Effect of gender on long-term survival after abdominal aortic aneurysm repair based on results from

18 Pages·2011·0.88 MB·English

View metadata, citation and similar papers at core.ac.uk brought to you by CORE
provided by Elsevier - Publisher Connector
CLINICAL RESEARCH STUDIES
Fromthe Peripheral Vascular SurgerySociety
Effect of gender on long-term survival after
abdominal aortic aneurysm repair based on results
from the Medicare national database
NataliaN.Egorova,PhD,MPH,aAgelikiG.Vouyouka,MD,bJamesF.McKinsey,MD,c
PeterL.Faries,MD,bK.CraigKent,MD,dAlanJ.Moskowitz,MD,aandAnnetineGelijns,PhD,a
NewYork,NY;andMadison,Wisc
Objectives:Historically,womenhavehigherprocedurallyrelatedmortalityratesthanmenforabdominalaorticaneurysm
(AAA)repair.Althoughendovascularaneurysmrepair(EVAR)hasimprovedtheseratesformenandwomen,effectsof
genderonlong-termsurvivalwithdifferenttypesofAAArepair,suchasEVARvsopenaneurysmrepair(OAR),need
furtherinvestigation.Toaddressthisissue,weanalyzedsurvivalinmatchedcohortswhoreceivedEVARorOARforboth
elective(eAAA)andrupturedAAA(rAAA).
Methods:UsingtheMedicareBeneficiaryDatabase(1995-2006),wecompiledacohortofpatientswhounderwentOAR
or EVAR for eAAA (n (cid:1) 322,892) or rAAA (n (cid:1) 48,865). Men and women were matched by propensity scores,
accountingforbaselinedemographics,comorbidconditions,treatinginstitution,andsurgeonexperience.Frailtymodels
wereusedtocomparelong-termsurvivalofthematchedgroups.
Results:PerioperativemortalityforeAAAswassignificantlyloweramongEVARvsOARrecipientsforbothmen(1.84%
vs4.80%)andwomen(3.19%vs6.37%,P<.0001).Onedifference,however,wasthatthesurvivalbenefitofEVARwas
sustainedforthe6yearsoffollow-upinwomenbutdisappearedin2yearsinmen.Similarly,thesurvivalbenefitofmen
vswomenafterelectiveEVARdisappearedafter1.5to2years.ForrAAAs,30-daymortalitywassignificantlylowerfor
EVARrecipientscomparedwithOARrecipients,forbothmen(33.43%vs43.70%P<.0001)andwomen(41.01%vs
48.28%,P(cid:1).0201).Six-yearsurvivalwassignificantlyhigherformenwhoreceivedEVARvsthosewhoreceivedOAR
(P(cid:1).001).However,thesurvivalbenefitforwomenwhoreceivedEVARcomparedwithOARdisappearedin6months.
SurvivalwasalsosubstantiallyhigherformenthanwomenafteremergentEVAR(P(cid:1).0007).
Conclusions:Genderdisparityisevidentfromlong-termoutcomesafterAAArepair.InthecaseforrAAA,wherethelong-term
outcomeforwomenwassignificantlyworsethanformen,thelessinvasiveEVARtreatmentdidnotappeartobenefitwomen
tothesameextentthatitdidformen.Althoughthelong-termoutcomeafteropenrepairforelectiveAAAwasalsoworsefor
women,EVARbenefitforwomenwassustainedlongerthanformen.Theseassociationsrequirefurtherstudytoisolate
specificriskfactorsthatwouldbepotentialtargetsforimprovingAAAmanagement. (JVascSurg2011;54:1-12.)
Elective or urgent endovascular repair (EVAR) of vascular Repair (OVER) clinical trials,2,3 as well as
abdominal aortic aneurysms (AAA) has been widely ac- population-based observational studies,4,5 have demon-
cepted as the treatment of choice for these lesions.1 The strated improved 30-day mortality and morbidity with
DutchRandomisedEndovascularAneurysmManagement EVARcomparedwithopenaneurysmrepair(OAR).How-
(DREAM),ComparisonofEndovascularAneurysmRepair ever,theyfailedtoshowsustainedsurvivalbenefitafter2to
withOpenRepairinPatientswithAbdominalAorticAn- 3yearsfromtheinitialsurgery.2-4
eurysm(EVAR1),andVeteransAffairsOpenversusEndo- Historically,thenaturalhistory,aswellasthemanage-
ment of AAA in women has been associated with worse
From the Department of Health Evidence and Policy,a and Division of overalloutcomes.6-11EVARhasimprovedimmediateout-
Vascular Surgery,b Mount Sinai School of Medicine, New York; the
comesinwomencomparedwithOAR;however,agender
DivisionofVascularSurgery,ColumbiaUniversity,MedicalCenter,New
Yorkc;andtheDepartmentofSurgery,UniversityofWisconsinSchoolof disparity in mortality persists.6,9,12,13 Whether this initial
MedicineandPublicHealth,Madison.d benefitissustainedinlong-termfollow-upforwomenand
Competitionofinterest:none.
Presented at the Thirty-fifth Annual Spring Meeting of the Peripheral
VascularSurgerySociety,June11-12,2010,Boston,Mass. Theeditorsandreviewersofthisarticlehavenorelevantfinancialrelationships
Additionalmaterialforthisarticlemaybefoundonlineatwww.jvascsurg.org. todisclosepertheJVSpolicythatrequiresreviewerstodeclinereviewofany
Reprint requests: Ageliki G. Vouyouka, MD, FACS, Division of Vascular manuscriptforwhichtheymayhaveacompetitionofinterest.
Surgery,MountSinaiMedicalCenter/SchoolofMedicineand5E98thSt. 0741-5214/$36.00
4th Flr, Box 273, New York, NY 10029 (e-mail: ageliki.vouyouka@ Copyright©2011bytheSocietyforVascularSurgery.
mountsinai.org). doi:10.1016/j.jvs.2010.12.049
1
JOURNALOFVASCULARSURGERY
2 Egorovaetal July2011
TableI. Demographicsandcomorbiditiesforpatientswhounderwentelectiveendovascular(EVAR)oropen(OAR)
repairofabdominalaorticaneurysm
Overall EVARa
Men Women Men Women
Variable (n(cid:1)250,438) (n(cid:1)72,454) P (n(cid:1)63,712) (n(cid:1)13,644) P
Age,years 74.97 75.86 (cid:2).0001 76.16 77.45 (cid:2).0001
Agegroups,%
65-69 22.8 18.5 (cid:2).0001 18.9 13.8 (cid:2).0001
70-74 29.8 28.1 (cid:2).0001 25.7 23.0 (cid:2).0001
75-79 27.1 28.7 (cid:2).0001 27.5 28.2 .0928
80-84 15.1 17.6 (cid:2).0001 19.4 22.5 (cid:2).0001
(cid:1)85 5.2 7.1 (cid:2).0001 8.5 12.5 (cid:2).0001
Race,%
White 95.2 92.1 (cid:2).0001 95.2 90.8 (cid:2).0001
Black 2.4 5.4 (cid:2).0001 2.4 6.6 (cid:2).0001
Hispanic 0.6 0.5 .541 0.6 0.6 .4167
Other 1.9 2.0 (cid:2).0001 1.9 2.0 .223
Comorbidities,%
Coronary 51.3 38.9 (cid:2).0001 57.0 42.3 (cid:2).0001
Valvular 8.7 9.9 (cid:2).0001 9.3 10.7 (cid:2).0001
Arrhythmia 27.8 22.5 (cid:2).0001 26.6 20.6 (cid:2).0001
CHF 15.8 17.6 (cid:2).0001 15.0 16.1 .0017
Pulmonary 36.3 41.6 (cid:2).0001 35.9 41.8 (cid:2).0001
Cerebrovascular 7.6 8.2 (cid:2).0001 7.3 6.8 .0518
Neurologic 4.2 4.0 .01 4.9 5.0 .6815
Renalfailure 4.8 4.5 .0002 6.3 6.1 .3921
PAD 20.6 27.9 (cid:2).0001 21.6 27.9 (cid:2).0001
Diabetes 12.8 11.0 (cid:2).0001 16.9 14.9 (cid:2).0001
Liverdisease 1.2 1.2 .3769 1.1 1.1 .8037
Cancer 7.4 4.1 (cid:2).0001 8.2 4.8 (cid:2).0001
Hypertension 63.4 70.1 (cid:2).0001 72.6 78.2 (cid:2).0001
Hyperlipidemia 28.0 27.6 .0385 42.5 41.3 .009
Procedure,%
EVAR 25.4 18.8 (cid:2).0001 100.0 100.0
Conversion 2.0 4.6 (cid:2).0001
OAR 74.6 81.2 (cid:2).0001
CHF,Congestiveheartfailure;PAD,peripheralarterialdisease.
aReflectingtheintroductionoftheEVARInternationalClassificationofDisease,9thedition,procedurecodein2000,EVARprocedureswereonlyidentified
duringthelast6yearsofthedataset.
whethertheinitialsurvivaladvantageassociatedwithmale hospitalizationswereusedtoassessbaselinecomorbidities.
genderafterEVARorOARpersistsovertimeneedsfurther The comorbidities included all those identified in prior
investigation. This study addressed these issues from a hospitalizationsandthechronicconditionsreportedatthe
populationperspectiveanddeterminedtheeffectofgender index hospitalization. The comorbidities that were ana-
on long-term survival differences after OAR and EVAR lyzedarereportedinAppendixTableO1(onlineonly).
underelectiveandurgentcircumstances. Thefollowingperioperativeoutcomeswerereviewed:
cardiac,respiratory,andacuterenalfailurerequiringdialy-
METHODS
sis, patient disposition, and 30-day mortality. Long-term
Data sources and population selection. The data adverse events included ruptured AAA, reoperation of
source for analysis was the Medicare Inpatient Standard AAA, arterial reinterventions (thrombectomy, embolec-
Analytical and Denominator files for years 1995-2006. tomy,catheter-based,andopenreconstructivereinterven-
Patientswereselectedthroughacombinationofdiagnoses tionsforreadmissionswithaprimaryorsecondarydiagnosis
codes(InternationalClassificationofDisease,9thClinical of AAA) and interventions for incision- and laparotomy-
Modification [ICD-9-CM] codes 441.4—AAA, without related complications, including intestinal postoperative
mentionofrupture,or441.3—rupturedAAA)andICD-9 obstruction,postoperativeadhesionsandincisional,groin
procedure codes (39.71 for EVAR and 38.44, 39.25, or abdominal wall hernia repairs (Appendix Table O1,
39.52,38.34,38.64,38.40,38.60forOAR)intheprimary onlineonly).
or any secondary position. The study excluded patients Readmissionrateswerecalculatedper1yearofsurvival.
withthoracicorthoracoabdominalaneurysms. Adischargeandreadmissionthatoccurredonthesameday
The hospitalization with the first AAA procedure was was considered one episode of care. Annual hospital and
identifiedastheindexhospitalization.Preindexandindex surgeonvolumeofAAAoperationsweredefinedasnumber
JOURNALOFVASCULARSURGERY
Volume54,Number1 Egorovaetal 3
TableI. Continued
OAR EVARvsOAR
Men Women Women Men
(n(cid:1)186,726) (n(cid:1)58,810) P P P
74.56337 75.49 (cid:2).0001 (cid:2).0001 (cid:2).0001
24.2 19.5 (cid:2).0001 (cid:2).0001 (cid:2).0001
31.2 29.3 (cid:2).0001 (cid:2).0001 (cid:2).0001
26.9 28.8 (cid:2).0001 .1557 .0074
13.6 16.4 (cid:2).0001 (cid:2).0001 (cid:2).0001
4.0 5.9 (cid:2).0001 (cid:2).0001 (cid:2).0001
95.2 92.4 (cid:2).0001 (cid:2).0001 0.4911
2.4 5.1 (cid:2).0001 (cid:2).0001 0.5267
0.6 0.5 .2929 .1627 0.9368
1.9 2.0 .0217 .8891 0.675
49.4 38.1 (cid:2).0001 (cid:2).0001 (cid:2).0001
8.5 9.7 (cid:2).0001 .0006 (cid:2).0001
28.3 22.9 (cid:2).0001 (cid:2).0001 (cid:2).0001
16.0 17.9 (cid:2).0001 (cid:2).0001 (cid:2).0001
36.5 41.6 (cid:2).0001 .5936 .0106
7.7 8.6 (cid:2).0001 (cid:2).0001 (cid:2).0001
4.0 3.8 .0199 (cid:2).0001 (cid:2).0001
4.3 4.1 .0331 (cid:2).0001 (cid:2).0001
20.3 27.9 (cid:2).0001 .9015 (cid:2).0001
11.3 10.1 (cid:2).0001 (cid:2).0001 (cid:2).0001
1.3 1.2 .1701 .3094 (cid:2).0001
7.1 3.9 (cid:2).0001 (cid:2).0001 (cid:2).0001
60.3 68.2 (cid:2).0001 (cid:2).0001 (cid:2).0001
23.1 24.5 (cid:2).0001 (cid:2).0001 (cid:2).0001
ofEVAR,OAR,orrupturedAAA(rAAA)repairs(EVAR
andOAR)peryear.5Theanalyseswerestratifiedasfollows:
(1)women:OARvsEVAR,(2)men:OARvsEVAR,(3)
EVAR:menvswomen,and(4)OAR:menvswomen. TableII. Five-yearrelativesurvivalestimatesformen
Statistics. Tocontrolforlifeexpectanciesinthegen- andwomenafteropen(OAR)andendovascular(EVAR)
eralpopulation,weusedrelativesurvival.Relativesurvival repairofabdominalaorticaneurysm(AAA)
wascalculatedusingthelife-tablemethod14bycomparing
the observed survival after AAA repair (including and ex- Men Women
cluding operation-related mortality (cid:2)90 days)15,16 with Variable %(95%CL) %(95%CL) P
theexpectedsurvivalofthepopulation17adjustedforback-
ElectiveAAA
grounddeathbasedonage,sex,andcalendaryear. OAR 86(86,87) 77(76,77) (cid:2).0001
Because each treatment group was vastly different, to EVAR 87(87,88) 80(78,81) (cid:2).0001
control for baseline characteristics and assess solely the Excluding90-daymortality
OAR 93(92,93) 84(84,85) (cid:2).0001
effect of procedure or gender on survival, we used a
EVAR 91(90,91) 85(83,86) (cid:2).0001
propensity-matchingtechnique.18-22Toincreasethepool
RupturedAAA
ofdataformatching,all1995to2006hospitalizationsfor OAR 34(33,35) 19(18,21) (cid:2).0001
OAR and hospitalizations from 2000 through 2006 for EVAR 43(39,48) 32(25,39) (cid:2).0001
EVARwereincluded.Todeterminethepropensityscore,a Excluding90-daymortality
logisticregressionmodelwasdevelopedwherethedepen- OAR 83(82,84) 72(70,74) (cid:2).0001
EVAR 75(67,82) 65(51,78) (cid:2).0001
dentvariablewasthetypeofprocedureorgender.Allbaseline
confounders,includinggender,race,comorbidities(ascate- CL,Confidencelimit.
JOURNALOFVASCULARSURGERY
4 Egorovaetal July2011
TableIII. Perioperativeandlong-termcomplicationsforwomenandmenafterelectiveendovascular(EVAR)oropen
(OAR)repairofabdominalaorticaneurysm
Women Men
EVAR,% OAR,% EVAR,% OAR,%
Variable (n(cid:1)9080) (n(cid:1)9080) P (n(cid:1)33,240) (n(cid:1)33,240) P
Disposition
Home/homecare 82.7a 64.8 (cid:2).0001 92.3a 77.6 (cid:2).0001
SNF/rehabilitation 12.3 26.2a (cid:2).0001 5.1 15.5a (cid:2).0001
Perioperativecomplications
Cardiac 4.45 8.96a (cid:2).0001 3.13 8.67a (cid:2).0001
Pulmonary 10.76 30.12a (cid:2).0001 7.01 26.14a (cid:2).0001
Renalfailurew/dialysis 0.54 1.30a (cid:2).0001 0.36 1.15a (cid:2).0001
30-daymortality 3.19 6.37a (cid:2).0001 1.84 4.80a (cid:2).0001
Long-termadverseevents/reinterventions
Readmissionrate 0.75a 0.69 (cid:2).0001 0.63a 0.57 (cid:2).0001
RupturedAAA 0.29a 0.12 .0001 0.29a 0.10 (cid:2).0001
AAArevision 1.81a 0.53 (cid:2).0001 1.94a 0.71 (cid:2).0001
Open 0.64a 0.43 (cid:2).0001 0.79a 0.52 (cid:2).0001
Endovascular 1.09a 0.10 (cid:2).0001 1.11a 0.19 (cid:2).0001
Arterialreinterventions 1.77a 0.50 (cid:2).0001 1.72a 0.63 (cid:2).0001
Otherreinterventionsb 2.28 4.29a (cid:2).0001 2.32 4.69a (cid:2).0001
SNF,Skillednursingfacility.
aRepresentsasignificantlyhigherrate.
bReinterventionsforincisionorlaparotomy-relatedcomplications,includingintestinalpostoperativeobstruction,postoperativeadhesionsandincisional,groin
orabdominalwallherniarepairs.
goric variables), age, year of surgery, hospital, and surgeon
volume(ascontinuousvariables)wereincludedinthemodel
as independent variables. The patients were matched by
greedy algorithm using the 8- to 1-digit matching scheme
withoutreplacement.18Pairedttestswereusedforcontinu-
ous variables and McNemar tests for categoric variables to
ensure that there were no significant differences in demo-
graphics and comorbid characteristics between matched
groups.Theresultsofmatchingbypropensityscorearesum-
marizedinAppendixTablesO2andO3(onlineonly).Peri- Fig 1. Long-term survival of (A) women and (B) men treated
operativecomplicationsand30-daymortalitywerecompared with elective endovascular (EVAR) and open (OAR) repair of
inmatchedgroupsusingtheMcNemartest. aorticabdominalaneurysm(AAA).Caseswerematchedbypro-
SurvivalcurveswereconstructedwithCoxmodels.Frailty pensityscore.Thehazardratio(HR)isthequotientofthehazard
ratesoftheEVARandOARcomparisongroupsandisexpressedas
modelswereusedtocontrolformatching(clustering).Haz-
apointestimatewitha95%confidenceinterval.
ardratio(HR)oflong-termsurvivalwasestimatedusingthe
SAS PROC PHREG procedure with the robust sandwich
estimate option.23,24 Propensity score and type of repair or
genderwereincludedinthisregressionanalysis.Weusedthe TableI.The30-daymortalityratesforwomenwere6.43%
Martingale methods to check the proportional hazard as- afterOARand2.96%afterEVARandformenwere4.64%
sumption. Readmission rates were compared with Poisson and 1.63%, respectively. After adjusting for life expecta-
regression analysis. Statistical significance was expressed as ncyandcalendaryear,the5-yearrelativesurvivalforboth
both P values and 95% confidence intervals. P (cid:2) .05 was men and women was less than in the general population
considered statistically significant. P values were reported andwassignificantlylowerforwomenthanmenaftereither
without adjusting for multiple comparisons. SAS 9.13 soft- procedure(TableII).Theworst5-yearrelativesurvivalwas
ware(SASInstituteInc,Cary,NC)wasused. observedinwomenundergoingOAR(84%).
Women: OAR vs EVAR. Women who underwent
RESULTS
EVARdifferedfromthosewhounderwentOAR:theywere
Elective AAA repair. There were 322,892 patients older, had a higher frequency of coronary artery disease
with elective AAA repair from 1995 through 2006, and (CAD) and diabetes, and a lower frequency of cardiac
245,536underwentOAR(70.5%men,29.5%women)and arrhythmiasandcerebrovasculardisease(TableI).Adjust-
77,356 underwent EVAR (82.4% men, 17.6% women). mentforthesedifferenceswasaccomplishedbypropensity
Baselinecharacteristicsofmenandwomenarereportedin scorematching,whichresultedin9080pairedwomenwho
JOURNALOFVASCULARSURGERY
Volume54,Number1 Egorovaetal 5
TableIV. Disposition,perioperative,andlong-termcomplicationsafterelectiveendovascular(EVAR)oropenrepair
(OAR)ofabdominalaorticaneurysmbygender
EVAR OAR
Men,% Women,% Men,% Women,%
Variable (n(cid:1)13,710) (n(cid:1)13,710) P (n(cid:1)59,279) (n(cid:1)59,279) P
Disposition
Home/homecare 91.6a 83.7 (cid:2).0001 80.0a 70.5 (cid:2).0001
SNF/rehabilitation 6.1 12.0a (cid:2).0001 11.1 17.7a (cid:2).0001
Perioperativecomplications
Cardiac 2.67 4.13a (cid:2).0001 8.41 9.26a (cid:2).0001
Pulmonary 6.23 9.85a (cid:2).0001 24.47 26.78a (cid:2).0001
Renalfailurew/dialysis 0.39 0.50 .205 0.96 1.27a (cid:2).0001
30-daymortality 1.77 2.96a (cid:2).0001 4.98 6.43a (cid:2).0001
Long-termadverseevents/reinterventionsa
Readmissionrate 0.66 0.77a (cid:2).0001 0.57 0.66a (cid:2).0001
RupturedAAA 0.33 0.28 .99 0.57 0.66 .1616
AAArevision 2.04 1.91 .1168 0.63a 0.48 .0049
Open 0.79 0.63 .6306 0.52 0.43 .3077
Endovascular 1.20 1.21 .0444 0.11a 0.05 (cid:2).0001
Arterialreinterventions 2.01 1.81 .3489 0.60 0.49 .1766
Otherreinterventionsb 2.23 2.08 .1125 4.40a 3.96 .0325
SNF,Skillednursingfacility.
aRepresentsasignificantlyhigherrate.
bReinterventionforincisionalorlaparotomy-relatedcomplications,includingintestinalpostoperativeobstruction,postoperativeadhesionsandincisional,
groin,orabdominalwallherniarepairs.
were balanced in baseline characteristics (Appendix Table
O2,onlineonly).
The 30-day mortality in these matched groups was
6.37% after OAR and 3.19% after EVAR (P (cid:2) .0001).
Higher rates of complications were observed in the OAR
group (Table III). EVAR recipients had a significant sur-
vivalbenefitthatwassustainedforthe6yearsofobserva-
tion(Fig1,A).Therelativeriskofdyingwas12%lessinthe
EVAR than in the OAR group (HR, 0.88; P (cid:2) .0001).
However, EVAR recipients had higher rates of AAA and
Fig 2. Long-termsurvivalafterelective(A)open(OAR)and(B)
arterial reinterventions, and OAR recipients had a higher endovascular(EVAR)repairofmenandwomenwithaorticab-
rate of laparotomy or incision-related reinterventions dominal aneurysms (AAAs). Cases were matched by propensity
(TableIII). score.Thehazardratio(HR)isthequotientofthehazardratesof
Men:OARvsEVAR. Thebalancedsubgroupscon- thetwocomparisongroups(menandwomen)andisexpressedas
sisted of 33,240 pairs (mean age, 75.5). Comorbidities apointestimatewith95%confidenceinterval.
included CAD in 55%, congestive heart failure (CHF) in
16%, renal failure in 6%, and neurologic diseases in 4.6% available in Appendix Table O2 (online only). Mean age
(AppendixTableO2,onlineonly). was75.4;comorbiditieswereCADin38%,CHFin18%,
Perioperative mortality in these balanced groups was andrenalfailurein4%.
1.84%afterEVARand4.80%afterOAR(P(cid:2).0001;Table The30-daymortalityrateinthesematchedgroupswas
III).Similartowomen,ratesofperioperativecomplications 6.43%forwomenand4.98%formen(P(cid:2).0001).Women
werelowerintheEVARgroup. alsohadhigherratesofpostoperativecomplications(Table
Distinctfromwhatwasseenwithwomen,thesurvival IV). Men experienced superior survival throughout the
advantageforEVARwasgreatestduringthefirst2postop- 6-yearfollow-upperiod(HR,1.06;P(cid:2).0001;Fig2,A).
erativeyears,butthenthecurvesconverged(Fig1,B).The The survival curves remained parallel, indicating that the
relativeriskofdyingwasabout4%lowerforEVARrecipi- survivalsforthecomparisongroupsasawholewerediffer-
ents(HR,0.96;P(cid:1).0049).Likewomen,menundergoing ent,conditionalontheinitial2-yearsurvival.Womenex-
EVAR had higher readmission, AAA, and arterial reinter- perienced higher rates of readmission (0.66 vs 0.57 for
ventionrates,whereasOARrecipientshadhigherlaparotomy- men, P (cid:2) .0001), and slightly lower AAA revision rates
relatedreinterventions(TableIII). (TableIV).
Elective OAR: Men vs women. Demographics and ElectiveEVAR:Menvswomen. Forthisanalysiswe
comorbiditiesofthebalancedsubgroups(59,279pairs)are constructedsubgroupsof13,710patientswithamatched
JOURNALOFVASCULARSURGERY
6 Egorovaetal July2011
TableV. Baselinedemographicsandcomorbiditiesforpatientswhohadendovascular(EVAR)oropen(OAR)repair
ofrupturedabdominalaorticaneurysm
Overall EVAR
Men Women Men Women
Variable (n(cid:1)37,371) (n(cid:1)11,494) P (n(cid:1)1497) (n(cid:1)510) P
Age,years 75.9 78.2 (cid:2).0001 77.5 78.8 .0003
Agegroups,%
65-69 20.9 13.1 (cid:2).0001 15.8 13.5 .2245
70-74 26.9 21.0 (cid:2).0001 23.2 17.1 .0037
75-79 25.5 25.8 .5083 24.1 23.5 .7891
80-84 16.9 23.0 (cid:2).0001 22.4 26.7 .0484
(cid:1)85 9.8 17.1 (cid:2).0001 14.6 19.2 .0127
Race,%
White 94.5 90.5 (cid:2).0001 92.3 87.5 .0009
Black 3.1 6.9 (cid:2).0001 5.6 10.6 .0001
Hispanic 0.6 0.5 .525 0.7 0.6 .7312
Otherrace 1.9 2.1 .189 1.3 1.4 .9507
Comorbidities,%
Coronary 28.6 24.7 (cid:2).0001 37.6 30.6 .0043
Valvular 5.2 6.8 (cid:2).0001 8.6 8.0 .7196
Arrhythmia 30.5 27.0 (cid:2).0001 33.7 33.3 .8903
CHF 20.9 24.5 (cid:2).0001 22.8 26.5 .0905
Pulmonary 35.0 37.3 (cid:2).0001 39.4 38.0 .5831
Cerebrovascular 5.4 6.4 (cid:2).0001 6.9 7.6 .5957
Neurologic 5.6 6.7 (cid:2).0001 6.2 6.9 .6037
Renalfailure 6.3 6.7 .1409 11.8 11.8 .9962
PAD 11.3 11.8 .1342 13.7 15.7 .2656
Diabetes 8.5 9.6 .0001 10.8 11.8 .5576
Liverdisease 1.3 1.2 .5139 1.2 1.6 .5276
Cancer 4.8 3.1 (cid:2).0001 7.4 6.5 .4755
Hypertension 44.8 53.4 (cid:2).0001 59.3 64.1 .0555
Hyperlipidemia 9.9 11.2 (cid:2).0001 19.8 22.5 .1797
Procedure,%
EVAR 4.0 4.4 .0416 100.0 100.0
Conversion 6.9 7.3 .7743
OAR 96.0 95.6 .0416 0.0 0.0
CHF,Congestiveheartfailure;PAD,peripheralarterialdisease.
meanageof77.4yearsandmajorcomorbiditiesofCADin otherhand,menhadhigherratesofCAD,cardiacarrhyth-
42%,CHFin16%,diabetesin15%,andrenalfailurein6% mia,andcancer(TableV).Thus,weconstructedfoursets
(AppendixTableO2,onlineonly). of matched subgroups to compare survival after aortic
The 30-day mortality rate was 2.96% for women and ruptureformenandwomenundergoingEVARandOAR.
1.77% for men (P (cid:2) .0001), with women experiencing Detailsofpatients’baselinecharacteristicsarereportedin
higher rates of perioperative complications (Table IV). AppendixTableO3(onlineonly).
Interestingly, although overall survival was superior for Ruptured AAA women: OAR vs EVAR. Demo-
men(HR,1.07;P(cid:1).0136),theinitialsurvivaladvantage graphicsandcomorbiditiesof495matchedpairsofwomen
inmenlessenedovertime,andthesurvivalcurvesbeganto (EVARvsOAR)areavailableinAppendixTableO3(on-
convergeafter2years(Fig2,B). lineonly).Themeanagewas78years,and18%wereaged
Ruptured AAA. Among the 48,865 patients with (cid:1)85.CADwaspresentinapproximately30%,CHFin25%,
rAAA repairs, 76.5% were men and 23.5% were women andrenalfailurein12%.
(TableV),andonly2007underwentEVAR.Theratioof Perioperative mortality was significantly lower after
men to women by either approach was about 3:1 (OAR: EVAR(41.01%)comparedwithOAR(48.28%,P(cid:1).0201)
76.6% vs 23.4%; and EVAR: 74.6% vs 25.4%). Overall inthesematchedsubgroups.Ratesofpostoperativecardiac,
30-daymortality52.7%afterOARand41.05%afterEVAR pulmonarycomplications,andrenalfailurerequiringdial-
for women was and 44.06% after OAR and 33.07% after ysiswerehigherintheOARrecipients(TableVI).
EVARformen.Regardlessrepairtype,relativesurvivalfor Analysis of early data from 2000 to 2004 showed
women was inferior to survival for men, with the worst equivalentsurvivalforOARandEVARrecipientsevenafter
survival after EVAR (65%; Table II). Women were older, 1,6,and12monthsofrepair.However,whenweincluded
with a mean age of 78 vs 76 years (P (cid:2) .0001) and had morerecentyears(2005and2006),wediscoveredasur-
higherratesofvalvularandpulmonarydisease,CHF,renal vival benefit of EVAR over OAR that persisted 6 months
failure,diabetes,hypertension,andhyperlipidemia.Onthe aftersurgery(Fig3,A).Overallsurvivalratesweresimilar
JOURNALOFVASCULARSURGERY
Volume54,Number1 Egorovaetal 7
TableV. Continued
OAR EVARvsOAR
Men Women Women Men
(n(cid:1)35,874) (n(cid:1)10,984) P P P
75.9 78.2 (cid:2).0001 .0564 (cid:2).0001
21.1 13.1 (cid:2).0001 .7839 (cid:2).0001
27.1 21.2 (cid:2).0001 .0239 .0008
25.5 25.9 .4551 .2331 .2162
16.7 22.8 (cid:2).0001 .0412 (cid:2).0001
9.6 17.0 (cid:2).0001 .1914 (cid:2).0001
94.6 90.6 (cid:2).0001 .0167 .0002
2.9 6.7 (cid:2).0001 .0008 (cid:2).0001
0.6 0.5 .5623 .8319 .396
1.9 2.1 .1796 .2525 .1101
28.2 24.5 (cid:2).0001 .0017 (cid:2).0001
5.1 6.8 (cid:2).0001 .2755 (cid:2).0001
30.4 26.7 (cid:2).0001 .001 .0073
20.8 24.4 (cid:2).0001 .2942 .0608
34.8 37.2 (cid:2).0001 .7169 .0003
5.3 6.3 (cid:2).0001 .223 .0057
5.6 6.6 (cid:2).0001 .8478 .3278
6.1 6.5 .1515 (cid:2).0001 (cid:2).0001
11.2 11.7 .215 .006 .0033
8.4 9.5 .0002 .0919 .0008
1.3 1.2 .4195 .4103 .8375
4.7 3.0 (cid:2).0001 (cid:2).0001 (cid:2).0001
44.2 52.9 (cid:2).0001 (cid:2).0001 (cid:2).0001
9.5 10.7 (cid:2).0001 (cid:2).0001 (cid:2).0001
0.0 0.0 (cid:2).0001 (cid:2).0001
0.0 0.0
100.0 100.0 (cid:2).0001 (cid:2).0001
for EVAR and OAR recipients (HR, 0.94; P (cid:1) .41). 6.5%(AppendixTableO3,onlineonly).Themortalityrate
Readmission, AAA revision, and arterial reinterventions (cid:2)30daysafterOARforrAAAinthesebalancedgroupswas
ratesweresignificantlyhigherafterEVAR(TableVI). 52.61% for women and 46.57% for men (Table VII). No
RupturedAAAmen:OARvsEVAR. Wecompared differences in perioperative cardiac, pulmonary, or renal
OAR with EVAR outcomes in 1421 pairs of matched complications between men and women were recorded.
men with rAAA. The mean age was 77; 14% were aged Long-termsurvivalwassignificantlyhigherformen(HR,
(cid:1)85. Pulmonary disease was present in almost 40%, 1.17; P (cid:2) .0001; Fig 4. A), whereas women had higher
CAD in about 38%, and renal failure in 11% (Appendix readmissionandrAAArates(TableVII).
TableO3,onlineonly).Similartowomen,meninthese Ruptured AAA EVAR: men vs women. We com-
matchedgroupshadhigher30-daymortalityafterOAR paredtheoutcomesofmatchedpairsofmenandwomen
than after EVAR (43.70% vs 33.43% P (cid:2) .0001) and a afterEVARwith499individualsineachgroup.Thedistri-
higher rate of perioperative pulmonary complications butionofdemographiccharacteristicswassimilarbetween
(Table VI). Distinct from what was observed with genders(AppendixTableO3,onlineonly).Theirmeanage
women, overall survival was better for men with rAAA was79years,and46%ofmenandwomenwereaged(cid:1)80.
after EVAR than after OAR (Fig 3, B). The overall Therateofunderlyingpulmonarydiseasewas38%,CAD
relativeriskofdyingwas15%higherintheOARgroup wasabout32%,CHFwasabout26%,andrenalfailurewas
(HR,0.85;P(cid:1).0008),withlowerreadmissionandAAA about10%.
reinterventionrates(TableVI). The perioperative mortality rate was about 10%
RupturedAAAOAR:menvswomen. Wecompared higherforwomen(41.08%)thanformen(30.66%;P(cid:2)
the outcomes of men and women after OAR in two .0001), and women had substantially lower long-term
matched subgroups, each with 10,911 individuals. They survival (Fig 4, B), with a 33% higher relative risk of
wereameanageof78years,andCADwaspresentin24%, dying (HR, 1.33; P (cid:1) .0005) and higher readmission
CHFin24%,pulmonarydiseasein37%,andrenalfailurein rates(TableVII).
JOURNALOFVASCULARSURGERY
8 Egorovaetal July2011
TableVI. Perioperativeandlong-termcomplicationsforwomenandmenafterendovascular(EVAR)oropen(OAR)
repairofrupturedabdominalaorticaneurysm(AAA)
Women Men
EVAR,% OAR,% EVAR,% OAR,%
Variable (n(cid:1)495) (n(cid:1)495) P (n(cid:1)1421) (n(cid:1)1421) P
Disposition
Home/homecare 27.7a 20.0 .0061 38.6a 26.7 (cid:2).0001
SNF/rehabilitation 25.9 29.5 .2053 22.9 25.0 .1883
Perioperativecomplications
Cardiac 16.16 21.41a .0326 17.80 19.92 .1499
Pulmonary 44.65 54.55a .0017 43.07 53.48a (cid:2).0001
Renalfailurew/dialysis 3.23 7.07a .0078 3.66 5.00 .0739
30-daymortality 41.01 48.28a .0201 33.43 43.70a (cid:2).0001
Long-termcomplications/reinterventions
Readmissionrate 1.02a 0.79 .0002 0.92a 0.76 (cid:2).0001
RupturedAAA 1.72 3.48 .1451 2.45 2.81 .8465
AAArevision 2.19a 0.61 .0226 2.35a 0.79 .0001
Open 0.94 0.61 .4089 0.82 0.72 .4488
Endovascular 1.25a 0.0 .0081 1.53a 0.07 .0013
Arterialreinterventions 2.19a 0.20 0.0134 1.79a 0.42 .0003
Otherreinterventionsb 2.97 2.46 .3350 2.81 3.67 .6273
SNF,Skillednursingfacility.
aRepresentsasignificantlyhigherrate.
bReinterventionforincisionalorlaparotomy-relatedcomplications,includingintestinalpostoperativeobstruction,postoperativeadhesionsandincisional,
groin,orabdominalwallherniarepairs.
pared with men and that it occurs in smaller sized aneu-
rysms.32,33
Another concern is that women have higher rates of
undiagnosed comorbid conditions that increase operative
risk,ashasbeendocumentedforcardiovasculardisease.34
Evenwhendiagnosed,treatmenttendstobeunderutilized:
women with CAD are less likely to receive statins and
(cid:3)-blockers,35,36andthosewithcarotidstenosisarelesslikely
toreceiveantiplatelettreatment.37Strongemergingevidence
Fig 3. Long-term survival of (A) women and (B) men treated fromclinicaltrials(DutchEchographicCardiacRiskEvalua-
with endovascular (EVAR) and open (OAR) repair of ruptured tion Applying Stress Echo II and IV [DECREASE II and
aortic abdominal aneurysm. Cases were matched by propensity IV]) suggests that lack of medical optimization before
score.Thehazardratio(HR)isthequotientofthehazardrateof
procedures has a major adverse effect on perioperative
theEVARandOARcomparisongroupsandisexpressedasapoint
cardiovascularmortalityandmorbidity.38,39
estimateand95%confidenceinterval.
Themorechallenginganatomy,higherchancesofun-
diagnosedriskfactors,andthehigherlikelihoodofinade-
DISCUSSION quatecardiovascularmedicaloptimization,maypartlyex-
plain why women with AAA have worse outcomes than
Therearesignificantanatomicandbiologicdifferences
men. Indeed, the perioperative mortality rates in women
betweenmenandwomenthatchallengethemanagement
afterelectiverepairwasreportedashighas10.6%in1980
of aortic aneurysmal disease in women. The abdominal
aortaandiliacarteriesinwomenare,onaverage,smaller;25-27 through1990intheMichiganregistry,3.8%higherthan
women tend to have a shorter infrarenal aortic neck,28,29 menduringthesameperiod.10Morerecentreports,which
andtheiraneurysmmorefrequentlyinvolvesthejuxtarenal reflecttheintroductionofEVARandimprovedperiopera-
and suprarenal aorta.30 Moreover, women with AAA are tive management, continue to demonstrate this gender
older than men by at least 4 to 5 years7 and have more disparity, with a mortality rate of 2.6% to 3.2% in men vs
cerebrovascular31andiliacocclusivedisease.7,12,13,28 4.8%to5.45%inwomen.9,40
ThenaturalhistoryofAAAdiseaseinwomenisnotas WithrAAA,thedifferencesinmortalityaremorepro-
well defined because of worse representation in clinical nounced.40,41IntheMichiganregistry,themortalityrate
trials.However,thereisanindicationthatAAAmightbe for women was 60.5%, 12% higher than for men. More
more dangerous in women. Data derived from United recently, these rates have dropped in both groups, but a
KingdomSmallAneurysmTrial(UKSAT)indicatethatthe significant gender disparity persists (52.9% in women vs
rupturerateisatleastthreetimeshigherinwomencom- 43%inmen).9,40,41
JOURNALOFVASCULARSURGERY
Volume54,Number1 Egorovaetal 9
TableVII. Perioperativeandlong-termcomplicationsforwomenandmenafterendovascular(EVAR)oropen(OAR)
repairofrupturedabdominalaorticaneurysm(AAA)
EVAR OAR
Men,% Women,% Men,% Women,%
Variable (n(cid:1)499) (n(cid:1)499) P (n(cid:1)10,911) (n(cid:1)10,911) P
Disposition
Home/homecare 38.3a 27.9 .0005 28.2a 19.6 (cid:2).0001
SNF/rehabilitation 25.9 25.9 1 17.6 20.1a (cid:2).0001
Perioperativecomplications
Cardiac 18.44 16.23 .3576 19.78 19.79 .9864
Pulmonary 39.68 43.69 .2059 49.95 48.93 .1319
Renalfailurew/dialysis 3.41 3.41 1 5.13 4.83 .3041
30-daymortality 30.66 41.08a .0005 46.57 52.61a (cid:2).0001
Long-termcomplications/reinterventions
Readmissionrate 0.88 1.04a .0052 0.66 0.80a (cid:2).0001
RupturedAAA 2.70 1.76 .1721 2.34 2.43a .0104
AAArevision 1.67 2.08 .6938 0.81a 0.45 .0023
Open 0.39 0.80 .3615 0.74a 0.41 .0051
Endovascular 1.29 1.28 .8779 0.07 0.02 .0688
Arterialreinterventions 1.71 2.13 .3808 0.47 0.32 .1989
Otherreinterventionsb 3.22 3.04 .6752 4.70 4.06 .7936
SNF,Skillednursingfacility.
aRepresentsasignificantlyhigherrate.
bReinterventionforincisionalorlaparotomy-relatedcomplications,includingintestinalpostoperativeobstruction,postoperativeadhesionsandincisional,
groin,orabdominalwallherniarepairs.
Thisspeculationobviouslyneedsfurtherinvestigation;
however,itledustoconstructmatchingpairsadjustedfor
known comorbidities to evaluate the effect of gender in
long-term outcomes. We observed a survival benefit for
EVARthatpersistedfortheentire6yearsoffollow-upfor
womenwhounderwentelectiverepair(Fig1).Thispattern
is distinct from what had been seen for EVAR use in a
mixed cohort of men and women in clinical trials. It is
possible that long-term survival benefit after EVAR for
Fig 4. Long-termsurvivalafter(A)open(OAR)and(B)endo- women was not seen in clinical trials due to the under-
vascular(EVAR)repairofrupturedaorticabdominalaneurysmin representationofwomen.
menandwomen.Caseswerematchedbypropensityscore.The Wealsoobservedsuperiorlong-termsurvivalwithelec-
hazardratio(HR)isthequotientofthehazardratesofthetwo tive EVAR in men compared with women. However, the
comparisongroups(menandwomen)andisexpressedasapoint survivaldifferencewasgreatestinthefirst2postoperative
estimateand95%confidenceinterval.
yearsanddisappearsafterward.Thisfindingregardingthe
long-termequalizationofsurvivalcurvesafterAAArepairis
TheintroductionofEVARhasreducedtherateofearly
reported here for the first time. It may reflect anatomic
mortality with elective AAA repair.6,7,9,40,41 EVAR has
differencesbetweengendersthatfavorwomeninthelong-
been frequently used for patients with multiple systemic
term. Women with AAA are less likely than men to have
diseasesanddecreasedlifeexpectancywhowouldnottol-
dilated iliac arteries.13,28 Iliac artery dilation is often re-
eratethemajorstressfromOAR.Indeed,the5-yearrelative
sponsible for late reintervention and poor long-term out-
survival for EVAR recipients compared with OAR recipi-
entswasuniformlyworse.Eventhoughwomengenerally comeafterprimaryintervention.Theaorticwallinwomen
havelongerlifeexpectancythanmen,theobserved5-year has different compliance than in men25,42 and probably
relativesurvivalafterAAArepairwaslowerforwomenthan responds differently to exclusion after insertion of a stent
men(TableII),whichwasconsistentwithpreviousobser- graft. In fact, Ouriel et al12 have reported more rapid
vations.11WomenwhoreceivedEVARdidtheworst:85% shrinkageoftheaorticsacinwomenundergoingEVAR.
afterelectiveandonly65%afterrAAAEVARreachedthe For open procedures, both eAAA and rAAA, and ur-
expectedsurvivalfortheiragegroup.Thissuggeststhatthe gent EVAR, the survival benefit for men persisted. The
EVARgroupconsistsofpatientswithhigherratesofsys- recovery after OAR and rAAA repair appears to be more
temicdiseases,suchascancer,cardiac,andchronicpulmo- complicated for women. They experience higher rates of
nary diseases than in the general population, and these postoperative complications and are less likely to be dis-
factorsareresponsibleforlowersurvival. chargedhome.Suchhighrateswerenotseenforwomen
JOURNALOFVASCULARSURGERY
10 Egorovaetal July2011
after elective EVAR, reflecting that elective EVAR is less world”medicinebecausetheyarederivedfromalltypesof
invasiveandlessdisruptivetobodyfunction.Thefavorable medicalcenters.
anatomic factors for women may play a lesser role in the
case of elective open and urgent repair of rAAA. Women CONCLUSIONS
withAAAareelderlyandmorelikelythanmentolivealone,
Our study is unique in identifying gender-related dif-
havesickerandolderspousesincapableofcaringforthem,
ferences not only postoperatively but long-term after dif-
or have lower income. For all these reasons, women after
ferenttypesofAAArepair.Weobservedsustainedgender-
OAR or rAAA repair are less likely to go home.9 Institu-
associated disparities in mortality favoring men after all
tionalization and poor functional status after intervention
types of repairs in elective and urgent clinical settings.
mayhaveaneffectonlong-termsurvival.Furtheranalysis
However,itisencouragingthatEVARhasdecreasedlong-
ofthecausesofdeathandreasonsforreadmissionswillbe
termmortalityinwomenafteraorticaneurysmrepairand
veryimportanttounderstandthesefindings.
that the women’s survival begins to equal men’s after 2
Analysis of the long-term effect of gender on survival years.Moreover,EVARhasbeguntohaveapositiveeffect
after urgent repair alone uncovered additional findings of in the urgent clinical setting of acute rupture with a pro-
interest.Initialdataanalysisfrom2000to2004showedno longedbenefitfromthistypeofintervention.Asexperience
benefit in survival of EVAR over OAR in women whose withendovascularinterventionsgrowsanddevicesimprove
aneurysmhadruptured.However,whenweexpandedthe tobetterfitthefemaleanatomy,webelievethatthehistoric
analysisandincludeddatafrom2005and2006,wenoticed disparities in aortic outcomes between men and women
a 6-month advantage in survival in women undergoing maysoondisappear.
urgentEVAR.Itislikelythatthisbenefitwilllastlongerin
thefutureassmaller-profiledevicesareusedandoperators AUTHORCONTRIBUTIONS
becomemoreexperiencedinselectingtherightdevicesfor
Analysis and interpretation: NE, AV, AM, PF, AG,
thefemaleanatomyinthefaceoftheurgentnatureofaortic
CK,GM
rupture.
Datacollection:NE
One limitation of our study is that it is observational
Writingthearticle:NE,AV
and carries all the potential biases inherent to such study
Critical revision of the article: NE, AV, AM, PF, AG,
designs. For instance, the subgroups of men and women
CK,GM
differed substantially (Tables I and IV). We adjusted our
Final approval of the article: NE, AV, AM, PF, AG,
comparisonsforthesedifferencesbycomparingpropensity
CK,GM
score-matched cohorts; however, there may still be some
Statisticalanalysis:NE
bias due to confounders that were not recognized. In
Obtainedfunding:AG,CK
addition,menandwomenwerematchedaccordingtoage
Overallresponsibility:AG
and comorbidities and therefore some “healthier” men
were excluded from comparisons. Thus, the men in the
matchedgroupmaynolongerberepresentativeofallmen REFERENCES
undergoingaorticprocedures.Nevertheless,wethinkthat 1. KentKC.Endovascularaneurysmrepair--isitdurable?NEnglJMed
the matched comparisons are necessary to determine the 2010;362:1930-1.
effect of gender alone, without the distorting effects of 2. GreenhalghRM,BrownLC,PowellJT,ThompsonSG,EpsteinD,
Sculpher MJ. Endovascular versus open repair of abdominal aortic
otherfactorsthatcouldalterlong-termsurvival.
aneurysm.NEnglJMed2010;362:1863-71.
Anotherlimitationisthatthedatawereoriginallycol- 3. DeBruinJL,BaasAF,ButhJ,PrinssenM,VerhoevenEL,CuypersPW,
lectedforadministrativepurposesandaresubjecttocoding etal.Long-termoutcomeofopenorendovascularrepairofabdominal
errors. In addition, it is often difficult to distinguish aorticaneurysm.NEnglJMed2010;362:1881-9.
4. SchermerhornML,O’MalleyAJ,JhaveriA,CotterillP,PomposelliF,
whetherasecondinterventionisrelatedtotheinitialaneu-
LandonBE.Endovascularvs.openrepairofabdominalaorticaneu-
rysmrepair.Forinstance,weassumedthatallpostoperative rysmsintheMedicarepopulation.NEnglJMed2008;358:464-74.
arterial reconstructions were related to the initial AAA 5. EgorovaN,GiacovelliJ,GrecoG,GelijnsA,KentCK,McKinseyJF.
repair, although they might have been performed for an National outcomes for the treatment of ruptured abdominal aortic
aneurysm:comparisonofopenversusendovascularrepairs.JVascSurg
entirelydifferentcondition.
2008;48:1092-100:100:e1-2.
Finally,theICD9-CMcodesusedforEVARhavemore
6. McPheeJT,HillJS,EslamiMH.Theimpactofgenderonpresentation,
generalusethanthoseusedforOAR.Therefore,readmis- therapy, and mortality of abdominal aortic aneurysm in the United
sionandreinterventionrateswereprobablyoverestimated, States,2001-2004.JVascSurg2007;45:891-9.
7. VouyoukaAG,KentKC.Arterialvasculardiseaseinwomen.JVascSurg
and it is quite possible our analysis carries a bias against
2007;46:1295-302.
EVAR.Giventhattheselimitationsaffectmenandwomen
8. Johnston KW. Influence of sex on the results of abdominal aortic
equally,theobserveddifferencesinreinterventionandre- aneurysm repair. Canadian Society for Vascular Surgery Aneurysm
admissionratesshouldbeaccurateandthereportedgender StudyGroup.JVascSurg1994;20:914-23;discussion:923-6.
9. DillavouED,MulukSC,MakarounMS.Adecadeofchangeinabdom-
effectshouldbereal.AdistinctbenefitofusingMedicare
inalaorticaneurysmrepairintheUnitedStates:Haveweimproved
datasetsisthatthereisamplerepresentationofwomento
outcomes equally between men and women? J Vasc Surg 2006;43:
address issues of gender and that data sets reflect “real 230-8;discussion:238.