Psychometric Properties of the Consumer Assessment of Healthcare Providers and Systems (CAHPS®) Clinician and Group Adult Visit Survey
Published in: Medical Care, v. 50, no. 11, suppl 3, Nov. 2012, p. S28-S34
Posted on RAND.org on October 28, 2012
BACKGROUND: The Consumer Assessment of Healthcare Providers and Systems (CAHPS®) Clinician and Group Adult Visit Survey enables patients to report their experiences with outpatient medical offices. OBJECTIVE: To evaluate the factor structure and reliability of the CAHPS Clinician and Group (CG-CAHPS) Adult Visit Survey. DATA SOURCE: Data from 21,318 patients receiving care in 450 clinical practice sites collected from March 2010 to December 2010 were analyzed from the CG-CAHPS Database. RESEARCH DESIGN AND PARTICIPANTS: Individual level and multilevel confirmatory factor analyses were used to examine CAHPS survey responses at the patient and practice site levels. We also estimated internal consistency reliability and practice site level reliability. Correlations among multi-item composites and correlations between the composites and 2 global rating items were examined. MEASURES: Scores on CG-CAHPS composites assessing Access to Care, Doctor Communication, Courteous/Helpful Staff, and 2 global ratings of whether one would Recommend their Doctor, and an Overall Doctor Rating. RESULTS: Analyses provide support for the hypothesized 3-factor model assessing Access to Care, Doctor Communication, and Courteous/Helpful Staff. In addition, the internal consistency reliabilities were ≥0.77 and practice site level reliabilities for sites with >4 clinicians were ≥0.75. All composites were positively and significantly correlated with the 2 global rating items, with Doctor Communication having the strongest relationship with the global ratings. CONCLUSIONS: The CG-CAHPS Adult Visit Survey has acceptable psychometric properties at the individual level and practice site level. The analyses suggest that the survey items are measuring their intended concepts and yield reliable information.