Review of published cost-effectiveness studies on tuberculosis treatment programmes.

This article aims to summarise key factors in the published literature associated with cost-effective tuberculosis (TB) treatment programmes and to make recommendations on how techniques for future studies could be improved. There is consistent evidence that fully ambulatory, short course chemotherapy programmes are currently the most cost-effective option, although this may depend on the cost of providing an effective community-based service. Direct supervision may be more cost-effective than self-administration because of the reduced need for monitoring and follow-up; more studies are needed, however, that include real outcome figures and household measures of cost. For studies taking a provider perspective, the methods used for measuring costs will be dependent on sources of information, but centralised accounts are the most preferred source. Effects should be measured in terms of actual outcome, and should preferably not be taken from the literature. Most of the studies reviewed did not consider the difficulties of introducing a theoretically cost-effective change into a health service. More studies are required that document actual changes in programme cost and outcome associated with the introduction of different types of treatment delivery. Future work could consider measuring cost in terms of resources (e.g., staff) rather than only finances, and more work is needed on household perspectives.