Cost-effectiveness of directly observed versus self-administered therapy for tuberculosis

Decision analysis was used to compare three alternative strategies for a 6-mo course of treatment for tuberculosis: directly observed drug therapy (DOT), self-administered fixed-dose combination drug therapy, and self-administered conventional individual drug therapy. Estimates of effectiveness were obtained from the published literature. Estimates of costs were obtained from the literature and the Baltimore City Health Department. Both DOT and fixed-dose combination therapy were less costly and more effective than conventional therapy, although DOT was most cost-effective. In total, the average cost per patient treated was $13,925 for DOT, $13,959 for fixed-dose combination therapy, and $15,003 for conventional therapy. Per 1,000 patients treated, 31 relapses and three deaths could be expected for DOT, 96 relapses and eight deaths for fixed-dose combination therapy, and 133 relapses and 13 deaths for conventional therapy. The marginal cost-effectiveness of DOT relative to fixed-dose combination therapy was most sensitive to variability in the direct cost of DOT and less sensitive to relapse rates for DOT and fixed-dose combination therapy. The inferior cost-effectiveness of conventional therapy was not sensitive to plausible variability in cost or effectiveness. Both DOT and fixed-dose combination therapy were cost-effective relative to conventional therapy, although DOT is probably most cost-effective.