottawa

ORTHOPEDIC SURGERYGraham Elder, MD, Sault Ste. Marie

PEDIATRIC SURGERY, GENERALSigmund Ein, MD, Toronto

PEDIATRIC SURGERY, ORTHOPEDICJames G. Wright, MD, MPh, Toronto

SPINAL SURGERYRaja Rampersaud, MD, Toronto

SPORTS MEDICINEPaul Martineau, MD, Montréal

SURGICAL ONCOLOGYDebrah A. Wirtzfeld, MD, Winnipeg

SURGICAL ONCOLOGY, MUSCULOSKELETALFrank O’Dea, MD, St. John’s

SURGICAL ONCOLOGY, SOFT TISSUEChris de Gara, MB MS, Edmonton

THORACIC SURGERYJohn Yee, MD, Vancouver

TRAUMA SURGERY, ORTHOPEDICWilliam Dust, MD, Saskatoon

TRAUMA SURGERY, SOFT TISSUEMary vanWijngaarden-Stephens, MD, EdmontonMorad Hameed, MD, Vancouver

VASCULAR SURGERYKent Mackenzie, MD, Montréal

The Canadian Journal of Surgery aims to contribute to the effective continuing medical education of Canadiansurgical specialists, and to provide surgeons with an effective vehicle for the dissemination of observationsin the areas of clinical, basic science and education research.Readers can find CJS online at cma.ca/cjs.Submission of new manuscripts can be made at http://mc.manuscript central.com/cjs.

Le Journal canadian de chirurgie vise à dispenser une éducation médicale continue efficace aux spécialistesen chirurgie au Canada, et fournir aux chirurgiens un mécanisme efficace pour diffuser les constatations de larecherche clinique, fondamentale et éducative.Les lecteurs trouveront en direct le JCC à l'adresse cma.ca/cjs-f.Nous favorisons l'envoi électronique de manuscrits. Veuillez visiter le http://mc.manuscriptcentral.com/cjs.

EDITORIAL STAFFÉQUIPE DE RÉDACTION

MANAGING EDITORDIRECTRICE DE LA RÉDACTIONWendy Carroll ([email protected])tel 800 663-7336 (Canada, USA)or 613 731-8610 x2198; fax 613 565-5471

EDITORIAL COORDINATORCOORDONNATRICE DE LA RÉDACTIONElisa Makadi ([email protected])tel 800 663-7336 (Canada, USA)613 731-8610 x2018; fax 613 565-5471

TRANSLATION • TRADUCTIONMarie Saumure

PRODUCTION STAFFÉQUIPE DE PRODUCTION

PRODUCTION MANAGERGESTIONNAIRE DE LA PRODUCTIONKathryn A. Freamo

PRODUCTION DESIGNERSINFOGRAPHISTES DE LA PRODUCTIONCarole Lalonde, Clara Walker

WEB PUBLISHERS ÉDITION EN LIGNEJames Harbinson, James Manship

ADVERTISINGPUBLICITÉ

ADVERTISING ASSISTANTSADJOINTES, ANNONCES PUBLICITAIRESLaurie McLeod, Susan Ritchie, DeborahWoodman613 731-8610 x2107/ 2041 fax 613 [email protected]

ADVERTISING SALESVENTES PUBLICITAIRESBlair Graham ([email protected])416 850-1485; cell 416 560-0202; fax 416 850-2242

Trish Sullivan ([email protected])cell 905 330-8770

PUBLISHING STAFFÉQUIPE DES SERVICES D’ÉDITION

DIRECTOR AND PUBLISHER, CMA PUBLICATIONSDIRECTRICE ET ÉDITRICE, PUBLICATIONS AMCGlenda Proctor

ASSOCIATE DIRECTOR, PUBLICATIONSDIRECTRICE ASSOCIÉE, PUBLICATIONSHolly Bodger

PROGRAM MANAGER, JOURNALSGESTIONNAIRE DE PROGRAMMES, JOURNAUXJanis Murrey

CMA EXECUTIVES DIRIGEANTS DE L'AMC

PRESIDENT • PRÉSIDENTLouis Francescutti

SECRETARY GENERAL AND CEOSECRÉTAIRE GÉNÉRAL ET CHEF DE LA DIRECTIONPaul-Émile Cloutier

VP, PROFESSIONAL SERVICES AND LEADERSHIPVP, SERVICES PROFESSIONNELS ET LEADERSHIPTim Smith

All prescription drug advertisements have beencleared by the Pharmaceutical Advertising AdvisoryBoard.Toutes les annonces de mé dicaments prescrits ontété approuvés par le Conseil consultatif de publicitépharmaceutique.

Owned and published by CMA1867 prom. Alta Vista Dr.Ottawa ON K1G 5W8

mastheadcjs-oct13_Layout 1 13-09-16 10:56 AM Page 292



Resident work conditions under the microscope

T his edition of CJS has a commentary by Imrie andcolleagues1 from the Royal College of Physiciansand Surgeons of Canada referencing the report

Fatigue, Risk and Excellence: Towards a Pan-Canadian Consen-sus on Resident Duty Hours2 that has recently been releasedby the National Steering Committee on Resident DutyHours. A large group of engaged physicians and otherexperts have reviewed literature on the effect of residentduty hour restrictions in relation to patient safety, residentwellness, training and educational outcomes and the educa-tional needs of surgeons. The working group highlightedevidence suggesting bad outcomes in care delivery andtraining — particularly for surgical specialties — whentotal or consecutive hours of resident duty are mandated.Restricting the number of hours, which in reality may be areflection of an increase in the number of handovers inpatient care, seems to have resulted in some bad outcomes.Imrie and colleagues comment on recent studies thatreported more complications in high-acuity surgicalpatients and increasing failures in certification examina-tions with rigid resident hours.

The actual conclusions of this report, as I interpret it,are that a tired doctor may not necessarily be a bad doctor,that there are no conclusive data to show that restrictionson consecutive resident duty hours are necessary for pa -tient safety, that there is no clear evidence that residentduty hour hours have an overall cross-medicine impact onacademic performance, and that there is evidence suggest-ing suboptimal patient care and educational outcomes insurgery resulting from the restriction of resident dutyhours. There have been observations that resident wellnesssuffered owing to longer hours under previous trainingschemes and that any changes in hours need to incorporate“other” metrics. The National Steering Committee onResident Duty Hours report is timely, as we enter a phaseof determining what is appropriate training or even whatdefines a normal practice after residency. We are certainlystruggling at our national association with the definition ofwhat constitutes a full-time staff member. We are hardpressed to show what constitutes an average call scheduleor operating room rotation across Canada. It does seemthat most staff surgeons are putting in long hours anddoing a fair amount of call.

We are not the only people interested in what defines anormal workload for physicians. Medical working condi-tions remain a hot topic in the popular media. A recent(Aug. 20, 2013) blog post from the New Yorker,3 “Why

doesn’t medical care get better when doctors rest more?,”by Dr. Lisa Rosenbaum examined her personal experience.In the article, she, and indirectly her physician mother,recount their take on the enforced resident hours and theattendant physician errors they are experiencing. She exam-ines some of the same literature as the National WorkingGroup on Resident Duty Hours and reaches the conclusionthat patient wellness has to come first, because that is whathappens in practice after residency. Closer to home,Maclean’s ran a story on a similar subject with an emphasison the fact that changing the hours worked might haveincreased errors.4

We definitely need to look at how these observationsand this landmark report should change the way we designtraining programs, especially in surgery. I think adding yetanother novel layer, such as competency-based training, ontop of restricted resident hours will not mitigate the short-comings of the current system. But something has to bereconfigured. It may be that we need better studies andthat the current results reflect inherent bias. However, weneed to step up and do something quickly because it seemsthe patients may be paying a price for our new programs.

Edward J. Harvey, MDCoeditor, Canadian Journal of Surgery

Competing interests: None declared.

DOI: 10.1503/cjs.021713

References

1. Imrie K, Frank J.R., Ahmed N, et al. A new era for resident dutyhours in surgery calls for greater emphasis on resident wellness. CanJ Surg 2013;56:295-6.

2. National Steering Committee on Resident Duty Hours. Fatigue, Riskand Excellence: Towards a Pan-Canadian Consensus on Resident Duty Hours.Ottawa (ON): Royal College of Physicians and Surgeons of Canada;2013. Available: www.residentdutyhours.ca/documents/fatigue _risk_and _excellence.pdf (accessed 2013 Aug. 12).

3. Rosenbaum L. Why doesn’t medical care get better when doctors restmore? The New Yorker 2013 Aug. 20. Available: www .newyorker .com/online /blogs/elements/2013/08/hospital-residency-hour -limits-problem-with-medical-care.html (accessed 2013 Aug. 20).

4. Belluz J. Did reducing hours for residents increase medical errors?MacLean’s 2013 Mar. 31. Available: http://www2.macleans .ca/2013 /03/31 /did-reducing-resident-hours-increase-medical-errors/ (accessed2013 Aug. 12).

edit-oct13_Layout 1 13-09-16 9:56 AM Page 293



Les conditions de travail des résidents sous lemicroscope

D ans ce numéro du JCC figure un commentaire duDr Imrie et de ses collaborateurs1 du Collège royaldes médecins et chirurgiens du Canada concernant

le rapport Fatigue, risque et excellence : À la recherche d’un con-sensus pancanadien sur les heures de travail des résidents2 récem-ment publié par le Comité directeur national sur les heuresde travail des résidents. Un grand nombre de médecins etautres experts ont passé en revue la littérature sur l’effet desrestrictions des heures de travail des résidents sur la sécuritédes patients et le mieux-être des résidents, ainsi que sur lesrésultats de la formation des chirurgiens et les besoins à cetégard. Le groupe de travail a mis en évidence des preuvessuggérant que lorsque les heures de travail totales ou con-sécutives des résidents sont imposées, cela entraîne de mau-vais résultats dans la prestation de soins et la formation —en particulier pour les spécialités chirurgicales. Limiter lenombre d’heures, ce qui en réalité peut refléter une aug-mentation du nombre de transferts des soins, semble avoirdonné de mauvais résultats. Le Dr Imrie et ses collabora-teurs commentent de récentes études qui ont signalé unnombre accru de complications chez les patients à la suited’interventions chirurgicales d’urgence et des taux de plusen plus élevés d’échec des examens de certification en raisondes heures de travail inflexibles des résidents.

Selon mon interprétation, les conclusions réelles de ce rap-port sont qu’un médecin fatigué n’est pas nécessairement unmauvais médecin, qu’il n’y a pas de données concluantes pourprouver qu’il est nécessaire de restreindre les heures de travailconsécutives des résidents pour assurer la sécurité des patients,qu’il n’y a pas de données claires démontrant que la réglemen-tation des heures de travail des résidents a une incidence néga-tive sur les résultats scolaires dans leur ensemble, et qu’il existedes preuves suggérant que la restriction des heures de travaildes résidents a un effet négatif sur le soin des patients et lerendement académique en chirurgie. On a observé que leslongues heures passées dans les programmes de formationanté rieurs avaient eu un effet négatif sur le mieux-être desrésidents et que tout changement dans les heures de travaildoit intégrer « d’autres » mesures. Le rapport du Comitédirecteur national sur les heures de travail des résidents arriveà point nommé, car nous entreprenons une phase de détermi-nation de ce qui constitue une formation appropriée, voiremême ce qui constitue une pratique normale après la rési-dence. Notre association nationale a certes de la difficulté àdéfinir ce qu’est un membre du personnel à temps plein. Ilnous est difficile de démontrer ce qui constitue un horaire degarde moyen ou une rotation moyenne en salle d’opérationdans l’ensemble du Canada. Il semble que la plupart deschirurgiens membres du personnel travaillent de longuesheures et font beaucoup d’heures de garde.

Nous ne sommes pas les seuls à vouloir définir ce quiconstitue une charge normale de travail pour les médecins.Les conditions de travail des médecins demeurent un sujet

brûlant dans les médias. Dans un récent billet de blogue (le20 août 2013) du New Yorker3 intitulé « Why doesn’t med-ical care get better when doctors rest more? », la Dre LisaRosenbaum a examiné son expérience personnelle. Dansson billet, elle, et indirectement sa mère médecin, racon-tent leur point de vue sur les horaires imposés aux résidentset les erreurs des médecins de garde qui en découlent. Elleexamine certains des mêmes documents que le Comitédirecteur national sur les heures de travail des résidents eten vient à la conclusion que le bien-être du patient doitpasser en premier, parce que c’est ce qui prévaut dans lapratique après la résidence. Ici au Canada, Maclean’s a pub-lié un article sur un sujet similaire, dans lequel l’accent étaitmis sur le fait que le changement des heures travailléesaurait en fait augmenté le nombre d’erreurs4.

Nous devons absolument regarder comment ces obser-vations et ce rapport déterminant devraient changer lafaçon dont nous concevons des programmes de formation,en particulier dans les disciplines chirurgicales. Je pensequ’ajouter un autre élément, tel que la formation axée surles compétences, à la restriction des heures de travail desrésidents n’atténuera pas les lacunes du système actuel.Mais il faut reconfigurer quelque chose. Il se peut que nousdevions réaliser de meilleures études et que les résultatsactuels reflètent un biais inhérent. Nous devons cependantagir sans plus tarder, car il semble que les patients payentun prix pour nos nouveaux programmes.

Edward J. Harvey, MDCo-rédacteur, Journal canadien de chirurgie

Intérêts concurrents: Aucuns déclarés.

DOI: 10.1503/cjs.021913

Références

1. Imrie K, Frank J.R., Ahmed N, et al. A new era for resident dutyhours in surgery calls for greater emphasis on resident wellness. CanJ Surg 2013;56:295-6.

2. Comité directeur national sur les heures de travail des résidents.Fatigue, risque et excellence : À la recherche d’un consensus pancanadien surles heures de travail des résidents. Ottawa (Ont.) Collège royal desmédecins et chirurgiens du Canada; 2013. Disponible ici : www.heuresdetravaildesresidents.ca/documents/fatigue_risque_et_excellence.pdf (consulté le 12 août 2013).

3. Rosenbaum L. Why doesn’t medical care get better when doctors restmore? The New Yorker, le 20 août 2013. Disponible ici : www .newyorker.com /online/blogs/elements/2013/08/hospital-residency -hour -limits-problem -with-medical-care.html (consulté le 20 août 2013).

4. Belluz J. Did reducing hours for residents increase medical errors?Maclean’s, le 31 mars 2013. Disponible ici : http://www2 .macleans .ca/2013 /03/31/did-reducing-resident-hours-increase-medical-errors/(consulté le 12 août 2013).

edit-oct13_Layout 1 13-09-16 9:56 AM Page 294


COMMENTARY • COMMENTAIRE

A new era for resident duty hours in surgery callsfor greater emphasis on resident wellness

I n Canada, physicians and surgeons in training have dual roles as learnersand care providers. While they have graded responsibility for the provisionof patient care, they are also training for independent practice. Over the

last several decades and against the backdrop of rapid changes in patient careand surgical practice, there has been ongoing discussion and evolution in theresidency education system and in residents’ schedules. Historically, residentsliterally resided within the hospitals in which they provided care and receivedtraining. Today’s health care and medical education systems are marked byshifts and broader changes in the characteristics of patient care, workflow andscheduling of residents and by an increasing professional diversity in thehealth care team. While residents of the 21st century do not live in the hospi-tals in which they work and train, they are nevertheless key members of healthcare delivery teams in our busiest hospitals, and it is in these hospitals that fac-ulty members and training programs are tasked with ensuring the safe matura-tion of residents into independent physicians and surgeons.

The ongoing dialogue surrounding resident duty hours has led to animportant debate in residency education worldwide. Until recently, there hasbeen no cohesive approach, inclusive of multiple stakeholders, toward issuessurrounding the regulation of resident duty hours in Canada. However, overthe last 16 months, a new era of the resident duty hours debate has begun.Since March 2012, supported with funding from Health Canada, concertedefforts have been made to arrive at a pan-Canadian consensus on this impor-tant topic. This work has resulted in an unprecedented report, entitled Fatigue,Risk and Excellence: Towards a Pan-Canadian Consensus on Resident Duty Hours,launched earlier this month.1

The project was directed by a senior consortium, the National SteeringCommittee on Resident Duty Hours, comprising stakeholders from Canadianpostgraduate medical education and representatives from collaborating part-ner organizations, including the Association of Canadian Academic HealthcareOrganizations; the Association of Faculties of Medicine of Canada; the Canad -ian Association of Internes and Residents; the Canadian Medical Association;the Collège des Médecins du Québec; the Federal, Provincial, and TerritorialCommittee on Health Workforce; the Fédération des Médecins Résidents duQuébec; and the Royal College of Physicians and Surgeons of Canada.

From the time of the project’s inception, the National Steering Committeerecognized the importance of identifying and addressing the challenges posedby resident duty hour regulations for training in surgical and procedural disci-plines. To do so, the committee established a working group to review andcompile the evidence related to the unique issues and challenges faced in sur-gical and procedural disciplines and the training of residents within such spe-cialties. Chaired by a trauma and acute care surgeon who is the program direc-tor of the general surgery residency training program at the University ofToronto (N.A.), the group reviewed a large volume of research related to theeffect of resident duty hour restrictions on the inter-related domains of patientsafety, resident wellness and training and educational outcomes, as well as the

Kevin Imrie, MD*†

Jason R. Frank, MD, MA(Ed)*‡

Najma Ahmed, MD, PhD†

Lisa Gorman, MA*

Kenneth A. Harris, MD*‡

From the *Royal College of Physiciansand Surgeons of Canada, Ottawa, Ont.,the †University of Toronto Faculty ofMedicine, Toronto, Ont., and the ‡Univer-sity of Ottawa, Faculty of Medicine,Ottawa, Ont.

Accepted for publicationJuly 16, 2013

Correspondence to:K. Harris774 Echo Dr.Ottawa ON K1S [email protected]

DOI: 10.1503/cjs.017713

era-imrie_Layout 1 13-09-16 9:47 AM Page 295

296 J can chir, Vol. 56, No 5, octobre 2013

COMMENTAIRE

varying educational needs of surgeons who provide patientcare across the large and diverse landscape of Canada. Inaddition to the Special Considerations for Procedural Dis-ciplines working group, 5 other expert working groups(Patient Safety, Medical Education, Resident and FacultyHealth and Wellness, Professionalism, and Health SystemsPerformance and Health Economics) were created toexplore themes across all medical disciplines.

One of the key findings articulated in the project’s finalreport pertains to the unique challenges faced by the sur gic -al and procedural disciplines in an environment of strictlyregulated resident duty hours. The literature reviewed bythe Special Considerations for Procedural Disciplinesworking group highlighted a body of evidence suggestingsuboptimal outcomes in surgical care delivery and surgicaltraining when total or consecutive hours of resident dutyare rigidly restricted, such as they have been in the UnitedStates and the European Union. Recent studies suggestincreased complications in high-acuity surgical patients anddeclining performance on some certification examinationsas a result of rigidly controlled resident duty hours.2–9 Theworking group’s analysis of the literature led to their con-clusion that there is a vital need for more research on theeffect of duty hours on surgical and procedural skills acqui-sition and performance and for the development of trainingand care delivery paradigms appropriate for higher-acuitypatient care areas, where such research findings more fre-quently emerge. It would seem appropriate that proceduraland surgical training programs work collaboratively with allstakeholder organizations to develop strategies that supportexcellent training outcomes, resident wellness and patientsafety among these disciplines.

The final project report outlines a series of 5 principlesand 5 recommendations for the future direction of residentduty hours in Canada with implications for all disciplinesand residency programs. Rather than outlining specificrestrictions regarding a set number of hours or shift length,chief in the National Steering Committee proposal is theneed for a comprehensive, rigorous and tailored approachto the management and mitigation of physician and sur-geon fatigue and burnout. Neither role — learner or healthcare provider — is well served if un checked fatigue impairscognitive functions or threatens professional satisfaction orhealth.

The Royal College of Physicians and Surgeons ofCanada is dedicated to facilitating the implementation andsupporting the launch of all recommendations proposed bythe National Steering Committee across all residency pro-

grams and all disciplines. Such changes focus on fatiguemitigation and stress management during residency andhighlight strategies that could involve new models of sched-uling; greater protection for sleep at night; adjustments toresident workload; and broader, educational innovations,such as competency-based training and evaluation, thatecho changes across the continuum of medical education.Also included as a key recommendation is the proposeddevelopment of a national consortium to study and dis -seminate new knowledge related to resident duty hour regu -lation. These and other innovations will support a transitionto a new era of resident duty hours that maximizes educa-tion and patient care outcomes in the 21st century.

References

1. National Steering Committee on Resident Duty Hours. Fatigue, Riskand Excellence: Towards a Pan-Canadian Consensus on Resident DutyHours. Ottawa (ON): Royal College of Physicians and Surgeons ofCanada; 2013. Available: www .residentdutyhours .ca /documents/fatigue _risk _and _excellence .pdf (accessed 2013 Aug 12).

2. Browne JA, Cook C, Olson SA, et al. Resident duty‐hour reformassociated with increased morbidity following hip fracture. Bone JointSurg Am 2009;91:2079-85.

3. Dumont TM, Rughani AI, Penar PL, et al. Increased rate of compli-cations on a neurological surgery service after implementation of theAccreditation Council for Graduate Medical Education work-hourrestriction. J Neurosurg 2012;116:483-6.

4. Gopaldas RR, Chu D, Dao TK, et al. Impact of ACGME work‐hourrestrictions on the outcomes of coronary artery bypass grafting in acohort of 600,000 patients. J Surg Res 2010;163:201-9.

5. Hoh BL, Neal DW, Kleinhenz DT, et al. Higher complications andno improvement in mortality in the ACGME resident duty‐hourrestriction era: an analysis of more than 107000 neurosurgical traumapatients in the Nationwide Inpatient Sample database. Neurosurgery2012; 70:1369-81.

6. Jamal MH, Doi SA, Rousseau M, et al. Systematic review and meta-analysis of the effect of North American working hours restrictionson mortality and morbidity in surgical patients. Br J Surg 2012;99:336-44.

7. Kaderli R, Businger A, Oescha A, et al. Morbidity in surgery: impactof the 50-hour work-week limitation in Switzerland. Swiss Med Wkly2012; 142:w13508.

8. Lewis F, Klingensmith M. Issues in general surgery residency training —2012. Ann Surg 2012;256:553-9.

9. Poulose BK, Ray WA, Arbogast PG, et al. Resident work hour limitsand patient safety. Ann Surg 2005;241:847-56.

era-imrie_Layout 1 13-09-16 9:47 AM Page 296



Laparoscopic versus open surgery for thetreatment of colorectal cancer: a literature reviewand recommendations from the Comité del’évolution des pratiques en oncologie

Background: Adoption of the laparoscopic approach for colorectal cancer treatmenthas been slow owing to initial case study results suggesting high recurrence rates atport sites. The use of laparoscopic surgery for colorectal cancer still raises a number ofconcerns, particularly with the technique’s complexity, learning curve and longer dura-tion. After exploring the scientific literature comparing open and laparoscopic surgeryfor the treatment of colorectal cancer with respect to oncologic efficacy and short-term outcomes, the Comité de l’évolution des pratiques en oncologie (CEPO) maderecommendations for surgical practice in Quebec.

Methods: Scientific literature published from January 1995 to April 2012 wasreviewed. Phase III clinical trials and meta-analyses were included.

Results: Sixteen randomized trials and 10 meta-analyses were retrieved. Analysis ofthe literature confirmed that for curative treatment of colorectal cancer, laparoscopy isnot inferior to open surgery with respect to survival and recurrence rates. Moreover,laparoscopic surgery provides short-term advantages, including a shorter hospital stay,reduced analgesic use and faster recovery of intestinal function. However, this ap -proach does require a longer operative time.

Conclusion: Considering the evidence, the CEPO recommends that laparoscopicresection be considered an option for the curative treatment of colon and rectal can-cer; that decisions regarding surgical approach take into consideration surgeon experi-ence, tumour stage, potential contraindications and patient expectations; and thatlaparoscopic resection for rectal cancer be performed only by appropriately trainedsurgeons who perform a sufficient volume annually to maintain competence.

Contexte : L’adoption de la laparoscopie pour traiter le cancer colorectal se fait lente-ment à cause des résultats des premières études de cas qui indiquent des taux élevés derécidive aux sites d’intervention. La laparoscopie pour traiter le cancer colorectal soulèvetoujours de nombreuses préoccupations, particulièrement en raison de la complexité de latechnique, de la courbe d’apprentissage, et de la durée de la chirurgie. Après avoir étudiédes publications scientifiques comparant l’efficacité oncologique et les résultats à courtterme de la laparoscopie à ceux de la chirurgie ouverte pour le traitement du cancer colo -rectal, le Comité de l’évolution des pratiques en oncologie (CEPO) a formulé des recom-mandations pour la pratique chirurgicale au Québec.

Méthodes : Une revue des écrits scientifiques publiés entre janvier 1995 et avril 2012 aété effectuée. Seuls les essais cliniques de phase III et les méta-analyses ont été répertoriés.

Résultats : Seize essais randomisés et 10 méta-analyses ont été retenus. L’analyse despublications a confirmé que pour le traitement curatif du cancer colorectal, la laparoscopien’est pas inférieure à la chirurgie ouverte pour ce qui est des taux de survie et de récidive.La laparoscopie offre de plus des avantages à court terme, y compris une hospitalisation demoins longue durée, une réduction de l’usage d’analgésiques et un rétablissement plusrapide de la fonction intestinale. Cette intervention prend toutefois plus de temps.

Conclusion: Compte tenu des données probantes, le CEPO recommande d’envisager larésection laparoscopique comme technique curative possible du cancer colorectal et queles décisions sur la méthode chirurgicale tiennent compte de l’expérience du chirurgien, dustade de la tumeur, des contre-indications possibles et des attentes du patient. Dans le casde la résection laparoscopique du cancer du rectum, le CEPO recommande qu’elle ne soitpratiquée que par des chirurgiens ayant reçu la formation nécessaire et qui pratiquent suf -fi samment d’interventions par année pour maintenir leur compétence.

Mélanie Morneau, MSc*

Jim Boulanger, PhD†

Patrick Charlebois, MD‡

Jean-François Latulippe, MD§

Rasmy Lougnarath, MD¶

Claude Thibault, MD**

Normand Gervais, MD††

For the Comité de l’évolutiondes pratiques en oncologie

From the *Direction québécoise du can-cer, Ministère de la Santé et des Servicessociaux du Québec (MSSS), the †Institutnational d’excellence en santé et ser-vices sociaux (INESS), the ‡McGill Uni-versity Health Centre (MUHC), MontrealGeneral Hospital, the §Hôpital Maison-neuve-Rosemont, the ¶Centre hospitalieruniversitaire de Montréal (CHUM), Hôpi-tal Saint-Luc, the **Centre hospitalieruniversitaire de Québec (CHUQ), HôpitalSaint-François d’Assise, and the ††Centrede Santé et de Services sociaux (CSSS)de Rivière-du-Loup, Centre hospitalierrégional du Grand-Portage. See end ofarticle for committee members.

Accepted for publicationJuly 24, 2012

Correspondence to:N. GervaisCentre hospitalier régional du Grand-Portage75, rue Saint-HenriRivière-du-Loup QC G5R [email protected]

DOI: 10.1503/cjs.005512

lapa-morneau_Layout 1 13-09-17 8:16 AM Page 297


RECHERCHE

C olorectal cancer is the third most commonly diag-nosed cancer and the second leading cause of cancer-related death in Canada. The Canadian Cancer

Society estimated that approximately 23 800 new colorec-tal cases will be diagnosed in Canada in 2013 (6300 inQuebec), and that 9200 related deaths will be reported(2450 in Quebec).1

Surgery is the only curative treatment for colorectal can-cer. Curative surgery requires resection of the primarytumour with negative margins and a complete oncologiclymphadenectomy. The resected colic segment depends onvascularization and lymphatic drainage at the tumour siteand, according to the American Joint Committee on Cancer,a minimum of 12 lymph nodes should be retrieved in sur -gical specimens. Otherwise, tumour stage could be under -estimated, and a suboptimal treatment could be offered.2

The surgical approach for rectal cancer is affected bytumour stage and localization.3 Generally, 5 types of resec-tion can be performed: local excision of the tumour, anter -ior resection of the rectum, proctectomy with coloanalanastomosis or with terminal colostomy (Hartmann), andabdominoperineal resection. A major improvement in sur-gical technique was achieved in 1982 when total mesorectalexcision was first described and resulted in a substantialreduction in recurrence rates.4

Traditionally, colorectal cancer resection has been per-formed exclusively through open surgery. However, fol -lowing successful laparoscopic procedures, such as chole -cystectomy, appendectomy and treatment of incisionalhernias, this surgical approach has gradually been intro-duced first in the treatment of colon cancer and then in thetreatment of rectal cancer.5 Surgical preparation forlaparoscopy is similar to that for open surgery. However,small lesions need to be carefully localized preoperativelyby means of colonoscopy (with a tattoo or clip) or bariumenema. Laparoscopic resection should result in theremoval of the colon or rectal segment containing thetumour and associated lymphatic drainage to the sameextent as open surgery. Surgery can be performed entirelyby laparoscopy, be laparoscopy-assisted (anastomosis isthen performed extracorporally) or be hand-assisted (inwhich case a sufficiently long incision is made to allow thesurgeon’s hand to enter the abdominal cavity). For all3 strategies, the abdominal wall incision should be pro-tected to prevent tumour dissemination.6

Laparoscopic resection of the sigmoid colon was firstdescribed by Jacobs and colleagues7 in 1991. However,generalized adoption of the laparoscopic procedure wasslower for colorectal cancer than for other pathologies.This can be explained by the disappointing results of initialcase studies on laparoscopic colon cancer resection, whichrevealed high recurrence rates at port sites.8,9 Despite thefact that more recent studies did not reproduce theseresults,10,11 many concerns still persist about the use oflaparoscopic surgery in colon and rectal cancer treatment,

notably with respect to the technique’s complexity, theassociated learning curve and the longer operative time.5

About half the general surgeons in Canada performlaparoscopic colorectal surgery. The highest rate of adop-tion of laparoscopic surgery is in Quebec, where it is esti-mated at 67%. Significant predictors for offering a laparo-scopic approach are recent graduation, male sex, provinceof practice, university affiliation and minimally invasivesurgery training, whereas constraints for adoption of thistechnique include the lack of available operative time andlack of formal training programs.12

The present review explores the relevant scientific liter-ature comparing open and laparoscopic surgery in thetreatment of colon and rectal cancer with respect to onco-logic efficacy and short-term risks and benefits. Based onthe best available evidence, recommendations have beenmade for surgical practice in Quebec.

METHODS

Published clinical trials comparing open and laparoscopicsurgery in colon and rectal cancer treatment were retrievedusing the medical subject headings “colorectal neoplasms,”“rectal neoplasms,” “general surgery,” “colorectal surgery,”“laparoscopy” and “colectomy” as well as the keywords“colorectal cancer,” “colon cancer,” “rectal neoplasms,”“rectal cancer,” “open surgery,” “resection,” “laparoscopy,”“colectomy,” “rectal surgery” and “total mesorectal exci-sion.” Only English- and French-language phase III ran-domized trials and meta-analyses were selected. Specificallyfor colon cancer, in light of the large number of studiesretrieved and the variations in study quality, only trialsinvolving more than 200 patients were retained. Theperiod covered was from January 1995 to April 2012, inclu-sively. Studies reporting data on colorectal cancer wereconsidered only if specific data on colon and rectal cancerwere presented separately. Economic studies, trials pertain-ing to metastatic disease of the colon or rectum and trialsaddressing chemotherapy- or radiotherapy-based treat-ments were excluded. Abstracts presented at the AmericanSociety of Clinical Oncology (ASCO) and European Soci-ety for Medical Oncology (ESMO) meetings from 2008 to2012 were also reviewed, and only those reporting phaseIII trial results were included.

The level of evidence of selected studies and thestrength of recommendation were evaluated using theASCO and ESMO gradation system (Table 1).13 The ori -ginal guideline was developed by a CEPO subcommittee,reviewed by independent experts and, finally, adopted bythe CEPO by consensus.

RESULTS

Eight phase III randomized clinical trials10,11,14–27 and5 meta-analyses28–32 comparing laparoscopy and open


Can J Surg, Vol. 56, No. 5, October 2013 299

RESEARCH

Table 1. Levels of evidence and grades of recommendations

Level or grade Description

Level of evidence I Evidence demonstrated by means of meta-analyses of well-designed controlled trials or large randomized trials with clear-cut

results (low rate of false-positive and false-negative errors, high power) II Evidence demonstrated by means of small randomized trials with uncertain results (high rate of false-positive and false-negative

errors, low power) III Evidence demonstrated by means of nonrandomized concurrent cohort comparisons with contemporaneous controls IV Evidence demonstrated by means of nonrandomized historical cohort comparisons V Evidence demonstrated by means of case series without controls Grade of recommendation A Supported by Level I evidence or multiple Level II, III or IV trials presenting concordant observations B Supported by Level II, III or IV trials presenting generally concordant observations C Supported by Level II, III or IV trials presenting nonconcordant observations D Supported by few trials or no empiric evidence

Adapted from Cook and colleagues13 with permission from the American College of Chest Physicians.

Table 2. Main characteristics of randomized clinical trials on colon cancer

Variable COLOR10,14 COST16,17 CLASICC11,20,21* LAPKON II26 ALCCaS27 Barcelona22,23 Liang24 LAFA-study25†

Study design, Phase III multi multi multi multi multi single single multi

Concealment Computer-generated random numbers

Centralized Centralized by telephone

Centralized by telephone Revealed

during operation

Centralized Sealed envelopes

Random-sized blocks 2–10

2 × 2 Internet

randomization module

Median follow-up, mo. 53 60 37 § § 43 40 —

Primary outcomes 3-year DFS Time to recurrence

3-year DFS 3-year OS

LR

§ § Cancer-related survival

Time to recurrence

Total postop hospital stay

Level of evidence‡ I I I I I II II I

Population, no

OP 542 428 140 222 298 102 134 108

LAP 534 435 273 250 294 106 135 110

Tumour stage

I 24% 26% (OP) 35% (LAP)

NA 28% (OP) 35% (LAP)

23% 22% — NA

II 43% 34% (OP) 31% (LAP)

NA 38% (OP) 32% (LAP)

40% 43% 49% NA

III 33% 28% (OP) 26% (LAP)

NA 33% (OP) 33% (LAP)

30% 35% 51% NA

IV — 4% (OP) 2% (LAP)

NA — 2% — — NA

Postoperative chemotherapy

According to surgeon

(p = 0.99)

According to surgeon

29% (OP) 28% (LAP)

NA NA 55% (OP) 61% (LAP)

For stage III patients

NA

Surgery

Surgical procedure

Right 47% 54% 45% 29% 58% 45% — 48%

Left 11% 7% 13% ¶ 4% 2% 70% 49%

Sigmoid 38% 38% 21% ¶ — 45% 30% —

Anterior — — 11% — 38% 5% — —

Conversion rate 19% 21% 25% 11% 15% 11% 3% 11%

Surgeon experience ≥ 20 LAP colectomies

≥ 20 LAP colectomies

≥ 20 LAP resections

≥ 20 LAP colectomies

52% surgeons treated

> 10 patients

Experienced team

Experienced surgeon

≥ 20 LAP for benign disease

CLASICC = Conventional Versus Laparoscopic-Assisted Surgery in Patients with Colorectal Cancer; COLOR = Colon Cancer Laparoscopic or Open Resection; COST = Clinical Outcomes of Surgical Therapy; DFS = disease-free survival; LAP = laparoscopy; LR = local recurrence; multi = multicentred; NA = not available; OP = open surgery; OS = overall survival; single = single-centred. *This trial included patients with colon and rectum cancers. When available, only data speci!c to colon cancer are presented. †This trial evaluated fast-track versus standard care and LAP versus OP (4 arms). Only the 2 arms with standard care (LAP v. OP) are presented. ‡As evaluated according to the American Society of Clinical Oncology and European Society for Medical Oncology gradation system (see Table 1). §Only short-term outcomes are published. ¶71% for left plus rectosigmoid.



RECHERCHE

surgery for the treatment of colon cancer were identified.For rectal cancer, 9 phase III randomized clinical trials11,20,21,33–42 and 7 meta-analyses were selected.28,29,43–47 Nomeeting abstracts satisfied the inclusion criteria. The maindesign characteristics of each randomized trial are sum -mar ized in Table 2 for colon cancer and in Table 3 for rec-tal cancer.

Oncologic outcomes: colon cancer

Phase III randomized trialsIn 2009, Buunen and colleagues10 presented the long-termresults of the Colon cancer Laparoscopic or Open Resec-tion (COLOR) noninferiority trial. The primary outcomewas 3-year disease-free survival, which was 74.2% with thelaparoscopic procedure and 76.2% with open surgery.Noninferiority thresholds were set at a 7% differencebetween the 2 procedures at a level of significance of p =0.025. These 2 criteria were not met, since the superior

limit of the 95% confidence interval (CI) of the observeddifference reached 7.2% (p = 0.03). Three-year overall sur-vival was 81.8% after laparoscopy and 84.2% after opensurgery (p = 0.45).

In 2004, the long-term results of the Clinical Outcomesof Surgical Therapy (COST) noninferiority trial16 werepresented. The cut-offs to declare the laparoscopic pro -cedure noninferior to open surgery regarding time torecurrence at 3 years were set at a hazard ratio (HR) of lessthan 1.23 and p ≥ 0.41. According to these criteria, thelaparoscopic procedure was not inferior to open surgery(p = 0.83). The cumulative incidence of recurrence did notsignificantly differ between the 2 procedures (HR 0.86,95% CI 0.63–1.17, p = 0.32). No differences in overall sur-vival (HR 0.91, 95% CI 0.68–1.21, p = 0.51) or disease-freesurvival (HR 0.95, 95% CI 0.74–1.23, p = 0.70) were ob -served. In 2007, Fleshman and colleagues17 publishedupdated results after 5 years of follow-up and confirmed thenoninferiority of laparoscopy in terms of time to recurrence

Table 3. Main characteristics of randomized clinical trials on rectal cancer

Variable Liang33 COREAN34 CLASICC11,20,21* Lujan36 Ng37,38 Ng39 Pechlivanides40 Braga41* Zhou42

Study design,Phase III

single multi multi single single single multi single single

Concealment Opaqueenvelopes

Computer-generated

Centralized, by phone

Computer-generated

Computer-generated

Computer-generated

Computer-generated

Computer-generated

NA

Median follow-up, mo.

44 NA 37 34 (OP)33 (LAP)

113 (OP)109 (LAP)

91 (OP)87 (LAP)

NA 54 NA

Primary outcomes 3-year OS DFS 5-year DFS,5-year OS,

LR

No. lymph nodes

retrieved,integrity ofmesorectalresection margin

Long-termmorbidity

Postoperative recovery

No. lymph nodes

retrieved

Short-termpostopmor

bidity

Short-termresults

Level of evidence† I I I I I I I I I

Population, no

OP 174 170 128 103 77 48 39 89 89

LAP 179 170 253 101 76 51 34 82 82

Tumour stage

I NA NA NA 15% (OP)11% (LAP)

15% (OP)17% (LAP)

NA NA NA NA

II NA NA NA 38% (OP)35% (LAP)

38% (OP)38% (LAP)

NA NA NA NA

III NA NA NA 43% (OP)45% (LAP)

26% (OP)36% (LAP)

NA NA NA NA

IV NA NA NA 5% (OP)10% (LAP)

21% (OP)9% (LAP)

NA NA NA NA

Postoperative chemotherapy

NA Recom-mended for 4 mo

29% (OP)28% (LAP)

Stage III orIV disease

33% (OP)14% (LAP)

NA NA NA NA

Surgery

Tumour distancefrom AV, cm

NA 5.3 (OP)5.6 (LAP)

NA 6.2 (OP)5.5 (LAP)

12–15 cm ≥ 5 cm 8 (OP)6 (LAP)

8.6 (OP)9.1 (LAP)

NA

Conversion rate < 1% 1% 34% 8% 30% 10% 3% 7% NA

Surgeonexperience

Experiencedsurgeon

28–150 LAP

≥ 20 LAP Experiencedteam

Experienced surgeon

Experienced surgeon

Experiencedsurgeon

Experiencedteam

Experienced surgeon

AV = anal verge; CLASICC = Conventional Versus Laparoscopic-Assisted Surgery in Patients with Colorectal Cancer; COREAN = Comparison of Open versus laparoscopic surgery for mid and low Rectal cancer After Neoadjuvant chemoradiotherapy; DFS = disease-free survival; LAP = laparoscopy; LR = local recurrence; NA = not available; OP = open surgery;OS = overall survival. *This trial included patients with colon and rectum cancers. When available, data speci!c to rectal cancer are presented. †As evaluated according to the American Society of Clinical Oncology and European Society for Medical Oncology gradation system (Table 1).



RESEARCH

(p = 0.75) and cumulative incidence of recurrence (HR0.84, 95% CI 0.62–1.13, p = 0.25).

In 2007, Jayne and colleagues11 presented the long-termresults of the Conventional versus Laparoscopic-AssistedSurgery in Colorectal Cancer (CLASICC) trial. The objec-tive was to assess overall survival, disease-free survival andlocal recurrence at 3 years in patients with colon or rectalcancer treated with laparoscopic or open surgery. Forcolon cancer, the local recurrence rates were 7.3% withlaparoscopy and 6% with open surgery (p = 0.68). Differ-ences between the 2 approaches with respect to 3-yearoverall survival (p = 0.51) and disease-free survival (p =0.75) were not significant. Updated data showed similarresults between groups assigned to laparoscopy and opensurgery for 5-year overall survival (55.7% v. 62.7%, p =0.25) and disease-free survival (57.6% v. 64%, p = 0.40).20

In 2008, Lacy and colleagues22 presented long-termresults of a phase III randomized trial conducted inBarcelona, Spain, updating data initially published in2002.23 The primary outcome was cancer-related mortality.After 3.5 years of follow-up, cancer-related mortality was9% with laparoscopy and 21% with open surgery (p =0.03); after a median follow-up of 8 years, the rate was16% and 27%, respectively, (p = 0.07). Recurrence rates of18% with laparoscopy and 28% with open surgery werealso observed (p = 0.07).

In 2007, Liang and colleagues24 published results of arandomized trial conducted in Taiwan by a single surgeon.Time to recurrence after colon cancer resection was notsignificantly different between the laparoscopic and openprocedure (p = 0.36). The cumulative incidence of recur-rence was 17% with laparoscopy and 21.6% with opensurgery.

In 6 trials, the extent of resection, as measured by resec-tion margins and the number of lymph nodes harvested,did not significantly differ between laparoscopic and opensurgery. The number of lymph nodes harvested variedbetween 10 and 17.14,16,21,23,24,26 In the studies reporting recur-rence rates at wound or port sites, the rates were not statis-tically different between the groups (1.3% v. 0.4%, p =0.09;10 0.9% v. 0.5%, p = 0.43;17 0.9% v. 0%, p value notavailable;23 and 0.7% v. 0.7%, p value not available,24 for thelaparoscopic and open procedures, respectively).

Meta-analysesIn 2011, Ma and colleagues29 conducted a meta-analysiscomparing laparoscopy with open resection for colorectalcancer. Data from 6 studies (n = 1800) specific to coloncancer showed that cancer-related mortality was 17.7%with laparoscopy and 19.7% with open surgery (odds ratio[OR] 0.85, 95% CI 0.66–1.09, p = 0.20).

In 2010, Bai and colleagues30 conducted a meta-analysisincluding 3 trials (n = 2147)10,17,22 that reported long-termoutcome data following laparoscopic and open colonresection. Overall mortality was similar for laparoscopy

and open surgery (24.9% v. 26.4%, OR 0.92, 95% CI0.76–1.12, p = 0.41). Overall recurrence rates of 19.3%and 20% (OR 0.96, 95% CI 0.78–1.19, p = 0.71), localrecurrence rates of 4% and 4.4% (OR 0.91, 95% CI 0.59–1.39, p = 0.66) and distal recurrence rates of 12.8% and14% (OR 0.90, 95% CI 0.70–1.16, p = 0.41) were alsoobserved with laparoscopic and open surgery, respectively.

In 2008, Kuhry and colleagues31 conducted a Cochranecollaboration meta-analysis comparing survival and recur-rence rates in patients with colorectal cancer treated withlaparoscopic or open surgery. This meta-analysis included12 trials. Four trials (n = 938) presented results of recur-rence in patients with colon cancer. The local recurrencerate was 5.2% with laparoscopic surgery and 5.6% withopen resection (OR 0.84, p = 0.57), whereas distant recur-rence was 11.3% and 13.6%, respectively (OR 0.82, p =0.32). In 5 trials (n = 1575), cancer-related mortality was14.6% with laparoscopy and 16.4% with open surgery (OR0.80, p = 0.12). The combined results of 4 trials (n = 1162)showed overall mortality of 20.4% with laparoscopy and23.6% with open surgery (OR 0.82, p = 0.17).

In 2007, Bonjer and colleagues32 presented a meta-analysis combining individual data on patients recruitedbefore March 2000 in the COLOR, COST, CLASICCand Barcelona trials (n = 1536). The 3-year overall survival(82.2% v. 83.5%, p = 0.56) and disease-free survival (75.8%v. 75.3%, p = 0.83) were similar between laparoscopic andopen colectomy, respectively.

A recent meta-analysis28 showed that the number oflymph nodes harvested was similar with laparoscopic andopen colon resection (weighted mean difference [WMD] –0.18, p = 0.82).

Oncologic outcomes: rectal cancer

Phase III randomized trialsIn 2011, Liang and colleagues33 presented results of a ran-domized trial conducted in a single centre in China toevaluate 3-year overall survival following laparoscopic oropen surgery for rectal cancer. After a median follow-up ofabout 44 months, overall survival was similar for laparo-scopic and open surgery (76% v. 82.8%, p = 0.46). Therewas no difference between these 2 procedures in themedian number of lymph nodes harvested (7.1 v. 7.4, p =0.47) or the distance between the inferior border of thetumour and the incised margin in the lower anteriorresection operation (3.2 cm v. 3.1 cm, p = 0.15).

In 2010, Kang and colleagues34 presented the short-termoncologic results of the Comparison of Open versuslaparoscopic surgery for mid and low REctal cancer AfterNeoadjuvant chemoradiotherapy (COREAN) trial. Thetumour had to be localized at no more than 9 cm from theanal verge. No difference was shown between laparoscopicand open surgery with respect to the macroscopic qualityof the resected mesorectum (complete 72.4% v. 74.7%;



RECHERCHE

almost complete 19.4% v. 13.5%; incomplete 4.7% v.6.5%; p = 0.41), the median number of lymph nodes har-vested (17 v. 18, p = 0.09), the negative circumferentialmargins (97.1% v. 95.5%, p = 0.77) and the rate of negativeproximal, distal and radial margins (p = 0.44, p = 0.54 andp = 0.31, respectively).

In 2010, Jayne and colleagues20 presented updatedresults of the CLASICC trial after a median follow-up of56.3 months. For rectal cancer, overall survival (60.3% v.52.9%, p = 0.13) and disease-free survival (53.2% v. 52.1%,p = 0.95) were similar for laparoscopic and open surgery,respectively. No difference was observed between laparo-scopic and open procedures in local and distal recurrencerates. Distal recurrence rates were 21.9% with both laparo-scopic and open anterior resection, whereas they were35.7% with laparoscopic abdominoperineal resection and40.8% with open abdominoperineal resection. Local recur-rence rates were 9.4% with laparoscopic and 7.6% withopen anterior resection. Positive resection margins weremore frequent with laparoscopic than open anterior resec-tion (12.4% v. 6.3%, p = 0.01), but not with abdominoperi -neal resection (20% v. 26%, p value not available).

In 2009, Lujan and colleagues36 presented results of anoninferiority randomized trial that evaluated efficacy oflaparoscopy compared with open resection of the low ormid rectum. Five-year disease-free survival (84.8% v. 81%,p = 0.90), overall survival (72.1% v. 75.3%, p = 0.98) andlocal recurrence (4.8% v. 5.3%, p = 0.78) were similar withlaparoscopic and open surgery, respectively. The meannumber of lymph nodes harvested was greater withlaparoscopy than with open surgery (13.6 v. 11.6, p = 0.03).Integrity of the resection margins was similar with laparo-scopic and open procedures (2.8% v. 4%, p = 0.42).

In 2009, Ng and colleagues37 presented results of a ran-domized trial conducted in a single centre in Hong Kongto evaluate long-term oncologic efficacy of laparoscopicsurgery for proximal rectal cancer (12–15 cm from the analverge). After a median follow-up of about 110 months, nodifference was found between laparoscopic and opensurgery in terms of overall survival (p = 0.30), cancer-related survival (p = 0.60) and disease-free survival (p =0.70) in patients with stage I–III rectal cancer. Mean sur-vival was not different for stage IV cancer (p = 0.16). Dur-ing the 10-year follow-up period, 37.3% of patientsassigned to laparoscopy and 38.8% of patients assigned toopen surgery died; 15.3% and 16.4% were rectal cancer-related deaths, and 18.6% and 19.4% were other cancer-related deaths, respectively. Recurrence rates (local 7.1% v.4.9%, p = 0.68; distal 12.3% v. 18.1%, p = 0.37), meannumber of lymph nodes harvested (11.5 v. 12, p = 0.70) andpositive resection margins (2.6% v. 1.3%, p = 0.62) weresimilar with laparoscopic and open surgeries, respectively.

In 2008, Ng and colleagues39 presented results of a ran-domized trial conducted in a single centre in Hong Kongthat evaluated laparoscopic surgery in distal rectal cancer

treatment (up to 5 cm from the anal verge). After a medianfollow-up of about 90 months, no difference was foundbetween laparoscopic and open surgery for 5-year overallsurvival (75.2% v. 76.5%, p = 0.20) and disease-free survival(78.1% v. 73.6%, p = 0.55) in patients with stage I–III rectalcancer. In patients with stage IV disease, mean survival was32.6 months after laparoscopy and 13.9 months after opensurgery (p = 0.05). During the follow-up period, 30% ofpatients assigned to laparoscopy and 47.2% of patientsassigned to open surgery died; 15% and 22.2%, respectively,were cancer-related deaths. Recurrence rates were similarwith laparoscopic and open surgery (p = 0.60), as were themean number of lymph nodes harvested (12.4 v. 13, p = 0.72)and the positive resection margins rate (6.3% v. 3.9%).

In 2007, Pechlivanides and colleagues40 presented theresults of a randomized trial conducted by a single surgeonto compare laparoscopic and open surgeries in the treat-ment of rectal cancer localized at less than 12 cm from theanal verge. The mean number of lymph nodes harvestedwas 19.2 in both groups (p = 0.2). No data on survival orrecurrence were presented.

In 2007, Braga and colleagues41 presented results of arandomized trial, conducted in Italy by a single team, aftera median follow-up of 53.6 months. No difference wasobserved between groups assigned to laparoscopy or opensurgery for 5-year overall survival (stage I: p = 0.93, II: p =0.37, III: p = 0.98, and IV: p = 0.95), 3-year local recurrence(4% v. 5.2%, p = 0.97) and mean number of lymph nodesharvested (12.7 v. 13.6, p value not available). The distalmargins were negative in all patients whereas positive cir-cumferential margins were observed in 1 patient treatedwith laparoscopy (1.3%) and 2 patients treated with opensurgery (2.4%).

Meta-analysesIn 2011, Huang and colleagues43 conducted a meta-analysisincluding 6 clinical trials (n = 1033) that evaluated efficacy oflaparoscopic surgery in rectal cancer treatment. Three-yearoverall survival (p = 0.11, 4 trials) and disease-free survival(p = 0.11, 3 trials) were not significantly different after

Table 4. Comparison of oncologic results in rectal cancer44

Result OR (95% CI) p value

Progression-free survival

3-year 0.90 (0.66–1.24) 0.53

5-year 1.17 (0.85–1.61) 0.35

Recurrence

Total 0.93 (0.68–1.25) 0.61

Local 0.83 (0.52–1.31) 0.41

Wound sites 1.34 (0.07–24.10) 0.84

Distal metastasis 0.89 (0.63–1.27) 0.52

Mortality

Overall 0.80 (0.60–1.07) 0.13

Cancer-related 0.71 (0.45–1.12) 0.14

CI = con!dence interval; OR = odds ratio.



RESEARCH

laparoscopy or open surgery. After a follow-up ranging from32.8 to 112.5 months, local recurrence rates after laparo-scopic and open surgery were not statistically different (p =0.21, 4 trials). No difference was observed betweenlaparoscopy and open surgery for the mean number oflymph nodes harvested (p = 0.43, 5 trials); positive circumfer-ential resection margins were also similar (7.9% v. 5.4%, p =0.63, 5 trials).

In 2011, Ohtani and colleagues44 conduted a meta-analysis comparing the oncologic efficacy of laparoscopicand open surgery for rectal cancer. Twelve trials wereincluded (n = 2095), and results showed no differencebetween the 2 procedures for the oncologic outcomesmeasured (Table 4).

In the meta-analysis comparing laparoscopy with openresection for colorectal cancer by Ma and colleagues29 in2011, data specific to rectal cancer from 5 studies were pre-sented (n = 991). Cancer-related mortality was 13.1% withlaparoscopy and 15.3% with open surgery (OR 0.76, 95%CI 0.53–1.11, p = 0.16).

In 2008, Anderson and colleagues45 conducted a meta-analysis including 22 clinical trials that evaluated the efficacyof laparoscopic surgery in rectal cancer treatment. After4.4 years of follow-up (13 trials), overall survival was 72%with laparoscopy and 65% with open surgery (p = 0.5). At amedian follow-up of 35 months, local recurrence rates afterlaparoscopic and open surgery were 7% and 8%, respect -ively (16 trials), whereas distal recurrence rates were 12%and 14%, respectively (p = 0.54, 9 trials). The mean numberof lymph nodes harvested was lower with laparoscopy thanwith open surgery (10 v. 12, p = 0.001, 17 trials); however,3 trials showed that more lymph nodes were harvested with

laparoscopy. The rate of positive resection margins was simi-lar with laparoscopic and open surgery (5% v. 8%, 10 trials).

Three additional meta-analyses compared short-termoncologic outcomes after laparoscopic and open surgeryfor rectal cancer and showed no difference between theprocedures in terms of the mean number of lymph nodesharvested and the rate of positive resection margins.28,46,47

Short-term outcomes

Duration of operationThirteen trials presented data on the duration of surgery forcolon and rectal cancer (Table 5). In all but 1 study, operativetime was longer for laparoscopic than for open surgery. TheCOLOR trial investigators showed that differences in oper-ative time between the 2 procedures for colon cancer tendedto be smaller in centres with high volumes (p = 0.027).14

Intraoperative blood lossFive trials assessed median intraoperative blood loss duringlaparoscopic and open surgeries for colon cancer.10,23–25,27

Four trials observed less blood loss during laparoscopy:COLOR (100 mL v. 175 mL, p = 0.003),10 Barcelona(105 mL v. 193 mL, p = 0.001),23 Liang and colleagues24

(54 mL v. 240 mL, p < 0.001) and LAFA-study (100 mL v.200 mL, p < 0.001).25 Only the Australasian LaparoscopicColon Cancer Study (ALCCaS) trial showed no signifi-cant difference between the 2 procedures (median bloodloss of 100 mL in both cases, p = 0.17).27

Six trials assessed intraoperative blood loss during rectalcancer surgery.34,36,37,39,41,42 All trials showed a trend towardless blood loss with laparoscopy; this trend was significant

Table 5. Operative time of open and laparoscopic surgeries

Trial

Operative time; median (range) or mean ± SD, min

p value Open surgery Laparoscopic surgery

Colon cancer

COLOR10 115 (70–180) 145 (102–230) < 0.001

COST16 95 (27–435) 150 (35–450) < 0.001

Barcelona23 118 ± 45 142 ± 52 0.001

Liang et al.24 184 ± 31 224 ± 45 < 0.001

LAPKON II26 138 ± 45 183 ± 61 < 0.001

ALCCaS27 107 (45–250) 158 (49–365) < 0.001

Rectal cancer

Liang et al.33 119 ± 22 138 ± 24 < 0.001

COREAN34 197 ± 63 245 ± 75 < 0.001

Lujan et al.36 173 ± 59 194 ± 45 0.020

Ng et al.37 154 ± 70 213 ± 59 < 0.001

Ng et al.39 164 ± 43 214 ± 46 < 0.001

Braga et al.41 209 ± 70 262 ± 72 < 0.001

Zhou et al.42 106 (80–230) 120 (110–220) 0.05

ALCCaS = Australasian Laparoscopic Colon Cancer Study; COLOR = Colon cancer Laparoscopic or Open Resection; COREAN = Comparison of Open versus laparoscopic surgery for mid and low Rectal cancer After Neoadjuvant chemoradiotherapy; COST = Clinical Outcomes of Surgical Therapy; SD = standard deviation.



RECHERCHE

in 4 trials: COREAN (200 mL v. 217.5 mL, p = 0.006),34

Lujan and colleagues36 (127.8 mL v. 234.2 mL, p < 0.001),Braga and colleagues41 (150 mL v. 350 mL, p < 0.001) andZhou and olleagues42 (20 mL v. 92 mL, p = 0.05).

Two trials compared transfusion use during laparoscopicand open surgeries for the treatment of colon cancer.26,27

Sixteen patients undergoing laparoscopy (5.4%) versus18 pa tients undergoing open surgery (6%) received atransfusion in the ALCCaS trial,27 whereas 11.6% and17.6% of patients assigned to laparoscopic and opensurgery, respetively, received transfusion in LAPKON II.26

Three trials evaluated transfusion rates during rectal can-cer resection.33,34,41 Only 1 patient, assigned to open surgery,needed a transfusion in the COREAN trial (p > 0.99),34

whereas 4 patients (2.4%) assigned to laparoscopy and8 patients (4.6%) assigned to open surgery needed a trans-fusion in the trial by Liang and colleagues33 (p = 0.38).Braga and colleagues41 observed transfusion rates of 7.2%with laparoscopy and 26.8% with open surgery (p = 0.002).

Postoperative painLiang and colleagues24 measured postoperative pain using avisual analogue scale of 0–10. Less pain was recorded afterlaparoscopy than open surgery for colon cancer (median3.5 v. 8.6, p < 0.001). In the COREAN trial, mean postop-erative pain was less after laparoscopy than open surgeryup to 3 days after surgery for rectal cancer (p < 0.05).34 Ngand colleagues39 reported no difference in pain, accordingto the visual analogue scale, on the first day after laparo-scopic and open surgery for rectal cancer (p = 0.41).

In 2 trials, less analgesic use was reported after lapa -roscopy than open surgery for colon cancer. In theCOLOR trial, 8%–14% fewer patients needed analgesics inthe first 3 days after laparoscopy than open surgery(p < 0.001 to p = 0.008).14 In the COST trial, this differencecorresponded to a median of 1 day less needing analgesics.16

There was also less analgesic use after laparoscopy thanopen surgery for rectal cancer, according to 4 trials. Inthese trials, lower doses of morphine (median 107.2 mg v.156.9 mg, p < 0.001),34 and fewer injections of analgesics(mean 6 v. 11.4, p = 0.007 and 4.9 v. 8.3, p = 0.001)37,39 wereused by patients assigned to laparoscopy than open surgery.The median duration of analgesic treatment was alsoshorter (3.9 d v. 4.1 d, p = 0.23).42

Recovery of intestinal functionFor colon cancer, Liang and colleagues24 found that postop-erative ileus lasted for a mean of 48 hours after laparoscopycompared with 96 hours after open surgery (p < 0.001). Inthe Barcelona trial, peristalsis began at a mean of 36 hoursafter laparoscopic surgery compared with 55 hours afteropen surgery (p = 0.001).23 In the ALCCaS trial, the meantime to passing first flatus was 3.2 days after laparoscopycompared with 3.5 days after open surgery (p = 0.027).27

Finally, 3 trials measured time to first bowel movement: 3.6,

5.0 and 4.4 days, respectively, after laparoscopy and 4.6, 6.0and 4.9 days, respectively, after open surgery (p < 0.001,14

p value not available21 and p = 0.011,27 respectively).For rectal cancer, peristalsis also began sooner after

laparoscopy than open surgery, as found in 5 trials: Liang andcolleagues33 (3.9 d v. 4.2 d, p < 0.001), Ng and colleagues39

(4.3 d v. 6.3 d, p < 0.001), Ng and colleagues37 (4.1 d v. 4.7 d, p = 0.06), CLASICC (5 d v. 6 d, p value not available)21 andZhou and colleagues42 (1.5 d v. 2.7 d, p = 0.009). This conclu-sion is supported in a meta-analysis (WMD –1.52 d, 95% CI–2.20 to –1.01).46 The mean time to passing first flatus wasalso shorter after laparoscopy (38.5 hr v. 60 hr, p < 0.00134 and3.1 d v. 4.6 d, p < 0.00139), as was the mean time for first stool(3 d v. 3.3 d, p < 0.00133 and 96.5 hr v. 123 hr, p < 0.00134).Return to normal diet after laparoscopic and open surgeryfor rectal cancer was evaluated in 4 trials: COREAN (85 hr v.93 hr, p < 0.001),34 Ng and colleagues39 (4.3 d v. 6.3 d, p =0.001), Ng and colleagues37 (4.3 d v. 4.9 d, p = 0.001) andCLASICC (6 d v. 6 d; p value not available).21 Faster return tonormal diet after laparoscopy was confirmed in a meta-analysis (WMD –0.92 d, 95% CI –1.35 to –0.50).46

Length of hospital staySeven trials presented data on length of hospital stay aftercolon cancer surgery (Table 6). In all cases, patients treatedwith laparoscopy had a shorter hospital stay than patientstreated with open surgery. However, the ALCCaS trialshowed that among laparoscopy-treated patients, those whowere converted to open surgery had a significantly longerhospital stay than those who were not converted (14.6 d v.8.6 d, p < 0.001).27 As for rectal cancer, 7 trials presenteddata on length of hospital stay (Table 6). Although all trialsreported a shorter hospital stay after laparoscopy than opensurgery, the difference was significant in only 3 trials. Inaddition, a meta-analysis showed that hospital stay wasshorter after laparoscopy than open surgery for rectal can-cer (WMD –2.67 d, 95% CI –3.81 to –1.54).46

Morbidity and mortalityOverall complication rates following colon cancer resec-tion were evaluated in 4 trials. Only in the Barcelona trialwas the complication rate lower for laparoscopy than opensurgery (11% v. 29%, p = 0.001),23 whereas COLOR,14

LAFA-study25 and the study by Liang and colleagues24

reported no significant difference after laparoscopic oropen procedures (p = 0.88, p = 0.20 and p = 0.15, respect -ively). Reported intraoperative complications includedhemorrhage, cardiac or pulmonary insufficiency, adverseanesthetic events and injury of bowel or adjacent organs.The intraoperative complication rate was statisticallyhigher following laparoscopy in 1 trial (10.5% v. 3.7%, p =0.001),27 whereas 3 trials showed no difference betweenlaparoscopic and open procedures (4% v. 2%, p = 0.10,16

7% v. 8%, p value not available,21 and 5.6% v. 2.3%, p =0.1026). No difference in postoperative complication rates



RESEARCH

was found in these 4 trials between laparoscopic and opensurgery (37.8% v. 45.3%, p = 0.06,27 19% v. 19%, p =0.98,16 26% v. 27%, p value not available,21 and 25.2% v.23.9%, p = 0.7526). Wound or urinary tract infections,anastomotic leakage, prolonged ileus, hemorrhage, andvarious cardiac, pulmonary or vascular complications werethe most frequent postoperative complications reported.

Complication rates following rectal cancer resectionwere presented in 7 trials.21,34,36,37,39,41,42 The intraoperativecomplication rates ranged from 6.1% to 21.2% for lapa -roscopy and from 12.4% to 23.5% for open surgery (p =0.016 to p = 0.60),21,34,42 and the postoperative complicationrates ranged from 2.4% to 45.1% and from 10.6% to52.1%, respectively (p = 0.012 to p = 0.96).36,37,39,41 Results ofa meta-analysis showed that patients assigned to lapa -roscopy presented less morbidity than those assigned toopen surgery (OR 0.63, 95% CI 0.41–0.96, p = 0.030).47

Reported intraoperative complications included hemor-rhage, cardiac or pulmonary insufficiency, anesthesia-related complications and injury of serosa or adjacentorgans, whereas postoperative complications includedanastomotic leakage, wound and urinary tract infectionsand various cardiac, pulmonary or vascular complications.Four trials reported the need for reoperation, and noneobserved any difference between patients undergoinglaparoscopic and open surgery for the frequency of secondoperation.34,37,39,41

Six trials reported short-term mortality following laparo-scopic or open resection for colon cancer.10,16,23,25–27 In all of

these studies, mortality was 2% or less and was not signifi-cantly different between the 2 groups (p = 0.40 to p = 0.99).Six trials reported the short-term mortality for rectal cancerand showed no difference between laparoscopic and openprocedures.21,33,36,37,39,41 The highest mortality was observed inthe CLASICC trial (4% following laparoscopy and 5% fol-lowing open surgery, p = 0.57).

Quality of lifeThree trials evaluated quality of life following laparo-scopic and open surgery in patients with colon cancer. In asubgoup of the COLOR trial (n = 285), social function at2 (p = 0.046), 4 (p = 0.031) and 12 (p = 0.05) weeks and rolefunction at 2 weeks (p = 0.006) were better followinglaparoscopic than open surgery.15 In the COST trial (n =428), only the global quality of life score 2 weeks aftersurgery was higher following laparoscopy (80 v. 75, p valuenot available),18 whereas only the global quality of lifescore (p = 0.05) and the subdomain of outlook (p = 0.02)were improved 18 months after surgery.19 The LAFA-study showed no statistically significant differences on anyof the scales evaluated between laparoscopic and opensurgery at any time point.25

For rectal cancer, 3 trials evaluated quality of life fol-lowing laparoscopic and open surgery. The COREANtrial showed that sleep (p = 0.004), physical condition (p =0.007) and fatigue (p = 0.021) were better 3 months afterlapa roscopy than open surgery. A greater frequency of sexualproblems after surgery than before the intervention was

Table 6. Length of hospital stay following open and laparoscopic surgeries

Trial

Length of hospital stay, d

Measure Open surgery Laparoscopic

surgery p value

Colon cancer

COLOR14 mean ± SD 9.3 ± 7.3 8.2 ± 6.6 < 0.001

COST16 median (IQR) 6 (5–7) 5 (4–6) < 0.001

Barcelona22 mean ± SD 7.9 ± 9.3 5.2 ± 2.1 0.005

Liang et al.23 mean ± SD 14.0 ± 2.0 9.0 ± 1.0 < 0.001

LAPKON II26 mean ± SD median (range)

13.0 ± 8.6 12 (4–109)

13.4 ± 12.0 10 (1–123)

0.032

ALCCaS27 median (range) mean ± SD

8 (4–59) 10.6 ± 7.2

7 (1–55) 9.5 ± 7.4

< 0.001 0.07

CLASICC21 median (IQR) 9 (8–13) 9 (7–12) NA

Rectal cancer

CLASICC21 median (IQR) 11 (9–15) 13 (9–18) NA

COREAN34 median (IQR) 9 (8–12) 8 (7–12) 0.06

Lujan et al.36 mean ± SD 9.9 ± 6.8 8.2 ± 7.3 0.11

Ng et al.37 median (range) 11.5 (3–38) 10.8 (5–27) 0.55

Ng et al. 39 median (range) 10 (3–39) 8.4 (2–32) 0.013

Braga et al.41 mean ± SD 13.0 ± 10.0 10.0 ± 4.9 0.004

Zhou et al.42 mean ± SD 13.3 ± 3.4 8.1 ± 3.1 0.001

ALCCaS = Australasian Laparoscopic Colon Cancer Study; CLASICC = Conventional Versus Laparoscopic-Assisted Surgery in Patients with Colorectal Cancer; COLOR = Colon cancer Laparoscopic or Open Resection; COREAN = Comparison of Open versus laparoscopic surgery for mid and low Rectal cancer After Neoadjuvant chemoradiotherapy; COST = Clinical Outcomes of Surgical Therapy; IQR = interquartile range; NA = not available; SD = standard deviation.



RECHERCHE

observed for both groups (p < 0.001), with no differencebetween groups (p = 0.29). More miction problems wereobserved after laparoscopy than open surgery (p < 0.001).34

Braga and colleagues41 showed better general health status(p < 0.001), physical condition (p < 0.001) and social function(p = 0.003) with laparoscopy than open surgery, but only inthe first year after the operation.41 In the CLASICC trial, itwas shown that men treated with laparoscopy tended to haveworse global sexual (p = 0.06) and erectile function (p = 0.07);however, this difference was not significant. Conversion andtotal mesorectal excision were identified as prognostic fac-tors negatively affecting sexual function. Although a decreasein global sexual function was generally observed in womenfollowing surgery, no difference was noted between groups.Urinary function was also similar following laparoscopic andopen surgery in both men and women.35

DISCUSSION

Oncologic outcomes

The first case series comparing laparoscopic and opensurgeries for colon cancer treatment reported high recur-rence rates at wound and port sites, raising doubts as tothe oncologic efficacy of laparoscopy. Large randomizedtrials that followed were therefore attentive to oncologicoutcomes, such as positive resection margins, number oflymph nodes harvested and recurrence rates at port sites.None of the trials included in this review showed any dif-ference between open and laparoscopic proceduresregarding these outcomes. Only in the COLOR trial werethere more recurrences in the abdominal wall observedfollowing laparoscopy than open surgery for colon cancer(5 at the port site and 2 at the tumour extraction site), butthe difference was not significant (p = 0.09).10 Three trialspresented data on recurrence at trocar and scar sites fol-lowing surgery for rectal cancer and reported only 1 re -currence (in the open surgery group).33,37,39 Thus, it appearsthat initial concerns were not justified.

All but 1 trial studying colon cancer concluded thatlaparoscopy is noninferior to open surgery in terms of over-all survival, disease-free survival and recurrence rate. TheCOLOR study group had set the noninferiority thresholdat 7%, and the upper limit of the 95% CI for the 3-year disease-free survival difference just exceeded this, at 7.2%.10

Though they could not totally exclude the possibility ofinferiority, the authors still concluded that lapa roscopycould be performed safely for the treatment of colon cancer.Moreover, when data are analyzed according to treatmentreceived, as recommended in the CONSORT statement,48

the noninferiority of laparoscopy is statistically confirmed.10

On the other hand, only 1 trial concluded that laparoscopywas superior to open surgery for colon cancer treatment.22

However, a number of biases were identified in that trial.First, trial design was based on the hypothesis that nonin -

feriority would be declared if cancer-related survival afterlaparoscopy was less than 15% inferior to that after opensurgery, which is a clinically unacceptable threshold accord-ing to the oncologic experts consulted. Second, equivalenceof the 2 groups must be questioned, since the groupassigned to laparoscopy received more postoperativechemotherapy than the group assigned to open surgery(61% v. 55%). It therefore seems inappropriate to conclude,based on the results of this single trial, that long-term out-comes of patients with colon cancer after laparoscopicresection are superior to those after open surgery.

The combined results of the 5 randomized trials in -cluded in this review confirm that laparoscopy is nonin -ferior to open surgery for colon cancer treatment withrespect to overall survival, disease-free survival and diseaserecurrence.10,11,16,22,24 Results presented in 3 meta-analysesalso strongly support these conclusions.30–32

Trials evaluating the laparoscopic procedure for rectalcancer had generally lower power than those evaluatinglaparoscopic colon cancer resection, thus explaining whytrials accruing fewer than 200 patients were included inthis review. Consequently, 6 randomized trials presentingdata on survival and recurrence were retrieved and showednoninferiority of laparoscopic compared with open resec-tion for rectal cancer in terms of overall and disease-freesurvival.20,33,36,37,39,41 The equivalence of these 2 procedures issupported by 3 meta-analyses,43–45 and was confirmed forboth anterior and abdominoperineal resections.20

All trials retrieved included resections of the left, right orsigmoid colon for adult patients with stage I, II or III coloncancer or anterior or abdominoperineal resection of rectalcancer of any stage. A laparoscopic procedure is thus appro-priate for these populations. In contrast, patients presentingwith 1 or more of the following conditions were generallyexcluded: transverse colon cancer; morbid obesity; adjacentorgan invasion; metastatic disease; cardiovascular, pulmonaryor hepatic disease; inflammatory bowel disease; or need foremergency surgery. These conditions can be consideredpotential contraindications for laparoscopy, depending onthe surgeon’s experience. Notably, transverse colon resectionposes many challenges, such as resection of the colic vesselsand mobilization of the 2 colic segments before anastomosis.For these reasons, laparoscopic transverse colon resectionshould only be performed by surgical teams with extensiveexperience in laparoscopic colon resections. On the otherhand, while patients with metastatic disease have generallybeen excluded from trials to avoid survival bias, clinical ex -perience shows that laparoscopy can successfully be per-formed in some symptomatic patients presenting withobstruction or bleeding.

Short-term outcomes

Selected trials showed some short-term benefits oflaparoscopy compared with open colorectal cancer



RESEARCH

resection. These benefits include reduced need for anal-gesics, less postoperative pain, faster recovery of intest -in al function and shorter hospital stay. While these dif-ferences were reported to be statistically significant,they represent only modest clinical benefits. However,1 trial showed that partial and complete return to regu-lar activities of daily life and return to work occurredtwice as quickly after laparoscopy than open surgery.24

This aspect could be a decisive factor in favour of lapa -roscopy for some patients.

Minor benefits regarding quality of life were alsoreported for patients who underwent laparoscopicsurgery.15,18,19,34,41 However, the validity of these results isopen to question, since bias could not be excluded becausepatients were not blinded to treatment allocation andbecause no validated instrument specific to laparoscopicsurgery is available. The questionnaires used in these trialsaddressed general health or cancer-related outcomes,which may have been less appropriate for quality of lifeassessment for patients undergoing surgery. Specifically forrectal cancer, it was shown that both urinary and sexualfunctions were not worse 3 months after laparoscopy com-pared with open surgery. Only in the COREAN trial weremore problems with micturition after laparoscopyobserved (p = 0.002);34 however, this result was not repli-cated in a meta-analysis.44

Laparoscopy requires longer operative time than opensurgery. Mean operative times for colon cancer resectionvaried between 95 and 184 minutes for open surgery andbetween 142 and 224 minutes for laparoscopy. The differ-ence in duration for the 2 procedures thus ranges between24 and 55 minutes. For rectal cancer, mean or median oper-ative times ranged between 106 and 209 minutes with opensurgery and between 120 and 262 minutes with laparoscopy,corresponding to a difference of 14–59 minutes betweenthe 2 procedures. In the COLOR trial, differences betweenoperative times tended to be smaller in centres with highvolumes, suggesting that this disadvantage of laparoscopy isattenuated as the surgeon acquires more experience.14 Thistrend was also observed for rectal cancer surgery.49–51

No clear conclusion can be drawn regarding complica-tion rates following open or laparoscopic procedures forcolon cancer, since 1 trial observed more complications fol-lowing open surgery,23 another showed more complicationsfollowing laparoscopy,27 and other trials observed similarcomplication rates for the 2 procedures.14,16,21,24–26 In mosttrials, however, complication rates remained at an accept-able level following both procedures and were generallylow grade (I and II).16 In rectal cancer trials, intraoperativeand postoperative complication rates ranged from 14% to24% and from 6% to 52%, respectively, and were not sta-tistically different between laparoscopic and open surgery.It is also presumed that the complication rate would beaffected by the surgeon’s experience, the complexity of thesurgery performed and the patient selection criteria.

Few patients died as a result of colorectal cancer resec-tion, regardless of the operative technique.

Conversion rate

In the trials included in this review, conversion ratesranged from 3% to 25% for colon cancer and from 0% to33% for rectal cancer. Conversion was not uniformlydefined across trials, making comparison difficult. Reasonsfor colon cancer conversion included advanced disease,large tumour size, inability to localize the tumour, inabilityto mobilize the colon, extensive adhesions, obesity andintraoperative complications, whereas reasons for rectalcancer conversion included substantial bleeding, difficultyin performing a safe resection, anastomotic leakage, rectalperforation, damaged ureter, obesity, advanced disease andnarrowness of the pelvic cavity. Moreover, body massindex was identified as a significant factor for conversionin the context of rectal cancer resection (p = 0.026).52 Thelowest conversion rates were generally observed in trialswhere a single surgeon or team performed all laparoscopicsur ger ies,23,24,33,34,36,40,41 suggesting that conversion rate isinversely correlated with surgeon experience. Further-more, in the CLASICC trial, the conversion rate to opensurgery for colorectal cancer decreased as more experiencewas gained: 38% in the first year of accrual to 16% in thesixth year.21

Conversion to open surgery is believed to have a nega-tive impact on survival and morbidity outcomes. A sub-group analysis performed by the authors of the COSTstudy showed that patients who underwent conversionexperienced more complications than those whose colonsurgery had been completed laparoscopically (7.8% v.2.9%). Five-year overall survival was also lower in thesepatients (69% v. 80%).17 In the CLASICC trial, convertedsurgery was associated with a hospital stay of up to 2 weekslonger than laparoscopy and with a higher complicationrate (93% v. 50% following open surgery and 59% follow-ing laparoscopy).21 On intention-to-treat analysis, patientswho were converted intraoperatively to open surgery wereappropriately included in the laparoscopic group. Conver-sion is a reality of normal practice and thus has to be considered in the evaluation of the safety of laparoscopy. Nevertheless, surgeries most suscept ible to being convertedshould be identified preoperatively whenever possible sothat these patients can be treated with open surgery toreduce the likelihood of complications.

Minimum training requirements

No randomized clinical trial has evaluated the minimumtraining requirements for safely performing laparoscopyfor colon cancer treatment, despite the need to betterinform surgeons in this regard. Most practice guidelinesand consensus statements agree that surgeons need to



RECHERCHE

acquire a certain level of training.53–56 In 2 trials, surgeonswere required to have performed at least 20 laparoscopiccolectomies in order to be recruited.10,16 A few trials havetried to define the learning curve of laparoscopic colonresection, but these are primarily case series constituting alow level of evidence.57–60 Despite the absence of evidence,it is reasonable to assume that the skills acquired from per-forming simple laparoscopic procedures are transferableto more complex surgeries.59 In this sense, surgeons shouldfirst acquire experience through simple laparoscopicresections of benign lesions and then progressively inte-grate more complex procedures into their practices, in -cluding those involving cancer.

One prospective61 and 3 retrospective49–51 trials evalu-ated the impact of surgeon experience on oncologic out-comes following rectal cancer resection. Three of themshowed that operative duration decreased significantlywith the number of interventions performed.49–51 Park andcolleagues51 observed a plateau after 90 interventions fol-lowed by a decrease in operative duration, whereas Ito andcolleagues50 reported that operative duration de creasedfrom 228 to 179 minutes after more than 40 inter ventionshad been performed. All 4 trials also showed a significantdecrease in postoperative morbidity as the surgeon gainedmore experience (after 30–60 interventions had been per-formed, depending on the trial). Since inclusion criteriafor patients and for surgeons were not the same for all trials, a general conclusion regarding the minimum num-ber of cases to be performed to gain sufficient experiencecannot be drawn. Nonetheless, the available evidence un -doubtedly shows that surgeon experience and competencein laparoscopy for rectal cancer have a major impact onoutcome.

CONCLUSION

Analysis of the scientific literature confirmed that for thecurative treatment of colon and rectal cancer, laparoscopy isnot inferior to open surgery with respect to overall survival,disease-free survival and rate of recurrence. In addition,laparoscopic surgery provides short-term advantages overopen surgery, particularly a shorter hospital stay, reducedneed for analgesics, faster recovery of intestinal function, andan earlier return to activities of daily life. In contrast, laparo-scopic surgery requires a longer operative time.

Considering the evidence currently available, theCEPO recommends that laparoscopic resection be con-sidered an option for the curative treatment of colon andrectal cancer (grade A recommendation); that decisionsregarding surgical approach (laparoscopic or opensurgery) for the curative treatment of colon cancer takeinto consideration the surgeon’s experience, tumour stage,potential contraindications and patient expectations (gradeD recommendation); and that laparoscopic resection forrectal cancer be performed only by appropriately trained

surgeons who perform a sufficient volume annually tomaintain competence (grade D recommendation).

Acknowledgements: The CEPO thanks Dr. Paul Belliveau, surgeonat Hotel-Dieu Hospital, Kingston, Ont.; Dr. Michèle Brie, surgeon atCentre hospitalier régional de Baie‐Comeau, Baie-Comeau, Que.;Dr. Émilie M. Comeau, surgeon at Centre hospitalier universitaire deSherbrooke, Sherbrooke, Que.; Dr. Jean Couture, surgeon atHôtel‐Dieu de Lévis, Lévis, Que.; Dr. Roger C. Grégoire, surgeon atCHUQ, Hôpital Saint‐François d’Assise, Québec, Que.; Dr. AriMeguerditchian, surgeon at MUHC, Royal Victoria Hospital, Mon-tréal, Que.; Dr. Marc Pocard, surgeon at Hôpital Lariboisière, Paris;and Dr. Éric Poulin, surgeon at Ottawa Hospital, Ottawa, Ont., for theircontribution as external reviewers of the clinical practice guidelines Résection du cancer du côlon: comparaison de la chirurgie par laparoscopie à lachirurgie ouverte,” and “Résection du cancer du rectum: comparaison de la chi-rurgie par laparoscopie à la chirurgie ouverte, from which this article wasadapted.

The CEPO comprises Dr. Félix Couture, Hôtel‐Dieu de Québec(CHUQ); Dr. Isabelle Roy, Hôpital Notre‐Dame (CHUM); MélanieKavanagh, Direction québécoise du cancer (MSSS); Gino Boily,INESSS; Alain Bureau, Hôpital Sainte‐Croix (CSSS Drummond); JimBoulanger, INESSS; Dr. Ghislain Cournoyer, Hôpital régional deSaint‐Jérôme (CSSS de Saint‐Jérôme); Nicole Déry, INESSS; SuzanneFrenette, Hôpital Maisonneuve-Rosemont; Dr. Normand Gervais, Cen-tre hospitalier régional du Grand-Portage (CSSS de Rivière-du-Loup);Stéphanie Goulet, Direction québécoise du cancer (MSSS); Marie-Pas-cale Guay, Jewish General Hospital; Jean-Marie Lance, INESSS;Dr. Julie Lemieux, Hôpital Saint-Sacrement (CHAUQ); Dr. BernardLespérance, Hôpital du Sacré-Cœur de Montréal, Groupe d’étude enoncologie du Québec; Sylvain L’Espérance, INESSS; Nathalie Letarte,Hôpital Saint-Luc (CHUM), Programme de gestion thérapeutique desmédicaments; Mélanie Morneau, Direction québécoise du cancer(MSSS); Dr. Jean-François Ouellet, Hôtel-Dieu de Québec (CHUQ);Dr. Gilles Pineau, INESSS; Mireille Poirier, Hôtel-Dieu de Québec(CHUQ); Éric Potvin, INESSS; Dr. Raghu Rajan, Montreal GeneralHospital (MUHC); Dr. Benoît Samson, Hôpital Charles Lemoyne;Dr. Lucas Sideris, Hôpital Maisonneuve-Rosemont; Lucie Surprenant,St. Mary’s Hospital; Dr. François Vincent, Pavillon Sainte-Marie (CSSSde Trois-Rivières).

The CEPO is a group of specialists in oncology that reports to theCancer Control Department of Quebec’s Ministry of Health and SocialServices. The Quebec Cancer Control Department provided financialsupport for the work reported in this article, which is an adaptation ofCEPO’s original clinical practice guidelines, freely available in theirFrench version at www.msss.gouv.qc.ca/cancer.


Contributors: M. Morneau, J. Boulanger and N. Gervais designed thestudy and acquired the data, which all authors analyzed. M. Morneau,J. Boulanger, P. Charlebois and N. Gervais wrote the article, which allauthors reviewed and approved for publication.

References

1. Canadian Cancer Society’s Advisory Committee on Cancer Statistics.Canadian Cancer Statistics 2013. Toronto (ON): Canadian CancerSociety; 2013.

2. Wilkinson N, Scott-Conner CE. Surgical therapy for colorectaladen ocarcinoma. Gastroenterol Clin North Am 2008;37:253-67.

3. Berardi R, Maccaroni E, Onofri A, et al. Multidisciplinary treatmentof locally advanced rectal cancer: a literature review. Part 1. ExpertOpin Pharmacother 2009;10:2245-58.

4. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancersurgery — The clue to pelvic recurrence? Br J Surg 1982;69:613-6.



RESEARCH

5. Martel G, Boushey RP. Laparoscopic colon surgery: past, present andfuture. Surg Clin North Am 2006;86:867-97.

6. Koopmann MC, Heise CP. Laparoscopic and minimally invasiveresection of malignant colorectal disease. Surg Clin North Am 2008;88: 1047-72.

7. Jacobs M, Verdeja JC, Goldstein HS. Minimally invasive colon resec-tion (laparoscopic colectomy). Surg Laparosc Endosc 1991;1:144-50.

8. Nduka CC, Monson JR, Menzies-Gow N, et al. Abdominal wallmetastases following laparoscopy. Br J Surg 1994;81:648-52.

9. Berends FJ, Kazemier G, Bonjer HJ, et al. Subcutaneous metastasesafter laparoscopic colectomy. Lancet 1994;344:58.

10. Buunen M, Veldkamp R, Hop WC, et al. Survival after laparoscopicsurgery versus open surgery for colon cancer: long-term outcome ofa randomised clinical trial. Lancet Oncol 2009;10:44-52.

11. Jayne DG, Guillou PJ, Thorpe H, et al. Randomized trial of laparo-scopic assisted resection of colorectal carcinoma: 3-year results of theUK MRC CLASICC Trial Group. J Clin Oncol 2007;25:3061-68.

12. Moloo H, Haggar F, Martel G, et al. The adoption of laparoscopiccolorectal surgery: a national survey of general surgeons. Can J Surg2009;52:455-62.

13. Cook DJ, Guyatt GH, Laupacis A, et al. Rules of evidence and clin -ical recommendations on the use of antithrombotic agents. Chest1992;102(Suppl):305S-11S.

14. Veldkamp R, Kuhry E, Hop WC, et al. Laparoscopic surgery versusopen surgery for colon cancer: short-term outcomes of a randomisedtrial. Lancet Oncol 2005;6:477-84.

15. Janson M, Lindholm E, Anderberg B, et al. Randomized trial ofhealth-related quality of life after open and laparoscopic surgery forcolon cancer. Surg Endosc 2007;21:747-53.

16. Clinical Outcomes of Surgical Therapy Study Group. A comparisonof laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 2004;350:2050-9.

17. Fleshman J, Sargent DJ, Green E, et al. Laparoscopic colectomy forcancer is not inferior to open surgery based on 5-year data from theCOST Study Group trial. Ann Surg 2007;246:655-62.

18. Weeks JC, Nelson H, Gelber S, et al. Short-term quality-of-life out-comes following laparoscopic-assisted colectomy vs open colectomyfor colon cancer: a randomized trial. JAMA 2002;287:321-8.

19. Stucky CC, Pockaj BA, Novotny PJ, et al. Long-term follow-up andindividual item analysis of quality of life assessments related tolaparoscopic-assisted colectomy in the COST trial 93-46-53 (INT0146). Ann Surg Oncol 2011;18:2422-31.

20. Jayne DG, Thorpe HC, Copeland J, et al. Five-year follow-up of theMedical Research Council CLASICC trial of laparoscopically assistedversus open surgery for colorectal cancer. Br J Surg 2010;97: 1638-45.

21. Guillou PJ, Quirke P, Thorpe H, et al. Short-term endpoints of con-ventional versus laparoscopic-assisted surgery in patients with colo -rectal cancer (MRC CLASICC trial): multicentre, randomised con-trolled trial. Lancet 2005;365:1718-26.

22. Lacy AM, Delgado S, Castells A, et al. The long-term results of a ran-domized clinical trial of laparoscopy-assisted versus open surgery forcolon cancer. Ann Surg 2008;248:1-7.

23. Lacy AM, Garcia-Valdecasas JC, Delgado S, et al. Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomised trial. Lancet 2002;359:2224-9.

24. Liang JT, Huang KC, Lai HS, et al. Oncologic results of laparoscopicversus conventional open surgery for stage II or III left-sided coloncancers: a randomized controlled trial. Ann Surg Oncol 2007;14:109-17.

25. Vlug MS, Wind J, Hollmann MW, et al. Laparoscopy in combinationwith fast track multimodal management is the best perioperativestrategy in patients undergoing colonic surgery: a randomized clinicaltrial (LAFA-study). Ann Surg 2011;254:868-75.

26. Neudecker J, Klein F, Bittner R, et al. Short-term outcomes from aprospective randomized trial comparing laparoscopic and opensurgery for colorectal cancer. Br J Surg 2009;96:1458-67.

27. Hewett PJ, Allardyce RA, Bagshaw PF, et al. Short-term outcomes ofthe Australasian randomized clinical study comparing laparoscopicand conventional open surgical treatments for colon cancer: theALCCaS trial. Ann Surg 2008;248:728-38.

28. Wu Z, Zhang S, Aung LH, et al. Lymph node harvested in laparo-scopic versus open colorectal cancer approaches: a meta-analysis.Surg Laparosc Endosc Percutan Tech 2012;22:5-11.

29. Ma Y, Yang Z, Qin H, et al. A meta-analysis of laparoscopy comparedwith open colorectal resection for colorectal cancer. Med Oncol 2011;28: 925-33.

30. Bai HL, Chen B, Zhou Y, et al. Five-year long-term outcomes oflaparoscopic surgery for colon cancer. World J Gastroenterol 2010;16:4992-7.

31. Kuhry E, Schwenk W, Gaupset R, et al. Long-term outcome of laparo-scopic surgery for colorectal cancer: a cochrane systematic review ofrandomised controlled trials. Cancer Treat Rev 2008;34:498-504.

32. Bonjer HJ, Hop WC, Nelson H, et al. Laparoscopically assisted vsopen colectomy for colon cancer: a meta-analysis. Arch Surg 2007;142: 298-303.

33. Liang X, Hou S, Liu H, et al. Effectiveness and safety of laparoscopicresection versus open surgery in patients with rectal cancer: a ran-domized, controlled trial from China. J Laparoendosc Adv Surg Tech A2011;21:381-5.

34. Kang SB, Park JW, Jeong SY, et al. Open versus laparoscopic surgeryfor mid or low rectal cancer after neoadjuvant chemoradiotherapy(COREAN trial): short-term outcomes of an open-label randomisedcontrolled trial. Lancet Oncol 2010;11:637-45.

35. Jayne DG, Brown JM, Thorpe H, et al. Bladder and sexual functionfollowing resection for rectal cancer in a randomized clinical trial oflaparoscopic versus open technique. Br J Surg 2005;92:1124-32.

36. Lujan J, Valero G, Hernandez Q, et al. Randomized clinical trialcomparing laparoscopic and open surgery in patients with rectal can-cer. Br J Surg 2009;96:982-9.

37. Ng SS, Leung KL, Lee JF, et al. Long-term morbidity and oncologicoutcomes of laparoscopic-assisted anterior resection for upper rectalcancer: ten-year results of a prospective, randomized trial. Dis ColonRectum 2009;52:558-66.

38. Leung KL, Kwok SP, Lam SC, et al. Laparoscopic resection of rectosig-moid carcinoma: prospective randomised trial. Lancet 2004;363: 1187-92.

39. Ng SS, Leung KL, Lee JF, et al. Laparoscopic-assisted versus openabdominoperineal resection for low rectal cancer: a prospective ran-domized trial. Ann Surg Oncol 2008;15:2418-25.

40. Pechlivanides G, Gouvas N, Tsiaoussis J, et al. Lymph node clearanceafter total mesorectal excision for rectal cancer: laparoscopic versusopen approach. Dig Dis 2007;25:94-9.

41. Braga M, Frasson M, Vignali A, et al. Laparoscopic resection in rectal



RECHERCHE

cancer patients: outcome and cost-benefit analysis. Dis Colon Rectum2007;50:464-71.

42. Zhou ZG, Hu M, Li Y, et al. Laparoscopic versus open total mesorec-tal excision with anal sphincter preservation for low rectal cancer.Surg Endosc 2004;18:1211-5.

43. Huang MJ, Liang JL, Wang H, et al. Laparoscopic-assisted versusopen surgery for rectal cancer: a meta-analysis of randomized con-trolled trials on oncologic adequacy of resection and long-term onco-logic outcomes. Int J Colorectal Dis 2011;26:415-21.

44. Ohtani H, Tamamori Y, Azuma T, et al. A meta-analysis of the short-and long-term results of randomized controlled trials that comparedlaparoscopy-assisted and conventional open surgery for rectal cancer.J Gastrointest Surg 2011;15(8):1375-85.

45. Anderson C, Uman G, Pigazzi A. Oncologic outcomes of laparoscopicsurgery for rectal cancer: a systematic review and meta-analysis of theliterature. Eur J Surg Oncol 2008;34:1135-42.

46. Aziz O, Constantinides V, Tekkis PP, et al. Laparoscopic versus opensurgery for rectal cancer: a meta-analysis. Ann Surg Oncol 2006;13:413-24.

47. Gao F, Cao YF, Chen LS. Meta-analysis of short-term outcomes afterlaparoscopic resection for rectal cancer. Int J Colorectal Dis 2006;21: 652-6.

48. Piaggio G, Elbourne DR, Altman DG, et al. Reporting of noninferior-ity and equivalence randomized trials: an extension of the CONSORTstatement. JAMA 2006;295:1152-60.

49. Balik E, Asoglu O, Saglam S, et al. Effects of surgical laparoscopicexperience on the short-term postoperative outcome of rectal cancer:results of a high volume single center institution. Surg Laparosc EndoscPercutan Tech 2010;20:93-9.

50. Ito M, Sugito M, Kobayashi A, et al. Influence of learning curve onshort-term results after laparoscopic resection for rectal cancer. SurgEndosc 2009;23:403-8.

51. Park IJ, Choi GS, Lim KH, et al. Multidimensional analysis of thelearning curve for laparoscopic resection in rectal cancer. J GastrointestSurg 2009;13:275-81.

52. Thorpe H, Jayne DG, Guillou PJ, et al. Patient factors influencingconversion from laparoscopically assisted to open surgery for colo -rectal cancer. Br J Surg 2008;95:199-205.

53. National Institute for Health and Clinical Excellence. Colorectalcancer — laparoscopic surgery (review): guidance. 2006 Aug. 23.Available: http://guidance.nice.org.uk/TA105/Guidance/pdf/English.(accessed 2009 Oct. 1).

54. Smith A, Rumble RB, Langer B, et al. Laparoscopic Colon CancerSurgery Expert Panel and Program in Evidence-based Care. Toronto:CCO Evidence-based series; 2005. Available: www .cancercare .on .ca/common /pages /UserFile .aspx?fileId=13988 (ac ces sed 2009 Oct. 1).

55. Society of American Gastrointestinal and Endoscopic Surgeons. Guide-lines for laparoscopic resection of curable colon and rectal cancer. 2009.Available: www .sages .org /publications /guidelines /guidelines -for -laparoscopic-resection-of-curable-colon-and-rectal-cancer/ (acces -sed 2009 Oct. 1).

56. National Comprehensive Cancer Network. Colon Cancer. V.3.2009.Available: www.nccn.org/professionals/physician_gls/PDF /colon.pdf(accessed 2009 Oct. 1).

57. Choi DH, Jeong WK, Lim SW, et al. Learning curves for laparo-scopic sigmoidectomy used to manage curable sigmoid colon can-cer: single-institute, three-surgeon experience. Surg Endosc 2009;23:622-8.

58. Dincler S, Koller MT, Steurer J, et al. Multidimensional analysis oflearning curves in laparoscopic sigmoid resection: eight-year results.Dis Colon Rectum 2003;46:1371-8, discussion 1378-9.

59. Schlachta CM, Mamazza J, Seshadri PA, et al. Defining a learningcurve for laparoscopic colorectal resections. Dis Colon Rectum 2001;44:217-22.

60. Tekkis PP, Senagore AJ, Delaney CP, et al. Evaluation of the learningcurve in laparoscopic colorectal surgery: comparison of right-sidedand left-sided resections. Ann Surg 2005;242:83-91.

61. Bege T, Lelong B, Esterni B, et al. The learning curve for the laparo-scopic approach to conservative mesorectal excision for rectal cancer:lessons drawn from a single institution’s experience. Ann Surg 2010;51:249-53.

Online manuscriptsubmission and peer review

AVAILABLE at

http://mc.manuscriptcentral.com/cjs




Comparison of the major intraoperative andpostoperative complications between unilateraland sequential bilateral total knee arthroplasty ina high-volume community hospital

Background: Total knee arthroplasty (TKA) is a common surgical treatment forarthritis. In the event of bilateral knee symptoms, a patient may elect for bilateralTKA (BTKA) under 1 anesthetic or 2 separate unilateral TKAs (UTKA). Controversyexists in the literature regarding the safety of BTKA versus UTKA. We compared therate of major intraoperative and postoperative complications for BTKA versus UTKAat a high-volume community hospital.

Methods: We compared 373 patients who underwent BTKA with 966 who under-went UTKA between May 2008 and May 2011. Health records were used to deter-mine patient characteristics and major intraoperative and postoperative complications.The BTKA and UTKA cohorts were matched for demographic characteristics andcomorbidities with the exception of previous transient ischemic attack and previousknee surgery (UTKA > BTKA).

Results: Rates of intraoperative and postoperative complications, including cardio-vascular, thromboembolic and neurologic complications; deep wound infections; andmortality, did not differ significantly between groups. Bilateral TKA was associatedwith a greater proportion of patients requiring blood transfusion than UTKA (29.8%v. 8.9%, p < 0.001). Among those transfused, there was no significant differencebetween the groups in the mean number of units required (1.72 ± 0.77 v. 1.53 ±0.85 units, p = 0.68).

Conclusion: Bilateral TKA was not associated with statistically greater rates of intra-operative and postoperative complications than UTKA, barring the proportion ofpatients requiring transfusion. Our results support the use of BTKA to treat bilateralknee arthritis in a high-volume community hospital setting.

Contexte : La prothèse (ou arthroplastie) totale du genou (PTG) est un traitementchirurgical courant contre l’arthrite. Quand les 2 genoux sont atteints, le patient peutchoisir entre une PTG bilatérale (PTGB), qui ne nécessitera qu’une seule anesthésie,ou 2 interventions unilatérales distinctes (PTGU). Dans la littérature, on ne semblepas s’entendre sur l’innocuité de la PTGB contre la PTGU. Nous avons comparé lestaux de complications peropératoires et postopératoires majeures associées aux PTGBet aux PTGU dans un hôpital communautaire où s’effectue un volume élevé de tellesinterventions.

Méthodes : Nous avons comparé 373 patients qui ont subi une PTGB à 966 qui ontsubi une PTGU entre mai 2008 et mai 2011. Nous avons consulté les dossiers médi-caux pour établir les caractéristiques des patients et relever les complications per-opératoires et postopératoires majeures. Les cohortes soumises à la PTGB et à laPTBU ont été assorties en fonction des caractéristiques démographiques et descomorbidités, à l’exception des antécédents d’accidents ischémiques transitoires etd’interventions chirurgicales du genou (PTGU > PTGB).

Résultats : Les taux de complications peropératoires et postopératoires, y compris car-diovasculaires, thromboemboliques et neurologiques, les infections de plaies profondeset la mortalité n’ont pas varié significativement entre les groupes. Une proportion plusgrande de patients soumis à la PTGB a nécessité une transfusion sanguine comparative-ment aux patients soumis à la PTGU (29,8 % c. 8,9 %, p < 0,001). Parmi les receveursde transfusions, on n’a noté aucune différence significative entre les groupes quant aunombre moyen d’unités requises (1,72 ± 0,77 c. 1,53 ± 0,85 unité, p = 0,68).

Erin Spicer, MSc*†‡

Garry Robert Thomas, MSc*†‡

Edward John Rumble, MD+

From the *Faculty of Medicine, Univer-sity of Toronto, Toronto, Ont. and the†Division of Orthopaedic Surgery, NorthYork General Hospital, Toronto, Ont.

‡Co–first authors

Accepted for publicationOct. 24, 2012

Correspondence to:E.J. Rumble5 Fairview Mall Dr. Suite 377North York ON M2J [email protected]

DOI: 10.1503/cjs.012912

compar-spicer_Layout 1 13-09-16 9:42 AM Page 311


RECHERCHE

T otal knee arthroplasty (TKA) is the most commonlyperformed joint replacement surgery in Canada.1,2 Itis considered a safe and reliable procedure for pro-

viding pain relief, correcting limb alignment and improvingquality of life in patients with moderate to severe arthritis.3

A substantial proportion of these patients present withbilateral symptomatic joint disease and may benefit frombilateral replacements. In these circumstances, the patientand orthopaedic surgeon must together decide whether topursue a bilateral knee replacement under a single anes-thetic (BTKA) or 2 separate unilateral TKAs (UTKA).3

Critics of BTKA argue that the procedure increases therisk for serious complications.3–9 The TKA complicationsmost often referenced in the literature are cardiovascular(i.e., new-onset arrhythmia, myocardial infarction, conges-tive heart failure), thromboembolic (i.e., deep vein throm-bosis, pulmonary embolism), neurologic (i.e., transientischemic attack, stroke, delirium), blood loss measureddirectly or indirectly through blood transfusion require-ments and mortality.6 The true risk associated with BTKAremains controversial, as many studies opposing the pro ced -ure have reported predominantly nonstatistical differencesin complication rates between BTKA and UTKA.3,10–13

Proponents of BTKA, however, report numerous ad -van tages, including decreased total anesthetic, total rehab -ilitation time, length of hospital stay and institution costs.These patients have also been found to have lower rates ofsuperficial wound infections and to report increased con -ven ience and satisfaction.11–19 One study also found that at10 years postoperatively the BTKA group had a signifi-cantly higher rate of survival than the UTKA group, inde-pendent of other contributing factors.10

The number of TKAs being performed in Canada isprojected to continue rising given the prevalence of degen-erative osteoarthritis in the aging population. While mostTKAs are performed in community hospitals, we are notaware of any literature comparing the rates of major com-plications of BTKA with UTKA within the context of acommunity hospital.20

Our objective was to compare the rates of major intra-operative and postoperative complications between pa -tients undergoing BTKA and those undergoing UTKA ata high-volume community hospital.

METHODS

Study population

We searched the surgical database to retrospectively iden-

tify all patients who had undergone a BTKA or a UTKAbetween May 2008 and May 2011. Patients who had anincomplete charted medical history, multiple proceduresunder the same anesthetic or a revision TKA wereexcluded from the analysis. Patients who underwentstaged TKA (2 distinct surgeries on both knees within a 1-year period) were excluded from the study because of thesmall population size and owing to suggestions in the lit-erature that retrospective analysis of staged procedurescan be misleading.21 Furthermore, patients who were sub-ject to blood conservation techniques, including autolo-gous blood transfusion, erythropoietin, and tranexamicacid (TXA) were excluded from the study. All patientsunderwent a preoperative assessment by an anesthesiolo-gist to ensure suitability for surgery, and those patientswho were found to need optimization of their comorbid -ities were referred to internal medicine. All candidateswith bilateral knee symptoms who were deemed eligiblefor surgery were given the option of BTKA or 2 UTKAs.Our institution’s research ethics board approved our studyprotocol.

Operative and postoperative procedures

A pool of 7 staff surgeons performed the TKAs using astandardized approach. Whereas in some centres, 2 sur-geons operate in parallel in patients undergoing BTKA,22

the BTKAs in our study were performed sequentially by asingle surgeon, with the patient under a single anesthetic.The patients received either general or regional anestheticwith sedation and were given routine preoperative intra-venous antibiotics. Procedures were performed undertourniquet control. All TKAs were done with intra -medullary femoral alignment and extramedullary tibialalignment. The prostheses were produced by Zimmeror Biomet and were either cruciate-sparing or posteriorstabilized.

Postoperatively, all patients were under the care of thesame hospitalist. Supplemental oxygen was administereduntil oxygen saturation was maintained at a minimum of92% on room air. Pain control was achieved using oralanalgesics supplemented by patient-controlled analgesia.All patients received routine thromboprophylaxis for 2–3 weeks. Continuous passive motion was started within24 hours of surgery. On the first postoperative day, patientswere mobilized, fully weight bearing, under the supervisionof physiotherapists. Patients were typically discharged bythe fourth day or transferred to a rehabilitation institutionas required.

Conclusion : La PTGB n’a pas été associée à des taux statistiquement plus élevés decomplications peropératoires et postopératoires comparativement à la PTGU, à l’ex-ception de la proportion de patients ayant nécessité une transfusion. Nos résultatsappuient le recours à la PTGB pour traiter l’arthrite bilatérale du genou dans le con-texte d’un hôpital communautaire ou le volume de ces interventions est élevé.



RESEARCH

Medical records

Two of us (E.S. and G.R.T.) collected and independentlyreviewed and verified all data from the patients’ hospitalrecords. Patient demographic characteristics and medicalhistory were determined by reviewing all standard preop-erative orthopedic, anesthesiology and internal medicinerecords. Intraoperative complications, if any, were identi-fied in operative and/or consultation notes. Patient pro -gress notes, consultation notes, discharge summaries,readmission records and follow-up appointment notes forup to 3 months were screened to identify any postopera-tive complications. Transfusion data were obtained fromthe institution blood bank records.

Statistical analysis

Continuous variables are expressed as means ± standarddeviation. We compared these data with an unpaired, 2-tailed Student t test using Microsoft Excel software.Dichotomous variables are expressed as numbers and per-centages, and we analyzed these data with a 2-tailed Fisherexact test using GraphPad Prism. We considered results tobe significant at p < 0.05. We calculated odds ratios (OR)and 95% confidence intervals (CIs) for all data.

RESULTS

Based on our study criteria, we excluded 36 patients fromthe UTKA group and 12 from the BTKA group. Of thosepatients who met our inclusion criteria, 966 underwentUTKA (65.9% women v. 34.1% men, mean age 69.6 ±10.2 yr) and 373 underwent BTKA (71.0% women v.29.0% men, mean age 69.1 ± 9.1 yr). No patients were lostto follow-up. There were no significant differencesbetween the cohorts in terms of sex, age, body mass index(BMI) or smoking history. In both cohorts, there werenearly twice as many women as men (Table 1).

We compared the groups for preoperative comorbid -ities, including hypertension, arrhythmia, coronary arterydisease (CAD), myocardial infarction (MI), congestiveheart failure (CHF), deep vein thrombosis (DVT), pul-monary embolism (PE), stroke, transient ischemic attack

(TIA), dementia, asthma, chronic obstructive pulmonarydisease (COPD), renal disease, diabetes mellitus, previousknee surgery, preoperative hemoglobin and anemia (hemo-globin < 130 g/L).23 No significant differences existedbetween cohorts except for incidence of TIA and previousknee surgery. A greater proportion of patients undergoingUTKA had a history of TIA than those undergoing BTKA(3.4% v. 1.1%, p = 0.015). The patients underoing UTKAhad significantly more previous knee operations, includingarthroscopic débridement, than patients undergoingBTKA (20.3% v. 5.6%, p < 0.001; Table 2).

There were no significant differences between theUTKA and BTKA cohorts for intraoperative and postop-erative cardiovascular, thromboembolic or neurologiccomplications; deep wound infections; or mortality. Inaddition, there were no significant differences in the ratesof respiratory failure, renal failure or length of hospital staybetween cohorts. The sole significant difference identifiedbetween cohorts was the greater proportion of BTKApatients requiring blood transfusions (29.8% v. 8.9%,p < 0.001). Among those patients in the BTKA and UTKAcohorts who were transfused, there was no significant dif-ference in the number of blood units required (1.72 ± 0.77v. 1.53 ± 0.85 units, p = 0.68; Table 3).

As part of a further analysis, we assessed the UTKA andBTKA cohorts to identify factors that may have predis-posed patients to transfusion. All patients who receivedtransfusion had preoperative anemia (p < 0.001). PreviousMI was significantly more common among transfused thannontransfused patients in the UTKA cohort (14.0% v.5.5%, p = 0.008); we did not observe a similar trend in theBTKA cohort. Transfused patients were significantly olderthan nontransfused patients in the UTKA cohort (76.1 ±8.60 yr v. 69.2 ± 10.12 yr, p < 0.001); again, we did notobserve a similar trend in the BTKA cohort.

DISCUSSION

The number of TKAs being performed in Canada is pro-jected to continue rising, given the aging population.1,2 Theliterature regarding the relative safety of BTKA remainscontroversial. While some studies supporting the use ofBTKA have shown comparable complication rates to that

Table 1. Demographic characteristics of patients undergoing bilateral or unilateral total knee arthroplasty

Characteristic UTKA, n = 966 BTKA, n = 373 p value OR (95% CI)

Sex

Male, no. (%) 329 (34.1) 108 (29.0) 0.08 1.3 (1.0–1.6)

Female, no. (%) 637 (65.9) 265 (71.0) 0.08 0.8 (0.6–1.0)

Age, mean ± SD yr 69.6 ± 10.2 69.1 ± 9.1 0.26 1.01 (0.99–1.02)

BMI, mean ± SD 32.5 ± 7.0 32.2 ± 6.0 0.38 1.01 (0.99–1.03)

Smoking, no. (%) 86 (8.9) 24 (6.2) 0.12 1.4 (0.9–2.3)

BMI = Body Mass Index; BTKA = bilateral total knee arthroplasty; CI = con"dence interval; OR = odds ratio; SD = standard deviation; UTKA= unilateral total knee arthroplasty.



RECHERCHE

of UTKA, critics of BTKA maintain that the procedureincreases the risk of serious complications, including car-diovascular, thromboembolic and neurologic events anddeath.3–9,13,14,24,25 Therefore, we conducted the present studyto determine whether it is safe to perform BTKA at ahigh-volume community hospital.

Our study demonstrated no significant differences inintraoperative and postoperative complications betweenpatients undergoing BTKA and those undergoing UTKAwith the exception of an increased proportion of patientsin the BTKA cohort requiring blood transfusions. It hasbeen suggested that patients undergoing BTKA are atincreased risk for cardiovascular complications due togreater intraoperative blood loss leading to substantialhemodynamic shifts and ischemia.4 A potential explanationfor the comparable cardiovascular outcomes betweencohorts in our study may be that the blood loss experi-enced by patients in the BTKA group was not substantialenough to precipitate more cardiovascular events. Supportfor this explanation may lie in the fact that the meantransfusion requirement for our patients undergoingBTKA was 1.72 ± 0.77 units, whereas the literaturereports requirements of 3.9 units for these patients.26 Fur-thermore, 1 study that identified a greater proportion ofpostoperative CHF in the BTKA cohort examined apopu lation aged older than 80 years,4 which was markedlyolder than the mean age of both cohorts in the present

study (69.6 ± 10.2 yr in the UTKA group and 69.1 ±9.1 yr in the BTKA group).

The reported incidence of thromboembolic events fol-lowing a TKA procedure varies widely in the literature,ranging from 0.4% to 71% for DVT and 0% to 3% forPE.6 Studies demonstrating high rates, however, used radi-ological imaging to detect thromboembolism. Radiologicalimaging often detects clinically asymptomatic thrombosis,whereas our study considered only symptomatic events.Consistent with our findings, several studies consideringonly symptomatic DVT or PE have demonstrated no sta-tistically significant differences between UKTA and BTKAcohorts.4,15 One possible explanation for this may be thatwhile BTKA increases recovery time and predisposespatients to immobility-related thromboembolic risk,patients undergoing BTKA tend to be more hypocoagu -lable in the postoperative period due to a greater intraop-erative consumption of coagulation factors.27

The literature reports a 3-month mortality of 0.46%associated with TKA.28 In addition, 1 study of in-hospitalmortality found that 0.5% of patients who had BTKA diedcompared with 0.3% of patients who had UTKA.29 In con-trast, our study identified lower mortality among patientswho had UTKA, with no significant difference betweencohorts (0.3% in the BTKA group v. 0.1% in the UTKAgroup, p = 0.48). Although the literature predominantlyreflects higher mortality in patients who had BTKA, many

Table 2. Comorbidities of patients undergoing bilateral or unilateral total knee arthroplasty

Group, no. (%)*

Comorbidity UTKA, n = 966 BTKA, n = 373 p value OR (95% CI)

Cardiovascular

Hypertension 553 (57.2) 222 (59.5) 0.46 0.9 (0.7–1.2)

Arrhythmia 107 (11.1) 29 (7.8) 0.09 1.5 (1.0–2.3)

Coronary artery disease 108 (11.2) 38 (10.2) 0.63 1.1 (0.8–1.6)

Myocardial infarction 60 (6.2) 25 (6.7) 0.71 0.9 (0.6–1.5)

Chronic heart failure 14 (1.4) 4 (1.1) 0.79 1.4 (0.4–4.2)

Thromboembolic

Deep vein thrombosis 23 (2.4) 11 (2.9) 0.56 0.8 (0.4–1.7)

Pulmonary embolism 8 (0.8) 3 (0.8) > 0.99 1.0 (0.3–3.9)

Neurologic

Transient ischemic attack 33 (3.4) 4 (1.1) 0.015 3.3 (1.1–9.3)

Stroke 25 (2.6) 7 (1.9) 0.55 1.4 (0.6–3.2)

Dementia 8 (0.8) 1 (0.3) 0.46 3.1 (0.4–24.9)

Respiratory

Asthma 97 (10.0) 34 (9.1) 0.68 1.1 (0.7–1.7)

COPD 40 (4.1) 10 (2.7) 0.26 1.6 (0.8–3.2)

Renal failure 32 (3.3) 6 (1.6) 0.10 2.1 (0.9–5.1)

Diabetes 172 (17.8) 72 (19.3) 0.53 0.9 (0.7–1.2)

Previous knee surgery 196 (20.3) 21 (5.6) < 0.001 4.3 (2.7–6.8)

Preop Hb status

Preop Hb, mean ± SD, g/L 134.7 ± 13.5 133.1 ± 13.7 0.21 1.01 (1.0–1.02)

Preop anemia,† mean ± SD, g/L 320.0 ± 33.1 118.0 ± 31.6 0.65 1.1 (0.8–1.4)

BTKA = bilateral total knee arthroplasty; CI = con"dence interval; COPD = chronic obstructive pulmonary disease; Hb = hemoglobin; OR = odds ratio; SD = standard deviation; UTKA= unilateral total knee arthroplasty. *Unless otherwise indicated. †Hb < 130 g/L.



RESEARCH

authors have acknowledged that there may be no differ-ence between procedures if performed at high-volumeinstitutions with specialized TKA surgeons.29,30 At our insti-tution, 500–600 BTKAs and UTKAs are collectively per-formed each year, making it a high- volume, specializedfacility.31 Furthermore, critics of BTKA claim that the studies demonstrating no difference in mortality betweencohorts are misleading, as BTKA candidates are oftenselected for their benign medical history, thereby introduc-ing a selection bias.32 Both populations in the present study,however, were comparable in terms of demographic char-acteristics and comorbidity profiles. In addition, ourpatients were not assigned to a particular procedure basedon comordbidities; rather, they chose the procedure theywould undergo.

The only significant difference between the BTKA andUTKA cohorts in the present study was the proportion ofpatients requiring blood transfusion of at least 1 unit (8.9%v. 29.8%, p < 0.001). All transfused patients, regardless ofcohort, had preoperative anemia, which is the strongestpredictor of transfusion.23 Studies show that the blood lossassociated with TKA can be as high as 2.2 units per kneereplaced due to extensive bone and soft tissue cuts; there-fore, it would be expected that operating on 2 knees ratherthan 1 knee predisposes the patient to a greater risk ofblood loss and the need for transfusion.9,33 In fact, it hasbeen reported that during BTKA, the blood loss from thesecond knee replacement is greater than that from the first,

likely due to decreased tissue clotting factors associatedwith intraoperative tissue trauma, hypothermia or hypox-emia.34 Interestingly, our study demonstrated that amongpatients being transfused, those in the BTKA cohortrequired a greater number of blood units than those in theUTKA cohort, but this finding was not significant (1.72 ±0.77 v. 1.53 ± 0.85 units, p = 0.68). While this outcome mayseem paradoxical, it is important to note that transfusionrequirement is a useful, albeit imperfect, substitute meas -urement for blood loss. Blood transfusion is a second linetreatment for hemodynamically unstable patients owing totransfusion-related risks; therefore, the trigger for transfu-sion is often reserved for patients with a hemoglobin levelof less than 70 g/L, and patients are often transfused untilstable rather than euvolemic.35 Prophylactically, the use ofantifibrinolytics has been shown to markedly reduce bloodloss and transfusion requirements during TKA.36 Intraop-erative use of antifibrinolytic TXA has recently been incor-porated into the TKA procedures at our institution toaddress the need for blood transfusion, particularly inpatients undergoing BTKA. Preliminary data demonstratethat use of TXA has resulted in lowered transfusion ratesfor patients undergoing BTKA.

In contrast to the present study, several larger bodies ofwork have found that while the absolute number of com-plications associated with any TKA procedure is small, therelative risk of cardiovascular complications, PE and deathis greater for BTKA.32,37 The results of our community

Table 3. Procedure-related complications in patients who had bilateral or unilateral total knee arthroplasty

Group, no. (%)*

Complication UTKA, n = 966 BTKA, n = 373 p value OR (95% CI)

Cardiovascular

Arrhythmia 8 (0.8) 1 (0.3) 0.46 3.1 (0.4–24.9)

Myocardial infarction 4 (0.4) 4 (1.0) 0.23 0.4 (0.1–1.5)

Chronic heart failure 10 (1.0) 0 (0) 0.07 NA

Total 22 (2.3) 5 (1.3) 0.39 1.7 (0.6–4.6)

Thromboembolic

Deep vein thrombosis 5 (0.5) 2 (0.5) > 0.99 1.0 (0.2–5.0)

Pulmonary embolism 3 (0.3) 1 (0.3) > 0.99 1.2 (0.1–11.2)

Total 8 (0.8) 3 (0.8) > 0.99 1.0 (0.3–3.9)

Neurologic

TIA/stroke 2 (0.2) 0 (0.0) > 0.99 NA

Delirium 12 (1.2) 3 (0.8) 0.77 1.6 (0.4–5.5)

Total 14 (1.4) 4 (1.1) 0.79 1.4 (0.4–4.1)

Respiratory failure 4 (0.4) 2 (0.5) 0.67 0.8 (0.1–4.2)

Renal failure 1 (0.1) 0 (0.0) 0.28 NA

Deep infection 4 (0.4) 1 (0.3) > 0.99 1.5 (0.2–13.9)

Mortality 1 (0.1) 1 (0.3) 0.48 0.4 (0–6.2)

Hospital stay, mean ± SD, d 4.9 ± 2.9 5.3 ± 2.1 0.36 0.92 (0.88–0.98)

Transfusion

Transfused 86 (8.9) 111 (29.8) < 0.001 0.2 (0.2–0.3)

Units, mean ± SD 1.53 ± 0.85 1.72 ± 0.77 0.68 0.89 (0.84–0.94)

BTKA= bilateral total knee arthroplasty; CAD = coronary artery disease; CI = con"dence interval; NA = not available; OR = odds ratio; SD = standard deviation; TIA = transient ischemic attack; UTKA= unilateral total knee arthroplasty. *Unless otherwise indicated.



RECHERCHE

hospital–based studymay thereforenotbeuniversallyapplicabletoallcentreswhereBTKAisperformed.Giventhecontroversial statusofBTKAsafety in the literature,patientsshouldbemadeawareofthemajorintraoperativeandpostoperative complications identified in the largercohortstudieswhendecidingwhethertoundergoelectivebilateralorunilateralkneereplacementsurgery.

Limitations

Ourstudyhadseverallimitationsthatshouldbeacknow-ledged.First,despitehavingastandardTKAapproachatourinstitution,ourstudywasbasedonthedataof7sur-geons,eachofwhommayhaveintroducedaminordegreeofprocedure variability.Nevertheless,we felt that com-biningtheresultsfromdifferentsurgeonsyieldedamoreaccuraterepresentationofTKAoutcomesatacommunityhospital andwouldenhance the external validityofourstudy.Second, reviewingonlyhospital recordsmaynotidentify all patient complications thatoccurred afterdis-charge.For instance, if apatient visited anotherhospitalwith an acute complication,our institutiondatabasemaynothavearecordofthatvisit.Thepotentialforthiserrorwasminimized,however, asno studypatientwas lost tofollow-upandall patientswere seenwithin thefirst3 postoperativemonths tomonitorkneeprogress andscreenforcomplications.Finally,itmaybesuggestedthata comparisonbetweenBTKAand stagedTKA ismorevaluablethanacomparisonofBTKAandUTKA.Wein-itially incorporated stagedTKA,but the sample sizewastoo small for the analysis tohave adequate statisticalpower. In addition, stagedTKA introduces confoundingfactors, suchas2distinctanestheticexposuresand2dis-tinctpostoperativerecoveryperiods,thatdonotexistinacomparisonbetweenBTKAandUTKA.Inaddition, theliteraturesuggeststhattheoperativeriskofstagedTKAisunderestimated in retrospective studies, aspatientswhoexperiencecomplicationsduring thefirst surgerymaybeunlikely toparticipate in a second surgery and therebybecomeaUTKApatient.21 Individualswhodecline thesecondoperation are likely to continueexperiencingsymptomsofarthritis,whichsuggeststhattheymayhavebeenbetterservedbya1-stepBTKA.This studyhas several strengths comparedwith the

existingbodyofliterature.First,havingthestudybasedatahigh-volumehospitalgenerateda largenumberof eli-giblepatientsinbothcohortsforcomparison.Second,thisstudyuniquelyexaminedtheTKAoutcomesatacommun-ity institutionwhereall surgeries areperformedby staffsurgeons,whereasmoststudiesintheliteraturewerecon-ductedatacademiccentreswheretraineesalsoparticipatedinoperations.Finally, the samehospitalistmanagedallpatientsinthepostoperativesettingatourinstitution,min-imizingthepossibiltythatvariabilityinpostoperativecareinfluencedcomplicationrates.

CONCLUSION

Thepurposeofourstudywas todeterminethesafetyofperformingBTKAprocedures at ahigh-volumecom-munityhospital.While several large scale studies,whichuseddata fromacademic centres,have identified relativeriskscorrelatedwithBTKA,ourstudydidnotfindBTKAtobe associatedwith increased cardiovascular, throm-boembolic,orneurologic complications;deepwoundinfection;ordeath comparedwithUTKA.Thepropor-tionofpatients requiringblood transfusionswas signifi-cantlygreater in theBTKAcohort,but thenumberofunitsrequiredpertransfusedpatientwasnotsignificantlydifferent.Recenteffortshavebeenmadeatourinstitutiontoeffectivelyreducetheneedfortransfusionbyintroduc-ingtheuseofTXA.WeconcludethatBTKAcanbeasafetreatment forbilateral knee arthritis in the contextof ahigh-volumecommunityhospitalwith experiencedTKAsurgeons,hospitalists, paramedical staff and ready accesstopostoperativeinpatientrehabilitation.

Acknowledgements: Wearegrateful to themembersof theDepart-mentofOrthopaedicsandDr.I.FefermanatNorthYorkGeneralHos-pital forproviding studydirection.Wearegrateful to theDr.EmilOrsiniMemorialFundforprovidingfinancialsupportintheformofascholarship for authorsE.Spicer andG.R.Thomas.We thankDr. C. Peskun for assistingwith thedata analysis and reviewing thepaperandDrs.G. ChanandH.Mehdianforprovidingadditionalguid-anceintheprojectdesign.

Competing Interests:Nonedeclared.

Contributors: E.SpicerandG.R.Thomascontributedtoallaspectsofthiswork.E.J.Rumbledesignedthestudy,analyzedthedata,reviewedthearticleandapprovedthefinalversionforpublication.

References

1. CanadianInstituteforHealthInformation.Specialized Services: JointReplacements. Available: www.cihi.ca/CIHI-ext-portal/internet/EN/TabbedContent/types+of+care/specialized+services/joint+replacements/cihi021359(accessed2011Sept.4).

2. LawrenceRC,HelmickC,ArnettF,etal.EstimatesoftheprevalenceofarthritisandselectedmuscloskeletaldisordersintheUnitedStates.Arthritis Rheum 1998;41:778-99.

3. BullockDP,SporerSM,ShirreffsTGJr.Comparisonof simultan-eousbilateralwithunilateraltotalkneearthroplastyintermsofperi-operativecomplications.J Bone Joint Surg Am 2003;85A:1981-6.

4. LynchNM,TrousdaleRT, IlstrupDM.Complications after con-comitantbilateral totalkneearthroplasty inelderlypatients.MayoClin Proc 1997;72:799-805.

5. AdiliA,BhandariM,PetruccelliD,et al.Sequentialbilateral totalkneearthroplastyunder1anestheticinpatients>or=75yearsold:complicationsandfunctionaloutcomes.J Arthroplasty 2001;16:271-8.

6. BurnettR,MacDonaldS.Complications inbilateral totalkneearthroplasty.Semin Arthroplasty 2003;14:245-62.



RESEARCH

7. Worland RL, Jessup DE, Clelland C. Simultaneous bilateral totalknee replacement versus unilateral replacement. Am J Orthop 1996;25: 292-5.

8. Lombardi AV, Mallory T, Fada R, et al. Simultaneous bilateral total kneearthroplasties: Who decides? Clin Orthop Relat Res 2001;392:319-29.

9. Lane GJ, Hosack WJ, Shah S, et al. Simultaneous bilateral versus uni-lateral total knee arthroplasty: outcomes analysis. Clin Orthop RelatRes 1997;345:106-12.

10. Ritter MA, Hardy LD, Davis KE, et al. Simultaneous bilateral, stagedbilateral, and unilateral total knee arthroplasty. A survival analysis. J Bone Joint Surg Am 2003;85A:1532-7.

11. Soudry M, Binazzi R, Insall J, et al. Successive bilateral total kneereplacement. J Bone Joint Surg Am 1985;67:573-6.

12. Stanley D, Stockley I, Getty C. Simultaneous or staged bilateral totalknee replacement in rheumatoid arthritis: a prospective study. J BoneJoint Surg Br 1990;72:772-4.

13. Morrey BF, Adams RA, Ilstrup DM, et al. Complications and mortalityassociated with bilateral or unilateral total knee arthroplasty. J BoneJoint Surg Am 1987;69:484-8.

14. Brotherton SL, Roberson JR, de Andrade JR, et al. Staged versussimultaneous bilateral total knee replacement. J Arthroplasty 1986;1:221-8.

15. Cohen RG, Forrest CJ, Benjamin JB. Safety and efficacy of bilateraltotal knee arthroplasty. J Arthroplasty 1997;12:497-502.

16. Dennis DA. Trends in total knee arthroplasty. Orthopedics 2006;29:S13-6.

17. Jankiewicz JJ, Sculco TP, Ranawat CS, et al. One-stage versus 2-stagebilteral total knee arthroplasty. Clin Orthop Relat Res 1994;309:94-101.

18. Kolettis GT, Wixson RL, Peruzzi WT, et al. Safety of 1-stage bilat-eral total knee arthroplasty. Clin Orthop Relat Res 1994;309:102-9.

19. Vince KG. Bilateral total knee arthroplasty under one anesthesia: asafe protocol. Mayo Clin Proc 1997;72:883-5.

20. Gandhi R, Tso P, Davis A, et al. Outcomes of total joint arthroplastyin academic versus community hospitals. Can J Surg 2009;52:413-6.

21. Kim S, Meehan J, White R. Operative risk of staged bilateral kneearthroplasty is underestimated in retrospective studies. J Arthroplasty2011;26:1198-204.

22. Dimitris CN, Taylor BC, Mowbray JG, et al. Perioperative morbidityand mortality of 2-team simultaneous bilateral total knee arthro-plasty. Orthopedics 2011;34:e841-6.

23. Keating EM, Meding JB, Faria PM, et al. Predictors of transfusionrisk in elective knee surgery. Clin Orthop Relat Res 1998;357:50-9.

24. Barrett J, Baron J, Losina E, et al. Bilateral total knee replacement:staging and pulmonary embolism. J Bone Joint Surg Am 2006;88:2146-51.

25. Alemparte J, Johnson G, Worland R, et al. Results of simultaneousbilateral total knee replacement: a study of 1208 knees in 604 patients.J South Orthop Assoc 2002;11:153-6.

26. Bould M, Freeman B, Pullyblank A, et al. Blood loss in sequentialbilateral total knee arthroplasty. J Arthroplasty 1998;13:77-9.

27. Patil N, Wakankar H. Pros and cons of simultaneous bilateral totalknee arthroplasty including morbidity and mortality rates. Orthopedics2008; 31:780-9.

28. Gill GS, Mills D, Joshi AB. Mortality following primary total kneearthroplasty. J Bone Joint Surg Am 2003;85A:432-5.

29. Urban MK, Chisholm M, Wukovits B. Are postoperative complica-tions more common with single-stage bilateral (SBTKR) than withunilateral knee arthroplasty: guidelines for patients scheduled forSBTKR. HSS J 2006;2:78-82.

30. Memtsoudis SG Besculides MC, Reid S, et al. Trends in bilateraltotal knee arthroplasties: 153,259 discharges between 1990 and 2004.Clin Orthop Relat Res 2009;467:1568-76.

31. Kreder HJ, Grosso P, Williams JI, et al. Provider volume and otherpredictors of outcome after total knee arthroplasty: a populationstudy in Ontario. Can J Surg 2003;46:15-22.

32. Stefansdóttir A, Lidgren L, Robertsson O. Higher early mortalitywith simultaneous rather than staged bilateral TKAs: results from theSwedish knee arthroplasty register. Clin Orthop Relat Res 2008;466:3066-70.

33. Berman AT, Geissele AE, Bosacco SJ. Blood loss with total kneearthroplasty. Clin Orthop Relat Res 1988;234:137-8.

34. Robinson AJ, Holcroft JW, Olcott C, et al. Pulmonary and coagula-tion changes in tourniquet shock. J Surg Res 1975;19:65-70.

35. Bong MR, Patel V, Chang E, et al. Risks associated with blood trans-fusion after total knee arthroplasty. J Arthroplasty 2004;19:281-7.

36. Ho KM, Ismail H. Use of intravenous tranexamic acid to reduce allo-geneic blood transfusion in total hip and knee arthroplasty: a meta-analysis. Anaesth Intensive Care 2003;31:529-37.

37. Restrepo C, Parvizi J, Dietrich T, et al. Safety of simultaneous bilateraltotal knee arthroplasty. A meta-analysis. J Bone Joint Surg Am2007;89:1220-6.




Impact of a regional acute care surgery model onpatient access and outcomes

Background: The consolidation of acute care surgery (ACS) services at 3 of 6 hospi-tals in a Canadian health region sought to alleviate a relative shortage of surgeons ableto take emergency call. We examined how this affected patient access and outcomes.

Methods: Using the generalized linear model and statistical process control, we ana-lyzed ACS-related episodes that occurred between 39 months prior to and 17 monthsafter the model’s implementation (n = 14 713).

Results: Time to surgery increased after the consolidation. Wait times increasedprimarily for patients presenting at nonreferral hospitals who were likely to requiretransfer to a referral hospital. Although ACS teams enabled referral hospitals tohandle a much higher volume of patients without increasing within-hospital waittimes, overall system wait times were lengthened by the growing frequency ofpatient transfers. Wait times for inpatient admission were difficult to interpretbecause there was a trend toward admitting patients directly to the ACS service,bypassing the emergency department (ED). For patients who did go through theED, wait times for inpatient admission increased after the consolidation; however,this trend was cancelled out by the apparently zero waits of patients who bypassedthe ED. Regionalization showed no impact on length of stay, readmissions, mortal-ity or complications.

Conclusion: Consolidation enabled the region to ensure adequate surgical coveragewithout harming patients. The need to transfer patients who presented at nonreferralhospitals led to longer waits.

Contexte : Le regroupement des services chirurgicaux d’urgence (SCU) dans3 hôpitaux sur 6 d’une région sanitaire canadienne visait à contrer une relativepénurie de chirurgiens capables d’effectuer les interventions d’urgence. Nous enavons analysé l’impact sur l’accessibilité des services et sur les résultats chez lespatients.

Méthodes : À l’aide du modèle linéaire généralisé et d’un contrôle statistique desprocédés, nous avons analysé les cas adressés aux SCU entre 39 mois précédant et17 mois suivant l’entrée en vigueur du regroupement des services (n = 14 713).

Résultats : L’intervalle avant l’intervention chirurgicale s’est allongé après leregroupement des services. Les temps d’attente ont principalement augmenté pourles patients qui consultaient dans un hôpital de premier recours d’où ils étaient sus-ceptibles d’être réorientés vers un hôpital de référence. Même si les équipes des SCUont permis aux hôpitaux de référence de gérer un volume beaucoup plus importantde patients sans augmentation du temps d’attente à l’hôpital même, le temps d’at-tente dans son ensemble s’est prolongé à l’échelle du système en raison de l’ac-croissement du nombre de transferts. Les temps d’attente pour les hospitalisationsont été difficiles à interpréter parce qu’on avait tendance à admettre les patientsdirectement aux SCU, en contournant les services d’urgences. Pour les patients quipassaient par les urgences, les temps d’attente pour une hospitalisation ont augmentéaprès le regroupement; toutefois, cette tendance a été compensée par l’attente pourainsi dire nulle des patients qui contournaient les services d’urgence. La régionalisa-tion n’a exercé aucun impact sur la durée du séjour, les réadmissions, la mortalité oules complications.

Conclusion : Le regroupement a permis à la région d’assurer une couverture chirur-gicale adéquate sans nuire aux patients. La nécessité de réorienter des patients vers leshôpitaux de référence a contribué à prolonger les temps d’attente.

Sara A. Kreindler, DPhil*†

Liping Zhang, MD, BSc†

Colleen J. Metge, PhD*†

Richard W. Nason, MD, MSc†

Brock Wright, MD, MBA*†

Wendy Rudnick, MN*

Michael E.K. Moffatt, MD, MSc*†

From the *Winnipeg Regional HealthAuthority, and the †University of Mani-toba, Community Health Services, Win-nipeg, Man.

Accepted for publicationNov. 14, 2012

Correspondence to: S. KreindlerWinnipeg Regional Health Authority,Research & Evaluation Unit200-1155 Concordia Ave.Winnipeg MB R2K [email protected]

DOI: 10.1503/cjs.007012

impact-krein_Layout 1 13-09-17 8:25 AM Page 318


RESEARCH

A ccess to emergency or urgent surgery has becomea serious concern across North America andbeyond.1,2 With the evolution of surgical subspe-

cialties there has been a relative decline in the availabilityof surgeons who provide emergency care, particularly afterhours.3–5 Many health care organizations and systems havesought to cope with the problem by consolidating surgicalservices. There is growing international interest in theacute care surgery (ACS) model, in which designatedresources, including the time of designated surgeons, areset aside for emergency and/or urgent surgery. The modelmay be implemented within a hospital, among hospitals(regionalization), or both.

Past research has suggested that within-hospital ACSconsolidation can reduce wait times for emergency surgery,and may sometimes improve patient outcomes.6–9 However,the literature is dominated by uncontrolled pre–post studies, which cannot rule out secular trends and otherconfounders, such as adoption of new surgical practices, asexplanations for the observed results. Very few studies haveevaluated ACS consolidation on a regional level involvingmultiple institutions. One study found that mortality andlength of stay (LOS) decreased over time within a region-alized ACS service, but it did not compare these rates withthose observed before regionalization.10 Another studyfound no significant change in mortality when high-acuitysurgery (including many procedures outside the scope ofemergency general surgery) was regionalized; there was anongoing decline in LOS, but this decline had begun beforeregionalization.11 To our knowledge, no study has examinedwait times at the pan-regional level, taking into accounthow transfers affect the patient journey. As the ACS modelbecomes increasingly widespread, there is an urgent needfor further evidence on the impacts of ACS consolidation,especially regional consolidation.

The Winnipeg Regional Health Authority (WRHA), anurban health region in Western Canada (catchment popu-lation 1.2 million), implemented the ACS model to sta -bilize and sustain call schedules. The region had beenunable to fill the gaps through recruitment because therewas neither the volume of work nor the resources to sup-port new recruits. This 6-hospital system consolidatedemergency general surgery at 3 referral sites, each of whichnow in cludes a designated ACS team. A clinical lead or ser-vice chief (funded position) is responsible for addressingadministrative needs and maintaining call schedules. Onesurgeon covers the day shift (7:30–16:30) from Monday toSunday, providing continuity throughout the week. Thenight shift (16:30–7:30, home-call) rotates among partici-pating surgeons on a daily basis. Remuneration is ablended model of a day and night stipend and fee-for- service. A hospitalist was added to the ACS team at thenonteaching site; hospitalists and/or physician assistantsplay a smaller role at the teaching sites. The referral hospi-tals have added daytime ACS slates, but not dedicated

emergency operating rooms (ORs). To offset the increasedvolume of ACS patients at the referral hospitals, both elec-tive general surgery and orthopedic surgery cases havebeen relocated to the nonreferral hospitals.

One hospital became a referral centre in April 2008,another in December 2008. (A third offered ACS through-out the study period but did not invite additional referrals.)The consolidation succeeded in filling call schedules andensuring reliable access to an on-call surgeon. This studysought to determine the impact of consolidation on patientaccess and outcomes.

METHODS

Our analyses of administrative data included all adult pa -tients (aged 20 years and older) with an ACS-related in -patient stay at a WRHA hospital between 2005 and 2009.The University of Manitoba Health Research Ethics Boardapproved our use of this deidentified data for research pur-poses. We defined “ACS-related” as an emergent admissionto the general surgery service, where the most responsiblediagnosis was gastrointestinal (GI)-related (ICD-10 Kcodes [diseases of the GI system] or R10 codes [abdominalpain]). This criterion was used because the most commonacute surgical conditions are GI-related (e.g., appendicitis,cholecystitis), whereas the most common types of emer-gency general surgery outside the scope of the ACS con-solidation (trauma and cancer-related surgery) are not. Wedefined an “episode of care” as an acute care admissionwith an eligible diagnosis and emergency department (ED)visit that occurred within 24 hours (or, if a patient transferwas reported, within 72 h) of another admission or visit.Based on these criteria, we arrived at a sample of 14 735 epi -sodes of care. In the interest of statistical independence ofobservations, we then excluded episodes where the patienthad a recent prior episode. For tests of system responsive-ness (time to surgery, time to inpatient admission) andcomplications, we considered a recent episode to haveoccurred in the previous 72 hours; for other tests of patientmanagement, we considered a recent episode to haveoccurred in the previous 30 days.

The dependent variables included time to surgery (timefrom first presentation at a WRHA hospital to firstsurgery), time to inpatient admission (time from first pre-sentation at a WRHA hospital to first inpatient admission),LOS (total time from first inpatient admission to last in -patient discharge), readmission (all-cause readmission toany WRHA hospital within 30 d of being discharged alive),death (in hospital or within 30 d of discharge) and compli-cations (presence of 1 or more complications in surgicalpatients; the data source included generalized complica-tions, such as infection, hemorrhage and iatrogenic injury,but not disease-specific issues, such as ruptured appendix).

The intervention period began on Apr. 1, 2008, with thecreation of the first referral site. Given the stepped nature



RECHERCHE

of implementation, we considered dividing the interven-tion period into 2 phases; however, phase 2 and its interac-tion with linear time never reached significance in anymodel and were therefore removed. Covariates includedmonth of last hospital discharge, sex, age, diagnostic cat -egory (appendicitis, cholecystitis, intestinal obstruction,pancreatitis, diverticulitis, other), operative status (under-going an ACS-related procedure within 7 d of the start ofthe episode; the list of eligible procedures was determinedby a surgeon who was blind to all other data), non-WRHAfacility (transfer to or from out-of-region facility duringthe episode) and referral hospital (presenting at a hospitalthat was or became a referral centre). The inclusion andexclusion criteria and related sample sizes for each analysisare provided in Table 1, and characteristics of the sampleare summarized in Table 2.


Analytic methods included statistical process control andregression modelling. Statistical process control (SPC)involves plotting the data on a control chart to examinethe timing and magnitude of any changes.12,13 Results aretested for significance according to rules that include1 data point outside the upper and lower control limits,6 consecutive data points ascending or descending and9 consecutive data points above or below the mean. Onthe control charts, the solid line represents the mean andthe dotted lines represent the upper and lower controllimits. We calculated these values based on the preinter-vention period.

We used multiple linear regression for continuous (log-transformed) variables and logistic regression for binaryvariables. Before choosing this method, we used theDurbin–Watson test to check for autocorrelation of errors,using data at the levels of both individual cases andmonthly aggregates.14 These tests did not show significantresults (the Durbin–Watson statistic approached 2), indi-cating that it was unnecessary to use a procedure, such asARIMA, to control for autocorrelation.

RESULTS

Implementation of ACS

Between 2007 and 2009, ACS all but stopped at the 3 non-referral centres, while the proportion of ACS episodeshandled at the 2 new referral centres increased by 71.5%.The overall volume of general surgery patients rose per-manently at 1 referral centre and temporarily at the other(it should be noted that the consolidation also redistrib-uted non–general surgery patients). A 4.5-fold increasewas observed in patient transfers among WRHA hospitalsfor surgical consultation or treatment. Ambulance datasuggested that this increase was somewhat, but not fully

offset by a decrease in the number of there-and-backtransfers for tests, such as computed tomography (see theAppendix, Figs. S1 and S2, available at cma.ca/cjs).

Time to surgery

Our SPC analysis suggested some increase in wait timesafter the intervention, although this trend did not reachsignificance (Fig. 1). However, Figure 2 shows that whenthe analysis was repeated for patients within the WRHA,the finding clearly reached significance (more than 9 pointsabove the preintervention mean, with the rise beginningaround December 2008). This finding was explained by theobservation that the vast majority of non-WRHA patientspresented to 1 of the 3 ACS referral hospitals directly,whereas about one-third of WRHA patients presented to anon-ACS site (both before and after the intervention). Fur-ther analysis confirmed that, whereas wait times stayedfairly constant at referral hospitals, they rose sharply atfeeder hospitals after the intervention began (see theAppendix, Figs. S3 and S4). This rise appeared to reflect

Table 1. Inclusion and exclusion criteria

Description; inclusion criteria No.

Total potential sample

All ACS episodes of care 14 735

Descriptive analyses

No ACS hospital stays within the past 72 h 14 713

Time to surgery, log


ACS episode is not incidental 14 659

Surgical patient 9 056

Date and time of surgery available 7 122

Surgery occurred during the recorded episode 7 052

Surgery occurred fewer than 7 d after start of episode 6 722

No missing “transfer from” visit 6 605

Time to inpatient, log



No missing “transfer from” visit 14 453

Length of stay, log

No ACS hospital stays within the past 30 d 14 182


No missing “transfer to” visit 14 107

30-d readmission


Discharge prior to the last month of data collection 13 911

Discharged alive 13 654

30-d mortality (includes inhospital)


Discharge prior to the last month of data collection 13 911

Complications


Surgical patient 9 093

Surgery occurred during the recorded episode 7 082

Surgery occurred fewer than 7 d after start of episode 6 748

ACS = acute-care surgery.



RESEARCH

the increasing proportion of patients who were transferred(from 18.8% to 70.6% of surgical patients presenting atnonreferral hospitals), not an increase in the length of timeassociated with a transfer (which did not change over thestudy period). Supplementary analyses indicated that trans-ferred patients waited a median of 5.5 hours longer thannontransferred patients; depending on whether the patienttravelled by ambulance, 2–3 of these hours could beaccounted for by the transfer process. The remaining timeappeared to reflect pre- and posttransfer delays, such aswaiting for access to a bed or OR suite. The increased waitdid not seem to occur as a result of arriving at a particularhospital (patients arriving at all nonreferral hospitals hadsimilar wait times), nor as a result of delays in being admit-ted to a referral hospital (transfer patients were admittedquickly and were fast-tracked to surgery). The region-wideimpact was a 1- to 2-hour increase in median wait timesevery half-year after the consolidation.

Multiple linear regression confirmed these results (seethe Appendix, Table S1). Both for the sample as a whole andmore strongly for WRHA patients, there was a significantinteraction between the intervention and the episode date.

This indicated that time to surgery began to get longer afterthe intervention was implemented. Significant interventioneffects were apparent at 12 and 18 months. There was also asignificant interaction between the intervention and type ofhospital: after the intervention, waits became longer atfeeder hospitals and slightly and temporarily shorter at refer-ral hospitals. When the analysis was restricted to patientswho presented at referral hospitals (99% of whom remainedthere), no intervention effect appeared. To ensure that miss-ing data for “time to surgery” had not biased our results, weperformed a Poisson regression using days to surgery as thedependent variable. The effects for intervention and monthas well as the interaction between them remained in thesame direction and reached significance in the WRHA sub-sample (p = 0.041). The significant result is striking, consid-ering the imprecision of the dependent variable, days tosurgery, whose value was 0 or 1 for 72% of episodes.

Time to inpatient admission

Wait times for inpatient admission were difficult to inter-pret because the study period coincided with a trend toward

Table 2. Characteristics of the study sample

Group; no. (%)

Variable

Pre–model implementationJanuary 2005–March 2008

n = 8 994

Post–model implementation April 2008–December 2009

n = 5 719 Total, n = 14 713

Sex

Male 4 175 (46.4) 2 642 (46.2) 6 817 (46.3)

Female 4 819 (53.6) 3 077 (53.8) 7 896 (53.7)

Age, yr

20–34 1 876 (20.9) 1 129 (19.7) 3 005 (20.4)

35–49 2 140 (23.8) 1 303 (22.8) 3 443 (23.4)

50–64 2 109 (23.4) 1 338 (23.4) 3 447 (23.4)

65–79 1 841 (20.5) 1 195 (20.9) 3 036 (20.6)

≥ 80 1 028 (11.4) 754 (13.2) 1 782 (12.1)

Operative status*

Nonsurgical 3 478 (38.7) 2 142 (37.5) 5 620 (38.2)

Surgical 5 516 (61.3) 3 577 (62.5) 9 093 (61.8)

Diagnostic category

Appendicitis 1 433 (15.9) 946 (16.5) 2 379 (16.2)

Cholecystitis 1 561 (17.4) 1 072 (18.7) 2 633 (17.9)

Intestinal obstruction 1 075 (12.0) 703 (12.3) 1 778 (12.1)

Pancreatitis 586 (6.5) 331 (5.8) 917 (6.2)

Diverticulitis 555 (6.2) 375 (6.6) 930 (6.3)

Other 3 784 (42.1) 2 292 (40.1) 6 076 (41.3)

Origin

WRHA 7 915 (88.0) 4 913 (85.9) 12 828 (87.2)

Non-WRHA 1 079 (12.0) 806 (14.1) 1 885 (12.8)

WRHA = Winnipeg Regional Health Authority.*In this table, patients who had surgery at any time are counted as operative.



RECHERCHE

admitting selected patients directly to the inpatient surgeryward, bypassing the ED. There was no apparent wait timefor such patients, but this would seem to be an underesti-mate, since we do not know when they really presented inthe system. To avoid underestimating post intervention waittimes, we ran the analyses twice: once for patients who wentthrough the ED and again for the full sample.

For patients admitted through the ED, SPC analysisshowed an unmistakable increase in time to inpatientadmission (Fig. 3). The most dramatic increase coincidedwith the jump in direct admissions, but some increase hadalready become apparent by early 2008, before the inter-vention period. However, the full-sample analysis sug-gested a possible decrease in wait times during the inter-vention period (Fig. 4).

The regression models echoed the findings that afterthe intervention, wait times increased for admissionsthrough the ED but decreased for admissions as a whole,although these effects varied in significance betweenWRHA patients and the full sample (see the Appendix,Table S2). As with time to surgery, the increased waits foradmissions through the ED were more apparent forpatients who presented to a nonreferral hospital (data not

shown). Decreased waits for overall admissions were moreapparent at referral hospitals, because only these hospitalsadopted a policy of bypassing the ED.

These results imply that ACS patients who visited theED postconsolidation spent longer there than they wouldhave preconsolidation. However, an increasing number ofACS patients spent no time in the ED at all. The 2 trends

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Tim

e to

sur

gery

, log

Month of discharge

Time to surgery: all patients

Fig. 1.Time to surgery (log) for all patients.

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Tim

e to

sur

gery

, log

Month of discharge

Time to surgery: all WRHA patients

Fig. 2. Time to surgery (log) for Winnipeg Regional HealthAuthority (WHRA) patients.

-1.4

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Tim

e to

inpa

tient

, log

Month of discharge

Time to inpatient admission: patients admitted from ED

Fig. 3. Time to inpatient admission (log) for patients admittedfrom the emergency department (ED).

-2.5

-2

-1.5

-1

-0.5

0

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Tim

e to

inpa

tient

, log

Month of discharge

Time to inpatient admission: all patients

Fig. 4.Time to inpatient admission (log) for all patients.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

LOS

, d, l

og

Month of discharge

Length of stay: all patients

Fig. 5. Length of stay (log) for all patients.



RESEARCH

may balance each other out, or it could be argued that1 trend is more important than the other. It is difficult todraw overall conclusions without making assumptionsabout the journey of patients who do not visit the ED.

Length of stay

On control charts, LOS seemed slightly shorter after theintervention, but SPC analysis did not detect a significantchange (Fig. 5). However, when the sample was split byoperative status, this effect appeared only for nonsurgicalpatients, with no intervention effect for surgical patients(see the Appendix Figs. S5 and S6). In addition, the con-trol chart for nonsurgical patients showed that anydecrease in LOS began in late 2007, which was well beforethe intervention. Multiple regression revealed that LOSbegan to rise after the intervention, but again only fornonsurgical patients (see the Appendix, Table S3). Thesefindings suggest that any variation in LOS did not imply achange in surgical outcomes.

Readmission, mortality and complications

Neither the SPC nor logistic regression detected an inter-vention effect on 30-day readmission rates, 30-day mortal-ity or surgical complications (see the Appendix, Table S4).These outcomes occurred at virtually identical ratesbefore and after the intervention (13.5% v. 13.5% for re -admissions, 2.4% v. 2.5% for mortality and 6.0% v. 6.2%for complications). We also found that, compared withpatients who presented at an ACS referral hospital, pa -tients who transferred there did not have longer LOS oran elevated risk for complications, readmissions or death.A supplementary analysis revealed that the proportion ofsurgeries performed after hours did not change with theintervention.

DISCUSSION

To our knowledge, this study is the first to examine waittimes for ACS in a multi-hospital system, linking togetherepisodes of care that were separated by transfers to meas -ure the entire patient journey. The main limitations of thiswork, the lack of randomization and the inability to con-trol for illness severity, are offset by the long interruptedtime series design. This represents an advance over priorstudies, which have featured simple pre–post or year-to-year comparisons. Through SPC, we were able to pin-point the timing of observed effects, reducing the risk ofbias from secular trends and unrelated developments.

The ACS intervention included both the region-wideconsolidation of ACS and hospital-specific efforts toimprove patient access through a designated ACS team. Atthe hospital level, strategies to improve access for ACSpatients do appear to have helped. The new referral hospi-

tals are handling a much higher volume of ACS patientsthan before, without increasing the amount of timebetween presenting at that hospital and receiving surgery.This achievement may be partly attributed to a reducedload of non-ACS patients, but may also reflect improvedprocesses. There has also been a move toward streamlininginpatient admissions by allowing referred patients tobypass the ED. However, for the entire patient populationserved by the 6 hospitals, this study showed a longer aver-age wait for access to emergent surgery following the ACSconsolidation. The efficiency gains that may have occurredat the hospital level do not outweigh the extra time spenttransferring patients who present at a nonreferral hospital.

The ACS intervention does not appear to have affectedpatient outcomes, with no changes in LOS, complications,readmissions or mortality. Thus, there is no evidence thatthe observed increase in wait times has harmed ACSpatients. Since we relied on broad definitions of both thepatient population and the outcomes of interest, it remainspossible that we failed to detect certain positive and/ornegative effects. In particular, since we were unable to con-trol for severity of illness, we cannot determine whetherthe consolidation might have beneficially or adverselyimpacted the subgroup of complex and unstable patients(we would note that the policy of direct admission to a sur-gical ward does not apply to unstable patients, who aretriaged to the ED; however, such patients may have beenaffected by other aspects of the consolidation). In addition,it is difficult to use LOS as a proxy for outcomes when theintervention involves shifting patients among hospitals thatmay differ in average LOS. Moreover, although our dataincluded all of the most common ACS conditions, we mayhave missed some relevant diagnoses, and we found it toodifficult to track the journey of patients whose ACSepisodes were incidental to another hospital admission. Infuture, the creation of an ACS registry would permit thedefinitive identification of ACS patients. However, the bestavailable data suggest that the consolidation did not affectbroad outcomes across the ACS population.

This study did not track the potential impacts of ACSconsolidation on non-ACS patients. Benchmarking datacollected annually from 2007/08 through 2010/11 detectedno significant changes in adverse outcomes (readmission,in-hospital mortality) for acute myocardial infarction orstroke patients at the referral hospitals, nor was there anincrease in the number of hospital admissions for ambula-tory care–sensitive conditions. These data provide no a priori evidence of patient harm, but firm conclusions can-not be drawn without more sensitive and frequent mea-sures of emergency patient outcomes.

CONCLUSION

Our findings suggest that a regional ACS model canensure adequate emergency surgical coverage without


threatening patient outcomes. This makes it a viable solu-tion to a health human resources problem that has be -come increasingly serious in Canada, the United Statesand elsewhere. Patient outcomes were not impacted;however, the growing number of patient transfersbrought an overall increase in wait times. The WRHAhas recognized a need to improve the current system sothat all patients experience a smooth journey withoutdelays; it is currently working to streamline the transferprocess while also reducing surgical volume at referralhospitals by redirecting other types of surgery. The sys-tem is still evolving, and benefits may be realized as itdevelops. With time, ACS patients may gravitate and bedirected toward the referral sites, with a lower proportionpresenting at non-ACS facilities. What is clear, however,is that regionalization brings its own challenges, and evi-dence on the outcomes of within-hospital ACS consolida-tion cannot necessarily be generalized to the regionallevel. Returning to a model with unpredictable gaps inthe emergency call schedule would not be an option forthe WRHA, nor for other regions in similar situations;however, our findings do not provide grounds for whole-sale adoption of the model regardless of context. It re -mains essential for jurisdictions that implement a regionalACS model to carefully monitor its impacts — bothintended and unintended.

Acknowledgements: This research was undertaken as part of an evalua-tion of the Winnipeg Regional Health Authority (WRHA) Acute-CareSurgical Service. We thank Trevor Strome, Miroslava Svitlica, Jill Evison,Trudy Wilgosh, Ray Larkins, Cameron Robertson, Susan Gerlach,Wendy Rogocki, Evelyn Fondse, Anne Haknasson, Crystal Letain,Victor Cho and Elaine Pelletier for their help with data acquisition. Wealso appreciate the suggestions we received from the MNU/WRHAWorking Group and from Dr. David Hochman.


Contributors: S.A. Kreindler, C.J. Metge, R.W. Nason and M.E.K. Moffattdesigned the study. L. Zhang acquired the data, which all authors inter-preted and analyzed. S.A. Kreindler wrote the article, which all authorsreviewed and approved for publication.

References

1. Division of Advocacy and Health Policy. A growing crisis in patientaccess to emergency surgical care. Bull Am Coll Surg 2006;91:8-19.

2. Bhagvan S, Civil I. Acute care surgery: Can New Zealand afford towait? N Z Med J 2009;122:71-6.

3. Sanchez M, Sariego J. The general surgeon shortage: causes, conse-quences, and solutions. South Med J 2009;102:291-4.

4. Sheldon GF. Workforce issues in general surgery. Am Surg 2007;73:100-8.

5. Gough IR. Sub-specialisation in surgery and the continuing chal-lenge of providing emergency surgery services. Med J Aust 2008;189:358-9.

6. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery modelimproves outcomes in patients with appendicitis. Ann Surg 2006; 244:498-504.

7. Maa J, Carter JT, Gosnell JE, et al. The surgical hospitalist: a newmodel for emergency surgical care. J Am Coll Surg 2007;205:704-11.

8. Sorelli PG, El-Masry NS, Dawson PM, et al. The dedicated emer-gency surgeon: towards consultant-based acute surgical admissions.Ann R Coll Surg Engl 2008;90:104-8.

9. Gandy RC, Truskett PG, Wong SW, et al. Outcomes of appen-dicectomy in an acute care surgery model. Med J Aust 2010 ;193:281-4.

10. Diaz JJ, Norris PR, Gunter OL, et al. Does regionalization of acutecare surgery decrease mortality? J Trauma 2011;71:442-6.

11. Hamilton SM, Letourneau S, Pekeles E, et al. The impact of region-alization on a surgery program in the Canadian health care system.Arch Surg 1997;132:605-9.

12. Hart MK, Hart RF. Statistical Process Control for Health Care. PacificGrove (CA): Wadsworth Group; 2002.

13. Diaz, M, Neuhauser MD. Pasteur and parachutes: when statisticalprocess control is better than a randomized controlled trial. Qual SafHealth Care 2005;14:140-3.

14. Durbin J, Watson GS. Testing for serial correlation in least squaresregression, II. Biometrika 1951;38:159-78.

RECHERCHE


Change of addressWe require 6 to 8 weeks’ notice to ensure uninterruptedservice. Please send your current mailing label, newaddress and the effective date of change to:

CMA Member Service Centre

1870 Alta Vista Dr.Ottawa ON K1G 6R7

tel 888 855-2555 or 613 731-8610 x2307 fax 613 [email protected]




Incremental value and clinical impact of necksonography for primary hyperparathyroidism: a risk-adjusted analysis

Background: Despite the different preoperative imaging modalities available forparathyroid adenoma localization, there is currently no uniform consensus on themost appropriate preoperative imaging algorithm that should be routinely followedprior to the surgical management of primary hyperparathyroidism (PHPT). Wesought to determine the incremental value of adding neck ultrasonography to scintigraphy-based imaging tests.

Methods: In a single institution, surgically naive patients with PHPT underwent thefollowing localization studies before parathyroidectomy: 1) Tc-99m sestamibi imagingwith single photon emission computed tomography/computed tomography(SPECT/CT) or Tc-99m sestamibi imaging with SPECT alone, or 2) ultrasonog raphyin addition to those tests. We retrospectively collected data and performed a multivari-ate analysis comparing group I (single study) to group II (addition of ultrasonography)and risk of bilateral (BNE) compared with unilateral (UNE) neck exploration.

Results: Our study included 208 patients. Group II had 0.45 times the odds of BNEversus UNE compared with group I (unadjusted odds ratio [OR] 0.45, 95% confi-dence interval [CI] 0.25–0.81, p = 0.008). When adjusting for patient age, sex, preop-erative calcium level, use of intraoperative PTH monitoring, preoperative PTHlevel, adenoma size, and number of abnormal parathyroid glands, Group II had 0.48times the odds of BNE versus UNE compared with group I (adjusted OR 0.48, 95%CI 0.23–1.03, p = 0.06). In a subgroup analysis, only the addition of ultrasonographyto SPECT decreased the risk of undergoing BNE compared with SPECT alone(unadjusted OR 0.40, 95% CI 0.19–0.84, p = 0.015; adjusted OR 0.38, 95% CI 0.15–0.96, p = 0.043).

Conclusion: The addition of ultrasonography to SPECT, but not to SPECT/CT, hasincremental value in decreasing the extent of surgery during parathyroidectomy, evenafter adjusting for multiple confounding factors.

Contexte : Malgré l’existence de diverses modalités d’imagerie préopératoire pour lalocalisation de l’adénome parathyroïdien, on déplore actuellement l’absence de con-sensus en ce qui concerne l’algorithme le plus approprié à suivre au chapitre de l’ima -gerie préalable à une prise en charge chirurgicale de l’hyperparathyroïdie primaire(HPTP). Nous avons voulu vérifier si l’ajout de l’échographie du cou aux tests d’ima -gerie scintigraphique offrait une valeur ajoutée.

Méthodes : Dans un établissement, des patients atteints d’HPTP n’ayant jamais subid’intervention chirurgicale ont été soumis à des examens de localisation préparathy-roïdectomie : 1) imagerie au moyen du sestamibi marqué au Tc-99m avec tomogra-phie par émission monophotonique/tomodensitométrie (SPECT/CT), ou imagerie aumoyen du sestamibi marqué au Tc-99m avec SPECT seule, our 2) échographie enplus de ces tests. Nous avons recueilli les données rétrospectivement et effectué uneanalyse multivariée pour comparer le Groupe I (examen seul) au Groupe II (ajout del’échographie) et la probabilité qu’ils subissent une exploration cervicale bilatérale(ECB) plutôt qu’unilatérale (ECU).

Résultats : Notre étude a recruté 208 patients. Le Groupe II s'est trouvé exposé àun risque 0,45 fois plus grand d’être soumis à une ECB plutôt qu’à une ECU, com-parativement au Groupe I (rapport des cotes [RC] non ajusté 0,45, intervalle deconfiance [IC] de 95 % 0,25–0,81, p = 0,008). Après ajustement pour tenir comptede l’âge et du sexe des patients, de leur taux préopératoire de calcium, de la surveil-lance peropératoire de l’HPT, du taux préopératoire de l’HPT, de la taille del’adénome et du nombre de ganglions parathyroïdiens anormaux, le Groupe II s’estrévélé exposé à un risque 0,48 fois plus grand à l’égard de l’ECB plutôt que de

May C. Tee, MD, MPH*

Simon K. Chan, MSc†

Vy Nguyen, MD*

Scott S. Strugnell, MSc*

Jonathan Yang, BSc*

Steven Jones, MD†

Pari Tiwari, MD‡

Daniel S. Levine, MB, ChB§

Sam M. Wiseman, MD*

From *St. Paul’s Hospital, Department ofSurgery and University of British Colum-bia, †Department of Medical Genetics,University of British Columbia andMichael Smith Genome Sciences Center,British Columbia Cancer Research Cen-ter, ‡St. Paul’s Hospital, Department ofRadiology, and University of BritishColumbia, and §St. Paul’s Hospital, Divi-sion of Nuclear Medicine, Department ofRadiology, and University of BritishColumbia, Vancouver, BC


This study was presented as a poster atthe Pacific Coast Surgical Association’s83rd Annual Meeting, Napa Valley, Calif.,February 2012.

Correspondence to:S.M. WisemanRoom C303, Burrard BuildingDepartment of Surgery, St. Paul’s Hospital,University of British Columbia1081 Burrard StreetVancouver BC V6Z [email protected]

DOI: 10.1503/cjs.015612

inc-tee_Layout 1 13-09-16 11:21 AM Page 325


RECHERCHE

P rimary hyperparathyroidism (PHPT) is a commonendocrine disorder characterized by an elevatedparathyroid hormone (PTH) level and hypercal-

cemia.1–3 Currently, the most common presentation ofPHPT is an asymptomatic individual who is incidentallyidentified through routine biochemical laboratory test-ing.3,4 When PHPT is symptomatic, its most prevalentclinical presentations is nephrolithiasis, followed by muscu-loskeletal complaints, neuropsychiatric disorders andabdominal symptoms.1–3 With an incidence of about 1% inthe adult population, increasing to 2% in adults aged55 years and older, PHPT tends to be a disease of middle-aged women.1,5

The most common cause of PHPT is a solitary parathy-roid adenoma, which accounts for up to 90% of cases.3,6

Approximately 2% of these adenomas may be found inectopic locations, including within the mediastinum,carotid sheath or thyroid gland.3,5 Other causes of PHPTinclude multiple parathyroid adenomas (5%), 4 glandhyperplasia (5%) and parathyroid carcinoma (< 1%).1,3,5

Parathyroidectomy is the treatment of choice for PHPTand offers an enduring cure.1,3,5,6

Historically, the gold standard approach to parathy-roidectomy has been a bilateral neck exploration.3 How-ever, the advent of increasingly available and accurate preoperative imaging, along with intraoperative PTHmeasurement, has allowed surgeons to carry out a morefocused surgical approach, especially since the vast major-ity of PHPT cases are caused by a single adenoma.2,3,7

Combined with other tools, such as the intraoperativeparathyroid hormone assay, there has been an increasingnumber of surgeons advocating for focused parathyroidec-tomy, such as unilateral neck exploration and minimallyinvasive parathyroidectomy.8–14 Moreover, focused parathy-roidectomy is associated with decreased surgical dissectionand risk of bilateral recurrent laryngeal nerve injury, min -imized postoperative pain and shortened length of stay inhospital.10,15

Accurate preoperative imaging has facilitated the use offocused parathyroidectomy.10 Several imaging modalitiesare available for preoperative localization, including Tc-99m sestamibi imaging (with planar and single photonemission computed tomography; SPECT), neck ultra-sonography, and combined Tc-99m sestamibi single pho-

ton emission computed tomography with computedtomography (SPECT/CT).13 Other imaging modalitiesbeing studied for parathyroid preoperative localizationinclude magnetic resonance imaging (MRI), positronemission tomography combined with CT (PET/CT), and4-dimensional angiography-enhanced CT.16,17

However, despite the different preoperative imagingtests available for parathyroid adenoma localization, thereis currently no uniform consensus on the most appropriatepreoperative imaging algorithm that should be routinelyfollowed before the surgical management of PHPT. More-over, the literature regarding preoperative imaging ofPHPT has been primarily focused on the sensitivity andspecificity of each imaging test.1,18 Several studies haveexamined performance of multiple preoperative imagingtests compared with a single imaging test and have focusedtheir analysis on sensitivity and specificity rather thanactual clinical outcomes.14,19–21 Therefore, the objective ofthis study was to evaluate whether there is incrementalvalue in adding ultrasonography to a functional imagingtest, such as SPECT or SPECT/CT, in reducing the riskof carrying out a bilateral (BNE) versus unilateral (UNE)neck exploration at the time of parathyroidectomy.

METHODS

Study hypothesis and objective

We hypothesized that there was an association betweenpreoperative imaging test for localization of parathyroidadenoma in PHPT and extent of surgery. The interven-tion of interest was the addition of ultrasonography to scintigraphy-based imaging (group II) compared with onlyscintigraphy-based imaging (group I). The outcome ofinterest was the relative risk of BNE versus UNE. Toaccount for spurious associations brought on by confound-ing factors, we used multivariate logistic regression analysisto evaluate this intervention–outcome relationship.

Participants

We retrospectively reviewed all operations carried out fortreatment of PHPT between January 2002 and August2011 at a single Canadian tertiary care centre. St. Paul’s

l’ECU comparativement au Groupe I (RC ajusté 0,48, IC de 95 % 0,23–1,03,p = 0,06). Selon une analyse de sous-groupe, seul l’ajout de l’échographie à laSPECT a réduit le risque de subir une ECB comparativement à la SPECT seule(RC non ajusté 0,40, IC de 95 % 0,19–0,84, p = 0,015; RC ajusté 0,38, IC de 95 %0,15–0,96, p = 0,043).

Conclusion : L’ajout de l’échographie à la SPECT, mais non à la SPECT/CT, aoffert une valeur ajoutée pour ce qui est de réduire l’étendue de l’opération durant laparathyroïdectomie, même après ajustement pour tenir compte de plusieurs facteursde confusion.



RESEARCH

Hospital (Vancouver, BC) is a referral centre for head andneck endocrine surgery, and subspecialty-trained head andneck surgeons performed all operations. There is no for-mal guideline for preoperative imaging for PHPT at ourcentre, and there has generally been an even distributionof ultrasonography, SPECT and SPECT/CT for preoper-ative localization. Frequently, patients undergo eitherSPECT or SPECT/CT, and some of these patients alsoundergo ultrasonography of the neck. Our operativeapproach to PHPT is to begin with a UNE for cases thathave localized preoperatively and to proceed with a BNEif a UNE does not identify the adenoma.

For cases that do not localize with preoperative im -aging, a BNE is carried out if an adenoma is not on theside of the neck initially explored. Inclusion in the studycohort required the patient to have met diagnostic criteriafor PHPT, undergone preoperative localization imagingand undergone neck exploration with pathological confir-mation of parathyroid adenoma removal. Individuals whounderwent a prior surgical parathyroid exploration and hadpersistent or recurrent PHPT were excluded. There were268 cases of PHPT that were reviewed. Nine patients whodid not undergo preoperative imaging were excluded. Twopatients who did not have an adenoma identified at surgerywere also excluded, as were 11 patients who had prior neckexplorations. Thirty-eight patients were excluded as theyunderwent imaging tests in addition to SPECT, SPECT/CT, and/or ultrasonography.

We reviewed patient demographic characteristics, pre-operative laboratory studies, imaging findings, operativefindings (including intraoperative PTH measurements),extent of operative exploration (UNE v. BNE), adenomasize, postoperative laboratory studies and pathologic diag-noses. We obtained the data by retrospectively reviewingpatient charts; imaging, biochemical and pathology results;and operative records. Operative identification of a para -thyroid adenoma with pathologic confirmation determinedwhether a preoperative imaging test resulted in correctpreoperative lateralization (left v. right). Cure of PHPTwas confirmed by normalization of serum calcium levelswithin 30 days postoperatively.

Patient information was collected and stored in a de -identified database for data analysis. The study was ap -proved by our institutional review board

Imaging protocols

The protocols for SPECT (comprising planar andSPECT imaging) and SPECT/CT (comprising planarand SPECT/CT imaging) at our centre have been previ-ously described.18 Imaging of the neck and upper thorax isobtained in the supine position with a low-energy, high-resolution collimator, 10 minutes after intravenous injec-tion of 700 MBq Tc-99m sestamibi. Planar images areobtained over 5 minutes, both immediately and 2 hours

postinjection, using a 128-square matrix. Noncircular orbitSPECT is obtained immediately and at 2 hours postinjec-tion using a 128-square matrix with 128 stops, 15 secondsper stop for the early acquisition and 20 seconds per stopfor the delayed acquisition. For patients undergoingSPECT/ CT, a volumetric CT acquisition is obtained afterthe delayed SPECT through the same anatomic region(80 kVp, 60 mAs, using CAREDOSE dose modulation).Sets of 1 mm thick contiguous slices are acquired andreconstructed as contiguous 1.25 mm thick slices using amedium-smooth reconstruction kernel. The SPECT/CTimages are then generated using iterative reconstructionwith 8 subsets and 12 iterations. All imaging was carriedout with a Siemens Symbia T6 hybrid SPECT/CT scan-ner. Finally, the SPECT/CT fused multiplanar images arereviewed on a Siemens Leonardo workstation (SiemensMedical Systems) by a nuclear medicine/radiology dualcertified physician.


As the outcome of interest was the extent of surgicalexploration based on UNE versus BNE, we chose our def-inition of correct localization based on correct lateraliza-tion. We evaluated the proportion of correct lateralizationin group I (SPECT v. SPECT/CT) compared with groupII (1 of these tests with the addition of ultrasonography)using a Fisher exact test. We used multivariate logisticregression to determine the odds of BNE versus UNE forgroup I (1 imaging test) compared with group II (additionof ultrasonography to the imaging test). We controlled forthe following covariates: age, sex, preoperative calcium,preoperative PTH, use of intraoperative PTH assay, aden -oma size (greatest dimension) and number of abnormalparathyroid glands (solitary adenomas v. abnormal para -thyroid glands). All data analyses were performed usingStata software version 11.2 (Stata Corp.).

RESULTS

Patients

We reviewed 268 cases of PHPT. Nine patients who didnot undergo preoperative imaging were excluded. Twopatients who did not have an adenoma identified atsurgery were also excluded, as were 11 patients who hadprior neck explorations. We excluded a further 38 patientsbecause they underwent imaging tests in addition toSPECT, SPECT/CT and/or ultrasonography. Our finalcohort consisted of 208 patients who had undergoneSPECT, SPECT/CT or the addition of ultrasonographyto these tests preoperatively.

The clinical and pathological characteristics of patientsare summarized in Table 1. There were 156 (75.0%)women and 52 (25.0%) men. The mean age was 58.6 ±



RECHERCHE

13.6 years. There were 8 (3.8%) cases of multiglandulardisease (5 double adenomas and 3 sporadic 4-gland hyper-plasias) and 2 (1.0%) cases of para thyroid carcinoma.Ectopic adenomas accounted for 7 (3.4%) cases, and allwere solitary. Excluding the para thyroid carcinomas andectopic adenomas, there were 191 (91.8%) cases with apathological diagnosis of solitary parathyroid adenoma.The mean patient follow-up, as defined by the date of thelast serum calcium measurement, was 140 days.

Baseline preoperative laboratory data were collected andall patients underwent biochemical testing post operatively.Ionized calcium levels were corrected for the patients’blood pH. The preoperative calcium levels were 2.75 ±0.19 mmol/L for total calcium level (reference range 2.18–2.58 mmol/L) and 1.52 ± 0.13 mmol/L for ionized calciumlevel (reference range 1.17–1.29 mmol/L). Postoperativefollow-up calcium levels were 2.33 ± 0.17 mmol/L for totalcalcium level and 1.26 ± 0.09 mmol/L for ionized calciumlevel.

The majority of study patients (67.4%) underwentparathyroidectomy with intraoperative parathyroid hor-mone measurement during which a 50% drop at 5 or10 min utes after adenoma removal, from baseline or pre-excision levels, was considered predictive of cure. BaselinePTH measurements were 19.8 ± 25.9 pmol/L (referencerange 1.3–6.8 pmol/L). Intraoperative PTH measurementswere 22.9 ± 20.5 pmol/L, 19.0 ± 20.9 pmol/L, 8.5 ±8.9 pmol/L and 6.3 ± 6.4 pmol/L for baseline pre-excision

at 0 minutes, 5 minutes and 10 minutes after adenomaresection, respectively.

Correct localization by imaging tests

We evaluated the incremental value of ultrasonographywith respect to proportion of correct localization inpatients who underwent only 1 preoperative imaging test(group I) versus patients who had the addition of ultra-sonography to a preoperative imaging test (group II).There were 75 patients in group I and 133 patients ingroup II. Group II had 78.9% correctly localized para -thyroid adenomas compared with only 54.7% in group I(p < 0.001). In a subgroup analysis, the addition of ultra-sonography to SPECT improved correct localizationfrom 45.9% to 75.8% (p < 0.001). The addition ofultrasonog raphy to SPECT/CT, however, did not signifi-cantly improve correct preoperative localization (p = 0.30;Table 2).

Impact of preoperative imaging test on extent of surgery

We assessed the clinical impact of adding ultrasonographyto a preoperative imaging test using logistic regressionmodels evaluating the odds of BNE versus UNE duringparathyroidectomy. These results are summarized inTable 3. Univariate analyses are illustrated as unadjusted

Table 1. Descriptive statistics for unilateral (UNE) versus bilateral (BNE) neck exploration

Group; mean ± SD or no. (%)*

Factor Overall, n = 208 UNE, n = 138 BNE, n = 70 p value†

Age, yr 58.6 ± 13.5 59.3 ± 14.2 57.1 ± 12.2 0.26

Sex, female 156 (75.0) 103 (74.6) 53 (74.7) 0.87

Preoperative Ca, mmol/L 2.75 ± 0.19 2.76 ± 0.18 2.73 ± 0.19 0.19

Intraoperative PTH assay 139 (66.8) 101 (73.2) 38 (54.3) 0.006

PTH, nmol/L 19.8 ± 25.9 22.6 ± 28.9 14.4 ± 17.5 0.032

Greatest dimension, cm 1.76 ± 1.02 1.91 ± 1.17 1.50 ± 0.61 0.011

Adenoma type 0.002

Solitary 191 (91.8) 133 (96.3) 58 (82.9) —

Double 5 (2.4) 0 (0) 5 (7.1) —

Ectopic 7 (3.4) 4 (3.0) 3 (4.3) —

Hyperplasia 3 (1.4) 0 (0) 3 (4.3) —

Carcinoma 2 (1.0) 1 (0.7) 1 (1.4) —

Imaging studies 0.007

Group I 75 (36.0) 41 (29.7) 34 (48.6) —

SPECT only 61 (29.3) 28 (20.3) 33 (47.2) —

SPECT/CT 14 (6.7) 13 (9.4) 1 (1.4) —

Group II 133 (64.0) 97 (70.3) 36 (51.4) —

SPECT+ ultrasonography 62 (29.8) 42 (30.4) 20 (28.6) —

SPECT/CT+ultrasonography

71 (34.1) 55 (39.9) 16 (22.8) —

BNE = bilateral neck exploration; Ca = calcium; CT = computed tomography; PTH = parathyroid hormone; SD = standard deviation;SPECT = single photon emission computed tomography; UNE = unilateral neck exploration.*Unless otherwise indicated.†The p value for differences between UNE and BNE was calculated using a 2-sided t test (continuous data) and χ2 test (categorical data).



RESEARCH

odds ratios (ORs) with associated 95% confidence inter-vals (CIs) and p values for association. Parallel representa-tion is demonstrated for multivariate analyses, which havebeen adjusted for the following confounders: age, sex, pre-operative serum calcium, intraoperative PTH assay, pre-operative PTH levels, size of adenoma and number ofabnormal parathyroid glands (solitary v. multiple).

Group II had 0.45 times the odds of BNE versus UNEcompared with group I (unadjusted OR 0.45, 95% CI0.25–0.81, p = 0.008). The adjusted analysis demonstratesthat group II had 0.48 times the odds of BNE versus UNEcompared with group I (adjusted OR 0.48, 95% CI 0.23–1.03, p = 0.06). We performed a subgroup analysis to deter-mine the incremental value of ultrasonography to SPECTand to SPECT/CT. Adding ultrasonography to SPECTfor preoperative imaging of PHPT reduces the odds ofBNE by 62% compared with SPECT alone (adjusted OR0.38, 95% CI 0.15–0.96, p = 0.043). However, the additionof ultrasonography to SPECT/CT for preoperative local-ization of PHPT does not reduce the odds of undergoingBNE compared with SPECT/CT alone (adjusted OR2.93, 95% CI 0.18–46.5, p = 0.31).

DISCUSSION

The present study evaluates the incremental value ofuntrasonography on preoperative parathyroid adenomalocalization tests in terms of correct lateralization andclinical impact on extent of parathyroid surgery (BNE v.UNE). The overall results suggest that there is incremen-tal value in adding ultrasonography to an imaging test tolocalize a parathyroid adenoma preoperatively, as well asto reduce the risk of BNE and the extent of operative neckexploration. Interestingly, a subgroup analysis demon-strated that ultrasonography provides added value in cor-rectly localizing a parathyroid adenoma preoperativelyand reduces the chance of BNE when combined withSPECT but not when combined with SPECT/CT.

Other studies have also reported incremental valuewhen combining preoperative localization studies. Purcelland colleagues14 evaluated ultrasonography and planar

scintigraphy (conventional 2-dimensional imaging withoutSPECT) for parathyroid localization and found that theirsensitivity was 57% and 54% for the tests, respectively, butthat combining the tests increased the sensitivity to 78%.In a prospective clinical study, the combination of ultra-sonography and planar scintigraphy had 96% sensitivity,83% specificity, 88% positive predictive value and 94%negative predictive value.19 A recent review by Johnson andcolleagues20 also concluded that preoperative PHPTpatient imaging with ultrasonography and planar scintig -raphy most accurately predicted the location of solitaryadenomas when compared with either imaging test alone.These findings are consistent with the results of the pre-sent study, although our study further validated the valueadded of combining ultrasonography and planar imagingwith either SPECT or SPECT/CT scintigraphy by evalu-ating the impact on the extent of parathyroid exploration.

The protection against BNE afforded by ultrasonog -raphy is supported by reports of surgeon-performed ultra-sonography for preoperative parathyroid adenoma local-ization.22–24 These studies underscore the importance ofoperator experience for clinically informative neck ultra-sonography and suggest that the selective addition of ses-tamibi imaging (for cases in which the ultrasound wasnegative or equivocal) is the most cost-effective preopera-tive approach and the most protective against BNE.22,23,25

Our results echo the utility of ultrasonography in preoper-ative imaging for PHPT, especially for imaging modalitiesthat alone may not adequately facilitate a focused neckexploration.

A retrospective case series recently evaluated the incre-mental value of combining ultrasonography and SPECT/ CTwith software fusion for preoperative localization of para -thyroid adenomas.21 Fifty-nine patients were evaluated inthis study, and the sensitivity of ultrasonography and SPECT/CT was 64% and 90%, respectively, with concordant

Table 2. Degree of correctlateralization, by imaging test

Imaging test No. (%)

Group I, n = 75 41 (55)

SPECT only, n = 61 28 (46)

SPECT/CT, n = 14 13 (93)

Group II, n = 133 105 (79)

SPECT + ultrasonography, n = 62

47 (76)

SPECT/CT + ultrasonography, n = 71

58 (82)

CT = computed tomography; SPECT = single photon emission computed tomography.

Table 3. Odds of bilateral versus unilateral neck exploration

Model* OR (95% CI) p value

Group I v. group II

Group I 0.45 (0.25–0.81) 0.008

Group II 0.48 (0.23–1.03) 0.06

SPECT v. SPECT + ultrasonography

SPECT 0.40 (0.19–0.84) 0.015

SPECT + ultrasonography

0.38 (0.15–0.96) 0.043

SPECT/CT v. SPECT/CT + ultrasonography

SPECT/CT 3.78 (0.46-31.1) 0.22

SPECT/CT + ultrasonography

2.93 (0.18-46.5) 0.31

CI = con!dence interval; CT = computed tomography; OR = odds ratio; SPECT = single photon emission computed tomography. *Model adjusted for age, sex, preoperative calcium, intraoperative parathyroid hormone (PTH) monitoring, preoperative PTH level, size of adenoma, and type of adenoma (solitary v. nonsolitary).



RECHERCHE / RESEARCH

ultrasonographyandSPECT/CTfindings in59%ofcases.21 Theobservation that combinedultrasonographyandSPECT/CThadanoverall sensitivityof95%andaccuracyof91%ledthisgrouptoconcludethattherewasincrementalvalueincombiningthesetestsforpreoperativelocalization.21 Ourfindingsarenotentirelyconsistentwiththisgroup’sobservations,aswedidnotobserveanysignifi-cantly increasedvalueadded incombiningSPECT/CTwithultrasonography.Apotentialreasonforthisdifferencemightbe the fairlyhigh rate atwhichSPECT/CTwasabletocorrectlylocalizeaparathyroidadenomapreopera-tively(92.8%).TheriskofionizingradiationfromSPECT/CTmustbe

considered.ArecentstudyquantifiedtheradiationexposureassociatedwithSPECT/CToftheneckandpostulatedthatitwould takeanywhere from2390 to8030 scans tocause1 caseof radiation-inducedcancer,dependingonpatientageandsex.26 Thus,themostappropriateroleforSPECT/CTmaybeaspartofthepreoperativeevaluationofcom-plexPHPT.SeveralrecentstudieshavesupportedSPECT/CTasavaluablepreoperativeimagingmodalityforPHPT,particularlyincasesofmulti-glandulardiseaseandectopicparathyroidadenoma.5 Caseserieshavesuggestedpromisefor this techniqueas anevolvingmethodofpreoperativelocalization forPHPT.18,27 Our results are consistentwiththesefindings.Despite thevalue incombiningpreoperative imaging

studies,thereisalsoevidencetosupporttheuseofasinglepreoperativeimagingtest.Moureandcolleagues11 evaluatedtheutilityofplanarscintigraphyaloneinlocalizingparathy-roidadenomasandguidingUNE.TheresultsoftheirstudysuggestedthatpatientswithPHPTandunequivocallyposi-tiveplanarscintigraphyscansmaysafelyundergoafocusedparathyroidectomywithoutadditionalpreoperative im-aging.11 Incontrast,ourresultssuggestthatbyaddingultra-sonographytoSPECT,thereisanincreasedprobabilityofaccuratepreoperativeparathyroidlocalization,andwehavedemonstratedthatthiswillleadtodecreasedextentofneckexplorationduringparathyroidectomy.While severaloutcomes (e.g., cureofPHPT)could

havebeenevaluated,we specifically choseextentofpara-thyroidsurgery.Thisisanimportantoutcometoevaluate,as it is clinically relevant forboth surgeonsandpatients,doesnotdependoncomplete1-yearfollow-upandisnotsubjecttocensoringofdataasaresultofcasesbeinglosttofollow-up.Oneofthestrengthsofthepresentstudyisour100%

completedataon theprimaryoutcomeof interest.Per-haps themost important contributionof thepresentreport is translating the impactofpreoperative imagingtestsonclinicallyrelevantoutcomes,specificallytheextentofparathyroidsurgerythatmustbecarriedoutforasuc-cessfulparathyroidectomy.Our results suggest that com-biningSPECTwithultrasonography is a safe and in-formativepreoperative imagingpractice forPHPT that

reduces the riskofBNEandhelps facilitate a focusedparathyroidoperation.

Limitations

Thepresentstudyhasseverallimitations.Twoareitsrela-tively small sample size and its retrospectivenature.Theretrospectivenatureofourstudydidnotallowforimple-mentationof a standardizedprotocol forpreoperativeimaginginPHPT.Weattemptedtoaccountforthistypeofbiasbyincorporatingthetypeof imagingtestorderedintoourmultivariatemodel and the effectonextentofsurgery.While this approach is reasonable inmost in-stances,wehad a relatively small sample size (n =208)withonly70 outcomesofinterest(BNE).Giventhenum-berofcovariatesinourmultivariateanalysis,ouradjustedmodelmaybesusceptibletobeingoverfitted.Wealsodidnotvalidateourmodel,aswecouldnotstablymodelouroutcomeofinterestbysplittingourdatabaseintotestandvalidationsubsets.

CONCLUSION

Weexamined the clinicalutilityof addingultrasonog-raphy to2 commonlyperformedpreoperative imaginginvestigations forPHPT—SPECTandSPECT/CT—assessedonthebasisofaccuracyofpreoperativelocaliza-tion and the extentof surgical exploration subsequentlyrequired.Our analysisdemonstrates thatgreater incre-mental value is gainedby addingultrasonography toSPECTrather than toSPECT/CT.Further studyoftechniquesforaccuratelylocalizingabnormalparathyroidglandspreoperatively,particularlyinaprospectiveorran-domized controlled studydesign, is an important futuredirection thatwill ultimately lead to improvedoutcomesforindividualswithPHPT.Competing interests: Nonedeclared.

Contributors: M.C.Tee,D.S.LevineandS.M.Wisemandesigned thestudy.M.C.Tee,V.Nguyen,S.S.Strugnell,J.Yang,P.Tiwari,D.S. LevineandS.M.Wisemanacquired thedata,whichM.C.Tee,S.K. Chan,S. Jones andS.M.Wisemananalyzed.M.C.Tee,V. NguyenandS.M.Wiseman wrote the article, whichM.C.Tee, S.K. Chan,S.S.Strugnell,S. Jones,P.Tiwari,D.S.Levine andS.M.Wisemanreviews.Allauthorsapprovedthearticleforpublication.

References

1. AACE/AAESTaskForceonPrimaryHyperparathyroidism.TheAmeri-canAssociationofClinicalEndocrinologistsandtheAmericanAssocia-tionofEndocrineSurgeonspositionstatementonthediagnosisandman-agementofprimaryhyperparathyroidism.Endocr Pract 2005;11:49-54.

2. FraserWD.Hyperparathyroidism.Lancet 2009;374:145-58.

3. Suliburk JW,PerrierND.Primaryhyperparathyroidism.Oncologist2007;12:644-53.

4. BilezikianJP,SilverbergSJ.Clinicalpractice.asymptomaticprimaryhyperparathyroidism.N Engl J Med 2004;350:1746-51.



RESEARCH

5. Levine DS, Wiseman SM. Fusion imaging for parathyroid localiza-tion in primary hyperparathyroidism. Expert Rev Anticancer Ther2010; 10:353-63.

6. Sitges-Serra A, Bergenfelz A. Clinical update: sporadic primaryhyperparathyroidism. Lancet 2007;370:468-70.

7. Dillavou ED, Jenoff JS, Intenzo CM, et al. The utility of sestamibiscanning in the operative management of patients with primaryhyperparathyroidism. J Am Coll Surg 2000;190:540-5.

8. Russell C. Unilateral neck exploration for primary hyperparathy-roidism. Surg Clin North Am 2004;84:705-16.

9. Howe JR. Minimally invasive parathyroid surgery. Surg Clin North Am2000;80:1399-426.

10. Carling T, Udelsman R. Focused approach to parathyroidectomy.World J Surg 2008;32:1512-7.

11. Moure D, Larranaga E, Dominguez-Gadea L, et al. 99MTc-sestamibias sole technique in selection of primary hyperparathyroidismpatients for unilateral neck exploration. Surgery 2008;144:454-9.

12. Palazzo FF, Delbridge LW. Minimal-access/minimally invasiveparathyroidectomy for primary hyperparathyroidism. Surg Clin NorthAm 2004;84:717-34.

13. Lew JI, Solorzano CC. Surgical management of primary hyper-parathyroidism: state of the art. Surg Clin North Am 2009;89:1205-25.

14. Purcell GP, Dirbas FM, Jeffrey RB, et al. Parathyroid localizationwith high-resolution ultrasound and technetium tc 99m sestamibi.Arch Surg 1999;134:824-8; discussion 828-30.

15. Lumachi F, Tregnaghi A, Zucchetta P, et al. Technetium-99m sestamibiscintigraphy and helical CT together in patients with primary hyper-parathyroidism: a prospective clinical study. Br J Radiol 2004; 77:100-3.

16. Abikhzer G, Levental M, Rush C. High resolution MRI in the detectionof an intrathymic parathyroid adenoma. Br J Radiol 2006;79:e78-80.

17. Rodgers SE, Hunter GJ, Hamberg LM, et al. Improved preopera-tive planning for directed parathyroidectomy with 4-dimensionalcomputed tomography. Surgery 2006;140:932-40; discussion 940-1.

18. Levine DS, Belzberg AS, Wiseman SM. Hybrid SPECT/CT imagingfor primary hyperparathyroidism: case reports and pictorial review.Clin Nucl Med 2009;34:779-84.

19. De Feo ML, Colagrande S, Biagini C, et al. Parathyroid glands:Combination of (99m)tc MIBI scintigraphy and US for demonstra-tion of parathyroid glands and nodules. Radiology 2000;214:393-402.

20. Johnson NA, Tublin ME, Ogilvie JB. Parathyroid imaging: techniqueand role in the preoperative evaluation of primary hyperparathy-roidism. AJR Am J Roentgenol 2007;188:1706-15.

21. Patel CN, Salahudeen HM, Lansdown M, et al. Clinical utility ofultrasound and 99mTc sestamibi SPECT/CT for preoperative local-ization of parathyroid adenoma in patients with primary hyper-parathyroidism. Clin Radiol 2010;65:278-87.

22. Jabiev AA, Lew JI, Solorzano CC. Surgeon-performed ultrasound: asingle institution experience in parathyroid localization. Surgery2009; 146:569-75; discussion 575-7.

23. Siperstein A, Berber E, Barbosa GF, et al. Predicting the success oflimited exploration for primary hyperparathyroidism using ultra-sound, sestamibi, and intraoperative parathyroid hormone: analysis of1158 cases. Ann Surg 2008;248:420-8.

24. Adler JT, Chen H, Schaefer S, et al. What is the added benefit of cer-vical ultrasound to (9)(9)mTc-sestamibi scanning in primary hyper-parathyroidism? Ann Surg Oncol 2011;18:2907-11.

25. Tublin ME, Pryma DA, Yim JH, et al. Localization of parathyroidadenomas by sonography and technetium tc 99m sestamibi single-photon emission computed tomography before minimally invasiveparathyroidectomy: Are both studies really needed? J Ultrasound Med2009; 28:183-90.

26. Smith-Bindman R, Lipson J, Marcus R, et al. Radiation dose associ-ated with common computed tomography examinations and theassociated lifetime attributable risk of cancer. Arch Intern Med 2009;169:2078-86.

27. Profanter C, Prommegger R, Gabriel M, et al. Computed axialtomography-MIBI image fusion for preoperative localization in pri-mary hyperparathyroidism. Am J Surg 2004;187:383-7.

Readers of CJS can subscribe to email alerts toreceive the table of contents by email when a newissue appears. Sign up now at cma.ca/cjs.

Les lecteurs du JCC peuvent souscrire aux infocourriels pour recevoir un avis par courriel pourchaque nouveau numéro. Inscrivez-vous dèsmaintenant à amc.ca/cjs.




Total hip arthroplasty using a combined anteriorand posterior approach via a lateral incision in patients with ankylosed hips

Background: For most patients with severely ankylosed hips, traditional surgicalapproaches do not provide sufficient exposure during THAs. We report our experi-ence with a combined anterior and posterior approach using a lateral incision for totalhip arthroplasty (THA) in patients with severe, spontaneous bony hip ankylosis.

Methods: Between January 2004 and December 2008, patients with severe, sponta-neous bony hip ankylosis underwent THA via a combined anterior and posteriorapproach using a lateral incision.

Results: We included 47 patients (76 hips) with a mean age of 53 (range 22–72) yearsin our study. All surgeries were successful, and no significant postoperative complica-tions occurred. The mean operative duration was 1.5 (range 1.3–1.7) hours, and meanblood loss was 490 (range 450–580) mL. The mean duration of follow-up was 5.5(range 2–11) years. Harris hip score improved from 53 to 88 points postoperatively,and the outcome was good to excellent in 88.37% of cases. Heterotopic ossificationoccurred in 6 hips, and infection, which resolved with antibiotics, occurred in 1 patient.

Conclusion: This combined anterior and posterior approach to THA using a lateralincision in patients with severe, spontaneous ankylosis provides very good exposure,protects the abduction unit and results in good to excellent postoperative recovery.

Contexte : Pour la plupart des patients atteints d’une grave ankylose de la hanche, lesapproches chirurgicales classiques n’exposent pas suffisamment l’articulation lors desinterventions pour prothèse totale de la hanche (PTH). Nous faisons état de notreexpérience d’une approche antéro-postérieure à l’aide d’une incision latérale pour laPTH chez des patients atteints d’une grave ankylose ostéoarticulaire spontanée de lahanche.

Méthodes : Entre janvier 2004 et décembre 2008, nous avons effectué des interven-tions pour PTH par l’approche antéro-postérieure à l’aide d’une incision latérale chezdes patients présentant une grave ankylose ostéoarticulaire spontanée de la hanche.

Résultats : Nous avons inclus dans notre étude 47 patients (76 hanches) âgés enmoyenne de 53 (entre 22 et 72) ans. Toutes les interventions ont réussi et on n’a eu àdéplorer aucune complication postopératoire significative. La durée moyenne de l’in-tervention a été de 1,5 (entre 1,3 et 1,7) heure et la perte de sang moyenne a été de490 (entre 450 et 580) mL. La durée moyenne du suivi a été de 5,5 (entre 2 et 11) ans.Le score de Harris pour la hanche s’est amélioré, passant de 53 à 88 points après l’in-tervention, et les résultats ont été de bons à excellents dans 88,37 % des cas. On aobservé une ossification hétérotopique dans 6 hanches, et 1 patient a présenté uneinfection traitée avec succès par antibiothérapie.

Conclusion : Cette approche antéro-postérieure à l’aide d’une incision latérale pourla PTH chez des patients présentant une grave ankylose ostéoarticulaire spontanéeoffre une excellente visibilité. Elle protège le système d’abduction et permet un réta -blissement postopératoire de bon à excellent.

A nkylosis, spontaneous fusion of a joint after injury or disease, is causedby a number of conditions, including rheumatoid arthritis, tuberculo-sis, septic arthritis, ankylosing spondylitis, severe osteoarthritis and hip

trauma.1 Regardless of whether a patient has fibrous or bony ankylosis, theclinical manifestation is complete loss of both active and passive motion of thehip joint.2 Considering the negative impact on quality of life, an appropriate

Jian Li, MD*†

Zhiwei Wang, MD*

Ming Li, MD*

Yuesong Wu, MS*

Weidong Xu, MD*

Zimin Wang, MD*

From the * Department of Orthopedics,the Second Military Medical UniversityChanghai Hospital and the †Departmentof Orthopedics, Hainan Division of Gen-eral Hospital of PLA, Sanya, China


Correspondence to: Zhiwei WangDepartment of OrthopedicsSecond Military Medical University Changhai Hospital168 Changhai RoadShanghai, China [email protected]

DOI: 10.1503/cjs.000812

total-li_Layout 1 13-09-16 10:36 AM Page 332


RESEARCH

surgical intervention should be considered. Total hiparthroplasty (THA) has become the most common surgicalprocedure performed in developed nations.2,3 Good out-comes are reported following THA in patients with anky-losed hips;4 however, the procedure is technically demand-ing, and postoperative complications are not uncommon,particularly when compared with primary THAs.5

Common approaches to the nonankylosed hip includethe anterior, posterior and lateral approaches. Each has itsown advantages and disadvantages. For example, theSmith–Peterson approach cannot ideally expose the hipjoint unless either the gluteus medius is partially de tachedfrom the greater trochanter or an additional trochantericosteotomy is performed. The Watson–Jones approachoffers poor operative exposure and requires partial des -truction of the abductors, and the space pos ter ior to theacetabulum is limited for intraoperative man ipulation.The Hardinge approach provides limited exposure andthus is not suitable for patients with bony hip ankylosis,especially for those in whom there is difficulty exposingthe posterior margin of the acetabulum. Finally, withthe posterolateral approach, abduction of the hip is lim-ited and anterior exposure of the acetabulum is relativelydifficult.6

For most patients with severely ankylosed hips, tradi-tional surgical approaches do not provide sufficient expos -ure during THAs.7–10 Femoral external rotation or adduc-tion deformities are common with ankylosed hips, and thusa simple anterior or posterior approach cannot satisfac -torily expose the surgical field. The purpose of this reportis to relay our experience with a combined anterior andposterior approach using a lateral incision for THA inpatients with severely ankylosed hips. We hypothesizedthat this combined approach would improve exposure ofthe hip joint capsule compared with both the posterior andanterior approaches and that it would result in good toexcellent clinical outcomes.

METHODS

Patient selection

Between January 2004 and December 2008, patients withspontaneous bony hip ankylosis undergoing THA weretreated using a combined anterior and posterior approachby the same group of orthopedic surgeons at the Chang-hai Hospital, Shanghai, China. Surgical indications forTHA included persistent low back pain, ipsilateral kneepain and instability, contralateral hip pain, nonfunctionalhip fusion preventing ipsilateral knee arthroplasty, flexiondeformity greater than 30°, adduction deformity greaterthan 10° and heterotopic fusion. Young patients requiringaesthetic or functional correction were also included. Weexcluded patients if there was either absence or severefibrosis of the abductors owing to prior surgeries, recentsigns of infection or marked atrophy of the quadricepsand/or abductors. We also excluded patients with fibrousankylosis. Patients were consecutively enrolled, and theyall provided informed consent. The ethics committee ofthe hospital approved our study protocol.

We calculated Harris hip scores for all patients bothpre- and postoperatively (at the time of last follow-up). Ascore of 90–100 points was considered excellent, 80–89 points was considered good, 70–79 points was consid-ered fair and a score of less than 70 points was consideredpoor.

Surgical procedure

Patients were placed in the lateral recumbent position(Fig. 1A). The surgeon made a skin incision along the cen-tre of the femoral shaft, with the superior end 8 cm abovethe tip of the greater trochanter and the inferior end 7 cmbelow the tip of the greater trochanter (Fig. 1B). Thus, thetotal incision length was 15 cm. The subcutaneous tissues

Fig. 1. (A) The patients were placed in a lateral recumbent position with the diseased limb upward. Thepubic symphysis and sacrum were fixed anteriorly and posteriorly, and the motion of the lower limb waspreserved. (B) Illustration of the surgical incision with reference to the greater trochanter.

BA



RECHERCHE

were dissected to expose the fascia lata, which was thenlongitudinally split distally to proximally. The trochantericbursa over the greater trochanter was incised, and then thegluteus maximus was cut proximally along the intervalbetween the gluteus maximus and the tensor fascia lata tothe proximal skin incision. The surgeon then made a 2 cmincision and bluntly dissected the fascia lata superiorly andinferiorly. Two thyroid retractors were used to slightly pullthe tensor fascia lata to maintain tension. The posteriorside was exposed first (Fig. 2A). The operating table wasrotated toward the prone position (away from the sur-geon) by 15°, and gauze was packed inside the deep sur-face of the gluteus maximus to control bleeding. The glu-teus maximus and fascia lata were retracted posteriorly toexpose the gluteus medius and external rotators. Inpatients with ankylosed hips, it is not possible to flex theknee joint to 90° and internally rotate the hip joint tomake the external rotators tense. Therefore, the piriformistendon was carefully elevated with a long, curved vascularclamp at the junction of the posterior margin of the inser-tion site of the gluteus medius and the greater trochanter.The piriformis was then incised at the insertion site of thetendon. Along the intertrochanteric fossa and the poster -ior side of the greater trochanter, the superior gemelluse,obturator internus, inferior gemellus and the insertion siteof the quadratus femoris were cut close to the bone withan electric scalpel. The surgeon further dissected theexternal rotators on the posterior and inferior surface ofthe joint capsule with a long-handled periosteal elevator.The retractor was adjusted to expose the posterior andinferior joint capsule. The junction between the deep partof the gluteus minimus and the superficial surface of theupper joint capsule was bluntly dissected with a long- handled periosteal elevator, and the gluteus minimus andthe gluteus medius were retracted superiorly with a Hoff-man retractor. Most parts of the superior, posterior andinferior joint capsule were removed, and the attached softtissues were removed with an electric scalpel along the

femoral neck. Large gauze was used to pack the incisionsite to complete the posterior exposure.

To expose the anterior joint capsule, the operating tablewas rotated toward the supine position (toward the sur-geon) by 15°, and the fascia lata was retracted anteriorlyand posteriorly. An incision was made along the junctionbetween the tensor fascia lata and the gluteus medius(Fig. 2B). The gluteus medius was retracted posteriorly, thefascia lata was pulled anteriorly, and the anterior part of thejoint capsule was exposed after releasing the soft tissueadhering to the femoral side of the joint capsule. Dissec-tion was performed along the space between the gluteusmedius and tensor fascia lata toward the joint capsule(Fig. 2C). The soft tissue on the anterior joint capsule wasthen dissected with a periosteal elevator, and the anteriorand inferior aspect of the joint capsule was exposed com-pletely by adjusting the position of the retractors. A Hoff-man retractor was inserted between the deep part of thegluteus minimus and the superficial surface of the upperjoint capsule to pull the gluteus minimus and the gluteusmedius superiorly. Injury to the superior gluteal nerve dueto excessive retraction was avoided. The anterior joint cap-sule and remnants of the superior and inferior joint capsulewere removed. To improve exposure, the femoral vastuslateralis was split in the direction of the femur. The anky-losed hip joint was totally exposed and the external abduc-tors (i.e., gluteus medius and minimus) were preserved.

The femoral neck osteotomy (Figs. 3 and 4) wasachieved in 2 steps. First, via the anterior approach, theosteotomy was performed 2 cm posteroinferior to thejunction of the femoral head and neck with an oscillatingsaw. Via the posterior approach, the osteotomy was per-formed 2 cm interoinferior to the junction of the femoralhead and neck. A V shape was formed by the anterior andposterior clefts to make the previously ankylosed hip amobile hip. Then, a standard femoral neck osteotomy wasperformed 1 cm superior to the lesser trochanter. The sci-atic nerve (identified via the loosely packed adipose tissue

Fig. 2. (A) The operating table was rotated away from the surgeon by 15°, and the posterior joint capsule of the hip joint wasexposed. (B) The operating table was then rotated toward the surgeon by 15° to allow access to the anterior structures of the hipjoint. (C) The anterior hip joint capsule was exposed by approaching the spatium intermusculare between the gluteus medius andtensor fascia lata.

A B C



RESEARCH

surrounding the nerve) was located in the gap between thegluteus maximus and the group of external rotation mus-cles. Under normal circumstances, no routine release wasperformed; however, for patients with severely contractedjoint capsules, the soft tissue surrounding the sciatic nervewas released by 4–5 cm along the nerve to provide addi-tional room. The newly formed hip joint was moved re -peatedly and the tense soft tissue, particularly the inferiorand medial soft tissue, around the hip joint was released.

The exact location of the acetabulum was identified withthe assistance of preoperative radiographs and remainingstumps following the femoral neck osteotomy. This is animportant step for ankylosed hips with in situ fusion; how-ever, it does not work for the ankylosed hip with deformedfusions. A small curved osteotome was then inserted approx-

imately 2 cm posterior and inferior to the junction of thefemoral head and neck in the anterior surgical field. Preop-erative radiographs faciliated proper positioning. The kneejoint was flexed 90°, and the hip joint was internally rotatedto expose the acetabulum. The acetabulum was reamed at45° inclination and 15° anteversion, and an acetabular trialinsert was placed. Next, the femoral canal was reamed and asuitable femoral neck prosthesis was selected. The femoralcomponents used were coated anatomic medullary locking(DePuy), and the acetabular components used were Duraloc(Depuy). Of the implants, 9 were cemented and 67 werenoncemented. Hip joint range of motion and soft tissue ten-sion were observed after reduction. If the soft tissue, es -pecially the gluteus medius, was tense, then a secondaryrelease was made along the superior margin of the acetabu-lum until the result was satisfactory. Hip joint reduction wasperformed, and again the hip joint range of motion and softtissue tension were assessed. The surgical wound was closedvia a layer-by-layer method, and continuous drainage withnegative pressure was applied.

We defined blood loss as the volume of intraoperativesuction plus the change in weight of the sponges used(postoperative sponge weight – preoperative spongeweight), with the sum divided by the density of blood(1.055 g/cm3). Postoperative blood loss through drainagewas not included.

Postoperative management and follow-up

Preoperative parenteral antibiotics and prophylaxis fordeep vein thrombosis (4 mg/d of fraxiparine, subcuta-neous) were administered to all patients. Postoperatively,the affected limb was placed in a neutral abduction posi-tion and a special shoe was applied to avoid adduction andinternal rotation, as this can lead to the development ofearly postoperative hip joint dislocation. A pillow was

Fig. 4. (A) The femoral neck was exposed via a combined anterior and posterior approach. (B) The femoral head has been amputated.

A B

Fig. 3. Result of the femoral neck exposure. The abductors werepreserved after exposing the femoral neck by gently applyingretractors after the combined anterior and posterior approach.There was no need to perform an osteotomy of the greatertrochanter in any of the included patients.



RECHERCHE

placed under the knee to flex the knee joint 30°, and a tri-angular pillow was placed between the legs. The drain wasremoved 24–48 hours postoperatively.

Non–weight bearing exercise and crutch ambulationwere initiated 1 and 2 weeks postoperatively, respectively.Partial weight-bearing began 6–8 weeks postoperatively; itwas delayed to provide sufficient time to ensure maximalsoft-tissue healing because the extent of the soft-tissue dis-section in ankylosed hips is greater than in healthy hips.We obtained radiographs before discharging the patients.

All patients underwent follow-up examination 1, 3 and6 months after discharge and yearly thereafter. Two in -dependent observers, 1 surgeon and 1 physician not in -volved in the arthroplasty, performed the follow-up exam -inations. Pain, function, deformities and range of motionusing the Harris hip score were assessed at the last follow-up, and the sum range of motion was measured based onflexion, extension, abduction, internal adduction internaland external rotation movements.

RESULTS

We included 47 patients (76 hips) in our study: 29(48 hips) were men, and the mean patient age was 53(range 22–72) years. Patient characteristics are summar -ized in Table 1. All cases were primary THAs in patientswith severe, spontaneous ankylosed hip joints. A represen-tative case is shown in Figure 5. The mean operative dura-tion was 1.5 (range 1.3–1.7) hours, and mean blood losswas 490 (range 450–580) mL. Fractures limited to theproximal femoral metaphysis occurred in 5 patients, andwiring was required in 2 of them. Immediate postopera-tive radiographs revealed excellent location of all prosthe-

ses, as the rotation centre of the hip joints were recon-structed and the fixation status of the prostheses wasdeemed acceptable.

The mean duration of follow-up was 5.5 (range 2–11)years. Four patients (6 hips) were lost to follow-up afterdischarge. In the remaining 43 patients, there were nopostoperative dislocations or hip impingements. The meanHarris hip score increased from 53 points preoperatively to88 points postoperatively. The outcome was excellent in14 patients, good in 24, fair in 4 and poor (because ofinfection) in 1 patient. In total, 88.37% of the outcomeswere considered good to excellent.

Postsurgically, 4 hips were associated with mild pain and

Table 1. Demographic and clinical characteristics of patientsundergoing total hip arthroplasty for ankylosed hip joints

Characteristic No. (%)*

No. of patients (no. of hips) 47 (76)

Sex

Male 29 (62)

Female 18 (38)

Affected side

Right 36 (77)

Left 40 (85)

Age, mean (range) yr 53 (22–72)

Duration of disease at time of surgery, mean (range) yr 7.5 (5–9)

Patient follow-up time, mean (range) yr 5.5 (2–11)

Presence of hip !exion contracture 67 (88)

Cause of ankylosis

Rheumatoid arthritis 9 (19.1)

Tuberculosis of the hip 6 (12.9)

Sequela of the septic arthritis 7 (14.9)

Anklylosing spondylitis 16 (34.0)

Osteoarthritis 5 (10.6)

Hip trauma 4 (8.5)

*Unless otherwise indicated.

Fig. 5. (A) Preoperative radiographs of a 35-year-old man. Hehad ankylosing spondylitis associated with bilateral ankylosedhip joints. Because the patient had lower extremity deformities,the pelvis could not be oriented in an ideal anteroposterior posi-tion for radiographs. The range of motion in every direction was0 in both hip joints and the patient experienced severe pain inboth hip joints. (B) Anteroposterior radiographs of the samepatient 3 years postoperatively. The position of the prosthesiswas excellent. The pain in both hip joints was completelyresolved, and 89° and 100° flexion was achieved in the left andright hip joints, respectively. The patient’s gait was normal postoperatively.

A

B



RESEARCH

discomfort, which was successfully managed via patient-controlled analgesia. In 1 case there was pain in the middleand upper thigh, which gradually improved over 3–4 weeks.Heterotopic ossification occurred in 6 hips: 3 were grade Iand the rest were grade II according to the Brooker classi-fication. Hip function was not affected in any of thesepatients. Mild nerve traction injury occurred in 2 patients;it presented as numbness of the skin over the lateral aspectof the limb and foot on the operated side. Nonsteroidalanti-inflammatory drugs and vitamin B were prescribed,and the condition resolved in both patients within6 months of surgery. Neither patient had evidence ofmotor neuron injury.

Postoperative measurements of the leg revealed a leglength difference of ≤ 1 cm in all but 1 patient, in whom1 limb was lengthened by 1.3 cm and the other limb short-ened by 0.9 cm. The patient had difficulty walking, whichwas resolved by adjusting the patient’s shoes. The maintype of nerve injury was sciatic nerve palsy, and all patientswere negative for Trendelenburg gait during the last fol-low-up visit. Preoperatively, flexion contracture was notedin 67 (88%) hips, the mean deformity angle was 37° (range12°–89°), and range of motion was 0°. At the last follow-up, mean range of motion was 119.7° (range, 100°–135°) inthe replaced joints (Table 2).

DISCUSSION

This report describes the results of a THA technique in47 patients with severe, spontaneous ankylosed hip jointsusing a combined anterior and posterior approach via alateral incision. This combined approach preserves theabductors with no damage to either the gluteus medius orgluteus minimus, thereby improving postoperative jointfunction; does not increase the length of the surgical inci-sion, but rather alters the position of the surgical field forperforming both the anterior and posterior procedures;clearly exposes the acetabulum and adjacent tissues; andallows the incision to be extended superiorly or inferiorlyat any time during the surgery, if indicated.

This approach is markedly different from the standardsurgical approaches to THA. Two common approaches forTHAs are a posterolateral approach (e.g., the Gibson orOsborne approaches) and a lateral approach (e.g. the Ollierapproach).2,5 When performing a primary THA, theseapproaches are simple and convenient, the anatomic struc-tures are distinct and surgical exposure is sufficient toachieve a satisfactory outcome. In ankylosed hip joints,however, these approaches are not optimal. The Gibsonapproach easily exposes the femoral side, but the externalrotators and posterior joint capsule are difficult to expose ifthe ankylosis is associated with an external rotation de -form ity of the femur. Furthermore, the exposure andrelease of the anterior joint capsule is difficult if an adduc-tion deformity exists. The Watson–Jones approach doesnot clearly expose the medial and inferior margin of theacetabulum. Thus, extensive soft tissue release (removingmost parts of the insertion site of the gluteus medius) or agreater trochanter osteotomy are often inevitable, whichcan contribute to the development of postoperative com-plications, such as muscle weakness in the abductors, su -per ior gluteal nerve injury (innervating the tensor fascialata, gluteus medius and minimus) and greater trochanternonunion.11 The Hardinge approach requires no greatertrochanter osteotomy, and the approach maintains theintegrity of the gluteus medius. However, the approachprovides limited exposure, and thus is not suitable forpatients with bony hip ankylosis, especially for those inwhom there is difficulty exposing the posterior margin ofthe acetabulum, which would affect placement of theacetabular prosthesis.

A transtrochanteric approach is the most commonlyused approach for ankylosed hips.12,13 However, it can resultin massive bleeding and has a relatively long operativeduration. Furthermore, fixation of the greater trochanter isdifficult, heterotopic ossification is severe, postoperativelocal pain occurs, bursitis of the greater trochanter candevelop, and the rate of trochanteric nonunion followingthe transtrochanteric approach has been reported to be0.8%–5%.14,15

Table 2. Mean patient hip movements at various time points postoperatively

Postoperative time; mean ± SD

Function 8 wk 12 wk 6 mo

Flexion, °

Left 65.08 ± 8.16 90.57 ± 5.55 115.54 ± 9.74

Right 68.13 ± 7.09 90.66 ± 6.82 111.66 ± 12.00

Extension, °

Left 7.97 ± 1.91 10.51 ± 2.45 19.84 ± 6.04

Right 8.41 ± 2.01 10.59 ± 1.92 20.63 ± 7.49

Abduction, °

Left 24.38 ± 4.52 27.92 ± 2.69 34.73 ± 4.24

Right 24.09 ± 4.11 28.09 ± 2.47 34.53 ± 2.65

SD = standard deviation.



RECHERCHE

Tab

le 3

. Su

mm

ary

of

key

stu

die

s o

f to

tal h

ip a

rth

rop

last

y in

pat

ien

ts w

ith

an

kylo

sed

hip

s

Stud

y P

atie

nts

Surg

ical

app

roac

h O

utco

mes

Pres

ent s

tudy

47

pat

ient

s (7

6 hi

ps) w

ith s

pont

aneo

us

anky

losi

s.

A c

ombi

ned

ante

rola

tera

l and

pos

tero

late

ral

appr

oach

via

a la

tera

l inc

isio

n. M

ean

follo

w-u

p w

as

5.5

year

s.

Har

ris h

ip s

core

impr

oved

fro

m 5

3 to

88

poin

ts p

osto

pera

tivel

y. T

he o

utco

me

was

exc

elle

nt t

o go

od in

88.

37%

of

case

s. F

emor

al p

rost

hesi

s su

bsid

ence

oc

curr

ed in

3 h

ips.

The

max

imum

sub

side

nce

was

2 m

m. N

o tr

eatm

ent

was

re

quire

d as

the

re w

ere

no s

ympt

oms

rela

ted

to t

he s

ubsi

denc

e. H

eter

otop

ic

ossi

!cat

ion

occu

rred

in 6

hip

s.

Bha

n et

al.16

R

etro

spec

tive

revi

ew o

f 54

pat

ient

s (9

2 hi

ps) w

ho u

nder

wen

t ce

men

tless

TH

A f

or b

ony

anky

losi

s du

e to

an

kylo

sing

spo

ndyl

itis.

Post

erio

r ap

proa

ch. M

ean

follo

w-u

p w

as 8

.5 y

ears

. H

arris

hip

sco

res

impr

oved

fro

m 4

9.5

to 8

2.6.

Com

plic

atio

ns in

clud

ed p

ain

(n

= 1

0), a

nter

ior d

islo

catio

n (n

= 4

), sc

iatic

ner

ve p

alsy

(n =

1),

revi

sion

ar

thro

plas

ty d

ue t

o as

eptic

loos

enin

g (n

= 1

3).

Kim

et a

l.18

12 p

atie

nts

with

bila

tera

l any

klos

is

(24

hips

). A

ll su

rger

ies

perf

orm

ed b

y a

sing

le s

urge

on v

ia t

he

stan

dard

tra

nstr

ocha

nter

ic a

ppro

ach.

H

arris

hip

sco

res

incr

ease

d fr

om 5

5.4

to 8

2.3.

Com

plic

atio

ns in

clud

ed

oste

olys

is (n

= 3

) and

loos

enin

g (n

= 2

), an

d 11

cup

s w

ere

outs

ide

the

safe

ra

nges

of

Lew

inne

k.

Abd

el-A

al e

t al

.19

12 p

atie

nts

(15

hips

) with

sur

gica

lly

(n =

5) a

nd s

pont

aneo

usly

fus

ed h

ips

(n =

7).

Har

ding

e ap

proa

ch.

Har

ris h

ip s

core

s in

all

patie

nts

impr

oved

fro

m 4

2 pr

eope

rativ

ely

to 7

6 po

stop

erat

ivel

y. C

ompl

icat

ions

incl

uded

fem

oral

art

ery

inju

ry d

ue t

o H

ofm

ann

retr

acto

rs in

fro

nt o

f th

e ac

etab

ulum

, 1 f

aile

d TH

A.

Josh

i et

al.5

103

patie

nts

unde

rwen

t 18

1 TH

As

for

anky

losi

ng s

pond

yliti

s.

Eith

er d

irect

late

ral a

ppro

ach

via

troc

hant

eric

os

teot

omy

or t

he H

ardi

nge

appr

oach

. All

patie

nts

rece

ived

cem

ente

d lo

w-f

rictio

n TH

As.

No

Har

ris h

ip s

core

s w

ere

repo

rted

. Ins

tead

, aut

hors

indi

cate

d th

at 9

6% o

f hi

ps h

ad a

n ex

celle

nt (l

ow) p

ain

scor

e an

d 29

.25

of t

he h

ips

had

norm

al o

r ne

ar-n

orm

al f

unct

ion.

Rut

z et

al.20

22

pat

ient

s (2

2 an

kylo

sed

hips

) se

cond

ary

to p

ostt

raum

atic

os

teoa

rthr

itis,

cox

itis,

hip

dys

plas

ia a

nd

prim

ary

oste

oart

hriti

s.

19 c

onve

rsio

ns w

ere

via

a la

tera

l app

roac

h an

d 3

wer

e vi

a an

ant

erol

ater

al a

ppro

ach;

20

patie

nts

wer

e av

aila

ble

at a

mea

n fo

llow

-up

of 1

3.2

year

s.

Har

ris h

ips

scor

es p

osto

pera

tivel

y w

ere

84.9

(pre

surg

ical

sco

res

wer

e no

t ci

ted)

. Com

plic

atio

ns o

ccur

red

in 3

0% o

f pa

tient

s an

d in

clud

ed t

rans

ient

ap

raxi

a of

the

fem

oral

ner

ve, a

lesi

on o

f th

e sc

iatic

ner

ve, 2

dee

p in

fect

ions

, an

d as

eptic

loos

enin

g of

2 f

emor

al s

tem

s.

Ham

adou

che

et a

l.21

45 p

atie

nts

(45

hips

) und

erw

ent T

HA

for

tr

eatm

ent o

f spo

ntan

eous

ank

ylos

is

(n =

20)

and

pos

tope

rativ

e an

kylo

sis

(n =

35)

. Cas

es o

f ank

ylos

ing

spon

dylit

is

and

!bro

us a

nkyl

osis

wer

e ex

clud

ed.

Late

ral a

ppro

ach

with

a s

tand

ard

troc

hant

eric

os

teot

omy.

M

erle

d’A

ubig

né h

ip s

core

s in

crea

sed

from

11.

3 po

ints

pre

oper

ativ

ely

to

16.5

poi

nts

at t

he t

ime

of la

st f

ollo

w-u

p. S

atis

fact

ory

hip

func

tion

was

ac

hiev

ed in

41

(91%

) pat

ient

s. C

ompl

icat

ions

incl

uded

dee

p ve

in t

hrom

bosi

s (n

= 2

), co

mm

on p

eron

eal n

erve

pal

sy (n

= 1

), an

d pe

ripro

sthe

tic a

nd in

guin

al

absc

ess

1 ye

ar p

osto

pera

tivel

y (n

= 1

).

Raj

arat

nam

et

al.22

15

pat

ient

s (1

6 hi

ps) u

nder

goin

g ce

men

tless

TH

A t

o tr

eat

spon

tane

ous

anky

losi

s (in

clud

ing

anky

losi

ng

spon

dylit

is) o

r su

rgic

al a

rthr

odes

is.

Har

ding

e ap

proa

ch in

5 p

atie

nts

and

the

rem

aini

ng

10 p

atie

nts

had

a po

ster

ior

appr

oach

with

an

enha

nced

pos

terio

r re

pair.

Med

ian

follo

w-u

p w

as

10.7

5 ye

ars.

Har

ris h

ip s

core

s im

prov

ed f

rom

70

to 8

3 an

d M

erle

d’A

ubig

né h

ip s

core

s ha

d a

scor

e of

4.8

for

pai

n, 3

.9 f

or m

obili

ty, a

nd 4

.4 f

or f

unct

ion.

The

ove

rall

Mer

le d

’Aub

igné

hip

sco

re p

osto

pera

tivel

y w

as a

mea

n of

13.

Ase

ptic

lo

osin

g re

quiri

ng a

ceta

bula

r cu

p re

visi

on o

ccur

red

in o

ne p

atie

nt, a

nd d

eep

vein

thr

ombo

sis

occu

rred

in a

noth

er.

Ban

gjia

n et

al.23

12

pat

ient

s w

ith b

ilate

ral a

nkyl

osed

hip

jo

ints

(24

hips

) cau

sed

by la

te

anky

losi

ng s

pond

yliti

s.

The

mos

t ap

prop

riate

app

roac

h de

pend

ing

on p

atie

nt

anat

omy

and

the

inci

sion

was

via

the

pos

tero

late

ral

appr

oach

. Mea

n fo

llow

-up

was

4.2

yea

rs.

All

hip

join

t fu

nctio

n im

prov

ed, a

nd #

exio

n de

form

ity w

as c

orre

cted

. Fle

xion

ra

nges

wer

e 75

–105

° (m

ean

84.4

°) a

nd e

xten

sion

ran

ges

wer

e 10

° to

20°

(m

ean,

18.

7°).

Har

ris h

ip s

core

s ra

nged

fro

m 1

5.21

poi

nts

preo

pera

tivel

y to

86

.25

poin

ts p

osto

pera

tivel

y. N

o pa

tient

exp

erie

nced

hip

pai

n po

stop

erat

ivel

y, a

nd p

reop

erat

ive

knee

and

low

er b

ack

pain

wer

e cl

early

re

lieve

d po

stop

erat

ivel

y.

THA

= t

otal

hip

art

hrop

last

y.



RESEARCH

The combined approach presented herein may appearsimilar to the technique described for periacetabularosteotomies and to some of the surgeries retrospectivelyreviewed by Bhan and colleagues,16 and to the “minimallyinvasive 2-incision approach” described by Berger;17 how-ever, there are some major differences. The Berger methodapproaches the hip between the tensor fascia lata and thesartorius muscle, and the rectus femoris is pulled inwardlyto expose the capsula articularis coxae. This method doesnot adequately expose the acetabulum in patients withseverely ankylosed hips. In addition, Bhan and colleaguesand Berger reported the use of fluoroscopy and 2 skin inci-sions for each hip.

As our study was not a controlled clinical trial, directlycomparing these results with a similar group of patientstreated via alternate approaches is difficult. As demon-strated in Table 3, the published studies on THA inpatients with ankylosed hips include variable etiologies ofthe ankylosis and variable outcome measures, use differenttypes of prostheses and may or may not involve performinga trochanteric osteotomy.5,16,18–23 Nonetheless, some com-parisons can be made, as Harris hip scores have beenreported in several studies. In our study, the Harris hipscore improved by a mean of 35 points. This is marginallygreater than the improvement reported by Bhan and col-leagues ( 33.1),16 Kim and colleagues (26.9)18 and Abdel-Aaland colleagues (34),19 and markedly greater than theimprovement reported by Rajaratnam and colleagues(13).22 However, the improvement in our study is notablylower than that reported by Bangjian and colleagues (71).23

In addition, postoperative complications in our study wereminimal compared with those noted in the other publishedstudies. For example, nerve injury to the sciatic, femoraland common peroneal nerves were reported in a numberof studies, as were deep vein thrombosis, aseptic looseningand deep infections.

The mean range of motion (119.7°) in our study is bet-ter than that reported in many primary THA studies. Thismay be because of differences in the procedures and thesubjectivity in measurements. In our practice, range ofmotion is in part determined by evaluating what kinds ofdaily activities the patients can perform. For patients whocould squat for toileting (and in this study all patientscould do so), performing such action would require atleast 120° of hip flexion. After a primary THA, patientswould likely be able to satisfy the criterion of sitting onstools, which typically requires 100°–120° range ofmotion.

Despite being unable to directly compare the publishedstudies on THA in ankylosed hips, our study presentssuperior outcomes in terms of low dislocation rate, goodhip abduction function and full correction of preoperativehip deformities than the other series listed in Table 3.Futhermore, the approach described herein optimizes theprosthetic insertion by fully exposing the acetabular and

femoral sides of surgical fields and prevents excessive softtissue traction that could lead to blunt soft tissue injury.Finally, this approach avoids injuring the gluteal muscula-ture that are essential in postoperative recovery of gait andhip abduction.

To ensure successful outcomes using this approach, anumber of key points should be considered. First, be par-ticularly diligent about protecting the gluteus medius andminimus and avoiding excessive traction of the gluteusmedius and minimus to protect the superior gluteal nerve.Next, during the femoral neck osteotomy, protect theperipheral soft tissue. Third, repair the anterior and pos -terior soft tissues around the hip joint carefully, as surgicalrelease leads to the loss of the stabilizing tissues, such asthe joint capsule and tendons. Finally, when preparing theacetabulum and femur, protect the peripheral muscles bypacking with gauze to avoid seeding bone marrow com -ponents into the peripheral muscles. This may decrease theincidence of heterotopic ossification.

Limitations

As noted, a limitation of this study is that it lacked a con-trol group. In addition, the sample was relatively small andthe follow-up relatively short. Finally, the Harris hipscore, although commonly used, has many limitations as aresearch tool. However, any scoring method has limita-tions, and the Harris hip score has been adopted by ourinstitution as a standard for assessment. Therefore, weused this method for the purpose of standardization anduniformity in our collected data.

CONCLUSION

This combined anterior and posterior approach using alateral incision provides very good exposure of ankylosedhip joints, protects the abduction unit and results in goodto excellent postoperative recovery. This THA approachshould be considered in the management of patients with1 or more naturally ankylosed hips.Competing interests: None declared.

Contributors: Zhiwei Wang designed the study. J. Li, Zhiwei Wang,Y. Wu and W. Xu acquired the data, which Zhiwei Wang, M. Li, Y. Wuand Zimin Wang analyzed. J. Li and Zimin Wang wrote the article,which all authors reviewed and approved for publication.

References

1. Kim YH, Oh SH, Kim JS, et al. Total hip arthroplasty for the treat-ment of osseous ankylosed hips. Clin Orthop Relat Res 2003;414:136-48.

2. Sathappan SS, Strauss EJ, Ginat D, et al. Surgical challenges incomplex primary total hip arthroplasty. Am J Orthop 2007; 36:534-41.

3. Mizra SB, Dunlop DG, Panesar SS, et al. Basic science considerations


in primary total hip replacement arthroplasty. Open Orthop J 2010;4:169-80.

4. Kilgus MD, Amstutz HC, Wolgin MA, et al. Joint replacement forankylosed hips. J Bone Joint Surg Am 1990;72:45-54.

5. Joshi AB, Markovic L, Hardinge K, et al. Total hip arthroplasty inankylosing spondylitis: an analysis of 181 hips. J Arthroplasty 2002; 17:427-33.

6. Palan J, Beard DJ, Murray DW, et al. Which approach for total hiparthroplasty: Anterolateral or posterior? Clin Orthop Relat Res 2009;467: 473-7.

7. Bottner F, Delgado S, Sculco TP. Minimally invasive total hip replace-ment: the posterolateral approach. Am J Orthop 2006;35:218-24.

8. Bertin KC, Röttinger H. Anterolateral mini-incision hip replacementsurgery. A modified Watson-Jones approach. Clin Orthop Relat Res2004; 429:248-55.

9. Basad E, Ishaque B, Stürz H, et al. The anterolateral minimally inva-sive approach for total hip arthroplasty: technique, pitfalls, and wayout. Orthop Clin North Am 2009;40:473-8.

10. Ritter MA, Harty LD, Keating ME, et al. A clinical comparison ofthe anterolateral and posterolateral approaches to the hip. ClinOrthop Relat Res 2001;385:95-9.

11. Amstutz HC, Dennis N. Total joint replacement for ankylosed hips.Indications, technique, and preliminary results. J Bone Joint Surg Am1975; 57:619-25.

12. Kilgus DJ, Amstutz HC, Wolgin MA, et al. Joint replacement forankylosed hips. J Bone Joint Surg Am 1990;72:45-54.

13. Kevin B. Cleveland KB. Nontraumatic soft tissue disorders. In:

Campbell’s Operative Orthopaedics. 11th ed. (1 pt VII). Philadelphia(PA): Mosby Elsevier; 2007.

14. Schinsky MF, Nercessian OA, Arons RR, et al. Comparison of com-plications after transtrochanteric and posterolateral approaches forprimary total hip arthroplasty. J Arthroplasty 2003;18:430-4.

15. Cashman JP, Cashman WF. Comparison of complications intranstrochanteric and anterolateral approaches in primary total hiparthroplasty. Orthopedics 2008;31:1085.

16. Bhan S, Eachempati KK, Malhotra R. Primary cementless total hiparthroplasty for bony ankylosis with ankylosing spondylitis. J Arthroplasty2008;23:859-66.

17. Berger RA. Total hip arthroplasty using the minimally invasive two-incision approach. Clin Orthop Relat Res 2003; (417):232-41.

18. Kim YL, Shin SI, Nam KW, et al. Total hip arthroplasty for bilater-ally ankylosed hips. J Arthroplasty 2007;22: 1037-1041.

19. Abdel-Aal A, Bakr H, Mahran M. Total hip arthroplasty for fusedhips. Orthopedics 2010;33:400-4.

20. Rutz E, Schäfer D, Valderrabano V. Total hip arthroplasty after hipjoint ankylosis. J Orthop Sci 2009;14:727-31.

21. Hamadouche M, Kerboull L, Meunier A, et al. Total hip arthroplastyfor the treatment of ankylosed hips: a five to twenty-one-year follow-up study. J Bone Joint Surg Am 2001;83-A:992-8.

22. Rajaratnam SS, Sexton SA, Waters TS, et al. Long term results ofcementless total hip replacement for reversal of hip ankylosis. Hip Int2009; 19:120-7.

23. Bangjian H, Peijian T, Ju L. Bilateral synchronous total hip arthro-plasty for ankylosed hips. Int Orthop 2011 Jul. 13 [Epub ahead of print].

RECHERCHE


How you can get involved in the CMA!The CMA is committed to providing leadership for physicians and promoting the highest standard of health and health care forCanadians. To strengthen the association and be truly representative of all Canadian physicians the CMA needs to hear from mem-bers interested in serving in elected positions and on appointed committees and advisory groups. The CMA structure comprises both governing bodies and advisory bodies either elected by General Council or appointed by the CMA Board of Directors. The Board ofDirectors — elected by General Council — has provincial/territorial, resident and student representation, is responsible for the over-all operation of the CMA and reports to General Council on issues of governance.

CMA committees advise the Board of Directors and make recommendations on specific issues of concern to physicians and the pub-lic. Five core committees mainly consist of regional, resident and student representation while other statutory and special commit-tees and task forces consist of individuals with interest and expertise in subject-specific fields. Positions on one or more of thesecommittees may become available in the coming year.

For further information on how you can get involved please go to http://www.cma.ca/membercentre/how-you-can-get-involved, orcontact

Cherise AraujoCorporate and Governance Services

Canadian Medical Association1867 Alta Vista Drive, Ottawa ON K1G 5W8Fax 613 526-7570, Tel 800 663-7336 x1949

[email protected]

By getting involved, you will have an opportunity to make a difference.

We hope to hear from you!




Clinical outcomes compared between laparoscopicand open appendectomy in pregnant women

Jun Chul Chung, MD, PhDGyu Seok Cho, MDEung Jin Shin, MDHyung Chul Kim, MD Ok Pyung Song, MD

From the Department of Surgery,Soonchunhyang University College ofMedicine, Soonchunhyang UniversityBucheon Hospital, Bucheon, Korea

Correspondence to: J.C. ChungDepartment of Surgery Soonchunhyang University Bucheon Hospital1174 Jung-dong, Wonmi-gu, Bucheon-si,Gyeonggi-do, South Korea [email protected]

DOI: 10.1503/cjs.022112

Background: Despite the initial absolute or relative contraindication of laparoscopicsurgery during pregnancy, in the last decade, laparoscopic appendectomy (LA) hasbeen performed in pregnant women. But few studies compare the outcomes of LAcompared with open appendectomy (OA). We investigated clinical outcomes to evalu-ate the safety and efficacy of LA compared with OA in pregnant women.

Methods: We recruited consecutive pregnant patients with a diagnosis of acuteappendicitis who were undergoing LA or OA between May 2007 and August 2011into the study.

Results: Sixty-one patients (22 LA and 39 OA) enrolled in our study. There were nosignificant differences in duration of surgery, postoperative complication rate andobstetric and fetal outcomes, including incidence of preterm labour, delivery type,gestation age at delivery, birth weight and APGAR scores between the 2 groups. How-ever, the LA group had shorter time to first flatus (2.4 ± 0.4 d v. 4.0 ± 1.7 d, p = 0.034),earlier time to oral intake (2.3 ± 1.6 d v. 4.1 ± 1.9 d, p = 0.023) and shorter postopera-tive hospital stay (4.2 ± 2.9 d v. 6.9 ± 3.7 d, p = 0.043) than the OA group.

Conclusion: Laparoscopic appendectomy is a clinically safe and effective procedurein all trimesters of pregnancy and should be considered as a standard treatment alter-native to OA. Further evaluation including prospective randomized clinical trials com-paring LA with OA are needed to confirm our results.

Contexte : Malgré une contre-indication initiale absolue ou relative de l’opérationlaparoscopique durant la grossesse au cours de la dernière décennie, on a eu recours àl’appendicectomie laparoscopique (AL) chez des femmes enceintes. Mais peu d’étudesont comparé les résultats de l’AL à ceux de l’appendicectomie ouverte (AO). Nousavons analysé les résultats cliniques pour comparer l’innocuité et l’efficacité de l’AL àcelles de l’AO chez les femmes enceintes.

Méthodes : Pour la présente étude, nous avons recruté des patientes enceintes con-sécutives porteuses d’un diagnostic d’appendicite aiguë qui ont dû subir une AL ouune AO entre mai 2007 et août 2011.

Résultats : Soixante-et-une patientes (22 soumises à l’AL et 39 à l’AO) ont étéinscrites à notre étude. Nous n’avons noté aucune différence significative pour ce quiest de la durée de l’intervention chirurgicale, du taux de complications postopératoireset des résultats obstétricaux et fœtaux, y compris l’incidence du travail prématuré, letype d’accouchement, l’âge gestationnel et le poids à la naissance et les indices d’APGAR entre les 2 groupes. Toutefois, le groupe soumis à l’AL a moins tardé àprésenter des flatuosités (2,4 ± 0,4 j c. 4,0 ± 1,7 j, p = 0,034), a commencé à s’alimenterpar la bouche plus tôt (2,3 ± 1,6 j c. 4,1 ± 1,9 j, p = 0,023) et a connu un séjour hospi-talier postopératoire plus bref (4,2 ± 2,9 j c. 6,9 ± 3,7 j, p = 0,043) comparativement augroupe soumis à l’AO.

Conclusion : L’appendicectomie laparoscopique est une intervention sécuritaire etefficace au plan clinique durant les 3 trimestres de la grossesse, et il faut l’envisagercomme solution de rechange thérapeutique standard à l’AO. Il faudra une évaluationplus approfondie pour confirmer nos résultats, y compris au moyen d’essais cliniquesrandomisés prospectifs comparant l’AL à l’AO.

A cute appendicitis is the most common nonobstetric condition requiringemergency surgery during pregnancy, with an estimated incidencebetween 0.05% and 0.13%.1 Although the incidence of acute appen-

dicitis during pregnancy is similar to that in nonpregnant women, the rate of

clinical-chung_Layout 1 13-09-16 10:41 AM Page 341


RECHERCHE

complicated appendicitis is much higher in pregnantwomen.2 Delay in diagnosis increases the risk of complica-tions in the mother and fetus. When acute appendicitis issuspected, an aggressive approach is recommended.3 Innonpregnant patients, the advantages of laparoscopicappendectomy (LA), including reduced postoperative pain,fewer wound infections, earlier hospital discharge and fasterreturn to work, are widely accepted.4 Despite the initialabsolute or relative contraindication of laparoscopic surgicalprocedures during pregnancy, in the last decade LA hasbeen performed in pregnant women.5 Although there havebeen no prospective randomized controlled trials studyingLA in pregnant women, several reports have documentedthe feasibility, safety and effectiveness of LA in this popula-tion.1,2,4–9 However, these reports include case reports andsmall series, so their results are limited. In particular, thereis very limited information on the outcomes of LA com-pared with open appendectomy (OA). Therefore, we inves-tigated clinical outcomes to evaluate the safety and efficacyof LA compared with OA in pregnant women.

METHODS

We recruited consecutive pregnant women with a diagnosisof acute appendicitis who underwent LA (study cohort) orOA (control cohort) between May 2007 and August 2011.Ultrasounds and complete blood counts (CBC) were rou-tinely obtained preoperatively. We confirmed the diagnosisof acute appendicitis by clinical examination and ultra-sonography. All patients received preoperative and postop-erative obstetric consultations and fetal monitoring. We ret-rospectively analyzed the medical records to compare the2 groups. Preoperative clinical data included age, body massindex (BMI), gestation age at operation, incidence of previ-ous Caesarean section and accuracy of the diagnostic ultra-sound. Perioperative data included the duration of surgery,return to normal bowel movement, return to adequate oralintake, length of stay in hospital (LOS), postoperative com-plications and final pathology. Obstetric and fetal dataincluded gestation age at delivery, incidence of pretermlabour, delivery type, birth weight, APGAR scores at 1 min -ute and 5 minutes, and fetal mortality.

Laparoscopic appendectomy technique in pregnantwomen

The patient was placed on the table in the supine positionwith a slight left side tilt (20°–30°). We performed the pro-cedure with the patient under general anesthesia and main-tained continuous end-tidal CO2 monitoring within thephysiologic range (30–40 mm Hg). Routinely, we inserted aFoley catheter, used pneumatic compression devices on thelegs and administered prophylactic antibiotics intravenouslybefore the incision. In all patients, we entered the peritonealcavity using the Hasson open technique, and the procedure

was performed using 3 ports. First, we inserted a 10 mmballoon trocar (telescope route) supraumbilically accordingto the size of the uterus (3–4 cm above the uterine fundus),and pneumoperitoneum (10–12 mm Hg) was achieved byCO2 insufflation. Subsequently, 2 working 5 mm trocarswere inserted, depending on the gestation age (Fig. 1). Theappendix was then elevated using a left-handed forceps, andthe mesoappendix was divided using a harmonic scalpel(Ethicon). The appendiceal stump was ligated using endo-loop (Ethicon) and transected using endo-scissors. Thespecimen was then placed in a Lap Bag® (SJM) andremoved through the umbilical port site. One closed suc-tion drain was left next to the appendiceal stump.


Data are presented as means ± standard deviations. Wecompared the groups using the Mann–Whitney U test orχ2 test, as appropriate. We used SPSS version 14.0 forWindows for all statistical comparisons, and we consid-ered results to be significant at p < 0.05.

RESULTS

Participants

Sixty-one patients enrolled in our study. In all 22 patientswho had LA, the procedure was completed laparoscopically

5 mm

10 mm

5 mm

Fig. 1.Trocar placement for laparoscopic appendectomy in differ-ent stages of pregnancy.



RESEARCH

without conversion, and 39 OAs were performed inpatients who refused LA. The mean age of the LA groupwas 29.3 ± 3.1 years, and that of the OA group was 31.4 ±4.3 years (p = 0.45). There were no significant differencesin BMI (22.7 ± 2.8 v. 22.8 ± 3.6, p = 0.83) or the gestationage at operation (16.4 ± 5.7 wk v. 16.7 ± 4.8 wk, p = 0.80)between the groups. In the LA group, 6 (27.3%) patientswere in the first trimester, 13 (59.1%) were in the secondtrimester and 3 (13.6%) were in the third trimester. Inthe OA group, 8 (20.5%) patients were in the first tri -mester, 20 (51.3%) were in the second trimester and 11(28.2%) were in the third trimester. The rate of previous Caesarean section in the LA group was 18.2%, and thatof the OA group was 20.5% (p = 0.64). In all patients, weobtained an ultrasound preoperatively to clarify the diag-nosis of acute appendicitis. There were no significant differences in the rate of false positive (9.1 v. 10.3%,

p = 0.65) and false negative results (22.7 v. 17.9%, p = 0.16)between the groups (Table 1). Table 2 demonstrates thepostoperative histopathological diagnoses in both groups.In the LA group, 2 (9.1%) patients had a normal appen-dix, 12 (54.6%) had focal appendicitis, 7 (31.8%) had sup-parative appendicitis and 1 (4.5%) had gangrenousappendicitis. In the OA group, 4 (10.3%) patients had anormal appendix, 19 (48.7%) had focal appendicitis, 10(25.6%) had sup parative appendicitis and 6 (15.4%) hadgangrenous appendicitis.

Perioperative outcomes

The mean duration of surgery in the LA group was 44.2 ±16.4 minutes, and that in the OA group was 47.3 ±14.7 minutes. The mean time to normal bowel movementin the LA group was significantly shorter than that in the

Table 1. Demographic and clinical characteristics of pregnant women undergoing laparoscopic oropen appendectomy

Group; mean ± SD or no. (%)

Characteristic LA, n = 22 OA, n = 39 p value

Age, yr 29.3 ± 3.1 31.4 ± 4.3 0.45

BMI 22.7 ± 2.8 22.8 ± 3.6 0.83

Gestation age at operation, wk 16.4 ± 5.7 16.7 ± 4.8 0.80

First trimester 6 (27.3) 8 (20.5)

Second trimester 13 (59.1) 20 (51.3)

Third trimester 3 (13.6) 11 (28.2)

Previous Caesarean section 4 (18.2) 8 (20.5) 0.64

Preoperative ultrasonography 22 (100) 39 (100) —

False-positive ultrasound 2 (9.1) 4 (10.3) 0.65

False-negative ultrasound 5 (22.7) 7 (17.9) 0.16

BMI = body mass index; LA = laparoscopic appendectomy; OA = open appendectomy; SD = standard deviation.

Table 2. Final histopathological diagnosis after operation

Group; no. (%)

Stage; diagnosis LA, n = 22 OA, n = 39

First trimester 6 (27.3) 8 (20.5)

Normal appendix 1 1

Focal appendicitis 3 4

Supparative appendicitis 2 2

Gangrenous appendicitis 0 1

Second trimester 13 (59.1) 20 (51.3)

Normal appendix 1 2




Third trimester 3 (13.6) 11 (28.2)

Normal appendix 0 1




LA = laparoscopic appendectomy; OA = open appendectomy.



RECHERCHE

OA group (2.4 ± 0.4 d v. 4.0 ± 1.7 d, p = 0.034). Also, themean time to adequate oral intake in the LA group wasearlier than in the OA group (2.3 ± 1.6 d v. 4.1 ± 1.9 d, p =0.023). The mean LOS in the LA group was 4.2 ± 2.9 days,and that of the OA group was 6.9 ± 3.7 days (p = 0.043).Three patients experienced complications, including intra-abdominal abscess and wound infection. In the LA group,an intra-abdominal abscess developed in 1 (4.5%) patientafter surgery. This 29-year-old woman with a gestation ageof 18 weeks was treated successfully with antibiotics anddelivered a healthy boy weighing 3720 g at a gestation ageof 40 weeks. In the OA group, 2 (5.1%) patients experi-enced complications (intra-abdominal abscess, woundinfection) after surgery. The intra- abdominal abscess de -veloped in a 33-year-old woman at 28 weeks’ gestation.She was treated successfully with percutaneous drainageand delivered a healthy girl weighing 2670 g at a gestationage of 37 weeks. Wound infection was treated conserva-tively. No statistical difference was observed in the rate ofthe complication between the groups (Table 3).

Obstetric and fetal outcomes

No patients were lost to follow-up, and they all haduncomplicated deliveries. There were no significant differ-ences in the incidence of preterm labour (9.1 v. 10.3%, p =0.66) or delivery type (p = 0.41) between the groups. The

LA and OA groups had equivalent fetal outcomes, asdemonstrated by gestational age at delivery, birth weightand APGAR scores (Table 4).

DISCUSSION

Although acute appendicitis is the most common cause ofnonobstetric abdominal surgery during pregnancy and itsincidence is similar to that in nonpregnant women, thediagnosis is difficult because of the anatomic and physio-logic changes that occur during pregnancy.5 The risk forappendicitis does not appear to be increased by pregnancy,but the incidence of perforated appendicitis in pregnantwomen is much higher than in the general population.2

The reported rate of appendiceal perforation during preg-nancy can be as high as 43%, compared with 19% in thegeneral population.1 Complicated appendicitis can lead tomaternal and fetal morbidity and even fetal loss, so preg-nant women with suspected appendicitis should undergosurgery immediately, regardless of the gestation age of thefetus.3,10,11 The number of negative laparoscopic and openexploration rates during pregnancy ranges from 0% to50% and 15% to 50%, respectively.1 In our study, theoverall negative appendectomy rate was 9.8% (9.1% forthe LA group and 10.3% for the OA group).

Diagnostic imaging studies are often used to clarify aconfusing clinical picture. Ultrasonography is widely used

Table 3. Perioperative outcomes of laparoscopic or open appendectomy in pregnant women

Group; mean ± SD*

Outcome LA, n = 22 OA, n = 39 p value

Operation time, min 44.2 ± 16.4 47.3 ± 14.7 0.48

Time to !rst "atus, d 2.4 ± 0.4 4.0 ± 1.7 0.034

Time to oral intake, d 2.3 ± 1.6 4.1 ± 1.9 0.023

Postoperative LOS, d 4.2 ± 2.9 6.9 ± 3.7 0.043

Complication rate, n (%) 1 (4.5) 2 (5.1) 0.76

LA = laparoscopic appendectomy; LOS = length of stay in hospital; OA = open appendectomy; SD = standard deviation. *Unless otherwise indicated.

Table 4. Obstetric and fetal outcomes of laparoscopic or open appendectomy in pregnant women

Group; mean ± SD or no. (%)

Outcome LA, n = 22 OA, n = 39 p value

Gestation age at delivery, wk 37.1 ± 1.7 38.2 ± 3.5 0.48

Preterm labour 2 (9.1) 4 (10.3) 0.66

Delivery type 0.41

Vaginal delivery 16 (72.7) 25 (64.1)

Caesarean section 6 (27.3) 14 (35.9)

Birth weight, g 2810 ± 293 2790 ± 312 0.79

APGAR score, 1 min 9.2 ± 0.1 9.3 ± 0.2 0.70

APGAR score, 5 min 9.8 ± 0.2 9.9 ± 0.1 0.73

Fetal loss 0 (0) 0 (0) —

LA = laparoscopic appendectomy; OA = open appendectomy; SD = standard deviation.



RESEARCH

as a first-line diagnostic test because of its safety for themother and fetus and its relatively high sensitivity andspecificity for many intra-abdominal processes.12 In ourstudy, ultrasonography was performed in all patients; acuteappendicitis was found in 15 (68.2%) patients in the LAgroup and 28 (71.8%) in the OA group.

Conventionally, the treatment of choice for acuteappendicitis during pregnancy has been OA. But there isno evidence that the benefits of OA outweigh those of LAin pregnant women with respect to perioperative morbid-ity and mortality.13 In the present study, the absence of sig-nificant differences in the analysis of clinical details sug-gests that our 2 groups had relatively similar preoperativeconditions. Our study also supports the safety of LAbecause we found no differences in perioperative morbidityand mortality compared with OA. Moreover, some provenadvantages of LA, including better intraoperative visualiza-tion, decreased surgical trauma, decreased gravid uterinemanipulation, shorter postoperative LOS and faster returnto work, may be even more important in pregnant women.7

In our study, the LA group had significantly shorter post-operative LOS, earlier recovery of bowel function andshorter time to oral intake than the OA group.

It has been recommended to position the patient on herleft side during surgery to prevent uterine compression of theinferior vena cava and to facilitate access to the appendix.6

Morrell and colleagues14 have suggested lateral rotation ofthe operating table to displace the uterus for better venousreturn. In our hospital, all pregnant patients were placed in asupine position with a slight left side tilt (20°–30°).

There is consensus that laparoscopic procedures aresafest in the second trimester of pregnancy because theuterus, owing to its small size, is less susceptible to trau-matic injuries.1 Some authors have suggested that laparo-scopic procedures performed during the first trimester areusually associated with greater risk for fetal loss because ofteratogenicity of medications and decreased uterine blooddue to the pneumoperitoneum.15 An advanced gestationage was previously regarded as a contraindication for LAbecause of expected technical difficulties.6 However, Upadhyay and colleagues10 demonstrated that laparoscopicsurgery in the third trimester of pregnancy is feasible andcan be performed safely, and they recommended laparo-scopic surgery in all 3 trimesters. In our study, LA was per-formed safely in all 3 trimesters without fetal mortality.

One of the most important concerns during LA in preg-nancy is the potential risk of injury to the gravid uterus.Some debate exists about the best method to access theabdomen. The Veress needle or the Hasson open tech-nique can be used to gain initial abdominal access. Eventhough complications have been described for all methods,spontaneous puncture of the uterus with a Veress needle isthe most serious.16 Friedman and colleagues17 reportedresults in a young pregnant woman at 21 weeks’ gestationwho underwent LA for suspected appendicitis. A Veress

needle injury to the serosa of the gravid uterus resulted inpostoperative pneumoamnion and subsequent fetal loss. Inour hospital, the peritoneal cavity was entered using theHasson open technique, and there was no injury to thegravid uterus. We consider the Hasson open technique tobe completely safe and suggest that it should be the stan-dard method in pregnant women, in accordance with theSociety of American Gastrointestinal Endoscopic Surgeons(SAGES) guidelines.12

Achieving pneumoperitoneum during laparoscopy isanother concern. The CO2 used for pneumoperitoneum isassociated with pulmonary effects in pregnant women anda potential risk for acidosis in the fetus. It has been recom-mended that intra-abdominal pressure should be main-tained at less than 12 mm Hg to avoid worsening pul-monary physiology in pregnant women.18 Previouslypublished animal studies reported no adverse fetal effectsof CO2 insufflation when the maximal intra-abdominalpressure was limited to 10–12 mm Hg for less than60 minutes.13 Although studies have demonstrated thatlaparoscopic surgery can be performed safely during anytrimester with good maternal and fetal outcomes, the long-term effects on the child after delivery have not been wellstudied.16 In the present study, intra-abdominal pressurewas maintained at 10–12 mm Hg and the duration ofsurgery was less than 60 minutes (mean duration of LA44.2 ± 16.4 min).

Stasis of blood in the lower extremities is common dur-ing pregnancy, so pregnant women are at high risk forthromboembolitic complications. According to the SAGESguidelines, intraoperative and postoperative pneumaticcompression devices and early postoperative ambulationare recommended to prevent deep vein thrombosis inpregnant patients.12 In the present study, we routinely usedpneumatic compression devices, and there were no throm-boembolitic complications.

CONCLUSION

Our results show that LA is safe and effective in alltrimesters and that it is associated with good maternal andfetal outcomes similar to those of OA. In addition, LA isassociated with shorter postoperative LOS, earlier recov-ery of bowel function and shorter time to oral intake. Onthe basis of our results and considering the general advan-tages of laparoscopic surgery, LA in pregnant womenshould be considered as a standard treatment alternativeto OA. Further evaluation, including prospective random-ized clinical trials comparing LA with OA, are needed toconfirm these results.


Contributors: J.C. Chung designed the study and wrote the article.G.S. Cho and H.C. Kim acquired the data, which J.C. Chung, E.J. Shinand O.P. Song analyzed. All authors reviewed the article and approved itfor publication.



RECHERCHE

References

1. Kirshtein B, Perry ZH, Avinoach E, et al. Safety of laparoscopicappendectomy during pregnancy. World J Surg 2009;33:475-80.

2. Corneille MG, Gallup TM, Bening T, et al. The use of laparoscopicsurgery in pregnancy: evaluation of safety and efficacy. Am J Surg2010; 200:363-7.

3. McGory ML, Zingmond DS, Tillou A, et al. Negative appendec-tomy in pregnant women is associated with a substantial risk of fetalloss. J Am Coll Surg 2007;205:534-40.

4. de Bakker JK, Dijksman LM, Donkervoort SC. Safety and outcomeof general open and laparoscopic procedures during pregnancy. SurgEndosc 2011;25:1574-8.

5. Moreno-Sanz C, Pascual-Pedreño A, Picazo-Yeste JS, et al.Laparoscopic appendectomy during pregnancy: between personalexperiences and scientific evidence. J Am Coll Surg 2007;205:37-42.

6. Curet MJ, Allen D, Josloff RK, et al. Laparoscopy during pregnancy.Arch Surg 1996;131:546-50.

7. Lyass S, Pikarsky A, Eisenberg VH, et al. Is laparoscopic appendec-tomy safe in pregnant women? Surg Endosc 2001;15:377-9.

8. Sadot E, Telem DA, Arora M, et al. Laparoscopy: a safe approach toappendicitis during pregnancy. Surg Endosc 2010;24:383-9.

9. Hannan MJ, Hoque MM, Begum LN. Laparoscopic appendectomy inpregnant women: experience in Chittagong, Bangladesh. World J Surg2012;36:767-70.

10. Upadhyay A, Stanten S, Kazantsev G, et al. Laparoscopic manage-ment of a nonobstetric emergency in the third trimester of preg-nancy. Surg Endosc 2007;21:1344-8.

11. Carver TW, Antevil J, Egan JC, et al. Appendectomy during earlypregnancy: What is the preferred surgical approach? Am Surg2005;71: 809-12.

12. Yumi H. Guidelines for diagnosis, treatment, and use of laparoscopyfor surgical problems during pregnancy: this statement was reviewedand approved by the Board of Governors of the Society of AmericanGastrointestinal and Endoscopic Surgeons (SAGES), September2007. It was prepared by the SAGES Guidelines Committee. SurgEndosc 2008;22:849-61.

13. Park SH, Park MI, Choi JS, et al. Laparoscopic appendectomy per-formed during pregnancy by gynecological laparoscopists. Eur J ObstetGynecol Reprod Biol 2010;148:44-8.

14. Morrell DG, Mullins JR, Harrison PB. Laparoscopic cholecystec-tomy during pregnancy in symptomatic patients. Surgery 1992;112:856-9.

15. Cohen-Kerem R, Railton C, Oren D, et al. Pregnancy outcome follow-ing non-obstetric surgical intervention. Am J Surg 2005;190:467-73.

16. Machado NO, Grant CS. Laparoscopic appendectomy in all trimestersof pregnancy. JSLS 2009;13:384-90.

17. Friedman JD, Ramsey PS, Ramin KD, et al. Pneumoamnion andpregnancy loss after second-trimester laparoscopic surgery. ObstetGynecol 2002;99:512-3.

18. Malangoni MA. Gastrointestinal surgery and pregnancy. GastroenterolClin North Am 2003;32:181-200.

Comment vous pouvez vous impliquer dans l’AMC !L’AMC est vouée à jouer un rôle de chef de file auprès des médecins et à promouvoir les normes les plus élevées de santé et de soinsde santé pour les Canadiens. Afin de renforcer l’Association et pour qu’elle représente véritablement tous les médecins du Canada,l’AMC a besoin de membres intéressés à occuper des charges élues et à siéger à des comités et des groupes consultatifs. La structurede l’AMC se compose d’organes de régie et d’entités consultatives élus par le Conseil général ou nommés par le Conseil d'admini -stration. Le Conseil d’administration, dont les membres sont élus par le Conseil général et représentent les associations médicalesprovinciales et territoriales, les résidents et les étudiants en médecine, est chargé de l’administration générale de l’AMC. Il rendcompte des questions de régie au Conseil général.

Les comités de l’AMC jouent le rôle de conseillers auprès du Conseil d’administration et présentent des recommandations au sujetde questions particulières intéressant les médecins et la population. Cinq comités principaux sont constitués principalement dereprésentants des régions, des résidents et des étudiants, tandis que les autres comités statutaires et spéciaux et les groupes de travailréunissent des personnes qui s’intéressent à des sujets précis et possèdent des compétences spécialisées. Des postes pourront devenirvacants dans un ou plusieurs de ces comités en cours d'année.

Pour obtenir plus d’information au sujet des façons de participer, veuillez consulter http://www.cma.ca/centredesmembres/comment-vous-impliquer-dans-lamc ou communiquer avec

Cherise AraujoCorporate and Governance Services

Canadian Medical Association1867 Alta Vista Drive, Ottawa ON K1G 5W8Fax 613 526-7570, Tel 800 663-7336 x1949

[email protected]

Votre participation peut faire la différence.

Nous espérons avoir de vos nouvelles !


© 2013 Canadian Medical Association Can J Surg, Vol. 56, No. 5, October 2013 E105


Adding an endovascular aortic surgery program toa rural regional medical centre

Background: Abdominal aortic aneurysms requiring surgical intervention are gener-ally treated by endovascular means. Such procedures are not always offered in ruralhospitals, possibly leaving patients underserved. We reviewed our experience initiatingan endoaortic surgery program.

Methods: A surgeon in a rural centre was credentialed to perform endovascular aor-tic aneurysm repair through collaboration with a university centre and was proctoredlocally for the first 5 abdominal aneurysm repairs. Web-based image storage was usedto review complex cases as part of an ongoing partnership. Referred patients werescreened for multiple aneurysms and underwent long-term monitoring.

Results: In all, 160 patients were evaluated for 176 aortic pathologies. Twenty-fivepatients (17 men) aged 55–89 years underwent 26 endovascular abdominal (n = 23) orthoracic (n = 3) aortic procedures. Emergent endovascular procedures were not per-formed. There were no operative deaths, requirements for dialysis or conversions toopen repair. Two endoleaks required early reintervention. The median length of stayin hospital for endovascular procedures was 2.5 days. Chronic endoleaks wereobserved in 7 patients. An additional 8 patients underwent open abdominal aneurysmrepair locally and 15 patients were referred to the university program.

Conclusion: Creation of an endovascular aortic surgery program in a rural hospital isfeasible through collaboration with a high-volume centre. Patient safety is enhanced byobtaining second opinions using web-based image review. Most interventions are forabdominal aortic aneurysms, but planning for a comprehensive aortic clinic is preferable.

Contexte : Les anévrismes de l’aorte abdominale justiciables d’une interventionchirurgicale sont généralement traités par voie endovasculaire. Ce type d’interventionn’est toutefois pas toujours pratiqué dans les hôpitaux ruraux. Les patients peuventdonc s’en trouver moins bien desservis. Nous avons passé en revu notre expérienceaprès la mise sur pied d’un programme de chirurgie endoaortique.

Méthodes : Grâce à une collaboration avec un centre universitaire, un chirurgiend’un centre rural a reçu l’agrément nécessaire pour effectuer la réparation endovascu-laire des anévrismes de l’aorte et il a été supervisé localement pour les 5 premièresréparations d’anévrisme de l’aorte abdominale. Une banque d’images sur le Web apermis de passer en revue des cas complexes dans le cadre d’un partenariat continu.On a fait subir aux patients adressés en consultation un dépistage d’anévrismes multi-ples et ils ont fait l’objet d’un suivi à long terme.

Résultats : En tout, 160 patients ont été examinés pour 176 anomalies aortiques.Vingt-cinq patients (17 hommes) âgés de 55 à 89 ans ont subi 26 interventionsendovasculaires de l’aorte abdominale (n = 23) ou thoracique (n = 3). Aucune autreintervention endovasculaire n’a été effectuée. On n’a eu à déplorer aucun décès en lienavec les interventions, aucun recours à la dialyse ni conversion vers une chirurgieouverte. Deux endofuites ont nécessité une réintervention précoce. La durée médianedu séjour hospitalier dans les cas d’intervention endovasculaire a été de 2,5 jours. Desendofuites chroniques ont été observées chez 7 patients. Huit autres patients ont subiune réparation ouverte d’anévrisme abdominal localement et 15 patients ont étéréférés au programme universitaire.

Conclusion : La création d’un programme de chirurgie de l’aorte endovasculairedans un hôpital rural est réalisable grâce à une collaboration avec un centre dont levolume d’interventions est élevé. La sécurité des patients est renforcée par l’obtentionde secondes opinions facilitées par une banque d’images sur le Web. La plupart desinterventions concernent des anévrismes de l’aorte abdominale, mais il est préférablede planifier la mise en place d’une clinique où on pourrait intervenir sur toutes lesportions de l’aorte.

Daniel Marelli, MD*

Edward Woo, MD†

Michael Watson, BSc*

Paul Fedalen, MD*

Grace Wang, MD†

Megan Stallings, PA-C*

Ronald Fairman, MD†

John Mannion, MD*

From the *Division of Cardiac Surgery,Bayhealth Medical Center, Dover, Del.,and the †Division of Vascular Surgery,University of Pennsylvania, Philadelphia,Penn.


Correspondence to:D. MarelliBayhealth CV Surgery Associates540 South Governors Ave., Suite 101ADover DE [email protected]

DOI: 10.1503/cjs.017912

E106 J can chir, Vol. 56, No 5, octobre 2013

RECHERCHE / RESEARCH

O pen aortic aneurysm repairs can have high compli-cation rates because many patients undergoingthis type of surgery have comorbidities, mostly

related to advanced age.1,2 Since Parodi and colleagues3

reported their experience using intraluminal graft implantsto treat aortic aneurysms, less invasive endovascular aorticaneurysm repair (EVAR) of abdominal aortic aneurysms(AAA) has become applicable in up to 75% of cases requir-ing intervention.4–6 Such technology is also applicable,albeit to a lesser extent, for thoracic endovascular aorticaneurysm repair (TEVAR). Despite questions about long-term benefit and potentially higher up-front costs, theseprocedures are more appealing because they have signifi-cantly lower 30-day morbidity and mortality.7–11

Adding an endovascular aortic repair program to anexisting surgical product line may not be possible becausethere could be too few vascular cases to attract a vascularsurgeon.12,13 In such instances, local general surgeons oftenhave provided open vascular surgery services, includingabdominal aortic aneurysm repair. These vascular servicesmay be complemented by interventional radiology and cardiology for catheter-based treatment of iliac and lowerextremity arterial disease. Alternatively, a rural hospital maytry to attract a dually specialized surgeon who has qual i fiedin both general and vascular surgery. Such surgeons, how-ever, frequently wish to limit their practices to vascular sur gery. The majority of general surgeons do not receivesuf ficient vascular training to provide endovascular aortican eurysm repairs.

Under these circumstances, rural physicians must referpatients requiring treatment for aortic aneurysm disease todistant hospitals.13–15 Unless a local physician or surgeon ismotivated to regularly follow such patients before and afterintervention, there is a risk of creating an underserved popu -lation. An appropriately qualified surgeon at a rural med-ical centre would allow the hospital to manage these pa -tients comprehensively. In this paper we describe how aregional hospital serving a population of 250 000–500 000,Bayhealth Medical Center (BMC), implemented such anaortic clinic and stent graft surgery program. A partnershipwith a distant university medical centre, the Hospital of the

University of Pennsylvania (HUP), was initiated to facili-tate the introduction of endovascular aortic procedures.

METHODS

An accredited institutional review board approved this ret-rospective study and waived the requirement for patientconsent.

Endovascular working group and timeline

Before creation of the endovascular surgery program, itwas necessary to establish a hospital working group. Anadministrator oversaw representatives from surgery, anes-thesia, clinical engineering, the cardiovascular operatingroom and diagnostic imaging. The group was tasked withthe generation of a timeline, credentialing guidelines andcontracting with a high-volume vascular surgery program.Credentialing guidelines were developed with the guidanceof the director of the local cardiac catheterization labora-tory. An 8-month timeline laid out monthly goals for theendovascular working group. It included the credential-ing process, the purchase of a new C-arm fluoroscopymachine, training the operating room (OR) nursing staffand ordering necessary endovascular inventory items. Theinventory list was compiled by the surgeon being creden-tialed (D.M.). Our timeline is presented in the Appendix(available at cms.ca/cjs). The working group met every2 weeks and reported its progress toward starting theendovascular service.

Joint venture with high-volume vascular surgeryservice

The BMC is in a county that is code 3 on the U.S.Department of Agriculture’s census-based Rural–UrbanContinuum Codes (RUCC) and serves a small metro areawith a population of about 250 000; BMC also partiallyserves an adjacent county that is code 4. The HUP is in acounty that is code 1 on the RUCC system. It is in alarge, metro area with a population of about 6 million and

Table 1. Procedures completed by D.M. during training with HUP surgeons

Procedure Assist at HUP

Operator at HUP

Operator at BMC Total procedures Required total

EVAR procedures 9 2 5* 9 assist 7 operator

5 assist 5 operator

Open AAA 4 4 2 suggested

Diagnosic aortoiliac procedures 41 20 10* 41 assist 30 operator

15 assist 15 operator

Interventional aortoiliac procedures 20 9 10* 20 assist 19 operator

15 assist 15 operator

TEVAR procedures 6 6 10

AAA = abdominal aortic aneurysm; BMC = Bayhealth Medical Center; EVAR = endovascular aortic aneurysm repair; HUP = Hospital of the University of Pennsylvania; TEVAR = thoracic endovascular aortic aneurysm repair. *Cases completed during the !rst 5 proctored cases at BMC.

Can J Surg, Vol. 56, No. 5, October 2013 E107

RESEARCH

is located about 80 miles from BMC.16 The HUP hasmade a major commitment to heart and vascular care.The partnership between BMC and HUP increases theiroutreach and facilitates the referral of BMC patients withcomplex aortic pathology to be treated at HUP. Thesepatients can then be treated by specialized surgeons withstate-of-the-art equipment that is not available at BMC.An example of this would be a hybrid operating roomsuite. Such a suite can have better imaging equipment andusually has a larger inventory of specialized cathetersneeded for unusual and complex cases. In return forfinancial compensation, BMC may use HUP’s brandidentity in advertisements and community outreach. TheHUP surgeons obtained credentials that allow them toassist or proctor up to 6 cases per year at BMC. TheBMC surgeons may also work in partnership with HUPsurgeons on complex cases by having privileges at bothhospitals. The BMC surgeons can consult rapidly withHUP surgeons on complex cases using web-based im -aging services, deciding whether to treat patients locallyor refer them.

Credentialing guidelines and process

Patient evaluation and selection started 6 months beforethe first expected EVAR procedure during the credential-ing period. During this time, urgent, emergent and morecomplex cases were referred to the university centre forearlier treatment or enrolment into newer graft trials. ABMC cardiac surgeon (D.M.) spent 3 days per week atHUP assisting in or performing endovascular procedures.During the 6-month training period, a list of usual wireand catheters needed for EVAR and TEVAR procedureswas developed so that the working group could make con-tact with vendors.

Because endovascular aortic surgery was a new proced -ure for our hospital, there was no established credentialingpolicy. We therefore developed a credentialing policy forEVAR and TEVAR. The policy included both a didacticand procedural portion. The didactic portion includedreading a relevant textbook and taking an accredited orapproved education program on EVAR procedures. Forthe procedural portion, the surgeon was required to par -tici pate in a specified number of endovascular procedures.The number and type of required procedures for eachmodule are listed in Table 1. These were based on Societyof Vascular Surgeons and American College of Cardiologyguidelines.

The surgeon must act as operator (while proctored by aqualified surgeon) in at least half of the required proced -ures (excluding TEVAR procedures). Each endovascularabdominal aneurysm repair case was counted as 1 EVARprocedure, 2 aortoiliac intervention procedures and 2 aorto -iliac diagnostic procedures. The TEVAR cases countedsimilarly, but without aortoiliac intervention. The surgeon

(D.M.) satisfied most of the procedural credentialingrequirements on cases at HUP then completed theserequirements in the first 5 cases at BMC. These were pre-planned with an HUP surgeon (E.W.) acting as a proctor.This satisfied the standard focused professional practiceevaluation (FPPE) required by the BMC medical staffoffice. A detailed count of completed cases is presented inTable 1. Collaboration with HUP surgeons allowed thesurgeon to complete his training and fulfill his proceduralcredentialing requirements. The TEVAR credentials werenot rigorously met, but the hospital credentialing commit-tee felt the overall experience was sufficient to sign off.

Imaging

Because of the intensive planning required before anendovascular procedure, high-quality computed tomog -raphy (CT) imaging studies are essential to the surgeon.The collaborative nature of the BMC endovascular serviceline necessitated rapid sharing of imaging studies withHUP surgeons. When starting our endovascular serviceline, we considered several imaging programs. One of themost important criteria for our imaging system was that itmust allow for easy and seamless remote image sharing.We chose the PEMS software package (M2S). With thissoftware, BMC computed tomography angiography (CTA)films can be uploaded and processed into 3-dimensional(3D) images upon email request. Viewing software can bedownloaded on any computer, and M2S-generated 3Dimages can be easily downloaded and stored locally forphysician evaluation. M2S acts as an intermediary between

A B C

Fig. 1. Example images from the M2S software package. (A) Apatient who will undergo endovascular aortic aneurysm repair(EVAR) at Bayhealth Medical Center. (B) A high-risk EVAR due tothrombus in the proximal neck. This patient would be referred tothe Hospital of the University of Pennsylvania. Note how atomographic slice can be placed onto the aortic contour, orthog-onal to the bloodpath. (C) Anatomy requiring open repair or anew generation graft adapted to severe neck angulation. Allpatients were screened for multiple aneurysms with chest,abdomen and pelvis imaging.


RECHERCHE

BMC and HUP surgeons, as BMC films can be droppeddirectly into a HUP surgeon’s (E.W.) folder, upon request.Isolated surgeons can then view the same film from theirrespective login points. This avoids having to physically sendand receive films, which can be costly and time-consuming.Another important criterion for the imaging program ishigh-quality image reformatting. The program we chosehas the ability to map thrombus and calcium along blood-paths. This feature facilitates endovascular graft planning, asseen in Figure 1.

Local surgeon partnership

Before starting this program, patients with AAA werereferred by primary care physicians to the local privategeneral surgery groups for open repair or were referred tovarious city hospitals. After instituting this program, manypatients underwent endovascular repair by the BMC(D.M.) or HUP surgeons. The ability of the endovascularprogram to treat simple cases has left the more complexcases for open repair. Partnering with a local general sur-geon became important to treat complicated cases requir-ing open repair. This has helped refine surgical techniquesfor open AAA repair.

Clinical pathway and operative technique

A complete clinical pathway was an integral part of start-ing the aortic service. One midlevel practitioner (M.S.)assisted the BMC surgeon (D.M.) on all aspects of the aortic clinic and surgery. Patients referred to this servicehave an initial consult and are counselled on aorticaneurysms or other pathology for which they werereferred. The patient’s aortic disease is then broadly clas-

sified as sporadic, degenerative, syndromic or familial. Incases of suspected hereditary aortic disease, patients’ fam-ily members are also urged to undergo screening. Allpatients are screened for multiple aneurysms and undergolifelong monitoring before and after surgical treatment.Patients with less severe disease and those who are unfitfor surgery are also followed long-term. Based on eachpatient’s needs and the extent of disease, a course of inter-vention is chosen as either medical treatment with βblockers17,18 and angiotensin converting enzyme inhib -itors19 or surgical treatment, when required. An eurysmsand intramural hematomas distal to the aortic arch wereevaluated with an intention to treat by endovascularmeans as a first choice. In the thoracic aorta, ideal can -didates had ulcers with large intramural hematoma or asaccular aneurysm. In the abdominal aorta, ideal can -didates had an infrarenal aneurysm with an adequateneck. In all cases, we looked for a 20 mm proximal anddistal seal. All borderline cases were reviewed with HUPsurgeons using web-based image sharing. A representativefrom the graft vendor brought in the stent grafts on a percase basis and was present as a resource throughout theoperative procedure.

Prior to surgical treatment, preoperative evaluation forall patients included echocardiography, nuclear medicine–based stress testing and anesthesia consultation. Patientswere admitted on the day of surgery and acetylsalicylicacid and β blockers were continued perioperatively. TheEVAR procedures were carried out with bilateral opencommon femoral artery access. The EVAR planning wasdone preoperatively using 3D image processing to evalu-ate the aneurysm as well as iliac and femoral access size.All patients received general anesthesia and underwentcomparison of upper body and lower extremity arterialline tracing after repair of femoral artery access. TheTEVARS were carried out with unilateral open femoral orcommon iliac artery access and contralateral percutaneousaccess. Depending on the degree of calcification, superfi-cial femoral artery access was occasionally preferred. In allcases, a purse string suture consisting of 4–0 polypropy-lene with a small red tourniquet was used to encircle theentry site (Fig. 2). This minimized blood loss and permit-ted downsizing access when exchanging the delivery sys-tem sheath for a moulding balloon access sheath. Thepurse string suture is removed at the end of the operationand the femoral artery is repaired with interrupted 5–0polypropylene sutures. All cases were completed usingCook Zenith (Cook Medical) or Endurant (Medtronic)graft systems. Postoperatively, all patients went to the car-diovascular surgical care unit, which has the ability to listpatients as intensive, intermediate or surgical care. Pa -tients could be discharged home or to a rehabilitationfacility from the unit without first going to a different hos-pital. Follow-up CTA with 2- and 5-minute delay films wereobtained 4–8 weeks postoperatively and at 6–12 month

Fig. 2. The femoral artery entry site is encircled with a doublepurse string using a single suture. The first layer is in the media,while the second layer is in the overlying adventitia. The entrysite can be controlled by applying gentle traction on the suture.It can be secured with a clamp applied to the tourniquet when itis pushed against the entry site.


RESEARCH

intervals thereafter. Ultrasonography was used selectivelyto monitor aneurysm sac size when appropriate to de -crease the risk of radiation exposure.20,21

Community awareness

To increase awareness of the new endovascular service, weused a variety of outreach programs. These included massmailings, meet and greet lunches with primary care phys -icians and senior citizen education at community centres.Special emphasis was placed on the importance of aorticaneurysm screening for high-risk patients. Patient recruit-ment increased community awareness about the dangersof aortic aneurysm, the benefits of early detection, andpossible options for AAA repair.


Data were collected and analyzed with Microsoft Excel2010. We compared continuous variables using a Student ttest, when appropriate.

RESULTS

In all, 160 patients in the BMC aortic clinic were evalu-ated and followed for a total of 176 aortic pathologiesover 46 months. Of these pathologies, 155 were aneur -ysms. Table 2 shows the anatomic distribution of these155 aneurysms and the average age of the corresponding

Table 2. Number of patients evaluated for aortic aneurysmpathology, n = 155

Pathology No. patients (male) Age, median (range) yr

Thoracic

TAA ascending 55 (42) 66 (47–88)

TAAA descending 8 (3) 67 (47–76)

Type I/II TAAA 2 (1) 66.5 (66–67)

Abdominal

AAA 79 (62) 75 (51–90)

Type III TAAA 6 (2) 65 (60–77)

Type IV TAAA 5 (0) 73 (66–73)

AAA = abdominal aortic aneurysm; TAA = thoracic aortic aneurysm; TAAA = thoracoabdominal aortic aneurysm.

Number of aneurysms evaluated n = 155

Thoracic Ascending TAA,

descending TAA and type I/II TAAA

n = 65

AbdominalInfrarenal AAA, pararenal AAA, type III TAAA and

type IV TAAAn = 90

Refer to HUP n = 3

EVAR at BMC n = 23

Open repair at BMC n = 8

Un!t for surgery n = 12

Refer to HUP n = 12

Continued surveillance n = 35

TEVAR at BMC n = 3

Open repair at BMC n = 29

Un!t for surgery n = 4

Continued surveillance n = 90

Fig. 3. Patients evaluated and treated in the aortic clinic at Bayhealth Medical Center (BMC). AAA = abdominal aortic aneurysm; EVAR = endovascular aortic aneurysm repair; HUP = Hospital of the University of Pennsylvania; TAA = thoracic aortic aneurysm; TAAA = thoracoabdominal aortic aneurysm; TEVAR = thoracic endovascular aortic aneurysm repair.


RECHERCHE

patients. Other pathologies included dissection, intra-mural hematoma and mesenteric ischemia. Figure 3shows the number of patients evaluated for aneurysmsurgery who had open repair, had endovascular repair,were referred to HUP or underwent further surveillance.Of the 90 abdominal aortic pathologies, 23 were treatedwith EVAR and 8 were treated with open repair. Of the65 thoracic aortic pathologies, 3 were treated withTEVAR and 29 were treated with open repair.

Overall, 9.7% of pathologies (15 of 155) were referredto HUP for treatment. Two of these were for persistentendoleaks and growing aneurysm sacs in patients previ-ously treated with EVAR at other hospitals. In total, 49.7%of evaluated aortic pathologies (77 of 155) underwent fur-ther surveillance, with 16 of these patients deemed unsuit-able for EVAR or open repair. Many of these patients hadadvanced frailty or dementia.

A total of 23 EVAR and 3 TEVAR procedures werecarried out at BMC. None was emergent. Iliac arteryaccess was required in 4 cases because of small femoralartery size. One patient with an aortoiliac aneurysmrequired hypogastric artery coiling, which was performedby an interventional cardiologist 1 week before surgery.Table 3 shows the results of endovascular and openabdominal aortic aneurysm repair at BMC during thestudy period. All 8 open abdominal procedures involved aretroperitoneal approach with suprarenal or supraceliacclamping. Three were for the diagnosis of type III (n = 2)or IV (n = 1) thoracoabdominal aortic aneurysms (TAAA).These patients were treated with adjunct distal perfusion,when appropriate. In this study, perioperative morbiditieswere considered separate from endoleaks. The mean sizeof aneurysms operated on for both EVAR and openrepair was comparable (mean 5.6 ± 0.8 cm v. 5.6 ± 0.9 cm;p = 0.97) and there was no significant difference in theaverage age of patients who underwent each procedure(mean 71.7 ± 9.0 yr v. 74.4 ± 7.7 yr, p = 0.46). The lengthof stay in hospital of patients undergoing EVAR was sig-nificantly shorter than that of patients undergoing openrepair (mean 3.0 ± 1.3 d v. 9.8 ± 3.8 d, p = 0.002). Therewere no conversions to open surgery during the studyperiod. Vascular morbidities are summarized in Table 4.Two patients required femorofemoral bypass periopera-tively. In 1 case, this was because of the inability to can-

nulate the contralateral main body gate, causing conver-sion to a uni-iliac stent graft repair. In another, it wasbecause of iliac artery stenosis that was aggravated by thestent graft. Three patients required iliac artery angio-plasty, 2 of which included stent placement. There wereno perioperative deaths in both EVAR and open repairgroups. Instances of perioperative morbidity occurred in4 of 23 EVAR repairs, including a patient with a post -operative ileus, a patient in whom renal dysfunction de -velop ed because pre-existing artery stenosis became oc -cluded due to plaque shift, a patient with graft infectionand a patient in whom a GI bleed developed 2 weekspostoperatively due to clopidogrel being taken for a pre-vious carotid artery stent. The graft infection requiredexcision and extra-anatomical reconstruction after refer-ral to HUP. A pre-existing aortoduodenal fistula was sus-pected. Perioperative morbidity occurred in 2 of 8 openrepairs, including a patient who experienced renal dys-function and another patient who experienced dialysis-dependent renal failure. The latter patient had a para renalabdominal aortic aneurysm. There were no perioperativedeaths or morbidities in the TEVAR group.

Endoleaks were classified as types I, II or III. These rep-resent seal area, retrograde fill and graft overlap area leaks,respectively. Reintervention for endoleak was requiredwithin 30 days in 2 patients. One, a type I, was post-TEVAR and presented with sudden back pain before dis-charge. The other, a type III, was post-EVAR and pre-sented with paroxysmal tachycardia and mild anemiabefore discharge. Both patients underwent uneventful re -intervention with additional stent graft placement. Fivepatients had type II endoleaks. Late type III endoleakswere suspected in another 2 EVAR patients, but these werenot confirmed by brachial accress angiograms, and therehas been no sac growth on follow-up imaging to date. OneTEVAR patient had a late aortic dissection distal to thegraft within 1 year of treatment.

No patient in this series required reintervention forlate endoleaks to date. Type II endoleaks are evaluatedwith serial CTA and managed conservatively as long asthe aneurysm sac size is stable or shrinking. Complextype II leaks that seem progressive or that could be con-fused with type III leaks were reviewed with a universitysurgeon.

Table 3. Results of abdominal aneurysm repair

Procedure Aneurysm size, mean ± SD, cm Age, mean ± SD, yr No. patients (male) LOS, mean ± SD, d

Perioperative morbidity*

Follow-up time, median (range) mo

EVAR, n = 23 5.6 ± 0.8 71.7 ± 9.0 23 (17) 3.0 ± 1.3 4† 6.7 (1.0–22.7)

Open repair, n = 8 5.6 ± 0.9 74.4 ± 7.7 8 (4) 9.8 ± 3.8 2‡ 4.0 (1.7–37.6)

EVAR = endovascular aortic aneurysm repair; LOS = length of stay; SD = standard deviation. *There were no deaths. Endoleaks were considered separately from perioperative morbidities. †The instances of morbidity include a patient with a postoperative ileus, a patient with renal dysfunction due to plaque shift, a patient with graft infection and a patient with a gastrointestinal bleed 2 weeks postoperatively due to clopidogrel being taken for a previous carotid artery stent. ‡One patient had renal dysfunction while another patient had dialysis-dependent renal failure postoperatively.


RESEARCH

DISCUSSION

The BMC created its heart program because of a need toserve a growing regional population in the range of250 000–500 000. Until we initiated this program, abdom -inal aneurysm care was provided by the general surgeons.Patients underwent open repair or were referred forendovascular stent grafting. The local interventional cardi-ologists serve most of the peripheral vascular needs for thecommunity. The local general surgeons treat the remainingpatients who require open surgery for peripheral arterial orvenous pathology, including emergent thromboembolec-tomy and femorofemoral bypass of the lower extremity.There is no local vascular surgeon. The BMC endoaorticprogram was established by a cardiac surgeon. However,the present model could be established by a general orolder vascular surgeon willing to undergo training and cre-dentialing, as described in the Methods section. In addi-tion, the model we describe for a comprehensive aorticclinic can be used by a vascular surgeon wishing to partnerwith a distant university program. Clearly, hospital orhealth system support is required to ensure that incentivesand goals of all parties involved are aligned. The purposeof this article was to describe how a region that was under-served for aortic care could become partially or mostlyserved locally. We noted that more than 80%–90% of thepathology we evaluated was aneurysmal disease.

The majority of patients requiring endovascular treat-ment had infrarenal AAAs. There were no conversions toopen surgery, reflecting careful patient selection. Initially,almost all cases were reviewed with the university surgeonspreoperatively. With time this proportion fell to about 1 in4. Adjunct surgical procedures commonly used werefemoral artery exposure, iliac artery exposure via theretroperitoneum and femorofemoral bypass. Along withcatheter and wire skills, such procedures can be performedby general surgeons willing to undergo a short trainingperiod. The cognitive requirements for clinical evaluationand follow-up are also within the reach of the generalsurgery knowledge base. Most abdominal aortic aneurysmsare currently treated by endovascular repair, which has lessperioperative morbidity and shorter hospital stay.7–9 Open

aneurysm cases have more complex neck anatomy and mayrequire referral to a university centre, depending on thelocal surgery team. At BMC, this was not always necessary,as the surgeon (D.M.) had previous experience withretroperitoneal exposures, which is more commonly need -ed with complex neck anatomy.

The initial results of this small data set show that,through collaboration with a university hospital, startingsuch a program at a rural regional medical centre can besuccessful. The results of AAA repair by both endovascularand open means at BMC are similar to results reported inmajor studies and reviews.7–11 The patient population forour study was similar to that in these trials. These studiesreport 30-day mortality of 0.5%–1.2% for endovascularrepair and 3.0%–4.8% for open repair. Although the oper-ative mortality at BMC (0% for both endovascular andopen repair) was lower than these findings, this could bebecause of our relatively small number of procedures aswell as our affiliation with a major university centre, whichtends to take on our higher-risk or atypical anatomypatients. For our study, the mean length of stay in hospitalfor endovascular and open repairs was also similar to thatof other studies.7–11 The rate of morbidities and endoleaksin our study was similar to results of these other trials.

To date, reintervention was required in 3 of 26 (11.5%)endoaortic procedures. These were early reinterventions,and there have been no late reinterventions to date. Thiscould be because of the use of newer generation stentgrafts, which have improved fixation. We also found that asour experience grew we became more comfortable withiliac artery stent angioplasty to treat residual stenosis afterEVAR, reducing the risk of unplanned femorofemoralbypass. Similarly we had 1 conversion to uni-iliac repairearly in our experience because we could not cannulate thecontralateral limb gate. We have since learned to capture awire from the ipsilateral side with a snare and pull it downthrough the contralateral gate as an alternative.

Although the endovascular surgery program was createdto fill a need, it has led to the creation of an aortic diseaseclinic. Patients are therefore followed from initial consultto intervention and screened in a lifelong monitoring pro-gram. Hopefully this will impact AAA-related mortality inthe BMC service area. There have been 160 patients evalu-ated in the clinic within a 46-month period. This has led toan increased expertise in imaging of aortic pathologies andcollaboration with local radiologists. It was necessary to usemany imaging modalities, depending on purpose of theimaging study. For example, noncontrast CT or magneticresonance angiography (MRA) studies were used mainlyfor aneurysm screening and monitoring and contrast-enhanced CT studies with 3D image reconstruction wereused for all patients needing surgery.

We tried to base our approach to patient selection onguidelines as well as common community practice. In thecase of thoracic aortic aneurysms, this meant limiting our

Table 4. Vascular morbidities for EVAR (n = 23) or TEVAR (n = 3)

Vascular comorbidity No.

Femorofemoral bypass 2

Early type I endoleak 1

Early type III endoleak 1

Type II endoleak 5

Undetermined endoleak 2

Late type B aortic dissection 1

EVAR = endovascular aortic aneurysm repair; TEVAR = thoracic endovascular aortic aneurysm repair.


RECHERCHE

selection to intamural hematoma with ulcer or saccularaneurysm. Our series included 1 of each and 1 reinterven-tion, as noted previously. More complex or trauma caseswere referred to the university centre for enrolment in trials or received surgery via a conventional open approach.This was in keeping with the 2010 American Heart Associ-ation guidelines on thoracic aortic disease.22 In the case ofabdominal aortic aneurysms our patients usually had theexpectation of endovascular repair. Randomized trials havedemonstrated the safety of EVAR versus open repairs withlower early mortality, but some of the data remain contro-versial as to overall superiority in patients younger than65 years.8,11 Our patient selection (average age 71.7 yr) wasin keeping with this as well as the 2009 Society of VascularSurgery guidelines.20 More than half the abdominal an -eurysm repairs in the United States are currently treatedby endovascular means. This trend has been correlatedwith the observation that the number of annual deathsfrom intact and ruptured AAA has substantially decreasedin the United States. This has coincided with an increase inelective AAA repair after the introduction of EVAR and adecrease in the diagnosis and repair of ruptured AAA.6 Thequestion of treating infrarenal aneurysms with containedruptures merits consideration. In our hospital, such pa -tients are treated by the on-call general surgeon with anopen operation or are referred. As we gain experience andas newer generation ruptured aneurysm uni-iliac kits be -come available, it may be possible for our centre to offeremergent EVAR to such patients. The limitation that weface is that it is not cost-effective for a small program tostock stent grafts in all sizes locally. A vendor representativemust be brought in for every case at this stage. We have

been able to offer urgent EVAR (within several days) forsymptomatic aneurysms. We also found that there arefewer TEVAR operations needed than EVARs. Our hospi-tal is unique since we have perfusion services to back upTEVAR procedures should open conversion be needed.Even with such backup available, we carefully select onlythe most simple TEVAR operations, typically intramuralhematoma or saccular anurysms meeting criteria for inter-vention. It can be expected that many similar sized centreswould not have such a service and would be referringTEVARs to the university centre. Such patients could stillbe followed in a local aortic clinic. This illustrates thenotion that a comprehensive aortic clinic that partners witha university centre is preferable.

Some reports suggest that many rural centres refertheir aortic aneurysm cases to high-volume, urban cen-tres.5,13,14 However, it has been shown that patients preferto see physicians closer to home, despite a possiblyhigher mortality.23 Primary care physicians may also wishto refer patients to specialists based on qualities such asappointment timeliness and communication.24 There is apossibility that rural patients are underserved by urbancentres despite idealistic intentions. We overcame theseobstacles through the use of a “hub-and-spoke” model inwhich patients are able to access a single, local aorticclinic. Patients at the clinic are either referred to thehigh-volume university centre or treated and followedlocally. In cases where patients require referral, they arecounselled locally. Their films are prereviewed by theuniversity surgeon and discussed with the local surgeon.The local surgeon can communicate with local primarycare physicians and provide follow-up after surgery forthe university centre. Also, in this way, referrals arescreened for fitness for surgery before being seen at theuniversity centre, preventing patients from travelling un -necessarily. Cardiology clearance is obtained locallybefore referral by the comprehensive aortic clinic, assist-ing the local primary care physician with this potentiallycomplex decision tree. This model is also advantageousfor the high-volume university centre. Treatment isstreamlined because patients are essentially ready forsurgery when they are seen by accepting surgeons. The“hub-and-spoke” model described here is really a double“hub-and-spoke” model and is illustrated in Figure 4.

Other literature describes the introduction of an endo -vascular aortic surgery program into a nonuniversity hospi-tal using vascular surgeons and interventional radiologists.25

In our model, a surgeon spent 3 days per week over a 6-month period at the university hospital. This allowed thesurgeon to receive the necessary training and experiencewithout unacceptably disrupting local services. The BMCnursing and endovascular team also travelled to HUP toobserve cases intermittently. This was then crystallized byhaving a HUP surgeon proctor the BMC team for its first5 cases and sign off on their competency. Credentialing

High-volumehospital centre

Rural aortic clinic

Fig. 4. Illustration of the double “hub-and-spoke” model. Therural aortic clinic acts as a weigh station for patients with aorticpathologies. Patients of all local providers can be treated locallyor referred to the high-volume hospital centre in a streamlinedprocess. Local providers are served by a single, local aortic clinic.


RESEARCH

policies must be set by the appropriate hospital committeeusing nationally recognized guidelines. In our case, we setthe bar fairly high for initial credentialing. Recredentialingguidelines are not as well defined. Our team found that try-ing to do at least 1 case per month was a comfortable pacefor familiarity. As patient load grew, the comprehensiveapproach we strived for required additional personnel, suchas a physician assistant or nurse practitioner, to consolidatecare and assist in seeing patients.

CONCLUSION

Creation of the endovascular aneurysm repair program atBMC allowed for many of the less complicated cases to betreated with EVAR, where they previously would have beentreated with open repair or referred. As experience wasgained with endovascular procedures, more complex caseswere treated at BMC rather than referred to HUP. The col-laboration also facilitated BMC patient access to new grafttrials at HUP. The BMC gained brand recognition with arenowned medical centre, while the high-volume centrebroadened its outreach.

Acknowledgements: We thank Leslie McClements and Deborah Tricaricofor their assistance with instituting the Bayhealth Medical Center aortic clinicand its database. We also thank Richard Hageney for his advice and guidance.


Contributors: Each author helped design the study, reviewed the arti-cle and approved its publication. D. Marelli, M. Watson, M. Stallingsacquired the data. D. Marelli and M. Watson analyzed the data andwrote the article.

References

1. Lederle FA, Johnson GR, Wilson SE, et al. Aneurysm Detection andManagement (ADAM) Veterans Affairs Cooperative Study Group.Prevalence and associations of abdominal aortic aneurysm detectedthrough screening. Ann Intern Med 1997;126:441-9.

2. Chang JB, Stein TA, Liu JP, et al. Risk factors associated with rapidgrowth of small abdominal aortic aneurysms. Surgery 1997;121:117-22.

3. Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graftimplantation for abdominal aortic aneurysms. Ann Vasc Surg 1991; 5:491-9.

4. Mohan MP, Rozenfeld M, Kane RA, et al. Nationwide trends inabdominal aortic aneurysm repair and use of endovascular repair inthe emergency setting. J Vasc Interv Radiol 2012;23:338-44.

5. Mell MW, Bartels C, Kind A, et al. Superior outcomes for ruralpatients after abdominal aortic aneurysm repair supports a systematicregional approach to abdominal aortic aneurysm care. J Vasc Surg2012; 56:608-13.

6. Giles KA, Pomposelli F, Hamdan A, et al. Decrease in total aneurysm-related deaths in the era of endovascular aneurysm repair. J Vasc Surg2009;49:543-51.

7. Schermerhorn ML, O’Malley AJ, Jhaveri A, et al. Endovascular vs.open repair of abdominal aortic aneurysms in the Medicare popula-tion. N Engl J Med 2008;358:464-74.

8. Lederle FA, Freischlag JA, Kyriakides TC, et al. Outcomes followingendovascular vs open repair of abdominal aortic aneurysm: a ran-domized trial. JAMA 2009;302:1535-42.

9. Blankensteijn JD, de Jong SE, Prinssen M, et al.; Dutch RandomizedEndovascular Aneurysm Management (DREAM) Trial Group. Two-year outcomes after conventional or endovascular repair of abdom -inal aortic aneurysms. N Engl J Med 2005;352:2398-405.

10. Prinssen M, Verhoeven EL, Buth J, et al. A randomized trial compar-ing conventional and endovascular repair of abdominal aortic aneur -ysms. N Engl J Med 2004;351:1607-18.

11. Jackson RS, Chang DC, Freischlag JA. Comparison of long-term sur-vival after open vs endovascular repair of intact abdominal aorticaneurysm among medicare beneficiaries. JAMA 2012;307:1621-8.

12. Doty B, Zuckerman R, Finlayson S, et al. How does degree of rural-ity impact the provision of surgical services at rural hospitals? J RuralHealth 2008;24:306-10.

13. Hill JS, McPhee JT, Messina LM, et al. Regionalization of abdominalaortic aneurysm repair: evidence of a shift to high-volume centers inthe endovascular era. J Vasc Surg 2008;48:29-36.

14. Forbes TL, Lawlor DK, DeRose G, et al. Examination of the trendin Canada toward geographic centralization of aneurysm surgeryduring the endovascular era. Ann Vasc Surg 2006;20:63-8.

15. Chan L, Hart GL, Goodman DC. Geographic access to health carefor rural Medicare beneficiaries. J Rural Health 2006;22:140-6.

16. Economic Research Service. Rural–urban continuum codes. Washing-ton: United States Department of Agriculture, Economic ResearchService; 2012. Available: www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx (accessed 2012 June 27).

17. Gadowski GR, Pilcher DB, Ricci MA. Abdominal aortic aneurysmexpansion rate: effect of size and beta-adrenergic blockade. J VascSurg 1994;19:727-31.

18. Leach SD, Toole AL, Stern H, et al. Effect of beta-adrenergic block-ade on the growth rate of abdominal aortic aneurysms. Arch Surg1988;123:606-9.

19. Hackam DG, Thiruchelvam D, Redelmeier DA. Angiotensin- converting enzyme inhibitors and aortic rupture: a population-basedcase-control study. Lancet 2006;368:659-65.

20. Chaikof EL, Brewster DC, Dalman RL, et al. The care of patientswith an abdominal aortic aneurysm: the Society for Vascular Surgerypractice guidelines. J Vasc Surg 2009;50(Suppl):S2-49.

21. Motta R, Rubaltelli L, Vezzaro R, et al. Role of multidetector CTangiography and contrast-enhanced ultrasound in redefining follow-up protocols after endovascular abdominal aortic aneurysm repair.Radiol Med (Torino) 2012;117:1079-92.

22. Hiratzka LF, Bakris GL, Beckman JF, et al. 2010 guidelines for thediagnosis and management of patients with thoracic aortic disease.Circulation 2010;121:e266-369.

23. Finlayson SR, Birkmeyer JD, Tosteson AN, et al. Patient preferencesfor location of care: implications for regionalization. Med Care 1999;37: 204-9.

24. Kinchen KS, Cooper LA, Levine D, et al. Referral of patients to spe-cialists: factors affecting choice of specialist by primary care phys -icians. Ann Fam Med 2004;2:245-52.

25. Willoughby RP, Fenton JA, Pudupakkam SR, et al. Implementationof a successful endovascular surgical program in a non-teaching tertiary-care centre in Ontario. Can J Surg 2004;47:182-8

E114 J can chir, Vol. 56, No 5, octobre 2013 © 2013 Association médicale canadienne


Development and preliminary validation of aFunction IndeX for Trauma (FIX-IT)

Background: Assessing fracture healing in clinical trials is subjective. The new Func-tion IndeX for Trauma (FIX-IT) score provides a simple, standardized approach toassess weight-bearing and pain in patients with lower extremity fractures. We con-ducted an initial validation of the FIX-IT score.

Methods: We conducted a cross-sectional study involving 50 patients with lowerextremity fractures across different stages of healing to evaluate the reliability and pre-liminary validity of the FIX-IT score. Patients were independently examined by2 orthopedic surgeons, 1 orthopedic fellow, 2 orthopedic residents and 2 researchcoordinators. Patients also completed the Short Form-36 version 2 (SF-36v2) ques-tionnaire, and convergent validity was tested with the SF-36v2.

Results: For interrater reliability, the intraclass correlation coefficents ranged from0.637 to 0.915. The overall interrater reliability for the total FIX-IT score was 0.879(95% confidence interval 0.828–0.921). The correlations between the FIX-IT scoreand the SF-36 ranged from 0.682 to 0.770 for the physical component summaryscore, from 0.681 to 0.758 for the physical function subscale, and from 0.677 to 0.786for the role–physical subscale.

Conclusion: The FIX-IT score had high interrater agreement across multiple exam-iners. Moreover, FIX-IT scores correlate with the physical scores of the SF-36.Although additional research is needed to fully validate FIX-IT, our results suggestthe potential for FIX-IT to be a reliable adjunctive clinician measure to evaluate heal-ing in lower extremity fractures.

Level of evidence: Diagnostic Study Level I.

Contexte : Évaluer la guérison d’une fracture dans le cadre d’essais cliniques est unprocessus subjectif. Le nouveau score FIX-IT (pour Function IndeX for Trauma) con-stitue une approche simple et standardisée pour évaluer la mise en charge et ladouleur chez les patients ayant subi une fracture d’un membre inférieur. Nous avonsprocédé à une validation initiale du score FIX-IT.

Méthodes : Nous avons réalisé une étude transversale regroupant 50 patients qui ontsubi une fracture d’un membre inférieur, à différents stades de la guérison, pour éva -luer la fiabilité et la validité préliminaire du score FIX-IT. Les patients ont été exami -nés indépendamment par 2 chirurgiens orthopédistes, 1 chargé de cours en ortho -pédie, 2 médecins résidents en orthopédie et 2 coordonnateurs de recherche. Lespatients ont aussi répondu au questionnaire SF-36v2 (Short Form-36 version 2) et lavalidité convergente a été vérifiée au moyen du SF-36v2.

Résultats : En ce qui concerne la fiabilité interexaminateur, les coefficients de cor-rélation intraclasse ont varié de 0,637 à 0,915. La fiabilité interexaminateur pour lescore FIX-IT total a été de 0,879 (intervalle de confiance de 95 % 0,828–0,921). Lescorrélations entre le score FIX-IT et le SF-36 ont varié de 0,682 à 0,770 pour le scoresommaire de la composante physique, de 0,681 à 0,758 pour la sous-échelle du fonc-tionnement physique et de 0,677 à 0,786 pour la sous-échelle du rôle physique.

Conclusion : Le score FIX-IT a offert une concordance interexaminateur élevéeentre les multiples examinateurs. De plus, les scores FIX-IT sont en corrélation avecles scores physiques obtenus au SF-36. Même s’il faudra approfondir la recherchepour valider complètement le score FIX-IT, nos résultats donnent à penser que cetindice pourrait être une mesure clinique d’appoint fiable pour évaluer la guérison desfractures de membres inférieurs.

Niveau de preuve : Étude diagnostique de niveau I.

Mohit Bhandari, MD, PhD*†

Scott M. Wasserman, MD‡

Nicole Yurgin, PhD‡

Brad Petrisor, MSc, MD*

Sheila Sprague, MSc†

Ricardo E. Dent, MD‡

From the *Division of OrthopaedicSurgery, Department of Surgery, and the†Department of Clinical Epidemiologyand Biostatistics, McMaster University,Hamilton, Ont., and ‡Amgen Inc, Thou-sand Oaks, Calif.


Correspondence to:M. Bhandari293 Wellington St. N, Suite 110Hamilton ON L8L [email protected]

DOI: 10.1503/cjs.004312


RESEARCH

T he clinical assessment of fracture healing is a largelysubjective process without a gold standard.1,2 Al -though measures have been developed for hip and

ankle fractures,3 there is no validated measure that ade-quately describes functional healing for tibial fractures.1 Arecently published systematic review that evaluated vari-ability in the assessment of fracture healing in orthopedictrauma studies reported that the 3 most commonly usedclinical criteria were the absence of pain or tenderness onpalpation or examination, the absence of pain or tender-ness when bearing weight and the ability to bear weight.4

Similarly, another review evaluating the clinical criteriaused to define fracture union found that the 4 most com-mon criteria were the absence of pain or tenderness whenbearing weight, the absence of pain or tenderness on palpi-tation or examination, the ability to bear weight and theability to walk or perform activities of daily living with nopain.2

Pain at the fracture site is commonly regarded as a signthat a fracture has not yet healed.1,2 However, some patientsmay have persistent pain despite evidence of healing,whereas others may have no pain without evidence of heal-ing.2 Consequently, pain alone may be an inadequate meas -ure to determine if a fracture has healed. The ability tobear weight on injured appendages has been suggested toserve as an objective measure for healing of tibial fracturestreated by external fixation5 because weight-bearing abilityhas been shown previously to increase with time postfrac-ture6,7 and has been found to correlate well with bone stiff-ness.8 In tibial shaft fractures treated with intramedullarynailing, weight-bearing is possible from the day aftersurgery. Therefore, early weight-bearing in this contextmay not represent a healed fracture, and other dimensions,such as pain, should be considered when assessing fracturehealing.

The Function IndeX for Trauma (FIX-IT) assessmentprovides a simple standardized approach to the measure-ment of weight-bearing and pain assessment in patientswith lower extremity fractures, specifically tibia andfemoral fractures. The FIX-IT score is a clinical outcomesassessment measure ranging from 0 to 12 points in 2 do -mains: the ability to bear weight (maximum 6 points) andpain at the fracture site (maximum 6 points; see the Appen-dix, available at cma.ca/cjs). The ability to bear weight isas sessed through the single-leg stand and ambulation pro-cedures. Pain is assessed through palpation and stress pro-cedures. The scores in both domains, which are weightedequally, are summed to obtain the final total score; themaximum score of 12 indicates the highest level of func-tion. The measure was developed based on a review ofpublished literature on the assessment of tibial fracturehealing and discussion with regulatory professionals andcontent experts in orthopedic trauma surgery. The objec-tive of the present study was to evaluate the face validity,content validity, external validity, overall physician satisfac-

tion, interrater reliability and convergent validity of theFIX-IT measure.

METHODS

Overview of the study design

To assess face and content validity, the FIX-IT measurewas independently evaluated by 5 orthopedic trauma sur-geons before the clinical study. We conducted a cross- sectional study of patients with lower extremity fracturesacross different stages of healing to evaluate the interraterreliability of the FIX-IT measure. We obtained researchethics board approval before initiating the study. Patientswere enrolled from Hamilton Health Sciences — GeneralSite in Hamilton, Ont., and the sample was a nonrandom,convenience sample of patients with tibial or femoral frac-tures presenting to a fracture clinic. To assess interraterreliability, patients were independently examined by 7 re -viewers. Prior to performing the FIX-IT assessments, werecorded the demographic and fracture characteristics ofpatients, and the patients completed the Short-form 36,version 2 (SF-36v2) questionnaire.

Assessment of face and content validity

To assess face validity, 5 orthopedic trauma surgeons fromNorth America, Europe and Asia independently reviewedthe FIX-IT measure and determined whether it lookedlike it was going to measure what it was supposed to meas -ure. Specifically, the surgeons were asked to rate on a scaleof 1–5 the overall agreement with the validity of this meas -ure for understanding functional healing in patients withfractures.

We assessed content validity qualitatively by asking eachof the 5 surgeons to determine whether each item was“essential,” “useful, but not essential” or “not necessary” tothe performance of the construct. Each surgeon also ratedtheir overall satisfaction with the administration of themeasure and completed an open-ended question askingwhether any item essential for the performance of the con-struct was currently missing from the existing measure.

Assessment of interrater reliability

We assessed consecutive patients with lower extremityfractures attending a fracture clinic for inclusion in ourstudy. The inclusion criteria were lower extremity long-bone fracture, age 18 years or older, English language, ability to ambulate before the fracture and provision ofinformed consent. The exclusion criteria were bilaterallower extremity fractures, fractures of the axial skeletonlimiting weight-bearing, inability to complete question-naires or comply with functional tests, presence of pre -injury lower extremity pain syndrome, paralysis or sensory


RECHERCHE

deficit and prefracture use of assistive devices. We obtainedinformed consent from all participants, and baseline andfracture characteristics were recorded.

Two orthopedic surgeons, 1 orthopedic fellow, 2 ortho-pedic surgical trainees and 2 research coordinators in -dependently assessed patient function using the FIX-ITmeasure. Raters were selected before the study and eachrater participated in a training session on how to use theFIX-IT measure. The team then evaluated each patient,and each rater, unaware of the other raters’ responses,scored patient function in all participants.

Assessment of convergent validity

Each eligible patient completed the SF-36v2, which is ahealth-related quality of life measure.9 The SF-36v2 di -mensions were scored separately and transformed to a 0–100 scale. Domains were also grouped into the physicalcomponent summary (PCS) score and the mental com -ponent summary (MCS) score, as recommended by theSF-36v2 scoring manual.9 We chose to use the SF-36v2rather than other available instruments because of itsuse in previous studies evaluating fracture outcomes andthe hypothesis that the FIX-IT measure would correlatewith the SF-36v2 physical functioning scale, role-physicalscale and physical health component summary measurescores.2,10–13

Sample size considerations

The sample size is controlled in reliability studies by vary-ing the number of raters and the number of patients.Although increasing the number in either group will yielda more precise reliability estimate, the number of partici-pants has a much greater impact on the precision than thenumber of raters. The number of raters was chosen basedon generalizability and feasibility. Using 2 orthopedic sur-geons, 1 orthopedic fellow, 2 orthopedic residents and2 re search coordinators as raters, we determined that asample of 50 patients would provide sufficient precisionfor meaningful analysis of the FIX-IT measure. Assumingan expected intraclass correlation (ICC) of 0.8, a samplesize of 50 patients and 7 raters, the expected half width ofthe 95% confidence interval (CI) for the estimated ICCwas approximately 0.10.14


Reviewer assessment of face and content validity and over-all satisfaction with the administration of FIX-IT wasassessed with 5-point Likert-type scales ranging from 1(completely disagree) to 5 (completely agree). Scores weresummarized qualitatively for each assessment.

We used ICCs with 95% CIs to measure agreement inthe rater’s overall FIX-IT scores, including the 4 com -

ponent scores. The ICC, used to quantify agreement for acontinuous variable, is equivalent to the quadraticallyweighted κ for categorical data. The weighted κ, asdescribed by Fleiss,15 adjusts the observed proportion ofagreement by correction for the proportion of agreementthat could have occurred by chance alone. As they arenumerically equivalent, similar guidelines for interpreta-tion of κ values can be applied to the ICC. Landis andKoch16 suggest that κ of 0–0.2 represents slight agreement,0.21–0.40 fair agreement, 0.41–0.60 moderate agreement,and 0.61–0.80 substantial agreement. A κ value above 0.80is considered almost perfect agreement. The value of theICC ranges from +1, representing perfect agreement, to –1, representing absolute disagreement.

In addition, FIX-IT was compared with similar do -mains of a frequently used patient-reported outcomesscale. Specifically, the association between the FIX-IT

Table 1. Content and face validity of the FIX-IT measure

Question; responses No. (%), n = 5

How would you rate your overall agreement withthe validity of this measure for understandingfunctional healing for fracture patients?

5 (completely satis!ed) 0

4 4 (80)

3 1 (20)

2 0

1 (completely unsatis!ed) 0

How satis!ed are you with the administration ofthe measure?

5 (completely satis!ed) 0

4 3 (60)

3 2 (40)

2 0

1 (completely unsatis!ed) 0

How important is the ability measured by the single-leg stand score for understanding functionalhealing for fracture patients?

Essential 1 (20)

Useful, but not essential 4 (80)

Not necessary 0

How important is the ability measured by the ambulation score for understanding functionalhealing for fracture patients?

Essential 3 (60)


Not necessary 0

How important is tolerance measured by the palpation score for understanding functional healing for fracture patients?

Essential 3 (60)


Not necessary 0

How important is the tolerance measured by the stress score for understanding functional healing for fracture patients?

Essential 1 (20)


Not necessary 0

FIX-IT = Function IndeX for Trauma.


RESEARCH

scores and the SF-36v2 physical functioning scale, role–physical scale, and physical health component summarymeasure scores were assessed using Pearson correlation.The SF-36v2 was scored according to the SF-36v2 version2.0 scoring manuals.9

RESULTS

Assessment of face and content validity

The FIX-IT instrument demonstrated acceptable face andcontent validity, as measured by 5 experts who determinedthat the items were all either useful or essential (Table 1).When asked if there were any additional items to considerincluding in the FIX-IT measure, 2 reviewers suggestedincluding return to work, a patient-important outcome

that indicates the ability to resume both physical and mental activities.17 As a substantial proportion of patientswill not return to work even 2 years after the fractureoccurrs,18 return to work as a measure of fracture healingmay not be very responsive to change.17 Therefore, the de -velopers opted not to incorporate questions on returnto work.

Patient characteristics

Of the 50 patients enrolled in the study, 42 (84%) hadtibia fractures and 8 (16%) had femur fractures (Table 2).The mean time from injury to assessment for the studywas 34 (range 0.5–555) months. The majority of thepatients evaluated had already established problems withtheir fracture healing (Table 2).

Table 2. Participant and fracture characteristics

Participant characteristic No. (%)* n = 50 Fracture characteristic

No. (%)* n = 50

Age, yr (mean ± SD) 48 ± 16 Fracture type

Sex Closed 32 (64)

Male 36 (72) Open 18 (36)

Female 14 (28) Type of treatment

Ethnicity Plate "xation 22 (44)

White 40 (80) Reamed intramedullary nail 16 (32)

Native Canadian 5 (10) Conservative treatment 9 (18)

Hispanic 4 (8) Screws 2 (4)

Southeast Asian 1 (2) Nonreamed intramedullary nail 1 (2)

Time from injury, mo., mean ± SD (range) 34 ± 77 (0.5–555) Complications

< 6 9 (18) None 17 (34)

6–12 4 (8) Yes 33 (66)

12–24 19 (38) Type of complication, n = 42‡

> 24 18 (36) Nonunion 13 (31)

Mechanism of injury Delayed union 13 (31)

Fall 19 (38) Infection 7 (17)

Motor vehicle accident 9 (18) Wound healing problem 5 (12)

Motorcycle motor vehicle accident 6 (12) Hardware failure 2 (5)

Pedestrian motor vehicle accident 4 (8) Varus collapse 1 (2)

Recreational vehicle accident 4 (8) Posttraumatic arthritis 1 (2)

Crush 3 (6) Reoperations

Direct trauma — blunt 3 (6) None 22 (44)

Twist 2 (4) Yes 28 (56)

Additional injuries Type of reoperation, n = 47‡

None 27 (54) Hardware removal 13 (28)

Yes 23 (46) Bone graft 11 (23)

Total of 33 additional injuries† Revision open reduction internal "xation 8 (17)

Upper extremity injury 11 (33) Irrigation and debridement 6 (13)

Other lower extremity injury 10 (30) Sequestrectomy 4 (9)

Spinal injury 5 (15) Skin graft 2 (4)

Head/facial injury 5 (15) Arthrodesis 2 (4)

Thoracic injury 2 (6) Osteotomy 1 (2)

Location of fracture None 22 (44)

Tibia 42 (84)

Femur 8 (16)

SD = standard deviation. *Unless otherwise indicated. †Patients may have had more than 1 additional injury. ‡Patients may have had multiple complications and reoperations.


RECHERCHE

Functional status

The mean SF-36v2 PCS score was 35.06 ± 9.77, and themean SF-36v2 MCS score was 43.73 ± 15.55 (Fig. 1). Themean SF-36v2 physical functioning scale score was 39.32 ±29.23, the mean role–physical scale score was 33.92 ± 34.09,and the mean bodily pain scale score was 42.94 ± 24.00.

The FIX-IT measure and assessment of interraterreliability

The mean overall FIX-IT score was 7.97 ± 2.73 (Table 3).The overall interrater reliability for the FIX-IT score was

0.879 (95% CI 0.828–0.921; Table 4). The interrater reli -ability was 0.860 (95% CI 0.787–0.913) between the2 orthopedic surgeons and the orthopedic fellow, 0.878(95% CI 0.793–0.929) between the 2 residents and 0.893(95% CI 0.819–0.938) between the 2 research coordinators.

Assessment of convergent validity

The correlations between the FIX-IT score and the SF-36v2 PCS score ranged from 0.682 to 0.770 (Table 5).The correlations between the FIX-IT score and the SF-36v2 physical functioning scale ranged from 0.681 to0.758, and the correlation between the FIX-IT score andthe SF-36v2 role-physical scale ranged from 0.677 to0.786. The correlations between each procedure in theFIX-IT score and the SF-36v2 PCS are summarized inTable 5.

DISCUSSION

The use of a reliable, valid and responsive measure of frac-ture healing is essential for precisely estimating treatmenteffects in clinical trials.17 The FIX-IT measure is a recentlydeveloped, simple fracture healing assessment tool empha-sizing outcomes that are likely important to patients. Thispreliminary study has demonstrated the FIX-IT measurehas acceptable face and content validity and has shownthat overall interrater reliability for the FIX-IT scoreamong all 7 reviewers was 0.879 (95% CI 0.828–0.921),which demonstrates excellent agreement.14 The interraterreliability was above 80% among the 2 orthopedic sur-geons and the orthopedic fellow, between the 2 residents,

SF-36v2

60

50

40

30

20

10

0

US population norm

Study sample

Physical healthcomponent summary

score

Mental health componentsummary score

Fig. 1. Comparison of Short Form-36, version 2 (SF-36v2) scoresfrom the study sample with the American population norms.

Table 3. FIX-IT scores, n = 50

Score, minimum: 0, maximum: 3* Mean ± SD Median Min–max Quartiles

Single-leg stand score 1.63 ± 1.19 2 0–3 25 1 50 2 75 3

Ambulation score 2.09 ± 0.93 2 0–3 25 2 50 2 75 3

Total score for ability to bear weighton fractured limb Minimum: 0, maximum: 6

3.72 ± 1.95 4 0–6 25 2 50 4 75 6

Palpation score 1.91 ± 0.75 2 1–3 25 1 50 2 75 2

Stress score 2.33 ± 0.80 3 0–3 25 2 50 3 75 3

Total score for pain at the fracture siteMinimum: 0, maximum: 6

4.25 ± 1.35 4 1–6 25 3 50 4 75 5

Overall FIX-IT score Minimum: 0, maximum: 12

7.97 ± 2.73 8 1–12 25 6 50 8 75 10

FIX-IT = Function IndeX for Trauma; SD = standard deviation.*Unless otherwise indicated.


RESEARCH

and between the 2 research coordinators. This demon-strates that the FIX-IT measure has excellent reliabilityacross different raters with different levels of clinicalassessment skills and suggests that FIX-IT can be consist -ently administered by surgical trainees in clinical practicesand in clinical studies by research coordinators.

Although the FIX-IT measure has adequate convergentvalidity with the SF-36v2, there are a couple of reasonsthat the correlation may not be perfect. First, generichealth-related quality of life measures often lack sensitivityto detect smaller functional changes that may be affectedby an orthopedic injury,2 and it is possible that the FIX-ITmeasure better captured the patient’s abilities than the SF-36v2. Also, the SF-36v2 elicits the patient’s perspective onphysical function whereas FIX-IT elicits the clinician’s per-spective on fracture healing. The expectation of healingmay be different for the patient and the clinician, possiblyimpacting the ratings of their functioning.

Limitations

As the present study was a preliminary evaluation of theFIX-IT assessment, it had limitations. First, in the initialsurgeon assessment of content, it may have been unclear toexpert reviewers that the goal was not to further reduce the

items on the FIX-IT assessment. Surgeons may have feltthat they had to indicate that at least something was notessential. Second, this was a convenience sample of patientsfrom 1 surgeon’s fracture clinic, limiting the generalizabil-ity of the findings. Third, the majority of the patientsincluded in this study were assessed at least 12 monthsafter the fracture, and many patients were being seen at thefracture clinic for complications. This is also evident in thepatients’ SF-36v2 scores, as they were lower than antici-pated. The SPRINT study,19 a large randomized controlledtrial evaluating reamed versus unreamed intramedullarynails, reported a physical component score of 42.9 ± 11 inthe reamed group and 43.5 ± 11 in the unreamed group1 year postinjury. This score is higher than the physicalcomponent score of 35.06 ± 9.77 found in the presentstudy, indicating that patients in our study likely experi-enced more complications than the typical patient with alower extremity fracture. Fourth, patients were only as -sessed once in our study, as opposed to being assessed overtime to measure the progression of fracture healing, as inclinical practice or in a prospective clinical trial.

A strength of this study is that multiple raters with dif-ferent clinical backgrounds and training levels independ -ently assessed each included patient. The interrater agree-ment was acceptable among all raters, implying that the

Table 4. Interrater reliability of the FIX-IT measure

Item

Group; ICC (95% CI)

Overall, n = 7 Surgeons and orthopedic

fellow, n = 3 Residents, n = 2 Research coordinators,

n = 2

Single-leg stand score 0.834 (0.769–0.890) 0.811 (0.717–0.880) 0.825 (0.712–0.897) 0.805 (0.624–0.895)

Ambulation score 0.854 (0.795–0.904) 0.857 (0.784–0.911) 0.893 (0.819–0.938) 0.856 (0.760–0.916)

Total score, ability to bear weight on fractured limb

0.897 (0.852–0.933) 0.890 (0.831–0.932) 0.915 (0.855–0.951) 0.874 (0.763–0.931)

Palpation score 0.714 (0.620–0.803) 0.637 (0.494–0.759) 0.701 (0.529–0.818) 0.751 (0.599–0.851)

Stress score 0.725 (0.633–0.810) 0.685 (0.552–0.793) 0.737 (0.580–0.841) 0.771 (0.628–0.864)

Total score, pain at the fracture site

0.784 (0.705–0.854) 0.732 (0.612–0.827) 0.763 (0.618–0.858) 0.827 (0.716–0.898)

Total FIX-IT score 0.879 (0.828–0.921) 0.860 (0.787–0.913) 0.878 (0.793–0.929) 0.893 (0.819–0.938)

CI = con!dence interval; FIX-IT = Function IndeX for Trauma; ICC = intraclass correlation coef!cient.

Table 5. Correlation of the FIX-IT measure with the SF-36v2 and correlation of each procedure in the FIX-IT measure with the SF-36v2 — physical health component summary score

ItemPearson correlation, values across

individual assessors, range

Correlation of the FIX-IT measure with the SF-36v2

SF-36v2 physical health component summary score (range) 0.682–0.770

SF-36v2 physical function scale score (range) 0.681–0.758

SF-36v2 role-physical scale score (range) 0.677–0.786

Correlation of each procedure in the FIX-IT measure with the SF-36v2 physical health component summary score (range)

Single-leg stand score 0.499–0.594

Ambulation score 0.549–0.711

Palpation score 0.309–0.506

Stress score 0.444–0.503

FIX-IT = Function IndeX for Trauma; SF-36v2 = Short-form 36 version 2.

FIX-IT measure can be administered by study personnelor surgical trainees, reducing the demands of a clinical trialon the orthopedic surgeon.

CONCLUSION

The FIX-IT measure incorporates common clinical cri -teria into a simple assessment tool. The developers did notinclude questions about activities of daily living or returnto work into the tool. Such questions were excluded in theFIX-IT measure because it was developed to be a simpleindex, and these questions are often included in other vali-dated measures that are administered in patients partici-pating in clinical trials.17 The developers also did notincorporate radiographic parameters into the assessmenttool. As radiographic parameters are subjective, adjudica-tion of these outcomes is becoming the gold standard,17

thus the developers opted to exclude radiographic out-comes from the FIX-IT measure.

Future research on the FIX-IT assessment should beconducted at multiple centres in larger numbers ofpatients, should include patients with fresh fractures andshould measure the evaluation of fracture healing progres-sion over time.

Acknowledgments: We thank the 5 orthopedic surgeons who assessedthe FIX-IT index for face and content validity prior to the clinical study.We thank the reviewers for their support and commitment. We wouldalso like to acknowledge Amgen Inc. for funding the study.

Competing interests: This study was funded by Amgen Inc. M. Bhandariis funded by a Canada Research Chair. S.M. Wasserman, N. Yurgin andR. Dent are employees of Amgen Inc and have stock options. No othercompeting interests declared. This study received research ethics boardapproval.

Contributors: M. Bhandari, S.M. Wasserman, N. Yurgin, S. Spragueand R.E. Dent designed the study. M. Bhandari, S.M. Wasserman,B. Petrisor and S. Sprague acquired the data, which M. Bhandari,S.M. Wasserman, N. Yurgin, S. Sprague and R.E. Dent analyzed.M. Bhandari, S.M. Wasserman and S. Sprague wrote the article.M. Bhandari, S.M. Wasserman, N. Yurgin, B. Petrisor and R.E. Dentreviewed the article. All authors approved its publication.

References

1. Bhandari M, Guyatt GH, Swiontkowski MF, et al. A lack of consen-sus in the assessment of fracture healing among orthopaedic sur-geons. J Orthop Trauma 2002;16:562-6.

2. Morshed S, Corrales L, Genant H, et al. Outcome assessment in clin-ical trials of fracture-healing. J Bone Joint Surg Am 2008;90:62-7.

3. Agel J, Swiontkowski MF. Guide to outcomes instruments for muscu-loskeletal trauma research. J Orthop Trauma 2006;20(8 Suppl):S1-146.

4. Corrales LA, Morshed S, Bhandari M, et al. Variability in the assess-ment of fracture-healing in orthopaedic trauma studies. J Bone JointSurg Am 2008;90:1862-8.

5. Joslin CC, Eastaugh-Waring SJ, Hardy JR, et al. Weightbearing aftertibial fracture as a guide to healing. Clin Biomech (Bristol, Avon) 2008;23: 329-33.

6. Aranzulla PJ, Muckle DS, Cunningham JL. A portable monitoringsystem for measuring weight-bearing during tibial fracture healing.Med Eng Phys 1998;20:543-8.

7. Kershaw CJ, Cunningham JL, Kenwright J. Tibial external fixation,weight bearing, and fracture movement. Clin Orthop Relat Res 1993;293: 28-36.

8. Goodship AE, Kenwright J. The influence of induced micromove-ment upon the healing of experimental tibial fractures. J Bone JointSurg Br 1985;67:650-5.

9. Ware JE, Kosinski M, Turner-Bowker DM, et al. How to score version 2of the SF-12 Health Survey (with a supplement documenting version 1).Lincoln (RI): QualityMetric; 2002.

10. Dogra AS, Ruiz AL, Marsh DR. Late outcome of isolated tibial frac-tures treated by intramedullary nailing: the correlation between disease-specific and generic outcome measures. J Orthop Trauma2002; 16: 245-9.

11. Zlowodzki M, Obremskey WT, Thomison JB, et al. Functional out-come after treatment of lower-extremity nonunions. J Trauma 2005;58: 312-7.

12. Bhandari M, Sprague S, Hanson B, et al. Health-related quality of lifefollowing operative treatment of unstable ankle fractures: a prospec-tive observational study. J Orthop Trauma 2004;18:338-45.

13. Bhandari M, Guyatt G, Tornetta P 3rd, et al.; Study to ProspectivelyEvaluate Reamed Intramedullary Nails in Patients with Tibial Frac-tures (SPRINT) Investigators. Randomized trial of reamed andunreamed intramedullary nailing of tibial shaft fractures. J Bone JointSurg Am 2008;90:2567-78.

14. Giraudeau B, Mary JY. Planning a reproducibility study: how manysubjects and how many replicates per subject for an expected width ofthe 95 per cent confidence interval of the intraclass correlation coef-ficient. Stat Med 2001;20:3205-14.

15. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York:John Wiley & Sons; 1981.

16. Landis JR, Koch GG. The measurement of observer agreement forcategorical data. Biometrics 1977;33:159-74.

17. Kooistra BW, Sprague S, Bhandari M, et al. Outcomes assessment infracture healing trials: a primer. J Orthop Trauma 2010;24:S71-5.

18. Butcher JL, MacKenzie EJ, Cushing B, et al. Long-term outcomesafter lower extremity trauma. J Trauma 1996;41:4-9.

19. SPRINT Investigators. Functional, general, and disability outcomes aftertibial nailing: a randomized trial. Orthopaedic Trauma AssociationAnnual Meeting; 2007 Oct. 17–20; Denver, Colo. Available: www.hwbf.org/ota/am/ota07/otapa/OTA070422.htm (accessed 2010 Jan. 24).

RECHERCHE




Quality of narrative operative reports in pancreaticsurgery

Background: Quality in health care can be evaluated using quality indicators (QIs).Elements contained in the surgical operative report are potential sources for QI data,but little is known about the completeness of the narrative operative report (NR). Weevaluated the completeness of the NR for patients undergoing a pancreaticoduo-denectomy.

Methods: We reviewed NRs for patients undergoing a pancreaticoduodenectomyover a 1-year period. We extracted 79 variables related to patient and narrator charac-teristics, process of care measures, surgical technique and oncology-related outcomesby document analysis. Data were coded and evaluated for completeness.

Results: We analyzed 74 NRs. The median number of variables reported was 43.5(range 13–54). Variables related to surgical technique were most complete. Process ofcare and oncology-related variables were often omitted. Completeness of the NR wasassociated with longer operative duration.

Conclusion: The NRs were often incomplete and of poor quality. Important ele-ments, including process of care and oncology-related data, were frequently missing.Thus, the NR is an inadequate data source for QI. Development and use of alternativereporting methods, including standardized synoptic operative reports, should beencouraged to improve documentation of care and serve as a measure of quality ofsurgical care.

Contexte : Il est possible d’évaluer la qualité des soins de santé au moyen d’indica-teurs de qualité (IQ). Les éléments contenus dans les notes opératoires (NO) sont unesource potentielle de renseignements pouvant servir d’IQ, mais on en sait peu sur leurexhaustivité. Nous avons voulu évaluer l’exhaustivité des NO dans les dossiers depatients soumis à une pancréatoduodénectomie.

Méthodes : Nous avons passé en revue les NO dans les dossiers de patients soumis àune pancréatoduodénectomie sur une période d’un an. Par analyse des documents,nous avons extrait 79 variables liées aux caractéristiques des patients et aux rédacteursdes NO, aux mesures des protocoles opératoires, à la technique chirurgicale et auxrésultats oncologiques. Nous avons encodé et évalué ces données en fonction de leurexhaustivité.

Résultats : Nous avons analysé les NO pour 74 interventions. Le nombre médian devariables relevées était de 43,5 (entre 13 et 54). Les variables liées au protocole desoins et les variables oncologiques étaient souvent omises. L’exhaustivité des NO étaitproportionnelle à la durée de l’intervention.

Conclusion : Les NO sont souvent incomplètes et leur qualité laisse à désirer. Deséléments importants, dont le protocole opératoire et les données oncologiques, étaientsouvent manquants. Ainsi, les NO constituent une source inadéquate de données en cequi concerne les IQ. Il faudra encourager la mise au point et l’utilisation d’autres typesde rapports, dont des synopsis opératoires standardisés, pour mieux documenter lessoins chirurgicaux prodigués et pour en évaluer la qualité.

Q uality improvement is an important component of health care sys-tems. Quality in health care can be evaluated in terms of the struc-tures, processes and outcomes of care.1 Process and/or outcome data

are used as quality indicators (QIs) for performance management. Outcomedata, such as survival time, complication rates or quality of life data, are oftendifficult to obtain or take a long time to mature. As a result, process of care

Meagan E. Wiebe, BSc*

Lakhbir Sandhu, MD*

Julie L. Takata, MA*

Erin D. Kennedy, MD, PhD*†

Nancy N. Baxter, MD, PhD†‡

Anna R. Gagliardi, PhD*†

David R. Urbach, MD, MSc*†

Alice C. Wei, MDCM, MSc*†

From the *Division of General Surgery,University Health Network, the †Depart-ments of Surgery and the Institute ofHealth, Policy Management & Evalua-tion, University of Toronto, and the ‡Divi-sion of General Surgery, St. Michael’sHospital, Toronto, Ont.


Correspondence to:A.C. WeiDivision of General Surgery10EN-215, Toronto General Hospital200 Elizabeth St.Toronto ON M5G [email protected]

DOI: 10.1503/cjs.028611


RECHERCHE

data are frequently used as a surrogate for outcome datawhen measuring the quality of a health care system becauseprocess of care data are often available from administrativedata sources.

At present, few QIs are available that measure theprocesses of care that occur during a surgical procedure.There is interest is measuring components of the operativereport as a potential source of data for QIs. Completenessand accuracy of an operative report may be a means to assessthe quality of care delivery and to identify opportunitiesfor new quality initiatives. Thus, elements of the operativereport have the potential to be used as QIs, but to ourknowledge, this has not yet been established in the literature.

A narrative operative report (NR) is currently the stan-dard documentation method used for the vast majority ofsurgical procedures in North America. It is an open formatdescription of the operative steps performed during a sur-gical procedure dictated by a surgeon in narrative form.The content of the NR is not standardized or regulated.2

The primary function of the NR is to document proced -ural events for clinical and medicolegal reasons. An NRmay potentially be used to measure the quality of the sur-gical procedure if intraoperative process of care measurescan be extracted in a robust manner. But, at the presenttime little is known about the quality or the completenessof the NR. A study by Govindarajan and colleagues3 foundthat NRs can be used to extract data about nontechnicalcompetencies of a surgical procedure, but the authors didnot address issues related to the quality of health care.Others have reported that NRs are of variable qualityowing to incomplete and/or inadequate reporting.2,4–8 Forpatients with cancer, Edhemovic and colleagues4 reportedthat NRs failed to adequately document the oncologicallyrelevant elements that occur in rectal cancer procedures.Incomplete and inconsistent documentation in the NRmay compromise the ability of physicians to make optimaldecisions regarding further treatment.9

Newer documentation methods have been developedthat allow for standardized reporting of operative proced -ures. Proponents of standardized operative reports, alsoknown as synoptic operative reports (SRs), point to morecomplete documentation with fewer omissions in SRs thanNRs, resulting in higher quality data, as reasons to adopt theSR. Data from SRs may be used to improve communicationbetween different health care providers to optimize clinicalcare, resulting in higher quality of care. For example, SRsthat include details regarding the margin status of a proced -ure (i.e., R0, R1 or R2) may help inform the planning ofpostoperative adjuvant therapy, such as external beam radio-therapy. Also, data from an SR can be used for performanceevaluation, quality improvement and research purposes.

Pancreaticoduodenectomy is performed for pancreaticcancer. Institutions that perform a high volume of pan-creaticoduodenectomies have better outcomes than cen-tres that perform only a few procedures per year.10–14

However, surgeon-specific processes account for a substan-tial component of the observed volume–outcome associa-tions for pancreaticoduodenectomy.15–20 It is postulated thatprocesses related to the technical proficiency and adherenceto oncologic principles during the pancreaticoduodenec-tomy contribute to the improved outcomes observed. Thus,operative notes may be a useful source of intraoperativeprocess of care data for this procedure.

The objective of our study was to evaluate the potentialof the NR as a possible source of quality improvement data.Using document analysis, we assessed the completeness ofreporting of data in NRs from pancreaticoduodenectomyprocedures to evaluate the quality of data available in NRs.

METHODS

We identified the NRs of patients who underwent a pan-creaticoduodenectomy between Jan. 1, 2008, and Dec. 31,2008, from a prospective maintained database at the Uni-versity Health Network, Toronto, Ont. This academicinstitution has a high volume of hepato-pancreato-biliary(HPB) procedures yearly, performed by 10 subspecialty-trained surgeons. We obtained ethics approval fromthe institutional research ethics board before the study commencement.

We analyzed the contents of dictated NRs. A draftframework of data elements considered important for anNR for pancreaticoduodenectomy was developed based ona literature review of outcomes following pancreaticoduo-denectomy, operative variables that were collected in anexisting provincial clinical database and input from generalsurgeons with content expertise. Potential data elementswere pilot-tested for face validity by 5 surgical oncologists,including HPB surgical oncologists, and modified based onexpert input to create a final set of variables.

We evaluated 79 variables covering 3 domains of inter-est: process of care, surgical manoeuvres and oncology-related variables. Of the 79 variables, 60 were consideredmandatory and 19 were deemed optional.

The standard NR consists of a verbatim transcribedaccount of the procedure narrated by a physician memberof the surgical team. This document is created free-formand is unstructured in format and content. We analyzeddictated NRs from the patients’ electronic medical records;handwritten notes in the paper chart were excluded. Datawere extracted from the NRs by an independent dataextractor. Demographic and clinical characteristics of thepatients and the characteristics of the physician narratorwere also recorded.


We calculated summary statistics for patient demographicinformation and the level of training of the individual whonarrated the report. The variables were grouped into data


RESEARCH

elements, and we calculated the median number of vari-ables reported for each data element. Data pertaining toconcomitant procedures performed at the same time asthe pancreaticoduodenectomy were excluded from analy-sis, as these elements were unique to each situation andthe content was nonstandard.

We performed univariate analysis using the Mann–Whitney U test, χ2 test or Fisher exact test, as appropriate.The data were analyzed using SPSS 15.0. In addition, weperformed comparative analysis of the 5 most and leastcomplete dictated NRs for variables of interest as a form ofsensitivity analysis.

RESULTS

A total of 78 pancreaticoduodenectomies were performed,and 74 NRs were available for data extraction. In 4 cases,an NR was not dictated and was absent from the elec-tronic medical record. These cases were excluded fromour analysis.

Patient characteristics

There were 74 patients analyzed. In 61 patients (82%), astandard Whipple type pancreaticoduodenectomy wasperformed. Thirteen patients also underwent concomitantvascular procedures including portal vein resections (13 of74) and/or arterial resections (2 of 74). Twenty-onepatients had additional non-HPB procedures.

The majority of patients were men (43 of 74, 58%) andolder than 60 years (45 of 74, 61%; Table 1).

Narrative report characteristics

The average time to dictation of the NRs was 1.5 days.The physician team member who dictated the NR was theattending surgeon (14 of 74, 19%), clinical fellow (43 of74, 58%) or senior surgical resident (16 of 74, 22%;Table 1). There were no instances of duplicate NRs. Noneof the NRs was reviewed or verified by the staff surgeonvia the electronic records system.

The median number of variables reported was 43.5(range 13–54; Tables 2–4). No NR was complete for all60 mandatory variables. The processes of care and onco-logic variables were least complete, with several omissions(Tables 2 and 4). A median of 3 of 9 (range 0–7) processes ofcare variables were reported. No procedure was completefor all process of care variables. The most commonly omit-ted process of care variables were urgency of surgery (13 of74, 18%), time out performed (11 of 74, 15%) and Ameri-can Society of Anesthesiologists (ASA) status (0%; Table 2).

Oncology-specific findings were reported for a medianof 5 of 9 (range 0–9) variables. Tumour size (38 of 74,51%), lymphadenectomy performed (29 of 68, 43%), clin -ic al resection and/or margin status reported (18 of 74,24%) and lymphadenopathy (22 of 74, 30%) were the leastfrequently reported oncologic variables (Table 4).

Administrative variables and surgical technique variableswere most commonly complete and were reported a for

Table 1. Patient and narrator characteristics, n = 74

Characteristic No. (%)

Patient sex, male:female 43:31 (58:42)

Patient age, yr

≤ 60 29 (39)

> 60 45 (61)

Narrator level

Clinical fellow 43 (58)

Senior surgical resident 16 (22)

Attending surgeon 14 (19)

Unknown 1 (1)

Table 2. Note and procedures process characteristics, n = 74

Characteristic No. present (%)*

Administrative

Chart no. 74 (100)

Date of surgery 74 (100)

Date of dictation 74 (100)

Dictating physician 73 (99)

Median (range) 4 (3–4)

Procedure

Surgeon 73 (99)

Incision 73 (99)

Preoperative diagnosis 68 (92)

Assistants 67 (91)

Procedure performed 66 (89)

Postoperative diagnosis 57 (77)

Proposed procedure 51 (69)

Position 49 (66)


Clinical information

Indication for surgery listed

71 (96)

Comorbidities listed 60 (81)


Process

Patient disposition 66 (89)

Sponge/instrument count reported

45 (61)

Consent obtained 43 (58)

Specimen disposition 24 (32)

Preoperative antibiotics 22 (30)

DVT prophylaxis 19 (26)

Urgency of surgery 13 (18)

Time out performed 11 (15)

ASA status 0


ASA = American Society of Anesthesiologists; DVT = deep vein thrombosis. *Unless otherwise indicated.


RECHERCHE

median of 4 of 4 (range 3–4) and 20 of 28 (range 1–26)variables, respectively. All identifying patient informationwas re port ed. The procedure performed was reported in66 (89%) of the cases. Variables associated with surgicaltechnique most commonly reported were those related topancreatic mo bil ization and resection with a median of 5 of6 variables (range 1–6; Table 3).

Narrative report completeness and physiciandictator characteristics

The χ2 tests revealed no significant results when compar-ing narrator type (attending surgeon, clinical fellow, seniorsurgical resident) for each of the reported variables, exceptfor time out performed, bile duct anastomosis type andspecimen disposition (all p < 0.05). Fellows reported time

out performed more often, surgical house staff (clinical fel-lows or residents) reported the bile duct anastomosis typemore often, and senior residents reported specimen dispo-sition more often than the other narrator types.

Sensitivity analysis was performed for comparativeanalysis. The 5 most and least complete NRs were identi-fied and compared. The 5 most complete NRs included 54(87%) variables. These NRs were all dictated by the sameclinical fellow within a week of the procedure date. The5 least complete NRs included 36 (< 58%) variables. TheseNRs were dictated by surgical house staff within a week ofthe procedure date. None of the most or least completereports was dictated by attending surgeons.

Narrative report completeness and perioperativeoutcomes

We evaluated the association between perioperative out-comes and completeness of NRs. Completeness wasdivided into quartiles. Perioperative outcomes of the leastand most complete NRs were compared (Table 5) by uni-variate analysis. Completeness of the NR was positivelyassociated with operating room (OR) times (p = 0.048). Inthe most complete NR quartile, a median of 48 variableswere present and the median OR time was 396 (unknown–800) minutes versus the least complete quartile, which hada median of 39.5 variables present and a median OR time

Table 3. Surgical manoeuvres, n = 74

Manoeuvre No. present (%)*

Pancreatic mobilization

Duodenum kocherized 70 (95)

Lesser sac opened 67 (91)

Tunnel created under pancreatic neck 67 (91)

GDA identified/divided 65 (88)

Cholecystectomy performed 64 (86)

Colon mobilized 50 (68)


Resection

Pancreas divided 71 (96)

Distal GI margin divided 69 (93)

Common bile duct divided 69 (93)

Proximal GI margin divided 66 (89)

Level of bile duct division reported 52 (70)

Uncinate process divided 49 (66)


Anastomosis

Pancreatic anastomosis type 67 (91)

Sutures used 73 (99)

GI anastomosis type 60 (81)

Pancreatic texture 43 (58)

Bile duct anastomosis type 38 (52)

Sutures used 69 (93)


Closure/other details

Wound closure, type 73 (99)

Other procedures described, if performed 20/21 (95)

Intraoperative consult obtained, applicable only 8/9 (89)

Hemostasis performed 65 (88)

Patient condition at end of case 59 (80)

Drains left 52 (70)

Estimated blood loss 44 (59)

Complications, intraoperative 24 (32)

Transfusions received 21 (28)

Transfusion type, applicable only 10/66 (15)


GDA = gastroduodenal artery; GI = gastrointestinal. *Unless otherwise indicated.

Table 4. Operative variables, n = 74

Variable Present (%)*

Adhesions described, if present 24/25 (96)

Exploratory laparotomy performed 70 (95)

Frozen section, if performed 14/15 (93)

Tumour location 66 (89)

Tumour extension 64 (86)

Tumour size 38 (52)

Lymphadenopathy 18 (24)

Lymphadenectomy performed, applicable only 29/68 (43)

Clinical resection/margin status reported 22 (30)


*Unless otherwise indicated.

Table 5. Narrative operative report completeness and perioperative outcomes

Variable

Completeness

p value Lowest quartile Highest quartile

Operative duration, mean (range) min

347 (251–495) 396 (unk.–800) 0.048

Length of stay, mean (range) d

9 (5–33) 10 (6–22) 0.96

Perioperative complications, no. (%)

4/16 (20) 6/19 (32) 0.42

Unk. = unknown.


RESEARCH

of 347 (251–495) minutes. There was no associationbetween the completeness of the NR and length of stayin hospital (p = 0.96) or major perioperative complications(p = 0.42).

DISCUSSION

The NRs are the usual form of documentation used torecord the details of a surgical procedure. The purpose ofour study was to examine the completeness of NRs inorder to evaluate their potential as a source of quality as -surance data.

Our results demonstrate that NRs are frequently in -complete. Variables related to surgical technique and admin -istrative details were often present, whereas oncology-related and process of care details were commonly omittedin the majority of NRs. Thus, NRs are a poor source forquality assurance data.

We found that narrator characteristics were associatedwith NR completeness. When we compared the 5 mostand least complete NRs, the 5 most complete NRs were alldictated by the same clinical fellow, whereas the least com-plete NR (13 variables reported, 16%) was dictated by theleast experienced narrator, a senior surgical resident. Thissuggests that narrator training affects the quality of NRs.To date, surgical education places little, if any, emphasis onteaching trainees how to dictate NRs, and as a result thequality of NRs are expectedly variable.21,22

Also, we report on an association between the complete-ness of NRs and operative duration. The longer the pro -ced ure time, the more likely that the NRs were more com-plete. Longer operative durations may be associated withmore complex procedures, suggesting that the complete-ness of the NRs may be associated with procedure com-plexity. This finding is consistent with the results of previ-ous work by Stewart and colleagues,8 who reported thatprocedural quality was directly related to the operativedocumentation. We did not find a significant associationbetween NR completeness and patient-related outcomes,such as length of stay in hospital or perioperative compli-cations. However, our study was underpowered to explorethis question.

There is scant literature on the quality of operativedocumentation in surgery. Edhemovic and colleagues4

demonstrated that the most complete parts of NRs con-tained the least important information (patient informa-tion, indication for the procedure and closure technique).Our findings were similar. Information on elements with-out long-term implications (e.g., incision type, anasto-motic suture technique) was virtually always complete,whereas information on oncology-related variables (e.g.,clinical margin status, extent of lymphadenectomy per-formed) was often omitted.

To our knowledge, our study is the first evaluation ofthe NR for the pancreaticoduodenectomy operation.

Strengths of this study are that the documents analyzedreflect NRs from a large contemporaneous sample of anexperienced group of HPB surgical oncologists with auniform approach to the pancreaticoduodenectomy.Thus, the variations that we identified reflect variations inthe quality of the NRs rather than substantial variationsin the procedure.

Limitations

A limitation of our study is its retrospective nature. In par-ticular, we excluded 4 procedures for which no NR wasdictated. Thus, our results are skewed in the direction ofNRs being more complete than in the real-world setting.

Another limitation of this study is that we did not per-form veracity checks of the variables studied, as NR com-pleteness rather than NR correctness was the focus of thisstudy. Thus, the fidelity of the NR as a reflection of theactual conduct of the procedure could not be assessed inour study. However, other authors have examined this asso-ciation using cognitive task analysis. Stewart and col-leagues8 found differences in the reporting of key steps oflaparoscopic cholecystectomy in the NRs of patients whounderwent laparoscopic cholecystectomy and in whombile duct injury occurred. In patients who had a bile ductinjury — a “bad” outcome — key elements of the surgicalprocedure were omitted, suggesting that the complete-ness of NRs reflects differences in the quality of the pro-cedure performed.

Implications

Owing to the inadequacies of NRs that we have reported,we recommend the use of SRs to complement or replaceNRs as a quality initiative. An SR may improve the com-pleteness and quality of reports by minimizing inconsist -ent, inaccurate or missing information transfer betweencare providers, which can lead to suboptimal patient care.Several studies report that SRs provide more complete in -formation than the NRs.4,23–27 In other specialties, par ticu -larly in pathology, synoptic reporting has been widelyembraced, which has improved interdisciplinary com muni -cation and led to more effective coordination of clinicalcare for individual patients.28–30 There is wide acceptance ofSRs by clinicians who prefer the readability of SRs overNRs.27–30 Further structured synoptic reporting results pro-mote quality by standardizing the reporting processesamong patients and institutions. Several jurisdictions man-date the use of synoptic pathology reports as a perform -ance indicator.23,31–34

Little research has been conducted to evaluate thepotential benefits and/or limitations of synoptic re-porting of operative procedures.2,4–6,8 However, owingto the potential benefits of SRs, many groups are de -velop ing and using SRs at the institutional level as quality


RECHERCHE

improvement strategies. But there are also potentialproblems with SRs.

A major objection is that SRs, with their pro formastructure, may not be flexible enough for some procedures,particularly complex cancer procedures, that do not lendthemselves to standardization. For these cases, SRs maynot be able to accurately reflect the details of the pro -cedure. However, Park and colleagues35 have recently de -veloped and implemented an electronic SR for pancrea -tectomy. They established that an SR is feasible andacceptable to surgeons, even for this complex, multistepprocedure. In their study, the mean time for SR comple-tion was only 4 ± 1.6 minutes per case. Furthermore, theSR document was more complete and reliable than NRs.35

A possible remedy to the structure of the SR is to includean optional free text field. This would allow nonstandardiz-able information to be included within an SR.

Other objections to SRs that have been suggested is thatthey can be difficult to complete, take longer to completethan NRs and add to the surgeon’s workload. However,several studies have reported that SRs take less time tocomplete than NRs.28,35 This suggests that the surgeon’sworkload is actually decreased with SRs. Thus, the percep-tion of SRs being more work is likely related to poorimplementation strategies and/or existing knowledge gapsrather than being an intrinsic property of SRs.

In addition, the associations between SRs and patientoutcomes, such as complication rate, positive margin rateand/or survival, have not yet been established. A few studiessuggest that the quality of documentation is directly relatedto the quality of surgery,8 but the more likely mechanismfor quality improvement is through more efficient com -munication of information among care providers, whichallows for optimal treatment recommendations. However,more research should be done in this area. This topic wasbeyond the scope of this present study, which focused onthe completeness of NRs. However, in future work, we willexamine the association between the format (NR v. SR),quality of documentation and patient-related outcomes.

CONCLUSION

Unstructured NRs for cancer surgery are seldom completeand are of poor quality. Clinically important variables arefrequently missing from NRs. As a result, an NR cannot beused as a data source for quality assurance purposes. Simi-lar considerations also limit the use of the NR for researchand medicolegal applications. Development and use of anSR should be encouraged to improve documentation of careand serve as a measure of quality of surgical care.

Acknowledgements: We thank the HPB Surgeons at the UniversityHealth Network: Bryce Taylor, MD; David Grant, MD; Mark Cattral,MD; Ian McGilvray, MD, PhD; Paul Grieg, MD; Carol-anne Moulton,MD, PhD; Sean Cleary, MD, MSc; and Steven Gallinger, MD, MSc, forcontributing cases, as well as Ms. Marina Englesakis for her assistance

performing literature reviews, Mr. Harden Huang and David Chan forinformatics support and Ms. Stacey Stegienko for administrative support.


Contributors: M.E. Wiebe, L. Sandhu, E.D. Kennedy, N.N. Baxter, A.R.Gagliardi, D.R. Urbach and A.C. Wei designed the study. M.E. Wiebe,L. Sandhu and A.C. Wei acquired the data. M.E. Wiebe, J.L. Takata andA.C. Wei analyzed the data. M.E. Wiebe and A.C. Wei wrote the article.All authors reviewed the article and approved its publication.

References

1. Donabedian A. The quality of medical care. Science 1978;200:856-64.

2. Harvey A, Zhang H, Nixon J, et al. Comparison of data extraction fromstandardized versus traditional narrative operative reports for database-related research and quality control. Surgery 2007;141:708-14.

3. Govindarajan A, Fraser N, Cranford V, et al. Predictors of multivis-ceral resection in patients with locally advanced colorectal cancer.Ann Surg Oncol 2008;15:1923-30.

4. Edhemovic I, Temple WJ, de Gara C, et al. The computer synopticoperative report — a leap forward in the science of surgery. Ann SurgOncol 2004;11:941-7.

5. Elit L, Bondy S, Chen Z, et al. The quality of the operative report forwomen with ovarian cancer in ontario. J Obstet Gynaecol Can 2006;28:892-7.

6. Cowan DA, Sands MB, Rabizadeh S, et al. Electronic templates ver-sus dictation for the completion of Mohs micrographic surgery oper-ative notes. Dermatol Surg 2007;33:588-95.

7. Parikh JA, Yermilov I, Jain S, et al. How much do standardized formsimprove the documentation of quality of care? J Surg Res 2007;143:158-63.

8. Stewart L, Hunter JG, Wetter A, et al. Operative reports: form andfunction. Arch Surg 2010;145:865-71.

9. Urquhart R, Grunfeld E, Porter GA. Synoptic reporting and thequality of cancer care. Oncology Exchange 2009;8:28-31. Available:www.oncologyex.com/gif/archive/2009/vol8_no1/8_review_1.pdf(accessed 2012 Nov. 19).

10. Birkmeyer JD, Finlayson SR, Tosteson AN, et al. Effect of hospitalvolume on in-hospital mortality with pancreaticoduodenectomy.Surgery 1999;125:250-6.

11. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume andsurgical mortality in the United States. N Engl J Med 2002;346:1128-37.

12. Birkmeyer JD, Warshaw AL, Finlayson SR, et al. Relationshipbetween hospital volume and late survival after pancreaticoduodenec-tomy. Surgery 1999;126:178-83.

13. Simunovic M, To T, Theriault M, et al. Relation between hospitalsurgical volume and outcome for pancreatic resection for neoplasmin a publicly funded health care system. CMAJ 1999;160:643-8.

14. Chowdhury MM, Dagash H, Pierro A. A systematic review of theimpact of volume of surgery and specialization on patient outcome.Br J Surg 2007;94:145-61.

15. Gruen RL, Pitt V, Green S, et al. The effect of provider case volumeon cancer mortality: systematic review and meta-analysis. CA CancerJ Clin 2009;59:192-211.

16. Csikesz NG, Simons JP, Tseng JF, et al. Surgical specialization andoperative mortality in hepato-pancreatico-biliary (HPB) surgery. J Gastrointest Surg 2008;12:1534-9.


RESEARCH

17. Eppsteiner RW, Csikesz NG, McPhee JT, et al. Surgeon volumeimpacts hospital mortality for pancreatic resection. Ann Surg 2009;249:635-40.

18. Eppsteiner RW, Csikesz NG, Simons JP, et al. High volume and out-come after liver resection: Surgeon or center? J Gastrointest Surg2008;12:1709-16; discussion 16.

19. Nathan H, Cameron JL, Choti MA, et al. The volume-outcomeseffect in hepato-pancreato-biliary surgery: hospital versus surgeoncontributions and specificity of the relationship. J Am Coll Surg 2009;208:528-38.

20. Schmidt CM, Turrini O, Parikh P, et al. Effect of hospital volume,surgeon experience, and surgeon volume on patient outcomes afterpancreaticoduodenectomy: a single-institution experience. Arch Surg2010;145:634-40.

21. Novitsky YW, Sing RF, Kercher KW, et al. Prospective, blindedevalu ation of accuracy of operative reports dictated by surgical resi-dents. Am Surg 2005;71:627-31, discussion 31-2.

22. Gillman LM, Vergis A, Park J, et al. Structured operative reporting: arandomized trial using dictation templates to improve operativereporting. Am J Surg 2010;199:846-50.

23. Gill A, Johns AL, Eckstein R, et al. Synoptic reporting improveshistopathological assessment of pancreatic resection specimens.Pathology 2009;41:161-7.

24. Cox JL, Zitner D, Courtney KD, et al. Undocumented patient infor-mation: an impediment to quality of care. Am J Med 2003;114:211-6.

25. Langlotz CP. Structured radiology reporting: Are we there yet? Radiology2009;253:23-5.

26. Johnson AJ, Chen MY, Swan JS, et al. Cohort study of structured report-ing compared with conventional dictation. Radiology 2009;253: 74-80.

27. Donahoe L, Bennett S, Temple W, et al. Completeness of dictatedoperative reports in breast cancer-the case for synoptic reporting. J Surg Oncol 2012;106:79-83.

28. Branston LK, Greening S, Newcombe RG, et al. The implementa-tion of guidelines and computerised forms improves the complete-ness of cancer pathology reporting. The CROPS project: a ran-domised controlled trial in pathology. Eur J Cancer 2002;38:764-72.

29. Naik SS, Hanbidge A, Wilson SR. Radiology reports: examiningradio logist and clinician preferences regarding style and content. AJRAm J Roentgenol 2001;176:591-8.

30. Verleye L, Ottevanger PB, Kristensen GB, et al. Quality of pathologyreports for advanced ovarian cancer: are we missing essential inform -ation? An audit of 479 pathology reports from the EORTC-GCG55971/NCIC-CTG OV13 neoadjuvant trial. Eur J Cancer 2011;47:57-64.

31. Hammond EH, Flinner RL. Clinically relevant breast cancer report-ing: using process measures to improve anatomic pathology reporting.Arch Pathol Lab Med 1997;121:1171-5.

32. Austin R, Thompson B, Coory M, et al. Histopathology reporting ofbreast cancer in Queensland: the impact on the quality of reporting asa result of the introduction of recommendations. Pathology 2009;41:361-5.

33. Qu Z, Ninan S, Almosa A, et al. Synoptic reporting in tumor pathol-ogy. Am J Clin Pathol 2007;127:898-903.

34. Leong AS. Synoptic/checklist reporting of breast biopsies: Has thetime come? Breast J 2001;7:271-4.

35. Park J, Pillarisetty VG, Brennan MF, et al. Electronic synoptic opera-tive reporting: assessing the reliability and completeness of synopticreports for pancreatic resection. J Am Coll Surg 2010;211:308-15.

E128 J can chir, Vol. 56, No 5, octobre 2013 © 2013 Association médicale canadienne


The impact of a massive transfusion protocol(1:1:1) on major hepatic injuries: Does it increaseabdominal wall closure rates?

Background: Massive transfusion protocols (MTPs) using high plasma and plateletratios for exsanguinating trauma patients are increasingly popular. Major liver injuriesoften require massive resuscitations and immediate hemorrhage control. Current pub-lished literature describes outcomes among patients with mixed patterns of injury. Wesought to identify the effects of an MTP on patients with major liver trauma.

Methods: Patients with grade 3, 4 or 5 liver injuries who required a massive bloodcomponent transfusion were analyzed. We compared patients with high plasma:redblood cell:platelet ratio (1:1:1) transfusions (2007–2009) with patients injured beforethe creation of an institutional MTP (2005–2007).

Results: Among 60 patients with major hepatic injuries, 35 (58%) underwent resusci-tation after the implementation of an MTP. Patient and injury characteristics weresimilar between cohorts. Implementation of the MTP significantly improved plasma:red blood cell:platelet ratios and decreased crystalloid fluid resuscitation (p = 0.026).Rapid improvement in early acidosis and coagulopathy was superior with an MTP(p = 0.009). More patients in the MTP group also underwent primary abdominal fas-cial closure during their hospital stay (p = 0.021). This was most evident with grade 4injuries (89% vs. 14%). The mean time to fascial closure was 4.2 days. The overall sur-vival rate for all major liver injuries was not affected by an MTP (p = 0.61).

Conclusion: The implementation of a formal MTP using high plasma and plateletratios resulted in a substantial increase in abdominal wall approximation. Thisoccurred concurrently to a decrease in the delivered volume of crystalloid fluid.

Contexte : Les protocoles de transfusion massive (PTM) impliquant des rapportsplasma:plaquettes élevés sont de plus en plus populaires pour traiter les patientsatteints d’un traumatisme hémorragique. Les chirurgies majeures du foie requièrentsouvent le déclenchement de protocoles de transfusion massive et une maîtrise immé-diate de l’hémorragie. La littérature actuelle décrit les résultats chez des patients vic-times de divers types de traumatismes. Nous avons voulu mesurer les effets d’un PTMsur les patients ayant subi un traumatisme majeur au foie.

Méthodes : Nous avons analysé les dossiers de patients ayant subi des blessures aufoie de grade 3, 4 ou 5 qui ont nécessité des transfusions massives de composants san-guins. Nous avons comparé les patients ayant nécessité des transfusions importantesde plasma, de culots globulaires et de plaquettes selon un rapport (1:1:1; 2007–2009)à des patients ayant subi leur traumatisme avant la mise en œuvre d’un PTM parl’établissement (2005–2007).

Résultats : Sur 50 patients ayant subi des lésions hépatiques majeures, 35 (58 %) ontreçu des traitements de réanimation après la mise en place du PTM. Les caractéristiquespropres aux patients et à leurs blessures étaient similaires entre les cohortes. L’applicationdu PTM a significativement amélioré les rapports plasma:culots globulaires:plaquettes etréduit l’administration de cristalloïdes à des fins de réanimation liquidienne (p = 0,026).L’amélioration rapide de l’acidose naissante et de la coagulopathie a été meilleure avec lePTM (p = 0,009). Plus de patients du groupe soumis au PTM ont aussi subi une ferme-ture aponévrotique abdominale primaire durant leur séjour hospitalier (p = 0,021). Celas’est surtout observé avec les lésions de grade 4 (89 % c. 14 %). Le délai moyen avant lafermeture aponévrotique a été de 4,2 jours. L’application du PTM n’a pas modifié le tauxde survie global pour l’ensemble des traumatismes hépatiques majeurs (p = 0,61).

Conclusion : La mise en place d’un PTM officiel reposant sur des rapports plasma etplaquettes élevés a donné lieu à une augmentation substantielle des fermetures de laparoi abdominale. Cela s’est produit en parallèle avec une diminution du volume decristalloïdes administrés pour la réanimation liquidienne.

Chad G. Ball, MD, MSc*Christopher J. Dente, MD*Beth Shaz, MD‡Amy D. Wyrzykowski, MD*Jeffrey M. Nicholas, MD*Andrew W. Kirkpatrick, MD†David V. Feliciano, MD*

From the *Department of Surgery,Emory University, Grady Memorial Hos-pital, Atlanta, Ga.; †Department of Hema-tology, University of Calgary, Calgary,Alta., and ‡Emory University, GradyMemorial Hospital, Atlanta, Ga.

This paper was presented at the 69th

meeting of the American Association forthe Surgery of Trauma in Boston, Mass.

Accepted for publicationMay 28, 2013

Correspondence to:C.G. BallFoothills Medical Centre1403 – 29 St. NWCalgary AB T2N [email protected]

DOI: 10.1503/cjs.020412


RESEARCH

R ecent excitement surrounding the use of massivetransfusion protocols (MTPs) with high plasma andplatelet concentrations for injured patients in

physio logic extremis is substantial.1–20 While the effect onoverall mortality in the civilian population is still de -bated,21–24 massive resuscitations with high plasma:packedred blood cell (RBC) ratios remain promising for address-ing the early coagulopathy25 and acidosis frequently associ-ated with life-threatening injury.10 Additional benefits of aformal MTP include earlier administration of blood prod-ucts during the resuscitation phase, improved overall effi-ciency, decreased total blood product use during a patient’shospital stay and a substantial economic savings.24

Concurrent to the initiation of MTP blood componenttherapy, the concept of damage control resuscitation alsoincorporates principles of reduced crystalloid delivery, per-missive hypotension and immediate operative and/orangiographic hemorrhage control.1–26 This constellation oftechniques is directed at patients who present in physio-logic extremis (pH ≤ 7.1, base deficit ≥ 12.5, and/or coretemperature ≤ 34°C).2,10,26,27 Interestingly, these parametersare nearly identical to the risk factors for the developmentof primary abdominal compartment syndrome (ACS).28–30

As a result of improved recognition of the ACS phenomen -on as well as the widespread application of temporaryabdominal closures (silo) as a preventative measure, theincidence of primary ACS has decreased substantially overthe past 5 years.28,29,31 Unfortunately, the resultant “open”abdomen remains fraught with considerable short andlong-term morbidity.31–35 In the best case scenario thisincludes a poor quality of life and the need for majorreconstructive surgery.32–35

In addition to its effect on acidosis and coagulopathy,MTPs have also been shown to substantially reduce thevolume of crystalloid fluid delivered during the initialresuscitation period.1,15,17 Uncontrolled/excessive resuscita-tion is a clear risk factor for the development of ACS aswell as a major obstacle to obtaining definitive fascial clos -ure of the abdominal wall (visceral edema).28–30,32,35,36 As aresult, it can be postulated that the incidence of both pri-mary ACS and the open abdomen in severely injured

patients may be reduced with the use of a formal 1:1:1 ratioMTP. Anecdotally, this appeared to be particularly evidentin patients with high-grade hepatic injuries at our institu-tion. As a result, the primary goal of our study was to iden-tify the effects of a mature MTP on patients with majorliver injuries by comparing them to a control group whounderwent massive transfusions before initiation of a for-malized high plasma protocol.

METHODS

The primary study population consisted of all patients witha high grade liver injury (grade 3, 4 or 5), who presented toGrady Memorial Hospital (GMH), after the implementa-tion of a formal MTP (Feb. 1, 2007, to Feb. 1, 2009; Table 1).The hospital is a level 1 trauma centre located in an urbansetting. Massive transfusion was defined as transfusion of≥ 10 units of RBCs in any 24-hour period during a pa -tient’s hospital stay. We compared this population with acohort with high-grade hepatic injuries who also under-went a massive transfusion (≥ 10 units of RBCs) prior tothe initiation of the formal MTP (Jan. 1, 2005, to Jan. 31,2007). The massive transfusion prospective registry, thetrauma patient registry and chart reviews supplied all data.Although our institution does not have a formal protocolfor management of the open abdomen, individual clinicalpractice was essentially identical. All management andchallenges were also discussed daily at “Morning Report”by the faculty and senior leadership.

The MTP at GMH is initiated for patients who present inphysiologic extremis (acidosis, coagulopathy, hypothermia) asa result of high-grade injuries. It is designed to ensure im -medi ate availability of aggressive and early component ther-apy and is activated with a phone call to the blood bank. Thisactivation is restricted to an attending physician or fellowfrom the departments of surgery, anesthesia, emergencymedicine or critical care. Efforts are made by clinical person-nel to obtain and deliver a sample of the patient’s blood to theblood bank for blood typing. The blood bank responds to thecall for protocol activation by immediately placing 6 units ofgroup O or type-specific RBCs and 6 units of group AB fresh

Table 1. Massive transfusion protocol package contents*

Package PRBCs Plasma Platelets Cryoprecipitate

Initiation 6 units (UD/TS) 6 units (UD)

1 (0.5 h) 6 units (UD/TS) 6 units (UD) 1 apheresis§

2 (1 h) 6 units (UD/TS) 6 units (TS) 20 units

3 (1.5 h)† 6 units (UD/TS) 6 units (TS) 1 apheresis§

4 (2 h) 6 units (UD/TS) 6 units (TS) 10 units

5 (2.5 h) 6 units (UD/TS) 6 units (TS) 1 apheresis§

6 (3 h)‡ 6 units (UD/TS) 6 units (TS) 10 units

PRBCs = packed red blood cells; TS = type-speci!c; UD = universal donor. *PRBCs and plasma can be doubled to 12 units each per cycle by request. †Recombinant Factor VIIa may be used at attending physician discretion (dose: 3.6 mg, 1 repeat dose as needed in 30 minutes). ‡If MTP is still active, alternate packages identical to packages 5 and 6 until protocol terminated. §A single apheresis unit of platelets is considered to equal 8–10 standard units.


RECHERCHE

frozen plasma (FFP) in a cooler as the “initiation package.”For this purpose, the blood bank maintains an adequateinventory of thawed plasma products for immediate distribu-tion. The blood bank then continues to prepare predesig-nated packages of components to be picked up every 30 min-utes with a goal ratio of RBC:FFP:platelets of 1:1:1 (Table 1).The blood bank continues to issue group O RBCs, but,owing to limited group AB plasma inventory, will issue ABOtype compatible FFP once the patient’s blood type is known.If requested, the blood bank is able to double up the protocolto allow for 12 units of RBCs and 12 units of FFP to bedelivered every 30 minutes. In addition, if bleeding is uncon-trolled, the clinical service can request a 3.6 mg dose ofrFVIIa after package 2 (18 units of RBCs), with an identicalsecond dose, if needed, distributed 30 minutes later. Thecharge nurse in the area of resuscitation is responsible fordesignating a “runner,” who picks up a cooler every 30 min-utes from the blood bank, returns used coolers and deliversproduct to the patient area. In addition to hemorrhage con-trol, the attending physician is responsible for starting andstopping the protocol and for activating rFVIIa use.

The protocol dictates performing coagulation param -eters and blood gases at least every other hour to moni-tor the patient’s response to therapy. The blood bankmedical director, through the transfusion committee ofthe hospital, reviews the MTP quality indicators: 90%or higher percentage of MTP cycles in which bloodproducts are available within 30 minutes and deliveredto the resuscitation area in a timely manner; 100% ofMTPs in which blood typing specimen was received bythe blood bank before the second cycle; 5% or lesswaste of blood products; and 0% incidence of transfu-sion reactions.

Exclusion criteria for the study were limited to patientswho did not undergo a massive transfusion following ahigh-grade liver injury. Liver injuries were graded usingthe American Association for the Surgery of Trauma grading system.37,38


We performed our statistical analyses using Stata version8.0 (Stata Corp). Normally or near-normally distributedvariables are reported as means, and non-normally dis -tribu ted variables are reported as medians. We comparedmeans using the Student t test and medians using theMann–Whitney U test. Differences in proportions amongcategorical data were assessed using the Fisher exact test.We considered results to be significant at p < 0.05 for allcomparisons.

RESULTS

A total of 35 and 25 patients with major liver injuriesunderwent a massive RBC transfusion before and after the

initiation of a formal (1:1:1) MTP, respectively. For allgrades of major hepatic trauma, patient demographics,injury characteristics, mechanisms, initial hemodynamicstatus and presenting base deficits were similar betweenthe groups (Tables 2–4).

The overall survival rate for all patients with majorliver injuries (grades 3, 4, and 5) was not affected by theimplementation of a formal MTP (18 of 35 in the MTPgroup v. 11 of 25 pre-MTP, p = 0.61). Most patients in theMTP cohort died of massive exsanguinating hemorrhageand physiologic exhaustion (33% of grade 3, 86% ofgrade 4, 88% of grade 5 patients). The rate of primaryabdominal fascial closure prior to discharge was signifi-cantly higher in the patient cohort who received a higherFFP:RBC ratio (12 of 18 in the MTP group v. 3 of 11 inthe pre-MTP group, p = 0.02). This was a result of thelarge difference between patients with grade 4 injuries (8in the MTP group v. 1 in the pre-MTP group; Table 3).Of the 6 patients who did not achieve fascial closure priorto discharge, 5 had prolonged mechanical limitations ofthe abdominal wall following a massive crystalloid-basedresuscitation. The remaining patient required multipleoperative interventions for concurrent injuries and dis-played moderate intraperitoneal sepsis as a driving factor.Of the 8 patients with grade 4 liver injuries in the MTP

Table 2. Comparison of patients with grade 3 liver injuries after massive transfusion

Group; mean (%)*

Variable MTP Pre-MTP

Total patients 9 5

Age, median, yr 30 34

Male sex 100 80

Penetrating mechanism 56 80

Injury severity score 31 29

Hemodynamic instability 44 60

Presenting base de!cit –13.3 –16.0

Concurrent injuries 3.3 3.0

Mechanical ventilation, d 16 11

LOS, d 32 38

Overall mortality 3 (33) 3 (60)

Initial damage control procedure 6 (66) 5 (100)

No. of operations among all patients 4 5

Primary abdominal fascial closure 2/6 (33) 1/2 (50)

Achieved by survivors prior to discharge

PRBC:FFP transfusion 1.28 7.4†

PRBC:platelets transfusion 1.12 17†

Total PRBC units, 24 h 26 28

Total PRBC units, < 6 h 25 24

Crystalloid infusion, L 6 9

Factor VIIa 5 (56) 0

Postoperative INR 1.26 1.98

Postoperative base deficit –7.1 –9.7

FFP = fresh frozen plasma; INR = international normalized ratio; LOS = length of stay in hospital; MTP = massive transfusion protocol; PRBC = packed red blood cells. *Unless otherwise indicated. †p < 0.05.


RESEARCH

group who underwent successful abdominal fascial closure during their initial hospital stay, 4 were closedduring the initial operative procedure and 4 underwentprimary fascial approximation at a mean of 4.2 days afteradmission.

Most patients in the MTP group underwent initial peri-hepatic packing (66% of grade 3, 47% of grade 4, 73% ofgrade 5 patients). This was comparable to patients in thepre-MTP group (60% of grade 3, 77% of grade 4, 86% ofgrade 5 patients; p = 0.048).

DISCUSSION

Major hepatic trauma consists of large parenchymal lacer-ations, hematomas, juxtahepatic venous injuries and com-plete hepatic avulsions.37,38 Accordingly, these patientsoften require substantial transfusions, and high associatedmortality correlates with the grade of injury.37 When oper-ative therapy is required, major liver injuries can also bedescribed as some of the most challenging cases. Becauseall current massive transfusion literature describes patientmorbidity and mortality following generalized injuries, theprimary goal of our study was to evaluate the influence ofa high plasma ratio MTP on the outcomes of patientswith major hepatic trauma.

Injured military and civilian patients classically requiremassive transfusion of blood products in approximately 8%and 3% of cases, respectively.39,40 These patients most oftenpresent to the hospital in physiologic extremis.2,15,40 As aresult, acidosis was predictably impressive in our patientcohort, with mean base deficits ranging from –13.3 to –16.4, depending on the hepatic grade of injury (Tables 2–4). The severity of their injuries was also evident in thehigh mean injury severity score (ISS), duration of mechan -ical ventilation and hospital stay as well as the rate ofhemodynamic instability at admission. The observationthat 66%–100% of patients required an emergent damagecontrol operative procedure also highlights the extremis inthis cohort. When taken as a collective, these patients epit-omize the requirement for massive blood product resusci-tation and immediate hemorrhage control via damage con-trol principles.

Although there appears to be a clear reduction in mor-tality for injured soldiers,10,13–15,41 this finding has recentlybeen questioned within the civilian cohort.1,21–24 Prior to theimplementation of a formalized high plasma MTP (Table 1),mortality for patients who required a massive transfusion(≥ 10 units of RBCs) following grade 3, 4 or 5 hepaticinjuries at our institution was 60%, 46% and 71%, re -spectively. Given associated patient ISS of 31, 26 and 29,

Table 4. Comparison of patients with grade 5 liver injuriesafter massive transfusion

Group; mean (%)*


Total patients 11 7

Age, yr median 29 30

Male sex (82) (86)

Penetrating mechanism (64) (71)


Hemodynamic instability (91) (71)




LOS, d 15 10

Overall mortality, no. 8 (73) 5 (71)

Initial damage control procedure, no. 9 (82) 7 (100)





PRBC:platelets transfusion 2.03 3.8

Total PRBC units, 24 h 23 37†

Total PRBC units, < 6 h 17 29†

Crystalloid infusion, L 6 11†

Factor VIIa 2 (18) 2 (29)


Postoperative base deficit –13.4 –14.6

FFP = fresh frozen plasma; INR = international normalized ratio; LOS = length of stay inhospital; MTP = massive transfusion protocols; PRBC = packed red blood cells. *Unless otherwise indicated. †p < 0.05.

Table 3. Comparison of patients with grade 4 liver injuries after massive transfusion

Group; mean (%)*


Total patients 15 13

Age, median, yr 32 27

Male sex (93) (92)

Penetrating mechanism (73) (54)


Hemodynamic instability (80) (77)




LOS, d 29 25

Overall mortality, no. 6 (40) 6 (46)

Initial damage control procedure, no. 13 (87) 13 (100)





PRBC:platelets transfusion 1.23 16*

Total PRBC units, 24 h 31 24

Total PRBC units, < 6 h 28 20

Crystalloid infusion, L 6 13†

Factor VIIa 6 (40) 2 (15)


Postoperative base deficit –4.9 –10.6†

FFP = fresh frozen plasma; INR = international normalized ratio; LOS = length of stay in hospital; MTP = massive transfusion protocol; PRBC = packed red blood cells. *Unless otherwise indicated. †p < 0.05.


RECHERCHE

respectively, as well as at least 3 concurrent injuries amongall cohorts, these survival ratios appear appropriate(Tables 2–4). Unfortunately, mortality in this patientcohort remained unchanged after the implementation of ahigh plasma ratio MTP (33%, 47%, 73%, respectively).This observation was surprising, given the previous identi-fication of an improved overall survival among blunttrauma patients who received an MTP resuscitation at ourtrauma centre.42 Considering the strikingly similar mortal-ity between the pre- and post-MTP cohorts for grade 4and 5 liver injuries, however, we believe this is unlikely tobe the result of sample size bias. This lack of reduction inmortality despite the 1:1:1 massive transfusion has beenpreviously described in mixed cohorts of injured pa -tients.21,24 It remains consistent when the mortality bias iscorrected,23 despite a clear improvement in early coagu-lopathy.22 This area of debate is likely to continue until aprospective randomized trial is completed.1,4–7,9

Although the technique of perihepatic packing wasfounded in 1908 with Pringle’s discussion of hepatictrauma,43 the modern interpretation of this work occurredin the late 1970s,44–46 followed by the concept of truncatedoperations with concurrent intra-abdominal packing forpatients in physiologic extremis by Stone and colleagues in1983.47 This philosophy was then coined “damage control”by Rotondo and colleagues, given its obvious conceptualsimilarity to the Navy’s use of the same term.48,49 Althoughthis concept has resulted in a substantial improvement inmortality when applied to the correct patient population,49

it also commits the patient to a series of subsequent opera-tive procedures aimed at restoring gastrointestinal continu-ity and abdominal wall closure.26 Unfortunately, manypatients are eventually left with “open” abdomens becauseof generalized visceral edema caused by their initial resus-citation and the prevention of ACS. Although theseabdomens are covered by a skin graft, the short and long-term morbidity, economic and resource cost are substantialand mortality is high.32–35,50

The abdominal wall closure rate associated with theimplementation of a 1:1:1 MTP was higher than in patientswho received a low plasma massive transfusion (67% v.27%) despite the use of similar intraoperative techniques.While these rates did not vary in patients with either grade3 or 5 liver trauma, a large improvement was noted inpatients with grade 4 injuries (14% v. 89%; Table 3). Theobserved decrease in crystalloid resuscitation following theuse of an MTP in patients with grade 4 injuries (13 v. 6 L)was also striking. This supports the anecdotal observationthat generalized visceral edema is reduced when plasmaand blood products are delivered via a crystalloid sparingMTP.1,15,17 To this end, 4 patients with grade 4 trauma evenunderwent immediate fascial closure during the initial oper-ative procedure. This compares to a mean wait of 4.2 daysin the remaining patients who had successful approximationof their abdominal wall during the initial hospital stay. We

believe this decrease in both visceral and abdominal walledema played an important role in achieving higher rates ofdefinitive abdominal fascial closure. The 54% reduction to6 L of administered crystalloids is also interesting, giventhat the published threshold for increasing the risk for ACSis 7.5 L.28–30,32,35 The same statistical decrease in crystalloidresuscitation was not observed in patients with grade 3injuries, who failed to show an increase in closure ratesdespite implementation of the formal MTP (Table 2). Thispotential link has recently been suggested elsewhere.1

Unfortunately, most patients with grade 5 hepatic injuriesdied of exsanguinating hemorrhage (Table 4), likely obscur-ing any potential improvement in abdominal closure associ-ated with a crystalloid sparing MTP.

As an MTP quality control measure, we also evaluatedthe actual RBC:FFP ratios delivered to each patient. Forthose with grade 3 liver injuries, the mean RBC:FFP ratiosimproved from 7.4 to 1.28 with the implementation of aformal MTP (Table 2). This pattern was also observed inpatients with grade 4 (from 10.9 to 1.56; Table 3) andgrade 5 (from 6.1 to 1.65; Table 4) injuries; RBC: plateletratios also improved. Furthermore, the utility of the MTPwas particularly evident in the correction of acidosisamong patients with grade 4 injuries (–10.6 to –4.9).

Limitations

Limitations in this study are multiple. First, it was retro-spective; therefore, the possibility of bias cannot be elim -inated. Second, although mortality among patients withgrade 4 and 5 liver injuries pre- and post-MTP werenearly identical, our study was substantially underpoweredto assess overall mortality. While our primary goal was todescriptively evaluate the impact of an MTP on majorliver injuries, small sample size (grade 3 injuries) may haveobscured improvements in mortality. Finally, although anincreased abdominal fascial closure rate was evident,observed decreases in visceral and abdominal wall edemawere anecdotal. As a result, confirmatory abdominal wallmeasurements and intra-abdominal pressures would behelpful in future studies.

CONCLUSION

The implementation of a formal MTP using high plasmaand platelet ratios resulted in a substantial increase inabdominal wall approximation. This occurred concur-rently to a decrease in the volume of crystalloid fluiddelivered during the initial resuscitation for massive hem-orrhage. We hypothesize that this improvement wasrelated to an overall decrease in generalized edema ofboth the viscera and abdominal wall. It was particularlypronounced in patients with grade 4 injuries. Given therapid adoption and initiation of modern 1:1:1 MTPsacross the globe,3 the targeted effects of this strategy


RESEARCH

among civilians with specific organ injuries should be fur-ther defined prospectively.Competing interests: D.V. Feliciano declares having received speakerfees. No other competing interests declared.

Contributors: C.G. Ball, C.J. Dente, B. Shaz, A.D. Wyrzykowski,J.M. Nicholas and D.V. Feliciano designed the study. C.G. Ball, C.J. Dente,B. Shaz, A.D. Wyrzykowski acquired and analyzed the data. D.V. Felicianoalso acquired the data, and A.W. Kirkpatrick also participated in data analy-sis. C.G. Ball, C.J. Dente, B. Shaz, A.D. Wyrzykowski and D.V. Felicianowrote the article, which all authors reviewed and approved for publication.

References

1. Cotton BA, Au BK, Nunez TC, et al. Predefined massive transfusionprotocols are associated with a reduction in organ failure and post -injury complications. J Trauma 2009;66:41-9.

2. Lier H, Krep H, Schroeder S, et al. The influence of acidosis,hypocalcemia, anemia, and hypothermia on functional hemostasis intrauma. J Trauma 2008;65:951-60.

3. Hoyt DB, Dutton RP, Hauser CJ, et al. Management of coagulopathyin the patients with multiple injuries: results from an internationalsurvey of clinical practice. J Trauma 2008;65:755-65.

4. Sperry JL, Ochoa JB, Gunn SR, et al. An FFP:PRBC transfusionratio >/=1:1.5 is associated with a lower risk of mortality after massivetransfusion. J Trauma 2008;65:986-93.

5. Teixeira PG, Inaba K, Shulman I, et al. Impact of plasma transfusionin massively transfused trauma patients. J Trauma 2009;66:693-7.

6. Holcomb JB, Wade CE, Michalek JE, et al. Increased plasma andplatelets to red blood cell ratios improves outcomes in 466 massivelytransfused civilian trauma patients. Ann Surg 2008;248:447-58.

7. Moore FA, Nelson T, McKinley BA, et al. Is there a role for aggres-sive use of fresh frozen plasma in massive transfusion of civiliantrauma patients? Am J Surg 2008;196:948-60.

8. Nunez TC, Voskresensky IV, Dossett LA, et al. Early prediction ofmassive transfusion in trauma: Simple as ABC (assessment of bloodconsumption)? J Trauma 2009;66:346-52.

9. Gunter OL, Au BK, Isbell JM, et al. Optimizing outcomes in damagecontrol resuscitation: identifying blood product ratios associated withimproved survival. J Trauma 2008;65:527-34.

10. Hess JR, Dutton RP, Holcomb JB, et al. Giving plasma at a 1:1 ratiowith red cells in resuscitation: Who might benefit? Transfusion 2008;48: 1763-5.

11. McLaughlin DF, Niles SE, Salinas J, et al. A predictive model formassive transfusion in combat casualty patients. J Trauma 2008; 64:S57-63.

12. Gonzalez EA, Moore FA, Holcomb JB, et al. Fresh frozen plasmashould be given earlier to patients requiring massive transfusion.J Trauma 2007;62:112-9.

13. Borgman MA, Spinella PC, Perkins JG, et al. The ratio of productstransfused affects mortality in patients receiving massive transfusion

at a combat support hospital. J Trauma 2007;63:805-13.

14. Schreiber MA, Perkins J, Kiraly L, et al. Early predictors of massivetransfusion in combat casualties. J Am Coll Surg 2007;205:541-5.

15. Holcomb JB, Jenkins D, Rhee P, et al. Damage control resuscitation:Directly addressing the early coagulopathy of trauma. J Trauma 2007;62:307-10.

16. Malone DL, Hess JR, Fingerhut A. Massive transfusion practicesaround the globe and a suggestion for a common massive transfusionprotocol. J Trauma 2006;60:S91-6.

17. Cotton BA, Gunter OL, Isbell J, et al. Damage control hematology:the impact of a trauma exsanguination protocol on survival and bloodproduct utilization. J Trauma 2008;64:1177-82.

18. Stinger HK, Spinella PC, Perkins JG, et al. The ratio of fibrinogen tored cells transfused affects survival in casualties receiving massivetransfusions at an army combat support hospital. J Trauma 2008; 64:S79-85.

19. Niles SE, McLaughlin DF, Perkins JG, et al. Increased mortalityassociated with the early coagulopathy of trauma in combat casual-ties. J Trauma 2008;64:1459-63.

20. Spinella PC, Perkins JG, Grathwohl KW, et al. Effect of plasma andred blood cell transfusions on survival in patients with combat relatedtraumatic injuries. J Trauma 2008;64:S69-77.

21. Scalea TM, Bochicchio KM, Lumpkins K, et al. Early aggressive useof fresh frozen plasma does not improve outcome in critically injuredtrauma patients. Ann Surg 2008;248:578-84.

22. Kashuk JL, Moore EE, Johnson JL, et al. Postinjury life threateningcoagulopathy: Is 1:1 fresh frozen plasma:packed red blood cells theanswer? J Trauma 2008;65:261-70.

23. Snyder CW, Weinberg JA, McGwin G, et al. The relationship ofblood product ratio to mortality: Survival benefit or survival bias?J Trauma 2009;66:358-62.

24. O’Keeffe T, Refaai M, Tchorz K, et al. A massive transfusion protocolto decrease blood component use and cost. Arch Surg 2008;143:686-90.

25. Brohi K, Singh J, Heron M, et al. Acute traumatic coagulopathy.J Trauma 2003;54:1127-30.

26. Wyrzykowski AD, Feliciano DV. Trauma damage control. In: Feli-ciano DV, Mattox KL, Moore EE, editors. Trauma. 6th ed. New York(NY): McGraw-Hill Medical; 2008. p. 851-70.

27. Cushman JG, Feliciano DV, Renz BM, et al. Iliac vessel injury: opera-tive physiology related to outcome. J Trauma 1997;42:1033-40.

28. Balogh ZJ, van Wessen K, Yoshino O, et al. Postinjury abdominalcompartment syndrome: Are we winning the battle? World J Surg2009; 33:1134-41.

29. Balogh Z, McKinley BA, Holcomb JB, et al. Both primary and sec-ondary abdominal compartment syndrome (ACS) can be predictedearly and are harbingers of multiple organ failure. J Trauma 2003; 54:848-59.

30. Ball CG, Kirkpatrick AW. Intra-abdominal hypertension andthe abdominal compartment syndrome. Scand J Surg 2007;96:197-204.

31. Offner PJ, de Souza AL, Moore EE, et al. Avoidance of abdominalcompartment syndrome in damage-control laparotomy after trauma.Arch Surg 2001;136:676-81.

32. Balogh Z, Moore FA, Goettler CE. Surgical management of theabdominal compartment syndrome. In: Ivatury RR, Cheatham ML,Malbrain MLNG, et al., editors. Abdominal Compartment Syndrome.Georgetown (DC): Landes Biomedical; 2006. p. 266-9.

33. Cheatham ML, Safcsak K, Llerena LE, et al. Long-term physical,mental, and functional consequences of abdominal decompression.J Trauma 2004;56:237-41.

34. Cheatham ML, Safcsak K. Long-term impact of abdominal decom-pression: a prospective comparative analysis. J Am Coll Surg 2008;207: 573-9.

35. Ball CG, Kirkpatrick AW, McBeth P. The secondary abdominal com-partment syndrome: not just another post-traumatic complication.Can J Surg 2008;51:399-405.

36. Pruitt BA Jr. Protection from excessive resuscitation: “pushing thependulum back”. J Trauma 2000;49:567-8.

37. Tinkoff G, Esposito TJ, Reed J, et al. American Association for theSurgery of Trauma organ injury scale I: spleen, liver and kidney, vali-dation based on the National Trauma Data Bank. J Am Coll Surg2008;207:646-55.

38. Moore EE, Shackford SR, Pachter HL, et al. Organ injury scaling:spleen, liver, and kidney. J Trauma 1989;29:1664-6.

39. Como JJ, Dutton RP, Scalea TM, et al. Blood transfusion rates in the

care of acute trauma. Transfusion 2004;44:809-13.

40. Holcomb JB. Damage control resuscitation. J Trauma 2007;62:S36-7.

41. Dente CJ, Shaz B, Nicholas JM, et al. Improvements in early mortal-ity and coagulopathy are sustained better in blunt trauma patientsafter institution of a massive transfusion protocol in a civilian level Itrauma center. J Trauma 2009;66:1616-24.

42. Hess JR, Brohi K, Dutton RP, et al. The coagulopathy of trauma: areview of mechanisms. J Trauma 2008;65:748-54.

43. Pringle JH. V. Notes on the arrest of hepatic hemorrhage due totrauma. Ann Surg 1908;48:541-9.

44. Lucas CE, Ledgerwood AM. Prospective evaluation of hemostatictechniques for liver injuries. J Trauma 1976;16:442-51.

45. Calne RY, McMaster P, Pentlow BD. The treatment of major livertrauma by primary packing with transfer of the patient for definitivetreatment. Br J Surg 1979;66:338-9.

46. Feliciano DV, Mattox KL, Jordan DL. Intra-abdominal packing forcontrol of hepatic hemorrhage: a reappraisal. J Trauma 1981;21:285-90.

47. Stone HH, Strom PR, Mullins RJ. Management of the major coagu-lopathy with onset during laparotomy. Ann Surg 1983;197:532-5.

48. Rotondo MF, Schwab CW, McGonigal MD, et al. “Damage control”:an approach for improved survival in exsanguinating penetratingabdominal injury. J Trauma 1993;35:375-83.

49. Rotondo MF, Zonies DH. The damage control sequence and under-lying logic. Surg Clin North Am 1997;77:761-77.

50. Stone HH, Fabian TC, Turkleson ML, et al. Management of acutefull-thickness losses of the abdominal wall. Ann Surg 1981;193:612-8.

RECHERCHE




Differences between referred and nonreferredpatients in cancer research

Background: In Canada, provincial cancer registries have been established to providerigorous population-based data for patients with colorectal cancer. Databases main-tained by regional cancer agencies contain a broader scope of information and havebeen used as a surrogate source of information for colorectal cancer research. It isunclear whether these data can be reliably extrapolated to all patients affected by colo -rectal cancer. We sought to determine whether patients included in a referral-baseddatabase are systematically different from patients who are not included.

Methods: We conducted a retrospective cohort study to compare patients referred tothe British Columbia Cancer Agency with those who were not referred. Comparisonwas based on age, sex and geographic location. We used univariate and logistic regres-sion analysis to identify significant differences between the cohorts.

Results: Univariate analysis demonstrated that the referral and nonreferral cohortsdiffered in sex, age and geographic location. For patients with rectal cancer, the refer-ral and nonreferral cohorts varied in age and geographic location. Multivariate analy-sis demonstrated significant differences in age and geographic location but not sex forpatients with colon and rectal cancer.

Conclusion: Patients included in the referral database differed in age and geographiclocation from those included only in the provincial database. Studies using large datasets from referral centres must be interpreted with caution and may not be represent -ative of the entire patient population.

Background : Au Canada, on a établi des registres provinciaux en oncologie pourgénérer des données représentatives rigoureuses au sujet des patients atteints de cancercolorectal. Les bases de données maintenues par les agences régionales du cancer con -tiennent un éventail plus large de renseignements et ont servi de source de données desubstitution pour la recherche sur le cancer colorectal. Or, on ignore s’il est possibled’extrapoler ces données de manière fiable à tous les patients atteints de cancer colorec-tal. Nous avons voulu déterminer si les patients inclus dans une base de données deréférence sont systématiquement différents des patients qui n’y figurent pas.

Méthodes : Nous avons procédé à une étude de cohorte rétrospective pour comparerles patients référés à l’agence de lutte contre le cancer de la Colombie-Britannique àceux qui n’y avaient pas été référés. La comparaison reposait sur l’âge, le sexe et l’em-placement géographique. Nous avons utilisé une analyse de régression univariée etlogistique pour dégager les différences significatives entre les cohortes.

Résultats : L’analyse univariée a démontré que les cohortes référée et non référéedifféraient aux plans du sexe, de l’âge et de l’emplacement géographique. Pour lespatients atteints d’un cancer rectal, les cohortes référée et non référée variaient selonl’âge et l’emplacement géographique. L’analyse multivariée a révélé des différencessignificatives aux plans de l’âge et de l’emplacement géographique, mais non au plandu sexe en ce qui concerne les patients atteints de cancer du côlon et du rectum.

Conclusion : Les patients inclus dans la base de données de référence étaient dif-férents de ceux qui ne figuraient que dans la base de données provinciale, pour ce quiest de l’âge et de l’emplacement géographique. Il faut interpréter avec prudence lesétudes reposant sur d’importantes séries de données provenant de centres de référence,car elles pourraient ne pas être représentatives de toute la population de patients.

C olorectal cancer is a leading cause of cancer-related death in NorthAmerica.1,2 Worldwide, the annual incidence of colorectal cancer hasbeen estimated at 780 000 to 1 million cases, of which about

20 000 cases will be diagnosed in Canada.2–4 To monitor the burden of colorectal

Jason Faulds, MD*

Colleen E. McGahan, MSc*†

P.T. Phang, MD*

Manoj J. Raval, MD, MSc*

Carl J. Brown, MD, MSc*

From the *Department of Surgery, St. Paul’s Hospital and University ofBritish Columbia, the †Surgical Oncology Network, BC Cancer Agency,and ‡Cancer Surveillance & Outcomes,BC Cancer Agency, Vancouver, BC

Accepted for publicationJune 25, 2012

Correspondence to:C.J. Brown1081 Burrard St.Vancouver BC V6Z [email protected]

DOI: 10.1503/cjs.027511


cancer on population health, provincial cancer registrieshave been established.5

To best treat patients with cancer, specialized cancercentres are established in most jurisdictions. In BritishColumbia, branches of the British Columbia CancerAgency (BCCA) are the sole providers of radiotherapy anda major provider of chemo therapy for citizens of theprovince. Research based at these cancer centres is oftenused to guide cancer treatment for all patients with cancerbased on the assumption that the data can be extrapolatedto all patients. However, it is unclear whether these datacan be used as a substitute for population-based cancerregistries owing to health care access bias.6 The primarygoal of the present study was to determine whetherpatients with colorectal cancer who are referred to theBCCA are systematically different from those who are notreferred and are therefore not captured within the canceragency database.

METHODS

Cancer databases in British Columbia

In British Columbia, a registry of all patients with a con-firmed tissue diagnosis of cancer is maintained by theBritish Columbia Cancer Registry (BCCR) and is man-dated by provincial law. The scope of data containedwithin the registry is limited, and no stage-specific dataare recorded. To address this, the BCCA established theGastrointestinal Cancer Outcomes Unit in 2002. Thisunit maintains a colorectal cancer outcomes database(CRC database), which captures additional disease infor-mation, including stage, and treatment information for allpatients with colorectal cancer who have been referred tothe BCCA. The CRC database has been used in previouscancer outcome research and has been considered a rea-sonable approximation of population-based data.7–10

Data set comparison

We obtained institutional ethics review board approvalfrom the BCCA. We queried the BCCR to identify allpatients in the province with colon, rectosigmoid and rec-tal cancer diagnosed between January 2002 and December2004. Patients included in the BCCR represent the entirepopulation of patients with colorectal cancer in BritishColumbia during the study period. The CRC databasesolely captures data for patients who have been referred toand used the BCCA. Cross-referencing the BCCR andCRC databases formed a cohort of patients who had usedthe BCCA (the referred cohort) and a second cohort ofpatients who had not used the BCCA (nonreferred cohort).

We collected demographic information, including age,sex and regional health authority (representing geographicregion within British Columbia), for all patients in each

cohort. All adult patients who were residents of BritishColumbia at the time of colorectal cancer diagnosis wereincluded in the study. Patients who were identified as hav-ing rectosigmoid cancer were grouped with patients withcolon cancer for the analysis. The rationale for combiningpatients with rectosigmoid and colon cancer was that these2 groups of patients were most frequently treated withsurgery and chemotherapy and rarely received radiationtherapy, suggesting these were not true rectal cancers. Wedetermined and compared rates of chemotherapy betweencohorts by reviewing the records of the BCCA pharmacy,which maintains a record of all chemotherapy administeredwithin the province.


We initially recorded data in contingency tables and per-formed univariate analysis. Statistical analysis involvedStudent t tests to compare continuous variables and the χ2

test to assess categorical variables. We considered a 2-sided p value of 0.05 to be statistically significant for allcomparisons. To account for confounding between vari-ables, we performed logistic regression analysis. Thecovariates selected for analysis were limited to those forwhich data were available from both the provincial cancerregistry and the Gastrointestinal Cancer Outcomes Unit.Age at diagnosis, sex and regional health authority werethe only covariates for which comparative data were avail-able. Age was entered as a continuous variable, whereas sexand health authority were analyzed as categorical vari-ables. We used regression coefficients to determine theodds ratio for capture in the CRC database for eachcovariate entered in the model.

RESULTS

We identified 7305 cases of colorectal carcinoma diagnosedbetween 2002 and 2004 in the BCCR. There were 6749 casesof colorectal adenocarcinoma that met the inclusion criteriaand were included in the analysis. The referred cohort com-prised 3651 patients from the BCCR who had used theBCCA. The remaining 3098 patients from the BCCR hadnot used the BCCA and were not in the CRC database, thusforming the nonreferred cohort (Fig. 1).

There were 4940 patients with colon cancer diagnosedduring the years of the study. Overall, we included 2302(46.6%) patients with colon cancer in the referred cohortand 2638 (53.4%) in the nonreferred cohort. A total of1809 patients with rectal cancer were identified, with 1349(74.6%) in the referred and 460 (25.4%) in the nonre-ferred cohorts.

For patients with colon cancer, univariate analysisdemonstrated significant differences in sex, age and geo-graphic location. As demonstrated in Table 1, 1071 (44.8%)women with colon cancer were referred to the BCCA

RECHERCHE


RESEARCH

compared with 1231 (48.2%) men (p = 0.017). Withincreasing age, the likelihood of referral to the BCCAdecreased; 44 (75.9%) patients younger than 40 years wereincluded in the referred cohort compared with 327 (25.0%)of octogenarians (p < 0.001). Referral rates ranged from 61(26.6%) to 689 (60.3%) based on geographic locationwithin the province (p < 0.001). The percentage of patientsreferred to the BCCA increased from 2002 to 2003; a simi-lar increase was not observed between 2003 and 2004.

The univariate analysis for patients with rectal cancershowed significant differences between cohorts withrespect to age and regional health authority (Table 2).There was no difference in rates of referral by sex (p = 0.23).In this group, 855 (75.5%) men and 494 (73.0%) womenwith rectal cancer were included in the referred cohort.With respect to age, a similar trend to that seen amongpatients with colon cancer emerged; as age increased, ratesof referral decreased. Of patients aged younger than40 years, 24 (85.7%) were included in the CRC databasecompared with 184 (54.3%) of patients older than 80 years.Referral rates ranged from 74 (67.3%) to 304 (81.7%)based on geographic location within the province(p = 0.007). The rate of patient inclusion in the CRC data-base increased from 2002 to 2004.

The results of the logistic regression analysis for pa -tients with colon cancer revealed no significant differencein rate of in clusion in the provincial cancer agency data-base based on sex, accounting for geographic location and

Potential cases of colorectal cancer

identified, n = 7305

Patients included in the analysis, n = 6749

Referral cohort, n = 3651 (54.1%)

Nonreferral cohort, n = 3098 (45.9%)

Colon cancer, n = 2302 (46.6%)

Rectal cancer, n = 1349 (74.6%)

Colon cancer, n = 2638 (53.4%)

Rectal cancer, n = 460 (25.4%)

Excluded • Patients who were not

BC residents or whose cancers were not colorectal, n = 57

• Patients with carcinoma in situ, n = 303

• Patients with histology other than adenocarcinoma, n = 119

• Diagnosis established postmortem, n = 77

Fig. 1. Study population.

Table 1. Univariate comparison between patients with colon cancer in the nonreferral cohort and the referral cohort

Characteristic

Cohort; no. (%)

p value Nonreferral Referral

Total no. patients with colon cancer

2638 (53.4) 2302 (46.6)

Sex 0.017

Female 1317 (55.2) 1071 (44.8)

Male 1321 (51.8) 1231 (48.2)

Age, yr < 0.001

< 40 14 (24.1) 44 (75.9)

40–49 74 (32.7) 152 (67.3)

50–59 261 (39.5) 400 (60.5)

60–69 420 (40.8) 609 (59.2)

70–79 888 (53.6) 770 (46.4)

≥ 80 981 (75.0) 327 (25.0)

Health authority < 0.001

Fraser 854 (57.0) 643 (43.0)

Interior 462 (48.8) 485 (51.2)

Northern 168 (73.4) 61 (26.6)

Vancouver Coastal 682 (61.8) 422 (38.2)

Vancouver Island 454 (39.7) 689 (60.3)

Year

2002 903 (55.9) 712 (44.1)

2003 841 (51.4) 796 (48.6)

2004 894 (53.0) 794 (47.0)

Table 2. Univariate comparison of patients with rectal cancer in the nonreferral cohort and the referral cohort

Characteristic

Cohort; no. (%)

p value Nonreferral cohort Referral cohort

Total no. patients with rectal cancer

460 (25.4) 1349 (74.6)

Sex 0.23

Female 183 (27.0) 494 (73.0)

Male 277 (24.5) 855 (75.5)

Age, yr < 0.001

< 40 4 (14.3) 24 (85.7)

40–49 21 (15.9) 111 (84.1)

50–59 42 (14.2) 254 (85.8)

60–69 86 (17.8) 396 (82.2)

70–79 152 (28.6) 380 (71.4)

≥ 80 155 (45.7) 184 (54.3)

Health authority 0.007

Fraser 139 (26.5) 385 (73.5)

Interior 102 (26.6) 282 (73.4)

Northern 36 (32.7) 74 (67.3)

Vancouver Coastal 111 (26.7) 304 (73.3)

Vancouver Island 68 (18.3) 304 (81.7)

Year

2002 181 (29.8) 426 (70.2)

2003 143 (23.7) 460 (76.3)

2004 136 (22.7) 463 (77.3)


RECHERCHE

age (p = 0.33; Table 3). Variation in rates of referral based ongeographic location was noted (p < 0.001). The interaction ofage at diagnosis and health region was significant, indicatingthat rates of referral by age were different among the regions.

For patients with rectal cancer, the logistic regressionanalysis demonstrated no significant difference in rate ofreferral based on sex while adjusting for health region andage at diagnosis (p = 0.50; Table 4). For every 5-year increasein age above 60 years at diagnosis, referral rates were esti-mated to decrease by 30% (p < 0.001). Accounting for age atdiagnosis and sex, significant differences in rates of referralwere noted for different geographic regions (p < 0.001).

Across each health authority, patients in the referredcohort were more likely to receive chemotherapy thanpatients in the nonreferred cohort (Table 5). For patientsin the nonreferred cohort, the rate of chemotherapy washighest in the Northern Health Authority and the Vancou-ver Coastal Health Authority for patients with colon andrectal cancer. Figures 2 and 3 demonstrate the variation inchemotherapy rates for patients with colon and rectal can-cers, respectively, across regional health authorities in theprovince of British Columbia.

DISCUSSION

A major strength of population-based research is theexternal validity of conclusions based on study results.11

The collection of population-based, stage-specific, cancerdata requires substantial resources and is currently notavailable for patients with colorectal cancer within theprovince of British Columbia. Therefore, much of theresearch regarding colorectal cancer in British Columbia

has relied on analysis of data collected on patients referredto the BCCA.7–10 Referral-based data can be a robustsource of information for health care research because ofthe broad scope of data collected. However, conclusionsbased exclusively on analysis of referral-based data aresubject to health care access bias, whereby referredpatients may not be representative of the entire patientpopulation.6,12

Our study demonstrates that patients referred to theBCCA, and therefore captured in the CRC database, aresystematically different from those patients who are notreferred. The results of this study are important in thatthey demonstrate that patients of advanced age and thoseliving in different regions of the province are not uni-formly included in the CRC database. Conclusions basedsolely on analysis of referral-based data are therefore notnecessarily generalizable to the entire population of pa -tients with colorectal cancer in the province.

Previous studies have demonstrated conflicting results indetermining the effect of sex on referral to regional cancercentres.13–15 In our study, the univariate analysis suggested

Table 3. Logistic regression analysis for patients with colon cancer demonstrating the adjusted effects of each covariate on rate of referral to the BCCA

Effect

Odds ratio (95% CI) for inclusion in the CRC

database cohort p value

Age at diagnosis < 0.001

Odds of referral for a 5-yr increase in age

Northern 0.86 (0.85–8.86)

Vancouver Island 0.70 (0.69–0.70)

Fraser 0.77 (0.77–0.77)

Vancouver Coastal 0.77 (0.77–0.78)

Interior 0.76 (0.76–0.77)


Odds of referral with mean age 70 yr

Northern 1

Vancouver Island 5.880 (4.176–8.279)

Fraser 2.304 (1.656–3.206)

Vancouver Coastal 1.835 (1.309–2.572)

Interior 3.366 (2.396–4.729)

Sex 1.061 (0.941–1.197) 0.33

Interaction between age at diagnosis and health authority 0.021

BCCA = BC Cancer Agency; CI = confidence interval; CRC = colorectal cancer outcomes database.

Table 5. Chemotherapy treatment by health authority for patients in the referred and nonreferred cohorts

Cancer; health authority

Cohort; % who received chemotherapy

p value BCCR only CRC database

Colon cancer < 0.001

Fraser 15.2 59.9

Interior 18.2 55.9

Northern 29.8 62.3

Vancouver Coastal 27.9 62.8

Vancouver Island 1.8 38.8

Rectal cancer < 0.001

Fraser 5.0 54.8

Interior 10.9 55.3

Northern 22.2 62.2

Vancouver Coastal 18.9 55.3

Vancouver Island 1.5 30.6

BCCR = BC Cancer Registry; CRC = colorectal cancer outcomes database.

Table 4. Logistic regression analysis demonstrating the adjusted effects of each covariate on rate of referral to the BCCA for patients with rectal cancer

Factor

Odds ratio (95% CI) for inclusion in the CRC

database cohort p value

Increase in 5-yr for age at diagnosis 0.771 (0.733–0.811) < 0.001


Northern 1

Vancouver Island 2.829 (1.722–4.648) < 0.001

Fraser 1.601 (1.010–2.537) 0.045

Vancouver Coastal 1.681 (1.048–2.698) 0.031

Interior 1.652 (1.027–2.657) 0.039

Sex 1.082 (0.862–1.359) 0.50

BCCA = BC Cancer Agency; CI = confidence interval; CRC = colorectal cancer outcomes database.


RESEARCH

that rates of referral to the BCCA were significantly lowerfor women with colon cancers than men. However, whenthese data were corrected for age at diagnosis and healthcare region, the effect of sex was no longer significant. Forpatients with rectal cancer, there was no significant differ-ence in rates of capture based on sex in the univariate orlogistic regression analysis.

Our study demonstrates that with advanced age, referralto the BCCA decreased. This finding is consistent withprevious research showing that patients with advanced agewere less likely to be referred to regional cancer centres formultidisciplinary assessment and treatment and that evenwhen referred they were less likely to use available cancerresources than younger patients.13,16 As nonreferred patientsdo not have stage-specific data, it was not possible to deter-mine whether lower rates of referral with advanced age wasowing to stage, comorbidity or physician and patient atti-tudes. Recent research has shown a trend toward moreadvanced age at initial colorectal cancer diagnosis.16 Thus,research based on referral centres may not appropriatelyinform treatment of elderly patients.

Within the province of British Columbia, regionalhealth authorities have been established to administerhealth care resources over the large geographic area of theprovince. In our analysis, we determined that rates of refer-

ral to the BCCA were significantly different among theprovince’s 5 regional health authorities. Previous researchhas determined that geographic barriers prevent effectiveuse of available health care resources.16–18 Differences inaccess to cancer care have been shown to vary with ruraland urban status.18 Previous reports have demonstratedthat patients living in rural areas have delayed initiation ofchemotherapy and radiation therapy compared with thoseliving in urban areas.19 Travel distance invariably con-tributes to the discrepancy.20 Rural status may influenceaccess to cancer screening, cancer care and follow-up. Fur-ther research can help quantify the effect of rural status oncolorectal cancer outcomes within British Columbia. Toserve the vast geographic area of the province, the BCCAhas developed satellite sites at regional and communityhospitals where patients may access adjuvant therapy.Despite these efforts, the main centres of the BCCA andmost of the resources are located in major urban areas.Patients living in more remote regions of the province maybe less likely to access these resources and would be lesslikely to be captured within the CRC database. In additionto potential geographic barriers to assessment at regionalcancer centres, the available resources in each healthauthority may influence rates of BCCA referral and inclu-sion in the CRC database. Oncologists not affiliated with

BC chemotherapy utilization 2000–2004: rectal cancer

1 = East Kootenay 2 = Kootenay Boundary 3 = Okanagan 4 = Thompson/Cariboo/Shuswap 5 = Fraser East 6 = Fraser North 7 = Fraser South 8 = Richmond 9 = Vancouver

10 = North Shore/Coast/Garabaldi 11 = South Vancouver Island 12 = Central Vancouver Island 13 = North Vancouver Island 14 = North West 15 = Northern Interior 16 = North East

1 2 3

4

5 6 7

6

7 8

9

10

11

12

13

14

15

16

Chemotherapy utilization

<30%

30%–39%

40%–49%

≥ 50%

Fig. 3. Geographic variation in chemotherapy rates for patientswith rectal cancer in the province of British Columbia.

BC chemotherapy utilization 2000–2004: colon cancer

1 = East Kootenay 2 = Kootenay Boundary 3 = Okanagan 4 = Thompson/Cariboo/Shuswap 5 = Fraser East 6 = Fraser North 7 = Fraser South 8 = Richmond 9 = Vancouver

10 = North Shore/Coast/Garabaldi 11 = South Vancouver Island 12 = Central Vancouver Island 13 = North Vancouver Island 14 = North West 15 = Northern Interior 16 = North East

1 2

3

4

5

6

6

7

7

8

9

10

11

12

13

14

15

16

Chemotherapy utilization

< 30% 30%–39%

40%–49%

≥ 50%

Fig. 2. Geographic variation in chemotherapy rates for patientswith colon cancer in the province of British Columbia.


RECHERCHE

the BCCA manage patients with colorectal cancer and pre-scribe chemotherapy in certain health authorities; theseregions may have lower rates of patient inclusion in theCRC database but similar overall therapeutic options forpatients with colon cancer. Our results suggest potentialgeographic influence on referral to regional cancer centres,but further research is required to test this hypothesis.

Previous studies have noted that rates of referral to re -gion al cancer centres have increased over time, possiblyre lated to increased compliance with recommendationsfor chemo- and radiotherapy for certain stages of colonand rec tal cancer, or perhaps owing to educational pro-grams.15,21,22 We noted that referral rates varied over the 3-year period of our study, but an increased rate of referralwas not observed. To determine conclusively whetherpatient capture into the CRC database has increased overtime, analysis over a longer period is required.

As expected, patients referred to the BCCA had higherrates of chemotherapy across all health care authorities inthe province. For patients who were not referred to theBCCA, rates of chemotherapy were highest in the North-ern Health Authority and the Vancouver Coastal HealthAuthority. The higher rates of chemotherapy among nonre-ferred patients within the Northern and Vancouver CoastalHealth Authorities may be explained by medical oncologistsnot affiliated with the BCCA administering adjuvant ther-apy in these regions. As information on cancer stage isunavailable for the patients included only in the BCCR, itwas not possible to determine the effect of cancer stage onchemotherapy rates. The low rates of chemotherapy in theBCCR cohort may be explained by patients having early-stage disease and therefore not requiring chemotherapy.This is unlikely, and previous research has demonstratedthat substantial numbers of patients with stage III coloncancer were not referred for chemo therapy.23 The inabilityto collect stage-specific information from the provincialcancer registry limits the ability to interpret the differencein chemotherapy rates between the 2 cohorts. The BCCR iscurrently involved in the Canadian Partnership AgainstCancer National Cancer Staging Initiative, which has amandate to capture population-based stage information formajor cancers, including colorectal cancer, for cases diag-nosed from 2010 onward. The implementation of this in -itiative in British Columbia is planned for late 2011.

Limitations

The results of this study are subject to potential limita-tions that must be considered when interpreting ourresults. Incorrect entry of patient information or miscod-ing of diagnosis is possible and can lead to misclassifica-tion bias. The study was limited to a single Canadianprovince, and the results may be not be generalizable toother provinces. As stage-specific information is not avail-able for patients in the BCCR cohort, it was not possible

to determine the influence of cancer stage on patientinclusion in the CRC database.

CONCLUSION

Our results demonstrate that patients included in thereferral database are systematically different from theoverall population of patients with colorectal cancer inBritish Columbia with respect to age and geographic loca-tion. Patients of advanced age were less likely to be in -cluded in the CRC database than younger patients. Inaddition, geographic variation in rates of inclusion withinthe CRC database was demonstrated for both colon andrectal cancer. Studies using large data sets at tertiary refer-ral centres, even in places that have consolidated care atthese centres, must be interpreted with caution. Ourresults suggest that health care access bias related to pa -tient age and location within the province influences thevalidity of the referral database. Expansion of the provin-cial cancer registry to include stage-specific information aspart of the National Staging Initiative will greatly improvethe ability to analyze care provided to all patients withcolo rectal cancer and help with health care resource man-agement and clinical research.

Competing interests: None delcared.

Contributors: J. Faulds, P.T. Phang and C.J. Brown designed the study.C.E. McGahan and C.J. Brown acquired and analysed the data, which J. Faulds, P.T. Phang and M.J. Raval also analysed. J. Faulds, M.J. Ravaland C.J. Brown wrote the article, which C.E. McGahan, P.T. Phang,M.J. Raval and C.J. Brown reviewed. All authors approved the final ver-sion for publication.

References

1. Sewitch MJ, Fournier C, Ciampi A, et al. Adherence to colorectalcancer screening guidelines in Canada. BMC Gastroenterol 2007;7:39.

2. Smiljanic S, Gill S. Patterns of diagnosis for colorectal cancer:screening detected vs. symptomatic presentation. Dis Colon Rectum2008;51:573-7.

3. Kamangar F, Dores G, Anderson W. Patterns of cancer incidence,mortality, and prevalence across five continents: defining priorities toreduce cancer disparities in different geographic regions of the world.J Clin Oncol 2006;24:2137-50.

4. Boyle P, Langman J. ABC of colorectal cancer epidemiology. BMJ2000;321:805-8.

5. Parkin DM. The evolution of the population-based cancer registry.Nat Rev Cancer 2006;6:603-12.

6. Delgado-Rodríguez M, Llorca J. Bias. J Epidemiol Community Health2004;58:635-41.

7. Phang PT, Kennecke H, McGahan C, et al. Predictors of positiveradial margin status in a population-based cohort of patients with rectalcancer. Curr Oncol 2008;15:98-103.


RESEARCH

8. Renouf D, Kennecke H, Gill S. Trends in chemotherapy utilizationfor colorectal cancer. Clin Colorectal Cancer 2008;7:386-9.

9. Pao J, Woods R, Raval M, et al. A population-based study of trendsand outcomes in patients with rectal cancer treated with abdomino -perineal resection in British Columbia between 2000 and 2006.[abstract] Can J Surg 2009;52(Suppl):S45.

10. Ho C, Ng K, O’Reilly S, et al. Outcomes in elderly patients withadvanced colorectal cancer treated with capecitabine: a population-based analysis. Clin Colorectal Cancer 2005;5:279-82.

11. Szklo M. Population-based cohort studies. Epidemiol Rev 1998;20:81-90.

12. Palla S, Farella M. External validity: A forgotten issue? J Orofac Pain2009; 23:297-8.

13. Etzioni DA, El-Khoueiry A, Beart R. Rates and predictors ofchemotherapy use for stage iii colon cancer: a systematic review. Cancer2008; 113:3279-89.

14. Mahoney T, Kuo Y, Topilow D, et al. Stage III colon cancers: whyadjuvant chemotherapy is not offered to elderly patients. Arch Surg2000; 135:182-5.

15. Luo R, Giordano S, Freeman J, et al. Referral to medical oncology: acrucial step in the treatment of older patients with stage III coloncancer. Oncologist 2006;11:1025-33.

16. Oliveria SA, Yood M, Campbell U, et al. Treatment and referral pat-

terns for colorectal cancer. Med Care 2004;42:901-6.

17. Haas JS, Brawarsky P, Iyer A, et al. Association of area sociodemo-graphic characteristics and capacity for treatment with disparities incolorectal cancer care and mortality. Cancer 2011;117:4267-76.

18. Singh GK, Williams SD, Siahpush M, et al. Socioeconomic, rural-urban, and racial inequalities in US cancer mortality: Part I — Allcancers and lung cancer and Part II — colorectal, prostate, breast,and cervical cancers. J Cancer Epidemiol 2011;2011:107497.

19. Yu XQ. Socioeconomic disparities in breast cancer survival: relationto stage at diagnosis, treatment and race. BMC Cancer 2009;9:364.

20. Meden T, St. John-Larkin C, Hermes D, et al. MSJAMA. Relation-ship between travel distance and utilization of breast cancer treat-ment in rural Northern Michigan. JAMA 2002;287:111.

21. Jessup JM, Stewart A, Greene F, et al. Adjuvant chemotherapy forstage III colon cancer implications of race/ethnicity, age, and differ-entiation. JAMA 2005;294:2703-11.

22. Cronin DP, Harlan L, Potosky A, et al. Patterns of care for adjuvanttherapy in a random population-based sample of patients diagnosedwith colorectal cancer. Am J Gastroenterol 2006;101:2308-18.

23. Hill S, Sarfati D, Blakely T, et al. Survival disparities in Indigenousand non-Indigenous New Zealanders with colon cancer: the role ofpatient comorbidity, treatment and health service factors. J EpidemiolCommunity Health 2010;64:117-23.


REVIEW • REVUE

Complications associated with laparoscopic sleevegastrectomy for morbid obesity: a surgeon’s guide

Obesity is a common disease affecting adults and children. The incidence of obesity inCanada is increasing. Laparoscopic sleeve gastrectomy (LSG) is a relatively new andeffective procedure for weight loss. Owing to an increase in the number of bariatricsurgical procedures, general surgeons should have an understanding of the complica-tions associated with LSG and an approach for dealing with them. Early postoperativecomplications following LSG that need to be identified urgently include bleeding, staple line leak and development of an abscess. Delayed complications include stric-tures, nutritional deficiencies and gastresophageal reflux disease. We discuss the prin-ciples involved in the management of each complication.

L’obésité est une maladie fort répandue, tant chez les adultes que chez les enfants, etson incidence est en hausse au Canada. La gastrectomie longitudinale laparoscopique(GLL) est une intervention relativement nouvelle et efficace pour la perte de poids.Compte tenu de l’augmentation du nombre d’interventions chirurgicales bariatriques,les chirurgiens généralistes doivent se familiariser avec les complications associées à laGLL et avec leur prise en charge. Les complications postopératoires immédiates de laGLL qu’il faut savoir reconnaître sans retard sont l’hémorragie, les fuites le long de laligne d’agrafes et la formation d’abcès. Parmi les complications plus tardives, mention-nons les sténoses, les carences alimentaires et le reflux gastro-œsophagien. Nousprésentons les principes qui sous-tendent la prise en charge de chaque complication.

O besity is a common disease affecting more than 300 million adultsworldwide.1 It is defined as a body mass index greater than 30. InCanada, the prevalence of obesity has increased almost 3-fold in the

past 2 decades.2 Approximately 25% of the Canadian population is now classi-fied as obese.3

Current options for bariatric surgery are categorized by several principles.Purely restrictive procedures include laparoscopic adjustable gastric bandingand sleeve gastrectomy. Roux-en-Y gastric bypass is a restrictive surgery with aminor malabsorption approach. Largely malabsorptive procedures with arestrictive component include duodenal switch and biliopancreatic diversion.Almost purely malabsorptive procedures include jejuno-ileal bypass.

Laparoscopic sleeve gastrectomy (LSG), also known as longitudinal or ver-tical gastrectomy, is a relatively new and effective surgical option for the man-agement of morbid obesity (Fig. 1). It was initially introduced in 1990 as analternative to distal gastrectomy with the duodenal switch procedure to reducethe rate of complications.4,5 Sleeve gastrectomy was first performed laparo-scopically by Ren and colleagues in 1999.6 At the time, LSG was considered afirst-stage operation in high-risk patients before biliopancreatic diversion orRoux-en-Y gastric bypass.7 Laparoscopic sleeve gastrectomy was subsequentlyfound to be effective as a single procedure for the treatment of morbid obes -ity.8 Although LSG functions as a restrictive procedure, it may also cause earlysatiety by removing the ghrelin-producing portion of the stomach.9

The incidence of obesity in Canada is on the rise, and more patients areundergoing bariatric surgical procedures.10 This growth is compounded with theescalating incidence of medical tourism wherein patients are travelling abroadfor surgical care, particularly bariatric surgery.11 This inevitably translates to an

Kourosh Sarkhosh, MD, MSc*

Daniel W. Birch, MD, MSc*†

Arya Sharma, MD, PhD, DSc‡

Shahzeer Karmali, BSc, MD*†

From the *Centre for the Advancementof Minimally Invasive Surgery (CAMIS),Royal Alexandra Hospital, and theDepartments of †Surgery and ‡Medicine,University of Alberta, Edmonton, Alta.


Correspondence to: S. KarmaliCAMIS, Royal Alexandra Hospital10240 KingswayEdmonton AB T5H [email protected]

DOI: 10.1503/cjs.033511

comp-sark-rev_Layout 1 13-09-16 10:44 AM Page 347


REVUE

increased incidence of complications associated with suchprocedures. It is therefore essential for all general surgeons,including those practising in smaller communities, to beaware of these potential complications and to have a basicunderstanding of how to manage them and when to ask forguidance from a bariatric surgeon. The purpose of this arti-cle is to shed some light on basic principles in the manage-ment of complications after LSG. We present our operativeapproach to LSG and review the major acute (within 2 wk ofsurgery) and late complications that can arise in patients fol-lowing LSG (Table 1).

OPERATIVE TECHNIQUE

The patient is placed in a supine position with the armsspread apart. Pneumoperitoneum is achieved using aclosed technique with a Veress needle placed in the leftsubcostal area of the abdomen. Two 10 mm ports areplaced in the supraumbilical and left midabdominal areas.An additional 15 mm port is placed in the right mid -abdomen to pass the stapler. Finally, 2 additional 5 mm

ports are placed in the left and right upper quadrants of theabdomen. The left lobe of the liver is retracted mediallyusing a Nathanson retractor placed in the subxiphoid area.

The stomach is decompressed at the beginning of theoperation by placing an orogastric tube. The surgeon standsto the patient’s right with the first assistant standing to thepatient’s left. The angle of His is taken down bluntly usingthe Goldfinger dissector (Ethicon Endo-Surgery), exposingthe left crus of the diaphragm. Dissection is started about6 cm proximal to the pylorus by taking down the gastrocolicligament using the Harmonic Scalpel (Ethicon Endo-Surgery). Dissection is carried out proximally toward theshort gastric vessels. This releases attachments to the greatercurvature of the stomach and gastric fundus. The orogastrictube is then removed and replaced by a 50-French bougieplaced in the stomach by the anesthesiologist and guidedlaparoscopically to sit in the lesser curvature of the stomachjust distal to the pylorus. A 60 mm Endo GIA tri-stapler isthen used to divide the stomach. We use 2 black cartridgesinitially to divide the distal stomach, starting 6 cm proximalto the pylorus. Next, 4–6 60 mm purple cartridges are usedto complete the division of the remainder of the stomach.The specimen is then taken out of the abdominal cavitythrough the 15 mm port. The bougie is then removed, andintraoperative gastroscopy is performed with air insufflationand methylene blue to rule out any leaks. We routinely closeour 15 mm port fascia under direct vision before exsufflationof the abdomen. Patients receive nothing by mouth aftersurgery. On postoperative day 1, an upper gastrointestinalstudy is done using gastrografin to rule out staple line leak-age (Fig. 2). A clear fluids diet is subsequently initiated.

ACUTE COMPLICATIONS

Hemorrhage

The risk of postoperative bleeding has been reported to bebetween 1% and 6% after LSG.4,12 The source of bleedingcan be intra- or extraluminal. Intraluminal bleeding fromthe staple line usually presents with an upper gastrointes -tinal bleed. Common symptoms include hematemesis ormelena stools. Diagnosis and management of intraluminalbleeding follows the common algorithm taken for anFig. 1. Sleeve gastrectomy.

Table 1. Complications associated with laparoscopic sleeve gastrectomy

Complication Chronicity Diagnosis Management

Hemorrhage Acute Physical findings, serial CBC Transfusion with or without laparoscopy/laparotomy

Leak Acute/chronic Physical findings, UGI series Drainage (infrared laparoscopy), antibiotics with or without stenting and/or repair

Abscess Chronic CT scan, ultrasound Drainage, antibiotics

Stricture Chronic Endoscopy, UGI series Endoscopy (dilatation), surgery (seroyotomy)

Nutrient deficiency Chronic Physical findings, blood work Nutritional supplements

GERD Chronic History, endoscopy Treatment with proton pump inhibitor

CBC = complete blood count; CT = computed tomography; GERD = gastroesophageal reflux disease; UGI = upper gastrointestinal.



REVIEW

upper gastrointestinal bleed. This includes establishmentof large bore intravenous lines for fluid resuscitation,administration of packed red blood cells if necessary, accu-rate measurement of urine output with insertion of aFoley catheter and an urgent gastroscopy to diagnose andcontrol the source of bleeding.

Extraluminal bleeding usually presents with a serialdrop in serum hemoglobin levels or signs of tachycardia orhypotension. Common sources for extraluminal bleedinginclude the gastric staple line, spleen, liver or abdominalwall at the sites of trocar entry. We suggest a second-looklaparoscopy in any patient who presents with extraluminalbleeding with a sustained heart rate greater than 120 beatsper minute and a drop in hemoglobin of more than 10 g/Lpostoperatively. Urgent laparoscopy facilitates a diagnosisand allows evacuation of the clot as well as surgical controlof the source of bleeding. Many times the actual sourcecannot be identified, but we believe that evacuation of thehematoma and placement of a closed suction drain oftenserves as a helpful adjunct to patient resuscitation.

A number of buttressing materials are commerciallyavailable to attempt to reduce the rate of bleeding from thestaple line. These include glycolide trimethylene carbonatecopolymer (Gore Seamguard; W.L. Gore and Associates),bovine pericardium strips (Synovis Surgical Innovations) orporcine small intestinal submucosa (Surgisis Biodesign,

Cook Medical). Whether the use of buttressing materialreduces the rate of bleeding remains controversial. In arecent prospective randomized trial, Dapri and colleagues13

compared the rate of staple line bleeding after LSG using3 different techniques: stapling the stomach with no re -inforcement, or reinforcement with either suturing or but-tressing with Gore Seamguard. These investigatorsobserved a significantly lower rate of bleeding with the useof buttressing material. There was no difference in theincidence of a leak. Albanopoulos and colleagues,14 how-ever, did not observe a significant difference in their rate ofpostoperative bleeding between patients with staple linesuturing or buttressing with Gore Seamgard after LSG.These investigators had a 6% and 0% complication rate(bleeding and leak) in their suturing and Seamgard arms,respectively, but this did not reach significance. Thereforeat our institution we do not routinely use any reinforce-ment materials (sutures or buttresses) for LSG.

Staple line leak

Gastric leak is one of the most serious and dreaded com-plications of LSG (Fig. 3). It occurs in up to 5% ofpatients following LSG.8,12 Several classifications exist

Gastric sleeve

Fig. 2. Radiograph showing a normal image of the stomach afterlaparoscopic sleeve gastrectomy.

Gastric sleeve

Large leak

Fig. 3. Radiograph showing a leak following laparoscopic sleevegastrectomy.



REVUE

based on the radiologic findings and time of diagnosis.15

Based on upper gastrointestinal contrast study, gastric leakcan be classified into 2 types. A type I or subclinical leak iscontrolled either through a surgical drain or through a fis-tulous tract into the abdominal or chest cavity. A type II orclinical leak is a disseminated leak with diffusion of thecontrast into the abdominal or chest cavities.16 Based onthe time of diagnosis, gastric leaks are classified as early orlate. An early leak is generally diagnosed within the first3 days after surgery, whereas a delayed leak is usually diag-nosed more than 8 days after surgery.17

Gastric leaks can be diagnosed either incidentally on aroutine upper gastrointestinal series performed postopera-tively without any clinical signs or during exploratorylaparoscopy/laparotomy performed owing to unexplainedtachycardia. In a study by Kolakowski and colleagues,18 acombination of clinical signs of fever, tachycardia andtachypnea was found to be 58.33% sensitive and 99.75%specific for detection of anastomotic leaks. Diabetes melli-tus and sleep apnea were associated with a greater inci-dence of anastomotic leak. Therefore, we suggest anexploratory laparoscopy for diagnosis in patients who showthese signs in the early postoperative period. In the pres-ence of a leak, an abdominal washout with surgical repairof the leak (if technically feasible) and establishment of anenteral feeding route should be performed. Because thestomach is limited in size, the preferred choice for enteralfeeding is typically a feeding jejunostomy.

In contrast, treatment of a delayed gastric leak is morechallenging surgically owing to the presence of an inflam-

matory reaction. In this setting, attempts to repair the leakare usually futile. Treatment options include conservativeor surgical management. This depends on the patient’shemodynamic condition and on physical and radiologicfindings. In the absence of hemodynamic instability andphysical findings suggestive of peritonitis, conservativemanagement may be initiated. This entails fluid resuscita-tion, initiation of intravenous antibiotics, nothing bymouth, percutaneous drainage of intra-abdominal collec-tions (if drainable) and intraluminal stenting.19 In a septicpatient with radiological evidence of a leak with diffuseintra-abdominal fluid collections, surgical drainage of thefluid collection is warranted.

At our institution, we have successfully managed delay -ed gastric leaks with drainage (either surgical or percuta-neous), establishment of a feeding route (enteral or par-enteral) and placement of gastric stents for approximately2–4 weeks (Fig. 4). Other investigators have also usedintraluminal stents for the management of gastric leaks.20,21

Himpens and colleagues20 reported their experience in themanagement of 29 patients with gastric leak after sleevegastrectomy with stenting. These investigators left thestents in situ on average for 7 weeks. Immediate successwas observed in 19 patients after placement of the firststent, whereas 5 patients required placement of a secondstent. Two patients had persistent leaks requiring a surgicalintervention.

Abscess

Intra-abdominal abscess is another possible complicationafter LSG. It usually presents with symptoms of abdom -inal pain, fever/chills or nausea and vomiting. If there areclinical suspicions, one should obtain a computed tomog-raphy scan of the abdomen to rule out the presence ofintra-abdominal abscess. In a series of 164 patients under-going LSG, Lalor and colleagues22 reported 1 patient withan abscess (0.7%). Treatment includes percutaneous drain -age and antibiotics.

CHRONIC COMPLICATIONS

Stricture

Formation of stricture is another potential complicationoccurring after LSG. It could present either acutely aftersurgery due to tissue edema or more commonly in a delayedfashion. Presenting symptoms include food intolerance, dys-phagia or nausea and vomiting. Although kinking of thestomach following LSG has been reported,23 the most com-mon site of stenosis is at the incisura angularis.24 An uppergastrointestinal study or endoscopy is usually diagnostic.

Treatment options depend on the time of presentation.A stricture diagnosed acutely after surgery can sometimesbe treated conservatively with bowel rest (nothing by

Intraluminal stent

Fig. 4. Radiograph showing an intraluminal stent for the treat-ment of a leak following laparoscopic sleeve gastrectomy.



REVIEW

mouth), rehydration with intravenous fluids and closeobservation. In the absence of other pathologies (e.g.,abscess, leak), these strictures will spontaneously resolvewith no need for further intervention. Failure of conserva-tive management warrants endoscopic dilation.

In contrast, chronic strictures usually require furtherintervention. These include either endoscopic or surgicaltreatments. Treatment options depend on the length ofstenosis. Endoscopic dilatation is an invaluable tool used inthis setting of a short segment stenosis.25 Successive treat-ments in 4- to 6-week intervals are adequate to treat stric-ture and ameliorate patient symptoms. In contrast, longsegment stenosis and failure of endoscopic managementdemands a surgical intervention. Options include laparo-scopic or open seromyotomy or conversion to Roux-en-Ygastric bypass. Dapri and colleagues26 reported their experi-ence with laparoscopic seromyotomy in patients who hadLSG. These investigators reported successful results withthis treatment. Parikh and colleagues25 reported an inci-dence of 3.5% of symptomatic stenosis following LSG intheir series of 230 patients; 2 patients required conversionto a Roux-en-Y gastric bypass owing to failure of endo-scopic management.

Nutritional deficiencies

Nutritional deficiencies are common after bariatricsurgery. The etiology is multifactorial owing to impairedabsorption and decreased oral intake. In a recent study byGehrer and colleagues,27 the prevalence of vitamin B12,vitamin D, folate, iron and zinc deficiency were reportedto be 3%, 23%, 3%, 3% and 14%, respectively, after LSG.In general, these investigators found micronutrient defi-ciencies to be less prevalent after LSG than Roux-en-Ygastric bypass; however, folate deficiency was slightlymore common after LSG than Roux-en-Y gastric bypass(22% v. 12%).27 Routine blood work is therefore war-ranted after LSG to diagnose vitamin and mineral defi-ciencies. At our institution, we routinely monitor patients’serum vitamin B12, vitamin D, folate, iron and calciumlevels at 3, 6 and 12 months after surgery and treat themaccordingly, if necessary.

Gastroesophageal reflux disease

Gastroesophageal reflux disease (GERD) is a conditionseen commonly in the bariatric surgery population.Although some operations, such as Roux-en-Y gastricbypass, are known to be associated with a reduced inci-dence of reflux postoperatively, this is controversial forLSG. In a recent systematic review by Chiu and col-leagues,28 the authors found the data to be inconclusivewith respect to the effect of LSG on GERD. Of theincluded studies, 4 showed an increased incidence ofGERD postoperatively, whereas 7 showed a decrease in

the incidence of GERD. Carter and colleagues29 per-formed a retrospective study on patients who underwentLSG and found 47% of their patients to have persistent(> 30 d) GERD symptoms. The most common reportedsymptoms included heartburn (46%) and regurgitation(29%). Management of patients with persistent GERDinvolves treatment with proton pump inhibitors. Thesepatients require close clinical follow-up. If their symptomspersist despite the use of proton pump inhibitors, we usu-ally perform a gastroscopy for diagnosis.

CONCLUSION

Laparoscopic sleeve gastrectomy is a new and effectiveprocedure for the surgical management of morbid obesity.Therefore, the number of patients undergoing this proce-dure will continue to rise. Basic understanding of commoncomplications and available treatment options is essentialfor all practising general surgeons. This paper offers basicmanagement guidelines for the treatment of complica-tions after LSG. By early diagnosis and treatment of thesecomplications, patient morbidity and mortality might bereduced.

Competing interests: None declared for K. Sarkhosh and A. Sharma.D.W. Birch declares having received consultant and speaker fees fromJohnson & Johnson, Ethicon Endo-Surgery, Covidien, Olympus, Bardand Baxter. S. Karmali declares having received speaker fees fromEthicon and Covidien.

Contributors: All authors designed the study. K. Sarkhosh acquired thedata, which K. Sarkhosh and A. Sharma analyzed. K. Sarkhosh andS. Karmali wrote the article, which all authors reviewed and approvedfor publication.

References

1. Deitel M. Overweight and obesity worldwide now estimated toinvolve 1.7 billion people. Obes Surg 2003;13:329-30.

2. Katzmarzyk PT, Mason C. Prevelance of class I, II and III obesity inCanada. CMAJ 2006;174:156-7.

3. Shields M, Carroll MD, Ogden CL. Adult obesity prevalence inCanada and the United States. NCHS Data Brief 2011;56:1-8.

4. Frezza EE. Laparoscopic vertical sleeve gastrectomy for morbidobes ity. The future procedure of choice? Surg Today 2007;37:275-81.

5. Marceau P, Hould FS, Simard S, et al. Biliopancreatic diversion withduodenal switch. World J Surg 1998;22:947-54.

6. Ren CJ, Patterson E, Gagner M. Early results of laparoscopic bilio -pancreatic diversion with duodenal switch: a case series of 40 con -secutive patients. Obes Surg 2000;10:514-23.

7. Regan JP, Inabnet WB, Gagner M, et al. Early experience with two-stage laparoscopic Roux-en-Y gastric bypass as an alternative in thesuper-super obese patient. Obes Surg 2003;13:861-4.

8. Moon Han S, Kim WW, Oh JH. Results of laparoscopic sleeve gas-trectomy (LSG) at 1 year in morbidly obese Korean patients. ObesSurg 2005;15:1469-75.

9. Gumbs AA, Gagner M, Dakin D, et al. Sleeve gastrectomy for mor-bid obesity. Obes Surg 2007;17:962-9.



REVUE

10. Padwal�RS,�Lewanczuk�RZ.�Trends� in�bariatric� surgery� in�Canada1993–2003.�CMAJ 2005;172:735.

11. Birch�DW,�Vu�L,�Karmali�S,� et� al.�Medical� tourism� in�bariatricsurgery.�Am J Surg 2010;199:604-8.

12. Melissas�J,�Koukouraki�S,�Askoxylakis�J,�et�al.�Sleeve�gastrectomy:�Arestrictive�procedure?�Obes Surg 2007;17:57-62.

13. Dapri�G,�Cadiere�GB,�Himpens�J.�Reinforcing�the�staple�line�duringlaparoscopic� sleeve�gastrectomy:�prospective� randomized�clinicalstudy�comparing�three�different�techniques.�Obes Surg 2010;20:462-7.

14.� Albanopoulos K,�Alevizos L,�Flessas J,� et� al. Reinforcing� the� stapleline�during�laparoscopic�sleeve�gastrectomy:�prospective�randomizedclinical� study�comparing� two�different� techniques.�Preliminaryresults.�Obes Surg 2012;22:42-6.

15. Burgos�AM,�Braghetto� I,�Csendes�A,� et� al.�Gastric� leak� after�laparoscopic–sleeve�gastrectomy�for�obesity.�Obes Surg 2009;19:1672-7.

16. Csendes�A,�Diaz�JC,�Burdiles�P,�et�al.�Classification�and�treatment�ofanastomotic� leakage�after� extended� total�gastrectomy� in�gastriccarcin�oma.�Hepatogastroenterology 1990;37:174-7.

17. Csendes�A,�Burdiles�P,�Burgos�AM,�et�al.�Conservative�managementof�anastomotic�leaks�after�557�open�gastric�bypasses.�Obes Surg 2005;15:�1252-6.

18. Kolakowski�S,�Kirkland�ML,�Scuricht�AL.�Routine�postoperativeupper�gastrointestinal�series�after�Roux-en-Y�gastric�bypass.�Arch Surg2007;142:930-4.

19. Márquez�MF,�Ayza�MF,�Lozano�RB,�et�al.�Gastric� leak�after� laparo-scopic�sleeve�gastrectomy.�Obes Surg 2010;20:1306-11.

20.� Himpens�J,�Dapri�G,�Cadiere�GB.�Treatment�of� leaks�after�sleeve�gastrectomy.�Bariatric Times September� 2009.�Available:� �http://bariatrictimes��.com�/treatment�-of�-leaks�-after�-sleeve�-gastrectomy/�(ac�-cessed�2012�Oct. 29).

21. Oshiro�T,�Kasama�K,�Umezawa�A,�et� al.�Successful�management�ofrefractory�staple� line� leakage�at� the�esophagogastric� junction�after�asleeve�gastrectomy�using� the�HANAROSTENT.�Obes Surg 2010;20:530-4.

22. Lalor�PF,�Tucker�ON,�Szomstein�S,�et�al.�Complications�after�laparo-scopic�sleeve�gastrectomy.�Surg Obes Relat Dis 2008;4:33-8.

23. Uglioni B,�Wolnerhanssen B,�Peters T,�et�al. Midterm�results�of�pri-mary�versus�secondary� laparoscopic�sleeve�gastrectomy�(LSG)�as�anisolated�operation.�Obes Surg 2009;19:401-6.

24. Cottam�D,�Qureshi�F,�Mattar�S,� et� al.�Laparoscopic� sleeve�gastrec-tomy�as� an� initial�weight-loss�procedure� for�high-risk�patients�withmorbid�obesity.�Surg Endosc 2006;20:859-63.

25.� Parikh�A,�Alley� JB,�Peterson�RM,�et� al.�Management�options� forsymptomatic�stenosis�after�laparoscopic�vertical�sleeve�gastrectomy�inthe�morbidly�obese.�Surg Endosc 2012;26:738-46.

26. Dapri�G,�Cadiere�GB,�Himpens� J.�Laparoscopic� seromyotomy� forlong� stenosis� after� sleeve�gastrectomy�with�or�without�duodenalswitch.�Obes Surg 2009;19:495-9.

27. Chiu�S,�Birch�DW,�Shi�X,�et�al.�Effect�of�sleeve�gastrectomy�on�gastro�-esophageal� reflux�disease:� a� systematic� review.�Surg Obes Relat Dis2011;7:510-5.

28. Gehrer�S,�Kern�B,�Peters�T,�et� al.�Fewer�nutrient�deficiencies� afterlaparoscopic�sleeve�gastrectomy�(LSG)�than�after�laparoscopic�Roux-en-Y-gastric�bypass� (LRYGB)�—�a�prospective� study.�Obes Surg2010;20:447-53.

29. Carter PR,�LeBlanc KA,�Hausmann MG,�et�al. Association�betweengastroesophageal� reflux�disease�and� laparoscopic� sleeve�gastrectomy.Surg Obes Relat Dis 2011;7:569-72.

CJS’s top viewed articles*

1. Research questions, hypotheses and objectivesFarrugia�et�al.Can J Surg 2010;53(4):278-81

2. Tracheostomy: from insertion to decannulationEngels�et�al.Can J Surg 2009;52(5):427-33

3. Adhesive small bowel obstruction:epidemiology, biology and preventionAttard�and�MacLeanCan J Surg 2007;50(4):291-300

4. Parmacological management of postoperativeileusZeinali�et�al.Can J Surg 2009;52(2):153-7

5. All superior pubic ramus fractures are notcreated equalSteinitz�et�al.Can J Surg 2004;47(6):422-5

6. Hardware removal after tibial fracture has healedSidky�and�BuckleyCan J Surg 2008;51(4):263-8

7. Bizarre parosteal osteochondromatousproliferation (Nora lesion): a resport of 3 casesand a review of the literatureGruber�et�al.Can J Surg 2008;51(6):486-9

8. Soft-tissue nail-fold excision: a definitivetreatment for ingrown toenailsChapeskie�and�KovacCan J Surg 2010;53(4):282-6

9. Topical nifedipine with lidocaine ointment versusactive control for pain after hemmorhoidectomy:results of a multicentre, prospective, randomized,double-blind studyPerrotti�et�al.Can J Surg 2010;53(1):17-24

10. Complications associated with adjustable gastricbanding for morbid obesity: a surgeon’s guideEid�et�al.Can J Surg 2011;54(1):61-6

* Based�on�page�views�on�PubMed�Central�ofresearch,�reviews,�commentaries�and�continuingmedical�education�articles.�Updated�Sept.�12,�2013.



CONTINUING MEDICAL EDUCATIONFORMATION MÉDICALE CONTINUE

CAGS AND ACS EVIDENCE BASED REVIEWS IN SURGERY. 45

Cost-effectiveness of bariatric surgery for severelyobese adults with diabetes

The term “evidence-based medicine” was first coined by Sackett and colleagues as“the conscientious, explicit and judicious use of current best evidence in making deci-sions about the care of individual patients.”1 The key to practising evidence-basedmedicine is applying the best current knowledge to decisions in individual patients.Medical knowledge is continually and rapidly expanding. For clinicians to practiseevidence-based medicine, they must have the skills to read and interpret the medicalliterature so that they can determine the validity, reliability, credibility and utility ofindividual articles. These skills are known as critical appraisal skills, and they requiresome knowledge of biostatistics, clinical epidemiology, decision analysis and econom-ics, and clinical knowledge.

Evidence Based Reviews in Surgery (EBRS) is a program jointly sponsored bythe Canadian Association of General Surgeons (CAGS) and the American Collegeof Surgeons (ACS) and is supported by an educational grant from ETHICON andETHICON ENDO-SURGERY, both units of Johnson & Johnson Medical Prod-ucts, a division of Johnson & Johnson and ETHICON Inc. and ETHICONENDO-SURGERY Inc., divisions of Johnson & Johnson Inc. The primary objec-tive of EBRS is to help practising surgeons improve their critical appraisal skills.During the academic year, 8 clinical articles are chosen for review and discussion.They are selected for their clinical relevance to general surgeons and because theycover a spectrum of issues important to surgeons, including causation or risk factorsfor disease, natural history or prognosis of disease, how to quantify disease, diag-nostic tests, early diagnosis and the effectiveness of treatment. A methodologicalarticle guides the reader in critical appraisal of the clinical article. Methodologicaland clinical reviews of the article are performed by experts in the relevant areas andposted on the EBRS website, where they are archived indefinitely. In addition, alistserv allows participants to discuss the monthly article. Surgeons who participatein the monthly packages can obtain Royal College of Physicians and Surgeons ofCanada Maintenance of Certification credits and/or continuing medical educationcredits for the current article only by reading the monthly articles, participating inthe listserv discussion, reading the methodological and clinical reviews and com-pleting the monthly online evaluation and multiple choice questions.

We hope readers will find EBRS useful in improving their critical appraisal skillsand in keeping abreast of new developments in general surgery. Four reviews are pub-lished in condensed versions in the Canadian Journal of Surgery and 4 are published inthe Journal of the American College of Surgeons. For further information about EBRS,please refer to the CAGS or ACS websites. Questions and comments can be directedto the program administrator, Marg McKenzie, at mmckenzie @mtsinai.on.ca.

Reference

1. Evidence-Based Medicine Working Group. Evidence-based medicine. JAMA 1992; 268:2420-5.

SELECTED ARTICLE

Horerger TJ, Zhang P, Segel JE, et al. Cost-effectiveness of bariatric surgeryfor severely obese adults with diabetes. Diabetes Care 2010;33:1933–39.

Harry J. Henteleff, MD Daniel W. Birch, MDPeter T. Hallowell, MDfor the Members of theEvidence Based Reviewsin Surgery Group*

*The CAGS/ACS Evidence BasedReviews in Surgery Group comprisesDrs. N.N. Baxter, K.J. Brasel, C.J. Brown,P.K. Chaudhury, C.M. Divino, E. Dixon,G.W.N. Fitzgerald, S.M. Hameed,H.J. Henteleff, T.G. Hughes, L.S. Kao,A.W. Kirkpatrick, S. Latosinsky,T.M. Mastracci, R.S. McLeod, A.M. Morris, T.M. Tawlik, L.K. Temple andMs. M.E. McKenzie.

Correspondence to:M. McKenzie, RNAdministrative Coordinator, EBRSMount Sinai Hospital, L3-01060 Murray St., PO Box 23Toronto ON M5T 3L9fax 416 [email protected]

DOI: 10.1503/cjs.020713

cagsebrs-henteleff_Layout 1 13-09-16 10:46 AM Page 353


FORMATION MÉDICALE CONTINUE

ABSTRACT

Objective: To analyze the cost-effectiveness of bariatricsurgery in severely obese adults who have diabetes. Basecase: Patients with body mass index (BMI) ≥ 35 who havediabetes. Methods: The Centre for Disease Control(CDC)-RTI Diabetes Cost-Effectiveness Model, which is aMarkov simulation model of disease progression and cost- effectiveness for type 2 diabetes, was expanded to considerthe effects of bariatric surgery. Interventions considered:Gastric bypass and gastric banding compared with usualdiabetes care. Outcomes considered: Diabetes-relatedand surgical complications, diabetes remission and relapserates, deaths, costs and quality of life. Results: Bariatricsurgery increased quality-adjusted life years (QALYs) andincreased costs. Bypass surgery had cost-effectiveness ratiosof $7 000 per QALY and $12 000 per QALY for severelyobese patients with newly diagnosed and established dia-betes, respectively. Gastric banding had cost-effectivenessratios of $11 000 per QALY and $13 000 per QALY,respectively. In sensitivity analyses, the cost- effectivenessratios were most affected by assumptions about the directgain in quality of life and by BMI reduction followingsurgery. Conclusion: The analysis indicates that gastricbypass and gastric banding are cost-effective methods ofreducing mortality and diabetes-related complications inseverely obese adults with diabetes.

COMMENTARY

Bariatric surgery has a well-established role in the man-agement of obesity. Procedures such as gastric banding,gastric bypass and gastric sleeve resection have beenshown to result in sustained weight loss where medicalmanagement has failed. There are approximately 2 millionpeople in Canada and 20 million people in the UnitedStates who have type 2 diabetes. Of these, approximately80%–90% are obese. Bariatric surgery has been shown toreverse type 2 diabetes by a mechanism that is unclear ina substantial number of patients. This effect can be in -depend ent of weight loss.

The study by Horerger and colleagues attempts tomeas ure the cost-effectiveness of bariatric surgery com-pared with medical management of diabetes. It takes anestablished model of cost-effectiveness simulating diabetesprogression that was developed by the Centers for DiseaseControl (CDC) and adds the costs and outcomes associ-ated with bariatric surgery. This model measures costs andoutcomes in patients with type 2 diabetes from diagnosis todeath and simulates the development of diabetic complica-tions along 3 microvascular pathways (nephropathy, neuro -pathy and retinopathy) and 2 macrovascular pathways(coronary heart disease and stroke). The important out-comes related to surgery are the probability of remission ofdisease (complete and partial), the probability of relapse

and the probability of complications related to the surgery.Reported costs are in 2005 US dollars and are discountedat a rate of 3% per year, as is usually done in an economicanalysis. Costs associated with usual diabetic care werederived from the UK Prospective Diabetes study.1

The study population is diabetic patients with a BMIof 35 or more. This is the group that qualifies for bari-atric surgery under current National Institutes of Health guidelines.

The authors performed a comprehensive analysis ofmedical costs using U.S. data that include the upfront costsof surgery as well as the costs associated with subsequentyears of care and the need for revision surgery (e.g., chole-cystectomy, pannectomy). They compare the costs of gas-tric bypass and gastric banding, which are the 2 most com-mon procedures performed in the United States, althoughsleeve gastrectomy is being performed more commonlynow. Unfortunately, the exact derivation of the costs is onlyavailable in an online appendix to the paper and cannot beeasily reviewed. It is not clear if the authors studied costsfrom a societal perspective and included time off work andpatient costs.

The authors modelled the outcomes of surgery basedon 2 meta-analyses that included 982 published studies. Itshould be noted, however, that only 5% of the publicationswere randomized controlled trials (RCTs). The authorsdetermined the rate of remission of diabetes in patientswith new onset (56.7%–80.3%) or established diabetes(40%) for both gastric banding and gastric bypass. Theywere also able to model the complications of surgery withdata derived from the same studies. Bariatric surgery hadcost-effectiveness ratios ranging between $7 000 and$13 000 per QALY. Patients with new onset diabetesgained more QALYs after surgery than those with estab-lished diabetes, and they gained more QALYs from bypasssurgery than gastric banding but had a higher surgical cost.To put these costs in context, it is generally considered thatinterventions with a cost of less than $50 000 per QALYare cost-effective.2

Most importantly in this type of study, the authors wenton to perform sensitivity analyses. They varied the possibleoutcomes of surgery from one-half to 2 times the observedrates to see if this affected the cost-effectiveness of surgery.They found that the incremental cost-effectiveness of surgerydid not change much unless they varied the QALY improve-ment or the weight loss associated with surgery. Finding anoperation to be cost-effective even when the outcomes varywidely strongly suggests that the findings are robust.

Unfortunately the authors did not find surgery to becheaper overall than medical management. However, theymodelled only the costs of diabetic care and did not factorin potential cost savings for other conditions related toweight loss. For example, decreasing blood pressure ordecreasing the need for joint replacement could definitelyreduce costs for surgical patients.



CONTINUING MEDICAL EDUCATION

Bariatric surgery has become an important option in themanagement of diabetes and should be considered in obesepatients with this disease. The study by Horerger and col-leagues provides a thorough analysis of the costs and bene-fits of gastric bypass and gastric banding and finds the pro-cedures to be cost-effective. In addition 3 recent RCTs ofbariatric surgery versus medical management in obesepatients with type 2 diabetes have been published.3–5 Theone by Dixon and colleagues3 showed that gastric bandingis superior to medical management at 2 years. Mingroneand colleagues4 compared gastric bypass and biliopancre-atic diversion to medical management. No patients in themedically managed arm had remission of diabetes at2 years, while 75% of the patients who underwent gastricbypass and 95% of those who underwent biliopancreaticdiversion were in remission.4 Schauer and colleagues5 com-pared gastric bypass and gastric sleeve to intensive medicalmanagement. At 1 year, 12% of the medically managedpatients, 42% in the gastric bypass group and 37% in thegastric sleeve group had a hemoglobin A1C of 6.0% orless.5 Thus, all of the commonly performed bariatric pro -cedures have been shown to be effective in the short termfor the management of type 2 diabetes.

What remains to be further defined is the most appro-priate time to intervene surgically in patients with type 2diabetes (at what BMI, at what illness duration, and at whatage?). Also to be defined is the most appropriate procedure(gastric band, gastric bypass, biliopancreatic diversion, orgastric sleeve). While this will take some time to assess, it isclear that bariatric surgery should be discussed with pa -tients as an option for treatment of type 2 diabetes and

obesity. The results of the study by Horerger and col-leagues should encourage surgeons, physicians, insurersand policy makers to improve access to bariatric/metabolicsurgery for morbidly obese patients with type 2 diabetes.

The importance and relevance of this cost-effectivenessstudy lies in the way in which it promotes a unique oppor-tunity for general surgeons to play a pivotal role in thedefinitive management of diabetes. Departments of surgerywill have to prepare for a further increase in demand for aprocedure with already long wait lists and limited access.Competing interests: None declared.

References

1. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose contril with sulphonylureas or insulin compared with con-ventional treatment and risk of complications in patients with type 2diabetes. Lancet 1998;352:837-53.

2. Grosse SD. Assessing cost-effectiveness in healthcare: history of the$50,000 per QALY threshold. Expert Rev Pharmacoecon Outcomes Res2008; 8:165-78.

3. Dixon JB, O’Brien PE, Playfair J, et al. Adjustable gastric bandingand conventional therapy for type 2 diabetes: a randomized con-trolled trial. JAMA 2008;299:316-23.

4. Mingrone G, Panunzi S, De Gretano A, et al. Bariatric surgery versusconventional medical therapy for type 2 diabetes. N Engl J Med 2012;366: 1577-85.

5. Schauer PR, Kashyap SR, Wolski K, et al. Bariatric Surgery versusintensive medical therapy in obese patients with diabetes. N Engl JMed 2012;366:1567-76.

FORUM canadien de chirurgieLa réunion annuelle du FORUM canadien de chirurgie aura lieu du 18 au 21 septembre 2014 à la Ville de Vancouver, Colombie-Britannique. Cette réunion interdisciplinaire permet aux chirurgiens de toutes lesrégions du Canada qui s’intéressent à la pratique clinique, au perfectionnement professionnel continu, à larecherche et à l’édu cation médicale d’échanger dans un climat de collégialité. Un programme scientifiqueintéressera les chirurgiens universitaires et communautaires, les résidents en formation et les étudiants.

Les principales organisations qui parrainent cette réunion sont les suivantes :• L’ Association canadienne des chirurgiens généraux• La Société canadienne des chirurgiens du côlon et du rectum• La Société canadienne de chirurgie thoracique• La Société canadienne d’oncologie chirurgicale

Le American College of Surgeons, l’Association canadienne des médecins et chirurgiens spécialistes del’obésité, l’Association québécoise de chirurgie, le Canadian Association of University Surgeons, le CanadianHepato-Pancreato-Biliary Society, le Canadian Undergraduate Surgical Education Committee, le James IVAssociation of Surgeons et l’Association canadienne de traumatologie sont au nombre des sociétés quiappuient cette activité.

Pour vous inscrire ou pour plus de renseignements, veuillez consulter le site www.cags-accg.ca.



CORRESPONDENCE • CORRESPONDANCE

IDENTIFICATION AND USE OFOPERATING ROOM EFFICIENCYINDICATORS: THE PROBLEM OF NOTPERFORMING THE RIGHT SEARCHWITHIN PUBMED

Drs. Fixler and Wright1 should becommended for demonstrating thatoperating room (OR) performanceindicator definitions vary in literatureand among children’s hospitals. Unfor-tunately, I do not agree with their con-clusion that the most logical coursewould be for professional associationsto agree upon and develop commonmetrics and definitions. Their conclu-sion is based on a limited review ofpapers that are not always relevant.

First, the Procedural Times Glos-sary has been the leading source forOR definitions since 1997.2 Papersdescribing operational research inORs use this glossary.3 A bibliographyof papers concerning operationalresearch within the OR can be foundonline (http://www .franklindexter .net/bibliography _TOC.htm).

Based on this evidence, I concludethat there are clear definitions formonitoring OR performance indica-tors. An additional conclusion is thathospitals continue to use their owndefinitions. This needs to be solved bysending surgeons, anesthesiologistsand managers of ORs to courseswhere they can learn which indicatorsto use and how to use them.

Fixler and Wright call for us to usethe OR resources in both an efficientand effective way. Here they make amistake. Indeed, monitoring the oper-ational performance of the OR maycontribute to the use of OR resourcesin an efficient way. However, the callto use OR resources in an effective wayis a faulty statement. According to theInstitute of Medicine’s Committee onQuality Health Care in America,effective care “is based on providingservices based on scientific knowledgeto all who could benefit, and refrain-ing from providing services to those

not likely to benefit (avoiding under-use and overuse, respectively).”4 Herethe patient clinical parameters are ofinterest and not, for example, the uti-lization rate of the OR.

In conclusion, performing an accu-rate search in PubMed will show thatthe actual problem of agreed-upondefinitions in literature, as describedby the authors, does not exist.

Pieter Stepaniak, PhDCatharina HospitalOperating Room DepartmentEindhoven, the NetherlandsGdansk Medical UniversityDepartment of General, Endocrine and Transplant Surgery

Gdansk, Poland


DOI: 10.1503/cjs.019413

References

1. Fixler T, Wright JG. Identification and useof operating room efficiency indicators: theproblem of definition. Can J Surg 2013;56:224-6.

2. Donham RT, Mazzei WJ, Jones RL. Associ-ation of anesthesia clinical directors' pro ceduraltimes glossary: glossary of times used for sched-uling and monitoring of diagnostic and ther -apeutic procedures. Columbus (OH): Associ-ation of Anesthesia Clinical Directors;1996.

3. Stepaniak PS, Heij C, Buise MP, et al.Bariatric surgery with operating room teamsthat stayed fixed during the day: a multicen-ter study analyzing the effects on patientoutcomes, teamwork and safety climate, andprocedure duration. Anesth Analg 2012;115:1384-92.

4. Institute of Medicine, Committee onQuality Health Care in America. Crossingthe quality chasm: a new health system for the21st century. Washington (DC): NationalAcademic Press; 2001.

COMMENT ON “IDENTIFICATION ANDUSE OF OPERATING ROOMEFFICIENCY INDICATORS: THEPROBLEM OF DEFINITION”

It was with profound interest that weread the commentary written byTamas Fixler and James G. Wright in

the August 2013 issue of the CanadianJournal of Surgery. The commentarydeals with the identification andmeas urement of operating room (OR)performance indicators, addressingthe variation among hospitals in termsof which indicators are collected andanalyzed.

Common definitions among hospi-tals are essential for external bench-marking. Although the authors identi-fied 8 indicators as the most criticalfor monitoring OR performance in15 children’s hospitals in Canada, def-initions for these indicators vary in lit-erature and across hospitals.

In the Netherlands, OR depart-ments of all 8 university medical cen-tres (UMCs) established a nationwidebenchmarking collaboration in 2005that is still active today. The objectiveof the collaboration is to improve ORperformance by learning from eachother through exchanging best prac-tices. Each UMC provides recordsfor all performed surgical cases to acentral OR benchmark database. Thisextensive database, presently com-prising more than 1 million surgicalcase records, is used to calculate keyperformance indicators related to theutilization of OR capacity. The data-base is also used for multicentreresearch on OR scheduling topicsand OR efficiency.

At the start of this collaboration, aset of performance indicators, particu-larly from a utilization perspective,was identified. Next, data definitionsof time periods and methods of reg -istration, as well as definitions of performance indicators, were har mon -ized among all benchmarking par -ticipants, a process that took nearly2 years. An independent data manage-ment centre enters the longitudinaldata collection in the central ORbenchmark database. This centre pro-vides professional expertise by facili-tating and processing data, and by per-forming reliability checks before dataare deemed ready for analysis.


CORRESPONDANCE

Our collaboration frequently meetsto discuss data analysis results andexplore processes and practicesbeyond the data. Through promotingdialogue among UMCs, a learningenvironment has been created.

G. Kazemier, MD, PhDProfessor of Hepatobiliary Surgery and Transplantation

Department of SurgeryVU University Medical CenterAmsterdam, the NetherlandsE. van Veen-Berkx, MScErasmus University Medical CenterDepartment of Operating RoomsRotterdam, the Netherlandsfor the Dutch OR Benchmarking Collaboration


DOI: 10.1503/cjs.020813

THE AUTHORS RESPOND

We thank Dr. Stepaniak for his inter-est in our commentary on the identi-fication and use of operating room(OR) efficiency indicators. While theProcedural Times Glossary, develop -ed by the Association of AnesthesiaClinical Directors (AACD), is a lead-ing source of procedural time defini-tions in support of economic and effi-ciency analyses within the OR, thisdoes not negate the fact that variableperformance indicator definitionsnonetheless exist in the body of ORefficiency literature. Moreover, des -pite the availability of leading sourcesof definitions such as the ProceduralTimes Glossary, differences in howhospitals define key OR performanceindicators persist.

Furthermore, even the AACD’sProcedural Times Glossary may notalways be adequate if one wants to

ensure consistent performance indica-tor data collection across multiplehospitals. For example, the AACDdefines “turnover time” as the “timefrom prior patient out of room to suc-ceeding patient in room time forsequentially scheduled cases.”1 How-ever, while this definition is clearlymeant to exclude idle time betweennonsequentially scheduled cases, itdoes not entirely address potentialexclusions, such as delays betweensequentially scheduled cases unrelatedto room cleaning and preparation(e.g., patient arrives late); how thesesituations are handled varies signifi-cantly across hospitals and materiallyimpacts how the indicator is collected.

Another example is the definitionof “on-time starts,” defined as thepatient being in the OR at the sched-uled time.1 This does not consider,however, whether certain late startsshould be excluded (e.g., owing todelayed access to postoperative beds,as is the case at some hospitals).

Thus, we do believe that there isroom for professional associations toagree to develop common metrics andoperational definitions, perhaps usingthe AACD’s Procedural Times Glos-sary (or an equivalent source) as astarting point, closing any gaps fromthere.

Regarding Dr. Stepaniak’s secondpoint, while performance indicatorsmay not contribute to the effective useof resources as defined by the Instituteof Medicine’s Committee on QualityHealth Care in America, they may doso under another definition, such asthe Oxford English Dictionary, whichdefines “effective” as “having anintended or expected effect.” If usingresources efficiently leads to the mostpatients having surgery in the best way

(i.e., on time starts, no delays, no can-cellations), then use of OR perform -ance indicators to monitor operationalperformance can indeed lead to theeffective use of resources.

In addition, we also thank Dr.Kazemier and Ms. van Veen-Berkx fortheir interest in our commentary andnote that the Dutch experience,whereby it took 2 years to harmonizeOR performance indicator definitionsand reporting across 8 universitymedical centres, speaks to the com-plexity of the undertaking and thecontinuing lack of universal standardsfor indicator definitions.

Moreover, some Canadian prov -inces have also had some success inharmonizing OR performance indi-cators, such as the OR BenchmarksCollaborative in Ontario. As ourcommentary has demonstrated,though, variable indicator definitionspersist and harmonizing them nation-ally may be particularly challengingdue to the provincial delivery ofhealth care.

Tamas Fixler, MASc, MBAIBM Canada Ltd.Thornhill, Ont.James G. Wright, MD, MPHDepartment of SurgeryRobert B. Salter Chair of Pediatric Surgical Research

The Hospital for Sick ChildrenToronto, Ont.


DOI: 10.1503/cjs.020513

Reference

1. Donham RT. Defining measurable OR-PRscheduling, efficiency, and utilization dataelements: the Association of AnesthesiaClinical Directors’ procedural times glos-sary. Int Anesthesiol Clin 1998;36:15-29.


INSTRUCTIONS FOR AUTHORS

The Canadian Journal of Surgery (CJS)publishes original research and reviewarticles on aspects of surgical techniqueand practice (including evidence-basedand international surgery, surgical biologyfor the clinician, trauma and crit ic al care),as well as editorials, commentaries, caseseries and letters.

MANUSCRIPT SUBMISSION

CJS is pleased to announce the launch ofour online submission and review systemwith Manuscript Central at http ://mc.manuscriptcentral .com /cjs. We will nolonger accept manuscript submissions viamail or email.

To submit a manuscript online

If you are new to our system, register as anew user by clicking on the “Regis terhere” link on the right-hand toolbar.

Enter your Author Centre and under Auth or Resources, select “Click here tosubmit a new manuscript.” You will findit helpful to have the following infor -mation on hand:

• Manuscript type• Title• Abstract (you may rekey or copy and

paste this from your manuscript; 300-word limit)

• Each coauthor’s first and last names,email address and country and city ofresidence

• Names and email addresses of anypreferred or nonpreferred reviewers

• Cover letter• Manuscript word count• All relevant manuscript files

Your manuscript will be assigned a man-u script number that is its identity in thesystem. Use it to track your manuscript’sprogress online in your Author Centre.You will be notified as soon as possible ofthe outcome of the peer review process.Decisions and queries will be communi-cated via email.

MANUSCRIPT PREPARATION

Authors should consult “Uniform require-

ments for manuscripts submitted to bio-medical journals.” In general, the materialshould be origin al, the methods appropri-ate, the data valid and the conclusionssupported by the data. The topic shouldbe of interest to Canadian surgeons andsurgeons-in-training. Manuscripts shouldbe as brief as pos sible, and tables and fig-ures included only if they add to the text.

Title page

This page should contain

• the full title• the authors’ full names, academic

degrees and affiliations• any meeting(s) at which the work was

presented in whole or in part• the corresponding author’s contact

information (mailing address, phoneand fax numbers and email address)

Abstract

All articles except case series should con-tain an abstract of 150– 250 words, prefer-ably structured, with the same headingsappearing in the body of the paper.

• Background• Methods• Results • Conclusion

If such headings are inappropriate to thearticle, an unstructured abstract should be provided that is both descriptive andinformative.

References

These should be cited as superscriptsin numerical order of appearance inthe text. References in tables and figuresshould be numbered and cited wherethe table or figure is first mentioned in thetext. Include the surnames and initials ofup to 3 authors, followed by “et al.” whenthere are more than 3 auth ors.

Tables and figures

• Tables should not duplicate data pres -ented in figures and must be self-

explanatory. Kindly prepare tablesusing the Word or Excel table func-tion (not spaces and tabs).

• Figures should be provided in anelectronic format. Charts should besent as Excel or Powerpoint files; bar-ring that, please provide data points.Photographic and radio graphic imagesshould be sent as JPGs, min imum300 dpi at minimum 2 × 2 inches. Youmay wish to consult our Digital ArtSubmission for Editorial Articles formor Digital camera specifications forgreater detail.

Cover letter

A covering letter signed by all the auth -ors should state that the manuscripthas not been published and is not underconsideration by any other journal. Itshould explain the purpose of the paperand its intended readership. All authorsmust disclose the source of any financialor material support, any commercialinterest they may have in the subject ofthe study and any involvement with anorganization with a financial interest inthe research materials or topic.

REVIEW AND PUBLICATION PROCESS

Peer review

All submitted manuscripts are viewed byan editor. Manuscripts that are not suitable for CJS or that are of insufficientpriority for publication are rejectedpromptly. Other manuscripts are sentfor peer review; note that peer reviewers’identities are kept confidential, but auth ornames are made known to reviewers.Manuscripts under consideration are priv-i leged communications between auth orsand editors. Editorial staff will discussthem only with the corresponding authorand with peer reviewers. Once a decisionis made, authors are notified promptly andare sent a copy of the reviewer comments.

Editing

Accepted manuscripts are edited with aview to clarity, brevity, style and accur acy.An edited manuscript proof is sent to thecorresponding author for approval. Note

instructions-oct13_Layout 1 13-09-16 11:44 AM Page 356

INSTRUCTIONS FOR AUTHORS


that all authors are respon sible for thecontent of their work, including copyedi-tor changes authorized by the corres -ponding author.

Reprints

Commercial and author reprints can bepurchased through Sheridan Press. Topurchase commercial article reprints andePrints, or to request a quote, please con-tact Matt Neiderer, Content Sales, Sheri-dan Content Services at 800 635-7181,ext. 8265; matt .neiderer @sheridan .com.Authors can order reprints by submittingan author reprint order form available atthe Sheridan Press Electronic OrderCentre at sheridan .com /cma /eoc or bycontacting Lori Laughman, CustomerService Representative, Sheridan ReprintsServices; lori .laughman @sheridan .com

EDITORIAL POLICIES

Duplicate submission/ publication

CJS will consider manuscripts for publi-cation on the understanding that they arenot under simultaneous review withanother publication and that they havenot previously been published.

Authorship criteria

Authorship credit will be assigned only toindividuals who meet ALL of the follow-ing criteria. Each author must have

• contributed to study conception anddesign OR data acquisition OR data

analysis• contributed to writing OR critically

reviewing the article• approved the final version of the arti-

cle submitted for publication

CJS’s authorship criteria are in accor-dance with the statement on authorshipissued by the International Committee ofMedical Journal Editors (www .icmje.org). Authors of accepted originalresearch, reviews and case series will berequired to complete a Contributors’Information Form, in which they mustspecify their contributions to the workdescribed in the manuscript. This infor-mation will be published with the article.

The acquisition of funding, data collec-tion or general supervision of theresearch group does not justify auth or-ship. Indi viduals who have made such acontribution to the manuscript shouldinstead be listed in the acknowledge-ments section.

Legal requirements

Upon acceptance, all authors must sub-mit signed Copyright and Financial dis-closure forms. Authors of Research andReview articles (including Surgical Biol-ogy for the Clinician, Trauma and Crit - ical Care, International Surgery and CaseSeries) must also submit completed Con-tributors’ forms. Where applicable, youmay also be required to include

• a signed Acknowledgement form fromanyone whose contribution goes be -yond administrative assistance and

whom you want to name• a signed Personal communication

form if you wish to cite a personalcommunication in your article (pleaseinclude the person’s affiliation and theyear in which the information wascommunicated)

• a signed Patient consent form fromany patient whose case is describedand any person identifiable in an illustration

• permission to reproduce material fromanother source for previously pub-lished tables or figures, signed by thecopyright holder, usually the publisher

Obligation to register clinicaltrials reported in CJS

The CJS endorses the statement of theInter national Committee of Medical Jour-nal Editors (available at www .icmje .org)concerning the registration of clin ical trials.In brief, the journal requires, as a conditionof consideration for publication, registra-tion of clinical trials in a public trials reg-istry at or before the onset of patient enrol-ment. For CJS, this policy applies to anyclinical trial that starts enrolment afterJul. 1, 2006. A clinical trial is defined as anyresearch project that prospectively assignshuman subjects to inter vention or compari-son groups to study the cause-and-effectrelation between a medical interventionand a health outcome. Studies designed forother purposes (e.g., Phase I trials) areexempt. The journal does not endorse aspecific registry but, when selecting a reg-istry, auth ors should use the criteria men-tioned in the statement of the InternationalCommittee of Medical Journal Editors.

instructions-oct13_Layout 1 13-09-16 11:44 AM Page 357


DIRECTIVES AUX AUTEURS

Le Journal canadien de chirurgie (JCC) pu - blie des articles de recherche originale etdes articles de revue sur tous les aspects dela technique et la pratique chi rurgicales (ycompris la chirurgie fondée sur les don-nées probantes, la chirurgie à l’étranger, labiologie chirurgicale pour le clinicien, lessoins des traumatisés et les soins critiques),de même que des édi toriaux, commen-taires et les séries de cas.

PRÉSENTATION DES MANUSCRITS

Le JCC est heureux d’annoncer le lan - cement de son système électronique deprésentation et de révision des manu-scrits sur Manuscript Central (http://mc.manuscriptcentral .com /cjs). Veuillez noterque nous n’acceptons plus les manuscrits par laposte ou par courrier électro nique.

Pour présenter un manuscrit en ligne :

Si vous utilisez notre système pour la première fois, allez à http://mc.manuscriptcentral.com/cjs et cliquez surle lien «Register here» à droite.

Entrez dans votre « Author Centre » etsous « Author Resources », cliquez « Clickhere to submit a new manu-script ». Ayezles renseignements sui vants à portée de lamain :

• type de manuscript• titre• résumé (à retaper ou copier-coller de

votre manuscrit; maximum 300mots)• pour chacun des co-auteurs : pré nom

et nom de famille, adresse de courriel,pays et ville de résidence

• nom et adresse de courriel des exami -nateurs privilégiés ou non privi légiés

• lettre de présentation• compte de mots du manuscrit• tous les fichiers pertinents du manuscrit.

Votre manuscrit se verra assigner unnuméro qui lui servira d’identité dans lesystème. Ce numéro vous permettra desuivre l’évolution de votre manuscritdans votre « Author Centre ». Le résultatde l’exa men par les pairs vous sera com-muniqué dès que possible, par courrierélectronique, ainsi que toute question, lecas échéant.

PRÉPARATION DU MANUSCRIT

Les auteurs sont priés de consulter les« Exigences uniformes pour les manu-scrits présentés aux revues biomédicales ».Il faut toutefois retenir qu’en général, ilfaut présenter un matériel original, utiliserdes méthodes appropriées et présenter des données valides pour étayer les con-clusions. Le sujet devra intéresser leschirurgiens et les chirurgiens en forma-tion du Canada. Il faut rédiger le manu-scrit le plus brièvement possible et ajouterdes tableaux et illustrations uniquements’ils ajoutent au texte.

Page titre

Cette page doit comporter

• le titre intégral• le nom complet et les diplômes univer-

sitaires et affiliations des auteurs• le titre de toute rencontre à laquelle

l’article a été présenté, en tout ou enpartie

• les coordonnées de l’auteur corres -pondant (adresse postale, numéros detéléphone et de télécopieur et adresseélectronique)

Résumé

Tous les articles sauf les séries de casdoivent comporter un résumé de 150 à250 mots, de préférence structuré, avec lesmêmes rubriques qui apparaissent dans lecorps du texte, comme suit :

• Contexte• Méthodes• Résultats• Conclusion

S’il ne convient pas d’utiliser ces titresdans l’article, fournir un résumé nonstructuré à la fois descriptif et informatif.

Références

La liste de références doit être établienumériquement, en indice supérieur, selonl’ordre d’apparition dans le texte. Lesréférences citées dans les tableaux et dans leslégendes de figures doivent être numérotéesselon l’endroit où le texte en fait une pre-

mière mention. Pour citer les documents àauteurs multiples, donner le prénom et lesinitiales d’au plus 3 auteurs et ajouter lamention “et coll.” s’il y en a plus de 3.

Tableaux et figures

• Les tableaux ne doivent pas présenter ànouveau des données présentées dansdes illustrations et doivent être suffi sam-ment explicites. Veuillez préparer lestableaux à l’aide de la fonction de ta -bleaux de Word ou en Excel (ne pasutiliser des espaces et tabulateurs).

• Il faut présenter les illustrations en for-mat électronique. Présenter les graphi -ques en fichiers Excel ou Power point.Sinon, fournir les points de données.Envoyer les images photo graphiquesou radiographiques en format JPG,d’une résolution d’au moins 300 ppp et d’une taille d’au moins 2 × 2 po. Ilserait bon de consulter les documentssuivants pour plus de détails : Présenta-tion de documents numériques pour lesédito riaux; Spécifications techniquespour la photographie numérique.

Lettre de présentation

La lettre de présentation, signée par tousles auteurs, doit confirmer que le manu-scrit n’a jamais été publié et n’a été soumisà aucun autre journal. La lettre de présen-tation doit exposer l’objectif du documentet indiquer à quel public il s’adresse. Lesauteurs doivent préciser la source de touteaide financière ou matérielle, tout intérêtcommercial qui les aurait motivé envers lesujet à l’étude ainsi que toute affiliation outravail avec un organisme ayant un intérêtfinancier dans le matériel de recherche oudans le sujet lui-même.

ÉTUDE DU MANUSCRIT, EXAMEN PARLES PAIRS ET PUBLICATION

Examen par les pairs

Tous les manuscrits sont étudiés par unrédacteur. Si l’on juge que le manuscritne convient pas au JCC ou que le sujetn’est pas prioritaire, la décision de ne paspu blier est prise rapidement. Les autresmanuscrits sont envoyés à l’examen par lespairs. Veuillez noter que l’identité des

directives-oct13_Layout 1 13-09-16 11:39 AM Page 358

DIRECTIVES AUX AUTEURS


pairs examinateurs de meure confidentiellemais que le nom des auteurs est commu-niqué aux examinateurs. Les manuscrits àl’étude sont considérés comme étant descommunications pri vilégiées entre lesauteurs et les rédacteurs. Le personnel dela rédaction n’en discutera qu’avec l’au-teur correspondant et avec les pairs exami-nateurs. Une fois la décision prise, les au -teurs sont avisés rapidement et on leurenvoie un exemplaire des commentairesdes examinateurs.

Révision

Les manuscrits acceptés sont révisés. Larévision vise à les rendre plus clairs, à lesabréger au besoin, à faire les ajustementsde style nécessaires et à vérifier leur exacti-tude. Il faut noter que tous les auteurs sontres ponsables du contenu de leur travail, ycompris des modifications ré dactionnellesautorisées par l’auteur corres pondant.

Tirés à part

On peut acheter des tirés à part d’auteurou commerciaux auprès de SheridanPress. Pour les tirés à part commerciaux etles cyberimpressions (ePrints), ou encorepour demander un prix, veuillez commu-niquer avec Matt Neiderer, Ventes de con-tenu, Sheridan Content Services (800635-7181, poste 8265; matt .neiderer@sheridan .com). Les auteurs peuventcommander des tirés à part en remplissantle bon de commande de tirés à part d’au-teur disponible au centre de commandesélectroniques de Sheridan Press (sheridan.com/cma/eoc) ou en communiquant avecLori Laughman, repré sentante du serviceà la clientèle, Sheridan Reprints Services(lori .laughman @sheridan .com).

POLITIQUES DE LA RÉDACTION

Présentation ou publication en double

Le JCC étudiera des manuscrits en vuede les publier à condition qu’ils ne soient

pas à l’étude simultanément dans uneautre publication et qu’ils n’aient pas déjàété publiés.

Qualité d’auteur

La qualité d’auteur ne sera accordée qu’auxpersonnes répondant à TOUS les critèresénoncés ci-après. Chaque auteur doit

• avoir contribué à créer ou concevoirl’étude OU la collecte des données OUl’analyse des données

• avoir contribué à rédiger l’article OU àen avoir fait la révision critique

• avoir approuvé la version finale de l’ar-ticle présenté pour publication

Les critères que le JCC attache à la qua -lité d’auteur sont conformes à l’énoncépertinent publié par le Comité interna-tional de rédacteur de revues médicales(www.icmje.org). Les auteurs d’articlesportant sur une recherche originale, desynthèse ou de séries de cas acceptésdevront remplir un Formulaire d’infor-mation sur la contribution des auteursoù ils doivent préciser leur contributionau travail décrit dans le manuscrit. Cetteinformation paraîtra avec l’article.

L’acquisition de financement, la collectede données ou la surveillance générale dugroupe de recherche ne justifient pas laqualité d’auteur. Les personnes qui ontapporté une contribution au manuscritdoivent plutôt figurer dans la liste desremerciements.

Exigences légales

Lorsqu’un manuscrit est accepté, tous lesauteurs doivent soumettre les formulairessignés de cession du droit d’auteur et dedivulgation de renseignements financiers.Les auteurs d’articles de recherche et desynthèse (y compris en biologie chirurgi-cale pour le cli n i cien, trauma tologie etsoins intensifs, chirurgie internationaleet séries de cas) doivent aussi soumettredes formulaires d’information sur la contribution des auteur remplis. Le cas

éché ant, on pourra aussi vous demanderd’inclure les documents suivants :

• un formulaire de reconnaissance signéde toute personne dont la contributiondépasse l’aide administrative et quevous voulez mentionner nommément

• un formulaire sur une communicationpersonnelle signé si vous voulez citer unecommunication personnelle dans votrearticle; veuillez inclure l’affiliation de lapersonne en cause et l’année au cours delaquelle l’information a été diffusée

• un formulaire de consentement dupatient signé par tout patient dont lecas est décrit et toute personne identi-fiable dans une illustration

• une autorisation de reproduire dumatériel provenant d’autres sourcesdans le cas de tableaux ou de figuresdéjà publiés, signée par le titulaire dudroit d’auteur, habituellement l’éditeur

Obligation d’inscrire les étudescliniques faisant l’objet d’unepublication dans le JCC

Le JCC approuve l’énoncé du Comitéinternational de rédacteurs de revue médi-cale (disponible à www.icmje.org) au sujet del’inscription des études cli niques. En deuxmots, pour que le journal envisage de pu -blier les résultats d’une étude clinique, celle-ci devra avoir été inscrite à un registre publicdes études, au début de l’inscription despatients ou avant. Pour le JCC, cette poli-tique s’applique à toute étude clinique ayantcommencé à inscrire des patients après le1 juillet 2006. On entend par étude cliniqueou essai cli nique tout projet de recherchequi affecte de façon prospective des sujetshumains à une inter vention ou à des groupestémoins afin d’étudier les liens de cause àeffet entre une intervention médicale etl’évolution de l’état de santé. Les étudesconçues à d’autres fins (p. ex., les études dePhase I) sont exemptées. Le journal ne pri -vilégie aucun registre parti culier, mais dansle choix d’un registre, les auteurs devraientse fonder sur les critères mentionnés dansl’énoncé du Comité international des rédac-teurs de revue médicale.

directives-oct13_Layout 1 13-09-16 11:39 AM Page 359


SERVICE INFORMATIONRENSEIGNEMENTS AUX LECTEURS

Subscription rates (2014)Libraries, research establishments and other multiple-reader institutions:Canada, Can$310; United States and other countries, US$360. Indi vi d-ual: Canada Can$200; USA US$235. Can adian subscribers please addapplicable taxes. For other pricing in formation, please contact the Can - adian Medical Association Subscription Office, PO box 830350, Birm-ingham AL 35283-0350; phone 800 633-4931 (Canada, USA) or205 995-1567; fax 205 995-1588; [email protected].

Other subscription informationAll subscriptions are payable in advance, in Canadian currency to Can -adian addresses, in US currency to all other addresses. Payment shouldbe made to CMA in the funds speci fied, drawn on a Canadian or USbank, respectively. VISA, MasterCard and American Express are also accepted. Orders and requests for infor ma tion on other journals shouldbe addressed to the Member Service Centre.

Change of addressWe require 8 weeks’ notice to ensure uninterrupted service. Please sendyour current mailing label, your new address and the effective date ofchange to the Member Service Centre.

Replacing missing issuesClaims for missing issues must be made within 3 months of the date ofpublication to be honoured and replaced (subject to availability) free ofcharge. Replacement copies of older issues, when available, must beprepaid at the single-copy rates listed below. Please submit claims, witha copy of your mailing label, to the Member Service Centre.

Single-copy salesSubject to availability, single copies may be purchased for Can$22 orUS$22. Please submit your order and pay ment to the Member ServiceCentre.

Article reprintsCommercial and author reprints can be purchased through SheridanPress. To purchase commercial article reprints and ePrints, or torequest a quote, please contact Matt Neiderer, Content Sales, SheridanContent Services at 800 635-7181, ext. 8265; matt .neiderer @sheridan.com. Authors can order reprints by submitting an author reprint orderform available at the Sheridan Press Electronic Order Centre at sheridan .com /cma /eoc or by contacting Lori Laughman, CustomerService Representative, Sheridan Reprints Services; lori .laughman@sheridan .com

Microform, abstracting and indexingCJS appears in the following indexing/abstracting services: ASCA, AbHyg, CBCARef, CINAHL, CPerf, ChemAb, CurCont, DentInd,ESPM, ExerpMed, H&SSA, HelmAb, ISR, IndMed, Inpharma, MED -LINE, NRN, NutrAb, PE&ON, RM&VM, Reac, RefZh, SCI, TDB.— BLDSC (3035.800000), CISTI, GNLM, IDS, IE, infotrieve,ingenta, KNAW.CCC.

Electronic availabilityAll articles are available on the Internet (cma.ca/cjs) in PDF format.

PermissionsCopyright for all material is held by CMA or its licensors. The CMA isa member of Access Copyright, The Canadian Copyright LicensingAgency, and has an agreement in place permitting them to grant orga-nizations and individuals, on behalf of CMA, the right to respond tocopyright requests. Please contact Sue Petrykewcyz, Interim SeniorPermissions Associate, Access Copyright, The Canadian CopyrightLicensing Agency, 800 893-5777; permissions @accesscopyright .ca.

Tarifs des abonnements (2014)Bibliothèques, établissements de recherche et autres établissements à lecteursmultiples : Canada, 310 $Can; États-Unis et ailleurs, 360 $US. Individus: Canada200 $Can; États-Unis 235 $US. Pour les abonnés au Canada, veuillez ajouter les taxesapplicables. Pour obtenir des renseignements sur les prix, veuillez communiquer avecle Can adian Medical Association Subscription Office, PO box 830350, Birmingham AL35283-0350; tél 800 633-4931 (Canada, É-U) ou 205 995-1567; fax 205 995-1588;[email protected].

Autres renseignements sur les abonnementsTous les abonnements sont payables d’avance en dollars canadiens s’ils sont livrés auCanada, et en dollars américains ailleurs dans le monde. Prière d’envoyer sonpaiement, tiré sur une banque canadienne ou américaine res pectivement, à AMC.VISA, MasterCard et American Express sont aussi acceptées. Les commandes et lesdemandes de renseignements sur les autres journaux doivent être adressées au Centredes services aux membres.

Changement d’adresseUn préavis de 8 semaines est nécessaire pour assurer la livraison ininterrompue devotre abonnement. Veuillez envoyer votre étiquette postale actuelle, votre nou velleadresse et la date d’entrée en vigueur au Centre des services aux membres.

Remplacement de numéros manquantsPour recevoir gratuitement (sous réserve de sa disponibilité) un numéro manquant,vous devez présenter votre demande dans les 3 mois de la date de publication. Toutecommande d’anciens numéros doit être réglée d’avance selon les prix de vente aunuméro qui sont indiqués ci-après. Veuillez envoyer votre demande ainsi qu’une copiede votre étiquette postale au Centre des services aux membres.

Vente d’exemplaires à l’unitéOn peut se procurer des exemplaires à l’unité (sous réserve de leur disponibilité) pour20 $Can, 20 $US. Veuillez envoyer votre commande et votre paiement au Cen tre desservices aux membres.

Tirés à partOn peut acheter des tirés à part d’auteur ou commerciaux auprès de Sheridan Press.Pour les tirés à part commerciaux et les cyberimpressions (ePrints), ou encore pourdemander un prix, veuillez communiquer avec Matt Neiderer, Ventes de contenu,Sheridan Content Services (800 635-7181, poste 8265; matt .neiderer @sheridan .com).Les auteurs peuvent commander des tirés à part en remplissant le bon de commandede tirés à part d’auteur disponible au centre de commandes électroniques de SheridanPress (sheridan.com/cma/eoc) ou en communiquant avec Lori Laughman, repré -sentante du service à la clientèle, Sheridan Reprints Services (lori .laughman @sheridan.com).

Microcopies, résumés et indexLe JCC est résumé et fiché dans l’index des services spécialisés suivants : ASCA,AbHyg, CBCARef, CINAHL, CPerf, ChemAb, CurCont, DentInd, ESPM,ExerpMed, H&SSA, HelmAb, ISR, IndMed, Inpharma, MEDLINE, NRN, NutrAb,PE&ON, RM&VM, Reac, RefZh, SCI, TDB.— BLDSC (3035.800000), CISTI, GNLM, IDS, IE, infotrieve, ingenta,KNAW.CCC.

Accès électroniqueTous les articles sont disponibles sur internet en format PDF (cma.ca/cjs-f ).

PermissionsLe droit d’auteur de tout le matériel appartient à l’AMC ou à ses concédants. L’AMCest membre d’Access Copyright, l’agence canadienne de gestion du droit d’auteur,avec qui elle a conclu une entente qui permet à l’agence d’accorder à des organisationset à des personnes, au nom de l’AMC, le droit de répondre à des demandes dereproduction de matériel protégé par droit d’auteur. Veuillez communiquer à cettefin avec Sue Petrykewcyz, associée principale intérimaire, Service des autorisations,Access Copyright (800 893-5777; [email protected]).

serviceinfo-oct13_Layout 1 13-09-20 12:28 PM Page 360

© 2013 Olympus America Inc. Trademark or Registered Trademark of Olympus or its affiliate entities in the U.S. and/or other countries of the world. All patents apply. OCIAD412-200113

The World’s Only Fully-Integrated Bipolar & Ultrasonic Technology

Groundbreaking Integrated Technology

For more information contact your local Olympus representative or call 888-387-0437 or visit www.olympusamerica.com/thunderbeat.

THUNDERBEAT is the world’s ONLY integration of both advanced bipolar and ultrasonic energies delivered simultaneously from a single, multi-functional instrument. This integration delivers the widely recognized benefits of each individual energy; the ability to rapidly cut tissue with ultrasonic energy, and the ability to create reliable vessel seals with bipolar energy.

THE BENEFITS OF UNPRECEDENTED VERSATILITY� Fastest-in-class cutting speed � Reliable 7 mm vessel sealing � Highest-in-class tip grasping force� Precise dissection with fine jaw design � Minimal thermal spread � Fewer instrument exchanges� Always-available bipolar energy � Reduced mist generation for

for hemostasis without cutting improved visibility

Western University and London Health Sciences Centre/St. Joseph’s Health Care LondonInvite Applications for the Position of

CHAIR/CHIEF, DIVISION OF ORTHOPAEDIC SURGERY, DEPARTMENT OF SURGERYThe Division of Orthopaedic Surgery, Department of Surgery, Schulich School of Medicine & Dentistry at Western University is seeking a Chair/Chief,full-time clinical academic orthopaedic surgeon, at the rank of Associate or Full Professor. The Division of Orthopaedic Surgery has approximately 25 full-time clinical academic surgeons and three scientists located at three Hospital sites. It is the largest Division in the Department of Surgery and consists ofsix clinical teaching units: Adult Reconstruction; Hand and Upper Limb; Paediatric Orthopaedics; Spine; Sport Medicine; and Trauma. The Division has anational and international reputation for teaching, clinical and research excellence.

Candidates must be certified in Orthopaedic Surgery by the Royal College of Physicians and Surgeons of Canada and licensed in the Province of Ontarioand have a strong record of accomplishment in clinical care, research, education, leadership, and administrative experience.

This is a dual position as Chair, Division of Orthopaedic Surgery – Western University; and as Chief of the Division of Orthopaedic Surgery – LondonHealth Sciences Centre & St. Joseph’s Health Care London.

London is home to Western University, with a full-time enrollment of about 32,000 students in a range of academic and professional programs. LondonHealth Sciences Centre and St. Joseph’s Health Care London serve a population base of more than 2 million Ontario residents from approximately 35hospitals within the region. London’s affordable housing, numerous parks, trails, recreational, cultural and sporting opportunities make it an attractive placeto live. London is close to the Great Lakes and is within easy reach of Toronto, Ontario and Detroit, Michigan. Details about the Schulich School of Medicine& Dentistry can be found at www.schulich.uwo.ca; Western University at www.uwo.ca; London Health Sciences Centre at www.lhsc.on.ca; and St.Joseph’s Health Care at www.sjhc.london.on.ca.

Interested candidates should send a letter of interest, curriculum vitae, and the names and addresses of three references to:

John D. Denstedt, MD, FRCSC, FACS, Chair/Chief, Department of SurgerySJHC, 268 Grosvenor Street, Room E3-117, London, Ontario, Canada N6A 4V2

Telephone 519 663-3349 • Fax 519 646-6347

Applications will be accepted until the position is filled. Review of applications will begin after September 1st, 2013.

Positions are subject to budget approval. Applicants should have fluent written and oral communications skills in English. All qualified candidates are encouragedto apply; however, Canadians and permanent residents will be given priority. Western University, London Health Sciences Centre and St. Joseph’s Health CareLondon are committed to employment equity and welcome applications from all qualified women and men, including visible minorities, aboriginal people andpersons with disabilities.

S-746

PAEDIATRIC ORTHOPAEDIC SURGEONIWK Health Centre

The Division of Paediatric Orthopaedic Surgery, Department of Surgery at Dalhousie University and the IWK HealthCentre are seeking a new colleague to join our academic group of Paediatric Orthopaedic Surgeons. The successfulcandidate will have an accredited Fellowship in Paediatric Orthopaedic Surgery and be eligible to practice in Canada.Potential colleagues in all stages of their career are encouraged to apply.

The IWK Health Centre serves a catchment area of nearly two million people in the Maritime Provinces of Canada andis an active participant in the post-graduate training of orthopaedic surgeons. There are ample opportunities for thedevelopment of an academic career in the many paediatric sub-specialty areas.

The candidate would join an established and stable funding environment and participate in the competitive alternatefunding plan which will allow time to pursue research or other academic interests.

Historic Halifax, the business and cultural centre of Atlantic Canada, is home to five Universities and offers an enviablelifestyle opportunity.

All qualified candidates are encouraged to apply; however, Canadians and permanent residents will be given priority.Dalhousie University and the IWK Health Centre are Employment Equity/Affirmative Action employers. The Health Centreand University encourage applications from qualified Aboriginal people, persons with a disability, racially visible personsand women.

Applications accompanied by a curriculum vitae and the names of three referees should be sent to:

Dr. Ron El-Hawary, Head, Division of Paediatric Orthopaedic SurgeryIWK Health Centre, Halifax, Nova Scotia, Canada

5850/5980 University Ave., PO Box 9700, Halifax, NS B3K 6R8Email [email protected]

S-7

47

CAREER/CLASSIFIED ADVERTISING/CARRIÈRES ET ANNONCES CLASSÉES




The World’s ONLY




A

SPONSORS PARRAINSCanadian Association of General Surgeons

Canadian Society for Vascular SurgeryCanadian Society of Surgical Oncology

Canadian Association of Thoracic Surgeons

Association canadienne des chirurgiens générauxSociété canadienne de chirurgie vasculaireSociété canadienne d’oncologie chirurgicaleAssociation canadienne des chirurgiens thoraciques


Laparoscopic versus open surgery for the treatment of colorectalcancer: a literature review and recommendations

Comparison of the major intraoperative and postoperative complications between unilateral and sequential bilateral TKA

Incremental value and clinical impact of neck sonography for primary hyperparathyroidism: a risk-adjusted analysis

Complications associated with laparoscopic sleeve gastrectomyfor morbid obesity: a surgeon’s guide

Vol. 5

6, N

o. 5

Pages 2

89–360

Octo

ber / o

ctobre 2013

Canadian Jo

urnal o

f Surgery •Jo

urnal ca

nadien de ch

irurgie


Documents

Discover Where Innovation Can Take Youcanjsurg.ca/wp-content/uploads/2014/10/CJS_56-5.pdf · cma.ca/cjs Laparoscopic versus open surgery for the treatment of colorectal ... Bipolar