| 066 |
|
SHOULD ENDOSCOPIC ULTRASOUND BE ROUTINE AFTER AN EPISODE OF “IDIOPATHIC” PANCREATITIS? A COST-UTILITY ANALYSIS
SF Owayed, DA Lloyd, JC Gregor
Division of Gastroenterology, University of Western Ontario, London, Ontario
BACKGROUND: Up to 30% of episodes of acute pancreatitis are labeled as idiopathic. In many patients, the disease is related to occult choledocholithiasis. Endoscopic retrograde cholangiopancreatography (ERCP) is the gold standard test in these situations. However, it is sometimes done unnecessarily with the potential for complications. More recently, endoscopic ultrasound (EUS) has been used for the diagnosis of occult choledocholithiasis. Our objective is to examine the utility and cost-effectiveness of various diagnostic strategies in patients who have recovered from an acute episode.
METHODS: Using a decision model (Data 3.5, Tree Age Software Inc.) we considered three strategies for evaluating patients with idiopathic acute pancreatitis (1) expectant management (FOLLOW), (2) ERCP in all patients (ERCP), (3) EUS followed by ERCP if a stone was found (EUS). Probabilities of various outcomes such as recurrent pancreatitis, post-ERCP pancreatitis and diagnostic accuracy of EUS were estimated from the literature. Direct medical costs (Canadian dollars) from a third party payer perspective were calculated using local hospital cost and our previously published data.
RESULTS: FOLLOW was estimated to result in 1136.4 quality-adjusted life weeks per patient (QALWs/pt). ERCP and EUS were superior at 1140.5 and 1142.3 QALWs/pt respectively. FOLLOW was estimated to cost $1555 per patient. ERCP and EUS, on the other hand, were more expensive at $2298 and $2142 respectively. This translates into a cost-utility ratio of $9423 per quality-adjusted life year (QALY) for ERCP vs. FOLLOW and $5174/QALY for EUS vs. FOLLOW. EUS was dominant over ERCP as long as the probability of occult choledocholithiasis was less than 33% and the specificity was greater than 81%. EUS was also beneficial for patients as long as the probability of having a stone was greater than 1%.
CONCLUSION: EUS appears to be indicated in all patients with idiopathic acute pancreatitis and is more cost-effective than ERCP.