Measuring primary care performance in Fragmented Health Systems: Psychometric Validation of the Short-Form Primary Care Assessment Tool across Rural and Urban Peru
Background: Peru’s fragmented primary care system requires brief, valid instruments to monitor care performance across diverse settings. Evidence on the psychometric performance of the short-form Primary Care Assessment Tool (PCAT-SF) in Peru remains limited. Methods: We conducted a cross-sectional psychometric validation study among 1678 adults attending primary care facilities in five Peruvian regions. Participants completed the 23-item PCAT-SF and the ESCOMA outpatient satisfaction scale. We assessed structural validity using confirmatory factor analysis with WLSMV estimation, internal consistency using ordinal alpha and McDonald’s omega, measurement invariance by sex and age, and convergent validity through correlations with ESCOMA scores. Reporting followed COSMIN guidelines. Findings: A five-factor correlated model showed adequate fit (CFI=0.949; TLI=0.940; RMSEA=0.066). The comprehensiveness domain performed better as two correlated subdomains. Factor loadings were strongest for coordination (0.857-0.938), whereas first contact showed weaker loadings (0.350-0.447) and low reliability (alpha=0.35). Measurement invariance across sex and age groups was supported. PCAT dimensions were positively associated with patient satisfaction, with continuity showing the strongest correlation (r=0.41). Interpretation: The PCAT-SF showed acceptable structural validity, reliability for most domains, and comparability across sex and age groups in diverse Peruvian primary care settings. The instrument may support routine assessment of primary care performance, although the first-contact domain requires further refinement.