An instrument to measure patient satisfaction with healthcare in an observational database: results of a validation study using data from CaPSURE

Abstract

Objective: To validate a satisfaction measure for use in longitudinal, prospective studies of patient care.

Study design: Patients with biopsy-confirmed prostate cancer (n = 228) who were enrolled in CaPSURE (Cancer of the Prostate Strategic Urologic Research Endeavor) completed a self-administered questionnaire that included a health-related quality-of-life and satisfaction measure. A subset of patients completed the questionnaire again within 30 days.

Methods: The satisfaction measure contained 6 individual subscales: overall satisfaction with care, contact with providers, confidence in providers, communication skills, humaneness, and a summary scale. Six items surveyed patients' willingness to participate in decision making (participatory style), and these were averaged into a single score. Variability, reliability, stability, and validity were evaluated.

Results: Responses to the items varied substantially. The overall satisfaction scale demonstrated good internal consistency reliability (Cronbach alpha = 0.82) and moderate test-retest reliability (0.62), and it could discriminate between groups of individuals expected to differ with regard to satisfaction (by age and disease stage). Subscale internal consistency reliability (0.37-0.54) and stability (0.38-0.63) were weaker, suggesting that only a single scale should be reported. The participatory scale performed poorly and could not be recommended for future use.

Conclusion: The overall satisfaction measure developed for this study demonstrated good reliability and validity and should be useful in other population-based studies in conjunction with other outcome measures.