New Evidences About Multidimensionality of the Patient-Practitioner Orientation Scale (PPOS) Construct in Undergraduate Medical Students

Abstract Background Patient-centered care is becoming a paradigm in medicine. The Patient-Practitioner Orientation Scale (PPOS) is the only tool that measures the patient-centered attitude of healthcare students and professionals. Despite its spread, PPOS has had a poor process of scale modelling and validation and previous studies raised concerns about its psychometric robustness. Objective This study aims to investigate the PPOS psychometric properties, factor structure, and construct validity on a large sample of undergraduate medical students. Methods Participants of this cross-sectional single-center study are 1543 first-year medical students. PPOS item validity (internal consistency, test–retest reliability), factor structure (explorative principal axis factoring), and construct validity (convergent-discriminant validity, between-groups invariance) have been investigated. Results A three-factor not clearly defined solution explaining 34.4% of the variance and containing 14 items out of 18 was retained. Internal consistency was questionable for factor 1 (a = 0.657), poor for factor 2 (a = 0.566), and unacceptable for factor 3 (a = 0.399). Item-total correlations for factor 1 and factor 2 were > 0.3, except for item 6 (ITC = 0.218) and item 12 (ITC = 0.283). Item total-correlations for factor 3 were all < 0.3. Test–retest reliability was acceptable for factor 1 (ICC = 0.704) and factor 2 (ICC = 0.789) and questionable for factor 3 (ICC = 0.661). Construct validity and measurement invariance across groups were satisfactory. Conclusion Findings in this study corroborate previous evidences about PPOS psychometric limitations and provide new evidence about the multidimensionality of patient-centeredness construct.