common bile duct injury

7/25/2019 Common Bile Duct Injury

1/6

Common Bile Duct InjuryDuring Laparoscopic Cholecystectomyand the Use of Intraoperative Cholangiography

Adverse Outcome or Preventable Error?

David R. Flum, MD; Thomas Koepsell, MD; Patrick Heagerty, PhD; Mika Sinanan, MD; E. Patchen Dellinger, MD

Background: Common bile duct (CBD) injury is a se-rious complicationof laparoscopic cholecystectomy (LC).Predictors of this adverse outcome have not been welldocumented.

Hypothesis:Surgeon experience and the use of intra-operative cholangiography (IOC) are associated with a

decreased rate of major CBD injury during LC.

Design: A retrospective population-based cohort study.

Setting:Washington State hospital discharge databasereports from 1991 through 1998.

Patients: Discharge reports were searched forInterna-tional Classification of Diseases, Ninth Revision, proce-dure codes consistent with LC andthen evaluatedfor pro-cedure codes for CBD repair and reconstruction within90 days of LC.

Main Outcome Measure:The rate of CBD injury in

patients undergoing LC based on the surgeons experi-ence and IOC use.

Results: In all, 30630 LCs and 76 major CBD injuries(2.5/1000 operations) wereidentified in thisanalysis.Therewere no significant differences between injured and non-injured patients in demographics, disease, payer status, or

hospital variables. A CBD injury occurred in 3.2 of 1000LCs in the early case order of surgeons compared with 1.7per 1000 at later points (P=.01) (relative risk, 1.81; 95%confidence interval, 1.44-2.88). The rate of injury in LCsperformed without IOC was 3.3 per 1000 compared with2.0 per 1000 in LCs with IOC (P=.02) (relative risk, 1.7;95%confidence interval,1.1-2.6). Surgeons experience andIOC use were independent predictors of injury.

Conclusions: Therate ofCBD injuryis significantly lowerwhen IOCis used. This effect is magnified during theearlyexperience of surgeons. Systematic use of IOC may sig-nificantly reduce the rate of CBD injury.

Arch Surg. 2001;136:1287-1292

THE WIDESPREADacceptanceof laparoscopy as the pre-ferred approach to chol-ecystectomy was based onanticipated reductions in

postoperative pain and recuperative timeassociated with minimal access.1 Soon af-ter itsintroduction, however, it becameclearthatlaparoscopic cholecystectomy(LC)wasassociated with uniquecomplications com-pared with the open approach. Early re-

portedrates of commonbile duct (CBD) in-jury were2 to 15 times2,3 greater thanthoseidentified in historic series. The incidenceof CBD injury appeared to be highest dur-ing the introduction of LC and was relatedto early experience with LC, the so-calledlearning curve.1 As the procedure has be-come increasingly common, surgeons havetendedto cite a rateofinjury of1:300.How-ever, despite the broad experience of sur-geons with LC during thepast decade, rates

of injury as high as 1.4%4,5 have recentlybeen reported.

The impact of a major CBD injury isstaggering to both the patient and thehealth care system. Major CBD injury isassociated with 11% case fatality6 and al-most always requires a technically de-manding, expensive operative reconstruc-tion of the biliary tree.7 Depending on theexpertiseof the surgeon reconstructing thebile ducts,8 reoperations may be com-

mon. For example, of 57 injuries result-ing in litigation, an average of 2 proce-dures (range, 1-8) wererequired to defini-tively repair the ducts.9 Itisthethirdmostcommonlylitigatedgeneral surgical com-plication, with median jury awards of$500000.6,10 Despiteitsrelativeinfrequency,CBDinjuryisbyfarthemostimportantpos-sible complication of LC.

Cholecystectomy requires the liga-tion and division of the cystic duct at a

ORIGINAL ARTICLE

From the Department ofSurgery, University ofWashington, Seattle.

(REPRINTED) ARCH SURG/VOL 136, NOV 2001 WWW.ARCHSURG.COM1287

2001 American Medical Association. All rights reserved.

wnloaded From: http://archsurg.jamanetwork.com/ by a University of North Dakota User on 06/06/2015


2/6

point separate from its junction with the CBD.11 Injuryto the CBD is most often the result of misidentificationof the cystic duct related to the CBD. Intraoperative chol-angiography (IOC) is a simple technique12 that can beperformed during LC. Before transection of any biliaryducts, a small catheter is inserted into the presumed cys-tic duct (or gallbladder neck) and contrast material is in-jected. Real-time fluoroscopy or static films can confirm

or correct assumptions made about biliary architecture.Intraoperative cholangiography was originally used inopen procedures for the detection of CBD stones; how-ever, in LC it may also serve as aroad mapfor opera-tive dissection. Some have suggested that IOC can helpdecrease the frequency13 and severity8,14-17 of CBD inju-ries by defining the biliary anatomy before duct transec-tion. Despite its potential benefit, however, not all sur-

METHODS

STUDY DESIGN

A retrospective cohort study was conducted with the useof a statewide hospital administrative discharge databaseto determine the effects of IOC and surgeons experienceon CBD injury after LC.

SETTING

Data were taken from the Washington State Comprehen-sive Hospital Abstract Reporting System database. This datasetis derived from all public andprivate hospitals in Wash-ington State (Veterans Affairs and US military hospitalsexcluded). It contains demographic variables, admission anddischarge administrative details, payer status,InternationalClassification of Diseases, Ninth Revision(ICD-9) procedureand diagnosis codes, and coded surgeon and hospital iden-tifiers. This study was granted an exception by agreementof theUniversity of WashingtonHumanSubjectReviewCom-mittee and the Washington State Department of Health. Thedata set includes only anonymous data and is considered

within the public domain.

SUBJECTS

All Comprehensive Hospital Abstract Reporting System re-ports from January 1, 1991, through December 31, 1998,were searched for ICD-9 procedure codes pertaining to LC.This group wasfurther evaluatedfor ICD-9 procedurescodesthat describe CBDinjury andrepair subsequent to LC. Ex-clusion criteriaincluded records with ICD-9 diagnosis codesconsistent with biliarytract malignancy and repairsof bili-ary tract fistulas. The LCs with major CBD injuries weretherefore defined as LCs necessitating either major re-pairof the CBD or anastomosis of bile duct to intestinewithin 90 days of LC (Table 1). The 90-day period forbile duct repair after LC was selectedbecause, in one large

surveyof CBDinjuries, this correspondedto themean timeuntil bile duct reconstruction.15

VARIABLE DEFINITIONS AND ANALYSIS

Univariate analysis was used to compare groups of pa-tientsundergoingLC with andwithout CBDinjury by meansof Stata (Stata Corp, College Station, Tex). Variables werecompared with Pearson 2, Fisher exact test, and un-paired, 2-tailed t test. We calculated the number of LCsper-formedby each surgeon beforeany given operation andcre-ated the variable case order as a marker of surgicalexperience at that pointin time. Case order ratherthantotalexperience with LC was considered the more relevantmarker of surgical experience, because it represents

experience at the time of the index LC within the sur-geons total experience curve. The lowest quartile of caseorders was consideredearly case order (defined as thesurgeons first 20 procedures). Relative risk (RR) of CBDinjury was calculated in relation to the use of IOC and/orwhether the operation occurred early in the sequence of asurgeons LC experience.Case order wasconsideredas botha continuous variable and a categorical variable on the ba-sis of quartiles. To examine variation among surgeons useof IOC, a second, surgeon-level variable (percentage of IOCuse) was created. This variable was defined as the propor-tion of LCs in which a given surgeon used IOC. The per-centage of IOC use was further categorized in relation tothe frequency of IOC use among all surgeons and dividedinto4 groups(15%, 15%-45%, 46%-75%, and75%IOCuse). Surgeonsin thegreater than 75% IOCuse group wereconsidered to beroutineusers of cholangiography. TheMantel extension test for trend was used to examine over-all odds ratios of CBD injury among different percentageof IOC use groups, adjusting for LCs in which an IOC wasor was not used.

Other analysis variables included age and sex, payerstatus, source of admission (emergency department, trans-fer, clinic), diagnoses and history of previous admissions

forbiliary tract disease (evidenced by ICD-9 diagnosticcodesconsistent with cholecystitis and/or cholecystitis with cho-ledochal disease, during previous hospitalizations),and hos-pital type (rural and nonprofit status as defined by Stateof

Washington definition, teaching if surgical training pro-grams were known to work at the hospital).

A multivariate logistic regression model was con-structed that included the primary variables of early sur-geon experience with LC (1st through 20th operation) andthe use of IOC. Models were constructed in 4 steps. Model1 (unadjusted) included only the primary covariates, IOCuse, and less experienced(case order, 1-20) variables.Model 2 included the primary covariates and demo-graphic variables (sex and age). Model 3 further includedpayer variables (Medicare, Medicaid, health maintenance

organization, and private). Finally, model 4 included theprimary covariates, demographics, payer variables, and pa-tient acuity variables (admission from emergency depart-ment, total previous admissions for biliary disease basedon ICD-9 diagnosis codes, and ICD-9 diagnosiscode for acutecholecystitis and/or concurrent choledochal disease at thetime of admission). Separate analyses using hospital-levelvariables (teaching, for-profit,and rural status) and surgeon-level variables (percentage of IOC useas a continuous vari-able) were also performed. Thismodel accounted for clus-tering by surgeon, andall results are presented with robustSEs.22 Unadjusted and adjusted odds ratios and predictedprobabilities were obtained by means of these logistic re-gression models. Because of the infrequency of CBD in-

jury, odds ratios approximate RR.





3/6

geons use IOC consistently and many have disputed thistheory.18-21 More important, while advocates of cholan-giography have speculated that more routine IOC usemight reduce the rate of CBD injury, this has not beenestablished in clinical trials. The purpose of this studywas to determine the effect of IOC use on CBD injury bymeans of a population-based survey and to examine thiseffect with regard to surgical experience.

RESULTS

A total of 30630 LCs were identified from 1991 through1998. The LCs were performed by 3447 attending phy-sicians at 138 hospitals. The number of patients under-going LC increased substantially from 1991 to 1992 andremained relatively stable from 1992 to 1998 (Figure 1).Seventy-six major CBD injuries (2.5/1000 or 1/400) by66 surgeons (range of injury per surgeon, 0-4) were iden-tified. The incidence of injury decreased slightly duringthe study period. A test for trend indicates that there wasa significant overall decrease in the rate of injury overtime (P =.05). After 1991, the incidence of CBD injurydeclined by approximately 10% per year.

Patients undergoing LC with subsequent CBD in-jury were compared with patients undergoing LC with-

out injury (Table 2). There were no significant differ-ences between groups on the basis of demographics,previous hospitalizations for biliary tract diagnoses, payerstatus, or hospital variables.

SURGEONS EXPERIENCE

Surgeons total experience (number of LCsperformed dur-ing an 8-year period) varied widely (mean, 68.7; me-dian, 56; range, 1-256). Twenty-five percent of sur-geons performed fewer than 20 LCs during the study

period, while 25% performed more than 100 LCs. TheLCs were analyzed on the basis of whether the surgeonperforming the LC was in the early (20 cases preced-ing), middle (21-36), high (37-75), or highest (75) ex-perience level group (based on quartile). The LCs per-formed in the early portion of a surgeons experience wereat significantly higher risk for injury than those per-formed at later points in thecase order. Injury to the CBD

occurred in 3.2 per 1000 of LCs in this early case order,compared with 1.7 per 1000 at later points in case order(RR, 1.81; 95% confidence interval, 1.44-2.88). Despitethis finding, no significant differences in the rate of CBDinjury were identified between later points in the expe-rience curve, and an overall test of trend failed to iden-tify significant differences in injury based on case orderfor case orders greater than 20 (P=.80). For example, therate of injury in LCs performed after surgeons 100th LCwas 1.6 per 1000 compared with 1.4 per 1000 after the150th LC (P =.90).

6000

3000

5000

4000

2000

1000

0

0.010

0.006

0.008

0.004

0.002

01996 1997 19981991 19 92 1993 199 4 1 995

Year

TotalOperations R

ateofInjury

No. of OperationsInjury Incidence

Figure 1.Number of patients undergoing laparoscopic cholecystectomy andrate of common bile duct injury from 1991 through 1998. The numberof operations increased substantially from 1991 to 1992 and remainedrelatively stable thereafter. The incidence of injury declined byapproximately 10% per year.

Table 1.ICD-9Procedure and Diagnostic Codes Usedas Inclusion and Exclusion Criteria*

ICD-9Code Description

Inclusion Criteria

Initial procedure code

51.23 Laparoscopic cholecystectomy

Subsequent procedure code

51.36 Choledochoenterostomy

51.37 Anastomosis of hepatic ductto intestinal tract

51.39 Other bile duct anastomosis

51.71 Repair of bile ducts

51.79 Repair of other bile ducts

51.94 Revision anastomosis of biliarytract

51.98 Percutaneous transhepaticbiliary drainage

Exclusion Criteria

P ro cedu re cod e 5 1.72 R ep air of f istula of commonbile duct

Diagnosis codes 155.1, 156,156.1, 156.8, and 156.9

Biliary tract and/or gallbladdercancer

*ICD-9indicatesInternational Classification of Diseases, Ninth Revision.

Within 90 days of initial procedure code.

Table 2. Similarity of Patients Within Study Group,Based on Presence of Injury*

VariablesCBD Injury

(n = 76)No CBD Injury(n = 30 630)

PValue

Age, mean SD, y 48.7 2.2 50.9 0.1 .29

Sex, % F 78.9 74.4 .80

Payer, %

Medicare 33.3 37.1 .69

Medicaid 19.7 15.4 .31

Private 36.8 37.9 .85

Disease severity, %

ED admission 30.3 37.4 .19

Acute cholecystitis 35.5 37.5 .73

CBD stone 3.9 6.2 .42

Hospital, %

Teaching 18.4 13 .16

Rural 9.7 9.5 .95

For-profit 4.9 4.3 .78

*CBD indicates common bile duct; ED, emergency department. There wereno significant differences between groups.

Diagnosis codes according to International Classification of Diseases,Ninth Revision.





4/6

INTRAOPERATIVE CHOLANGIOGRAM

Intraoperative cholangiography was performed in 63.7%of LCs. The rate of CBD injury was significantly lowerwhen IOC was used. The rate of injury in LCs per-formed without IOC was 3.3 per 1000, compared with2.0 per 1000in LCs with IOC (P=.02) (RR, 1.7; 95% con-

fidence interval, 1.1-2.6). Case order and IOC use wereindependent predictors of injury.

When the effect of IOC use on injury rate duringthe early portion of thelearning curvewas examined, sig-nificant differences were identified (Table 3). For ex-ample, during the first 36 LCs of surgeons (median caseorder of LCs in survey was 36), the risk of CBD injurywas significantly higher if no IOC was obtained (RR, 2.2;95% confidence interval, 1.31-3.74). The difference in

CBD injury rates between LCs done with and withoutIOC appeared somewhat larger at lower case orders, buta regression model showed no significant interactionbetween IOC and case order.

ROUTINE IOC USE

Because surgeons whouse IOCmore frequently mighthavea lower rate ofCBD injury independent ofIOC use, wecon-sidered the percentage of time that a surgeon performedIOC as a separate, surgeon-level covariate. Fifty percent ofsurgeons performed IOC less than 18% of the time. Tenpercent of surgeons performed IOC more than 75% of thetime, and only 5% usedIOC in more than 92% oftheir LCs.

When surgeon percentage of IOC use was examined as acategorical variable (15%, 15%-45%, 46%-75%,75%)CBD injuryrateswere notsignificantly differentamongLCsperformed by surgeons in any of thepercentage of IOC cat-egories (Table 4). For example, surgeons using IOCs inmore than 75% of LCs had an injury rate of 2.2 per 1000compared with 2.8 of 1000 among those who used IOCless than 15% of the time (P=.33). Furthermore, when thesubgroup of surgeons whoperformedIOCin nearlyallLCs(90%)wasevaluatedseparately,no significant benefitwasidentified (rate of injury, 1.8/1000) when compared withsurgeons using IOC less than 15% of the time(P=.25). In-terestingly,among surgeons who used IOCin greater than75% of LCs, there was a significantly lower rate of injury

in LCs that included the IOC compared with LCs that didnot include the IOC (1.6/1000 compared with 6.3/1000,respectively;P.001, Fisher exact test). No similar effectwas identifiedin surgeon groupswith other frequencies ofIOCuse. The Mantel extension test showed no significanttrends in CBD injury rates across percentage of IOC usegroups, adjusting for IOC use (P=.52).

MULTIVARIATE ANALYSIS

A logistic regression model was constructed to deter-mine whether the effect of case order and IOC use wasaltered by the addition of possible confounding vari-ables (Table 5). The unadjusted model and 3 adjusted

models all identified the negative association of IOC useand CBD injury. The magnitude of the association re-mained largely unchanged in all models. Hospital-levelvariables, including rural location, teaching, and/or for-profit status, were also tested and were found to have es-sentially no effect on odds ratios (data not shown).

PREDICTED PROBABILITY OF CBD INJURY

Case order and the use of IOC were incorporated into apredictive model for CBD injury.Figure 2graphically

Table 3. Injury Based on Case Order in Surgeons Seriesand Use of Intraoperative Cholangiography (IOC)

CaseOrder

No. (%)

Without IOC With IOC Total

20 26/5872 (0.44) 23/9473 (0.24)* 49/15345(0.32)

20-36 6/1691 (0.35) 1/3072 (0.03) 7/4763(0.15)

37-75 5/2150 (0.23) 7/4110 (0.17) 12/6260(0.19)

75 0/1403 8/2859 (0.28) 8/4262(0.19)

Total 37/11116(0.33) 39/19514(0.20) 76/30630(0.25)

*P= .03 vs no IOC.P= .01 vs rate of injury in all other case order groups.P.01 vs no IOC.

Table 4. Incidence of Injury by Surgeons Percentage Useof Intraoperative Cholangiography (IOC)

% IOCUse

No. (No./1000)P

ValueOverall Rate* Without IOC With IOC

15 7/2539 (2.8) 7/2444 (2.9) 0/95 .60

15-45 13/5801 (2.2) 9/4007 (2.2) 4/1794 (2.2) .90

46-75 25/7901 (3.2) 10/2920 (3.4) 15/4981 (3.0) .7075 31/14 389 (2.2) 11/1745 (6.3) 20/12 644 (1.6) .001

*No significant differences among IOC use categories based onMantel-Haenszel procedure for injury rate (by group) adjusted by IOC use(P= .50).

With vs without IOC use (Fisher exact test).

Table 5. Logistic Regression Analysisof Major Common Bile Duct Injury*

Odds Ratio(95% Confidence Interval)

Use ofCholangiography

EarlyExperience

Model 1, unadjusted 0.61 (0.39-0.96) 1.77 (1.10-2.80)

Model 2, demographic adjustedfor sex and age

0.62 (0.39-0.97) 1.79 (1.20-2.80)

Model 3, payer, adjusted for aboveand HMO payer, Medicare payer,Medicaid payer

0.62 (0.40-0.98) 1.82 (1.20-2.80)

Model 4, severity of disease,adjusted for above and EDadmission source, acutecholecystitis,previousadmissions for biliary disease,choledochal disease

0.63 (0.40-0.90) 1.8 (1.10-2.80)

*HMO indicates health maintenance organization; ED, emergencydepartment.

Numbers 1 to 20 in a surgeons operative series.

Diagnosis codes according to International Classification of Diseases, NinthRevision.





5/6

represents the predicted probability of CBD injury at thefull range of possible case orders considered in this sur-vey (1-256) and based on IOC use. Although the 2 fit-ted curves appear to have different slopes, there was nostatistically significant interactive effect between IOC useand case order (P =.09).

COMMENT

The wide acceptance of LC in the early 1990s was basedon several case series rather than randomized con-trolled trials. Community awareness that a minimallyinvasive procedure could dramatically reduce postop-erative pain and improve recuperation made it impracti-cal to conduct trials that might have better refined thetechnique. As a result, even though IOC has beenoffered by some8,13-17 as an approach to prevent CBDinjury during LC, testing its efficacy has not provedpractical because of the relative infrequency of the com-plication. Testing for a difference in injury rate of 1:400to 1:200 related to IOC use would require more than6000 patients in each limb of a hypothetical prospec-tive, randomized trial.

Although CBD injury occurs more commonly in theearly portion of a surgeons learning curve,1,8 recent re-ports4,5 suggest that even surgeons with a great deal ofexperience with LC have high rates of CBD injury. Tech-nical modifications to decrease injury rates have been sug-gested.23 These include proper angulation of the cysticduct and the CBD during initial dissection, use of anangled laparoscope, and a lowered threshold for conver-sion to an open procedure.11,17,24 These technical recom-mendations and proctoring of surgeons during their firstLCs have been incorporated into themodern eraof lapa-roscopy, yet the reported rate of CBD injury remainshigh.4,5

In the era when IOC was applied strictly for CBD

stone detection, surgeons used it either routinely or se-lectively on the basis of their confidence in the preop-erative diagnostic modalities related to CBD stones. Asa result, a debate in the surgical community has contin-ued for the past30 years25 regarding the benefitsand draw-backs of routine IOC use. The debate continues today26

even thoughthe benefit of IOC in the era ofLC has beenexpanded by also providing a road map of the biliary sys-tem. Somesurgeons useIOC routinely for CBD stone iden-tification, confirmation of assumptions made during op-erative dissection, and/or training purposes. Selectivecholangiographers tend to use the IOC when they are un-sure of the anatomy or when CBD stones are suspected.Others choose to obtain a pre-LC or post-LC endo-

scopic cholangiogram if a CBD stone is suspected. Thesesurgeons believe that IOC does not offer significant ad-vantages in operative dissection sufficient to justify theassociated operative time and cost of IOC.

Theresults of this retrospective cohort evaluation in-dicate that the rate of CBD injury after LC is significantlylowerabout 40%when IOC is used. As expected, sur-gical experience was also associated with the rate of CBDinjury. Significantlyhigher injury risk wasidentifiedin LCsperformed early in a surgeons case series. Therelative riskof CBD injury based on IOC use did not vary according

to case order. However, more CBDinjuries occurred at ear-lier parts of the learning curve, and, therefore, more CBDinjuries have the potential to be prevented by using IOC

early in a surgeons experience.With the use of predicted probability based on thedata from this survey, it may be possible to better definethe risk of CBD injury for patients undergoing LC. Sincemany surgeons quote the general rate of injury (as re-ported in previous series) of 1:300 to 1:350 (0.33%-0.29%), these predictive models may be helpful in theinformed consent process. For example, a surgeon whohas performed more than 250 LCs and uses IOC has apredicted probability of CBD injury that is significantlylower (0.6/1000) than that of a surgeon in his or her first20 LCs who does not use IOC (3.9/1000).

To distinguish the technology from the technolo-gist, this analysis tried to separate the benefit of the IOC

from other benefits that a surgeon whofrequently uses IOCmight offer. Surgeons who used IOC in a higher propor-tion of operations didnot appearto have significantly lowerrates of CBD injury. This may be because surgeons whoperformed IOC frequently (75% of LCs) had a signifi-cantly higher rate of injury in situations when they didnotobtain an IOC. This suggests that, when those who useIOCmore routinely deviate fromtheir typical practicepat-tern, they are more prone to CBD injury, or that the riskof CBD injury is higher in patients in whom no IOC wasobtainedor completed. This relationshipwasfurther evalu-ated in regression analysis using the surgeon-level vari-able of percentage of IOCuse and controlling for whetheran IOCwas obtained. With this regression analysis,no ad-

ditional effectof the more routine IOC user, separate fromIOC use itself, was identified. However, a comprehensiveevaluation to separate theeffects of IOC use from IOC user(surgeon) would require a much larger dataset.

This issue highlights the limitations of this data set:important clinical and intraoperative variables, such aswhy an IOC was used or the timing of the IOC in theLC, are not included. This limits our conclusions aboutwhy rates of injury might be different in LCs when noIOC was obtained. For example, surgeons may have de-termined that there was complex,high-riskanatomy

0.0040

0.0015

0.0025

0.0030

0.0035

0.0020

0.0010

0.0005

0

150 200 2500 50 100

Without Cholangiogram

With Cholangiogram

Surgeon's Case Order

ProbabilityofCBD

Injury

Figure 2.Predicted probability of major common bile duct (CBD) injurybased on experience, with and without the use of intraoperativecholangiography. Less experienced surgeons who use cholangiography arepredicted to have a lower rate of CBD injury than those who do not usecholangiograms.





6/6

in a patient and obtained an IOC in operations that weremore likely to have a CBD injury or after a CBD injuryhad already occurred. This, however, would result in ahigher rate of injury found with IOC use and woulddiminish the effect of IOC demonstrated in this survey.Conversely, injuries that were detected after the fact byan IOC might be underrepresented in this samplebecause of an increased chance for misclassification asan open cholecystectomy if an open repair was under-

taken at the time of injury recognition. However, thispossibility is not relevant to the issue of injury preven-tion. An IOC performed for the purpose of injury pre-vention should be done before the complete division ofany bile duct.

Another potential hazard of using administrative da-tabases for research is that it is difficult to effectively iden-tify operative complications. For example, although anICD-9code for accidental puncture during surgery(E870 series and 998.2) does exist, this method of casedetection has been shown to dramatically overreport therate of clinically relevant CBD injury.27 Our techniqueof CBD injury detection was devised to avoid this pit-fall. However, by describing onlymajor CBD injury

(through selection of cases based on the ICD-9proce-dure codes suggesting CBD repair or bile ductto-bowelanastomosis), we have excluded the minor CBD inju-ries, cystic duct leaks, and accessory duct transectionsthat also occurred during the study period. This study,by focusing on more clinically relevant injuries, under-estimates the rate of all CBD injuries.

The data set is also limited in its inclusion and clas-sification criteria: patients with CBD injury who died be-fore reoperation or who left the state after being injuredwould not be included in this survey. Also, the first LCof a surgeon in Washington who had previously accu-mulated experience with LC elsewhere would be con-sidered the first in his or her their experience curve. Fi-

nally, while it is clear that IOC was negatively associatedwith CBD injury, it is less clear that all injuries wouldhave been prevented had an IOC been performed. Forexample, a thermal injury to the CBD during dissectionwould go unrecognized by IOC and might present onlydays later. As well, in several anecdotal cases of CBD in-jury the IOC was misinterpreted or disregarded, or dis-section followed a properly obtained and interpreted IOCthat then resulted in injury. The finding that injury wasnearly half as likely when IOC was used suggests that arandomized controlled trial shouldbe conductedto provethe hypothesis that many CBD injuries are preventablewith increased use of IOC.

The relationship of IOC and CBD injury should be

further investigated to determine whether more routineuse of IOC might actas a system-level modification inthe prevention of surgical error. There are many issuesthat influence recommendations about routine IOC use.These include efficacy, cost (D.R.F., Christopher Flow-ers, MD, and David L. Veenstra, PhD, unpuplished data,2001), resource utilization, and reimbursement consid-erations. In the interim, this study suggests that sur-geons should consider the increased use of IOC, espe-cially in their early experience with LC.

Corresponding author and reprints: David R. Flum, MD,Department of Surgery, Robert Wood Johnson ClinicalScholars Program, University of Washington, H-220Health Sciences Center, Box 357183, Seattle, WA 98195-7183 (e-mail: [email protected]).

REFERENCES

1. The Southern Surgeons Club. A prospective analysis of 1518 laparoscopic cho-

lecystectomies [published correction appears inN Engl J Med. 1991;325:1517-1518].N Engl J Med.1991;324:1073-1078.

2. Bernard HR, Hartman TW. Complications after laparoscopic cholecystectomy.

Am J Surg.1993;165:533-535.

3. Gouma DJ, Go PM. Bile duct injury during laparoscopic and conventional cho-

lecystectomy.J Am Coll Surg. 1994;178:229-233.

4. Windsor JA,Pong J. Laparoscopicbiliary injury:more thana learning curve prob-

lem.Aust N Z J Surg. 1998;68:186-189.

5. Calvete J, Sabater L, Camps B, et al. Bile duct injury during laparoscopic chole-

cystectomy: myth or reality of the learning curve?Surg Endosc.2000;14:608-

611.

6. Kern KA. Medicolegal analysis of bile duct injury during open cholecystectomy

and abdominal surgery.Am J Surg.1994;168:217-222.

7. Schol FP, Go PM, Gouma DJ. Outcome of 49 repairs of bile duct injuries after

laparoscopic cholecystectomy.World J Surg.1995;19:753-757.

8. Woods MS, Traverso WL, Kozarek RA, et al. Characteristics of biliary tract com-

plications during laparoscopic cholecystectomy: a multi-institutional study.Am

J Surg.1994;167:27-34.9. Kern KA. An overview of 711 general surgery liability cases: the anatomy of sur-

gical malpractice claims.Bull Am Coll Surg.1995;80:34-49.

10. Carroll BJ, Birth M, Phillips EH. Common bile duct injuries during laparoscopic

cholecystectomy that result in litigation. Surg Endosc.1998;12:310-314.

11. Chen XR, Lou D, Li SH, et al. Avoiding serious complications in laparoscopic

cholecystectomylessons learned from an experience of 2428cases. AnnAcad

Med Singapore. 1996;25:635-639.

12. Sabharwal AJ, Minford EJ, Marson LP, et al. Laparoscopic cholangiography: a

prospective study.Br J Surg.1998;85:624-626.

13. Flowers JL, Zucker KA, Graham SM, et al. Laparoscopic cholangiography:

results and indications. Ann Surg.1992;215:209-216.

14. Kullman E, Borch K, Lindstrom E, et al. Value of routine intraoperative cholan-

giography in detecting aberrant bile ducts and bile duct injuries during laparo-

scopic cholecystectomy.Br J Surg.1996;83:171-175.

15. Woods MS, Traverso LW, Kozarek RA, et al. Biliary tract complications of lapa-

roscopic cholecystectomy are detected more frequentlywith routine intraopera-

tive cholangiography.Surg Endosc.1995;9:1076-1080.16. Carroll BJ, Friedman RL, Liberman MA, Phillips EH. Routine cholangiography

reduces sequelae of common bile duct injuries. Surg Endosc. 1996;10:1194-

1197.

17. Hunter JG. Avoidance of bile duct injury during laparoscopic cholecystectomy.

Am J Surg.1991;162:71-76.

18. Borjeson J, Liu SK, Jones S, Matolo NM. Selective intraoperative cholangiog-

raphy during laparoscopic cholecystectomy: how selective? Am Surg.2000;

66:616-618.

19. Lorimer JW, Fairfull-Smith RJ. Intraoperative cholangiography is not essential

to avoid duct injuries during laparoscopic cholecystectomy. Am J Surg.1995;

169:344-347.

20. Hawasli A. Does routine cystic duct cholangiogram during laparoscopic chole-

cystectomy prevent common bile duct injury? Surg Laparosc Endosc.1993;3:

290-295.

21. Wright KD, Wellwood JM. Bile duct injury during laparoscopic cholecystectomy

without operative cholangiography. Br J Surg.1998;85:191-194.

22. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous out-comes.Biometrics. 1986;42:121-130.

23. Cox MR, Wilson TG, Jeans PL, et al. Minimizing the risk of bile duct injury at

laparoscopic cholecystectomy.World J Surg.1994;18:422-427.

24. PerissatJ. Laparoscopic cholecystectomy:the European experience. AmJSurg.

1993;165:444-449.

25. Way LW, Admirand WH, Dunphy JE. Management of choledocholithiasis. Ann

Surg. 1972;176:347-359.

26. PhillipsEH. Routineversus selectiveintraoperative cholangiography.AmJ Surg.

1993;165:505-507.

27. Taylor B. Common bile duct injury during laparoscopic cholecystectomy in On-

tario: doesICD-9coding indicate true incidence?CMAJ.1998;158:481-485.



common bile duct injury

Documents