The economics of TRACE - A cost-effectiveness analysis of trandolapril in postinfarction patients with left ventricular dysfunction

Objective: The objective of the study was to compute a cost-effectiveness ratio relating the economic cost of trandolapril to the number of effectiveness units (i.e. life-years) gained. Design and Setting: The trandolapril cardiac evaluation (TRACE) study was a prospective placebo-controlled clinical trial designed to determine the long term effect of the oral angiotensin-converting enzyme (ACE) inhibitor trandolapril in postinfarction patients with left ventricular dysfunction. We used the individual data of the TRACE trial to compute a cost-effectiveness ratio relating the economic cost of trandolapril to the number of life-years saved. The analysis was differential and was conducted from a payer perspective in a French setting. Costs were computed from individual data related to the use of resources during the TRACE trial. For drug treatments, we chose French public prices, and for hospitalisations, we used the mean cost as determined by the diagnosis related group (DRG) from the 1996 Programme de Medicalisation des Systemes d'Information (PMSI) database from the French Ministry of Health. Life expectancy was estimated through an accelerated failure-time model with an exponential distribution specification; we made the conservative hypothesis that the effect of trandolapril on mortality after the end of the trial was nil. We assessed the standard deviation and the 95% confidence interval (CI) of the ratio through its bootstrap distribution. Results: The incremental cost-effectiveness ratio of treating patients with trandolapril rather than with placebo was estimated as 4910 French francs (FF) per life-year saved. Discounting both costs and health effects led to a ratio of FF6950 per life-year saved. The bootstrap estimate of the ratio reached FF5950 and the 95% CI was FF5650 to FF6250 per life-year saved. Conclusions: These results could be considered as highly cost effective, even though our estimation was very close to the design and the conditions of the TRACE trial. Nevertheless, we showed that this trial constitutes a favourable case for economic evaluation.