DOI: 10.1016/j.athoracsur.2009.03.076 2009;88:40-46 Ann Thorac Surg

Alan M. Speir, Vigneshwar Kasirajan, Scott D. Barnett and Edwin Fonner, Jr Bypass Grafting Patients in Virginia

Additive Costs of Postoperative Complications for Isolated Coronary Artery

http://ats.ctsnetjournals.org/cgi/content/full/88/1/40located on the World Wide Web at:

The online version of this article, along with updated information and services, is

Print ISSN: 0003-4975; eISSN: 1552-6259. Southern Thoracic Surgical Association. Copyright © 2009 by The Society of Thoracic Surgeons.

is the official journal of The Society of Thoracic Surgeons and theThe Annals of Thoracic Surgery

by on July 22, 2013 ats.ctsnetjournals.orgDownloaded from

AIiAEIS

rbtsp

cfS0atdwmti

Ttac$9eccsah

(cCPV

A

PS

AK

©P

AD

ULT

CA

RD

IAC

dditive Costs of Postoperative Complications forsolated Coronary Artery Bypass Grafting Patientsn Virginialan M. Speir, MD, Vigneshwar Kasirajan, MD, Scott D. Barnett, PhD, anddwin Fonner, Jr, DrPH

nova Health System, Falls Church, Virginia Commonwealth University Pauley Heart Center, Richmond, and Virginia Cardiac

urgery Quality Initiative, Norfolk, Virginia

bdfiafi((

fseoaa

Background. Complications after open-heart surgeryesult in an increased length of stay and greater financialurdens for all. The purpose of this study was to measure

he additive costs of postoperative complications forelected subgroups of patients after coronary artery by-ass grafts in the Commonwealth of Virginia.Methods. A multiyear statewide data repository with

linical and billing data was used to measure outcomesor the period 2004 to 2007. The Society of Thoracicurgeons records matched with Universal Billing (UB-4) charge data for all payers were used to estimate thedditive costs of cardiac surgical outcomes using cost-o-charge ratios. Additive cost was defined as theifference between the baseline cost of an average caseith no complications and one with a postoperativeorbidity or mortality. Multivariate analysis was used

o account for important covariates and apportion

ncremental costs.

atpd

csua

M

SSC29marfcmh

ddress correspondence to Dr Fonner, 14808 West 81st Place, Lenexa,S 66215; e-mail: edfonner@gmail.com.

2009 by The Society of Thoracic Surgeonsublished by Elsevier Inc

ats.ctsnetjournDownloaded from

Results. The baseline cost of isolated coronary arteryypass grafting (CABG) cases with no complicationsuring the study period was $26,056. Isolated atrialbrillation was the most frequently cited complicationnd had the lowest additive cost ($2,574). Additive costsor isolated CABG patients were greatest for those casesnvolving prolonged ventilation ($40,704), renal failure$49,128), mediastinitis ($62,773), and operative mortality$49,242).

Conclusions. Additive costs can serve as an indicatoror pursuing quality improvement initiatives. Our resultsuggest additive costs vary according to type of postop-rative complication and comorbidities. Regional collab-rations of multidisciplinary groups in cardiac surgeryre an effective means to implement quality guidelinesnd drive down additive costs.

(Ann Thorac Surg 2009;88:40–6)

© 2009 by The Society of Thoracic Surgeons

he study of costs and outcomes in cardiac surgery iswell-documented. Brown and colleagues [1] studied

he occurrence of seven postoperative complicationsmong Medicare patients and found the average cost oforonary artery bypass graft (CABG) surgery to be32,201 � $23,059 and an average length-of-stay (LOS) of.9 days. In their study group, 13.6% of patients experi-nced some associated complication with an incrementalost increase of $15,468 and 5.2 days LOS. Chen andolleagues [2] found a myocardial infarction after CABGurgery doubled intensive care unit (ICU) time leading to

48% increase in hospital LOS and a 43% increase inospital costs.The Virginia Cardiac Surgery Quality Initiative

VCSQI) is a voluntary consortium of 17 hospitals and 13ardiac surgical practices performing over 99% of theommonwealth of Virginia’s open-heart procedures.reviously documented [3], the principal aim of theCSQI is to promote clinical quality improvement initi-

ccepted for publication March 25, 2009.

resented at the Forty-fifth Annual Meeting of The Society of Thoracicurgeons, San Francisco, CA, Jan 26–28, 2009.

tives through outcomes analysis. The group maintainshat collaborative, regional efforts focusing on loweringostoperative complications will improve efficiency, re-uce resource utilization, and contain costs.The purpose of this study was to measure the additive

osts of postoperative complications within the region inelected subgroups after isolated CABG. These data aresed by the group to identify quality improvement initi-tives that have the potential to generate savings.

aterial and Methods

tudy Populationubjects were 14,780 isolated CABG patients from theommonwealth of Virginia performed between January004 and December 2007. Cases represent approximately9% of all isolated CABG cases performed in the Com-onwealth during the study period. This secondary data

nalysis of registry data was exempt from the institutioneview board and participating institutions were exemptrom Health Insurance Portability and Accountability Actonsideration by the use of the Small Business Agree-ent for Business Associates agreements between each

ospital with their surgical providers and the VCSQI.

0003-4975/09/$36.00doi:10.1016/j.athoracsur.2009.03.076

by on July 22, 2013 als.org

CCc(rirccmiicac

ckTaipbcopgoetaomdcphwi

cecac

ATiscwnpw

TTv

oaifccmc

ABmtaatttopsp

SCdqpo(speeetcp(ccst

R

Ab1ftpefppp�p

41Ann Thorac Surg SPEIR ET AL2009;88:40–6 ADDITIVE COSTS OF POSTOP CABG PATIENTS

AD

ULT

CA

RD

IAC

linical and Financial Datalinical and financial data were linked by matching

ardiac surgery patients’ Society of Thoracic SurgeonsSTS) records with their universal billing dischargeecords. Matched data facilitate the analysis of outcomesn relation to incurred hospital costs and in-patientesource use for subgroups of patients. Postoperativeomplications studied include atrial fibrillation (isolatedases and those occurring with other complications),ediastinitis, permanent stroke, reoperations for bleed-

ng, renal failure, prolonged ventilation, and death. Eachnstitution was responsible for coding and submitting itsardiac data to VCSQI’s repository, while all participantsgreed to guidelines regarding data definitions, dataollection, and timely submission as set forth by the STS.

The standard uniform bill (UB) for institutional healthare providers used throughout the U.S., currentlynown as the UB-04, is the patient’s final hospital bill.he UB-04 record contains revenue codes, service codesnd descriptions, quantities, and charge data per unit fornpatient services. The VCSQI developed a two-steprocedure for making the financial data more compara-le from one provider organization to another. First,harges for all of the ICD-9 (International Classificationf Diseases, 9th Revision) revenue codes contained in theatient’s record were aggregated into 21 logicallyrouped cost categories (see list in the Appendix). Sec-nd, a set of cost-to-charge ratios (CCRs) submitted fromach provider organization was multiplied by the sums ofhe charges in each of the 21 categories. The product was

set of normalized charges relatively comparable fromne surgeon and hospital provider to another. An esti-ate of the total cost of an individual inpatient stay was

erived by summing the normalized charges over the 21ost categories. Differences between cost items from onerovider to another may be due to unique individualospital coding and billing patterns. Isolated CABG casesere combined over the four-year period and were not

nflation adjusted.Baseline costs were defined as the average cost of those

ases without postoperative complications. The differ-nce between the average baseline cost of uncomplicatedases and cases with a morbidity or mortality was defineds an additive cost. Three methods of estimating additiveosts were tested.

dditive Costs for Isolated Complicationshe first method of calculating additive costs examined

solated complications; ie, those cases that had only onepecific complication coded in the STS record. During theourse of our research, the cost estimates of this methodere judged to vary more widely due to the smallerumber of cases with isolated complications. For exam-le, rarely will a case of mediastinitis occur in isolationithout some other comorbidity.

otal Additive Costshe second method of estimating additive costs in-

olved examining cases where a specific complication a

ats.ctsnetjournDownloaded from

ccurs concurrently with other postoperative second-ry complications. The additive cost involves subtract-ng the baseline cost of cases without a complicationrom the average total cost for cases with a primaryomplication and other reported postoperative compli-ations. These estimates of additive costs are generallyore stable and greater than those for isolated

omplications.

llocated Costsecause we found that postoperative complications nor-ally occur with other secondary complications rather

han in isolation, we developed a method for allocatingdditive costs to the complication being studied and itsssociated complications. The method of allocating addi-ive costs between the primary and secondary complica-ions involves a linear system weighting each complica-ion occurring in a patient group proportional to itsccurrence. The total additive cost for a particular com-lication can then be allocated to the complication beingtudied and all other complications associated with thatrimary complication.

tatistical Analysisontinuous data are presented as mean � standardeviation (SD). Categoric data are presented as fre-uency and percent. Tabular cost data are presenteder thousands. Statistical comparisons were made withne-way analysis of variance. Observed to expectedO/E) ratios were calculated by dividing the number oftratum specific events by the stratum specific summedatient STS calculated predicted risk. A linear, mixedffects multivariate model with hospital as a randomffect was constructed to model the effects of postop-rative complications on total cost after adjustment forhe potential confounding effects of demographics andomplications to provide the estimated additive coster modeled covariate. The upper and lower deciles

10%) were excluded from the regression analysis toontrol for extreme cases. All statistical analyses wereonducted in SAS (Version 9.1; Cary, NC). Statisticalignificance was assumed for p values 0.05 or less,wo-tailed.

esults

breakdown of the isolated CABG study populationy demographics and risk factors can be found in Table. Nearly 80% of the patients were Caucasian, 25%emales, and an average age of 64.1 years. Over 78% ofhe cases were elective procedures. During the studyeriod several demographic trends were observed. Forxample, the percent of Caucasian patients declinedrom 81.2% in 2001 to 73.5% in 2007 (p � 0.05), theercent of New York Heart Association class III, IVatients declined (66.1% to 57.8%, p � 0.05), and theercent of female patients declined (26.4% to 24.7%, p

0.05). Several risk factors increased over the studyeriod: arrhythmia (6.5% to 8.7%, p � 0.05); peripheral

rtery disease (14.0% to 16.3%); and hypertension

by on July 22, 2013 als.org

(o

pqc(

tp3ihlsineC

ca�of(fia

tr4c

TC

V

ABCFCEECDHFDHHHLHNPPCAPPELL

CdA

T

C

PAMRPPRO

a

O

TA

C

EIAMRPPRO

a

C

42 SPEIR ET AL Ann Thorac SurgADDITIVE COSTS OF POSTOP CABG PATIENTS 2009;88:40–6A

DU

LTC

AR

DIA

C

77.2% to 82.9%, p � 0.05). No similar trends werebserved for postoperative complications.Seventy percent of the study group experienced no

ostoperative complications (Table 2). The most fre-uently reported complication was atrial fibrillationombined with some other postoperative complication15.2%). The least frequently reported was mediastini-

able 1. Clinical Characteristics and Postoperativeomplications

ariablesCABG-Onlyn � 14,780

ge (years), mean � SD 64.1 � 10.7ody mass index, kg/m2 29.6 � 10.7reatinine 1.2 � 1.0emale, n (%) 3,801 (25.7)aucasian 11,549 (78.1)jection fraction 0.518 � 0.124jection fraction � 30% 747 (5.1)hronic lung disease 251 (17.0)iabetes 5,737 (38.8)ypertension 11,722 (79.3)

amily history CAD 5,913 (40.0)yslipidemia 11,539 (78.1)istory smoking 8,451 (57.2)istory renal failure 520 (3.5)eart failure 1,713 (11.6)eft main disease 4,408 (29.8)istory of myocardial infarction � 7 days 3,178 (21.5)YHA class (III, IV) 9,041 (61.2)eripheral artery disease 2,190 (14.8)rior stroke 1,060 (7.2)erebral vascular disease 1,947 (13.2)rrhythmia 1,028 (7.0)rior CABG 530 (3.6)rior valve 36 (0.2)lective procedure 11,602 (78.5)OS, Admit to discharge 9.0 � 7.9OS, Surgery to discharge 6.6 � 7.0

ABG � coronary artery bypass grafting; CAD � coronary arteryisease; LOS � length of stay; NYHA � New York Heartssociation.

able 2. Frequency and Percent of Patients with Postoperative

omplication Any Complicatio

atients with no complications 13,852 (70.0%)trial fibrillationa 3,013 (15.2%)ediastinitis 73 (0.4%)

eoperation for bleeding 361 (1.8%)ermanent stroke 252 (1.3%)rolonged ventilation time 1,751 (8.8%)enal failure 738 (3.7%)perative mortality 344 (1.7%)

Includes other postoperative complications and (or) operative mortality

/E � observed to expected.

ats.ctsnetjournDownloaded from

is (73 patients, 0.4%). Operative mortality was re-orted in 1.7% of all isolated CABG procedures (n �44). Risk-adjusted outcomes measures are also shownn Table 2 for the postoperative complications studiedere along with operative mortality. The study popu-

ation had fewer cases of mediastinitis, permanenttroke, and operative mortality than expected. Thencidence of renal failure was equivalent to the U.S.orm (O/E � 0.98). The VCSQI had a higher-than-xpected incidence of prolonged ventilation cases afterABG surgery (O/E � 1.53).Average LOS for CABG-only cases with no compli-

ations (baseline) was 7.4 days � 5.8 (see Table 3). Theddition of isolated atrial fibrillation cases resulted in a2.2 day increase (p � 0.001). However, a complication

f either mediastinitis (�30.4 days, p � 0.001), renalailure (�16.3 days, p � 0.001), or permanent stroke�15.3 days, p � 0.001) resulted in the greatest increaserom baseline LOS. Cases involving operative mortal-ty averaged 18.3 days, nearly 11 days above theverage for cases with no complications.The average cost of an isolated CABG case during

he study period was $30,654 (median � $24,414),egardless of frequency or type of complication. Table

presents the additive cost per case above baselineost ($26,056) for cases with no postoperative compli-

plications

Only Complication O/E Ratios

— —2,264 (11.4%) Not available

12 (0.06%) 0.52191 (0.96%) Not available97 (0.49%) 0.76

915 (4.61%) 1.53264 (1.33%) 0.98

Not applicable 0.75

able 3. Unadjusted Average Length of Stay (Days) forddition of Postoperative Complications

ABG-only 2004 to 2007Length of Stay

Admission to Discharge

xpected LOS: No Complications 7.4 � 5.8solated atrial Fibrillation 9.6 � 9.3trial fibrillationa 12.8 � 13.3ediastinitis 37.8 � 29.0

eoperation for bleeding 12.5 � 10.9ermanent stroke 22.7 � 21.9rolonged ventilation time 20.5 � 18.1enal failure 23.7 � 21.6perative mortality 18.3 � 22.0

Includes other postoperative complications and (or) operative mortality.

ABG � coronary artery bypass grafting; LOS � length of stay.

Com

n

.

by on July 22, 2013 als.org

ci0Ocfvp

acea$tccTtvc

t

foatiw((crlr

C

Wfdthpha

TC

T

NIAMPRPROO

a mplic$

T(

C

AMPRPRO

a

43Ann Thorac Surg SPEIR ET AL2009;88:40–6 ADDITIVE COSTS OF POSTOP CABG PATIENTS

AD

ULT

CA

RD

IAC

ations. Cost increases ranged from a low of $2,744 forsolated atrial fibrillation (�10.3% above baseline, p �.001) to $62,744 for mediastinitis (�240%, p � 0.001).perative mortality added, on average, $49,244 to the

ost of uncomplicated care (�189%, p � 0.001). Renalailure added $49,144 (�188%, p � 0.001), prolongedentilation added $40,744 (�156%, p � 0.001), andermanent stroke added $34,144 (�131%, p � 0.001).Table 5 shows the average costs of major postoper-

tive complications and allocates costs to the primaryomplication and other secondary complications. Forxample, the total cost of mediastinitis ($88,829) had andditive cost of $62,773. Of this additional amount,23,499 was attributed to the infection; $39,274 was dueo other secondary complications. In most cases, theosts allocated to the associated complications ex-eeded the cost attributed to the primary complication.he exception was prolonged ventilation, where two-

hirds of the total additive cost was attributed toentilator use and one-third to other associatedomplications.

Table 6 estimates additive costs using a mixed mul-ivariate model to measure differences after controlling

able 4. Additive Cost of Patients With No Postoperative Comomplications

otal Cost No. of Patientsa

o complications (baseline) 10,515solated atrial fibrillation (no mortality) 1,575trial fibrillationb 2,092ediastinitis 51

ermanent stroke 185eoperation for bleeding 274rolonged ventilation time 1,236enal failure 520perative mortality 265ther complications 1,699

Patients with cost data available; b Includes other postoperative co5,000.

able 5. Total, Additive, and Allocated Costs of PostoperativeBaseline Cost � $26,056)

omplication Total Cost Additiv

trial fibrillationa $38.1 $1ediastinitis $88.8 $6

ermanent stroke $60.1 $3eoperation for bleeding $46.1 $2rolonged ventilation time $66.7 $4enal failure $75.1 $4perative mortality $75.3 $4

Includes other postoperative complications and (or) operative mortality.

ats.ctsnetjournDownloaded from

or institution. When patient characteristics and post-perative complications are included in a multivari-ble regression, 28.9% of the variation in cost was dueo hospital variation. Baseline risk-adjusted cost of ansolated CABG was $24,479. A myocardial infarctionithin 7 days ($1,721, p � 0.001), previous CABG

$3,278, p � 0.001), and preoperative renal failure$3,830, p � 0.001) resulted in the greatest risk-adjustedost increases. Postoperative complications such as aeoperation for bleeding ($5,088, p � 0.001) and pro-onged ventilation time ($4,420) resulted in the greatestisk-adjusted cost increases.

omment

hile the trend of catheter-based interventions per-ormed by cardiologists has increased over the pastecade, the annual volume of open heart procedures in

he U.S. remains a significant contributor to totalealth care costs. Acute awareness on the part ofayers, policy analysts, administrators, and providersas resulted in careful evaluation of the clinical resultsnd financial impact of cardiac surgical care. In re-

ations Compared With Patients With Postoperative

Mean % Increase Cost Difference p Value

$26.1 — — —$28.8 10.3% �$ 2.7 0.001$38.1 46.0% �$12.0 0.001$88.8 240% �$62.7 0.001$60.2 131% �$34.1 0.001$46.1 76.6% �$20.0 0.001$66.8 156% �$40.7 0.001$75.2 188% �$49.1 0.001$75.3 189% �$49.2 0.001$56.4 116% �$30.3 0.001

ations and (or) operative mortality. Excludes patients with total cost �

plications Due to Primary and Secondary Complications

st

Reason for Additive Cost

PrimaryComplication

SecondaryComplication

Cost % Cost %

$2.7 23.0% $9.3 77.0%$23.5 37.4% $39.3 62.6%$9.8 28.9% $24.2 71.1%$4.0 19.9% $16.1 80.1%

$25.7 63.4% $14.9 36.6%$22.9 46.7% $26.2 53.3%$11.0 22.4% $38.1 77.6%

plic

Com

e Co

2.12.74.10.00.79.19.2

by on July 22, 2013 als.org

schsh

wccogfsey

miteiob

atcaadtscootr

iimtot(cRcais

wocsatsaaRaiftr(wum

rtr

T

E

IENR

N

FA

HAC

PCDFHRHPSP

MLAMPRPO

a

Cd

44 SPEIR ET AL Ann Thorac SurgADDITIVE COSTS OF POSTOP CABG PATIENTS 2009;88:40–6A

DU

LTC

AR

DIA

C

ponse to resource constraints and increased scrutiny,ardiac surgeons along with their hospital partnersave been leaders in attempts to maintain quality inurgical outcomes while decreasing the costs of openeart surgical procedures.The Virginia Cardiac Surgery Quality Initiative hasorked to align physicians and hospitals in Virginia by

oordinating efforts to maintain quality while reducingosts. The group’s database combines clinical results inpen heart surgical procedures (standardized by STSuidelines) with uniform patient-specific financial dataor each hospitalization. Baseline costs for cardiacurgical procedures and the relative impact of postop-rative additive costs have been estimated. This anal-

able 6. Multiple Regression Analysis of Costa

ffectAdditive Cost

Estimate p Value

ntercept 24,848.0 �0.0001jection fraction �0.331 �0.0001onelective procedure 1,303.2 �0.0001ace:White (vs other) -436.4 0.0076Black (vs other) -86.0 0.6904ew York Heart Assoc. class III,

IV (vs I, II)724.1 �0.0001

emale 677.2 �0.0001ge:Age 70 to 79 years (vs � 70) 502.7 �0.0001Age � 80 years (vs � 70) 1,321.1 �0.0001eart failure 1,247.3 �0.0001rrhythmia 1,377.9 �0.0001hronic obstructive pulmonary

disease561.1 �0.0001

ulmonary artery disease 761.9 �0.0001erebrovascular disease 607.8 0.0002iabetes mellitus 581.9 �0.0001amily history of CAD -111.1 0.2789ypertension 188.9 0.1129enal failure 3,830.2 �0.0001istory smoking �76.0 0.4485revious CABG 3,277.6 �0.0001econd or greater CABG �2,283.8 �0.0001revious cardiovascular

intervention694.6 �0.0001

yocardial infarction � 7days 1,720.8 �0.0001eft main disease 296.8 0.0055trial fibrillation 1,252.3 �.0001ediastinitis 3,785.5 0.0063

ermanent stroke 2,931.6 �.0001eoperation for bleeding 5,088.1 �.0001rolonged ventilation time 4,420.0 �.0001perative death �463.6 0.3522

Upper and lower cost deciles (10%) were excluded.

ABG � coronary artery bypass grafting; CAD � coronary arteryisease.

sis has stimulated discussion on variations in perfor- r

ats.ctsnetjournDownloaded from

ance measures across hospitals and has helpeddentify evidenced-based best practices with the po-ential to improve outcomes and contain costs. Forxample, while the $2,574 additive cost associated withsolated atrial fibrillation is relatively low, its frequentccurrence led to the introduction of an amiodarone-ased atrial fibrillation prevention prophylaxis in 2006.Also, it was noted that the additive costs allocated toprimary complication in most cases is often less than

he amount attributed to associated complications (ex-ept in cases involving prolonged ventilation). Reevesnd Murphy [4] found that red blood cell transfusionsmong cardiac surgery patients increased the inci-ence of infections, renal failure, prolonged ventila-

ion, low cardiac index, myocardial infarction, andtroke. These complications were associated with in-reased ICU time, increased cost, and increased post-perative LOS. This led to a statewide implementationf an intraoperative and postoperative red blood cellransfusion protocol aimed at reducing transfusionates.

While analysis is underway to document the resultingnitial impact of these protocols on subgroups of patientsn Virginia, VCSQI is also tying postoperative glucose

anagement, extubation practices, and infection controlo evidenced-based guidelines and the formulation ofther protocols. The importance of these focused initia-ives is established by the additive costs of mediastinitisgreater than $60,000) and the higher-than-expected oc-urrence of prolonged ventilation use in our region.educing the incidence of relatively modifiable compli-ations such as prolonged ventilation time, transfusionssociated complications, and infection control can resultn significant cost reductions for hospitals and lengths oftay for patients.

There are several limitations to this study. First, dataere generated from a nonrandomized, retrospectivebservational cardiac surgical database (althoughases represent a nearly 100% census of all cardiacurgical procedures occurring in the state). The vari-nces attributed to the unique characteristics of par-icipating hospitals were not taken into account in thetatistical analysis. For example, hospital variance had

significant impact on the cost analysis of coronaryrtery bypass in the New York State Cardiac Surgeryeporting System [5]. Investigators found hospital siteccounted for an additional 8% variation beyond thempact of patient factors. Third, there were additionalactors of payer mix or socioeconomic status of patientshat were not considered in the analysis. Previouseports suggest health maintenance organizationHMO) patients have lower costs for bypass surgeryhen compared with Medicaid patients [6]. Fourth, these of cost-to-charge ratios in the cost estimationethodology has known benefits and weaknesses.A discussion on potential savings resulting from a

eduction in the incidence of postoperative complica-ions is incomplete without mentioning the investmentequired to achieve such goals. Some initiatives may

equire sizeable upfront expenditures followed by

by on July 22, 2013 als.org

lfcinvseiicV

TAa

R

1

2

3

4

5

6

A

CA

C

RIPIRCGCTAORIELBRDTO

P

Cit

D

DWttoQo

etocytf

mctit

45Ann Thorac Surg SPEIR ET AL2009;88:40–6 ADDITIVE COSTS OF POSTOP CABG PATIENTS

AD

ULT

CA

RD

IAC

ower on-going costs. Others may require limited up-ront investments while higher expenses may be in-urred over the course of an outcomes improvementnitiative. A comprehensive assessment of estimatedet savings should include the direct savings of pro-ided care and the hidden costs of the infrastructure,upport personnel, and technology necessary for suchndeavors. Nonetheless, the inclusion of additive costn the quality improvement dialogue has been thempetus for meaningful progress by hospitals andardiac surgical practices in the Commonwealth ofirginia.

he authors would like to thank Clifford E. Fonner and theRMUS Corporation for the preparation of tables, the statistics,nd the contribution to the study methodology.

eferences

. Brown PP, Kugelmass AD, Cohen DJ, et al. The frequency andcost of complications associated with coronary artery bypassgrafting surgery: results from the United States Medicareprogram. Ann Thorac Surg 2008:85:1980–7.

. Chen JC, Kaul P, Levy JH, et al. Myocardial infarction follow-ing coronary artery bypass graft surgery increases healthcareresource utilization. Critical Care Med 2007:35:1296–301.

. Rich JB, Speir AM, Fonner E Jr; Virginia Cardiac SurgeryQuality Initiative. Making a business case for quality byregional information sharing involving cardiothoracic sur-gery. Am Heart Hosp J 2006;4:142–7.

. Reeves BC, Murphy GJ. Increased mortality, morbidity, andcost associated with red blood cell transfusion after cardiacsurgery. Curr Opin Cardiol 2008:23:607–12.

. Cowper PA, DeLong ER, Peterson ED, et al. Variability in costof coronary bypass surgery in New York State: potential forcost savings. Am Heart J 2002:143:130–9.

. Mushinski M. Average charges for coronary artery bypass

grafts and percutaneous transluminal coronary angioplasties,1995. Stat Bull Metrop Insur Co 1997;78:20–8. i

he same clinical platform and common financial formatting.

Tampmag

s

a

rcit9q

fin

ats.ctsnetjournDownloaded from

ppendix

ost Categories and Revenue Codes Used in Costnalysis

ost Category Revenue Codes

egular room/step down 100-179CU/CCU 200-219harmacy 250-259

V 260-269espiratory therapy 410-419ardiac cathode lab 481eneral supplies 270-274, 276-277, 279ardiac diagnostics 480, 482-489, 730-731, 739herapies (PT, OT, cardiac rehab) 420-449, 460-469, 943nesthesia 370-379perating room 360-369, 490-499adiology (including MRI and CT) 320-359, 400-409

mplants (pacers, ICD, valve) 275, 278mergency room 450-459ab 300-319lood 380-399ecovery room 710-719ialysis 800-809, 820-859, 880-889elemetry 732ther – miscellaneous 180-199, 220-249, 280-299,

470-479, 500-679, 700-709, 740-799, 901-920,922-942, 944-999

eripheral vascular lab 921

T � computed tomography; CCU � cardiac care unit; ICU �ntensive care unit; ICD � implantable cardioverter defibrilla-or; IV � intravenous; MRI � magnetic resonance imag-

ng; OT � occupational therapy; PT � physical therapy.

ISCUSSION

R T. BRUCE FERGUSON (Greenville, NC): Dr Chitwood, Drood, members, and guests. I have no disclosures. I would like

o congratulate Dr Speir and his colleagues on this work and forheir timely sharing of their manuscript. I have been an admirerf the accomplishments of the VCSQI [Virginia Cardiac Surgeryuality Initiative] for a number of years and I appreciate the

pportunity to discuss this work.The reliable combination of clinical and cost data has been

lusive, even at the individual hospital level, for a variety ofechnical, administrative, and agenda reasons. The importancef clinical data linked to financial data is to give the right clinicalontext to the financial information. While this hypothesis haset to be fully tested against administrative data alone, I believehis accomplishment by the VCSQI is an important step forwardor clinical information use in medicine.

(Slide) Important attributes of this project are listed here fromy perspective. This is a sustainable cardiothoracic surgical

ollaboration for the purpose of improving quality. The innova-ions they have brought to the table include the breakthroughntegration of clinical and financial data across the group, using

he partnership between the local providers and their hospitalsnd the hospitals and providers together as participants in aultisite collaboration is absolutely critical to the success of this

rogram. The regional structure provides for important align-ent of agendas without which this kind of activity could not be

ccomplished, and I believe in some context this gives us alimpse of the future.(Slide) We are used to looking at risk adjustment for our

urgical procedures.(Slide) What we need to be able to move to is to look at risk

djustment in the context of the cost of care delivery processes.A second implication of these studies is the local level and

egional analysis that has been demonstrated here, where theombination of data from different data sets is brought togethern a unique way. We have found, as have others who have donehis locally, that the fidelity of data matching needs to exceed5% or greater for meaningful analysis to occur. I have twouestions for Dr Speir.First, has the VCSQI had a similar experience with clinical

nancial matching and data validity and have you examined this

ot only as a group but in terms of interhospital variability? And

by on July 22, 2013 als.org

sagi[t

Dttgaottag

qmccta

fswramc

DoFvs

Da$1ni

Dh

O

ETFTtigtS

itwwiiWevW

46 SPEIR ET AL Ann Thorac SurgADDITIVE COSTS OF POSTOP CABG PATIENTS 2009;88:40–6

©P

AD

ULT

CA

RD

IAC

econdly, given the availability of these data, particularly thege-related and gender-related information, have you as aroup discussed the potential impact on patient selection? That

s, have you begun to couple a financial risk with the STSSociety of Thoracic Surgeons] predicted risk? I would like tohank the Society for the privilege of discussing this paper.

R SPEIR: Thank you, Dr Ferguson, for your kind remarks. Ihink that data validation is the Achilles heel of any database. Inhe case of our database, we lost a half million dollars, as thisroup is self-funded, in the first four and a half years because oflignment with our initial third-party administrator, who couldnly match the costs with the clinical outcomes in 68% of theime. With the ARMUS Corporation, we are matching at 99% forhe past five years, and I think are back on track effectively andccurately. That is not a plug for ARMUS, but they have done areat job and have been worth it.In terms of our financial modeling, it is our project in the first

uarter of this year to attempt to use the STS risk assessmentethodology to formulate a financial model that is predictive of

osts depending on the patient’s preoperative risk factors. If wean do this successfully, we may then be able to use variation inhe hospitalization cost from the STS clinical database outcomes

nline Discussion Forum

eb site.

AS

LAS

Da

TTMRSQPFe

2009 by The Society of Thoracic Surgeonsublished by Elsevier Inc

ats.ctsnetjournDownloaded from

orm and clinical record. For example, if a patient’s cost ofurgical care is inconsistent with their STS database information,ere the risk factors or complications accurately entered? In

egard to your second question, we have not used predictive costs a criterion to exclude any patients for cardiac surgery. Thatay be in the future. Thank you again for your supportive

omments.

R W. RANDOLPH CHITWOOD, JR. (Greenville, NC): I havene question. The cost of the study and the work is self-funded.rom where does this funding come? Does it come from indi-idual hospitals? Could you possibly fund this program at thetate level? Have you tried to get state funding for this project?

R SPEIR: Thank you. Each of the hospitals contributes $5,000nd each of the cardiac surgical practices have gratefully put in2,500 annually. We have raised internally over $2 million in the5 years. We have not been able to get any state funding or anyational funding, but, President Chitwood, if you have any

nterest in this project, we would be grateful.

R CHITWOOD: Well, since they are putting $20 billion inealth IT [information technology] in the stimulus package,

s a trigger for internal or external audit of the STS database maybe a little dust will fall off the table for us. Thank you.

ach month, we select an article from the The Annals ofhoracic Surgery for discussion within the Surgeon’sorum of the CTSNet Discussion Forum Section.he articles chosen rotate among the six dilemma

opics covered under the Surgeon’s Forum, whichnclude: General Thoracic Surgery, Adult Cardiac Sur-ery, Pediatric Cardiac Surgery, Cardiac Transplanta-ion, Lung Transplantation, and Aortic and Vascularurgery.Once the article selected for discussion is published

n the online version of The Annals, we will post a no-ice on the CTSNet home page (http://www.ctsnet.org)ith a FREE LINK to the full-text article. Readersishing to comment can post their own commentary

n the discussion forum for that article, which will benformally moderated by The Annals Internet Editor.

e encourage all surgeons to participate in this inter-sting exchange and to avail themselves of the otheraluable features of the CTSNet Discussion Forum and

For July, the article chosen for discussion under thedult Cardiac and Pediatric Cardiac Surgery Dilemmaection of the Discussion forum is:

ong-Term Results of the Leaflet Extension Technique inortic Regurgitation: Thirteen Years of Experience in aingle Center

ong Seop Jeong, MD, PhD, Kyung-Hwan Kim, MD, PhD,nd Hyun Ahn, MD, PhD

om R. Karl, MDhe Annals Internet Editorater Children’s Hospital

aymond Terrace. Brisbane, 4101ueensland, Australiahone: (�61) 7-3163-7594ax: (�61) 7-3163-5308

mail: tom.karl@mater.org.au

Ann Thorac Surg 2009;88:46 • 0003-4975/09/$36.00

by on July 22, 2013 als.org

DOI: 10.1016/j.athoracsur.2009.03.076 2009;88:40-46 Ann Thorac Surg

Alan M. Speir, Vigneshwar Kasirajan, Scott D. Barnett and Edwin Fonner, Jr Bypass Grafting Patients in Virginia

Additive Costs of Postoperative Complications for Isolated Coronary Artery

& ServicesUpdated Information

http://ats.ctsnetjournals.org/cgi/content/full/88/1/40including high-resolution figures, can be found at:

References http://ats.ctsnetjournals.org/cgi/content/full/88/1/40#BIBL

This article cites 6 articles, 1 of which you can access for free at:

Citations http://ats.ctsnetjournals.org/cgi/content/full/88/1/40#otherarticles

This article has been cited by 16 HighWire-hosted articles:

Subspecialty Collections

http://ats.ctsnetjournals.org/cgi/collection/professional_affairs Professional affairs

http://ats.ctsnetjournals.org/cgi/collection/coronary_disease Coronary disease

following collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

.healthpermissions@elsevier.comemail: orhttp://www.us.elsevierhealth.com/Licensing/permissions.jsp

in its entirety should be submitted to: Requests about reproducing this article in parts (figures, tables) or

Reprints reprints@elsevier.com

For information about ordering reprints, please email:

by on July 22, 2013 ats.ctsnetjournals.orgDownloaded from