Abstract Rationale In high TB-burden settings, the diagnostic yield of contact tracing is often limited by implementation barriers. We previously conducted a multi-site implementation trial showing that a multi-component, user-centered contact tracing strategy designed to address barriers can improve yield. In this nested qualitative study, we explored community health workers (CHW) adherence to the user-centered strategy. Methods After obtaining verbal consent, two male Ugandan social scientists conducted focus-group discussions(FGDs) with CHWs who implemented user-centered TB contact investigation during a stepped-wedge, cluster-randomized trial in 12 TB clinics and surrounding communities in Central Uganda. The strategy included branded educational materials, an electronic survey and contact enumeration application, a sputum collection video, and free motorbike transport, as well as regular performance feedback and CHW-led community of practice meetings. Two analysts used thematic analysis to characterize implementation fidelity and adaptation. Results 30 CHWs (13 women, 17 men) from all 12 sites participated in six FGDs. CHWs reported that some strategy components enforced fidelity, including (1) the electronic survey tool, and (2) community of practice meetings. The electronic survey tool prompted them to complete contact tracing tasks, including sharing TB testing results with contacts and conducting HIV screening. CHWs also reported adapting strategy components for several reasons, including to (1) better meet client needs, (2) compensate for resource limitations, (3) reach more clients, and (4) accommodate their own needs. For example, several reported deprioritizing educational booklet review among contacts with low interest or literacy. When resources were scarce, such as in a district without HIV testing capabilities, CHWs skipped HIV screening questions. CHWs adapted the strategy in unanticipated ways to enhance efforts to reach close contacts of TB patients. A video intended to demonstrate how to produce a sputum sample for TB testing was repurposed to illustrate TB symptoms, helping contacts identify others who might be experiencing similar symptoms. Similarly, when travel distance limits prevented motorbike riders from reaching close contacts, some CHWs persuaded them to exceed those limits to reach the contacts. Finally, one CHW described using the user-centered strategy brand logo on printed materials to identify close contacts when they presented to clinic for testing, while others used a diary provided during user-centered training to track their contact tracing activities. Conclusion CHWs demonstrated both fidelity to core components and context-responsive adaptations of intervention tools to optimize effectiveness in different local settings. This abstract is funded by: National Institute of Allergy and Infectious Diseases through grant R01AI104824 (JLD)
Steinbrook et al. (Fri,) studied this question.