Contact tracing of tuberculosis: a systematic review of transmission modelling studies.

Abstract

The WHO recommended intervention of Directly Observed Treatment, Short-course (DOTS) appears to have been less successful than expected in reducing the burden of TB in some high prevalence settings. One strategy for enhancing DOTS is incorporating active case-finding through screening contacts of TB patients as widely used in low-prevalence settings. Predictive models that incorporate population-level effects on transmission provide one means of predicting impacts of such interventions. We aim to identify all TB transmission modelling studies addressing contact tracing and to describe and critically assess their modelling assumptions, parameter choices and relevance to policy. We searched MEDLINE, SCOPUS, COMPENDEX, Google Scholar and Web of Science databases for relevant English language publications up to February 2012. Of the 1285 studies identified, only 5 studies met our inclusion criteria of models of TB transmission dynamics in human populations designed to incorporate contact tracing as an intervention. Detailed implementation of contact processes was only present in two studies, while only one study presented a model for a high prevalence, developing world setting. Some use of relevant data for parameter estimation was made in each study however validation of the predicted impact of interventions was not attempted in any of the studies. Despite a large body of literature on TB transmission modelling, few published studies incorporate contact tracing. There is considerable scope for future analyses to make better use of data and to apply individual based models to facilitate more realistic patterns of infectious contact. Combined with a focus on high burden settings this would greatly increase the potential for models to inform the use of contract tracing as a TB control policy. Our findings highlight the potential for collaborative work between clinicians, epidemiologists and modellers to gather data required to enhance model development and validation and hence better inform future public health policy.