Long-term Outcomes of Endoscopic vs Surgical Drainage of the Pancreatic Duct in Patients With Chronic Pancreatitis

Transcription

1 GASTROENTEROLOGY 2011;141: CLINICAL PANCREAS Long-term Outcomes of Endoscopic vs Surgical Drainage of the Pancreatic Duct in Patients With Chronic Pancreatitis DJUNA L. CAHEN,* DIRK J. GOUMA, PHILIPPE LARAMÉE, YUNG NIO, ERIK A. J. RAUWS, MARJA A. BOERMEESTER, OLIVIER R. BUSCH, PAUL FOCKENS, ERNST J. KUIPERS,* STEPHEN P. PEREIRA, # DAVID WONDERLING, MARCEL G. W. DIJKGRAAF,** and MARCO J. BRUNO* *Department of Gastroenterology and Hepatology, Erasmus Medical Center, Rotterdam, The Netherlands; Departments of Gastroenterology and Hepatology, Surgery, and Radiology and **Clinical Research Unit, Academic Medical Center, Amsterdam, The Netherlands; National Clinical Guideline Center, Royal College of Physicians, London, England; and # Institute of Hepatology, University College London, London, England BACKGROUND & AIMS: A randomized trial that compared endoscopic and surgical drainage of the pancreatic duct in patients with advanced chronic pancreatitis reported a significant benefit of surgery after a 2-year follow-up period. We evaluated the long-term outcome of these patients after 5 years. METHODS: Between 2000 and 2004, 39 symptomatic patients were randomly assigned to groups that underwent endoscopic drainage or operative pancreaticojejunostomy. In 2009, information was collected regarding pain, quality of life, morbidity, mortality, length of hospital stay, number of procedures undergone, changes in pancreatic function, and costs. Analysis was performed according to an intention-to-treat principle. RESULTS: During the 79-month follow-up period, one patient was lost and 7 died from unrelated causes. Of the patients treated by endoscopy, 68% required additional drainage compared with 5% in the surgery group (P.001). Hospital stay and costs were comparable, but overall, patients assigned to endoscopy underwent more procedures (median, 12 vs 4; P.001). Moreover, 47% of the patients in the endoscopy group eventually underwent surgery. Although the mean difference in Izbicki pain scores was no longer significant (39 vs 22; P.12), surgery was still superior in terms of pain relief (80% vs 38%; P.042). Levels of quality of life and pancreatic function were comparable. CONCLUSIONS: In the long term, symptomatic patients with advanced chronic pancreatitis who underwent surgery as the initial treatment for pancreatic duct obstruction had more relief from pain, with fewer procedures, than patients who were treated endoscopically. Importantly, almost half of the patients who were treated with endoscopy eventually underwent surgery. Keywords: Pancreas; Clinical Trial; Comparison of Therapy; Ductal Decompression. In chronic pancreatitis, ductal decompression is advocated in patients with pain and a dilated pancreatic duct. Endoscopic drainage involves sphincterotomy, extracorporeal shock wave lithotripsy, removal of stones, and dilatation of strictures by means of temporary stent insertion. 1 Surgical drainage is performed by a pancreaticojejunostomy according to Partington and Rochelle. 2 In 2007, we reported a prospective randomized trial that compared endoscopic and surgical drainage of the pancreatic duct after 2 years of follow-up. 3 The results revealed that surgical drainage was more effective in terms of pain relief, the state of physical health, and required procedures. The only other randomized trial on this subject by Dite et al is different from the present trial because surgery encompassed more than just a drainage procedure and endoscopic therapy did not include lithotripsy, which later became the cornerstone of endoscopic drainage. 4 Therefore, our study remains the only trial that genuinely compared the 2 drainage options. The publication led to much debate regarding the implications for clinical practice A frequently heard comment was that the follow-up period was not long enough to show the real benefit of endoscopic drainage. Furthermore, it was speculated that the effectiveness of a pancreaticojejunostomy would wane over time. As a result, many still prefer endoscopic treatment and consider surgery only as a secondline therapy because they want to avoid a major operation Therefore, we have compared the 2 treatment groups again, 5 years after the first analysis. Patients and Methods Baseline Data Between 2000 and 2004, 39 symptomatic patients underwent randomization; 19 patients were assigned to endoscopic treatment and 20 to surgery. As reported before, the demographic and clinical characteristics of the patients in the 2 treatment groups were similar with the exception of ongoing alcohol abuse, which was present in 5 surgically treated patients and in none of the endoscopically treated patients. Chronic pancreatitis was associated with complex pathologic features in the studied population (with a combination of strictures and stones in 79% of patients). In 2004, the safety committee ended Abbreviations used in this paper: SF-36, Medical Outcomes Study 36-Item Short-Form General Health Survey by the AGA Institute /$36.00 doi: /j.gastro

2 November 2011 ENDOSCOPIC VS SURGICAL DRAINAGE OF PANCREATIC DUCT 1691 the study after an unscheduled interim analysis on the basis of a significant difference in outcome favoring the surgical group (P.001). At that time, the median follow-up was 24 months (range, 6 24). Treatment During Follow-up During the follow-up period, all endoscopic and surgical procedures took place at the Academic Medical Center in Amsterdam. The choice of treatment was at the discretion of the treating physician. In accordance with the original study protocol, an endoscopic cholangiopancreatography was performed every 3 months during endoscopic treatment and stents were removed in case of stricture resolution. Collection of Follow-up Data The present analysis was performed 5 years after the termination of the initial trial. Data were collected by the study coordinator during a visit to the outpatient clinic at the end of The Izbicki pain score 16 and the quality-of-life scores from the Medical Outcomes Study 36-Item Short-Form General Health Survey (SF-36) questionnaire 17 and the health score profile from the EuroQol (EQ-5D) 18 were obtained through written questionnaires completed by the patients at home. In addition, blood and stool samples were taken to evaluate pancreatic function. Information regarding procedures performed and hospital stay was retrieved from hospital records and treating physicians. Information regarding the disease course of deceased patients was collected from hospital records and completed by a telephone interview with family members and/or the treating physician. Outcome Measures In accordance with the first study, the primary outcome measure was pain, expressed as the mean Izbicki pain score. The secondary outcome measures were pain relief, physical and mental health, health utility, morbidity, mortality, length of hospital stay, number of procedures performed, and changes in exocrine and endocrine pancreatic function. Pain relief was classified as complete (Izbicki pain score 10) or partial (Izbicki pain score 10 after a decrease of 50% compared with baseline). Patients were considered to have endocrine insufficiency if they required treatment for glycemic control. During follow-up, treatment was initiated when the fasting glucose level was 6.7 mmol/l (121 mg/dl) and the glycated hemoglobin level was more than 6%. Exocrine insufficiency was defined as an elastase level of 200 g/g of feces. The use of pancreatic enzymes was not considered to indicate pancreatic insufficiency, because they had also been prescribed as part of pain management. Changes from baseline in pancreatic function (both endocrine and exocrine) were evaluated by dividing the patients into 4 groups: those in whom insufficiency persisted, those in whom insufficiency developed, those in whom insufficiency resolved, and those in whom sufficiency persisted. Cost Description With the data of the first trial, a cost-effectiveness analysis was conducted from a provider s perspective as part of the development process of a clinical guideline for alcohol use disorders, which was funded by the UK National Institute for Health and Clinical Excellence ( CG100). This analysis concluded that surgery was highly costeffective compared with endoscopy. For the present study, a second cost analysis was performed to assess the long-term cost difference between the 2 treatment options. Costs regarding procedures performed, hospital stay, and pancreatic insufficiency were calculated by combining patient-level resource use data from the trial with standard UK unit costs. 19,20 For the pancreatic function, only the costs of exocrine insufficiency were included, because no difference in endocrine function was observed between the 2 groups throughout the follow-up period. The difference in exocrine function was not significant at the end of the follow-up period but was almost significant at 2 years (P.05). Therefore, the treatment costs of exocrine insufficiency were included in the cost description for the first 2 years only. Statistical Analysis Analysis was performed according to the intention-totreat principle. Depending on the distributional properties, outcome measures were expressed as means SD or as medians with ranges. To adjust for baseline scores, analysis of covariance was performed for the Izbicki pain score. Determination of pain relief and of changes in pancreatic function at the end of follow-up was based on the baseline data and the data from the final follow-up in Statistical significance was assessed with the use of Student t test for normally distributed continuous data (length of follow-up, SF-36, EQ-5D, fecal elastase); either the 2 test for categorical data (changes in endocrine function), with Yates correction when appropriate (pain relief), or Fisher exact test for categorical data (changes in exocrine pancreatic function); and the median test for nonnormally distributed continuous data (length of hospital stay, hospital readmittance, diagnostic and therapeutic procedures). All reported P values are 2 sided and were not adjusted for multiple testing. For the cost description, a probabilistic analysis was modeled using a Monte Carlo approach. 21 Probability distributions were applied to each parameter (gamma for unit costs, beta for probabilities associated with pancreatic insufficiency, and lognormal for the mean number of procedures), and 10,000 Monte Carlo simulations were computed, with all model parameters set simultaneously, each selected at random from their respective parameter distribution. Results Patient Survival and Follow-up For the present study, the mean follow-up time was 79 months (SD, 24). One patient was lost to follow-up 6 months after undergoing surgery and was excluded from the analysis (Figure 1). One early death was reported previously. This endoscopically treated patient died of a perforated duodenal ulcer 4 days after shock wave lithotripsy. During the follow-up period, another 6 patients died from causes unrelated to pancreatitis, 2 in the endoscopically treated group and 4 in the surgically treated group, a median of 45 months after treatment (range, months). From the remaining 31 patients, prospective data were collected (of whom 16 were endoscopically treated and 15 had undergone surgery) (Table 1). Treatment During the Last 5 Years of Follow-up In the second follow-up period, 5 patients underwent further endoscopic treatment and the total median

3 1692 CAHEN ET AL GASTROENTEROLOGY Vol. 141, No. 5 both occasions. However, stent therapy was terminated in one of the patients after this event and a pancreaticojejunostomy was performed. In the surgically treated group, one patient was operated on in the last 5 years. In this patient with hereditary pancreatitis, an uncomplicated tail resection and closure of the pancreatic stump with a pancreaticojejunostomy was performed 92 months after the first operation because of ongoing pancreatitis. Figure1. Study enrollment and long-term follow-up. stent duration increased from 27 weeks after 2 years to 40 weeks (range, weeks) at the end of the follow-up period. Complications occurred in 4 patients during the second follow-up period. In 2 patients the pancreatic stent became occluded and in 2 patients a rupture of the pancreatic duct occurred, which was treated conservatively on Treatment Outcome Of the patients who were primarily endoscopically drained, 6 (32%) did not require any further pancreatic duct drainage during follow-up (Figure 2). Three patients in the endoscopic group were operated on because the initial endoscopic treatment failed. In 9 patients (47%), a recurrent pancreatic duct obstruction developed a median of 7 months after treatment was terminated (range, 1 58 months). In 5 patients the reobstruction was caused by a combination of strictures and stones and in 4 patients by a recurrent stricture. All of these 9 patients were again treated endoscopically. Subsequently, in 7 of these cases a second recurrent stenosis developed and in 2 patients the obstruction recurred 3 times. Three of the 9 patients were successfully managed endoscopically, and 6 were eventu- Table 1. Outcomes of Endoscopic and Surgical Treatment of the 31 Patients Alive at Long-term Follow-up Variable Endoscopy (n 16) Surgery (n 15) Endoscopic results vs surgical results (95% confidence interval) Follow-up (mo), mean (SD) ( 16 to 2).13 Ongoing alcohol abuse, n (%) 0 3 (20) Izbicki pain score, mean (SD) a ( 5 to 39).12 Pain relief, n (%) b 6 (38) 12 (80) 43 ( 66 to 8) c.042 Complete/partial pain relief 4/2 (25/13) 8/4 (53/27) No relief 10 (62) 3 (20) SF-36 quality-of-life scores d Physical health component ( 12 to 3).23 Mental health component ( 9 to4).46 EQ-5D based health utility scores e UK values ( 0.20 to 0.14).75 Dutch values ( 0.17 to 0.15).92 Exocrine function, n (%).13 Insufficiency persisted 10 (63) 11(73) Insufficiency developed 6 (38) 2 (13) Insufficiency resolved 0 2 (13) Fecal elastase ( g/g) ( 79 to 38).48 Endocrine function, n (%).32 Insufficiency persisted 4 (25) 4 (27) Insufficiency developed 7 (44) 3 (20) Sufficiency persisted 5 (31) 8 (53) P value a The Izbicki pain score ranges from 0 to 100, with higher scores indicating more severe pain. b Pain relief at the end of follow-up was classified as complete (Izbicki pain score 10) or partial (Izbicki pain score 10 after a decrease of 50%). c This is the difference between the percentages. d The scores on the SF-36 physical and mental health components range from 0 to 100, with higher scores indicating better quality of life. Linear transformations were performed to standardize the scores to a mean score of in a general Dutch population. e The utilities of the observed health score profiles on the EQ-5D are based on the time trade-off elicitation technique from interviews with adults from the UK general population (Dolan P. Modeling valuations for EuroQol health states. Med Care 1997;35: ) and the Dutch general population (Lamers LM, Stalmeier PF, McDonnell J, et al. [Measuring the quality of life in economic evaluations: the Dutch EQ-5D tariff]. Ned Tijdschr Geneeskd 2005;149: ), respectively. Utilities range from either (UK) or (Dutch), indicating serious health problems, to unity, indicating no problems at all.

4 November 2011 ENDOSCOPIC VS SURGICAL DRAINAGE OF PANCREATIC DUCT 1693 patient were $6006 higher in the endoscopic group, but this difference was not significant (95% confidence interval, $16,188 to $27,786; P.29). Figure 2. Treatment outcome. ally operated on. Overall, 9 patients (47%) had treatment converted to surgery (a pancreaticojejunostomy in 8 and a Beger procedure in one) a median of 24 months after randomization (range, 6 74 months). Seven of these 9 patients were still alive. Interestingly, only 2 of these patients had complete relief of pain after surgery. In the surgically treated group, none of the patients developed a recurrent pancreatic duct obstruction. Outcomes The primary and secondary outcomes are summarized in Table 1 (alive patients) and Table 2 (all patients). Before, the first analysis revealed a significant difference in the mean Izbicki pain score during the first 2 years of 24 points. In this second analysis, the pain score difference in favor of the surgically treated group was no longer significant (39 vs 22) (Table 1). After adjustment for baseline scores, the mean difference was 17 (95% confidence interval, 5 to39;p.12). However, the secondary outcome measure of pain relief, either complete or partial, was still significantly higher in the surgically treated group (80% vs 38%; P.042). In the first 2 years of follow-up, the physical health component of the SF-36 questionnaire was lower in endoscopically treated patients than in surgically treated patients. Five years later, no difference in quality-of-life (SF-36) and health utility (EQ-5D based) scores was observed. In 2009, almost all patients were exocrine insufficient in both groups. Endocrine insufficiency was observed less frequently with a trend toward preservation of this function after surgical treatment. The 2 treatment groups did not differ significantly with respect to length of hospital stay, but endoscopically treated patients underwent significantly more procedures than surgically treated patients (median, 12 vs 4; P.001). Cost differences are shown in Table 3. Results were in favor of surgery for diagnostic procedures and the treatment of pancreatic insufficiency. The overall costs per Discussion These long-term results indicate that in patients with advanced calcifying chronic pancreatitis and a symptomatic pancreatic duct obstruction, surgery is more effective than endoscopic treatment. An operative pancreaticojejunostomy at the time of initial presentation was superior to lithotripsy and stenting, not only in terms of pain relief, but also in terms of required reinterventions, without being associated with higher costs. In the first 2 years after randomization, the benefits of surgery were shown by a more rapid, effective, and sustained pain relief. This difference in pain relief was still present after a long-term follow-up period of 7 years. The absolute Izbicki pain score difference between the 2 groups was no longer significant. However, in contrast to the first analysis, which used multiple measurements during the 2-year follow-up, the second analysis was based on a single measurement. Therefore, the smaller sample size of the reduced 31 surviving patients lacks the power to prove a difference in pain score, although a trend toward a benefit of surgery was still observed. In both groups, a slight improvement in the Izbicki score was noticed after the second follow-up period, which may also be attributed to the natural course of the disease. 22 Not only was surgery more effective in terms of pain relief, it also required fewer procedures. In fact, for most of the surgically treated patients, a single operation accomplished immediate and permanent pain relief. This is in strong contrast to the endoscopically treated group, who generally had to undergo multiple procedures. One might debate that regardless of these long-term data, the Table 2. Procedures and Hospital Stay of All 38 Patients Who Underwent Endoscopic and Surgical Treatment Variable Endoscopy Surgery P value Total follow-up period Conversion to surgery, n (%) 9 (47) NA NA Hospital stay, median no. of 13 (2 237) 11 (5 345).33 days (range) Hospital readmittance, median 2 (0 9) 0 (0 29).194 no. (range) Procedures, median no. (range) 12 (1 59) 4 (1 25).001 Diagnostic 6 (0 40) 3 (0 17) Therapeutic 8 (1 21) 1 (1 14) Last 5 years of follow-up Conversion to surgery, n (%) 5 (26) NA NA Hospital stay, median no. of 0 (0 119) 0 (0 286).32 days (range) Hospital readmittance, median 0 (0 5) 0 (0 22).32 no. (range) Procedures, median no. (range) 2 (0 43) 0 (0 20).51 Diagnostic 1 (0 35) 0 (0 14).74 Therapeutic 1 (0 13) 0 (1 13).19 NA, not applicable.

5 1694 CAHEN ET AL GASTROENTEROLOGY Vol. 141, No. 5 Table 3. Long-term Cost Analysis of Endoscopic and Surgical Treatment Cost category a Endoscopy Surgery Cost difference (95% confidence interval) Diagnostic procedures (38 to 1749).043 Therapeutic procedures and hospital stay 28,327 23, ( 16,895 to 26,963).32 Pancreatic insufficiency (31 to 328).036 Total cost 31,048 25, ( 16,188 to 27,786).29 P value a Costs are presented in US dollars and were converted from pound sterling using the 2009 Purchasing Power Parity ($1 0.62) SNA_TABLE4. previous results, which revealed the substantial burden of endoscopic treatment during the first 2 years of follow-up, already justified the choice for surgery. In addition, it now becomes apparent that in almost half of the patients, endoscopic drainage fails altogether and treatment was converted to surgery. From these observations, we conclude that the general perception that endoscopic treatment is less invasive than surgery cannot be sustained. This is in accordance with a recent report by Rutter et al, who retrospectively evaluated the long-term outcome of 292 patients with chronic pancreatitis and showed that patients who undergo surgery require fewer subsequent interventions, require a shorter hospital stay, and have a better quality of life compared with endoscopic or conservative treatment. 23 Furthermore, the results of the cost description emphasize that surgery is also a good choice from an economic perspective. Of the endoscopically treated patients, almost half converted to surgery. Interestingly, salvage surgery was not very effective. Although the numbers are small, this observation raises the hypothesis that postponing surgery has a negative influence on treatment outcome. Because chronic pancreatitis is an ongoing inflammatory process, resulting in irreversible damage to the pancreas, it may have been attenuated by a more aggressive and effective drainage approach in the beginning of the disease course. Therefore, in the analysis, these converted patients were allocated to the endoscopic group on an intention-totreat basis and not to the surgically treated group in a per-protocol analysis. This study is unique because it is one of the few prospective trials on pancreatic duct drainage and provides the longest follow-up to date. The first study by Dite et al was in favor of surgical drainage but did not include lithotripsy and therefore did not meet the current standards of endoscopic treatment. 4 The only other prospective randomized study was published by Dumonceau et al, who compared conventional endoscopic treatment, consisting of shock wave lithotripsy combined with endotherapy, with lithotripsy alone. The patient population of this study was different from the present trial in that only 56% had pain at baseline. This might explain the better outcome, because pain relief was observed in 56% of the patients. 24 Surprisingly, the least invasive form of treatment, lithotripsy without endotherapy, was associated with the best treatment response (61% vs 54%). This is even more interesting, given that our own endoscopic treatment protocol was previously criticized for not being aggressive enough. 7,10 The question remains if there is a role for endoscopic treatment of pancreatic duct obstructions in chronic pancreatitis. First, further development of endoscopic techniques might improve results. To date, studies that focus on the technical aspects of endoscopic treatment are rare and conflicting. On the one hand, some investigators advocate a more aggressive approach; Costamagna et al have reported promising results of cumulative stenting with a success rate of 84%. 25 Others are investigating less extensive techniques, such as the group from Brussels that performed lithotripsy without endoscopic stenting. 24 Future prospective studies are needed to solve these issues. In addition, there is still an open debate regarding the choice of the surgical procedure. In this study, surgery consisted of a drainage procedure only, because patients with an enlarged pancreatic head were excluded. However, many believe that combining drainage with a partial resection of the pancreatic head (Beger or Frey procedure) offers even better pain relief, but this requires further investigation. 14 We believe that proper patient selection is of vital importance for the final outcome of endoscopic treatment. Now that endoscopic drainage seems inferior to surgery in symptomatic patients with complex pathology, the interest shifts to patients with less extensive disease. There is evidence that in patients with a single obstruction or stone, the course of the disease is favorably altered by an early intervention. Farnbacher et al found that the only parameter predictive of long-term pain relief after endoscopic pancreatic duct drainage was a short duration of disease. 26 Also, animal studies have shown that pancreatic insufficiency develops early in the course of obstructive pancreatitis and becomes permanent within several weeks. 27 Therefore, patients with minor or even without symptoms may benefit most from endoscopic drainage, but this should evidently be evaluated in future prospective studies before it can be applied in clinical practice. The results of this study confirm that initial surgical drainage of the pancreatic duct is superior to endoscopic treatment in symptomatic patients with advanced chronic pancreatitis, not only based on short-term outcomes but also in the long-term. These benefits apply for pain relief and the need for reinterventions. This reconfirms our original recommendation that surgery is the preferred treatment for this specific group of patients.

6 November 2011 ENDOSCOPIC VS SURGICAL DRAINAGE OF PANCREATIC DUCT 1695 References 1. Nguyen-Tang T, Dumonceau JM. Endoscopic treatment in chronic pancreatitis, timing, duration and type of intervention. Best Pract Res Clin Gastroenterol 2010;24: Partington PF, Rochelle RE. Modified Puestow procedure for retrograde drainage of the pancreatic duct. Ann Surg 1960;152: Cahen DL, Gouma DJ, Nio Y, et al. Endoscopic versus surgical drainage of the pancreatic duct in chronic pancreatitis. N Engl J Med 2007;356: Dite P, Ruzicka M, Zboril V, et al. A prospective, randomized trial comparing endoscopic and surgical therapy for chronic pancreatitis. Endoscopy 2003;35: Kleeff J, Friess H, Buchler MW. Endoscopic versus surgical treatment for chronic pancreatitis. N Engl J Med 2007;356:2103; author reply Elta GH. Is there a role for the endoscopic treatment of pain from chronic pancreatitis? N Engl J Med 2007;356: Wong T. Endoscopic versus surgical treatment for chronic pancreatitis. N Engl J Med 2007;356:2101; author reply Dumonceau JM. Endoscopic versus surgical treatment for chronic pancreatitis. N Engl J Med 2007;356:2102; author reply Siriwardena AK. Endoscopic versus surgical treatment for chronic pancreatitis. N Engl J Med 2007;356: ; author reply Delhaye M, Deviere J. Endoscopic versus surgical treatment for chronic pancreatitis. N Engl J Med 2007;356: ; author reply Kozarek RA. Pancreatic endoscopy. Endoscopy 2008;40: Tringali A, Boskoski I, Costamagna G. The role of endoscopy in the therapy of chronic pancreatitis. Best Pract Res Clin Gastroenterol 2008;22: Attasaranya S, Abdel Aziz AM, Lehman GA. Endoscopic management of acute and chronic pancreatitis. Surg Clin North Am 2007; 87: , viii. 14. Bachmann K, Kutup A, Mann O, et al. Surgical treatment in chronic pancreatitis timing and type of procedure. Best Pract Res Clin Gastroenterol 2010;24: Hirota M, Asakura T, Kanno A, et al. Endoscopic treatment for chronic pancreatitis: indications, technique, results. J Hepatobiliary Pancreat Sci 2010;17: Bloechle C, Izbicki JR, Knoefel WT, et al. Quality of life in chronic pancreatitis results after duodenum-preserving resection of the head of the pancreas. Pancreas 1995;11: Ware JE, Jr. SF-36 health survey update. Spine 2000;25: Rabin R, de Charro F. EQ-5D: a measure of health status from the EuroQol Group. Ann Med 2001;33: National Health Service reference costs Department of Health, United Kingdom, British Medical Association and Royal Pharmaceutical Society of Great Britain. British National Formulary. 59 th ed. London, England: UK BMJ Group and RPS Publishing, Doubilet P, Begg CB, Weinstein MC, et al. Probabilistic sensitivity analysis using Monte Carlo simulation. A practical approach. Med Decis Making 1985;5: Ammann RW. The natural history of alcoholic chronic pancreatitis. Intern Med 2001;40: Rutter K, Ferlitsch A, Sautner T, et al. Hospitalization, frequency of interventions, and quality of life after endoscopic, surgical, or conservative treatment in patients with chronic pancreatitis. World J Surg 2010;34: Dumonceau JM, Costamagna G, Tringali A, et al. Treatment for painful calcified chronic pancreatitis: extracorporeal shock wave lithotripsy versus endoscopic treatment: a randomised controlled trial. Gut 2007;56: Costamagna G, Bulajic M, Tringali A, et al. Multiple stenting of refractory pancreatic duct strictures in severe chronic pancreatitis: long-term results. Endoscopy 2006;38: Farnbacher MJ, Schoen C, Rabenstein T, et al. Pancreatic duct stones in chronic pancreatitis: criteria for treatment intensity and success. Gastrointest Endosc 2002;56: Lamme B, Boermeester MA, Straatsburg IH, et al. Early versus late surgical drainage for obstructive pancreatitis in an experimental model. Br J Surg 2007;94: OECD Stats Extracts. Purchasing Power Parities (PPPs) for OECD countries since Paris, France: Organisation for Economic Co-operation and Development, Received February 18, Accepted July 18, Reprint requests Address requests for reprints to: Djuna L. Cahen, MD, PhD, Department of Gastroenterology and Hepatology, Erasmus Medical Center, PO Box 2040, 3000 CA, Rotterdam, The Netherlands. d.cahen@erasmusmc.nl; fax: (31) Acknowledgments Current Controlled Trials number ISRCTN Conflicts of interest The authors disclose no conflicts.