Laboratory Errors and the Misdiagnosis of Tuberculosis

To the Editor. —We commend Dr Nitta and colleagues 1 for their analysis of factors that may contribute to the misdiagnosis of multidrug-resistant tuberculosis (MDR TB). The New York City Department of Health Bureau of Tuberculosis Control has identified several cases in which a misdiagnosis of TB resulted from contamination of a laboratory specimen. To minimize the possibility of misdiagnosis based on a contaminated laboratory specimen, we have instituted routine clinical and laboratory investigations of all cases that have only 1 positive culture for Mycobacterium tuberculosis , regardless of the drug susceptibilities of the microorganism and the patient's smear results for acid-fast bacilli. Medical records are reviewed to determine whether the patient in question has clinical evidence of TB, and laboratory log books are reviewed to determine whether any specimens from patients with confirmed TB were processed immediately before the specimen in question. If appropriate, DNA typing by restriction fragment length