[False diagnosis of tuberculosis by culture]

A remarkable input to the epidemiology of tuberculosis was not the only benefit of the molecular tools developed in the early nineties for Mycobacterium tuberculosis intra-species differentiation. These genotyping methods served also to unveil specimen cross-contamination, which was until then overlooked in laboratories culturing mycobacteria. This error consists in the accidental carry-over of bacilli from a specimen with high bacterial load to that, or those, processed subsequently. The ensuing detection of falsely positive cultures can result in a wrong diagnosis of tuberculosis and the initiation of a long-lasting treatment with potentially toxic drugs. This series of errors implies the mismanagement of patients, the distraction of public health system resources, and the distortion of epidemiological data. M. tuberculosis laboratory cross-contamination was detected wherever investigated systematically, with a median rate of 3% of all positive cultures. The confirmation of this error requires a critical appraisal of bacteriological, clinical, epidemiological and genotyping results. We present here a review of national and international information on laboratory cross-contamination and describe measures recommended for minimizing the risk, surveying the occurrence, and avoiding clinical consequences of this laboratory error that raises a question on the reliability of a positive culture.