Aims- To demonstate the feasability of using rapid PCR-based typing techniques on a large scale and to identify risk factors for and the extent of recent transmission of tuberculosis in inner London. Methods- All available isolates of Mycobacterium tuberculosis from 1993 from inner London were typed using PCR based-typing methods (spoligotyping and heminested inverse PCR). Clusters of isolates with indistinguishable typing patterns were investigated to identify epidemiological links. Information, including data from the 1993 National Survey of Tuberculosis Notifications was used to compare risk factors in clustered and non-clustered cases Results- Of 694 eligible isolates 564 (81%) were typed. 2% were due to laboratory cross contamination. Of the remaining isolates 148 (27%) fell into 43 clusters. Plausible links (including household clusters, suspected hospital transmission, transmission related to homelessness and geographical proximity) were identified in 54% of clusters. Clustering was independently associated with pulmonary tuberculosis and birth in the UK. Homelessness was independently associated with being in a large cluster. Discussion- Assuming that clusters represent chains of recent transmission and each contains 1 index case, our findings suggest that at least 19% (148-43/555) of tuberculosis in inner London is due to recent transmission. This is lower than in San Francisco and New York in the early 1990's but higher than the traditionally assumed 10% in developed countries. The study shows that homeless people play an important role in tuberculosis transmission in London. It demonstrates that rapid PCR-based techniques are sufficiently robust to allow typing of large numbers of isolates. Such techniques could help to identify laboratory cross-contamination, and community and hospital outbreaks sufficiently early to allow effective intervention.
|Issue number||SUPPL. 6|
|Publication status||Published - Dec 1997|
Copyright 2006 Elsevier B.V., All rights reserved.