Long-term effectiveness of intermittent therapy for tuberculosis: final report of three Denver studies.

The long-term effectiveness of intermittent non-rifampin-containing regimens for the treatment of tuberculosis is demonstrated. The direct administration of 170 to 190 doses during an extended period of time (18 to 20 months) permits a 10 to 20% rate of missed treatment doses during the continuation phase of therapy while retaining a cost-effectiveness ratio comparable to that reported for 6-month, rifampin-containing regimens administered daily. Although 4 patients from a highly noncompliant population of 165 persons relapsed during a follow-up period of 2.5 to 12 yr, after treatment, 3 of these failures were the result of the system's inability to administer the drugs adequately. These regimens are confirmed as effective treatment alternatives that may be used in the treatment of tuberculosis in potentially unreliable patients when rifampin is contraindicated.