Tuberculosis in a shopping centre, Portugal, 2004-5

Genotyping enables to confirm or exclude a tuberculosis (TB) cluster. Excluding the link between cases is particularly important in countries with intermediate/high incidence of TB where the emergence of several TB cases in a particular location in space or time (higher than the expected) could be explained by chance alone. During 2004 and 2005, five TB cases occurred in five shops of a Portuguese shopping centre which employed a total of about 1000 workers. After an epidemiological survey, 52 close contacts were identified and screened. Latent tuberculosis infection was diagnosed in 10 contacts (eight family members and two work colleagues of cases). Genotyping of the Mycobacterium tuberculosis isolates revealed no link between the cases. For this reason no screening of all staff of the shopping centre was carried out. However, close contacts (52) and all fellow workers (1000) were kept under surveillance for two years, and no additional cases were diagnosed. The present analysis demonstrates that the exclusion of a chain of ongoing transmission by genotyping for the investigation of a cluster is cost-effective from the perspective of the public health service, because it allows to avoid unnecessary large scale screening operation and instead to direct resources to more effective measures of TB control.