Quality control in tuberculosis bacteriology. 2. The origin of isolated positive cultures from the sputum of patients in four studies of short course chemotherapy in Africa.

During intensive bacteriological follow-up of 2123 patients who had received short course chemotherapy regimens in 4 controlled clinical trials in Africa and had not had a bacteriological relapse, 405 isolated positive cultures were obtained from 37 429 sputum specimens in 3 East African laboratories. These cultures might have arisen as a result of clerical error, of transfer of bacilli from positive to negative specimens in the laboratory or from the lesions of the patients. Clerical error in the labelling of specimens or the recording of results did not seem a frequent cause, since isolated positive cultures contained many fewer colonies than cultures from other positive specimens being processed at the same time. Several of evidence suggested that some isolated positive cultures arose from the lesions of patients: a decrease in their incidence occurred in successive time periods after chemotherapy; the number of isolated positives per patient departed significantly from the Poisson distribution; they were more often drug resistant than other cultures processed at the time; positive cultures were obtained less frequently from known autoclaved specimens inserted among the the study specimens than from the study specimens themselves; no association was found between the incidence of isolated positive and of specimens containing numerous viable M. tuberculosis being processed at the same time. Nevertheless some of these cultures probably arose by transfer in the laboratory, since the rates at which transfer were known to occur differed in the 3 laboratories and corresponded to the rates of obtaining isolated positive cultures.