Results of a directly observed intermittent isoniazid preventive therapy program in a shelter for homeless men.

We offered directly supervised isoniazid preventive therapy to middle-aged and elderly men who were exposed to tuberculosis during an outbreak of cases at a homeless shelter. Isoniazid, 900 mg, was administered twice weekly to men at the shelter and at the downtown Public Health Clinic. We report on compliance and adverse reactions associated with this preventive regimen. Isoniazid preventive therapy was offered to 64 men. Forty-seven men (73%) began therapy, and 23 of them (49%) completed the 6- to 12-month regimen. Men who failed to complete isoniazid therapy received a median of 11 biweekly doses (range, 1 to 41 doses) over a median of 9 wk (range, 3 to 37 wk). The most common reason for incomplete treatment was that the men no longer frequented the shelter. One hepatotoxic reaction occurred, for a 2.1% cumulative incidence among all men who started therapy and a 3.7% cumulative incidence among the 27 known alcoholic men who began therapy. In addition, we sought to identify personal characteristics of the men that might be associated with noncompliance. Out-of-state location of personal contacts in case of emergency was strongly associated with poor compliance. These data provide some assurance that directly supervised isoniazid preventive therapy may be offered safely to sheltered elderly men with a high prevalence of alcoholism. Further improvements in compliance with isoniazid preventive therapy in homeless populations may depend on the development of a therapeutic regimen of only 2 to 3 months duration.