Intensified Tuberculosis Case Finding Among HIV-Infected Persons From a Voluntary Counseling and Testing Center in Addis Ababa, Ethiopia

Abstract

OBJECTIVE: To evaluate commonly available screening tests for pulmonary tuberculosis (TB), using sputum bacteriology as a gold standard, in HIV-infected persons attending an urban voluntary counseling and testing clinic in Addis Ababa, Ethiopia. DESIGN: Prospective enrollment of HIV-infected persons, all of whom underwent TB screening, regardless of symptoms, with: (1) symptom screening and physical examination, (2) 3 sputum specimens for smear microscopy, and (3) chest radiograph. One sputum was also sent for concentrated smear microscopy and mycobacterial culture. Chest radiographs were reviewed by 2 independent radiologists. A confirmed TB diagnosis was defined as 1 positive sputum smear and/or 1 positive sputum culture. RESULTS: We enrolled 438 HIV-infected persons: 265 (61%) females, median age 34 years (range: 18-65), median CD4 cell count 181 cells per cubic millimeter (range: 2-1185). Overall, 32 (7%) persons were diagnosed with TB, of whom 5 (16%) were asymptomatic but culture-confirmed TB cases. Screening for cough >2 weeks would have detected only 12 (38%) confirmed TB cases; screening for cough or fever, of any duration, would have detected 24 (75%) cases, with specificity of 64%. Negative predictive value of screening for these 2 symptoms was 97%. Simulation of the current Ethiopian national guidelines had a sensitivity of 63% and specificity of 83% for diagnosing TB disease among study patients. CONCLUSIONS: Traditional symptom screening is insufficient for detecting TB disease among HIV-infected persons but may serve to exclude TB disease. More sensitive, rapid, and low-cost diagnostic tests are needed to meet the demand of resource-limited settings.