Setting: The Chest Clinic and the JICA (Japan International Cooperation Agency) Molecular Laboratories, University Teaching Hospital, Lusaka, Zambia, and the Department of Internal Medicine, Institute of Tropical Medicine, Nagasaki University, Nagasaki, Japan.
Objective: To evaluate the polymerase chain reaction (PCR) as a laboratory test for the rapid diagnosis of pulmonary tuberculosis in the African situation by identifying mycobacterial DNA in urine samples using two commonly described molecular methods.
Design: Prospective collection and laboratory analysis of urine samples from adult Zambian patients with culture-confirmed pulmonary tuberculosis and healthy controls.
Methods: Urine was obtained from 63 patients with culture-confirmed active pulmonary tuberculosis and 63 'healthy' control patients with no active tuberculosis. DNA was isolated from urine sediment and subjected to analyses by two well-described PCR-based methods, 'the Sechi method' and 'the Githui method', for the identification of Mycobacterium tuberculosis DNA. The sensitivity and specificity of the two tests were determined.
Results: The sensitivity and specificity of the Githui method were 55.6% (35/63) and 98.4% (62/63), respectively. The sensitivity and specificity of the Sechi method were 28.6% (18/63) and 98.4% (62/63), respectively. Of the 63 patients, 50 (79%) were HIV sero-positive and the frequency of positive PCR urines using the Githui method was greater in HIV-positive patients than in HIV-negative patients (32/50 = 64% vs. 3/13 = 23%; P = 0.05).
Conclusions: Neither the Githui method nor the Sechi method was sensitive enough to be recommended for routine use in clinical practice. PCR-based assays for the detection of M. tuberculosis DNA in urine will require further refinement before they can be recommended for use in clinical practice in Africa. The presence of mycobacterial DNA in urine samples of patients with pulmonary tuberculosis also requires further study.