M Conner, senior lecturer
P Norman, senior lecturer
D Willits, general practitioner
I Porter, research and development head
Background—In order that patient satisfaction may be assessed in a meaningful way, measures that are valid and reliable are required. This study was undertaken to assess the construct validity and internal reliability of the previously developed Patient Satisfaction Questionnaire (PSQ).
Method—A total of 1390 patients from five practices in the North of England, the Midlands, and Scotland completed the questionnaire. Responses were checked for construct validity (including confirmatory factor analysis to check the factor structure of the scale) and internal reliability.
Results—Confirmatory factor analysis showed that items loaded on the appropriate factors in a five factor model (doctors, nurses, access, appointments, and facilities). Scores on the specific subscales showed highly significant positive correlations with general satisfaction subscale scores suggesting construct validity. Also, the prediction (derived from past research) that older people would be more satisfied with the service was borne out by the results (F (4, 1312) = 57.10; p<0.0001), providing further construct validation. The five specific subscales (doctors, nurses, access, appointments, and facilities), the general satisfaction subscale, and the questionnaire as a whole were found to have high internal reliability (Cronbach's α = 0.74–0.95).
Conclusion—The results suggest that the PSQ is a valid and internally reliable tool for assessing patient satisfaction with general practitioner services.
(Quality in Health Care 2000;9:210–215)
Key Words: patient satisfaction; general practitioner services; questionnaire construction; construct validity; reliability