Outcomes research in endoscopy: current status and future directions

Abstract

The heightened interest among endoscopists in outcomes and effectiveness research during the past several years has paralleled that of the medical community at large. Outcomes research has been the subject of a 1992 ASGE Investigators Conference,1Keeffe EB McMahon Jr, LF Outcomes and effectiveness research: the 1992 investigator conference of the American Society for Gastrointestinal Endoscopy.Gastrointest Endosc. 1993; 39: 840-845Abstract Full Text PDF PubMed Scopus (10) Google Scholar at least two editorials in this journal,2Rabeneck L Why should gastroenterologists know about outcomes research?.Gastrointest Endosc. 1993; 39: 723-725Abstract Full Text PDF PubMed Scopus (5) Google Scholar, 3Brazer S Ensuring quality care for patients with digestive diseases.Gastrointest Endosc. 1996; 44: 204-207Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar a postgraduate course in surgical endoscopy,4Traverso LW Surgical endoscopy: outcomes and costs.Surg Endosc. 1995; 9: 1211-1248Crossref PubMed Google Scholar and two review articles in Gastroenterology.5American Gastroenterological Association Task Force on Outcomes Research A primer on outcomes research for the gastroenterologist.Gastroenterology. 1995; 109: 302-306Abstract Full Text PDF PubMed Scopus (26) Google Scholar, 6Bouchard S Barkun AN Barkun JS Joseph L Technology assessment in laparoscopic general surgery and gastrointestinal endoscopy: science or convenience?.Gastroenterology. 1996; 110: 915-925Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar Moreover, the American Digestive Health Foundation has committed significant financial resources to training and research in outcomes analysis in gastroenterology. Despite this outpouring of interest and commitment to outcomes research, there has been a relative paucity of studies in the endoscopic literature. Indeed, 44% of a sample of journal subscribers believed that there were not enough published endoscopic outcomes studies (personal communication, Michael V. Sivak Jr., MD).Outcomes research is an area of investigation that examines the effectiveness of medical interventions in everyday practice. Although it asks many of the same questions and employs similar research methods as traditional clinical research, several general distinctions can be made. First, outcomes research is generally observational and only occasionally involves randomized trials. Second, outcomes research examines questions of effectiveness (i.e., the impact of an intervention in unselected patients receiving standard medical care) as opposed to efficacy (i.e., impact of an intervention in a tightly controlled research protocol). Third, outcomes studies typically have less stringent entry criteria and examine patient populations with a more diverse spectrum of illness. Fourth, outcomes researchers often have less control over the intervention being studied and data are often collected for purposes other than research. Fifth, because data collection costs are often less, analyses frequently encompass thousands to even millions of patients. Finally, because patients are generally not randomly assigned to treatments, the outcomes researcher often depends on multivariable statistical techniques to adjust for baseline differences between patient groups.The current enthusiasm for outcomes research can be traced to several factors. The recent emphasis on cost containment has led to fears that quality of care may be adversely affected, and outcomes research provides a quantitative basis for examining potential trade-off between cost and quality. Health care has also become increasingly competitive and purchasers clearly need information on both price and patient outcome to make informed decisions. Recent advances in computer technology have led to health care databases that can aggregate and analyze medical care on a massive scale. Last, researchers who have focused on practice patterns have found enormous variation in the use of procedures, including gastrointestinal endoscopy.7Chassin MR Kosecoff J Park RE Winslow CM Kahn KL Merrick NJ et al.Does inappropriate use explain geographic variations in the use health care services?.JAMA. 1987; 258: 2533-2537Crossref PubMed Scopus (538) Google Scholar, 8Greenwald BD Mentnech RM Upper GI endoscopy in the United States: significantly different geographic practice patterns [abstract].Gastroenterology. 1994; 106: A9Google Scholar For example, 50% to 70% of differences in the rate of upper endoscopy were observed between communities7Chassin MR Kosecoff J Park RE Winslow CM Kahn KL Merrick NJ et al.Does inappropriate use explain geographic variations in the use health care services?.JAMA. 1987; 258: 2533-2537Crossref PubMed Scopus (538) Google Scholar and large regions.8Greenwald BD Mentnech RM Upper GI endoscopy in the United States: significantly different geographic practice patterns [abstract].Gastroenterology. 1994; 106: A9Google Scholar These findings have led to further studies to determine the reasons for this variation and have led health policy makers to ask “Which rate is correct?”Although the interest in endoscopic outcomes and effectiveness studies has significantly increased over the past few years, there is still great demand for additional research, given the high volume and aggregate costs of endoscopy, the rapidly evolving technology, and skepticism by some concerning the potential benefit of certain procedures. Furthermore, although a broad range of outcomes studies is desirable, there is particular need for investigations in five key areas: (1) defining the impact of endoscopic procedures in routine community practice (i.e., effectiveness); (2) measuring the benefit of endoscopy in relation to nontraditional endpoints (i.e., quality of life measures); (3) comparing endoscopic outcomes across groups of providers; (4) assessing the accuracy of both clinical and claims-based endoscopic terminology; and (5) developing and validating disease-specific severity of illness measures.Additional effectiveness studiesThrough the use of carefully designed randomized clinical trials, the benefit of several endoscopic treatments has been established. These include electrocautery and injection of bleeding ulcers,9Cook DJ Guyatt GH Salena BJ Laine LA Endoscopic therapy for acute nonvariceal upper gastrointestinal hemorrhage: a meta-analysis.Gastroenterology. 1992; 102: 139-148PubMed Google Scholar, 10Sacks HS Chalmers TC Blum AL Berrier J Pagano D Endoscopic hemostasis: an effective therapy for bleeding peptic ulcers.JAMA. 1990; 264: 494-499Crossref PubMed Scopus (304) Google Scholar variceal sclerotherapy and banding,11Laine L Cook D Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis.Ann Intern Med. 1995; 123: 280-287Crossref PubMed Scopus (606) Google Scholar and endoscopic sphincterotomy for cholangitis12Lai ECS Mok FPT Tan ESY Lo C Fan S You K et al.Endoscopic biliary drainage for severe acute cholangitis.N Engl J Med. 1992; 326: 1582-1586Crossref PubMed Scopus (474) Google Scholar and gallstone pancreatitis.13Neoptolemos JP London NJ James D Carr-Locke DL Bailey IA Fossard DP Controlled trial of urgent endoscopic retrograde cholangiopancreatography and endoscopic sphincterotomy versus conservative treatment for acute pancreatitis due to gallstones.Lancet. 1988; 2: 979-983Abstract PubMed Scopus (784) Google Scholar, 14Fan S Lai ECS Mok FPT Lo C Zheng S Wong J Early treatment of acute biliary pancreatitis by endoscopic papillotomy.N Engl J Med. 1993; 328: 228-232Crossref PubMed Scopus (719) Google Scholar However, because of potential differences in patient populations due to the typically narrow entry and exclusion criteria of randomized trials and/or case mix served by a specific hospital, analysis of the intervention in everyday practice may yield disparate results. For example, because many cases of biliary obstruction in the Far East are due to brown pigmented intrahepatic stones or Ascaris infection (whereas primary cholesterol stones are common in an American population), the effectiveness of routine urgent ERCP with sphincterotomy may differ. Moreover, as technical skills in endoscopy may vary substantially, the relative benefit of certain endoscopic procedures may be determined more by operator expertise than other clinical factors. Thus, the impact of endoscopy performed by the experts who typically conduct large trials may differ widely from that of routine practice. Current community-based projects in Cleveland15Cooper GS Chak A Hammar PJ Way LE Rosenthal GE The effectiveness of endoscopy in patients with upper gastrointestinal hemorrhage: a community-based study [abstract].Gastrointest Endosc. 1997; 45: AB89Google Scholar and Portland,16Fennerty MB Lieberman DA Magaret N Effectiveness of H. pylori treatment regimens in clinical practice: a community-based outcomes study [abstract].Gastroenterology. 1997; 112: A14Google Scholar as well as the multicenter Clinical Outcomes Research Initiative (CORI),17Lieberman DA Fleischer DE Eisen GM Helfand M Clinical outcomes research initiative: preliminary report of a clinical practice consortium [abstract].Gastrointest Endosc. 1997; 45: AB52Abstract Full Text PDF Scopus (3) Google Scholar are worthwhile efforts to collect data on routine endoscopic practice and outcomes, and these projects will potentially provide important effectiveness data. Finally, given the increasing emphasis on outpatient management for many common acute illnesses, sites for data collection should ideally include ambulatory endoscopy facilities as well as hospital-based units.Broadening the range of outcomesA second priority area of inquiry is the impact of endoscopic procedures on additional end points. Whereas comparison of mortality rates between alternative therapies often provides important data, for many conditions these measures may not be the most appropriate outcomes to study.18Cotton PB Therapeutic gastrointestinal endoscopy: problems in proving efficacy.N Engl J Med. 1992; 326: 1626-1628Crossref PubMed Scopus (19) Google Scholar For example, it is unlikely that endoscopic procedures alone will substantively affect survival for rapidly fatal diseases such as unresectable esophageal or pancreatic malignancies or for rarely terminal conditions such as erosive esophagitis. However, compared with surgical or empiric drug therapy, endoscopy may be associated with important benefits in patient quality of life, including symptomatology, need for additional procedures and hospitalization, and functional status. Where feasible, these measures should be incorporated as end points into endoscopic studies. Valid and reliable disease-specific quality of life measures have already been developed for inflammatory bowel disease19Irvine EJ Feagan B Rochon J Archambault A Fedorak RN Groll A et al.Quality of life: a valid and reliable measure of therapeutic efficacy in the treatment of inflammatory bowel disease.Gastroenterology. 1994; 106: 287-296Abstract PubMed Google Scholar and other gastrointestinal conditions, but are needed for all the common disorders treated in endoscopic practice. Also, because of increasing concern over costs of medical care, comparison of expenditures during treatment and follow-up may provide additional data on the benefit of endoscopic procedures. Several cost-effectiveness analyses comparing endoscopic management to other strategies have been performed,20Erikson RA Carlson B The role of endoscopic retrograde cholangiopancreatography in patients with laparoscopic cholecystectomies.Gastroenterology. 1995; 109: 262-263Google Scholar, 21Fendrick AM Chernew ME Hirth RA Bloom BS Alternative management strategies for patients with suspected peptic ulcer disease.Ann Intern Med. 1995; 123: 260-268Crossref PubMed Scopus (200) Google Scholar, 22Silverstein MD Petterson T Talley NJ Initial endoscopy or empirical therapy with or without testing for Helicobacter pylori for dyspepsia: a decision analysis.Gastroenterology. 1996; 110: 72-83Abstract Full Text PDF PubMed Scopus (175) Google Scholar, 23Offman JJ Etchason J Fullerton S Kahn KL Soll AH Management strategies for Helicobacter pylori seropositive patients with dyspepsia: clinical and economic consequences.Ann Intern Med. 1997; 126: 280-291Crossref PubMed Scopus (162) Google Scholar, 24Provenzale D Kowdley KV Arora S Wong JB Prophylactic colectomy or surveillance for chronic ulcerative colitis?.Gastroenterology. 1995; 109: 1188-1196Abstract Full Text PDF PubMed Scopus (102) Google Scholar but data on actual expenditures in everyday practice would also be valuable.Studies of practice variationAn important contribution of outcomes research has been the demonstration of substantial variation in outcome across providers. It is unlikely that gastrointestinal endoscopy will emulate coronary artery bypass surgery in New York25Green J Wintfeld N Report cards on cardiac surgeons.N Engl J Med. 1995; 332: 1229-1232Crossref PubMed Scopus (310) Google Scholar and Pennsylvania26Schneider EC Epstein AM Influence of cardiac surgery performance reports on referral practices and access to care.N Engl J Med. 1996; 3335: 251-256Crossref Scopus (363) Google Scholar with publication of individual hospital and physician statistics, but other more general comparisons would be of interest. For example, it is known that general internists and family practitioners perform approximately 15% to 20% of upper endoscopies and colonoscopies in the United States.27Meyer GS Jacoby I Krakauer H Powell DW Aurand J McCardle P Gastroenterology workforce modeling.JAMA. 1996; 276: 689-694Crossref PubMed Google Scholar, 28Ackermann RJ Performance of gastrointestinal tract endoscopy by primary care physicians.Arch Fam Med. 1997; 6: 52-58Crossref PubMed Scopus (15) Google Scholar Unfortunately, there are no data comparing endoscopic outcomes, including procedural indications, success and complication rates, and appropriate use of therapeutic maneuvers, between different types of providers. For example, post-ERCP complication rates have been shown to vary substantially depending on level of the endoscopic experience,29Freeman ML Nelson DB Sherman S Haber GB Herman ME Dorsher PJ et al.Complications of endoscopic sphincterotomy.N Engl J Med. 1996; 335: 909-918Crossref PubMed Scopus (2147) Google Scholar and missed diagnosis rates for cancer at colonoscopy vary between gastroenterologists and other examiners.30Rex DK Rahmani EY Haseman JH Lemmel GT Kaster S Buckley JS Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice.Gastroenterology. 1997; 112: 17-23Abstract Full Text PDF PubMed Scopus (515) Google Scholar Thus, similar studies for other common procedures and diseases would also be important. Similarly, for highly specialized procedures such as endoscopic ultrasonography, outcomes data are needed to support the assertion that the procedure should be restricted to referral centers.31Wang KK DiMagno EP Endoscopic ultrasonography: high technology and cost containment.Gastroenterology. 1993; 105: 283-286PubMed Google ScholarThe use and outcome of endoscopic procedures should also be compared between hospital types, including teaching, nonteaching, nonprofit and for profit, as well as between practice settings, including fee for service and capitated. For example, comparisons between teaching and nonteaching hospitals in Cleveland have shown significant differences in the use of endoscopic therapy for bleeding peptic ulcers, yet few differences in outcome.32Cooper GS Chak A Hammar PJ Way LE Rosenthal GE Treatment differences between teaching and nonteaching hospitals in patients with upper gastrointestinal hemorrhage [abstract].Gastrointest Endosc. 1997; 45: AB89Google Scholar Similarly, the potential for differences in outcomes among patients referred for open access endoscopy in community practice33Froehlich F Burnand B Pache I Vader J-P Fried M Schneider C et al.Overuse of upper gastrointestinal endoscopy in a country with open-access endoscopy: a prospective study in primary care.Gastrointest Endosc. 1997; 45: 13-19Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar, 34Hungin APS Thomas PR Bramble MG Corbett WA Idle N Contractor BR et al.What happens to patients following open access gastroscopy? An outcome study from general practice.Br J Gen Pract. 1994; 44: 519-521PubMed Google Scholar should also be investigated. Further studies are thus needed to better define the impact, if any, of practice environment on outcome.Standardization and validation of terminologyEndoscopic studies have often included multiple sites and endoscopists as a means to enhance external validity (generalizability) as well as to ensure adequate numbers of study subjects. Although multicenter studies have clear advantages, there are potential problems with standardization of endoscopic terminology. Investigators in diverse settings may have extremely divergent definitions for lesions such as “erosion” or “gastritis,” as well as potential difficulties in defining the location of lesions.35Crespi M Delvaux M Schapiro M Venables C Zwiebel F Minimal standards for a computerized endoscopic database.Am J Gastroenterol. 1994; 89: S144-S153PubMed Google Scholar, 36Muller AD Sonnenberg A Prevention of colorectal cancer by flexible endoscopy and polypectomy.Ann Intern Med. 1995; 123: 904-910Crossref PubMed Scopus (561) Google Scholar Moreover, there may be a high level of interobserver variability for findings such as “visible vessel” or “active bleeding.” The variability in the use of terminology is potentially more problematic for data that are not collected for predesigned clinical trials. In that regard, efforts by the European Society of Gastrointestinal Endoscopy35Crespi M Delvaux M Schapiro M Venables C Zwiebel F Minimal standards for a computerized endoscopic database.Am J Gastroenterol. 1994; 89: S144-S153PubMed Google Scholar and the ASGE to develop standardized endoscopic terminology are important.Outcomes investigators have also used large health claims databases, such as those maintained by Medicare and the Veterans Administration.36Muller AD Sonnenberg A Prevention of colorectal cancer by flexible endoscopy and polypectomy.Ann Intern Med. 1995; 123: 904-910Crossref PubMed Scopus (561) Google Scholar, 37Rabeneck L Wray NP Petersen NJ Long-term outcomes of patients receiving percutaneous endoscopic gastrostomy tubes.J Gen Intern Med. 1996; 11: 287-293Crossref PubMed Scopus (213) Google Scholar Although attractive because of the large sample sizes, small incremental research costs, and accurate representation of routine clinical practice that they offer, there are potential questions about the accuracy of these files for coding of endoscopic procedures. The sensitivity and positive predictive value of these codes for surgical procedures are typically 90% to 100%, but similar data for endoscopy are lacking. In a small sample of California discharge abstracts, the sensitivity and specificity for coding of colonoscopy were 91% and 100%, respectively,38Romano PS Luft HS Getting the most out of messy data: problems and approaches for dealing with large administrative data sets.in: Medical effectiveness research data methods. Agency for Health Care Policy and Research, Rockville, Maryland1992: 57-70Google Scholar but other procedures were not examined. Moreover, the reliability of coding for specific subtypes of procedures (i.e., upper endoscopy with control of hemorrhage, colonoscopy with polypectomy) is also not defined. Future investigations should assess the utility and limitations of both the International Classification of Diseases (ICD)-9 and Current Procedural Terminology (CPT)-4 coding systems for endoscopic procedures.Adequate severity adjustmentWhen patient outcomes are evaluated and compared across providers, there has to be adequate adjustment for severity of illness. The use of severity measures allows for appropriate adjustments for differences in case mix, because the sickest patients tend not to be evenly distributed. Hence, severity measures should be generalizable to different practice settings (i.e., academic and community practices) and include data that are routinely collected. Commonly used measures are often generic for all medical-surgical diagnoses, and are either code-based, including demographic characteristics and ICD-9 codes, or chart-based, such as clinical and laboratory data.39Hughes JS Iezzoni LI Daley J Greenberg L How severity measures rate hospitalized patients.J Gen Intern Med. 1996; 11: 303-311Crossref PubMed Scopus (31) Google Scholar Unfortunately, measures that were developed to predict inhospital mortality for intensive care units (APACHE scores) or hospital charges (APR-DRG) are probably not the best methods for risk-adjusting a heterogeneous cohort of patients who have undergone endoscopy. Thus, in assessing endoscopic outcomes, there is a need for additional disease-specific measures, which should ideally include key endoscopic findings as well as relevant clinical data. Because variables may not have the same relation with every outcome of interest, it may also be necessary to develop more than one severity measure.In conclusion, outcomes and effectiveness research, which has been termed the “third revolution in medical care,”40Relman AS Assessment and accountability: the third revolution in medical care.N Engl J Med. 1988; 319: 1120-1122Google Scholar is critically important for demonstrating the benefits of endoscopic procedures to patients, providers, and the medical community at large. To acquire this critical knowledge and information, there must be continued training of outcomes investigators, increased funding for such studies by the gastroenterology societies, and the performance of well designed studies, with emphasis on the five priority areas that have been identified. The heightened interest among endoscopists in outcomes and effectiveness research during the past several years has paralleled that of the medical community at large. Outcomes research has been the subject of a 1992 ASGE Investigators Conference,1Keeffe EB McMahon Jr, LF Outcomes and effectiveness research: the 1992 investigator conference of the American Society for Gastrointestinal Endoscopy.Gastrointest Endosc. 1993; 39: 840-845Abstract Full Text PDF PubMed Scopus (10) Google Scholar at least two editorials in this journal,2Rabeneck L Why should gastroenterologists know about outcomes research?.Gastrointest Endosc. 1993; 39: 723-725Abstract Full Text PDF PubMed Scopus (5) Google Scholar, 3Brazer S Ensuring quality care for patients with digestive diseases.Gastrointest Endosc. 1996; 44: 204-207Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar a postgraduate course in surgical endoscopy,4Traverso LW Surgical endoscopy: outcomes and costs.Surg Endosc. 1995; 9: 1211-1248Crossref PubMed Google Scholar and two review articles in Gastroenterology.5American Gastroenterological Association Task Force on Outcomes Research A primer on outcomes research for the gastroenterologist.Gastroenterology. 1995; 109: 302-306Abstract Full Text PDF PubMed Scopus (26) Google Scholar, 6Bouchard S Barkun AN Barkun JS Joseph L Technology assessment in laparoscopic general surgery and gastrointestinal endoscopy: science or convenience?.Gastroenterology. 1996; 110: 915-925Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar Moreover, the American Digestive Health Foundation has committed significant financial resources to training and research in outcomes analysis in gastroenterology. Despite this outpouring of interest and commitment to outcomes research, there has been a relative paucity of studies in the endoscopic literature. Indeed, 44% of a sample of journal subscribers believed that there were not enough published endoscopic outcomes studies (personal communication, Michael V. Sivak Jr., MD). Outcomes research is an area of investigation that examines the effectiveness of medical interventions in everyday practice. Although it asks many of the same questions and employs similar research methods as traditional clinical research, several general distinctions can be made. First, outcomes research is generally observational and only occasionally involves randomized trials. Second, outcomes research examines questions of effectiveness (i.e., the impact of an intervention in unselected patients receiving standard medical care) as opposed to efficacy (i.e., impact of an intervention in a tightly controlled research protocol). Third, outcomes studies typically have less stringent entry criteria and examine patient populations with a more diverse spectrum of illness. Fourth, outcomes researchers often have less control over the intervention being studied and data are often collected for purposes other than research. Fifth, because data collection costs are often less, analyses frequently encompass thousands to even millions of patients. Finally, because patients are generally not randomly assigned to treatments, the outcomes researcher often depends on multivariable statistical techniques to adjust for baseline differences between patient groups. The current enthusiasm for outcomes research can be traced to several factors. The recent emphasis on cost containment has led to fears that quality of care may be adversely affected, and outcomes research provides a quantitative basis for examining potential trade-off between cost and quality. Health care has also become increasingly competitive and purchasers clearly need information on both price and patient outcome to make informed decisions. Recent advances in computer technology have led to health care databases that can aggregate and analyze medical care on a massive scale. Last, researchers who have focused on practice patterns have found enormous variation in the use of procedures, including gastrointestinal endoscopy.7Chassin MR Kosecoff J Park RE Winslow CM Kahn KL Merrick NJ et al.Does inappropriate use explain geographic variations in the use health care services?.JAMA. 1987; 258: 2533-2537Crossref PubMed Scopus (538) Google Scholar, 8Greenwald BD Mentnech RM Upper GI endoscopy in the United States: significantly different geographic practice patterns [abstract].Gastroenterology. 1994; 106: A9Google Scholar For example, 50% to 70% of differences in the rate of upper endoscopy were observed between communities7Chassin MR Kosecoff J Park RE Winslow CM Kahn KL Merrick NJ et al.Does inappropriate use explain geographic variations in the use health care services?.JAMA. 1987; 258: 2533-2537Crossref PubMed Scopus (538) Google Scholar and large regions.8Greenwald BD Mentnech RM Upper GI endoscopy in the United States: significantly different geographic practice patterns [abstract].Gastroenterology. 1994; 106: A9Google Scholar These findings have led to further studies to determine the reasons for this variation and have led health policy makers to ask “Which rate is correct?” Although the interest in endoscopic outcomes and effectiveness studies has significantly increased over the past few years, there is still great demand for additional research, given the high volume and aggregate costs of endoscopy, the rapidly evolving technology, and skepticism by some concerning the potential benefit of certain procedures. Furthermore, although a broad range of outcomes studies is desirable, there is particular need for investigations in five key areas: (1) defining the impact of endoscopic procedures in routine community practice (i.e., effectiveness); (2) measuring the benefit of endoscopy in relation to nontraditional endpoints (i.e., quality of life measures); (3) comparing endoscopic outcomes across groups of providers; (4) assessing the accuracy of both clinical and claims-based endoscopic terminology; and (5) developing and validating disease-specific severity of illness measures. Additional effectiveness studiesThrough the use of carefully designed randomized clinical trials, the benefit of several endoscopic treatments has been established. These include electrocautery and injection of bleeding ulcers,9Cook DJ Guyatt GH Salena BJ Laine LA Endoscopic therapy for acute nonvariceal upper gastrointestinal hemorrhage: a meta-analysis.Gastroenterology. 1992; 102: 139-148PubMed Google Scholar, 10Sacks HS Chalmers TC Blum AL Berrier J Pagano D Endoscopic hemostasis: an effective therapy for bleeding peptic ulcers.JAMA. 1990; 264: 494-499Crossref PubMed Scopus (304) Google Scholar variceal sclerotherapy and banding,11Laine L Cook D Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis.Ann Intern Med. 1995; 123: 280-287Crossref PubMed Scopus (606) Google Scholar and endoscopic sphincterotomy for cholangitis12Lai ECS Mok FPT Tan ESY Lo C Fan S You K et al.Endoscopic biliary drainage for severe acute cholangitis.N Engl J Med. 1992; 326: 1582-1586Crossref PubMed Scopus (474) Google Scholar and gallstone pancreatitis.13Neoptolemos JP London NJ James D Carr-Locke DL Bailey IA Fossard DP Controlled trial of urgent endoscopic retrograde cholangiopancreatography and endoscopic sphincterotomy versus conservative treatment for acute pancreatitis due to gallstones.Lancet. 1988; 2: 979-983Abstract PubMed Scopus (784) Google Scholar, 14Fan S Lai ECS Mok FPT Lo C Zheng S Wong J Early treatment of acute biliary pancreatitis by endoscopic papillotomy.N Engl J Med. 1993; 328: 228-232Crossref PubMed Scopus (719) Google Scholar However, because of potential differences in patient populations due to the typically narrow entry and exclusion criteria of randomized trials and/or case mix served by a specific hospital, analysis of the intervention in everyday practice may yield disparate results. For example, because