Treatment results of DOTS in 1797 Sudanese tuberculosis patients with or without HIV co-infection.

SETTING Consecutive new tuberculosis (TB) patients, from eight states in Sudan, who had never been previously treated for as much as 1 month between 1998 and 2000. OBJECTIVE To determine the impact of human immunodeficiency virus (HIV) co-infection on tuberculosis treatment outcome. DESIGN All patients presenting with symptoms suggestive of tuberculosis underwent sputum microscopy for acid-fast bacilli (AFB). Treatment is free of charge, and directly observed for all smear-positive patients. Treatment outcomes were those defined by the World Health Organization. All patients were tested anonymously for human immunodeficiency virus (HIV) using the Bionor test. RESULTS Of 10 494 patients suspected of TB and referred for sputum microscopy, 1797 were TB cases; 983 had smear-positive pulmonary tuberculosis, 521 smear-negative pulmonary tuberculosis, and 293 extra-pulmonary tuberculosis. Smear-positive cases showed a cure rate of 77.2% and a failure rate of 1%. Smear-negative and extra-pulmonary patients had a completion rate of 79.4%. Cure rates for the smear-positive cases were 68.3% for HIV-positive and 77.6% for HIV-negative patients (P = 0.164). Case fatality was significantly higher among HIV-positive (12%) than among HIV-negative cases (1.8%) (OR 7.7, 95% CI 3.51-16.8). CONCLUSION To date, a relatively low proportion of tuberculosis patients in Sudan also have HIV infection. These patients are substantially more likely to die while on treatment for their tuberculosis, a fact that underlines their need for more comprehensive care if their lives are to be prolonged. In addition, every effort is required to diminish the transmission of HIV infection to prevent the tragedy this infection represents to the community.