175

A Barkun¹, R Crott², C Fallone¹, W Kennedy2, J Lachaine², C Levington², D Armstrong³, N Chiba^3,4, A Thomson⁵, SJO Van Zanten⁶, P Sinclair⁷, B Chakraborty⁸, S Escodebo⁸, S Smyth⁸, R White⁸, K Nevin⁸

¹Gastroenterology, McGill University, Montreal, Quebec, ²Pharmacy, University of Montreal, Montreal, Quebec, ³Gastroenterology, McMaster University, Hamilton, Ontario, ⁴Gastroenterology, Guelph General Hospital, Guelph, Ontario, ⁵Gastroenterology, University of Alberta, Edmonton, Alberta, ⁶Gastroenterology, Dalhousie University, Halifax, Nova Scotia, ⁷INSINConsulting, Inc, Guelph, Ontario, ⁸AstraZeneca, Inc, Mississauga, Ontario

AIM: The optimal cost-effective management of adult patients with uninvestigated dyspepsia (UD) remains unknown.

METHODS: We compared the CanDys approach (acid suppression if heartburn predominant symptoms, and a test-and-treat strategy for others, acid suppression for H. pylori (HP) negative patients and eradication for HP+) to empirical anti-secretory therapy (EAS) and prompt endoscopy (PE). All patients had pain/discomfort in the upper abdomen, with or without heartburn, acid regurgitation, excessive burping or belching, increased abdominal bloating, feeling of abnormal or slow digestion, early satiety, or nausea. EAS was 4 weeks of omeprazole 20 mg qd or ranitidine 150 mg bid. Effectiveness was the number of symptom-free (SF) months at 1 year. Probabilities stemmed from clinical trials, and direct costs (in 2001 CAN$) from provincially set fees. The Monte Carlo simulation used Data Pro, with sensitivity analyses carried out across clinically relevant ranges of probabilities and costs. 95% confidence intervals (CI) were calculated. Acceptability curves were used to compare strategies.

RESULTS: Looking at the medical costs, the least costly strategy per patient treated was CanDys-ranitidine ($239.50, 4.39 SF-months), followed by CanDys-omeprazole ($262, 4.71 SF-months), and PE-omeprazole ($2,245.60, 7.42 SF-months). The EAS-omeprazole, EAS-ranitidine, and PE-ranitidine approaches were dominated. Results were sensitive to variations in efficacy and cost, and there were wide overlaps in most point estimates of cost-effectiveness. Using a willingness-to-pay threshold value of $175 per symptom-free month, 78% of simulations led to choosing CanDys-omeprazole over alternatives.

CONCLUSION: The CanDys-omeprazole approach is more effective and more costly than CanDys-ranitidine, and is less effective but much less costly than PE-omeprazole. Other strategies are dominated. However, there exist wide overlaps in the 95% CI. Acceptability curve analysis favored CanDys-omeprazole over other approaches using a willingness to pay threshold of 175$ for each additional SF months.

Supported in part by AstraZeneca Canada Inc.

NEXT ABSTRACT >