Artificial Intelligence in Strengthening Primary Healthcare Delivery in Pakistan: A Feasibility and Application Assessment

Abstract

Background: Primary healthcare facilities in Pakistan may benefit from artificial intelligence (AI)-supported triage and clinical decision support, but provider readiness and contextual barriers remain uncertain. Objective: To assess AI readiness, perceived clinical utility, triage potential, ease of use, and implementation barriers among primary healthcare providers in Southern Punjab, Pakistan, and to compare outcomes across professional cadres. Methods: This cross-sectional survey included 96 medical officers, lady health supervisors, and allied health professionals working in Basic Health Units and Rural Health Centers between August and December 2025. Participants were recruited through non-probability purposive sampling. Readiness and technology acceptance were assessed using a modified Artificial Intelligence Readiness Scale for Medical Professionals and adapted Technology Acceptance Model domains. Descriptive statistics, Pearson correlations, one-way analysis of variance, and Tukey post-hoc comparisons were used. Results: The response rate was 91.4%. Mean AI readiness was 3.12 ± 0.64, perceived clinical utility was 3.78 ± 0.58, triage efficiency potential was 3.85 ± 0.61, perceived ease of use was 2.64 ± 0.72, and perceived barriers were 4.15 ± 0.53. Awareness differed across cadres (F = 8.14, p < 0.001), with medical officers reporting higher scores than lady health supervisors and allied health professionals. Technological competence correlated positively with triage-tool acceptance (r = 0.64, p < 0.001). Conclusion: Providers perceived substantial potential value in AI-supported primary care, but cadre-specific readiness gaps and infrastructural and ethical barriers may constrain implementation. Workforce preparation and infrastructure development should precede prospective pilot evaluation.