The impact of human immunodeficiency virus on presentation and diagnosis of tuberculosis in a cohort study in Zambia.

Two hundred and forty-nine patients with tuberculosis were recruited to a cohort study to investigate the interaction between tuberculosis and HIV in Lusaka, Zambia; findings at presentation are presented here. One hundred and eighty-two (73%; 95% confidence interval 67-79%) of the cases were HIV-1 antibody positive. The diagnosis of tuberculosis was confirmed by microscopy for acid-alcohol fast bacilli, culture of Mycobacterium tuberculosis, or histology in 74% of all cases. HIV negative and positive cases differed in site of disease: among HIV negative patients 72% had pulmonary disease alone, 16% extrapulmonary disease alone and 12% had both, whereas among HIV positive patients 40% had pulmonary disease alone, 34% extrapulmonary disease alone and 26% both (P < 0.001). HIV negative and positive cases were compared with regard to outcome of diagnostic procedures: 55% of HIV negative cases could be diagnosed at enrollment by sputum smear, but only 35% of HIV positive cases (P < 0.01). Among pulmonary cases confirmed by sputum culture, 76% of HIV negative patients had a positive sputum smear, compared with 57% of HIV positive patients (P = 0.09). Pleural and pericardial disease were difficult to confirm, but culture of pleural fluid was positive in 12/46 HIV positive patients, compared with 0/11 HIV negative patients. Lymph node disease was readily confirmed by biopsy. The tuberculin test was positive in only 30/110 (27%) of HIV positive cases, but in 21/38 (55%) of HIV negative cases (P < 0.01). Mycobacterium tuberculosis was cultured in 57% of HIV negative cases and 54% of HIV positive cases; no atypical mycobacteria were isolated. Initial resistance to isoniazid was present in isolates from 5% of cases with a positive culture.