P91 Incorporating tuberculosis strain typing data into routine contact tracing investigations: experience from the field

Strain typing of tuberculosis (TB) isolates by 24 loci mycobacterial interspersed repetitive unit-variable number tandem repeats (MIRU-VNTRs) is now a routine laboratory tool for TB control, but its utility in informing contact tracing and public health action has not been well reported in the United Kingdom. Since November 2011 we have routinely held typing meetings and undertaken cluster investigation. Over 18 months, 68 clusters were discussed. Fifty-five (81%) clusters were small (2–5 patients), 7 (10%) were medium (6–14 patients) and 6 (9%) were large (>15 patients, median = 42, IQR = 26–52). Enhanced epidemiological investigation was undertaken in 27/68 (40%) clusters. Typing meetings alone readily identified 20 definite epidemiological links between 46/458 (10%) cases. In 15 cases, 9 definite or probable links were not supported by genotyping, leading to expanded screening in one workplace. 112 extended interviews were done. A further 23 definite links between 77 (17%) cases, 2 probable links between 5 (1%) cases and 24 possible links between 72 (16%) cases were found. Expanded screening as a direct result of strain typing and cluster investigation occurred in 4/6 settings where non-household transmission was identified (a factory, 2 places of worship, a hospital, a hostel and a pub). An additional 124 contacts were identified. 65 attended screening, 21 latent TB cases were treated and 1 active TB case was found. Routine incorporation of strain typing data in contact tracing improves diagnosis of latent and active infection but requires investment in data management systems and human resource for enhanced epidemiological investigation.