Identifying and Prioritizing Barriers to TB Prevention and Care in High-Burden Countries: A Community-Engaged Approach Using Best-Worst Scaling

Objective: To comprehensively identify the multi-level barriers to tuberculosis prevention and care in high-burden settings, and prioritize them according to their perceived harm and modifiability. Methods: We conducted a multi-phase stakeholder-engaged study, synthesized barriers from two scoping reviews, and then convened regional workshops in Hyderabad, India, and Nairobi, Kenya, prioritizing TB survivors, community advocates, and frontline healthcare workers (HCWs). Participants refined existing barriers and added new ones. They then completed a Best-Worst Scaling (BWS) exercise to assess the perceived impact (harm) of each barrier, and Likert-based questions to assess perceived modifiability. Hierarchical Bayes (HB) modeling was used to generate mean impact scores (MIS) for each barrier, and mean scores for barrier modifiability (range, 1-4) were calculated. Findings: 81 stakeholders from 28 countries participated, and 65 completed the survey (24% community representatives/advocates, 36% frontline HCWs; 17% were TB survivors). Stakeholder input expanded the barrier set from 31 to 39, with most newly added barriers at the health systems level. The BWS identified systems-level drug and supply challenges (MIS=5.8), patient/community-level financial factors (MIS=5.2), and inadequate provision of holistic care (MIS=4.9) as the highest-impact barriers; 7 of the top 10 barriers overlapped across regions. Asia participants placed 5 barriers, and Africa participants 10 in the high-impact/high-modifiability priority quadrant; four were shared across both: patient/community knowledge, HCW knowledge, HCW attitudes, and lack of community engagement mechanisms. Conclusion: Incorporating the lived experience of TB-affected groups revealed substantial gaps in literature-derived TB care barriers. The high-impact, modifiable barriers identified provide actionable targets for programs in high-burden settings.