EQUITY UNDER THE KNIFE: JUSTICE AND EVIDENCE IN SURGERY

EQUITY UNDER THE KNIFE: JUSTICE AND EVIDENCE IN SURGERYbioe_1980 1..8

WENDY ROGERS, CHRISTOPHER DEGELING AND CYNTHIA TOWNLEY

Keywordssurgery,equity,justice,evidence,health technology

assessment,public health ethics

ABSTRACTSurgery is an increasingly common and expensive mode of medical inter-vention. The ethical dimensions of the surgeon-patient relationship, includ-ing respect for personal autonomy and informed consent, are muchdiscussed; but broader equity issues have not received the same attention.This paper extends the understanding of surgical ethics by considering thenature of evidence in surgery and its relationship to a just provision ofhealthcare for individuals and their populations.

JUSTICE AND EVIDENCE IN SURGERY

The aim of this paper is to refine current conceptions ofthe ways in which surgery can and should be responsiveto the requirements of justice. It is perhaps uncontrover-sial to claim that meeting the requirements of justiceentails surgeons providing effective treatments in an equi-table manner. We argue that this involves a broader set ofconsiderations and responsibilities than are customarilyassociated with surgical practice. To defend this claim, weexamine the relationship between evidence and justice,critically examine the use of evidence in surgery, presentexamples that show the importance of effectivenessstudies, and consider barriers to a justice orientation insurgery.

1 SURGERY AND JUSTICE

Many diseases are treated surgically. People living inindustrialized societies are likely to undergo at least fiveoperations in their lifetime.1 Surgeons are trained to tailortheir interventions to the individual needs of patients, butthe effects of surgery are not limited to the treatment ofpatients one-by-one. Surgical services must also meet theneeds of populations and serve the public good. They

should be accessible to people on the basis of clinicalneed, and they must be as safe and effective as possible.Patients, providers and payers all need information onhow, when and where it is best to seek and deliver surgicalcare, creating a prima facie responsibility for surgeons toparticipate in generating the necessary evidence.

While surgical interventions themselves are not thekinds of things that can be just or unjust, the provision ofsurgical services can be. Actions on the part of surgeonsand health service providers can generate, exacerbate orameliorate health inequities.2 Such health inequities mayarise from uneven distribution of surgical services, a moregeneral lack of access to effective surgical interventions,or from other causes such as skewing of service provisionsecondary to commercial interests. There has been con-siderable interest in identifying and investigating inequi-ties in access to surgical treatments.3 Our interest lies ininvestigating the nature of the evidence that can provide

1 P.H.U. Lee & A.A. Gawande. The number of surgical procedures inan American lifetime in 3 states. J Am Coll Surg 2008; 207: S75; A.Elixhauser & R.M. Andrews. Profile of Inpatient Operating RoomProcedures in US Hospitals in 2007. Arch Surg 2010; 145: 1201–1208.

2 We use the term ‘health inequities’ to refer to avoidable differences(disparities) in health status and health outcomes that occur betweenmore and less-advantaged social groups and which reflect the socialdeterminants of health. For further discussion see: P. Braveman & S.Gruskin. Defining Equity in Health. J Epidemiol Community Health2003; 57: 254–258; W. Rogers. 2007. Health Inequities and the SocialDeterminants of Health. In Principles of Health Care Ethics. R. Ash-croft et al. Chichester, West Sussex: John Wiley & Sons, Ltd: 585–591.3 See for example A.M. Epstein et al. Racial Disparities in Access toRenal Transplantation: Clinically Appropriate or Due to Underuse orUveruse? NEJM 2000; 343(21): 1537–1544; C.P. Brown et al. Dispari-ties in the Receipt of Cardiac Revascularization Procedures betweenBlacks and Whites: an Analysis of Secular Trends. Ethn Dis 2008;18(2 Suppl 2): S2-112–117.

Address for correspondence: Prof. Wendy Rogers, Professor of Clinical Ethics, Philosophy Department and Australian School of Advanced Medicine,Macquarie University, Sydney, Australia. T: +61 2 9850 8858. Email: [email protected] of interest statement: No conflicts declared

bs_bs_banner

Bioethics ISSN 0269-9702 (print); 1467-8519 (online) doi:10.1111/j.1467-8519.2012.01980.x

© 2012 Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

a sound foundation for informing service provision, andin assessing the effects of access to those services, as thereis no justice in achieving equitable access to interventionsthat are themselves ineffective or harmful. Questions ofevidence are therefore essential to the fair distribution ofeffective surgical treatments.

The claim that decisions about evidence have ethicalimplications is not new.4 Choices about the collection ofand reliance on evidence can clearly affect ethical issuessuch as how practice should be organized, how resourcesshould be distributed, and who should be the beneficia-ries of research.5 Nonetheless, the issue of what consti-tutes strong and appropriate evidence in surgery remainscontested. Despite sustained inquiry into the nature ofsurgical interventions, processes of surgical innovation,and systems for evaulating these, the relevance of thesedebates for justice has not been explored.6 Claims aboutsurgical evidence can concern efficacy or effectiveness.7

Each provides different types of knowledge about theusefulness, safety and effects of surgical interventions onindividuals and populations.8

Efficacy refers to a measurable difference made by theintervention in an experimental population. Effectivenessrefers to the difference it makes in real patients’ lives. Byfar the majority of surgical research concerns efficacy,not effectiveness. We argue that for both ethical andpragmatic reasons, effectiveness matters.9 Without thegeneration and collection of effectiveness data, we riskbeing oblivious to inequities in outcomes, and thereforeliable to exacerbating them.10 For surgical practice toachieve just outcomes, clinicians need to understand,interpret and act on the best possible evidence.11 The bestevidence includes effectiveness; this fact should play agreater role in the way surgeons choose to generate evi-dence and use it to guide their practices.

2 ETHICAL EVIDENCE: EFFECTIVENESSNOT EFFICACY

In this section we argue that while studies of efficacy areneeded to assess the utility of a surgical intervention, suchstudies do not provide all the evidence necessary for justprovision of surgical services. Interventions are effica-cious if the treatment produces the anticipated outcomeunder ideal conditions. In efficacy studies, (sometimescalled ‘explanatory’ trials) surgeons typically test thenovel device and/or technique in small relatively uniformgroups of patients who meet specific eligibility criteria, toensure that the results are not confounded by uncon-trolled variables. For example, patients with commonco-morbidities are generally excluded from these trials.Trials in which patients are randomized to receive eitherthe study intervention or the comparator (randomizedcontrolled trials – RCTs) are considered to yield thestrongest evidence of efficacy. Efficacy studies followbasic research, animal trials, or both and are usuallyconducted in specialist referral centres by one or twosurgeons who might well have been involved in develop-ing the technology or technique, often working withindustry funders.12

Claims about the efficacy of an intervention are almostalways based on measurements of specific outcomes suchas symptom scores, laboratory data, time-to-disease-recurrence, and mortality. These endpoints are imme-diate and intermediate measures of the effects of theintervention – surrogates rather than measures of overallhealth outcomes. Notably, the outcome measures usedin efficacy trials do not typically include standardized‘quality of life’ assessments, or even measurements offunctional capacity and patient satisfaction (described assoft endpoints),13 nor do they extend over time to identifydelayed sequelae. Although evidence of efficacy fromRCTs is generally regarded as the ‘gold standard’ formedical evidence, it does not exhaust the knowledgerequired for an evidence-based surgery, precisely becausethe excluded ‘soft endpoints’ matter to recipients oftreatment.14

The conditions for efficacy studies differ considerablyfrom the conditions under which ordinary patients will

4 J. Worrall. Evidence and Ethics in Medicine. Perspect Biol Med 2008;51: 418–431.5 W.A. Rogers & A. Ballantyne. Justice in Health Research: What is

the Role of Evidence-Based Medicine? Perspect Biol Med 2009; 52:188–202.6 See R. Lilford et al. Trials in Surgery. Br J Surg 2004; 91(1): 6–16; J.

Barkun et al. Evaluation and Stages of Surgical Innovations. Lancet2009; 374(9695): 1089–1096.7 B. Flay et al. Standards of Evidence: Criteria for Efficacy, Effective-

ness and Dissemination. Prevention Science 2005; 6: 151–175.8 U. Guller. Surgical Outcomes Research Based on Administrative

Data: Inferior or Complementary to Prospective Randomized ClinicalTrials? World J Surg 2006; 30: 255–266.9 S. Honeybul et al. Neurotrauma and the Rule of Rescue. J Med

Ethics 2011; 37(12): 707–710.10 D.R. Urbach & A.M. Morris. Health Care Reform and ComparativeEffectiveness: Implications for Surgeons. Arch Surg 2010; 145: 120–122.11 P.F. Ridgway & U. Guller. Interpreting Study Designs in SurgicalResearch: A Practical Guide for Surgeons and Surgical Residents. J AmColl Surg 2009; 208: 635–645.

12 In addition to potential conflicts arising from surgeons involved indeveloping the interventions under investigation, industry-fundedresearch also creates potential for bias (see C. Bailey et al. Industryand Evidence-based Medicine: Believable or Conflicted? A SystematicReview of the Surgical Literature. Can J Surg 2011; 54: 321–326).13 R.H. Brook & K.N. Lohr. Efficacy, Effectiveness, Variations, andQuality: Boundary-Crossing Research. Medical Care 1985; 23: 710–722;Flay, et al. op. cit. note 7.14 J.B. Semmens, et al. The Quality of Surgical Care Project: A model toEvaluate Surgical Outcomes in Western Australia using Population-based Record Linkage ANZ J Surg 1998; 68: 397–403; Guller. op. cit.note 8.

2 Wendy Rogers, Christopher Degeling and Cynthia Townley

© 2012 Blackwell Publishing Ltd.

likely encounter the intervention. Typical patients in thewider community might well present the co-morbiditiesthat were excluded from the study population. Trialparticipants must have the capacities to comply withresearch protocols, there may be age-related criteria andsocially based exclusions, such as fluency in English lan-guage, or having a fixed home address, but there are nosuch restrictions on ordinary patients.15 The trial sur-geons and their colleagues in the community may differ intheir enthusiasm for and expertise in the intervention.The environment may differ, as a specialist or researchcentre will likely involve high staffing ratios, close atten-tion to the protocols for the trial, and so forth. For thesereasons, efficacy studies demonstrate what a new proce-dure or technology can achieve in closely specifiedcircumstances, and suggest by inference what the inter-vention could possibly do when delivered to patientsunder different conditions.

Investigations into efficacy may be considered unjustwith respect to selection for research trials;16 howevertheir effect upon wider inequities depends upon how effi-cacy results are applied to the provision of surgical ser-vices more broadly. For such services to be providedequitably, information about efficacy is not sufficient;information about effectiveness is also needed.

Effective interventions are those that result in a favour-able outcome for patients with the index condition, inrelation to variations in their capacities, preferences andcontext.17 The focus of an inquiry about effectiveness is todetermine the average outcomes for patients, in uncon-trolled or ‘real world’ conditions. Evidence of effective-ness indicates how well a procedure produces the desiredresult (a beneficial effect) in ordinary practice, rather thanin research. Effectiveness studies should be conducted inprimary care environments where patients who have thetargeted condition are usually seen and treated.18 Studiesof effectiveness are (or should be) conducted on thegeneral population of surgical patients in all its diversity.They usually require widespread longterm observationand collection of data; and for the results to be meaning-ful, data should pertain to ‘everyday’ surgeons practisingin general surgical units, rather than to results achievedby their colleagues who work in referral or researchcentres.

Studies of effectiveness typically focus on generalhealth outcomes, rather than relying upon the symptom

scores or laboratory data used to demonstrate efficacy.Therefore the sample size in effectiveness studies must besufficiently large to detect small but important differencesin people’s health and quality of life. Data collectedabout the effectiveness of a specific procedure can alsoprovide information about how much it will cost, and theinfluence of provider factors on outcomes.

Trials of efficacy and trials of effectiveness bothprovide relevant evidence for the just provision of sur-gical care. We may think of this as a stepwise process –climbing the evidence mountain. Evidence of efficacy isthe first step – if the view is overwhelmingly negative,then the research intervention should be abandoned.If the view is promising, the next stage is to collect evi-dence of effectiveness in a pilot program, which will giveus a view from higher up.19 Here we can see whether thesunlit slopes predicted by efficacy studies are on theother side of an impenetrable chasm, or whether theyare easily attainable. Ongoing monitoring and datacollection can strengthen the knowledge base, leadingto clearer views from ever higher evidentiary vantagepoints.

Ethical and justice problems can arise when neces-sarily limited evidence of efficacy is taken to justify theintroduction and broad dissemination of new surgicalinterventions for the general population. If we are lucky,evidence about effectiveness and evidence about efficacyare consistent: the intermediate outcome measures ofefficacy studies accurately predict the longer term mea-sures of effectiveness. This happy congruence exists forsome interventions: emergency appendectomy and peni-cillin for the treatment of pneumococcal pneumoniaare examples. But the results of efficacy studies do notnecessarily translate to improvements in health out-comes for patient populations.20 Effectiveness studies areneeded to take the step from knowledge that an inter-vention works ‘somewhere’, to knowledge that ‘it willwork for us’.21

Unfortunately. however, the surgical literature dis-plays a strong bias towards evidence of efficacy, typifiedby surgical RCTs. As a result, knowledge that a proce-dure or implant works ‘somewhere’ is still presumed suf-ficient to justify its broad dissemination to other surgeonsand other patients. These low evidentiary standards donot meet the needs of patients and the general public,although they may reflect the interests of stakeholderssuch as surgeons and device manufacturers. On this

15 W. Rogers. Evidence Based Medicine and Justice: a Framework forLooking at the Impact of EBM upon Vulnerable or DisadvantagedGroups. J Med Ethics 2004; 30: 141–145.16 Ibid.17 R. Ashcroft. What is Clinical Effectiveness? Studies in History andPhilosophy of Science Part C: Studies in History and Philosophy ofBiological and Biomedical Sciences 2002; 33: 219–233.18 Brook & Lohr, op. cit. note 13; Flay et al., op. cit. note 7.

19 K. Spilsbury. Variations in Procedure Rates: A Need for‘Population-based’ Surgical Research ANZ J Surg 2004; 74: 404–40520 T.R. Fleming & D.L. DeMets. Surrogate End Points in ClinicalTrials: Are We Being Misled? Ann Intern Med 1996; 125: 605–613.21 N. Cartwright. A Philosopher’s View of the Long Road from RCTsto Effectiveness. Lancet 2011; 377: 1400–1401.

Equity Under the Knife: Justice and Evidence in Surgery 3


point, a comparison of two orthopaedic proceduresshows the value of effectiveness data. The efficacy of theBirmingham Hip Resurfacing procedure (BHR) wasestablished clinically in several small cohort trials. After apilot study that included monitoring patients for sixyears, the BHR was made available to surgeons treatingthe general patient population in 1997. In contrast,De Puy’s ASR© implant was released onto the generalmarket in 2003, without any comparable longitudinalclinical trials in patients. The efficacy of the ASR© waspresumed on the basis of mechanical simulations and itssimilarity to other approved devices. The De Puy productwas promoted aggressively to surgeons as a less invasivealternative treatment for osteoarthritis, specificallydesigned to meet the needs of younger patients whowanted to maintain a highly active lifestyle.

While the BHR has performed well since its marketrelease with relatively few complications,22 a significantnumber of patients with ASR© hips experienced ‘cata-strophic’ problems within 2-5 years after implantation.23

Complications included debilitating inflammationaround the implant and heavy metal poisoning. With93,000 units sold worldwide the ASR© was recalled in2010; the ongoing consequences of its failure described byregulators as ‘a public health nightmare’.24 A number ofmedical opinion leaders have since argued that manda-tory clinical trials for medical devices and a formal systemof longitudinal data collection will limit the risks ofsimilar events in the future.25 Clearly the ASR© shouldhave undergone some form of longitudinal clinical testingbefore its general release, but this case reveals a largerethical problem. Given the availability of a well estab-lished and extensively tested treatment why should thebenchmark for dissemination of a new procedure ordevice be set at the low standard of evidence of efficacy?Surely we would first wish to know that any replacementwas as effective as the old one. As this knowledge can

only be obtained in general surgical practice (rather thanin specialist RCTs) new surgical treatments should beintroduced in stages to minimize both the risk of harms topatients, and the risk of unproductive expenditure. As itstands, the evidentiary and regulatory framework thatpermitted the general use of the ASR© hip served theneeds of device manufacturers and surgeons, and notthose of patients. Aside from allowing unacceptableand avoidable adverse outcomes, the provision of thesesurgical services meant that resources were directedaway from an effective treatment. If the ASR© had beensubjected to a study of effectiveness comparable to thatundertaken in the BHR case, these unjust outcomeswould have been averted.

Taking evidence of efficacy to be sufficient grounds fordisseminating a new treatment will also fail to identifyany potential problems with service delivery. A highlyefficacious procedure ‘in the hands’ of one surgical team,can be an ineffective and even harmful response to thesame condition elsewhere, due to inevitable variations incase complexity in the general population, diverse hospi-tal and surgeon experience, and differences in the qualityof equipment. In health services research and clinical epi-demiology, knowledge of the efficacy of an intervention isonly an obligatory passage point in the process of acquir-ing knowledge of its effectiveness.26 Yet in surgery, ratherthan being recognized as the acme of surgical evidence,effectiveness studies are generally seen as an optionalcomplement to the dissemination of new techniques andtechnologies to every-day surgical practice.27 Rather thanbeing a mandatory component of surgical research,evaluation of the effectiveness of specific practices andprocedures is uncommon.

There are a number of reasons why this might be so.First, the concurrent and continuous evaluation of theeffectiveness of surgical procedures is time-consumingand resource intensive. When effectiveness data con-firms efficacy data, collection of the former can appearto be wasted effort. Second, effectiveness data is messy;there are problems in accounting for and meaning-fully analysing the effects of patient variation andco-morbidities on the results of complex treatment. It isdifficult to use the results of population-based observa-tional studies in decisions about individual patients; andit is not always clear how patient preferences should beincorporated and assessed. These points reflect someof the criticisms of EBM levelled by members of thesurgical community.28

22 A.D. Carrothers et al. Birmingham Hip Resurfacing: The Prevalenceof Failure. J Bone Joint Surg Br 2010; 92-B: 1344–1350.23 D.J. Langton et al. Blood Metal Ion Concentrations after HipResurfacing Arthroplasty: A Comparative Study of Articular SurfaceReplacement and Birmingham Hip Resurfacing Arthroplasties. J BoneJoint Surg Br 2009; 91-B: 1287–1295; D. Cohen. Out of Joint: The Storyof the ASR. BMJ 2011; 342.24 Other issues raised by the dramatic failure of ASR© hipsinclude: adverse events reporting, the relationship between surgeonsand device manufacturers, and indeed the role of advertising inthe rapid dissemination of the product. For further analysis seeJ. Meek. The Privatisation of the NHS. London Review of Books 2011;33: 3–10.25 G.D. Curfman & R.F. Redberg. Medical Devices – Balancing Regu-lation and Innovation. NEJM 2011; 365: 975–977; D.R. Challoner &W.W. Vodra. Medical Devices and Health – Creating a New Regula-tory Framework for Moderate-Risk Devices. NEJM 2011; 365:977–979; S.E. Graves. What is Happening with Hip Replacement?MJA 2011; 194: 620–621.

26 N. Freemantle & T. Strack. Real-world Effectiveness of New Medi-cines Should be Evaluated by Appropriately Designed Clinical Trials.J Clin Epidemiol 2010; 63: 1053–1058, Flay et al., op. cit. note 7.27 Urbach & Morris, op. cit. note 10; R.S. McLeod et al. 2008.Evidence-Based Surgery. In Surgery. Springer New York: 21–35.28 N. Black. Evidence-based Surgery: A Passing Fad? World J Surg1999; 23: 789–793.



Despite these concerns, effectiveness studies havecontributed and continue to contribute to the equitableprovision of surgical care. Publications from the WesternAustralian Quality in Surgical Care Project reveal theepistemic and normative value of population-basedstudies of surgical effectiveness. Early work focussed onthe effectiveness of surgical interventions for abdominalaortic aneurysm (AAA) in Western Australia between1985–94. At the time, AAA surgery was performed reluc-tantly in patients over 80 years, because of concernsabout procedural risk and the marginal benefits forpatients who may be nearing the end of their lives. Find-ings from the population study indicated that octogenar-ians should not automatically be excluded from surgicalmanagement, and that with careful case selection olderpatients with AAA may go on to enjoy many more yearsof life.29 The study also highlighted a gender disparityin outcomes that needed immediate research attention.Later work found that aneurysms tend to rupture at asignificantly smaller diameter in women than men, andthat the current recommendations for AAA operationleft women exposed to catastrophic ruptures, and worseoverall outcomes.30

Soon afterwards the West Australian Health Serviceadopted a model for care for AAA that aimed to ensurethat all patients with this condition ‘receive the rightcare, at the right time, by the right team and in the rightplace’.31 The key features of this program were the edu-cation of family doctors and the public to promote earlydetection and the elimination of gaps in service provisionfor high risk populations, managed through an integratedsurveillance and intervention system. People are nowmore likely to receive the appropriate form of surgicalcare, rather than being denied access to treatment on thebasis of presumptions about particular patient character-istics such as age or gender. This example shows someof the gaps that result from the uncritical application ofthe results of efficacy studies to the wider population,gaps that are invisible without comprehensive long termstudies of effectiveness. Achieving justice for women andfor octogenarians in access to appropriate treatment forAAA required effectiveness evidence.

Effectiveness studies can also contribute to just healthcare by identifying differences in outcomes across regionsbecause of a lack of access to efficacious care. In thisexample, the known efficacious interventions for breastcancer were found to be of variable effectiveness, result-ing in substantial inequities in outcomes of the treatment

for breast cancer in Western Australia.32 Women whowere initially treated in non-metropolitan hospitalsbetween 1982–2000 had a 50% greater chance of dyingfrom breast cancer than women treated in metropolitanhospitals. When controlled for other factors, analysisindicated that this disparity arose because the efficaciousprotocol of care was not always delivered. The causes ofthis failure were multifactorial but probably included dif-ferential access to intraoperative imaging technologiesand appropriate adjuvant therapies (radio- and chemo-therapy) for patients outside metropolitan areas, and,potentially, differences in the experience, level andapproaches to treatment between rural and urban sur-geons.33 In this case, effectiveness tracked efficacy; thetreatment protocol was good, but the problem was thatwomen outside metropolitan regions did not receive theefficacious treatment. However, this inequity would haveremained invisible, and/or the protocol may have beenabandoned as ineffective, without the evidence generatedby a longitudinal population-based study of effectiveness.Thus achieving equity in breast cancer outcomes reliedupon a study of effectiveness, which identified shortcom-ings in service delivery.34

In addition to investigating the transferability of effi-cacious treatments into the community setting, longtermstudies of effectiveness can help to identify barriersto providing effective care. Given that regulators andpolicymakers are increasingly reluctant to take actionwithout first seeing ‘good quality evidence’, such mea-sures are essential to ensure the just provision of surgicalcare. As the examples above illustrate, effectiveness datacan be collected despite the challenges, and when suchdata are obtained, they can reveal issues of justice andfairness in healthcare. Other studies by the Quality inSurgical Care Project have drawn attention to thesuperiority of specific types of surgical intervention forobesity35 and lower back pain,36 thus providing furthersupport for our claim that the just provision of surgicalservices requires such evidence.

29 P.E. Norman et al. Long term Relative Survival after Surgery forAbdominal Aortic Aneurysm in Western Australia: Population BasedStudy. BMJ 1998; 317: 852–856.30 P.E. Norman & J. Powell. Abdominal Aneurysm: The Prognosis isWorse in Women than in Men Circulation 2007; 115: 2865–2869.31 Department of Health Western Australia. 2008. Abdominal AorticAneurysm Model of Care. Perth: Health Networks Branch.

32 K. Spilsbury et al. Subsequent Surgery after Initial BreastConserving Surgery: A Population Based Study. ANZ J Surg 2005; 75:260–264.33 K.J. Mitchell et al. Rural-urban Differences in the Presentation,Management and Survival of Breast Cancer in Western Australia. TheBreast 2006; 15: 769–776.34 K. Spilsbury, et al. Long-term Survival Outcomes following BreastCancer Surgery in Western Australia. ANZ J Surg 2005; 75: 625–630;C. Clayforth et al. Five-year Survival from Breast Cancer in WesternAustralia over a Decade. The Breast 2007; 16: 375–381.35 F.J. Smith et al. Incidence of Bariatric Surgery and PostoperativeOutcomes: a Population-Based Analysis in Western Australia. MJA2008; 189: 198–202.36 R.E. Moorin & C.D.A.J. Holman. The Impact of the Evolution ofInvasive Surgical Procedures for Low Back Pain: a Population BasedStudy of Patient Outcomes and Hospital Utilization. ANZ J Surg 2009;79: 610–618.



3 BARRIERS TO JUSTICE IN SURGERY

The previous section discussed the importance of usingeffectiveness studies in order to supplement evidencegleaned from efficacy studies. However, improving theuse of effectiveness data in guiding the provision of sur-gical care is but one of the steps needed for surgery tomeet the requirements of justice. Barriers to justice insurgery may also be found within the surgical communityand culture. These include the individualistic nature ofsurgery with its commitment to look after the patient ‘onthe table’; the general disinclination of doctors to engagewith the social determinants of health; and the lack ofuptake of evidence in surgery.

In surgery, as in much of clinical medicine, there is acommitment to the individual patient currently seekingcare. Broader questions of access to care and resourceallocation are often seen as managerial or political issues,rather than clinical issues. Despite the widespread accep-tance of justice as an important principle of medicalethics, these issues are presumed beyond the responsibil-ity of clinicians.37 Accounts of surgical ethics seem to paylittle attention to justice. Aside from periodic discussionsabout the ethical implications of cost constraints andrationing of services, concerns for justice in surgery aretypically focussed on protecting the individual fromunequal treatment regardless of age, race or sex, ratherthan preventing health inequities across populations andcommunities.38 Miles Little describes this focus on indi-vidual patients in surgical ethics as ‘a defining elementin ethical relationships’.39 Yet the examples provided inSection 2 demonstrate that the ethical aspects of surgicalpractice go beyond individual doctor-patient relation-ships. Surgeons have responsibilities to individualpatients, but they also have a collective responsibility tothe general community.

Miles Little’s observation points to the second barrierto justice in surgery: the reluctance of doctors to engagewith broader issues concerning the nature and distribu-tion of morbidity and mortality. It is widely recognizedthat the health of individuals and populations is largelydetermined by social factors, such as location, education,employment and so forth. However, there is evidence(from general practice rather than from surgical practice)that doctors have a limited range of responses in relation

to obvious social inequities affecting their patients’health. These are:

1. Blaming the victim for their disadvantage;2. Feeling sympathy but excluding social problems

from their scope of practice;3. Feeling powerless about social forces affecting the

lives of patients; and4. Attempting to address social disadvantage in a piece

meal way, as best they can.40

Furler and Palmer, commenting on these findings,conclude that together these responses foster collectivesilence as to how physicians should respond to socially-mediated health problems. This silence is not benign.First, refusal to acknowledge that social inequities causeobservable health inequalities allows doctors to assume alevel playing field in which the demands of justice are metby providing equal health care to all who access theirservices. This approach assumes that justice is achievedby treating all patients equally, as if their capacities,access to resources, supports and social capital were allequal. It neglects to notice that patients are not all equalin these respects, and that justice requires us to take suchinequalities seriously. Second, silence suppresses evidencethat doctors themselves contribute to socially medi-ated health inequities through differential treatment ofpatients based upon age, gender, education, ethnicity andso forth.41 There is no comparable evidence concerningsurgeons’ responses to obvious social inequities affectingthe health of their patients; but it seems plausible thattheir responses would resemble those of their medicalcolleagues. Indeed, as surgery is very much an action-based practice, feelings of powerlessness in the face ofsocial factors may be exacerbated in surgeons, with aconcomitant tendency to ignore them or consider themexcluded from the scope of surgical practice. To theextent that surgeons, like other doctors, are reluctantto engage with social factors, inequities may remainunacknowledged and unaddressed.

A third barrier to justice in surgery relates to the roleof, and perceptions about research in surgery. As wehave argued, effectiveness data is essential to informsafe and equitable practice; however collecting, analys-ing and disseminating effectiveness data is painstaking

37 T. Beauchamp & J. Childress. 1996. Principles of biomedical ethics.4th edn: New York: Oxford University press.38 M. Little. The Fivefold Root of an Ethics of Surgery. Bioethics 2002;16: 183–201; N. Daniels. 2008. Just health: meeting health needs fairly.Cambridge: Cambridge University Press.39 M. Little. Ethonomics: The Ethics of the Unaffordable. Arch Surg2000; 135: 17–21,18.

40 M. Harris & S. Knowlden. 1999. Clinical Perspective: a GeneralPractitioner Response to Health Differentials. Sydney: Australian Centrefor Health Promotion; J. Furler & V.Palmer. The Ethics of EverydayPractice in Primary Medical Care: Responding to Social HealthInequities. Philos Ethics Humanit Med 2010; 5: 6.41 M. van Ryn M, Burke J. The Effect of Patient Race and Socio-economic Status on Physicians’ Perceptions of Patients. Soc Sci Med,2000; 50: 813–828; M. van Ryn & S.S. Fu. Paved with Good Intentions:Do Public Health and Human Service Providers Contribute to Racial/ethnic Disparities in Health? Am J Pub Health 2003; 93: 248–255.



work which must involve the whole of the profession,rather than a few research surgeons. There are severalreasons why achieving a shift towards this kind ofresearch culture may be difficult. First, there is nostrong tradition of research in surgery; surgical practicehas been widely criticized for being unscientific andlacking an evidence base.42 This is demonstrated by widevariations in procedures across comparable popula-tions, indicating that surgeons have quite individualisticresponses to similar conditions in their patients.43

Second, there is little agreement within the professionas to what constitutes appropriate research methods.While there is serious and ongoing debate on the issue,44

the focus has tended to be on the need for more RCTsin surgery, and the challenges of performing suchRCTs. This focus has directed attention away fromthe serious and prior question about what kinds ofresearch will best serve the needs of patients. Third,collecting effectiveness data lacks the prestige andexcitement of partaking in an RCT investigating anovel therapy. Effectiveness research does not requirethe imagination and creativity sometimes consideredintegral to the researcher role, and may thereby be lessappealing to researchers and practitioners alike. Next,effectiveness research potentially involves many inthe profession, rather than a few dedicated pioneers,creating challenges in motivating and organizinglarge numbers of individual surgeons working across arange of settings. Finally, funding is scarce for surgicalresearch, yet collecting the large data sets necessary forassessing effectiveness is costly. While device manufac-turers may fund research into a novel device, they areless likely to want to fund longitudinal studies, so thataccessing the resources to perform the research may bedifficult.

For all of these reasons, we are a long way from aculture in surgery in which research is the norm; the‘right’ research is performed; and the results reliably usedto inform practice.45 Yet we need studies of effectiveness,as without these, we are left with inappropriate interven-tions in use, wide variations in practice, and inequitableoutcomes.

CONCLUSION

To avoid creating or exacerbating health inequities,effective surgical interventions must be provided inan equitable manner across patient populations.Failure to do so risks abandoning patients to ineffectivetreatments, unwarrantedly expensive care or unjustpatterns of care. There are compelling reasons to tracksurgical interventions, and to insist that effectivenessresearch complements all promising efficacy research.While studies of effectiveness, like those of efficacy, havetheir limitations in directing individual treatments,46 theinformation they provide is essential for ethical andjust practice. Provision of a high quality and equitableservice requires that procedures be systematicallyappraised for effectiveness throughout their clinicaluse.47 However, achieving a culture in which effective-ness data is collected and acted upon requires a majorshift within the traditions and research practices ofsurgery, and current models of surgical training andeducation. It is beyond our scope here to do more thangesture at the changes necessary to achieve such a shift,which at a minimum would require: action at the level ofindividual surgeons in collecting data and using evi-dence; acceptance by the surgical fraternity of the needfor evidence to justify practice; institutional support forcollecting data and implementing effective treatments;government support for the costs incurred in collectingdata and making results available;48 and education andsupport for patients and the general population betterto understand why effectiveness matters. In particular,surgical practitioners have a responsibility to contributewhere they can towards generating evidence as to theeffectiveness of their preferred interventions; justicerequires more than the application of technical skills forthe benefit of individual patients.

Of course, collecting evidence of effectiveness aloneis no guarantee of equity; nevertheless it is an essentialstep. Further decisions about rationing would remain.However, irrespective of the method of rationing used,information about effectiveness is critical to the deci-sion.49 Only with widespread generation and use ofeffectiveness data can the best and the most just surgicalcare be provided.

42 Lilford et al., op. cit. note 6; P. Ergina et al. Challenges in EvaluatingSurgical Innovation. Lancet 2009; 374(9695): 1097–1104.43 J. Birkmeyer et al. Variation Profiles of Common Surgical Proce-dures. Surgery 1998; 124: 917–923.44 Barkun et al., op. cit. note 6; P. McCulloch et al. No Surgical Inno-vation without Evaluation: the IDEAL Recommendations. Lancet2009; 374(9695): 1105–1112; J. Cook. The Design, Conduct and Analy-sis of Surgical Randomised Trials. Trials. 2009; 10: 9.45 K. Slim et al. Half of the Current Practice of GastrointestinalSurgery is Against the Evidence: A Survey of the French Society ofDigestive Surgery. J Gastroint Surg 2004; 8: 1079–1082.

46 B.C. Reeves. Limitations of Analyses of Effectiveness Using Obser-vational Data. Eur Heart J 2006; 27: 1642–1643; U. Guller. Caveatsin the Interpretation of the Surgical Literature. Br J Surg 2008; 95:541–546.47 Guller, op. cit. note 8; McCulloch et al., op. cit. note 44.48 We note the recent moves towards effectiveness research in the US(J.M VanLare et al. Five Next Steps for a New National Programfor Comparative-effectiveness Research. N Engl J Med 2010; 362: 970–973).49 A.M.Buyx et al. Ethics and effectiveness: Rationing Healthcare byThresholds of Minimum Effectiveness. BMJ 2011; 342: d54.



Acknowledgement

This research was supported in part by an Australian Research GrantDP120101092: ‘How do we know what works? Ethics and evidence insurgical research’, CIs W. Rogers and C. Townley.

Wendy Rogers is Professor of Clinical Ethics at Macquarie Univer-sity. She has a long-standing interest in the ethics of evidence-basedmedicine, and is currently working on three related projects investi-gating the ethics and epistemology of surgical research and innovativepractices.

Christopher Degeling has a background as a practising veterinarian,and in the history and philosophy of science and public health ethics.His research focuses on unravelling the networks embedded withinhealth care practices, and investigating how these are shaped by events,processes, values and meaning to meet social, medical and technologicalneeds.

Cynthia Townley is an epistemologist who has recently published ADefense of Ignorance: its value for knowers and roles in feminist and socialepistemologies (Lanham MD: Lexington Books). Her interests includethe intersection between ethics and epistemology, and animal ethics.



Documents

EQUITY UNDER THE KNIFE: JUSTICE AND EVIDENCE IN SURGERY