Abstract
Background/Objectives: Patient-level factors strongly influence antimicrobial resistance (AMR) through the pressure applied to healthcare professionals to prescribe antibiotics even for self-limiting viral infections, enhanced by knowledge and attitude concerns. This includes Africa, with high levels of AMR. However, validated measurement tools for African primary healthcare (PHC) are scarce. This study evaluated the reliability, structural validity, and interpretability of the Community Antimicrobial Use Scale (CAMUS) in South Africa. Methods: A cross-sectional survey was conducted with 1283 adults across 25 diverse public PHC facilities across two provinces. The 30-item theory-based tool underwent exploratory and confirmatory factor analysis (EFA/CFA), reliability, and validity testing. Results: EFA identified a coherent five-factor structure: (F1) Understanding antibiotics; (F2) Social and behavioural norms; (F3) Non-prescribed use; (F4) Understanding of AMR; and (F5) Attitudes. Internal consistency was strongest for knowledge and misuse domains (alpha approximation 0.80). Test-retest reliability was good-to-excellent (ICC: 0.72-0.89). CFA confirmed acceptable composite reliability (CR ≥ 0.63). Although average variance extracted (AVE) was low for broader behavioural constructs, indicating conceptual breadth, it was high for AMR knowledge (0.737). Construct validity was supported by positive correlations with health literacy (r = 0.48) and appropriate use intentions (r = 0.42). Measurement error metrics (SEM = 1.59; SDC = 4.40) indicated good precision for group-level comparisons. Conclusions: CAMUS demonstrated a theoretically grounded structure with robust performance in knowledge and misuse domains. While social and attitudinal domains require refinement, we believe the tool is psychometrically suitable for group-level antimicrobial use surveillance and programme evaluation in South African PHC settings and wider to help with targeting future educational programmes among patients.