Long-Term Health-Related Quality of Life after Iatrogenic Bile Duct Injury Repair Aslam Ejaz, MD, MPH, Gaya Spolverato, MD, Yuhree Kim, MD, MPH, Rebecca Dodson, Jason K Sicklick, MD, FACS, Henry A Pitt, MD, FACS, Keith D Lillemoe, MD, FACS, John L Cameron, MD, FACS, Timothy M Pawlik, MD, MPH, PhD, FACS
MD,
Data on the effect of bile duct injuries (BDI) on health-related quality of life (HRQOL) are not well defined. We sought to assess long-term HRQOL after BDI repair in a large cohort of patients spanning a 23-year period. STUDY DESIGN: We identified and mailed HRQOL questionnaires to all patients treated for major BDI after laparoscopic cholecystectomy between January 1, 1990 and December 31, 2012 at Johns Hopkins Hospital. RESULTS: We identified 167 patients alive at the time of the study who met the inclusion criteria. Median age at BDI was 42 years (interquartile range 31 to 54 years); the majority of patients were female (n ¼ 131 [78.4%]) and of white race (n ¼ 137 [83.0%]). Most patients had Bismuth level 2 (n ¼ 56 [33.7%]) or Bismuth level 3 (n ¼ 40 [24.1%]) BDI. Surgical repair most commonly involved a Roux-en-Y hepaticojejunostomy (n ¼ 142 [86.1%]). Sixty-two patients (37.1%) responded to the HRQOL questionnaire. Median follow-up was 169 months (interquartile range 125 to 222 months). At the time of BDI, mental health was most affected, with patients commonly reporting a depressed mood (49.2%) or low energy level (40.0%). These symptoms improved significantly after definitive repair (both p < 0.05). Limitations in physical activity and general health remained unchanged before and after surgical repair (both p > 0.05). CONCLUSIONS: Mental health concerns were more commonplace vs physical or general health issues among patients with BDI followed long term. Optimal multidisciplinary management of BDI can help restore HRQOL to preinjury levels. (J Am Coll Surg 2014;219:923e932. 2014 by the American College of Surgeons)
BACKGROUND:
Cholelithiasis, the primary risk factor for benign gallbladder disease, affects between 10% and 15% of the adult US population.1 Consequently, >750,000 cholecystectomies are performed annually in the United States for benign gallbladder disease.1 Shortly after its introduction
in the late 1980s, laparoscopic cholecystectomy (LC) became the standard of care for acute uncomplicated cholecystitis and symptomatic gallbladder disease. The transition to the minimally invasive removal of the gallbladder has resulted in shorter hospital length of stay (LOS) and less postoperative pain, but consequently the incidence of iatrogenic bile duct injuries (BDI) has increased.2-4 In fact, recent studies have found a 2- to 4fold increase in the risk of BDI after LC as compared with traditional open cholecystectomy,3 with current BDI rates estimated between 0.4% and 0.6%.4-6 Minor BDI and biliary leaks can be treated through noninvasive modalities, including endoscopic or percutaneous interventions with good results.7 Major BDI, however, often requires temporizing interventions via percutaneous or endoscopic stent and drain placement with delayed surgical repair typically after several weeks to allow resolution of local inflammation.6,7 Several studies have shown that definitive surgical repair for BDI often results in
Disclosure Information: Nothing to disclose. Support: Dr Aslam Ejaz was supported in part by the Eleanor B Pillsbury Foundation for surgical research. Received February 10, 2014; Revised April 27, 2014; Accepted April 30, 2014. From the Department of Surgery, Johns Hopkins Hospital, Baltimore, MD (Ejaz, Spolverato, Kim, Dodson, Cameron, Pawlik), Department of Surgery, University of Illinois Hospital and Health Sciences Center, Chicago, IL (Ejaz), Department of Surgery, University of California San Diego School of Medicine, La Jolla, CA (Sicklick), Department of Surgery, Temple University School of Medicine, Philadelphia, PA (Pitt), and Department of Surgery, The Massachusetts General Hospital, Harvard Medical School, Boston, MA (Lillemoe). Correspondence address: Timothy M Pawlik, MD, MPH, PhD, FACS, Department of Surgery, Johns Hopkins Hospital, 600 N Wolfe St, Blalock 688, Baltimore, MD 21287. email: [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
923
http://dx.doi.org/10.1016/j.jamcollsurg.2014.04.024 ISSN 1072-7515/14
Ejaz et al
924
Quality of Life after Bile Duct Injury Repair
Abbreviations and Acronyms
BDI GIQLI HRQOL IQR LC LOS SF-36
¼ ¼ ¼ ¼ ¼ ¼ ¼
bile duct injury Gastrointestinal Quality of Life Index health-related quality of life interquartile range laparoscopic cholecystectomy length of stay short-form 36
acceptable perioperative outcomes; however, the longterm effects of BDI remain largely disputed.3,8-12 Iatrogenic BDI during LC has been shown to affect both short-term (longer hospital length of stay, increased hospital costs and perioperative morbidity)13,14 and longterm outcomes (increased mortality).14,15 Although data exist on commonly reported surgical outcomes, such as perioperative morbidity and mortality after BDI repair, patient reported outcomes on health-related quality of life (HRQOL) are more scarce. Patient-reported outcomes are, however, of equal if not greater significance in determining the true success of an intervention. Several heterogeneous studies of varying sample size and length of follow-up have attempted to evaluate the impact of BDI on HRQOL, with mixed results.9,16-20 More than a decade ago, our group published one of the earliest reports of HRQOL after BDI repair.20 The objective of the current study is to provide an updated assessment of the short- and long-term HRQOL after BDI repair at a major tertiary hepatopancreaticobiliary referral center in a large cohort of patients that now spans more than 20 years. In addition, we aimed to identify potential patient and clinical factors that might be associated with improved HRQOL after BDI repair.
METHODS Patients and data collection We identified all patients treated for major BDI after LC between January 1, 1990 and December 31, 2012 at the Johns Hopkins Hospital. Major BDI is defined as all transections or lacerations of the common bile duct, common hepatic duct, or major segmental ducts of the biliary system. All patients with minor leaks from the cystic duct or gallbladder bed and those who sustained a BDI after an entirely open cholecystectomy were excluded from analysis. Patients with bile duct disease from causes other than iatrogenic injury during LC were also excluded. Patients who were found to be dead from any cause as ascertained from the Social Security Death Index or whose addresses were unknown or outdated were omitted from all analyses.
J Am Coll Surg
Type of BDI was classified using the Bismuth classification system.21 Standard patient demographic data were collected. Operative information at the time of BDI and any pretransfer attempts at percutaneous or endoscopic intervention were recorded, if available. Operative data for the definitive BDI repair were recorded, including type of operation and the number of biliary stents and abdominal drains placed, if applicable. Perioperative complications after definitive BDI repair were recorded during the index hospitalization or if they occurred within 90 days from the date of surgery. Patients were classified based on their response status to the HRQOL questionnaire (responders vs nonresponders). The Johns Hopkins University Institutional Review Board approved the study. Quality of life assessment All patients treated for major BDI at our institution and known to be alive at the time of the study (as assessed through the Social Security Death Index) were mailed questionnaires designed to evaluate HRQOL (questionnaire is in the Appendix, online only; available at: http://www.journalacs.org). As approved from the Institutional Review Board, two mailings were conducted to all patients known to be alive at the time of the study. All patients could opt out of being contacted again. If no response was received after the second mailing, attempts were made to contact patients by phone to inform them of the study. A final third mailing was then attempted to all patients who had not yet responded. As no specific questionnaire exists to evaluate HRQOL after BDI repair, the survey was constructed using elements of the generic Short Form-36 (SF-36) questionnaire, as well as the disease-specific Gastrointestinal Quality of Life Index (GIQLI) and chronic liver disease questionnaire.22-24 The SF-36 was used to assess general health perception and well-being, and elements from the GIQLI and chronic liver disease questionnaire were used to assess specific gastrointestinal symptoms. The survey consisted of 70 questions designed to evaluate general physical and mental health, role functioning, general health perceptions, and current gastrointestinal and constitutional symptomatology using standard Likert scale assessments. For comparison purposes, patients were asked to evaluate their HRQOL pre (the time after BDI and before definitive repair) and post definitive surgical repair. The survey also included elements of interest specific to BDI, such as inquiries about litigation, attitudes toward percutaneous and endoscopic interventions, and previous and current work and financial status. A pretest analysis of the survey was performed on a random selection of 10 patients to ensure that all items and
Vol. 219, No. 5, November 2014
Ejaz et al
questions in the survey were clear and appropriate. On average, the survey took approximately 20 minutes to complete. Statistical analysis Continuous variables are presented as the median with the interquartile range (IQR), where appropriate. Categorical variables are displayed as whole numbers and percentages. Comparative analyses of continuous variables were performed using Wilcoxon test for parametric and nonparametric data and one-way ANOVA, as appropriate. Fisher’s exact test or chi-square test was used for comparing categorical variables. Logistic regression analysis was performed to test the association of clinicopathologic and operative characteristics with improved HRQOL. For statistical analyses, p values 0.05) (Table 1). Responders and nonresponders also did not differ by type of definitive operation, length from injury to definitive repair, LOS after repair, postoperative morbidity, readmission, or length of postoperative stenting (all p > 0.05). In fact, compared with responders, nonresponders
926
Table 1.
Ejaz et al
J Am Coll Surg
Quality of Life after Bile Duct Injury Repair
Clinicopathologic Characteristics of Cohort, Stratified by Questionnaire Response Status
Age at BDI, y, median (IQR) Female sex, n (%) Race, n (%) White African American Hispanic Initial operation location, n (%) Outside institution Johns Hopkins Hospital Type of initial operation, n (%) Laparoscopic cholecystectomy Laparoscopic to open cholecystectomy Laparoscopic to open cholecystectomy with CBD exploration Injury recognized during initial surgery, n (%) Attempted repair before transfer, n (%) Hepaticojejunostomy End to end ductal repair Choledochojejunostomy Choledochoduodenostomy Cholecystojejunostomy Level of injury, n (%) Bismuth 1 Bismuth 2 Bismuth 3 Bismuth 4 Bismuth 5 BDI presentation, n (%) Jaundice Bile leak Cholangitis Pain Postoperative procedures before referral, n (%) ERCP PTC PTC þ ERCP
Total (n ¼ 167)
Nonresponders (n ¼ 105)
Responders (n ¼ 62)
42.0 (31.0e54.0) 131 (78.4)
39.0 (31.0e50.5) 84 (80.0)
44.0 (33.0e57.0) 47 (75.8)
136 (83.0) 17 (10.2) 10 (6.0)
81 (78.1) 13 (12.4) 7 (6.7)
55 (88.7) 4 (6.5) 3 (4.8)
158 (94.6) 9 (5.4)
98 (93.3) 7 (6.7)
60 (96.8) 2 (3.2)
104 30 31 60
(62.3) (18.0) (18.6) (37.0)
64 18 22 42
(61.0) (17.1) (21.0) (42.0)
40 12 9 18
(64.5) (19.4) (14.5) (29.0)
29 25 3 3 2
(17.7) (15.2) (1.8) (1.8) (1.2)
17 18 2 2 2
(16.7) (17.7) (2.0) (2.0) (2.0)
12 7 1 1
(19.4) (11.3) (1.6) (1.6) 0
p Value
0.10 0.52 0.28
0.34
0.75
0.10 0.87
0.34 21 56 40 16 23
(12.7) (33.7) (24.1) (9.6) (13.9)
15 35 29 8 11
(14.4) (33.7) (27.9) (7.7) (10.6)
6 21 11 8 12
(9.7) (33.9) (17.7) (12.9) (19.4)
36 38 15 6
(36.4) (18.2) (16.2) (6.1)
18 18 12 6
(32.1) (32.1) (21.4) (10.7)
18 (41.9) 20 (46.5) 3 (7.0) 0
11 (11.2) 4 (4.1) 72 (73.5)
7 (11.5) 4 (6.6) 46 (75.4)
0.04
0.9 18 (11.3) 8 (5.0) 118 (74.2)
BDI, bile duct injury; CBD, common bile duct; IQR, interquartile range; PTC, percutaneous transhepatic cholangiography.
differed only by the number of stents (nonresponders: 1 stent; IQR 1 to 2 stents vs responders: 2 stents; IQR 1 to 2 stents; p ¼ 0.007) and drains (nonresponders: 2 drains; IQR 1 to 3 drains vs responders: 2 drains; IQR 2 to 3; p ¼ 0.03) used at the time of surgery (Table 2). Bile duct injuryespecific inquiries Among the 62 patients responding to the survey, 43 (70.5%) patients sought litigation for their injury, with the majority of those patients reporting that they had “won” their lawsuit (n ¼ 29 [70.7%]) (Table 3). Most
patients (n ¼ 45 [72.6%]) did not believe that they were adequately informed of the possibility of a BDI before surgery. Of specific interest to patients with BDI, the majority of responders said that biliary tubes interfered with their daily activities (n ¼ 36 [66.5%]) and intimate situations (n ¼ 36 [67.9%]). Although a minority of patients found caring for their biliary stents to be difficult (n ¼ 19 [33.8%]), most admitted to being embarrassed by their tubes (n ¼ 33 [60.0%]). With regard to the personal financial impact of BDI, 44.3% (n ¼ 27) of patients reported having substantial financial
Ejaz et al
Vol. 219, No. 5, November 2014
Table 2.
Quality of Life after Bile Duct Injury Repair
927
Factors Related to Definitive Bile Duct Injury Repair and Perioperative Outcomes Total
Nonresponders (n ¼ 105)
Responders (n ¼ 62)
17 (4e114) 48 (29.1) 74 (25.5e283.5)
14 (4e105) 32 (31.1) 78 (16e273)
21 (5e124) 16 (25.8) 59 (43e289)
Factors
Days from injury to transfer, median (IQR) Attempted repair at initial transfer, n (%) Days from BDI to definitive repair, median (IQR) Definitive operation type, n (%) Hepaticojejunostomy Primary end-to-end ductal repair Choledochojejunostomy No. of postoperative stents (IQR) No. of postoperative drains (IQR) Postoperative morbidity, n (%) Postoperative mortality, 90-d, n (%) Length of stay after repair, median (IQR) Postoperative readmission, all cause, median (IQR) Length of postoperative stenting, d, median (IQR) Failed repair requiring reoperation or stent reinsertion, n
142 1 3 2 2 59 1 8 0 208
(86.1) (0.6) (1.8) (1e2) (1e3) (33.5) (0.7) (7e10) (0e1) (113e377) 5
86 1 3 1 2 35 1 8 0 204
(82.7) (1.0) (2.9) (1e2) (1e3) (20.0) (1.0) (7e10) (0e1) (98e377) 2
p Value
0.94 0.47 0.06 0.54
56 (91.8) 0 0 2 (1e2) 2 (2e3) 24 (38.7) 0 9 (7e11) 1 (0e2) 239 (149e380) 3
0.007 0.03 0.48 0.44 0.91 0.24 0.34 0.47
BDI, bile duct injury; IQR, interquartile range.
hardship due to their BDI. Although the majority of patients were able to return to work after their BDI (n ¼ 43 [84.3%]), 27.1% (n ¼ 16) of patients believed their job was in jeopardy at some point during their treatment course. Health-related quality of life assessment Patients were asked to assess their physical, mental, and general health both before BDI repair and currently. At the time of BDI, mental health appeared to be most affected, with nearly half of all patients reporting a depressed mood (49.2%) or low energy level (40.0%) “most” or “all the time” before BDI repair (Fig. 2). These Table 3.
symptoms improved significantly after repair, as only 18.3% reported having a depressed mood and 18.0% having low energy “most” or “all the time” currently (both p < 0.05). Interestingly, limitations in physical activity and general health remained unchanged before and after surgical repair (both p > 0.05). The majority of patients reported having good to excellent health both before BDI repair (61.7%) and currently (65.6%). Pain levels on a 5-point Likert scale (0 ¼ no pain to 4 ¼ severe pain) were assessed and categorized based on time after BDI repair. The mean pain score of all patients preoperatively was 2.1 (SD 1.6) with the majority of patients (59.0%) reporting moderate to extreme pain
Response to Bile Duct InjuryeSpecific Inquiries
Inquiries
Were you informed of risk of BDI before surgery? Did you seek litigation?* Did you receive a positive verdict?* Did your biliary tubes interfere with daily activities? Did your biliary tubes interfere with intimacy? Did your biliary tubes embarrass you? Was it difficult caring for your biliary tubes? Were you affected by not being able to work? Were you affected financially? Was your job put in jeopardy Were you affected by changes in your physical appearance? Were you able to return to work? *Not at all indicates “no”; very much indicates “yes.” BDI, bile duct injury.
Total responders, %
62 61 41 55 53 55 56 59 61 59 59 51
Not at all n %
45 18 12 3 2 9 8 9 10 30 10 8
72.6 29.5 29.3 5.5 3.8 16.4 14.3 15.3 16.4 50.9 17.0 15.7
Mild to moderate amount n %
4 d d 16 15 13 29 11 24 13 26 d
6.5 d d 29.1 28.3 23.6 51.8 18.6 39.3 22.0 44.1 d
Very much n %
13 43 29 36 36 33 19 39 27 16 23 43
21.0 70.5 70.7 65.5 67.9 60.0 33.9 66.1 44.3 27.1 39.0 84.3
928
Ejaz et al
J Am Coll Surg
Quality of Life after Bile Duct Injury Repair
Figure 2. Evaluation of health-related quality of life before bile duct injury repair and currently. *Without limitation, **most or all the time, ***statistically significant difference (p < 0.05) between groups.
(Table 4). Interestingly, patients with 10 years of follow-up after BDI repair, mean pain scores were significantly decreased (10 to 15 years: 1.0 [SD 0.9]; >15 years: 1.0 [SD 0.7]; p < 0.05). In addition, only 8% of patients reported having severe or very severe pain 10 to 15 years postoperatively, and none of the patients with >15 years of follow-up reported having severe or very severe pain (Fig. 3). Other commonly reported symptoms were also assessed based on the number of events occurring within the past 4 weeks of evaluation. Most patients were relatively pain-free, with the majority reporting having 6 days/week, 3.4%) followed by frequent/urgent bowel movements (1 to 5 days/week, 40.0%; >6 days/week, 8.3%). Table 4.
Factors associated with improved health-related quality of life Various demographic, clinicopathologic, and management factors were analyzed for their effect on HRQOL (Table 5). Clinicopathologic and management factors such as age, sex, race, type of injury, and length of follow-up were not associated with predicting HRQOL. In addition, HRQOL did not differ among patients seeking litigation or among those who reported winning their lawsuit.
DISCUSSION More than 30 million Americans are affected by gallstones, resulting in nearly 2 million ambulatory care visits each year.1 Since the introduction of LC in 1985, the management of benign gallbladder disease has been revolutionized. As with any radical change, the new technology was cautiously and carefully adopted. In time, LC proved to be an effective method to treat benign
Pain Scores Stratified by Time from Bile Duct Injury Repair
Pain level
0 (None), n (%) 1 (Mild), n (%) 2 (Moderate), n (%) 3 (Severe), n (%) 4 (Extreme), n (%) Mean (SD)
Preoperative
14 11 9 8 19 2.1
(23.0) (18.0) (14.8) (13.1) (31.2) (1.6)
15 Years postoperative
10 (45.5) 8 (36.4) 4 (18.2) 0 0 1.0 (0.7)
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
929
Table 5. Univariate Analysis of Factors Related to Improved Quality of Life Factors
Figure 3. Pain levels as stratified by time of follow-up after definitive surgical bile duct injury repair.
gallbladder disease, resulting in less postoperative pain and decreased hospital LOS as compared with traditional open cholecystectomy.2-4 However, as the adoption of LC became more widespread, the incidence of BDI after surgery was reported to be considerably higher, with estimates ranging from 2 to 4 times the risk as compared with traditional open cholecystectomy. Bile duct injury has been shown to cause considerable morbidity and mortality in patients as compared with those undergoing uncomplicated LC.13,15 In addition, the financial impact from the management of BDI is substantial, resulting in costs up to 26 times greater than an uncomplicated LC.13 Recent improvements in surgical technique and care for patients with BDI have improved immediate and long-term outcomes, such as morbidity, mortality, and readmissions.7 Of greater consequence, however, is the potential impact of BDI on HRQOL. Studies analyzing patientreported HRQOL are necessary to truly evaluate the effectiveness of the management of this difficult injury. Our group reported one of the first analysis of HRQOL in patients undergoing BDI repair.20 In this study, 54 patients were followed for a median of 59 months and were
Age Younger than 42 y 42 y or older Sex Male Female Race White Other Earlier repair at outside hospital Yes No No. of postoperative stents 1 2 3 Level of injury Low (Bismuth 1, 2) High (Bismuth 3, 4, 5) Days from surgery to stent removal 10 45. How often do you require follow-up of your bile duct injury with a healthcare professional? a. Weekly b. Monthly c. 2 to 4 times per year d. Yearly e. No longer require follow-up for my injury
932.e5
932.e6
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
932.e7
932.e8
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
932.e9
932.e10
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
MD,
Data on the effect of bile duct injuries (BDI) on health-related quality of life (HRQOL) are not well defined. We sought to assess long-term HRQOL after BDI repair in a large cohort of patients spanning a 23-year period. STUDY DESIGN: We identified and mailed HRQOL questionnaires to all patients treated for major BDI after laparoscopic cholecystectomy between January 1, 1990 and December 31, 2012 at Johns Hopkins Hospital. RESULTS: We identified 167 patients alive at the time of the study who met the inclusion criteria. Median age at BDI was 42 years (interquartile range 31 to 54 years); the majority of patients were female (n ¼ 131 [78.4%]) and of white race (n ¼ 137 [83.0%]). Most patients had Bismuth level 2 (n ¼ 56 [33.7%]) or Bismuth level 3 (n ¼ 40 [24.1%]) BDI. Surgical repair most commonly involved a Roux-en-Y hepaticojejunostomy (n ¼ 142 [86.1%]). Sixty-two patients (37.1%) responded to the HRQOL questionnaire. Median follow-up was 169 months (interquartile range 125 to 222 months). At the time of BDI, mental health was most affected, with patients commonly reporting a depressed mood (49.2%) or low energy level (40.0%). These symptoms improved significantly after definitive repair (both p < 0.05). Limitations in physical activity and general health remained unchanged before and after surgical repair (both p > 0.05). CONCLUSIONS: Mental health concerns were more commonplace vs physical or general health issues among patients with BDI followed long term. Optimal multidisciplinary management of BDI can help restore HRQOL to preinjury levels. (J Am Coll Surg 2014;219:923e932. 2014 by the American College of Surgeons)
BACKGROUND:
Cholelithiasis, the primary risk factor for benign gallbladder disease, affects between 10% and 15% of the adult US population.1 Consequently, >750,000 cholecystectomies are performed annually in the United States for benign gallbladder disease.1 Shortly after its introduction
in the late 1980s, laparoscopic cholecystectomy (LC) became the standard of care for acute uncomplicated cholecystitis and symptomatic gallbladder disease. The transition to the minimally invasive removal of the gallbladder has resulted in shorter hospital length of stay (LOS) and less postoperative pain, but consequently the incidence of iatrogenic bile duct injuries (BDI) has increased.2-4 In fact, recent studies have found a 2- to 4fold increase in the risk of BDI after LC as compared with traditional open cholecystectomy,3 with current BDI rates estimated between 0.4% and 0.6%.4-6 Minor BDI and biliary leaks can be treated through noninvasive modalities, including endoscopic or percutaneous interventions with good results.7 Major BDI, however, often requires temporizing interventions via percutaneous or endoscopic stent and drain placement with delayed surgical repair typically after several weeks to allow resolution of local inflammation.6,7 Several studies have shown that definitive surgical repair for BDI often results in
Disclosure Information: Nothing to disclose. Support: Dr Aslam Ejaz was supported in part by the Eleanor B Pillsbury Foundation for surgical research. Received February 10, 2014; Revised April 27, 2014; Accepted April 30, 2014. From the Department of Surgery, Johns Hopkins Hospital, Baltimore, MD (Ejaz, Spolverato, Kim, Dodson, Cameron, Pawlik), Department of Surgery, University of Illinois Hospital and Health Sciences Center, Chicago, IL (Ejaz), Department of Surgery, University of California San Diego School of Medicine, La Jolla, CA (Sicklick), Department of Surgery, Temple University School of Medicine, Philadelphia, PA (Pitt), and Department of Surgery, The Massachusetts General Hospital, Harvard Medical School, Boston, MA (Lillemoe). Correspondence address: Timothy M Pawlik, MD, MPH, PhD, FACS, Department of Surgery, Johns Hopkins Hospital, 600 N Wolfe St, Blalock 688, Baltimore, MD 21287. email: [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
923
http://dx.doi.org/10.1016/j.jamcollsurg.2014.04.024 ISSN 1072-7515/14
Ejaz et al
924
Quality of Life after Bile Duct Injury Repair
Abbreviations and Acronyms
BDI GIQLI HRQOL IQR LC LOS SF-36
¼ ¼ ¼ ¼ ¼ ¼ ¼
bile duct injury Gastrointestinal Quality of Life Index health-related quality of life interquartile range laparoscopic cholecystectomy length of stay short-form 36
acceptable perioperative outcomes; however, the longterm effects of BDI remain largely disputed.3,8-12 Iatrogenic BDI during LC has been shown to affect both short-term (longer hospital length of stay, increased hospital costs and perioperative morbidity)13,14 and longterm outcomes (increased mortality).14,15 Although data exist on commonly reported surgical outcomes, such as perioperative morbidity and mortality after BDI repair, patient reported outcomes on health-related quality of life (HRQOL) are more scarce. Patient-reported outcomes are, however, of equal if not greater significance in determining the true success of an intervention. Several heterogeneous studies of varying sample size and length of follow-up have attempted to evaluate the impact of BDI on HRQOL, with mixed results.9,16-20 More than a decade ago, our group published one of the earliest reports of HRQOL after BDI repair.20 The objective of the current study is to provide an updated assessment of the short- and long-term HRQOL after BDI repair at a major tertiary hepatopancreaticobiliary referral center in a large cohort of patients that now spans more than 20 years. In addition, we aimed to identify potential patient and clinical factors that might be associated with improved HRQOL after BDI repair.
METHODS Patients and data collection We identified all patients treated for major BDI after LC between January 1, 1990 and December 31, 2012 at the Johns Hopkins Hospital. Major BDI is defined as all transections or lacerations of the common bile duct, common hepatic duct, or major segmental ducts of the biliary system. All patients with minor leaks from the cystic duct or gallbladder bed and those who sustained a BDI after an entirely open cholecystectomy were excluded from analysis. Patients with bile duct disease from causes other than iatrogenic injury during LC were also excluded. Patients who were found to be dead from any cause as ascertained from the Social Security Death Index or whose addresses were unknown or outdated were omitted from all analyses.
J Am Coll Surg
Type of BDI was classified using the Bismuth classification system.21 Standard patient demographic data were collected. Operative information at the time of BDI and any pretransfer attempts at percutaneous or endoscopic intervention were recorded, if available. Operative data for the definitive BDI repair were recorded, including type of operation and the number of biliary stents and abdominal drains placed, if applicable. Perioperative complications after definitive BDI repair were recorded during the index hospitalization or if they occurred within 90 days from the date of surgery. Patients were classified based on their response status to the HRQOL questionnaire (responders vs nonresponders). The Johns Hopkins University Institutional Review Board approved the study. Quality of life assessment All patients treated for major BDI at our institution and known to be alive at the time of the study (as assessed through the Social Security Death Index) were mailed questionnaires designed to evaluate HRQOL (questionnaire is in the Appendix, online only; available at: http://www.journalacs.org). As approved from the Institutional Review Board, two mailings were conducted to all patients known to be alive at the time of the study. All patients could opt out of being contacted again. If no response was received after the second mailing, attempts were made to contact patients by phone to inform them of the study. A final third mailing was then attempted to all patients who had not yet responded. As no specific questionnaire exists to evaluate HRQOL after BDI repair, the survey was constructed using elements of the generic Short Form-36 (SF-36) questionnaire, as well as the disease-specific Gastrointestinal Quality of Life Index (GIQLI) and chronic liver disease questionnaire.22-24 The SF-36 was used to assess general health perception and well-being, and elements from the GIQLI and chronic liver disease questionnaire were used to assess specific gastrointestinal symptoms. The survey consisted of 70 questions designed to evaluate general physical and mental health, role functioning, general health perceptions, and current gastrointestinal and constitutional symptomatology using standard Likert scale assessments. For comparison purposes, patients were asked to evaluate their HRQOL pre (the time after BDI and before definitive repair) and post definitive surgical repair. The survey also included elements of interest specific to BDI, such as inquiries about litigation, attitudes toward percutaneous and endoscopic interventions, and previous and current work and financial status. A pretest analysis of the survey was performed on a random selection of 10 patients to ensure that all items and
Vol. 219, No. 5, November 2014
Ejaz et al
questions in the survey were clear and appropriate. On average, the survey took approximately 20 minutes to complete. Statistical analysis Continuous variables are presented as the median with the interquartile range (IQR), where appropriate. Categorical variables are displayed as whole numbers and percentages. Comparative analyses of continuous variables were performed using Wilcoxon test for parametric and nonparametric data and one-way ANOVA, as appropriate. Fisher’s exact test or chi-square test was used for comparing categorical variables. Logistic regression analysis was performed to test the association of clinicopathologic and operative characteristics with improved HRQOL. For statistical analyses, p values 0.05) (Table 1). Responders and nonresponders also did not differ by type of definitive operation, length from injury to definitive repair, LOS after repair, postoperative morbidity, readmission, or length of postoperative stenting (all p > 0.05). In fact, compared with responders, nonresponders
926
Table 1.
Ejaz et al
J Am Coll Surg
Quality of Life after Bile Duct Injury Repair
Clinicopathologic Characteristics of Cohort, Stratified by Questionnaire Response Status
Age at BDI, y, median (IQR) Female sex, n (%) Race, n (%) White African American Hispanic Initial operation location, n (%) Outside institution Johns Hopkins Hospital Type of initial operation, n (%) Laparoscopic cholecystectomy Laparoscopic to open cholecystectomy Laparoscopic to open cholecystectomy with CBD exploration Injury recognized during initial surgery, n (%) Attempted repair before transfer, n (%) Hepaticojejunostomy End to end ductal repair Choledochojejunostomy Choledochoduodenostomy Cholecystojejunostomy Level of injury, n (%) Bismuth 1 Bismuth 2 Bismuth 3 Bismuth 4 Bismuth 5 BDI presentation, n (%) Jaundice Bile leak Cholangitis Pain Postoperative procedures before referral, n (%) ERCP PTC PTC þ ERCP
Total (n ¼ 167)
Nonresponders (n ¼ 105)
Responders (n ¼ 62)
42.0 (31.0e54.0) 131 (78.4)
39.0 (31.0e50.5) 84 (80.0)
44.0 (33.0e57.0) 47 (75.8)
136 (83.0) 17 (10.2) 10 (6.0)
81 (78.1) 13 (12.4) 7 (6.7)
55 (88.7) 4 (6.5) 3 (4.8)
158 (94.6) 9 (5.4)
98 (93.3) 7 (6.7)
60 (96.8) 2 (3.2)
104 30 31 60
(62.3) (18.0) (18.6) (37.0)
64 18 22 42
(61.0) (17.1) (21.0) (42.0)
40 12 9 18
(64.5) (19.4) (14.5) (29.0)
29 25 3 3 2
(17.7) (15.2) (1.8) (1.8) (1.2)
17 18 2 2 2
(16.7) (17.7) (2.0) (2.0) (2.0)
12 7 1 1
(19.4) (11.3) (1.6) (1.6) 0
p Value
0.10 0.52 0.28
0.34
0.75
0.10 0.87
0.34 21 56 40 16 23
(12.7) (33.7) (24.1) (9.6) (13.9)
15 35 29 8 11
(14.4) (33.7) (27.9) (7.7) (10.6)
6 21 11 8 12
(9.7) (33.9) (17.7) (12.9) (19.4)
36 38 15 6
(36.4) (18.2) (16.2) (6.1)
18 18 12 6
(32.1) (32.1) (21.4) (10.7)
18 (41.9) 20 (46.5) 3 (7.0) 0
11 (11.2) 4 (4.1) 72 (73.5)
7 (11.5) 4 (6.6) 46 (75.4)
0.04
0.9 18 (11.3) 8 (5.0) 118 (74.2)
BDI, bile duct injury; CBD, common bile duct; IQR, interquartile range; PTC, percutaneous transhepatic cholangiography.
differed only by the number of stents (nonresponders: 1 stent; IQR 1 to 2 stents vs responders: 2 stents; IQR 1 to 2 stents; p ¼ 0.007) and drains (nonresponders: 2 drains; IQR 1 to 3 drains vs responders: 2 drains; IQR 2 to 3; p ¼ 0.03) used at the time of surgery (Table 2). Bile duct injuryespecific inquiries Among the 62 patients responding to the survey, 43 (70.5%) patients sought litigation for their injury, with the majority of those patients reporting that they had “won” their lawsuit (n ¼ 29 [70.7%]) (Table 3). Most
patients (n ¼ 45 [72.6%]) did not believe that they were adequately informed of the possibility of a BDI before surgery. Of specific interest to patients with BDI, the majority of responders said that biliary tubes interfered with their daily activities (n ¼ 36 [66.5%]) and intimate situations (n ¼ 36 [67.9%]). Although a minority of patients found caring for their biliary stents to be difficult (n ¼ 19 [33.8%]), most admitted to being embarrassed by their tubes (n ¼ 33 [60.0%]). With regard to the personal financial impact of BDI, 44.3% (n ¼ 27) of patients reported having substantial financial
Ejaz et al
Vol. 219, No. 5, November 2014
Table 2.
Quality of Life after Bile Duct Injury Repair
927
Factors Related to Definitive Bile Duct Injury Repair and Perioperative Outcomes Total
Nonresponders (n ¼ 105)
Responders (n ¼ 62)
17 (4e114) 48 (29.1) 74 (25.5e283.5)
14 (4e105) 32 (31.1) 78 (16e273)
21 (5e124) 16 (25.8) 59 (43e289)
Factors
Days from injury to transfer, median (IQR) Attempted repair at initial transfer, n (%) Days from BDI to definitive repair, median (IQR) Definitive operation type, n (%) Hepaticojejunostomy Primary end-to-end ductal repair Choledochojejunostomy No. of postoperative stents (IQR) No. of postoperative drains (IQR) Postoperative morbidity, n (%) Postoperative mortality, 90-d, n (%) Length of stay after repair, median (IQR) Postoperative readmission, all cause, median (IQR) Length of postoperative stenting, d, median (IQR) Failed repair requiring reoperation or stent reinsertion, n
142 1 3 2 2 59 1 8 0 208
(86.1) (0.6) (1.8) (1e2) (1e3) (33.5) (0.7) (7e10) (0e1) (113e377) 5
86 1 3 1 2 35 1 8 0 204
(82.7) (1.0) (2.9) (1e2) (1e3) (20.0) (1.0) (7e10) (0e1) (98e377) 2
p Value
0.94 0.47 0.06 0.54
56 (91.8) 0 0 2 (1e2) 2 (2e3) 24 (38.7) 0 9 (7e11) 1 (0e2) 239 (149e380) 3
0.007 0.03 0.48 0.44 0.91 0.24 0.34 0.47
BDI, bile duct injury; IQR, interquartile range.
hardship due to their BDI. Although the majority of patients were able to return to work after their BDI (n ¼ 43 [84.3%]), 27.1% (n ¼ 16) of patients believed their job was in jeopardy at some point during their treatment course. Health-related quality of life assessment Patients were asked to assess their physical, mental, and general health both before BDI repair and currently. At the time of BDI, mental health appeared to be most affected, with nearly half of all patients reporting a depressed mood (49.2%) or low energy level (40.0%) “most” or “all the time” before BDI repair (Fig. 2). These Table 3.
symptoms improved significantly after repair, as only 18.3% reported having a depressed mood and 18.0% having low energy “most” or “all the time” currently (both p < 0.05). Interestingly, limitations in physical activity and general health remained unchanged before and after surgical repair (both p > 0.05). The majority of patients reported having good to excellent health both before BDI repair (61.7%) and currently (65.6%). Pain levels on a 5-point Likert scale (0 ¼ no pain to 4 ¼ severe pain) were assessed and categorized based on time after BDI repair. The mean pain score of all patients preoperatively was 2.1 (SD 1.6) with the majority of patients (59.0%) reporting moderate to extreme pain
Response to Bile Duct InjuryeSpecific Inquiries
Inquiries
Were you informed of risk of BDI before surgery? Did you seek litigation?* Did you receive a positive verdict?* Did your biliary tubes interfere with daily activities? Did your biliary tubes interfere with intimacy? Did your biliary tubes embarrass you? Was it difficult caring for your biliary tubes? Were you affected by not being able to work? Were you affected financially? Was your job put in jeopardy Were you affected by changes in your physical appearance? Were you able to return to work? *Not at all indicates “no”; very much indicates “yes.” BDI, bile duct injury.
Total responders, %
62 61 41 55 53 55 56 59 61 59 59 51
Not at all n %
45 18 12 3 2 9 8 9 10 30 10 8
72.6 29.5 29.3 5.5 3.8 16.4 14.3 15.3 16.4 50.9 17.0 15.7
Mild to moderate amount n %
4 d d 16 15 13 29 11 24 13 26 d
6.5 d d 29.1 28.3 23.6 51.8 18.6 39.3 22.0 44.1 d
Very much n %
13 43 29 36 36 33 19 39 27 16 23 43
21.0 70.5 70.7 65.5 67.9 60.0 33.9 66.1 44.3 27.1 39.0 84.3
928
Ejaz et al
J Am Coll Surg
Quality of Life after Bile Duct Injury Repair
Figure 2. Evaluation of health-related quality of life before bile duct injury repair and currently. *Without limitation, **most or all the time, ***statistically significant difference (p < 0.05) between groups.
(Table 4). Interestingly, patients with 10 years of follow-up after BDI repair, mean pain scores were significantly decreased (10 to 15 years: 1.0 [SD 0.9]; >15 years: 1.0 [SD 0.7]; p < 0.05). In addition, only 8% of patients reported having severe or very severe pain 10 to 15 years postoperatively, and none of the patients with >15 years of follow-up reported having severe or very severe pain (Fig. 3). Other commonly reported symptoms were also assessed based on the number of events occurring within the past 4 weeks of evaluation. Most patients were relatively pain-free, with the majority reporting having 6 days/week, 3.4%) followed by frequent/urgent bowel movements (1 to 5 days/week, 40.0%; >6 days/week, 8.3%). Table 4.
Factors associated with improved health-related quality of life Various demographic, clinicopathologic, and management factors were analyzed for their effect on HRQOL (Table 5). Clinicopathologic and management factors such as age, sex, race, type of injury, and length of follow-up were not associated with predicting HRQOL. In addition, HRQOL did not differ among patients seeking litigation or among those who reported winning their lawsuit.
DISCUSSION More than 30 million Americans are affected by gallstones, resulting in nearly 2 million ambulatory care visits each year.1 Since the introduction of LC in 1985, the management of benign gallbladder disease has been revolutionized. As with any radical change, the new technology was cautiously and carefully adopted. In time, LC proved to be an effective method to treat benign
Pain Scores Stratified by Time from Bile Duct Injury Repair
Pain level
0 (None), n (%) 1 (Mild), n (%) 2 (Moderate), n (%) 3 (Severe), n (%) 4 (Extreme), n (%) Mean (SD)
Preoperative
14 11 9 8 19 2.1
(23.0) (18.0) (14.8) (13.1) (31.2) (1.6)
15 Years postoperative
10 (45.5) 8 (36.4) 4 (18.2) 0 0 1.0 (0.7)
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
929
Table 5. Univariate Analysis of Factors Related to Improved Quality of Life Factors
Figure 3. Pain levels as stratified by time of follow-up after definitive surgical bile duct injury repair.
gallbladder disease, resulting in less postoperative pain and decreased hospital LOS as compared with traditional open cholecystectomy.2-4 However, as the adoption of LC became more widespread, the incidence of BDI after surgery was reported to be considerably higher, with estimates ranging from 2 to 4 times the risk as compared with traditional open cholecystectomy. Bile duct injury has been shown to cause considerable morbidity and mortality in patients as compared with those undergoing uncomplicated LC.13,15 In addition, the financial impact from the management of BDI is substantial, resulting in costs up to 26 times greater than an uncomplicated LC.13 Recent improvements in surgical technique and care for patients with BDI have improved immediate and long-term outcomes, such as morbidity, mortality, and readmissions.7 Of greater consequence, however, is the potential impact of BDI on HRQOL. Studies analyzing patientreported HRQOL are necessary to truly evaluate the effectiveness of the management of this difficult injury. Our group reported one of the first analysis of HRQOL in patients undergoing BDI repair.20 In this study, 54 patients were followed for a median of 59 months and were
Age Younger than 42 y 42 y or older Sex Male Female Race White Other Earlier repair at outside hospital Yes No No. of postoperative stents 1 2 3 Level of injury Low (Bismuth 1, 2) High (Bismuth 3, 4, 5) Days from surgery to stent removal 10 45. How often do you require follow-up of your bile duct injury with a healthcare professional? a. Weekly b. Monthly c. 2 to 4 times per year d. Yearly e. No longer require follow-up for my injury
932.e5
932.e6
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
932.e7
932.e8
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
Vol. 219, No. 5, November 2014
Ejaz et al
Quality of Life after Bile Duct Injury Repair
932.e9
932.e10
Ejaz et al
Quality of Life after Bile Duct Injury Repair
J Am Coll Surg
