Preventability of incident cases of tuberculosis in recently exposed contacts

C R MacIntyre,A J Plant

Abstract:Setting: Contacts of tuberculosis (TB) cases are at risk for TB. If contact screening and intervention are effective, one would expect a reduced incidence of TB in contacts who have been screened. Objective: To measure the incidence of TB in contacts during a 2-year follow up, and to estimate the preventability of incident cases. Methods: A retrospective cohort study of 783 contacts screened in Victoria, Australia, in 1991. Contacts were matched with the TB registry for the following 2 years. Screening records were reviewed. Results: The rate of TB in contacts was 511/100000 population/year for the first 2 years. In Poisson regression models the only significant variable predicting disease was skin test reaction size. Six of eight incident cases were potentially preventable, with a lowest achievable incidence rate of 128/100000/year. Conclusion: Contacts who underwent screening for TB through a state screening programme had a high incidence of TB during the 2 year follow up. Published rates of TB of 425-670/100000 in untreated contacts suggests that the Victorian screening programme had minimal impact on the natural history of disease progression. Intrinsic programme factors such as the appropriateness of the guidelines, adherence to guidelines and rates of preventive therapy need to be evaluated. The devolution of the TB programme in the 1980s also reduced its efficacy. Systematic assessment of screening programmes for efficacy and outcome is part of good public health practice.