Subscribe to RSS
DOI: 10.1055/s-2000-138
Analysis of the Risk Factors Associated with Endoscopic Sphincterotomy Techniques: Preliminary Results of a Prospective Study, with Emphasis on the Reduced Risk of Acute Pancreatitis with Low-Dose Anticoagulation Treatment
M.D. T. Rabenstein
Medizinische Klinik I mit Poliklinik Universität Erlangen-Nürnberg
Krankenhausstrasse 12
91054 Erlangen, Germany
Phone: +49-9131-8536909
Email: Thomas.Rabenstein@med1.med.uni-erlangen.de
Publication History
Publication Date:
25 September 2003 (online)
Background and Study Aims: The aim of the present study was to analyze the risk factors associated with complications of endoscopic sphincterotomy (ES).
Patients and Methods: In all consecutive endoscopic sphincterotomies carried out between September 1994 and December 1996, the possible risk factors (12 patient-related factors and 12 procedure-related ones), as well as the concomitant medical treatment, indications, techniques, and success of endoscopic sphincterotomy were evaluated prospectively. Risk factors were analyzed on an exploratory basis using univariate methods. “Potential risk factors” (univariate, P < 0.1) underwent multivariate analysis to determine independent “risk factors” (multivariate, P < 0.05). In addition, the complication rate was calculated according to the number of potential risk factors present.
Results: A total of 438 patients who underwent ES were analyzed. Complications occurred in 7.5 % (n = 33; acute pancreatitis 4.3 %, hemorrhage 2.3 %, cholangitis 0.9 %, technical 0.2 %). Statistical analysis of the complications identified three independent risk factors (coagulopathy, patient age (≤ 60 years, pancreas divisum), and one protective factor (pancreatic duct obstruction). The frequency of acute pancreatitis was increased by two independent risk factors (pancreas divisum, ES frequency < 40 procedures/year) and was reduced if low-dose anticoagulation (unfractionated heparin or low molecular weight heparin) was administered (0.9 %, one of 115 vs. 5.8 %, 18 of 313; P < 0.05). The effect of anticoagulation was not confounded by the presence or absence of other potential risk factors for acute pancreatitis. Neither the risk nor the severity of hemorrhage were increased by low-dose anticoagulation. Due to the low number of events, only potential risk factors for hemorrhage were identified (coagulopathy, intensive-care treatment). The overall complication rate and the incidence of pancreatitis and hemorrhage increased significantly depending on the number of simultaneous potential risk factors present (P < 0.0001).
Conclusions: Patients at risk for complications after endoscopic sphincterotomy can be identified by risk factor analysis. These data suggest the hypothesis that low-dose anticoagulation prior to endoscopic sphincterotomy reduces the risk of acute pancreatitis after sphincterotomy.
#Introduction
It is known from numerous retrospective investigations that endoscopic retrograde cholangiopancreatography (ERCP) combined with endoscopic sphincterotomy (ES) is one of the most difficult interventions in gastrointestinal endoscopy, and that it is associated with a high rate of complications. If the intervention is carried out by experienced endoscopists, the success rate usually reaches a level of 90 - 95 %. However, a complication rate of 5 - 10 % needs to be taken into account. On average, one in 100 patients dies of complication-related causes after endoscopic sphincterotomy [1 - 4].
In interventional gastrointestinal endoscopy, as in all types of surgery, it is necessary to carry out continuous assessment of the quality of the measures applied and to strive for continuous improvement through comparative evaluation of the benefits and risks [5]. The first consensus conference aiming to define complications after endoscopic sphincterotomy and to classify their severity met in 1990 - 16 years after the initial description of the technique [1] [6] [7] . The results of the conference currently represent the worldwide standard for assessing complications after ERCP and ES, and they are generally regarded as the foundation for the continuing development of these methods [5]. Any attempts to minimize the risks involved acquire particular importance in this context [5]. However, before an assessment of specific measures for reducing the risk of complications in endoscopic sphincterotomy can be considered, the major factors that determine the risks of the intervention need to be clarified.
The present report describes a prospective exploratory single-center investigation of the risk factors of endoscopic sphincterotomy and equivalent cutting techniques. In addition, the significance of these factors for the post-ES complication rate, and the effect of concomitant medical treatment, were evaluated.
#Methods
The study protocol and the selection of potentially relevant risk factors was based on experience derived from a number of prospective and retrospective studies regarding endoscopic sphincterotomy conducted since 1973 [3] [4] [6] [8 - 11] . All consecutive ES procedures carried out between 12 September 1994 and 31 December 1996 were evaluated prospectively. The term “endoscopic sphincterotomy” was used to cover various endoscopic incision procedures applied:
Biliary endoscopic sphincterotomy (B-ES). The technique of biliary endoscopic sphincterotomy was carried out as described in a variety of previously published reports, with the cutting route generally aiming in the 11-o'clock direction and using double-lumen guide-wire sphincterotomes in nearly all procedures [12].
Pancreatic sphincterotomy (PST). PST was performed as described elsewhere, either using a guide-wire sphincterotome or, if selective cannulation was not possible due to stones or strictures, using the needle knife without previous stenting [11].
Minor papilla sphincterotomy (MST) was conducted with the needle knife in all cases, with one exception (guide-wire sphincterotome).
Needle-knife papillotomy (NKP). A precut papillotomy using the needle knife consists of an incision of the roof of the papilla in order to achieve selective cannulation of the biliary or pancreatic duct with a catheter or a sphincterotome. As more than 95 % of all precut procedures in this study were performed with the needle knife, the term “needle-knife papillotomy” will be used exclusively below. In general, precut papillotomy using the needle knife was performed according to the technique described by Huibregtse [13].
Data acquisition included indications, concomitant medical treatment, technical performance, success, and complications. Patients were followed up using physical examinations and blood samples at four, 24, and 48 hours after ES. Clinical observations were recorded throughout the patients' entire hospital stay. After 30 days, all patients or their family doctors were contacted by phone or mail to monitor any later occurrence of complications. B-ES, PST, and MST were rated as successful if the cutting procedure achieved a sphincterotomy sufficiently wide for all the intended or required therapeutic measures to be carried out. An NKP was regarded as successful if selective cannulation of the bile duct or the pancreatic duct was subsequently possible. The definition of complications and the classification of their severity followed the commonly accepted criteria [1]. For a better overview, the complications designated as “minor” and “moderate” were combined to form the single category “moderate.” Since nearly all patients in Germany who undergo ERCP or ES remain hospitalized until the entire diagnosis or therapy is completed, use of the terms “unplanned hospitalization” or “period of hospitalization” is not helpful in defining or classifying the severity of complications. For this reason, the type and duration of the specific therapeutic measures used to manage complications in the present study served as criteria for assessing the degree of severity of a complication. Indirect complications (pneumonia, apoplexy, thrombosis, pulmonary embolism, etc.) were not taken into account in the present investigation. The method-related mortality and the overall mortality were determined 30 days after the last application of a cutting technique. All of the parameters assessed were fed into a computer-based documentation system (SPSS for Windows).
Statistical analysis of potential risk factors of endoscopic sphincterotomy. Age data were documented as mean values with standard deviations, and comparison of age between groups was performed using the Mann-Whitney U test. All other variables were categorical, and were compared using Fisher's exact test (univariate analysis) and logistic regression (multivariate analysis) [14]. A P value of less than or equal to 0.05 (two-sided) was rated as significant. Possible risk factors were selected for investigation on the basis of prior knowledge, either from our own group's studies or from published data. For all patients who underwent ES, a total of 24 possible risk factors were assessed, either prior to ES (patient-related risk factors, n = 12) or immediately after ES (procedure-related risk-factors, n = 12). These 24 variables were compared with three different types of complication, producing 72 statistical tests. The error rate within each set of 24 variables and the type of complications underwent Bonferroni correction to control for family-wise error rate, and only P values of less than 0.002 can be regarded as statistical significant [15]. However, with regard to the three different types of complications, only the error rate per comparison was controlled [15].
The present investigation must be interpreted as an exploratory study. P values of between 0.05 and 0.002 given on the tables for risk factor analysis are only descriptive summaries (the term “potential risk factor” will be used below). The number of selected variables for multiple testing was in accordance with the criterion proposed by Concato et al. [16], and selection of variables was performed in a forward manner [17]. The inclusion criterion for multiple testing was a univariate P value of less than 0.1; other variables were not included in the first step of the multivariate model. Variables with a P value greater than 0.05 in the last step of multivariate analysis were excluded via variable selection (final exclusion criterion). Thus, only variables with a P value of equal to or less than 0.05 are listed as results of multivariate analysis (the term “risk factor” will be used below). Factors that showed a reduced complication rate in the statistical analysis were given an inverse odds ratio (10-1 instead of 0.1). If there were empty cells in the cross-tables and no potential relevance of the analyzed factor for the complication rate (P > 0.1), no odds ratios were calculated. Due to the low number of events, multivariate analysis for hemorrhage was not performed. In order to determine the clinical significance of the potential risk factors as a whole, the complication rate was also assessed as a function of the number of simultaneously present potential risk factors.
#Results
In all, endoscopic sphincterotomy techniques were used in 438 patients during the observation period (Table [1]). There were 243 male patients (55.5 %) and 195 female ones (45.5 %); the average age of the patients was 61.3 ± 16.4 years (median 63.0, minimum 14, maximum 96 years), but the male patients were significantly younger (58.0 ± 15.4 years) than the female ones (65.4 ± 16.7 years; P < 0.001). The most frequent indications for endoscopic sphincterotomy consisted of bile duct stones (37.7 %), malignancies (23.3 %), and chronic pancreatitis (21.9 %). Acute pancreatitis and papillary stenosis were indications for ES in less than 5 %. Complications after laparoscopic cholecystectomy that required endoscopic therapy showed an incidence of 5.4 %, but were classified as “other indications.” Specifically, the following indications for ES were included in the category of “other indications” (overall occurrence 10.5 %): bile duct leakage, benign bile duct stenosis, cholangitis, primary sclerosing cholangitis, sphincter of Oddi disease, pancreas divisum, pancreatic leakage, and diagnostic clarification in case of jaundice, bile duct dilation, or malignant disease of unknown origin. Patients with bile duct stones or malignancy were significantly older than patients with chronic pancreatitis or “other indications” (P < 0.05). Whereas female patients underwent sphincterotomy significantly more frequently due to bile duct concrements (P < 0.005) or papillary stenoses (P < 0.001), in male patients sphincterotomy was carried out more frequently because of chronic pancreatitis (P < 0.0001).
On average, 1.3 endoscopic cutting techniques were used per patient. The most frequent cutting technique was B-ES, at 66.2 % of patients (290 of 438). PST was performed in 9.4 % (41 of 438). Needle-knife papillotomy was carried out prior to B-ES in 28.3 % (82 of 290) and in 26.8 % (11 of 41) in combination with PST. A total of 20.1 % of patients (88 of 438) received needle-knife papillotomy without any need for subsequent complete sphincterotomy. In the latter group of patients, the indication for ES consisted of placement of internal drainage tubes to treat bile duct stenoses, pancreatic duct stenoses, or cholangitis. In addition, this group included a small number of diagnostic cases in which there was a suspicion of biliary or pancreatic disorder due to clinical findings and laboratory tests, but attempts at cannulation using the usual set of instruments had failed. A minority of 4.3 % of the patients (19 of 438) underwent sphincterotomy of the minor papilla. In two cases, minor papilla sphincterotomy was accompanied by needle-knife papillotomy of the major papilla. Thus, a total of 41.9 % of the patients (183 of 438) underwent NKP.
The total success rate of endoscopic sphincterotomy was 96.0 % (n = 421), and there were no discrepancies between the various endoscopists' individual success rates. In 4.0 % of cases (15 of 378), biliary sphincterotomy of the major papilla and therapeutic biliary access were not successful even after using the needle knife. In 10.5 % (two of 19), sphincterotomy of the minor papilla remained unsuccessful. A success rate of 100 % was obtained with pancreatic sphincterotomy.
Post-ES complications were observed in 7.5 % of patients (n = 33) (Table [2]). The majority of the complications (31 of 33) were diagnosed within 48 hours of ES, but two cases of hemorrhage became manifest four and seven days after ES, respectively. All complications were managed nonsurgically, including ten severe events. Two patients, both inoperable due to their general state of health, died of post-ES hemorrhage, resulting in a method-related mortality of 0.5 %. Thirty days after ES, the overall mortality amounted to 6.4 % (n = 28). Delayed or undiagnosed complications of ES were not suspected as the cause of death in any cases of late mortality, since all the fatalities that occurred within the 30-day follow-up period were attributed to an accompanying disorder - in the majority of cases to a malignant disease.
Post-ERCP pancreatitis was the most frequently observed complication (4.3 %, n = 19). Acute pancreatitis occurred more often in female patients than in males (6.7 %, 13 of 195, vs. 2.5 %, six of 243; P < 0.05), and more often in patients aged under 60 years (6.9 %, 14 of 203, vs. 2.1 %, five of 235; P < 0.05). In the majority of cases (n = 16), “moderate” pancreatitis was noted, whereas “severe” pancreatitis was present in only three patients. One patient developed severe pancreatitis several hours after ES, and moderate bleeding in addition on the fourth day post-ES. Sphincter of Oddi dysfunction (SOD) is a rare indication for ES in our department, and SOD was therefore not taken into account in the risk factor analysis. However, it should be noted that two of four patients with SOD developed acute pancreatitis after ES. Hemorrhage occurred in a total of 2.3 % of cases (10 of 438). As seven events were “severe” and two patients died, post-ES hemorrhage proved to be the most hazardous complication of ES. The incidence of cholangitis and/or cholecystitis was 0.9 % (n = 4); all cases showed a moderate course. One technical complication was noted during the observation period (0.2 %): in a Billroth II situation, a biliary stent serving as a guide was placed before needle-knife papillotomy was performed. During the subsequent needle-knife papillotomy procedure, the plastic stent migrated proximally, and it could only be removed endoscopically in a later ERCP session.
Primary biliary endoscopic sphincterotomy showed a complication rate of 7.2 % (15 of 208) and primary pancreatic sphincterotomy exhibited only half of the complication rate of B-ES (3.3 %, one of 30; not significant). In cases in which different endoscopic sphincterotomy techniques were combined, the complication rate rose to 9.8 % (eight of 82) for NKP with secondary B-ES and to 18.2 % (two of 11) for NKP and secondary PST, respectively. In comparison with primary standard sphincterotomy, however, this difference was not significant. In the 88 patients who underwent needle-knife papillotomy alone, six complications occurred (6.8 %). Overall, the complication rate in cases in which NKP was involved (8.7 %, 16 of 183) did not significantly differ from that in cases in which NKP was not performed (6.7 %, 17 of 255; P = 0.46). Papillotomy of the minor papilla showed a complication rate of 5.3 % (one of 19). The sphincterotomies were performed by a total of seven endoscopists. While four of the endoscopists already had previous experience in endoscopic sphincterotomy before the start of the study (individual numbers of prior procedures: 77, 148, 149, and 369), three endoscopists began performing sphincterotomy after an approximately two-year training phase in ERCP. Endoscopists with previous experience carried out 93.2 % of the ES procedures evaluated in the present investigation (n = 408), and the “beginners” conducted 6.8 % of the interventions (n = 30) in selected patients and under the supervision of an experienced colleague. Under these conditions, the ES trainees showed a complication rate of 3.3 % (one of 30). Interventions in which an “ES-inexperienced” endoscopist performed ERCP, while ES was conducted by an experienced endoscopist, were credited to the experienced investigator. In all, 7.7 % complications (32 of 408) occurred in ES procedures carried out by experienced endoscopists. In the present investigation, three of the seven endoscopists had an annual ES frequency of more than 40 procedures per year, associated with individual complication rates of between 4.6 % and 7.7 %. However, the complication rates for endoscopists with a lower ES frequency reached up to 22.0 %. Overall, endoscopists with a high ES frequency (more than 40 procedures per year) caused significantly fewer complications (6.3 %, 23 of 364) than endoscopists with a lower ES frequency (13.5 %, 10 of 74; P < 0.05).
#Risk Factors for Complications after Endoscopic Sphincterotomy
The results of the exploratory univariate risk factor analysis for complications of endoscopic sphincterotomy are presented in Table [3], and the results of the multivariate analysis of possible risk factors for complications are shown in Table [4].
Several variables can be described as potential risk factors for overall complications: preexisting coagulopathy, patient age (≤ 60 years), pancreas divisum, previous laparoscopic cholecystectomy, intensive-care treatment, ES frequency of the endoscopist < 40/year, and insertion of a nasobiliary tube after sphincterotomy. Three variables were associated with a reduced complication rate if present: chronic pancreatitis with duct obstruction (stones or strictures), diabetes mellitus, and low-dose anticoagulation. Multivariate analysis demonstrated that coagulopathy, patient age (≤ 60 years), pancreas divisum, and pancreatic obstruction (protective) had an independent prognostic value with regard to the overall incidence of complications, but after Bonferroni correction, none of the variables analyzed reached confirmed statistical significance. A highly significant correlation was found between the number of simultaneously present potential risk factors in the same patient and the resulting complication rate (Table [5]). Patients with three or less potential risk factors had a complication rate of only 2.9 % (eight of 278), and if there were four or more potential risk factors, the complication rate reached 15.6 % (25 of 160).
The analysis of specific risk factors for post-ES pancreatitis showed that patient-related risk factors appear to have greater clinical significance for inducing pancreatitis than procedure-related risk-factors (Table [3]): Exploratory analysis identified nine potential risk factors for post-ES pancreatitis (pancreas divisum, patient age (≤ 60 years), female sex, previous laparoscopic cholecystectomy, preexisting anemia, ES frequency of the endoscopist < 40/year, pancreatic cannulation, use of the needle knife, and an incision length < 5 mm), and one potential protective factor (low-dose anticoagulation). According to the multivariate analysis, pancreas divisum seems to be the most important patient-related risk factor (odds ratio 8.2). A total of 12 patients had pancreas divisum (11 complete, one incomplete), and four complications occurred (three cases of pancreatitis, one of hemorrhage). In all patients with pancreas divisum, the minor papilla was cannulated for diagnostic reasons during the same ERCP session in which the endoscopic sphincterotomy techniques were performed. In one of two patients with pancreas divisum who had received biliary sphincterotomy, pancreatitis occurred in a large ventral pancreas. In another patient, severe pancreatitis combined with moderate hemorrhage occurred after needle-knife papillotomy, pancreatic sphincterotomy (major papilla) and minor papilla cannulation; minor papilla sphincterotomy was performed in a second session without complications. A third patient had pancreatitis following MST. Pancreatitis after minor papilla sphincterotomy was thus seen only in one of ten cases, but in three of 12 cases of pancreas divisum. The major procedure-related risk was a low ES frequency of the endoscopist, resulting in a pancreatitis incidence of 10.8 % (eight of 74) when endoscopists with a case frequency of less than 40 per year carried out ES (odds ratio 3.8). The incidence of post-ES pancreatitis significantly increased with the number of potential risk factors present (Table [5]). Patients with up to four potential risk factors developed pancreatitis in 1.7 % of cases (six of 349), whereas five or more potential risk factors resulted in a pancreatitis rate of 14.6 % (13 of 89).
Four potential risk factors were identified relating to post-ES hemorrhage. The highest risk for hemorrhage was observed in patients with preexisting coagulopathy, and in intensive-care patients. In addition, hemorrhage was associated with longer incisions and occurred after placement of nasobiliary tubes. Due to the low number of events, multivariate analysis was not performed. The incidence of hemorrhage significantly depended on the number of potential risk factors (Table [5]).
| Indication | Frequency | Age (mean) | Sex (%) | ||
| n | % | M | F | ||
| Bile duct stone | 165 | 37.7 | 67.7 | 46.1 | 53.9 |
| Malignancy | 102 | 23.3 | 65.7 | 52.9 | 47.1 |
| Chronic pancreatitis | 96 | 21.9 | 48.2 | 80.2 | 19.8 |
| Papillary stenosis | 19 | 4.3 | 63.6 | 26.3 | 73.7 |
| Acute pancreatitis | 10 | 2.3 | 60.1 | 80.0 | 20.0 |
| Other indications | 46 | 10.5 | 55.3 | 50.0 | 50.0 |
| Total | 438 | 100.0 | 61.3 | 55.5 | 44.5 |
| Complication | Classification | Total | |||
| Moderate | Severe | ||||
| n | n | n | % | 95 % Cl | |
| Pancreatitis | 16 | 3 | 19 | 4.3 | 2.8 - 6.6 |
| Hemorrhage* | 3 | 7 | 10 | 2.3 | 1.3 - 4.2 |
| Cholangitis | 4 | - | 4 | 0.9 | 0.3 - 2.3 |
| Technical | 1 | - | 1 | 0.2 | 0.0 - 1.2 |
| Perforation | - | - | - | 0.0 | - |
| Total (patients) | 23 | 10 | 33** | 7.5 | 5.4 - 10.4 |
| * Two patients with severe hemorrhage died (mortality 0.5 %). ** In one patient, severe pancreatitis was observed immediately after ES, and moderate hemorrhage was noticed four days after the procedure. | |||||
| Risk factor | Patients (n = 438) | Total complications (n = 33, 7.5 %) | Acute pancreatitis (n = 9, 4.3 %) | Hemorrhage (n = 10, 2.3 %) | ||||||||||||
| n | % | OR | 95 % Cl | P | n | % | OR | 95 % Cl | P | n | % | OR | 95 % Cl | P | ||
| Patient-related factors | ||||||||||||||||
| Age ≤ 60 years | 203 | 21 | 10.3 | 2.1 | 1.03 - 4.47 | 0.0458 | 14 | 6.9 | 3.4 | 1.20 - 9.63 | 0.0180 | 5 | 2.5 | 1.2 | 0.33 - 4.07 | 1.0000 |
| Female sex | 195 | 19 | 9.7 | 2.8 | 1.05 - 7.57 | 0.1450 | 13 | 6.7 | 2.7 | 1.05 - 7.57 | 0.0358 | 4 | 2.1 | 0.8 | 0.23 - 2.97 | 1.0000 |
| *Diabetes mellitus | 86 | 1 | 1.2 | 10.0-1 | 0.02 - 0.87 | 0.0104 | 1 | 1.2 | 5.0-1 | 0.03 - 1.66 | 0.1418 | 0 | 0.0 | |||
| Conventional cholecystectomy | 86 | 3 | 3.5 | 0.4 | 0.11 - 1.30 | 0.1686 | 1 | 1.2 | 0.2 | 0.03 - 1.66 | 0.1418 | 2 | 2.3 | 1.0 | 0.21 - 4.91 | 1.0000 |
| Laparoscopic cholecystectomy | 30 | 5 | 16.7 | 2.7 | 0.97 - 7.64 | 0.0644 | 4 | 13.3 | 4.0 | 1.25 - 13.02 | 0.0368 | 0 | 0.0 | |||
| Pancreas divisum | 12 | 3 | 25.0 | 4.4 | 1.13 - 17.11 | 0.0437 | 3 | 25.0 | 8.5 | 2.11 - 34.60 | 0.0192 | 1 | 8.3 | 4.2 | 0.49 - 36.20 | 0.2448 |
| Anemia | 18 | 3 | 16.7 | 2.6 | 0.71 - 9.48 | 0.1461 | 3 | 16.7 | 5.1 | 1.33 - 19.22 | 0.0374 | 1 | 5.6 | 2.7 | 0.32 - 22.43 | 0.3456 |
| Previous post-ERCP pancreatitis | 11 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | |||||||||
| Previous gastrectomy | 12 | 2 | 16.7 | 2.5 | 0.53 - 12.15 | 0.2265 | 0 | 0.0 | 1 | 8.3 | 4.2 | 0.49 - 36.20 | 0.2448 | |||
| Previous jaundice | 272 | 24 | 8.8 | 1.7 | 0.76 - 3.73 | 0.2625 | 11 | 4.0 | 0.8 | 0.33 - 2.11 | 0.8098 | 9 | 3.3 | 5.6 | 0.71 - 44.98 | 0.0975 |
| Intensive-care patient | 7 | 2 | 28.6 | 5.2 | 0.96 - 27.69 | 0.0911 | 0 | 0.0 | 2 | 28.6 | 21.2 | 3.56 - 125.76 | 0.0093 | |||
| Coagulopathy 1 | 7 | 3 | 42.9 | 10.0 | 2.14 - 46.87 | 0.0111 | 0 | 0.0 | 2 | 28.6 | 21.2 | 3.56 - 125.76 | 0.0093 | |||
| Procedure-related factors | ||||||||||||||||
| ES frequency < 40/year | 74 | 10 | 13.5 | 2.3 | 1.05 - 5.10 | 0.0494 | 8 | 10.8 | 3.9 | 1.50 - 10.04 | 0.0072 | 1 | 1.4 | 0.5 | 0.07 - 4.33 | 1.0000 |
| Emergency ES | 56 | 3 | 5.4 | 0.6 | 0.20 - 2.25 | 0.7853 | 0 | 0.0 | 3 | 5.4 | 3.0 | 0.76 - 12.09 | 0.1236 | |||
| Sphincterotomy procedures | ||||||||||||||||
| B-ES | 297 | 24 | 8.1 | 1.3 | 0.58 - 2.85 | 0.6987 | 12 | 4.0 | 0.8 | 0.31 - 2.09 | 0.6253 | 8 | 2.7 | 1.9 | 0.40 - 9.18 | 0.5119 |
| PST | 41 | 3 | 7.3 | 1.0 | 0.28 - 3.31 | 1.0000 | 3 | 7.3 | 1.9 | 0.52 - 6.74 | 0.4067 | 0 | 0.0 | |||
| MST | 19 | 1 | 5.3 | 0.7 | 0.09 - 5.20 | 1.0000 | 1 | 5.3 | 1.2 | 0.16 - 9.80 | 0.5772 | 1 | 5.3 | 2.5 | 0.30 - 21.07 | 0.3612 |
| Size of sphincterotomy 2 | 141 | 11 | 7.8 | 1.1 | 0.50 - 2.25 | 0.8490 | 10 | 7.1 | 2.4 | 0.97 - 6.15 | 0.0757 | 0 | 0.0 | |||
| NKP involvement | 183 | 16 | 8.7 | 1.4 | 0.67 - 2.79 | 0.4627 | 12 | 6.6 | 2.5 | 0.98 - 6.57 | 0.0577 | 2 | 1.1 | 0.3 | 0.07 - 1.66 | 0.2070 |
| Failed procedure | 17 | 1 | 5.9 | 0.8 | 0.16 - 7.25 | 1.0000 | 1 | 5.9 | 1.4 | 0.20 - 5.60 | 0.5363 | 0 | 0.0 | |||
| Pancreatic cannulation | 246 | 22 | 8.9 | 1.6 | 0.76 - 3.40 | 0.2736 | 15 | 6.1 | 3.0 | 0.99 - 9.35 | 0.0566 | 5 | 2.0 | 0.8 | 0.22 - 2.72 | 0.7540 |
| Pancreatic contrast | 276 | 23 | 8.3 | 1.4 | 0.64 - 2.98 | 0.4582 | 15 | 5.4 | 2.3 | 0.74 - 6.96 | 0.2231 | 7 | 2.5 | 1.4 | 0.35 - 5.41 | 0.7511 |
| *Pancreatic obstruction 3 | 72 | 1 | 1.4 | 5.0-1 | 0.20 - 1.09 | 0.0271 | 1 | 1.4 | 3.3-1 | 0.04 - 2.07 | 0.3370 | 1 | 1.4 | 1.7-1 | 0.07 - 4.48 | 1.0000 |
| *Anticoagulation 4 | 115 | 4 | 3.5 | 2.5-1 | 0.12 - 1.05 | 0.0637 | 1 | 0.9 | 6.4-1 | 0.02 - 1.13 | 0.0328 | 1 | 0.9 | 3.3-1 | 0.04 - 2.44 | 0.4660 |
| NSAID treatment | 16 | 1 | 6.3 | 0.8 | 0.10 - 6.35 | 1.0000 | 1 | 6.3 | 1.5 | 0.19 - 11.96 | 0.5144 | 0 | 0.0 | |||
| Nasobiliary tube | 24 | 4 | 16.7 | 2.7 | 0.85 - 8.28 | 0.0966 | 2 | 8.3 | 2.1 | 0.46 - 9.77 | 0.2795 | 2 | 8.3 | 4.6 | 0.92 - 23.03 | 0.0990 |
| * Protective factors in this analysis. 1 Coagulopathy: prothrombin test (Quick's value) < 50 %, platelet count < 50 000. 2 Small sphincterotomy/papillotomy (≤ 5 mm) vs. medium/large (>5 mm). 3 Pancreatic obstruction (stenosis or obstructing stones) vs. no pancreatic obstruction. 4 Low-dose anticoagulation treatment (low molecular weight heparin or unfractionated heparin) vs. no administration of heparin derivatives. B-ES: biliary endoscopic sphincterotomy; CI: confidence interval; ERCP: endoscopic retrograde cholangiopancreatography; ES: endoscopic sphincterotomy; MST: minor papilla sphincterotomy; NKP: needle-knife papillotomy; NSAID: nonsteroidal anti-inflammatory drug; OR: odds ratio; PST: pancreatic sphincterotomy. | ||||||||||||||||
| Risk factor | Total complications | Acute pancreatitis | ||||
| OR | 95 % Cl | P | OR | 95 % Cl | P | |
| Coagulopathy | 9.7 | 1.95 - 48.10 | 0.006* | |||
| Age ≤ 60 years | 2.9 | 1.33 - 6.21 | 0.007** | |||
| Pancreas divisum | 7.6 | 1.56 - 36.6 | 0.012** | 8.2 | 1.91 - 34.79 | 0.005* |
| ES frequency < 40/year | 3.8 | 1.44 - 10.00 | 0.007 | |||
| Pancreatic obstruction*** | 14-1 | 0.01 - 0.59 | 0.014** | |||
| * Reduction in the P value in the multivariate analysis compared to the univariate analysis (Table 3), due to different methods of statistical testing (Fisher's exact test vs. logistic regression analysis). ** Reduction in the P value in the multivariate analysis compared to the univariate analysis (Table 3) due to an increase in the odds ratio and to different methods of statistical testing (Fisher's exact test vs. logistic regression analysis). *** Protective in this analysis. 1 Coagulopathy: prothrombin test (Quick's value) < 50 %, platelet count < 50 000. | ||||||
| Potential risk | Patients | Complications | Complication rate | |
| factors* (n) | (n) | (n) | % | 95 % Cl |
| Total complications | ||||
| 0 | 2 | 0 | 0.0 | |
| 1 | 21 | 0 | 0.0 | |
| 2 | 86 | 1 | 1.2 | |
| 3 | 169 | 7 | 4.1 | |
| 0 - 3 | 278 | 8 | 2.9 | 1.5 - 5.6 |
| 4 | 109 | 10 | 9.2 | |
| 5 | 44 | 12 | 27.3 | |
| 6 | 7 | 3 | 42.9 | |
| 4 - 6 | 160 | 25 | 15.6 | 11.9 - 25.3 |
| Acute post-ES pancreatitis | ||||
| 0 | 14 | 0 | 0.0 | |
| 1 | 38 | 0 | 0.0 | |
| 2 | 96 | 0 | 0.0 | |
| 3 | 98 | 3 | 3.1 | |
| 4 | 103 | 3 | 2.9 | |
| 0 - 4 | 349 | 6 | 1.7 | 0.8 - 3.7 |
| 5 | 55 | 4 | 7.3 | |
| 6 | 27 | 5 | 18.5 | |
| 7 | 6 | 3 | 50.0 | |
| 8 | 1 | 1 | 100.0 | |
| 5 - 8 | 89 | 13 | 14.6 | 8.0 - 23.7 |
| Post-ES hemorrhage | ||||
| 0 | 61 | 0 | 0.0 | |
| 1 | 181 | 1 | 0.6 | |
| 2 | 187 | 6 | 3.2 | |
| 0 - 2 | 429 | 7 | 1.6 | 0.8 - 3.3 |
| 3 | 8 | 2 | 25.0 | |
| 4 | 1 | 1 | 100.0 | |
| 3 - 4 | 9 | 3 | 33.3 | 7.5 - 70.1 |
| * For protective factors, the absence of the factor concerned was used for analysis. ** P < 0.0001. | ||||
Low-Dose Anticoagulation and Acute Post-ES Pancreatitis
In 26.2 % of patients (115 of 438), low-dose anticoagulation (LDA) had been administered for various clinical reasons - e.g., to prevent thrombosis in case of immobilization, or due to prior thrombosis or prior pulmonary embolism; however, ERCP was not an indication for LDA. Patients received either heparin ( 15 000 IU/day (n = 57) subcutaneously or intravenously, or certoparin 3000 IU/day subcutaneously (n = 58), and routine controls of the partial thrombin time prior to ERCP were less than 40 seconds. In individual cases of acute thrombosis or acute pulmonary embolism, 25 000 - 30 000 IU/day heparin was administered intravenously prior to ERCP, and the infusion was discontinued from two hours before to six hours after the ERCP, without repeat checking of the partial thrombin time. In such cases, the therapeutic effect of heparin during the ERCP procedure was assumed to be equivalent to that of low-dose anticoagulation. The incidence of acute pancreatitis after ES was significantly lower in patients who had received low-dose anticoagulation, compared to those without anticoagulation (0.9 %, one of 115, vs. 5.8 %, 18 of 313; univariate relative risk 6-1, P < 0.05). The effect of LDA on reducing the rate of acute pancreatitis was most pronounced in patients aged below 60 years (0.0 % vs. 8.1 %). However, the potential protective effect of anticoagulation was not found to be significant in the multivariate analysis. Female patients received anticoagulation more often than male patients (62 of 195 vs. 57 of 243); none of the female patients with anticoagulation developed pancreatitis. Patients aged 60 years and less received anticoagulation significantly less frequently than older patients (47 of 203 vs. 78 of 235); no pancreatitis was noted in younger patients with anticoagulation. No other discrepancies between LDA and non-LDA patients were observed, although there was no randomization between LDA and non-LDA. The frequency and severity of hemorrhage, and the mortality due to bleeding, were not increased by anticoagulation (anticoagulation 1.7 %, two of 115, no fatalities, vs. non-anticoagulation: 2.6 %, eight of 313, two fatalities).
#Discussion
In the present study, the success rate of endoscopic sphincterotomy was found to be 96.0 %, a figure that is in good accordance with internationally published data [2] [4] [18 - 22] . The complication rate of 7.5 % observed in the study lies in the lower range of the rate in previously reported experience [2] [4] [18 - 22] .
In the current series, endoscopists with an ES frequency of more than 40 procedures per year caused significantly fewer complications than endoscopists with a lower ES frequency (6.3 % vs. 13.5 %), and the endoscopist was an important risk factor, particularly with regard to post-ES pancreatitis. These prospectively obtained results confirm those of a retrospective investigation conducted by our group [10], and largely corroborate the findings of the multicenter study by Freeman et al., which was the first to establish that the annual number of endoscopic sphincterotomies performed by an endoscopist has a significant effect on the complication rate [2]. In contrast to the study by Freeman et al., which focused on biliary sphincterotomy in particular and which was based on a multicenter design, the results of the present study on general endoscopic sphincterotomy procedures are derived from the experience of a single center. Thus, the patient series and indication spectra for the two investigations differ substantially. It should be noted that the high-risk indication group of patients with “sphincter of Oddi disease,” which was important in Freeman's study, did not feature in the present study. On the other hand, this study included a significantly larger number of patients with malignancy or chronic pancreatitis. Despite this, the two studies both show that endoscopic training, practice, and experience strongly influence the results of endoscopic sphincterotomy [2] [10] [22 - 25] .
Technical problems, such as the sphincterotomy procedure performed, the size of the sphincterotomy, needle-knife papillotomy, failed procedure, pancreatic cannulation, and pancreatic contrast injection were of no significance in relation to the overall complications in the present study. These results contrast slightly with the study by Freeman et al., in which several of these technical difficulties were rated as significant risk factors [2]. The present results may have been influenced by the patient population. More than 20 % of the study population was suffering from advanced chronic pancreatitis, and the complication rate in this patient group was very low. It is known, and confirmed by the findings of the present investigation, that patients with this type of underlying disease very rarely develop acute pancreatitis after endoscopic interventions such as ERCP, ES, PST, or extracorporeal shock-wave lithotripsy [2] [26 - 30] . However, one obvious difference in this study concerns the frequency with which needle-knife papillotomy was conducted. Overall, an increase in the frequency of NKP from 33 % between 1988 and 1993 [3] to a current level of 41.9 % has been observed [4]. A comparably high rate of NKP in relation to ES has been described only by a small number of other European centers, such as Amsterdam [31] and Hamburg [32]. The frequent use of the needle-knife in our department is a result of our approach of using an early precut procedure when there is an obvious indication for ES and a difficult cannulation situation. In a recently published retrospective study, we were able to show that early needle-knife papillotomy does not lead to an increased complication rate [9]. It was striking that, even with a prospective study design, early needle-knife papillotomy did not show an increased risk for complications in general compared to primary biliary endoscopic sphincterotomy or primary pancreatic sphincterotomy. Increasing numbers of reports have demonstrated that the use of the needle-knife in the hands of adequately trained and experienced endoscopists is by no means riskier than standard sphincterotomy [9] [13] [20] [31] [33] [34] ; in fact, NKP can actually enhance the effectiveness of ES by improving the success rate and reducing examination times [13] [35 - 38] . Since every attempt at cannulation is traumatic and carries a complication risk of its own, as unanimously confirmed by the risk factor analysis carried out by Freeman et al., as well as other reports [2] [39 - 41] , the risk of needle-knife papillotomy appears to be lower if this cutting technique is applied at an early stage - clearly before a papillary edema with obstructed drainage can develop as a result of prolonged manipulations at the papilla, if the endoscopist is sufficiently familiar with the needle-knife procedure and if the intervention is performed exactly according to the original description given by Huibregtse [13]. At many centers, needle-knife papillotomy is only used as a last resort in cases in which previous extensive cannulation attempts have failed, and the frequency of NKP compared to that of ES is consequently less than 10 %. In these circumstances, NKP has proved to be a risky technique [2] [42 - 44] . It must be assumed that this restrictive approach involves a selection of high-risk candidates for needle-knife papillotomy. With an underlying bias that is obviously linked to the different cannulation strategies, it is not possible to assign the occurrence of a complication to needle-knife papillotomy or to preceding cannulation attempts. The debate regarding the risks and benefits of precutting using the needle-knife will therefore continue to be controversial [38] [42] [45] until an unprejudiced prospective randomized investigation of the different so-called “access procedures” [41] leads to clarification.
Most of the other potential risk factors for “overall complications” did not correlate with technical aspects of ERCP or ES, but rather with patient-related factors: pancreas divisum, preexisting coagulopathy, absent signs of severe chronic pancreatitis, and younger age.
Since pancreatitis represents the most frequent complication after endoscopic sphincterotomy [2] [4] [25] [46 - 49] , the risk factors for this specific complication are of particular interest. The major risk factors for post-ES pancreatitis in the present investigation were an ES frequency of less than 40 procedures per year and pancreas divisum. Most of the other potential risk factors were patient-related: female sex, younger age, previous laparoscopic cholecystectomy, and preexisting anemia. The reduced risk of pancreatitis observed in patients with prior conventional cholecystectomy, and the increased risk in patients with prior laparoscopic cholecystectomy, might be explained by coincidental factors. Firstly, the usual reasons for ES after conventional cholecystectomy were long-term complications (bile duct stones), and these patients were significantly older; the sensitivity of the pancreas may therefore be reduced due to involution of the pancreatic gland in older patients. Secondly, the patients with a previous laparoscopic cholecystectomy were younger, and they underwent sphincterotomy due to severe early postoperative complications, such as strictures and fistulas. Some of these patients might have similarities with patients suffering from sphincter of Oddi disease, but the initial indication for laparoscopic cholecystectomy was not taken into account. In addition to an endoscopist with a low ES frequency, further potential procedure-related risks in the present analysis were pancreatic cannulation, an incision of less than 5 mm, and needle-knife papillotomy. Freeman et al. described sphincter of Oddi dysfunction and younger age, as well as female sex and a history of (ERCP-induced) pancreatitis, as being patient-related risk factors for post-ES pancreatitis [2]. Pancreas divisum was not noted in their study, probably because only biliary sphincterotomies were taken into account. Procedure-related risks were precut papillotomy, difficult cannulation, pancreatic contrast injections, pancreatic acinarization, sphincter of Oddi manometry, and guide-wire cannulation [2]. Unfortunately, the diameter of the bile duct was not assessed in the present study, and no information is therefore available regarding this previously described risk factor for post-ES pancreatitis, which seems to be important in patients with sphincter of Oddi disease [2] [20] . All in all, there is broad agreement between the present results and the data reported by Freeman et al. [2].
In the present study, hemorrhage after endoscopic sphincterotomy did not occur more frequently than in other reports, although more severe cases of bleeding were observed and 20 % of the events occurred more than 48 hours after the intervention [2] [4] [19 - 22] . The high proportion of severe bleeding might be an effect induced by the current setting used in ES, in which the cutting effect outweighed the coagulation effect - the cutting/coagulation ratio being 3 : 1. However, potential risk factors for hemorrhage were present in virtually all cases, and each of these factors was associated with a substantial risk of hemorrhage in the statistical analysis: preexisting coagulopathy (univariate odds ratio 21.2), intensive-care patient (OR 21.2), incision length of more than 10 mm (OR 6.8, not shown in the table), and placement of a nasobiliary drainage tube immediately after ES (OR 4.6). Thus, the higher incidence of hemorrhage was probably caused by the risk factors that were present.
What is the clinical significance of a single or of several combined risk factors for the resulting ES complication rate? The frequency of complications increased in dependence on the number of potential risk factors simultaneously present. For all complications, as well as for post-ES pancreatitis and post-ES hemorrhage, a threshold value for the number of potential risk factors was observed, above which a highly significant increase in the complication rate emerged (Table [5]). Although this was an exploratory investigation, these data therefore suggest that the presence of single risk factors does not increase the complication rate of ES. However, patients presenting with an accumulation of several risk factors are genuinely at risk, and are therefore good candidates for measures avoiding complications - i.e., procedures performed by the most experienced endoscopist available, or prophylactic medical treatment.
Low-dose anticoagulation treatment - in most cases administered in order to prevent thrombosis - appears to have a protective effect against the occurrence of acute pancreatitis after ES. Patients who did not receive anticoagulation showed a six-fold increased risk of acute pancreatitis compared to patients receiving low-dose anticoagulation. With regard to the distribution of general patient data and of other potential risk factors for acute pancreatitis, the underlying complication risk in the anticoagulation group and the non-anticoagulation group was comparable. Female patients, who have an increased risk for acute post-ES pancreatitis, received low-dose anticoagulation significantly more often than male patients, and none of the female patients with low-dose anticoagulation developed post-ES pancreatitis. Patients who were older than 60 received low-dose anticoagulation in one-third of cases. Low-dose anticoagulation therapy may be responsible for the low incidence of post-ES pancreatitis in the older patients. In the younger patient population, acute pancreatitis was not observed in the anticoagulation group. Although this was not a randomized trial, the data suggest that there is a protective effect against the occurrence of acute post-ES pancreatitis. In this series, the frequency of hemorrhage, its severity, and the resulting mortality rate were not increased by anticoagulation therapy.
There are some experimental data suggesting two different possible etiopathogenetic principles that might explain the protective effect of low-dose anticoagulation against the occurrence of post-ES pancreatitis. Firstly, heparin appears to have a directly inhibitory effect on pancreatic proteases [50 - 52]; secondly, heparin improves the microcirculation in pancreatic tissue by inhibiting proteases of the coagulation cascade [50] [53] . Microcirculation disturbances may play an important role in the course of acute pancreatitis, since fibrin thrombi were found in the capillaries of eight patients, and organized thromboses in the veins of six patients, in a Japanese histopathological study of nine patients with fat necrosis in acute pancreatitis [54].
#Conclusions
The preliminary data provided by this exploratory analysis suggest the following conclusions, although the confirmatory part of this ongoing study has yet to be completed.
Low-dose anticoagulation seems to have a protective effect against the occurrence of acute post-ERCP pancreatitis. A prospective randomized trial appears to be justified.
Patients with coagulopathy, younger age, pancreas divisum, and an absence of chronic pancreatitis are at risk for complications after endoscopic sphincterotomy. However, there are also other potential risk factors that increase the risk of the procedure if they accumulate in a single patient. The endoscopist, and in particular the frequency with which he or she performs ES, is an important risk factor in endoscopic sphincterotomy.
It is indispensable to assess, document, and be aware of the risk factors involved in endoscopic sphincterotomy, in order to evaluate the efficacy of complication-avoiding measures. These risk factors have a significant influence on the complication rate, and such assessment and documentation should allow scientific results to be compared.
#Acknowledgements
The authors are grateful to all the endoscopists in our unit who are not named as authors, for their extensive documentation of all sphincterotomy procedures (Dr. J. Hochberger, Dr. S. Mühldorfer, Dr. G. Nusko). We would also like to express our thanks to Prof. M. Grade for translating the manuscript and to B. Framke for additional statistical analysis.
Parts of this study were presented at the Annual Meeting of the American Gastroenterological Association, 11 - 14 May 1997 in Washington, DC and 17 - 20 May 1998 in New Orleans.
#References
- 1 Cotton PB, Lehmann GA, Vennes J, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc. 1991; 37 338-393
- 2 Freeman ML, Nelson DB, Sherman S, et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med. 1996; 335 909-918
- 3 Ell C, Rabenstein T, Ruppert T, et al. 20 Jahre endoskopische Papillotomie - Analyse der Erlanger Erfahrungen bei 2752 Patienten. Dtsch Med Wochenschr. 1995; 120 163-167
- 4 Rabenstein T, Schneider HT, Hahn EG, et al. 25 years of endoscopic sphincterotomy in Erlangen: assessment of experience in the treatment of 3498 patients. Endoscopy. 1998; 30 (Suppl 2) 201
- 5 Huibregtse K. Complications of endoscopic sphincterotomy and their prevention. N Engl J Med. 1996; 335 961-963
- 6 Classen M, Demling L. Endoskopische Sphinkterotomie der Papilla Vateri und Steinextraktion aus dem Ductus choledochus. Dtsch Med Wochenschr. 1974; 99 496-497
- 7 Kawai KY, Ahasaha K, Murakami M. Endoscopic sphincterotomy of the papilla of Vater. Gastrointest Endosc. 1974; 20 148-150
- 8 Riemann JF, Seuberth K, Demling L. Mechanical lithotripsy of common bile duct stones. Gastrointest Endosc. 1984; 30 140-141
- 9 Rabenstein T, Ruppert T, Schneider HT, Hahn EG, Ell C. Benefits and risks of needle-knife papillotomy. Gastrointest Endosc. 1997; 46 207-211
- 10 Rabenstein T, Schneider HT, Nicklas M, Hahn EG, Ell C. Significance of the endoscopist on success and complications of EST: comparison of experienced and learning endoscopists. Gastrointest Endosc. 1999; 50 628-636
- 11 Ell C, Rabenstein T, Schneider HT, et al. Safety and efficacy of pancreatic sphincterotomy in chronic pancreatitis. Gastrointest Endosc. 1998; 48 244-249
- 12 Classen M.
Endoscopic papillotomy. In: Sivak MV (ed). Gastroenterologic endoscopy. Philadelphia; Saunders, 1987: 631-653 - 13 Huibregtse K, Katon RM, Tytgat GNJ. Precut papillotomy via fine needle-knife papillotome: a safe and effective technique. Gastrointest Endosc. 1986; 32 403-405
- 14 Armitage P, Berry G.
Statistical methods in medical research. 3rd ed . Oxford; Blackwell Science 1994: 311 - 15 Hochberg Y, Tamhane AC (eds).
Multiple comparison procedures. New York; Wiley, 1987 - 16 Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariate models. Ann Intern Med. 1993; 118 201-210
- 17 Hosmer DW Jr, Lemeshow S (eds).
Applied logistic regression. New York; Wiley, 1989 - 18 Boender J, Nix GAJJ, de Ridder MAJ, et al. Endoscopic papillotomy for common bile duct stones: factors influencing the complication rate. Endoscopy. 1994; 26 209-216
- 19 Lambert ME, Betts CD, Hill J, et al. Endoscopic sphincterotomy: the whole truth. Br J Surg. 1991; 78 473-476
- 20 Sherman S, Ruffolo TA, Hawes RH, Lehman GA. Complications of endoscopic sphincterotomy: a prospective series with emphasis on the increased risk associated with sphincter of Oddi dysfunction and nondilated bile ducts. Gastroenterology. 1991; 101 1068-1075
- 21 Vaira D, Ainley C, Williams S, et al. Endoscopic sphincterotomy in 1000 consecutive patients. Lancet. 1989; 8660 431-434
- 22 Wojtun S, Gil J, Gietka W, Gil M. Endoscopic sphincterotomy for choledocholithiasis: a prospective single-center study on the short-term and long-term treatment results in 483 patients. Endoscopy. 1997; 29 258-265
- 23 Jowell PS, Baillie J, Branch MS, et al. Quantitative assessment of procedural competence: a prospective study of training in endoscopic retrograde cholangiopancreatography. Ann Intern Med. 1996; 125 983-989
- 24 Jowell PS, Branch MS, Afforti J, et al. At least 180 ERCPs are needed to attain competence in diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 314
- 25 Freeman ML. Complications of endoscopic biliary sphincterotomy: a review. Endoscopy. 1997; 29 228-297
- 26 Schneider HT, May A, Benninger J, et al. Piezoelectric shock wave lithotripsy of pancreatic duct stones. Am J Gastroenterol. 1994; 89 2042-2048
- 27 Cremer M, Devière J, Delhaye M, et al. Stenting in severe chronic pancreatitis: results of medium-term follow-up in seventy-six patients. Endoscopy. 1991; 23 171-176
- 28 Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic pancreatic duct sphincterotomy: indications, technique and analysis of results. Gastrointest Endosc. 1994; 40 592-598
- 29 Smits ME, Rauw EAS, Tytgat GJ, Huibregtse K. Endoscopic treatment of pancreatic stones in patients with chronic pancreatitis. Gastrointest Endosc. 1996; 43 556-560
- 30 Dumonceau JM, Devière J, Le Moine O, et al. Endoscopic pancreatic drainage in chronic pancreatitis associated with ductal stones: long-term results. Gastrointest Endosc. 1996; 43 547-555
- 31 Slot WB, Schoeman MN, Disario JA, et al. Needle-knife sphincterotomy as a precut procedure: a retrospective evaluation of efficacy and complications. Endoscopy. 1996; 28 334-339
- 32 Binmoeller KF, Seifert H, Gerke H, et al. Papillary roof incision using the Erlangen-type pre-cut papillotome to achieve selective bile duct cannulation. Gastrointest Endosc. 1996; 44 689-695
- 33 Dowset JF, Polydorou AA, Vaira D, et al. Needle-knife papillotomy: how safe and how effective?. Gut. 1990; 31 905-908
- 34 Foutch PG. A prospective assessment of results for needle-knife papillotomy and standard endoscopic sphincterotomy. Gastrointest Endosc. 1995; 41 25-32
- 35 Baillie J. Needle-knife papillotomy revisited. Gastrointest Endosc. 1997; 46 282-284
- 36 Siegel HJ. Precut papillotomy: a method to improve success of ERCP and papillotomy. Endoscopy. 1980; 20 130-133
- 37 Siegel HJ, Ben-Zvi JS, Pullano W. The needle-knife: a valuable tool in diagnostic and therapeutic ERCP. Gastrointest Endosc. 1989; 35 499-503
- 38 Vandervoort J, Carr-Locke DL. Needle-knife access papillotomy: an unfairly maligned technique?. Endoscopy. 1996; 28 365-366
- 39 Tham TCK, Vandervoort L, Wong RCK, et al. Therapeutic ERCP in outpatients. Gastrointest Endosc. 1997; 45 225-230
- 40 Vandervoort J, Tham TCK, Wong RCK, et al. Prospective analysis of risk factors for pancreatitis after diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 440
- 41 Vandervoort J, Tham TCK, Wong RCK, et al. Prospective study of post-ERCP complications following diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 441
- 42 Booth FV, Doerr RJ, Khalafi RS, et al. Surgical management of complications of endoscopic sphincterotomy with precut papillotomy. Am J Surg. 1990; 159 132-135
- 43 Tweedle DEF, Martin FF. Needle-knife papillotomy for endoscopic sphincterotomy and cholangiography. Gastrointest Endosc. 1991; 37 518-521
- 44 Shakoor T, Geenen JE. Precut papillotomy. Gastrointest Endosc. 1992; 38 623-627
- 45 Cotton PB. Precut papillotomy: a risky technique for experts only. Gastrointest Endosc. 1989; 35 578-579
- 46 Chen YK, Foliente RL, Santoro MJ, et al. Endoscopic sphincterotomy-induced pancreatitis: increased risk associated with nondilated bile ducts and sphincter of Oddi dysfunction. Am J Gastroenterol. 1994; 89 327-333
- 47 Sherman S, Lehmann GA. ERCP and endoscopic sphincterotomy-induced pancreatitis. Pancreas. 1991; 6 350-67
- 48 Geenen JE, Hogan WJ, Dodds WJ, et al. Long-term results of endoscopic sphincterotomy for treating patients with sphincter of Oddi dysfunction. Gastroenterology. 1989; 92 1401-1405
- 49 Tittobello A. Diagnosis and prevention of post-EST pancreatitis. Endoscopy. 1997; 29 285-287
- 50 Toulon P, Chadeuf G, Bouillot JL, et al. Involvement of heparin cofactor II in chymotrypsin neutralization and in pancreatic proteinase-antiproteinase interaction during acute pancreatitis in man. Eur J Clin Invest. 1991; 21 303-309
- 51 Kosidlo S, Gabryelewicz A, Prokopowicz J, Prokopowicz K. Pancreatic proteolytic and inhibitory activity during acute experimental pancreatitis in rats with reference to heparin treatment. Mater Med Pol. 1986; 18 136-143
- 52 Gabryelewicz A, Kosidlo S, Prokopowicz J, Prokopowicz K. Does heparin modify protease-antiprotease balance in acute experimental pancreatitis in rats?. Hepatogastroenterology. 1986; 33 79-92
- 53 Bleeker WK, Agterberg J, Rigter G, et al. Protective effect of antithrombin III in acute experimental pancreatitis in rats. Dig Dis Sci. 1992; 37 280-285
- 54 Suda K. Histopathological study on fat necrosis with comparison between agonal stage and acute pancreatitis. Rinsho Byori. 1991; 39 977-980
M.D. T. Rabenstein
Medizinische Klinik I mit Poliklinik Universität Erlangen-Nürnberg
Krankenhausstrasse 12
91054 Erlangen, Germany
Phone: +49-9131-8536909
Email: Thomas.Rabenstein@med1.med.uni-erlangen.de
References
- 1 Cotton PB, Lehmann GA, Vennes J, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc. 1991; 37 338-393
- 2 Freeman ML, Nelson DB, Sherman S, et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med. 1996; 335 909-918
- 3 Ell C, Rabenstein T, Ruppert T, et al. 20 Jahre endoskopische Papillotomie - Analyse der Erlanger Erfahrungen bei 2752 Patienten. Dtsch Med Wochenschr. 1995; 120 163-167
- 4 Rabenstein T, Schneider HT, Hahn EG, et al. 25 years of endoscopic sphincterotomy in Erlangen: assessment of experience in the treatment of 3498 patients. Endoscopy. 1998; 30 (Suppl 2) 201
- 5 Huibregtse K. Complications of endoscopic sphincterotomy and their prevention. N Engl J Med. 1996; 335 961-963
- 6 Classen M, Demling L. Endoskopische Sphinkterotomie der Papilla Vateri und Steinextraktion aus dem Ductus choledochus. Dtsch Med Wochenschr. 1974; 99 496-497
- 7 Kawai KY, Ahasaha K, Murakami M. Endoscopic sphincterotomy of the papilla of Vater. Gastrointest Endosc. 1974; 20 148-150
- 8 Riemann JF, Seuberth K, Demling L. Mechanical lithotripsy of common bile duct stones. Gastrointest Endosc. 1984; 30 140-141
- 9 Rabenstein T, Ruppert T, Schneider HT, Hahn EG, Ell C. Benefits and risks of needle-knife papillotomy. Gastrointest Endosc. 1997; 46 207-211
- 10 Rabenstein T, Schneider HT, Nicklas M, Hahn EG, Ell C. Significance of the endoscopist on success and complications of EST: comparison of experienced and learning endoscopists. Gastrointest Endosc. 1999; 50 628-636
- 11 Ell C, Rabenstein T, Schneider HT, et al. Safety and efficacy of pancreatic sphincterotomy in chronic pancreatitis. Gastrointest Endosc. 1998; 48 244-249
- 12 Classen M.
Endoscopic papillotomy. In: Sivak MV (ed). Gastroenterologic endoscopy. Philadelphia; Saunders, 1987: 631-653 - 13 Huibregtse K, Katon RM, Tytgat GNJ. Precut papillotomy via fine needle-knife papillotome: a safe and effective technique. Gastrointest Endosc. 1986; 32 403-405
- 14 Armitage P, Berry G.
Statistical methods in medical research. 3rd ed . Oxford; Blackwell Science 1994: 311 - 15 Hochberg Y, Tamhane AC (eds).
Multiple comparison procedures. New York; Wiley, 1987 - 16 Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariate models. Ann Intern Med. 1993; 118 201-210
- 17 Hosmer DW Jr, Lemeshow S (eds).
Applied logistic regression. New York; Wiley, 1989 - 18 Boender J, Nix GAJJ, de Ridder MAJ, et al. Endoscopic papillotomy for common bile duct stones: factors influencing the complication rate. Endoscopy. 1994; 26 209-216
- 19 Lambert ME, Betts CD, Hill J, et al. Endoscopic sphincterotomy: the whole truth. Br J Surg. 1991; 78 473-476
- 20 Sherman S, Ruffolo TA, Hawes RH, Lehman GA. Complications of endoscopic sphincterotomy: a prospective series with emphasis on the increased risk associated with sphincter of Oddi dysfunction and nondilated bile ducts. Gastroenterology. 1991; 101 1068-1075
- 21 Vaira D, Ainley C, Williams S, et al. Endoscopic sphincterotomy in 1000 consecutive patients. Lancet. 1989; 8660 431-434
- 22 Wojtun S, Gil J, Gietka W, Gil M. Endoscopic sphincterotomy for choledocholithiasis: a prospective single-center study on the short-term and long-term treatment results in 483 patients. Endoscopy. 1997; 29 258-265
- 23 Jowell PS, Baillie J, Branch MS, et al. Quantitative assessment of procedural competence: a prospective study of training in endoscopic retrograde cholangiopancreatography. Ann Intern Med. 1996; 125 983-989
- 24 Jowell PS, Branch MS, Afforti J, et al. At least 180 ERCPs are needed to attain competence in diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 314
- 25 Freeman ML. Complications of endoscopic biliary sphincterotomy: a review. Endoscopy. 1997; 29 228-297
- 26 Schneider HT, May A, Benninger J, et al. Piezoelectric shock wave lithotripsy of pancreatic duct stones. Am J Gastroenterol. 1994; 89 2042-2048
- 27 Cremer M, Devière J, Delhaye M, et al. Stenting in severe chronic pancreatitis: results of medium-term follow-up in seventy-six patients. Endoscopy. 1991; 23 171-176
- 28 Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic pancreatic duct sphincterotomy: indications, technique and analysis of results. Gastrointest Endosc. 1994; 40 592-598
- 29 Smits ME, Rauw EAS, Tytgat GJ, Huibregtse K. Endoscopic treatment of pancreatic stones in patients with chronic pancreatitis. Gastrointest Endosc. 1996; 43 556-560
- 30 Dumonceau JM, Devière J, Le Moine O, et al. Endoscopic pancreatic drainage in chronic pancreatitis associated with ductal stones: long-term results. Gastrointest Endosc. 1996; 43 547-555
- 31 Slot WB, Schoeman MN, Disario JA, et al. Needle-knife sphincterotomy as a precut procedure: a retrospective evaluation of efficacy and complications. Endoscopy. 1996; 28 334-339
- 32 Binmoeller KF, Seifert H, Gerke H, et al. Papillary roof incision using the Erlangen-type pre-cut papillotome to achieve selective bile duct cannulation. Gastrointest Endosc. 1996; 44 689-695
- 33 Dowset JF, Polydorou AA, Vaira D, et al. Needle-knife papillotomy: how safe and how effective?. Gut. 1990; 31 905-908
- 34 Foutch PG. A prospective assessment of results for needle-knife papillotomy and standard endoscopic sphincterotomy. Gastrointest Endosc. 1995; 41 25-32
- 35 Baillie J. Needle-knife papillotomy revisited. Gastrointest Endosc. 1997; 46 282-284
- 36 Siegel HJ. Precut papillotomy: a method to improve success of ERCP and papillotomy. Endoscopy. 1980; 20 130-133
- 37 Siegel HJ, Ben-Zvi JS, Pullano W. The needle-knife: a valuable tool in diagnostic and therapeutic ERCP. Gastrointest Endosc. 1989; 35 499-503
- 38 Vandervoort J, Carr-Locke DL. Needle-knife access papillotomy: an unfairly maligned technique?. Endoscopy. 1996; 28 365-366
- 39 Tham TCK, Vandervoort L, Wong RCK, et al. Therapeutic ERCP in outpatients. Gastrointest Endosc. 1997; 45 225-230
- 40 Vandervoort J, Tham TCK, Wong RCK, et al. Prospective analysis of risk factors for pancreatitis after diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 440
- 41 Vandervoort J, Tham TCK, Wong RCK, et al. Prospective study of post-ERCP complications following diagnostic and therapeutic ERCP [abstract]. Gastrointest Endosc. 1996; 43 441
- 42 Booth FV, Doerr RJ, Khalafi RS, et al. Surgical management of complications of endoscopic sphincterotomy with precut papillotomy. Am J Surg. 1990; 159 132-135
- 43 Tweedle DEF, Martin FF. Needle-knife papillotomy for endoscopic sphincterotomy and cholangiography. Gastrointest Endosc. 1991; 37 518-521
- 44 Shakoor T, Geenen JE. Precut papillotomy. Gastrointest Endosc. 1992; 38 623-627
- 45 Cotton PB. Precut papillotomy: a risky technique for experts only. Gastrointest Endosc. 1989; 35 578-579
- 46 Chen YK, Foliente RL, Santoro MJ, et al. Endoscopic sphincterotomy-induced pancreatitis: increased risk associated with nondilated bile ducts and sphincter of Oddi dysfunction. Am J Gastroenterol. 1994; 89 327-333
- 47 Sherman S, Lehmann GA. ERCP and endoscopic sphincterotomy-induced pancreatitis. Pancreas. 1991; 6 350-67
- 48 Geenen JE, Hogan WJ, Dodds WJ, et al. Long-term results of endoscopic sphincterotomy for treating patients with sphincter of Oddi dysfunction. Gastroenterology. 1989; 92 1401-1405
- 49 Tittobello A. Diagnosis and prevention of post-EST pancreatitis. Endoscopy. 1997; 29 285-287
- 50 Toulon P, Chadeuf G, Bouillot JL, et al. Involvement of heparin cofactor II in chymotrypsin neutralization and in pancreatic proteinase-antiproteinase interaction during acute pancreatitis in man. Eur J Clin Invest. 1991; 21 303-309
- 51 Kosidlo S, Gabryelewicz A, Prokopowicz J, Prokopowicz K. Pancreatic proteolytic and inhibitory activity during acute experimental pancreatitis in rats with reference to heparin treatment. Mater Med Pol. 1986; 18 136-143
- 52 Gabryelewicz A, Kosidlo S, Prokopowicz J, Prokopowicz K. Does heparin modify protease-antiprotease balance in acute experimental pancreatitis in rats?. Hepatogastroenterology. 1986; 33 79-92
- 53 Bleeker WK, Agterberg J, Rigter G, et al. Protective effect of antithrombin III in acute experimental pancreatitis in rats. Dig Dis Sci. 1992; 37 280-285
- 54 Suda K. Histopathological study on fat necrosis with comparison between agonal stage and acute pancreatitis. Rinsho Byori. 1991; 39 977-980
M.D. T. Rabenstein
Medizinische Klinik I mit Poliklinik Universität Erlangen-Nürnberg
Krankenhausstrasse 12
91054 Erlangen, Germany
Phone: +49-9131-8536909
Email: Thomas.Rabenstein@med1.med.uni-erlangen.de