Is early endoscopic retrograde cholangiopancreatography useful in the management of acute biliary pancreatitis?

Available online at www.sciencedirect.com

Digestive and Liver Disease 40 (2008) 379–385

Digestive Endoscopy

Is early endoscopic retrograde cholangiopancreatography useful in the management of acute biliary pancreatitis? A meta-analysis of randomized controlled trials A. Moretti, C. Papi ∗ , A. Aratari, V. Festa, M. Tanga, M. Koch, L. Capurso Gastroenterology Unit “San Filippo Neri” Hospital, Rome, Italy Received 23 August 2007; accepted 3 December 2007 Available online 19 February 2008

Abstract Aims. To compare early endoscopic retrograde cholangiopancreatography with conservative management for the treatment of acute biliary pancreatitis: a meta-analysis of prospective randomized trials. Method. Pertinent studies were selected from the Medline, Embase, and the Cochrane Library Databases, references from published articles and reviews. Conventional meta-analysis according to DerSimonian and Laird method was used for the pooling of the results. The rate difference (95% CI) and the number needed to treat were used as a measure of the therapeutic effect. Results. Five prospective randomized trials including 702 patients were selected. Overall complications and mortality rates were 31% and 6%, respectively. In predicted severe pancreatitis the pooled rate difference for complications in early endoscopic retrograde cholangiopancreatography was 38.5% (95% CI −53% to −23.9%); p < 0.0001; number needed to treat = 3. In predicted mild pancreatitis the pooled rate difference for complications in early endoscopic retrograde cholangiopancreatography was 1.8% (95% CI −5.6% to 9.3%); p = 0.6. No mortality was observed in predicted mild pancreatitis. In predicted severe pancreatitis the pooled rate difference for mortality in the early endoscopic retrograde cholangiopancreatography group was 4.3% (95% CI −16% to 7.5%); p < 0.24. Conclusions. Early endoscopic retrograde cholangiopancreatography reduces pancreatits-related complications in patients with predicted severe pancreatitis although mortality rate is not affected. In predicted mild pancreatitis early endoscopic retrograde cholangiopancreatography has no advantage compared to conservative management. © 2007 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. Keywords: Acute pancreatitis; Acute biliary pancreatitis; ERCP; Meta-analysis

1. Introduction Gallstones are the commonest cause of acute pancreatitis in western societies, accounting for 30–60% of all cases [1–4]. The clinical course of acute biliary pancreatitis (ABP) is similar to that of pancreatitis of other causes: approximately 80% of patients have a mild and self-limiting disease course, while 20% of patients will develop a severe form with an associated mortality as high as 30% [5–8]. Gallstones passing the papilla of Vater are considered the initial step for ABP [9] but ∗ Corresponding author at: UOC Gastroenterologia, A.C.O. San Filippo Neri, Via G. Martinotti 20, 00135 Rome, Italy. Tel.: +39 06 33062245; fax: +39 06 33062641. E-mail address: [email protected] (C. Papi).

the exact mechanism by which passage of gallstone through the ampulla of Vater initiates inflammation of the pancreas remains unclear [10]. Indeed the pathogenesis of ABP may be multifactorial [11–13], but clinical and experimental studies have suggested that impacted ampullary stones, and persistent biliopancreatic obstruction, are associated with a more severe disease course [11,14,15]. These observations provide the rationale for early removal of impacted stones in patients with ABP. In fact, at the beginning of the eighties, early surgery and biliary decompression via a transduodenal sphincteroplasty had been advocated as the treatment of choice in patients with ABP [16]. However, subsequent observational studies and clinical trials showed that early surgery resulted in a significant increase in morbidity and mortality [17,18].

1590-8658/$32 © 2007 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2007.12.001

380

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385

Since the late 1980s, early endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic sphincterotomy (ES) have been advocated in the management of ABP [19]. However, results of randomized controlled trials (RCTs) have been conflicting. Two RCTs comparing early ERCP versus conservative management in ABP have shown a benefit of early ERCP for reducing complication rate in patients with severe pancreatitis [20,21], however, another study failed to show any benefit [22]. Two meta-analyses concluded that early ERCP significantly reduces complication rate but not mortality only in severe ABP [23,24]. However, these meta-analyses include only three RCTs and another study published as an abstract [25]. We have performed a metaanalysis of RCTs addressing early ERCP in ABP updated to February 2007.

2. Methods We did a literature search using the main databases (MEDLINE 1966–2001, EMBASE 1980–2001, and the Cochrane Library). A computer-assisted search was conduced using the following combination of medical subject heading terms (MESH): “ERCP” or “sphincterotomy” and “acute disease” and “biliary pancreatitis”. We also did a manual search using references from the articles retrieved and main review articles and guidelines [1,26–29]. Prospective RCTs of early ERCP in ABP were selected. Studies were judged suitable for meta-analysis only if they met all the following criteria: (1) prospective RCTs of early ERCP versus conservative management in ABP; (2) well defined outcomes including at least one of the following: (a) local complications (b) systemic complications; (c) mortality. Morbidity and mortality rates were considered as the number of patients suffering at least one complication or death within 90 days after admission. Only results fully reported in journal articles were considered. All articles passed through a multilevel, systematic review by a team of four physicians (AM, CP, MT, AA) and methodological criteria and the results of each study were recorded. Studies that fulfilled the inclusion criteria were reviewed blindly and independently by the same four authors to tabulate subject demographics, study design, definition of primary and secondary outcomes, and frequencies of each end-point, using a standardized data abstract form. Disagreement was resolved by consensus. Each study was evaluated for quality: we used a modified Jadad scale [30] with total score ranging from 1 to 3. We did not contact study authors for additional data.

3. Statistical analysis Results were analysed by the DerSimonian–Laird method [31] for comparing and summarizing outcomes of individual RCTs. The term rate difference (RD), i.e. the difference in event rates between the treatment and control groups, was

used as a measure of the therapeutic effect. Confidence intervals (CI) were always calculated at 95%. Number needed to treat (NNT), that is the number of patients who must be treated in order to obtain one therapeutic effect, was also calculated [32]: mathematically, NNT is equivalent to the reciprocal of RD and the 95% confidence intervals for the NNT are the reciprocal of the 95% confidence intervals for RD. The alpha level was set at 0.05, for a two-tailed test. A statistical program published by T. Chalmers and us was used for this purpose [33]. Results were also checked using Epistat (copyright © Epistat Services, 1991), StatsDirect statistical tools (Copyright © 1990–2001) and an appropriate meta-analysis software [34]. Intertrial heterogeneity in treatment effect was evaluated using the Q statistic of DerSimonian–Laird [31]. To compensate somewhat for the lack of the power of the test, we decided not to accept evidence of a therapeutic effect, even if statistically significant, in the event of a p value < 0.10 for heterogeneity. Furthermore, to detect heterogeneity, a visual display was obtained, representing the results on a L’Abb`e plot [35]. Publication bias has been calculated according to Rosenthal’s method [36].

4. Results Eight trials, published between 1986 and 2007, met the inclusion criteria [20–22,37–41]. Three of these eight studies were excluded: two were preliminary reports [37,38] and the definitive results were published later by the same group of authors [20,22]; one trial was excluded because the study design allowed a change in the allocation arm after randomization according to the clinical evolution of biliary obstruction [39]. This review is, therefore, based upon the results from five studies [20–22,40,41]. Studies were reviewed blindly and independently by a team of four physicians (AM, CP, MT, AA) using a standardized data abstract form. Upon initial analysis, concordance implied 95% agreement between the authors. A total of 702 patients had been enrolled: 353 were randomized to early ERCP and 349 were randomized to conservative management and eventually elective ERCP if appropriate. In all studies the diagnosis of ABP was based according to clinical and biochemical data, and sonographic or CT scan findings. Assessment of severity was made using different clinical scores: Glasgow’s Score [20,22], Ranson’s Score [21] and APACHE II Score [40,41]. Of 702 patients, 232 (33%) met the criteria for severe pancreatitis and 470 (67%) had predicted mild pancreatitis. Early ERCP was performed within 72 h from admission in three studies [20,22,41] and within 24 h in two studies [21,40]. Success rate of ERCP ranged from 88% to 96%; overall ERCP and ES-related complication rate was 3.4%: most of these were self-limited post-procedure bleeding. Conservative management consisted of fluid replacement, antibiotics, analgesics, octreotide, O2 , nasogastric suction and parenteral

72 h from admission 103 2007 Oria [41]

2

45 2002 Zhou [40]

3

238 1997 Folsh [22]

3

195 1993 Fan [21]

AP: acute pancreatitis; ABP: acute biliary pancreatitis; AB: antibiotics; TPH: total parenteral nutrition.

APACHE II >6 AP unrelated to gallstones Acute cholangitis comorbidity

24 h from admission

72 h from symptoms

Fluid replacement AB, 02 TPN Onset Fluid replacement AB, octreotide Fluid replacement AB, 02 APACHE II >8

Ranson >3

AP unrelated to gallstones Acute cholangitis Acute cholecystitis cystitis Acute cholangitis AP unrelated to gallstones

Glasgow >2

Supportive

Fluid replacement

72 h from admission AB, 02 24 h from admission Glasgow >2 AP unrelated to gallstones 1988 Neoptolemos [20]

3

Patients no. Quality score Year

A further analysis was performed in patients with predicted mild and severe ABP according to clinical scores (Table 1). Pancreatitis was predicted as mild in 470 of

Author ref.

4.2. Mild versus Severe acute biliary pancreatitis

Table 1 Characteristics of randomized trial

Two primary end-points have been considered: complication rate (including local and systemic complications) and mortality rates. Overall complications and mortality rates were 31% and 6%, respectively. Results of primary outcomes measures through the five studies are reported in Table 2. Ninety-seven of 353 patients in the early ERCP group (27%) suffered at least one complication compared to 121 of 349 patients in the conservative group (36%). The pooled RD for complication rate in the early ERCP group was 8.7% (95% CI −15.8% to −1.5%); p = 0.01; NNT = 12. No heterogeneity was found (χ2 = 5.03, d.f. = 4, p = 0.28) (Fig. 1). Twenty-three of 353 patients in the early ERCP group (6%) died within 3 months from admission compared to 22 of 346 patients in the conservative group (6%). The pooled RD for mortality rate in early ERCP group was 0.2% (95% CI 4.6% to 4.2%); p = 0.9. No heterogeneity was found (χ2 = 6.8, d.f. = 4, p = 0.14) (Fig. 2).

Diagnostic criteria for ABP

4.1. Main outcomes

381

Typical clinical/biochemical data US/CT scan findings Typical clinical/biochemical data Typical clinical/biochemical data US/CT scan findings Typical clinical/biochemical data US/CT scan findings Typical clinical/biochemical data US/CT scan findings

Exclusion criteria

nutrition if clinically indicated. In the conservative management group elective ERCP was performed after 72 h from admission if clinically indicated. Patients with no gallstonerelated pancreatitis were excluded in all studies; patients with acute biliary complications such as acute cholecystitis or cholangitis, and patients with relevant comorbidity were also excluded. The characteristics of studies are shown in Table 1.

121

Fig. 1. Rate differences (RD) (95% CI) for complication rate in prospective randomized trials addressing early ERCP vs. conservative management in acute biliary pancreatitis. The pooled RD for complications in early ERCP is 8.7% (95% CI −15.8% to −1.5%); p = 0.01; NNT = 12.

2

Severity assessment

Timing ERCP

Conservative management

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385

23/353 (6%)

22/346 (6%)

97/353 (27%)

121/349 (36%)

ERCP n = 353 Control n = 34$ Control n = 349 45/702 (6%) 232/702 (33%) 702

218/702 (31%)

4/103 (4%) 38/103 (37%) 103

20/103 (19%)

0/45 (0%) 14/45 (31%) 45

6/45 (13%)

21/238 (9%) 46/238 (19%) 238

115/238 (48%)

81/195 (42%) 195

46/195 (24%)

14/195 (7%)

10/59 (17%) 21/62 (34%) 17/97 (18%) 29/98 (30%) 58/126 (46%) 57/112 (51%) 1/20 (5%) 5/25 (20%) 11/51 (22%) 9/52 (17%) ERCP n = 59 Control n = 62 ERCP n = 97 Control n = 98 ERCP n = 126 Control n = 112 ERCP n = 20 Control n = 25 ERCP n = 51 Control n = 52 6/121(5%) 53/121 (44%) 121

31/121 (26%)

Mortality Complications Randomization Mortality Complications Severe AP Patients

Fig. 2. Rate differences (RD) (95% CI) for mortality rate in prospective randomized trials addressing early ERCP vs. conservative management in acute biliary pancreatitis. The pooled RD for mortality in early ERCP is 0.2% (95% CI −4.6% to 4.2%); p = 0.9.

702 patients (67%) and severe in 232 of 702 patients (33%) (Table 2). Separate data concerning complication rate in mild and severe pancreatitis have been reported in three studies [20,21,40]; data concerning mortality in mild and severe pancreatitis have been reported in four studies [20,21,40,41]. In predicted severe pancreatitis, overall complication rate was 41%; conversely, in predicted mild pancreatitis, overall complication rate was 10% (p < 0.0001). Mortality rate was 13% in predicted severe pancreatitis while no mortality was observed in mild pancreatitis (p < 0.0001). In mild pancreatitis, the pooled RD for complications in early ERCP group was 1.8% (95% CI −5.6% to 9.3%); p = 0.6. No heterogeneity was found (χ2 = 0.2, d.f. = 2, p = 0.9). In severe pancreatitis the pooled RD for complications in early ERCP group was 38.5% (95% CI −53% to −23.9%); p < 0.0001; NNT = 3. No heterogeneity was found (χ2 = 0.85, d.f. = 2, p = 0.6) (Fig. 3). The pooled RD for mortality in early ERCP group was 4.3% (95% CI −16% to 7.5%); p < 0.24. No heterogeneity was found (χ2 = 4.73, d.f. = 3, p = 0.2) (Fig. 4).

1988

1993

1997

2002

2007

Neoptolemos [20]

Fan [21]

Folsh [22]

Zhou [40]

Oria [41]

Total

Year

4.3. Length of hospital stay

Author ref.

Table 2 Primary end points in randomized controlled trials

1/59 (2%) 5/62 (8%) 5/97 (5%) 9/98 (9%) 14/126 (11%) 7/112 (6%) 0/20 (0%) 0/25 (0%) 3/51(6%) 1/52 (2%)

382

Only two studies reported data concerning hospitalization in patients randomized to early ERCP or conservative management [20,40]. No difference in length of hospital stay was observed in patients with predicted mild pancreatitis. In patients with predicted severe pancreatitis hospitalization was significantly shorter in the early ERCP group: 9.5 days (range 6–36) versus 17.0 days (range 4–74); p < 0.03 in the study of Neoptolemos et al. [20], and 15.4 ± 3 days versus 63 ± 4.8 days; p < 0.05 in the study of Zhou et al. [40].

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385

Fig. 3. Rate differences (RD) (95% CI) for complication rate in prospective randomized trials addressing early ERCP vs. conservative management in acute biliary pancreatitis predicted as mild or severe according to clinical scores. In mild pancreatitis the pooled RD for complications in early ERCP is 1.8% (95% CI −5.6% to 9.3%); p = 0.6. In severe pancreatitis the pooled RD for complications in early ERCP group is 38.5% (95% CI −53% to −23.9%); p < 0.0001; NNT = 3.

Fig. 4. Rate differences (RD) (95% CI) for mortality rate in prospective randomized trials addressing early ERCP vs. conservative management in severe acute biliary pancreatitis. The pooled RD for mortality in early ERCP is 4.3% (95% CI −16% to 7.5%); p < 0.2.

4.4. Publication bias The number of null studies needed to lead our metaanalysis result to level of statistical non-significance is 3.

5. Discussion Gallstone disease has been recognized as one of the commonest cause of acute pancreatitis in western societies

383

[1–4]. Although the exact pathogenesis of ABP remains unclear [10,42] it has been postulated that small migrating stones can initiate an attack of ABP, and repeated passage of small stones, or impacted stones in the ampulla of Vater, can contribute to the progression of the disease towards a more severe course [43]. This “two phases theory” is supported by the observation that about 70% of patients with ABP who undergo biliary surgery or ERCP during an acute attack have stones in the common bile duct [44] and that impacted ampullary stones, and persistent biliopancreatic obstruction, are associated with a more severe disease course [11,14]. These observations provide the rationale for early biliary decompression and stone removal in patients with ABP. Between the end of the seventies and the beginning of the eighties urgent surgery with biliary decompression was proposed as the treatment of choice in patients with ABP [16]. However this practice was associated with an increased mortality [17,18] and, therefore, was no longer performed. Since the late 1980s, endoscopic interventions have widely replaced surgical bile duct exploration and have been increasingly included in the management of ABP. Although ERCP and ES allows effective removal of stones and biliary drainage with a success rate of approximately 90% in expert hands [20–24], injection of contrast medium into the pancreatic duct carries the risk to worsen pancreatitis [28,45]. So the question is: which patients can benefit from early ERCP and ES and which is the optimal time of endoscopic interventions in ABP? Results of RCTs comparing early ERCP versus conservative management in the management of ABP have been conflicting [20–22]. A meta-analysis published in 1999, showed a significant reduction in complication rate in patients treated with early ERCP compared to patients treated conservatively: the absolute risk reduction (ARR) was 13.2% (95% CI 6.9% to 19.5%) with NNT = 7.6. A slight but nonsignificant reduction in mortality rate was also observed (ARR = 3.9%, 95% CI 0.35% to 7.45%; NNT = 25.6) [23]. However, this meta-analysis included only three RCTs published as full papers and one abstract: a subgroup analysis based on the severity of ABP was not performed, although the authors concluded that it might be assumed that the greatest benefit from early ERCP would be found in patients with severe ABP. In a subsequent meta-analysis published in 2004, only three RCTs published as full papers have been included. A significant reduction of complication rate (OR = 0.27, 95% CI 0.14 to 0.53), but not mortality, was reported in patients with severe ABP treated with early ERCP [24]. We have performed a meta-analysis of RCTs addressing early ERCP versus conservative management in patients with ABP updated to February 2007. Overall five RCTs, all published as full-papers, have been included and 702 patients enrolled. Moreover, a subgroup analysis according to the severity of ABP has been performed. In the whole patient population, we have observed a significant reduc-

384

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385

tion of complication rate in the early ERCP group (RD −8.7%; 95% CI −15.8% to −1.5%, p = 0.01, NNT 12), but no difference in mortality has been observed (RD 0.2%; 95% CI 4.6% to 4.2% p = 0.9). The benefit of early ERCP for reducing complication rate is probably due to the high proportion of severe ABP in the whole patient population (33%). In fact, subgroup analysis, according to predicted severity of ABP, showed no benefit of early ERCP for reducing complication rate in mild pancreatitis but a reduction of complication rate of approximately 40% in severe ABP (RD −38.5%, 95% CI −53% to −23.9%; p < 0.0001; NNT = 3). In other words, if an ABP is predicted as severe, only three patients would need to be treated with early ERCP to prevent complications in one patient. Conversely, if the course of pancreatitis is predicted as mild, ERCP performed within 72 h from admission has no advantage compared to conservative management (and eventually elective ERCP if clinically indicated) for preventing local or systemic complications. In all studies, severity stratification of ABP has been performed at admission by means of clinical scores such as Ranson’s and Glasgow’ s criteria or APACHE II score. There are several drawbacks in predicting severity according to these scores. The ability to accurately predict which patient will have mild or severe pancreatitis is poor and the predictive value of tools such as Ranson’s and Glasgow’s criteria is limited by a time delay of at least 48 h from symptom onset [45–47]. The positive predictive value (PPV) of clinical scores is relatively low (50–60%) despite a high negative predictive value (NPV) (90%) [48,49]. This means that clinical scores overestimate a severe course; in fact, only 60% of patients predicted as having a severe course will turn out to be severe, but the clinical scores will miss 10% of severe cases (NPV 90%). In studies included in the present meta-analysis, complication and mortality rates are significantly higher in patients predicted as having severe ABP (41% versus 10%, p < 0.0001; and 13% versus 0%, p < 0.0001, respectively). This suggests that Ranson, Glasgow and Apache II score can identify a subgroup of patients with a more severe course. However, a complication rate of 10% in patients with predicted mild pancreatitis, suggests that these patients had a severe course despite they had been considered to have mild ABP, according to the NPV of 90% reported in the literature [48,49]. In conclusion, patients predicted as having severe ABP can benefit from early ERCP: a significant reduction of pancreatitis-related complications can be expected compared to conservative management, although no reduction in mortality is observed. In patients predicted as having mild ABP early ERCP has no advantage, compared to conservative management, for reducing complication rate. However more accurate and rapid predictors of severity are needed and a more accurate selection of patients with impacted stones by non or mini-invasive modalities, such as magnetic resonance cholangio-pancreatography (MRCP) and endoscopic ultrasound (EUS), should be considered [50].

Practice points • In acute biliary pancreatitis, early ERCP and sphincterotomy is recommended in patients with predicted severe pancreatitis. • In patients predicted as having mild pancreatitis, early ERCP has no advantage, compared to conservative management, for reducing complication rate. • Pancreatitis-related mortality is not affected by early ERCP.

Research agenda • Need of more accurate tests for predicting severity in acute pancreatitis. • Endoscopic ultrasound (EUS) and magnetic resonance cholangio pancreatography (MRCP) should be considered to better select patients who can benefit from therapeutic ERCP.

Conflict of interest statement None declared.

References [1] Toouli J, Brooke-Smith M, Bassi C, Carr-Locke D, Telford J, Freeny P, et al. Guidelines for the management of acute pancreatitis. J Gastroenterol Hepatol 2002;17(Suppl.):S15–39. [2] Toh SK, Phillips S, Johnson CD. A prospective audit against national standards of the presentation and management of acute pancreatitis in the South of England. Gut 2000;46:239–43. [3] Beckingham IJ, Bornman PC. ABC of diseases of liver, pancreas, and biliary system. Acute pancreatitis. Br Med J 2001:595–8. [4] Uomo G, Pezzilli R, Gabbrielli A, Castoldi L, Zerbi A, Frulloni L, et al., ProInf-AISP Study Group. Diagnostic assessment and outcome of acute pancreatitis in Italy: results of a prospective multicentre study. ProInf-AISP: Progetto informatizzato pancreatite acuta, Associazione Italiana Studio Pancreas, phase II. Dig Liver Dis 2007;39:829–37. [5] Steinberg W, Tenner S. Acute pancreatitis. N Engl J Med 1994;330:1198–210. [6] Baron TH, Morgan DE. Acute necrotizing pancreatitis. N Engl J Med 1999;340:1412–7. [7] Pezzilli R, Uomo G, Gabbrielli A, Zerbi A, Frulloni L, De Rai P, et al., ProInf-AISP Study Group. A prospective multicentre survey on the treatment of acute pancreatitis in Italy. Dig Liver Dis 2007;39:838–46. [8] Cavallini G, Frulloni L, Bassi C, Gabbrielli A, Castoldi L, Costamagna G, et al., ProInf-AISP Study Group. Prospective multicentre survey on acute pancreatitis in Italy (ProInf-AISP): results on 1005 patients. Dig Liver Dis 2004;36:205–11.

A. Moretti et al. / Digestive and Liver Disease 40 (2008) 379–385 [9] Forsmark CE. The clinical problem of biliary acute necrotizing pancreatitis: epidemiology, pathophysiology, and diagnosis of biliary necrotizing pancreatitis. J Gastrointest Surg 2001;5:235– 9. [10] Uomo G, Slavin J. Endoscopic sphincterotomy for acute pancreatitis: arguments in favour. Ital J Gastroenterol Hepatol 1998;5:557– 61. [11] Acosta JM, Pellegrini CA, Skinner DB. Etiology and pathogenesis of acute biliary pancreatitis. Surgery 1980;88:118–25. [12] Lee SP, Nicholls JF, Park HZ. Biliary sludge as a cause of Acute Pancreatitis. New Engl J Med 1992;326:589–93. [13] Blackstone MO. The opie century. Pancreas 1988;3:340–5. [14] Acosta JM, Rubio Galli OM, Rossi R, Chinellato AV, Pellegrini CA. Effect of duration of ampullary gallstone obstruction on severity of lesion of acute pancreatitis. J Am Coll Surg 1997;184:499– 504. [15] R¨unzi M, Saluja A, Lerch MM, Dawra R, Nishino H, Steer ML. Early ductal decompression prevents the progression of biliary pancreatitis: an experimental study in the opossum. Gastroenterology 1993;105:157–64. [16] Stone HH, Fabian TC, Dunlop WE. Gallstone pancreatitis: biliary tract pathology in relation to time of operation. Ann Surg 1981;194: 305–10. [17] Tondelli P, Stutz K, Harder F, Schuppisser JP, Allg¨ower M. Acute gallstone pancreatitis: best timing for biliary surgery. Br J Surg 1982;69:709–10. [18] Kelly TR, Wagner DS. Gallstone pancreatitis: a prospective randomized trial of the timing of surgery. Surgery 1988;104:600–5. [19] Safrany L, Neuhaus B, Krause S, Portocarrero G, Schott B. Endoscopic papillotomy for acute pancreatitis caused by biliary disease. Dtsch Med Wochenschr 1980;105:115–9. [20] Neoptolemos JP, Carr-Locke DL, London NJ, Bailey IA, James D, Fossard DP. Controlled trial of urgent endoscopic retrograde cholangiopancreatography and endoscopic sphincterotomy versus conservative treatment for acute pancreatitis due to gallstones. Lancet 1988;2:979–83. [21] Fan ST, Lai EC, Mok FP, Lo CM, Zheng SS, Wong J. Early treatment of acute biliary pancreatitis by endoscopic papillotomy. N Engl J Med 1993;328:228–32. [22] Folsch UR, Nitsche R, Ludtke R, Hilgers RA, Creutzfeldt W, The German Study Group on Acute Biliary Pancreatitis. Early ERCP and papillotomy compared with conservative treatment for acute biliary pancreatitis. N Engl J Med 1997;336:237–42. [23] Sharma VK, Howden CW. Metaanalysis of randomized controlled trials of endoscopic retrograde cholangiography and endoscopic sphincterotomy for the treatment of acute biliary pancreatitis. Am J Gastroenterol 1999;94:3211–4. [24] Ayub K, Imada R, Slavin J. Endoscopic retrograde cholangiopancreatography in gallstone-associated acute pancreatitis. Cochrane Database Syst Rev. 2004;18:CD003630. [25] Novak A, Novakowska-Dulawa E, Marek TA, Rybicka J. Final results of the prospective, randomized controlled study on endoscopic sphincterotomy versus conventional management in acute biliary pancreatitis. Gastroenterology 1995;108:A380. [26] UK guidelines for the management of acute pancreatitis. Working Party of the British Society of Gastroenterology; Association of Surgeons of Great Britain and Ireland; Pancreatic Society of Great Britain and Ireland; Association of Upper GI Surgeons of Great Britain and Ireland. Gut 2005;54(Suppl. 3):1–9. [27] Banks PA, Freeman ML, Practice Parameters Committee of the American College of Gastroenterology. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379–400.

385

[28] Heinrich S, Schafer M, Rousson V, Clavien PA. Evidence-based treatment of acute pancreatitis: a look at established paradigms. Ann Surg 2006;243:154–68. [29] Enns R, Baillie J. Review article: the treatment of acute biliary pancreatitis. Aliment Pharmacol Ther 1999;13:1379–89. [30] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1–12. [31] DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clin Trials 1986;7:177–88. [32] Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. N Engl J Med 1988;318:1728–33. [33] Chalmers TC, Blum A, Buyse M, Hewitt P, Koch M, editors. Data analysis for clinical medicine. The quantitative approach to patient care in gastroenterology. Rome, Manchester, New York, Hong Kong: International University Press; 1988. [34] Messori A, Rampazzo R. Meta-analysis of clinical trials based on censored end-points: simplified theory and implementation of the statistical algorithms on a microcomputer. Comput Methods Programs Biomed 1993;40:261–7. [35] L’Abb´e KA, Detsky AS, O’ Rourke K. Meta-analysis in clinical research. Ann Intern Med 1987;107:224–33. [36] Rosenthal R. The ‘File Drawer Problem’ and tolerance for null results. Psycho Bull 1979;86:638–41. [37] Nitsche R, Folsch UR, Ludtke R, Hilgers RA, Creutzfeldt W. Urgent ERCP in all cases of acute biliary pancreatitis? A prospective randomized multicenter study. Eur J Med Res 1995;1:127–31. [38] Neoptolemos JP, London N, Slater ND, Carr-Locke DL, Fossard DP, Moosa AR. A prospective study of ERCP and endoscopic sphincterotomy in the diagnosis and treatment of gallstone acute pancreatitis. A rational and safe approach to management. Arch Surg 1986;121:697–702. [39] Acosta JM, Katkhouda N, Debian KA, Groshen SG, Tsao-Wei DD, Berne TV. Early ductal decompression versus conservative management for gallstone pancreatitis with ampullary obstruction: a prospective randomized clinical trial. Ann Surg 2006;243:33–40. [40] Zhou MQ, Li NP, Lu RD. Duodenoscopy in treatment of acute gallstone pancreatitis. Hepatobil Pancreat Dis Int 2002;1:608–10. [41] Oria A, Cimmino D, Ocampo C, Silva W, Kohan G, Zandalazini H, et al. Early endoscopic intervention versus early conservative management in patients with acute gallstone pancreatitis and biliopancreatic obstruction: a randomized clinical trial. Ann Surg 2007;245:10–7. [42] Kelly TR, Swaney PE. Gallstone pancreatitis: the second time around. Surgery 1982;92:571–5. [43] Neoptolemos JP. The theory of ‘persisting’ common bile duct stones in severe gallstone pancreatitis. Ann R Coll Surg Engl 1989;71:326–31. [44] Kelly TR. Gallstone pancreatitis: the timing of surgery. Surgery 1980;88:345–9. [45] Vitale GC. Early management of acute gallstone pancreatitis. Ann Surg 2007;245:18–9. [46] Lempinen M, Puolakkainen P, Kemppainen E. Clinical value of severity markers in acute pancreatitis. Scand J Surg 2005;94:118–23. [47] Rettally CA, Skarda S, Garza MA, Schenker S. The usefulness of laboratory tests in the early assessment of severity of acute pancreatitis. Crit Rev Clin Lab Sci 2003;40:117–49. [48] Sandberg AA, Borgstrom A. Early prediction of severity in acute pancreatitis. Is this possible? JOP 2002;3:116–25. [49] Triester SL, Kowdley KV. Prognostic factors in acute pancreatitis. Clin Gastroenterol 2002;34:167–76. [50] Verma D, Kapadia A, Eisen GM, Alder DG. EUS versus MRCP for detection of choledocholitiasie. Gastrointest Endosc 2006;64:248–54.