[False-positive cultures due to cross contamination in tuberculosis laboratories].

Fifteen episodes of Mycobacterium tuberculosis laboratory cross-contamination suspected between 1996 and 2001 at 6 laboratories in Buenos Aires City and suburbs were investigated by IS6110 RFLP. Thirteen episodes were confirmed. Even though BACTEC 460 produced the highest number of confirmed episodes in a single laboratory, the most extended one occurred while employing conventional culture procedures in solid medium. The double repetitive element-polymerase chain reaction (DRE-PCR) was applied to 8 of these episodes and produced concordant results with those of the RFLP. The DRE-PCR appears to be a valuable tool for the prompt identification of false positive cultures. The timely rectification of defects in laboratory protocols can avert false diagnoses of tuberculosis and unnecessary prolonged treatments.