Transmission of multidrug-resistant tuberculosis in the UK: a cross-sectional molecular and epidemiological study of clustering and contact tracing.

BACKGROUND Between 2000 and 2012 the number of multidrug-resistant (MDR) tuberculosis cases in the UK increased from 28 per year to 81 per year. We investigated the proportion of MDR tuberculosis cases arising from transmission in the UK and associated risk factors. METHOD We identified patients with MDR tuberculosis notified in England, Wales, and Northern Ireland between Jan 1, 2004, and Dec 31, 2007, by linking national laboratory and surveillance data. Data for laboratory isolates, including drug sensitivities and 24-mycobacterial interspersed repetitive-unit-variable-number tandem repeat (MIRU-VNTR) typing were obtained routinely from the National Tuberculosis Reference laboratories as part of national tuberculosis surveillance. We investigated clusters of cases with indistinguishable MIRU-VNTR profiles to identify epidemiological links. We calculated transmission using the n-1 method and established associated risk factors by logistic regression. We also assessed the likelihood of transmission to additional secondary active tuberculosis cases, identified through conventional contact tracing. FINDINGS 204 patients were diagnosed with MDR tuberculosis in the study period; 189 (92·6%) had an MIRU-VNTR profile. We identified 12 clusters containing 40 individuals and 149 unique strains. The proportion of cases attributable to recent transmission, on the basis of molecular data, was 15% (40 cases clustered-12 clusters/189 with a strain type). The proportion of cases attributable to recent transmission (ie, transmission within the UK) after adjustment for epidemiological links was 8·5% (22 cases with epidemiological links-six clusters/189 cases with a strain type). Being UK born (odds ratio 4·81; 95% CI 2·03-11·36, p=0·0005) and illicit drug use (4·75; 1·19-18·96, p=0·026) were significantly associated with clustering. The most common transmission setting was the household but 21 of 22 of epidemiological links were missed by conventional contact tracing. 13 secondary active tuberculosis cases identified by conventional contact tracing were mostly contacts of patients with MDR tuberculosis from countries of high tuberculosis burden. 11 (85%) of 13 shared the same country of birth as the index case, of whom ten did not share a strain type or drug resistance pattern. INTERPRETATION Transmission of MDR tuberculosis in the UK is low and associated with being UK born or illicit drug use. MIRU-VNTR typing with cluster investigation was more successful at identifying transmission events than conventional contact tracing. Individuals with tuberculosis who have had contact with a known MDR tuberculosis source case from a country of high tuberculosis burden should have drug-sensitivity testing on isolates to ensure appropriate treatment is given. FUNDING Public Health England.