More on Treatment Outcomes in Multidrug-Resistant Tuberculosis

To the Editor: Zignol et al. (Sept. 15 issue) 1 suggest that within 10 years, molecular technologies will replace conventional phenotypic testing in surveys of drug resistance and surveillance of tuberculosis. To avoid the overestimation of drug resistance, wider implementation of testing for minimal inhibitory concentration (MIC) must occur first, for four main reasons. First, most commercial molecular assays do not provide information on the type of mutation. Second, different mutation types may be associated with different phenotypic MICs and different clinically relevant levels of resistance, which may lead to unnecessary discontinuation of a drug or incorrectly identifying a patient as . . .