ERCP economics

Abstract

It is not often that one prepares for an editorial by visiting the medical center's financial offices. However, the original articles related to ERCP1Ho KY Montes H Sossenheimer MJ Tham TCK Ruyman F Van Dam J Carr-Locke DL. Features which may predict hospital admission following outpatient therapeutic ERCP.Gastrointest Endosc. 1999; 49: 587-592Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar, 2Freeman ML Nelsen DB Sherman S Haber GB Fennerty MB DiSario JA et al.Same-day discharge after endoscopic biliary sphincterotomy: observations from a prospective multicenter complication study.Gastrointest Endosc. 1999; 49: 580-586Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 3Poon RT-P Yeung C Lo C-M Yuen W-K Liu C-L Fan S-T. Prophylactic effect of somatostatin on post-ERCP pancreatitis: a randomized controlled trial.Gastrointest Endosc. 1999; 49: 593-598Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar, 4Andriulli A Leandro G Niro G Mangia A Festa V Villani MR et al.Medical treatment can diminish pancreatic damages after ERCP: a meta-analysis.Gastrointest Endosc. 1999; (In press): 49Google Scholar published in this issue of Gastrointestinal Endoscopy caused me to do just that. How so? On the one hand, two of the articles deal with outpatient therapeutic ERCPs1Ho KY Montes H Sossenheimer MJ Tham TCK Ruyman F Van Dam J Carr-Locke DL. Features which may predict hospital admission following outpatient therapeutic ERCP.Gastrointest Endosc. 1999; 49: 587-592Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar, 2Freeman ML Nelsen DB Sherman S Haber GB Fennerty MB DiSario JA et al.Same-day discharge after endoscopic biliary sphincterotomy: observations from a prospective multicenter complication study.Gastrointest Endosc. 1999; 49: 580-586Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar and the other, costly but apparently effective, mechanisms to decrease postprocedural pancreatitis.3Poon RT-P Yeung C Lo C-M Yuen W-K Liu C-L Fan S-T. Prophylactic effect of somatostatin on post-ERCP pancreatitis: a randomized controlled trial.Gastrointest Endosc. 1999; 49: 593-598Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar, 4Andriulli A Leandro G Niro G Mangia A Festa V Villani MR et al.Medical treatment can diminish pancreatic damages after ERCP: a meta-analysis.Gastrointest Endosc. 1999; (In press): 49Google Scholar Taken together, this ERCP potpourri deserves comment regarding the delicate balance of procedural safety and cost. Is outpatient ERCP safe? Is it cost-effective? If it is cost-effective, for whom? Ho et al.1Ho KY Montes H Sossenheimer MJ Tham TCK Ruyman F Van Dam J Carr-Locke DL. Features which may predict hospital admission following outpatient therapeutic ERCP.Gastrointest Endosc. 1999; 49: 587-592Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar at Brigham and Women's Hospital prospectively studied admissions for post-ERCP events in 415 consecutive patients in whom outpatient therapeutic ERCP was the defined goal. Ten percent of patients (41) required immediate or subsequent admission for adverse events and another 15% (63) for observational purposes. There were two deaths (0.5%) in this series from severe pancreatitis and the article does not specify whether either of these patients was initially admitted or originally discharged from the endoscopy unit and subsequently readmitted. However, 84% of the adverse events requiring admission to the hospital occurred within 4 hours of completing the ERCP. Multivariate analyses suggested that there were three factors that were statistically more likely in the 63 patients without complications admitted for observation after procedure: intraprocedure pain, previous history of pancreatitis, and performance of endoscopic sphincterotomy. Perhaps more important were the factors not associated with need for hospitalization: suspected sphincter dysfunction, multiple pancreatic injections, acinarization, and procedural bleeding, among others. These authors stated that routine outpatient ERCP was safe and saved a $680 hospitalization charge in patients who did not require admission. Freeman et al.,2Freeman ML Nelsen DB Sherman S Haber GB Fennerty MB DiSario JA et al.Same-day discharge after endoscopic biliary sphincterotomy: observations from a prospective multicenter complication study.Gastrointest Endosc. 1999; 49: 580-586Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar in turn, report the results of a 17-center prospective study in which 614 of 2347 (26%) patients undergoing endoscopic sphincterotomy were discharged after procedure at the endoscopist's discretion. Thirty-five of these 614 patients (5.7%) were readmitted for adverse events (pancreatitis 20, miscellaneous 15) and an additional, unspecified number subsequently were admitted for observation. Eighty percent of patients developed an adverse event within 6 hours and the following clinical variables were statistically significant for subsequent adverse events using multivariate analysis: sphincter of Oddi dysfunction, cirrhosis, difficult cannulation, precut sphincterotomy, and combined percutaneous-endoscopic procedure. Taken together, these articles suggest that universal or selective outpatient therapeutic ERCP can be undertaken with an acceptable risk and a potential for cost savings. However, these studies also tell us that admissions for observation and adverse events increase proportionally contingent on the percentage of procedures in which outpatient therapy is attempted (25% versus 5.7%). Furthermore, some of the clinical or endoscopic factors found to predispose to subsequent hospitalization in the two series do not help us much. Previous pancreatitis predisposes to subsequent admission in the study by Ho et al,1Ho KY Montes H Sossenheimer MJ Tham TCK Ruyman F Van Dam J Carr-Locke DL. Features which may predict hospital admission following outpatient therapeutic ERCP.Gastrointest Endosc. 1999; 49: 587-592Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar but not in that by Freeman et al.2Freeman ML Nelsen DB Sherman S Haber GB Fennerty MB DiSario JA et al.Same-day discharge after endoscopic biliary sphincterotomy: observations from a prospective multicenter complication study.Gastrointest Endosc. 1999; 49: 580-586Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar Likewise, there is discordance with regard to the importance of suspected sphincter of Oddi dysfunction, as well as performance of precut sphincterotomy and multiple pancreatic injections. At best, these reports allow us to selectively delineate historical or procedural “warning signs” that may sensitize us to observe a subset of patients overnight. Does outpatient ERCP really save money? If so, whose? Data are sparse and despite claims to the contrary, one study found that direct costs of outpatient sphincterotomy were actually higher than overnight observation costs when patients were monitored in the recovery room for longer than 2.25 hours.5Chow S Bosco JJ Shea JA Heiss FW. Outpatient endoscopic sphincterotomy: analysis of safety and cost.Am J Gastroenterol. 1996; 91 ([abstract]): 1930Google Scholar Moreover, our group, as well as others, has previously documented that therapeutic ERCP may actually lose money, primarily as a consequence of the costs associated with disposable ERCP accessories.6Kim-Deobald J Kozarek RA Ball TJ. Prospective evaluation of costs of disposable accessories in diagnostic and therapeutic ERCP.Gastrointest Endosc. 1993; 39: 763-765Abstract Full Text PDF PubMed Scopus (28) Google Scholar, 7Walker RS Vanagunas AD Williams P Chodash HB. Therapeutic ERCP: a cost-prohibitive procedure?.Gastrointest Endosc. 1997; 46: 143-146Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar This was related to cumulative accessory expense and the vagaries of the insurance reimbursement system. As such, medical centers have traditionally charged for equipment that was not routinely being reimbursed but rather “globalized” into a room fee. Although reimbursement was a particular problem with Medicare and Medicaid, it was found to be equally problematic with many managed care insurance plans as well. Patients discharged after a therapeutic ERCP procedure were reimbursed for a geographically dependent room fee ($550 in the Seattle area) whether the procedure cost $500 or $1500. Patients hospitalized under a diagnosis-related group (DRG) or even those hospitalized for overnight observation (2359) were ultimately included under a complex Medicare plan in which a yearly reconciliation payment covered at least some of the medical center's accessory costs. The fact that selective outpatient therapeutic ERCP appears to be cost-effective for the insurer has to be placed into the context of both the patient and provider. From the latter standpoint, there is a reason that ambulatory endoscopy centers have been loathe to embrace outpatient ERCP. From the former standpoint, patient comfort and safety should remain our primary concerns and neither of the above-mentioned articles addresses the issue of postprocedure nausea and pain, anxiety, and other commonly defined outcomes measures. The other two ERCP articles in this issue of Gastrointestinal Endoscopy can also be construed as addressing safety and cost issues.3Poon RT-P Yeung C Lo C-M Yuen W-K Liu C-L Fan S-T. Prophylactic effect of somatostatin on post-ERCP pancreatitis: a randomized controlled trial.Gastrointest Endosc. 1999; 49: 593-598Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar, 4Andriulli A Leandro G Niro G Mangia A Festa V Villani MR et al.Medical treatment can diminish pancreatic damages after ERCP: a meta-analysis.Gastrointest Endosc. 1999; (In press): 49Google Scholar The article by Poon et al.3Poon RT-P Yeung C Lo C-M Yuen W-K Liu C-L Fan S-T. Prophylactic effect of somatostatin on post-ERCP pancreatitis: a randomized controlled trial.Gastrointest Endosc. 1999; 49: 593-598Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar randomized 220 patients to receive either a placebo or 3 mg somatostatin 30 minutes prior to and for 12 hours after diagnostic or therapeutic ERCP. There was a tendency for those patients treated with somatostatin to have lower serum amylase and lipase values and statistically significant decreases in postprocedural pain (1% versus 6%) and pancreatitis (3% versus 10%) were noted. This study and a recently published series by Bordas et al.8Bordas JM Toledo-Pimentel V Llach J Elena M Mondelo I Gines A et al.Effects of bolus somatostatin in preventing pancreatitis after endoscopic pancreatography: results of a randomized study.Gastrointest Endosc. 1998; 47: 230-234Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar using a single dose of somatostatin immediately prior to cannulation are the only two studies to date that have documented a decreased risk of postprocedural pancreatitis with this drug. In the final article, Andriulli et al.,4Andriulli A Leandro G Niro G Mangia A Festa V Villani MR et al.Medical treatment can diminish pancreatic damages after ERCP: a meta-analysis.Gastrointest Endosc. 1999; (In press): 49Google Scholar in turn, undertake a meta-analysis of previously published studies in which somatostatin, octreotide, or gabexate-mesilate were administered in an attempt to decrease post-ERCP pancreatitis.4Andriulli A Leandro G Niro G Mangia A Festa V Villani MR et al.Medical treatment can diminish pancreatic damages after ERCP: a meta-analysis.Gastrointest Endosc. 1999; (In press): 49Google Scholar Reviewing 24 controlled studies, the authors assessed three outcomes in these studies, which admittedly administered medications in various dosages and time intervals prior to or after ERCP: postprocedural pain, hyperamylasemia, and acute pancreatitis. In this meta-analysis, octreotide proved effective in reducing post-ERCP hyperamylasemia but not pain or subsequent pancreatitis, perhaps because of its stimulatory effect on the sphincter of Oddi. Somatostatin, in turn, was believed to be effective in reducing post-ERCP pain and pancreatitis, but not hyperamylasemia. Finally, gabexate-mesilate was the only medication associated with improvement in all three outcome parameters in this meta-analysis. Anything that decreases the risk of ERCP should be welcomed, but everything must be placed in the cost equation. For instance, utilizing the methodology of Poon et al.,3Poon RT-P Yeung C Lo C-M Yuen W-K Liu C-L Fan S-T. Prophylactic effect of somatostatin on post-ERCP pancreatitis: a randomized controlled trial.Gastrointest Endosc. 1999; 49: 593-598Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar all patients, even those undergoing a diagnostic ERCP, would have been hospitalized for the 12-hour infusional period. The decrease in postprocedural pain incidence (16% versus 5.7%), in turn, would suggest that for every 100 patients, a significant subset would have received the infusion needlessly, numbers that are magnified further when comparing the incidence of pancreatitis (10% versus 3%) in placebo versus somatostatin-treated patients. A similar rationale can be applied to use of the protease inhibitor gabexate. In the largest study to date, Cavallini et al.9Cavallini G Tittobello A Frulloni L Masci E Mariani A DiFrancesco V the Gabexate in Digestive Endoscopy Italian Group Gabexate for the prevention of pancreatic damage related to endoscopic retrograde cholangiopancreatography.N Engl J Med. 1996; 335: 919-923Crossref PubMed Scopus (349) Google Scholar randomized more than 400 patients to placebo or 1 gm gabexate for 30 minutes before and 12 hours after ERCP. Six percent of patients developed postprocedural pain versus 14% of control subjects, and there was a 2% incidence of acute pancreatitis versus an 8% incidence in patients randomized to placebo. How do we put these four articles together? How do we balance risk with cost? Is it better to infuse all patients with a substance that inhibits proteolytic activity or pancreatic secretion to minimize procedurally induced pancreatitis in a subset? Even if it entails an overnight hospitalization in the 20% to 30% of patients who may undergo a diagnostic procedure alone? Assuming that insurers will adopt adequate reimbursement payments for outpatient therapeutic ERCPs, is it preferable to attempt selective as opposed to universal same-day discharges? The answer to these questions obviously lies in some middle ground. As most of the readmissions for complications in the articles by Ho et al.1Ho KY Montes H Sossenheimer MJ Tham TCK Ruyman F Van Dam J Carr-Locke DL. Features which may predict hospital admission following outpatient therapeutic ERCP.Gastrointest Endosc. 1999; 49: 587-592Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar and Freeman et al.2Freeman ML Nelsen DB Sherman S Haber GB Fennerty MB DiSario JA et al.Same-day discharge after endoscopic biliary sphincterotomy: observations from a prospective multicenter complication study.Gastrointest Endosc. 1999; 49: 580-586Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar were for pancreatitis, it seems reasonable to pretreat certain high-risk groups with agents that may minimize pancreatitis or require overnight observation. Patients with previous ERCP-related pancreatitis, those with sphincter of Oddi dysfunction, particularly if undergoing manometry, and those undergoing pancreatic endotherapy (sphincterotomy, stent placement) in the setting of normal ductal anatomy seem most suitable for such prophylaxis. Other patients, although perhaps not candidates for pancreatitis prophylaxis, may nevertheless be candidates for extended observation. Ho et al. would suggest that intraprocedural pain is a risk factor, whereas Freeman et al. add factors such as cirrhosis, coagulopathy, difficult cannulation, or a combined percutaneous-endoscopic procedure. I can think of others from strictly a common sense or medical-legal standpoint: patients with significant postsphincterotomy hemorrhage necessitating cautery or injection therapy, those experiencing local perforation treated with stent placement, aged or infirm patients with significant comorbid disease, or finally, anyone without an adequate safety net or support system. Just as cost analysis needs to consider institutional and insurance expenditures, risk extends to the provider and to the patient. ERCP will be a safer and more cost-effective procedure in the future. Not only will newer technologies (magnetic resonance cholangiopancre-atography, EUS, spiral CT) decrease the number of purely diagnostic procedures undertaken, but risk stratification will allow us to treat patients at significant risk of developing postprocedural pancreatitis. This will, however, be given in bolus or infusional form immediately prior to ERCP as opposed to a prolonged infusion and is more likely to be somatostatin or an anticytokine than a protease inhibitor. Finally, we will develop outcomes analyses to include patient satisfaction measurements that will help to place same-day discharge of individuals undergoing therapeutic ERCP into broader clinical perspectives. In the meantime…