Could repeated prevalence surveys lead to decreasing tuberculosis prevalence in a community?

Abstract

Setting: Tiruvallur District, South India, where one baseline tuberculosis (TB) disease prevalence survey followed by three repeat prevalence surveys were conducted every 2.5 years between 1999 and 2008, and where the DOTS strategy was implemented in 1999. Objective: To rule out the possibility that the observed decline in TB prevalence was influenced by conducting repeat prevalence surveys, we compared the findings from two surveys: the third repeat survey conducted in 2006-2008 and an independent single survey in a neighbouring area conducted in 2008-2009. Design: An independent survey was conducted to estimate the prevalence of TB in the same district in 2008–2009 using a different set of villages and employing repeat survey methodology. The independent survey findings were compared with those of the third repeat survey. Results: The estimated prevalence rate of culture- and smear-positive TB was respectively 401 per 100,000 and 186 per 100,000 population in the third repeat survey area. The corresponding rates were 340 and 184/100,000 in the independent survey area. The difference in prevalence was not significant (culture P = 0.09; smear P = 0.93). Conclusion: The estimated prevalence rates in the two different sample survey areas were comparable, indicating that the repeated prevalence surveys in the study area did not influence the observed decline in TB disease prevalence.