Files
Abstract
Problem: Globally tuberculosis incidence is decreasing by approximately 1 2% per year. If these trends continue 2050 global targets for tuberculosis control will not be met. Supplementary interventions are needed to supplement current control strategies. Household contact tracing has been widely recommended but not implemented in low-income, high-burden settings. Goal: Expand the evidence-base on the effectiveness of household contact tracing of tuberculosis cases and make the use of household contact tracing more efficient for National Tuberculosis Programs in low-income, high-burden settings. Methods: We conducted four separate studies using separate outcomes and methodologies. First, we conducted a systematic review and meta-analysis of studies that measured either latent or active tuberculosis in both household contacts of tuberculosis cases and unexposed controls. We compared prevalence rates in these two groups across studies. In the three subsequent studies, we used data from theKawempe Community Health study set in Kampala Uganda. In aim 2, we derived a predictive risk score to identify household contacts most at risk for coprevalent tuberculosis disease. In the third aim, we investigated risk factors for incident tuberculosis disease amongst household contacts. In the fourth aim, we attempted to validate two clinical algorithms, the World Health Organizations symptom-based algorithm and the Chan Risk Score, to detect coprevalent and incident tuberculosis disease in child household contacts. Results: In Aim 1, household contacts were 9.8 (95% CI, 4.024.0) times more likely to have coprevalent tuberculosis disease compared to unexposed control groups. In Aim 2, we created two proposed scores that would minimize the number of contacts screened but found over 80% of contacts. In Aim 3, HIV-infection, young age, cavitary status of the index case, and past active tuberculosis disease in contacts were strong risk factors for incident tuberculosis disease. In Aim 4, the WHO symptom-based algorithm was high efficient at finding coprevalent and incident diseased children however the Chan Risk Score validated poorly. Conclusions: Household contact tracing is a highly effective tuberculosis case finding tool that must be used to supplement current tuberculosis control. New tools that optimize household contact tracing, such as predictive risk scores and symptomatic algorithms may entice National Tuberculosis Programs in low-income, high-burden settings to use this intervention.