A N Barkun1, K Herba, V Adam, W Kennedy, C A Fallone, M Bardou. 1. Division of Gastroenterology, McGill University Health Centre, Montreal General Hospital Site, Quebec, Canada. alan.barkun@muhc.mcgill.ca
Abstract
BACKGROUND: Recent data suggest a role for high-dose oral proton pump inhibition in ulcer bleeding. AIM: To compare the cost-effectiveness of oral high-dose proton pump inhibition to both high-dose intravenous proton pump inhibition and placebo administration. METHODS: The model adopted a 30-day time horizon, and focused on patients with ulcer haemorrhage initially treated endoscopically for high-risk stigmata. Re-bleeding rates were set a priori based on non-head-to-head data from the literature, and charges and lengths of stay from a national American database. Sensitivity analyses were carried across a broad range of clinically relevant assumptions. RESULTS: Re-bleeding rates for patients receiving intravenous, oral, or placebo therapies were 5.9%, 11.8%, and 27%, respectively. The mean lengths of stay and costs for admitted patients with and without re-bleeding were 4.7 and 3 days; $11,802, and $7993, respectively. High-dose intravenous proton pump inhibition was more effective and less costly (dominant) than high-dose oral proton pump inhibition with incremental savings of $136.40 per patient treated. The oral high-dose strategy in turn dominated placebo administration. Results remained robust according to one- and two-way sensitivity analyses. CONCLUSION: In patients undergoing endoscopic haemostasis, subsequent high-dose intravenous proton pump inhibition is more cost-effective than high-dose oral proton pump inhibition, which in turn dominates placebo. The results from this exploratory-type cost analysis require confirmation by head-to-head prospective trials performed in Western populations.
BACKGROUND: Recent data suggest a role for high-dose oral proton pump inhibition in ulcer bleeding. AIM: To compare the cost-effectiveness of oral high-dose proton pump inhibition to both high-dose intravenous proton pump inhibition and placebo administration. METHODS: The model adopted a 30-day time horizon, and focused on patients with ulcer haemorrhage initially treated endoscopically for high-risk stigmata. Re-bleeding rates were set a priori based on non-head-to-head data from the literature, and charges and lengths of stay from a national American database. Sensitivity analyses were carried across a broad range of clinically relevant assumptions. RESULTS: Re-bleeding rates for patients receiving intravenous, oral, or placebo therapies were 5.9%, 11.8%, and 27%, respectively. The mean lengths of stay and costs for admitted patients with and without re-bleeding were 4.7 and 3 days; $11,802, and $7993, respectively. High-dose intravenous proton pump inhibition was more effective and less costly (dominant) than high-dose oral proton pump inhibition with incremental savings of $136.40 per patient treated. The oral high-dose strategy in turn dominated placebo administration. Results remained robust according to one- and two-way sensitivity analyses. CONCLUSION: In patients undergoing endoscopic haemostasis, subsequent high-dose intravenous proton pump inhibition is more cost-effective than high-dose oral proton pump inhibition, which in turn dominates placebo. The results from this exploratory-type cost analysis require confirmation by head-to-head prospective trials performed in Western populations.
Authors: Alan N Barkun; Viviane Adam; Joseph J Y Sung; Ernst J Kuipers; Joachim Mössner; Dennis Jensen; Robert Stuart; James Y Lau; Emma Nauclér; Jan Kilhamn; Helena Granstedt; Bengt Liljas; Tore Lind Journal: Pharmacoeconomics Date: 2010 Impact factor: 4.981