Isoniazid prophylactic therapy for tuberculosis in HIV-seropositive patients - a least-cost analysis
The expected upsurge in the number of new cases of tuberculosis resulting from the HIV/AIDS epidemic prompted an examination of the feasibility of prevention strategies to limit the increase in clinical tuberculosis. A computer spreadsheet model was developed to estimate the costs and benefits that would result from isoniazid chemoprophylaxis for tuberculosis in a hypothetical cohort of 100 000 HIV-seropositive people in South Africa over a period of 8 years. At a 50% prevalence of tuberculosis infection among those at high background risk, and 5 - 10% among those at low risk, there would have been 34 000 cases of active tuberculosis in the cohort and their contacts if no prophylactic therapy had been used. On the other hand, a chemoprophylaxis policy would have meant only 12 200 cases of tuberculosis, if a patient compliance rate of 68,5% had been assumed. Such a policy would have prevented 21 800 cases of active tuberculosis. The estimated total discounted cost of a chemoprophylaxis programme would have been R51,3 million. In the absence of preventive therapy the discounted cost of treating patients with active tuberculosis would have been R91,9 million over the 8 year period. Therefore, if the benefits of chemoprophylaxis were defined in terms of averted health care costs, such a policy would have resulted in net savings of R40,6 million. This study did not estimate losses in production associated with tuberculosis treatment or the value of preventing tuberculosis per se, though such indirect costs would have increased the benefit of the prevention programme. Sensitivity analysis suggests that although a 6-month chemoprophylaxis policy appears justifiable on economic considerations, this is critically dependent on the annual risk of developing tuberculosis, patient compliance and the validity of assumptions on the efficacy and duration of protection of isoniazid prophylaxis.