The potential yield of geographically targeted tuberculosis contact investigation in urban Uganda

We investigated the potential yield of conducting active case finding for tuberculosis (TB) within a defined geographic radius (50 or 100 m) around the households of individuals diagnosed with TB at health facilities. In a well-defined geographic area within Kampala, Uganda, residential locations were determined for 85 people diagnosed with TB at local health facilities over an 18-month period and for 60 individuals diagnosed with TB during a subsequent community-wide door-to-door screening campaign. Ten of the individuals diagnosed through community screening lived within 50 m of an individual previously diagnosed with TB in a local health facility (TB prevalence: 0.98%), and 15 lived at a distance of 50-100 m (prevalence: 0.87%). The prevalence ratio was 1.4 (95% confidence interval (CI): 0.7-2.9) for those <50 m and 1.2 (95% CI 0.6-2.2) for those 50-100 m, compared to >100 m. Using TB notifications to identify areas for geographically targeted case finding is at most moderately more efficient than screening the general population in the context of urban Uganda.