166

AN Barkun^1,2, V Adam², W Kennedy³, CA Fallone¹, K Herba¹, M Bardou¹

McGill University¹, Divisions of Gastroenterology¹, Epidemiology² and the Université de Montréal, Faculty of Pharmacy³, Montréal, Québec

AIM: The management of patients with bleeding ulcers has evolved significantly over the past 5 years, with recent data suggesting a possible role for high dose oral PPI’s. We studied the cost effectiveness of this therapy.

METHODS: Patients with bleeding ulcers initially treated endoscopically for high risk stigmata of re-bleeding were considered. The model structure was defined and validated by a group of clinical experts. Probabilities were determined from the literature. Charges and lengths of stay (LOS) were obtained from a 20% random sample of the NIS2000 (Nation-wide Inpatient Sample) administrative database. Pharmacological costs of pantoprazole 40 mg Po BID for 5 days were tabulated separately, yielding conservative estimates. A complicated stay was when a patient experienced re-bleeding. A third party payer perspective was adopted, over a 30-day time horizon. All costs were expressed in 2001 US dollars. Sensitivity and threshold analyses were carried across a broad range of clinically relevant assumptions.

RESULTS: The mean duration of hospital stays and costs for patients admitted with uncomplicated and complicated ulcer bleeding were 3 and 4.7 days, and $7,993 and $11,802 respectively. The re-bleeding rates for patients receiving oral therapy or not were 7.3% and 21.43%, respectively. In the base-case analysis, the alternative of administering 5 days of high dose oral proton pump inhibition was dominant as it was both more effective (incremental effectiveness of 14.1%) and less costly (a savings of $497.60/per patient) than not giving an oral PPI. Only when the re-bleeding rate in the no oral PPI group dropped below 7% or rose over 21% in the oral PPI treated group did the conclusions about cost-effectiveness dominance change. No differences were noted over wide ranges of the adopted cost estimates.

CONCLUSION: A strategy of administering high dose oral proton pump inhibition for 5 days in patients with bleeding ulcers initially treated with endoscopic hemostasis for high risk ulcer stigmata for re-bleeding was both more effective and cheaper than not doing so. Prospective trials in Western patients are required to confirm the increased effectiveness, and subsequently, the cost-effectiveness of this approach.

The RUGBE registry was made possible through a partnership and support funding bringing together the Canadian Association of Gastroenterology and Altana Pharma Inc.

NEXT ABSTRACT >