Objective: To analyze the effect of human immunodeficiency virus (HIV) status and other variables on the outcome of tuberculosis treatment in Spain.
Patients and methods: Multicenter retrospective cohort study in 6 autonomous communities of Spain (from May 1996 to April 1997). Data on treatment outcome were collected for new cases of tuberculosis in accordance with European guidelines. Follow up of patients continued for 3 months after scheduled end of treatment.
Results: Of the 4899 patients included, 3417 (69.7%) had a satisfactory outcome, 438 (8.9%) died before or during treatment, and 1044 (21.4%) had a potentially unsatisfactory outcome. On stratification by HIV status, satisfactory outcome, mortality, and potentially unsatisfactory outcome were reported for 43.4%, 21.5%, and 35.1%, respectively, of HIV-positive patients; 71%, 6.2%, and 22.8%, respectively, of HIV-negative patients; and 74.3%, 7.5%, and 18.2%, respectively, of patients with no HIV status available. HIV modified the effect of several variables on the outcome of treatment, and so separate logistic regression models for each HIV category were constructed. Among HIV-positive patients, mortality increased in patients with neoplastic disease and in users of drugs by nonintravenous routes of administration, whereas potentially unsatisfactory outcomes increased in intravenous drug users and in women.
Conclusions: In Spain, the outcome of tuberculosis treatment is much worse in HIV-positive patients. Drug use and presence of neoplastic disease substantially affect mortality.