Concordance of drug resistance profiles between persons with drug-resistant tuberculosis and their household contacts: a systematic review and meta-analysis.

BACKGROUND: Household contacts of patients with drug-resistant tuberculosis are at high risk for being infected with Mycobacterium tuberculosis and for developing tuberculosis disease. To guide regimen composition for the empirical treatment of tuberculosis infection and disease in these household contacts, we estimated drug resistance profile concordance between index patients with drug-resistant tuberculosis and their household contacts. METHODS: We performed a systematic review and meta-analysis of studies published through July 24, 2018 and reported resistance profiles of drug-resistant tuberculosis index and secondary cases within their households. Using a random-effects meta-analysis, we estimated resistance profile concordance, defined as the percentage of secondary cases whose M. tuberculosis strains were resistant to the same drugs as strains from their index cases. We also estimated isoniazid/rifampin concordance, defined as whether index and secondary cases had identical susceptibilities for isoniazid and rifampin only. RESULTS: We identified 33 eligible studies, which evaluated resistance profile concordance between 484 secondary cases and their household index cases. Pooled resistance profile concordance was 54.3% (95% confidence interval [CI]: 40.7-67.6, I2=85%). Pooled isoniazid/rifampin concordance was 82.6% (95% CI: 72.3-90.9; I2=73%). Concordance estimates were similar in a sub-analysis of 16 studies from high tuberculosis-burden countries. There were insufficient data to perform a sub-analysis among pediatric secondary cases. CONCLUSION: Household contacts of drug-resistant TB patients should receive treatment for TB infection and disease that assumes that they, too, are infected with a drug-resistant M. tuberculosis strain. Whenever possible, drug susceptibility testing should be performed for secondary cases to optimize regimen composition.