A large, simple trial of a tuberculosis vaccine.

Although there are no new tuberculosis vaccines currently available, it is possible to estimate the infrastructure needed for efficacy trials of such a vaccine. A randomized, placebo-controlled vaccine strategy for community-wide vaccination of adults is proposed: a large, simple trial with recipients stratified at enrollment by human immunodeficiency virus serologic status and purified protein derivative-skin test status. The outcome, tuberculous disease, would be assessed by community-wide surveillance. Such a trial could be carried out in populations in developed countries where the annual incidence of tuberculous disease is >100 cases per 100,000 persons and in developing countries where the incidence is >400 per 100,000 persons. In developed countries, enrollment of 14,600-80,000 persons would be needed, depending on the initial assumptions; in developing countries, enrollment would be 4400-27,000 persons. Readiness for tuberculosis vaccine efficacy trials will require epidemiological field studies to identify potential trial sites and investment in local diagnostic, surveillance, and data management capabilities.