Geographical targeting of active case finding for tuberculosis in Pakistan using artificial intelligence software: a qualitative study embedded within the SPOT TB trial

Background: Tuberculosis (TB) remains a critical public health challenge in Pakistan. The SPOT-TB trial evaluated MATCH-AI; an AI tool designed to geographically target active case finding (ACF) by identifying sites for screening TB. Qualitative study was conducted to examine field team and stakeholder experiences to understand the human, organizational, and contextual factors effecting implementation. Methods: Five sub-recipients (SRs) were randomly selected; two districts per SR based on certain selection criteria. Thematic analysis was conducted on thirty In-Depth Interviews (IDIs) and two Focus Group Discussions (FGDs), guided by the Socio-Technical Systems (STS) framework. Findings: Themes included (1) MATCH-AI as a useful tool, (2) operational and contextual challenges, (3) challenges of the staff, (4) organizational readiness, and (5) stakeholder engagement across hierarchy. The staff valued MATCH-AI for reducing bias and external pressure, and it identified TB cases in previously overlooked areas. Local knowledge of staff was crucial as the AI didnot account for operational barriers and contextual issues in certain areas. Weak infrastructure, and inconsistent stakeholder engagement, the system lacked the readiness needed for a new technology to make optimal impact. Understanding of how MATCH-AI functioned varied across hierarchical levels diminishing the sense of ownership among field staff. Interpretation: MATCH-AI holds genuine potential to systematize TB screening and reduce selection bias. Yet it cannot replace the contextual intelligence of field staff like knowledge of community trust, gender norms, and security realities. Effective implementation demands reliable infrastructure, meaningful stakeholder engagement, and field staff orientation. AI integration succeeds only when technical solutions align with human and organizational readiness.