RESUMO
OBJECTIVE: To compare molecular techniques with conventional diagnostic methods for evaluating nosocomial transmission of multidrug-resistant tuberculosis (MDR-TB). DESIGN: We conducted a 12-week postexposure inception cohort study of health-care personnel who had been exposed to a patient with MDR-TB. MATERIAL AND METHODS: In addition to baseline and follow-up tuberculin skin tests and chest roentgenography, weekly pulmonary specimens were evaluated by (1) auramine-rhodamine fluorescent staining, (2) culture for mycobacteria, and (3) polymerase chain reaction (PCR) to amplify IS6110, a nucleic acid insertion sequence unique to the Mycobactrium tuberculosis complex. RESULTS: The index patient's isolate of M. tuberculosis showed a mutation in codon 531 of the RNA polymerase beta subunit (rpoB) gene of M. tuberculosis, which is associated with rifampin resistance and considered a marker for this MDR-TB strain. All pulmonary and gastric specimens from study participants had negative auramine stains and cultures for mycobacteria, One person, however, had separate specimens with repeatedly positive PCR results for IS6110 sequences, but the specimens contained a wild-type M. tuberculosis rpoB codon 531 dissimilar from the index patient's strain. CONCLUSION: Although both molecular and conventional testing showed that no exposed person was infected with the MDR-TB strain, molecular test results were available sooner and seemed more sensitive for detecting M. tuberculosis in one exposed person, presumably in a preinfection or "colonized" stage. Molecular methods provided information that helped distinguish this person's M. tuberculosis strain from the index patient's MDR-TB strain. Additional prospective studies should assess the value of these molecular techniques in similar clinical settings.