OBJECTIVE: To validate a patient-reported outcomes (PROs) measure for patients with hypertension. DESIGN: This was a validation study that employed mixed methods including classical test theory, item response theory, and reliability and validity evaluation. SETTING: The community health centers in two Chinese cities. PARTICIPANTS: Fifteen patients participated in the pilot survey, and 641 patients completed the PRO survey. MAIN OUTCOME MEASURES: Classical test theory and item response theory were applied to the item selection. Evaluations and adjustments were based on results of reliability, validity, responsiveness and feasibility analyses. RESULTS: The item selection process produced a final scale with 4 domains (physiology, psychology, society and treatment), 12 subdomains and 56 items. According to confirmatory factor analysis, the construct validity was adequate. The split-half reliability, Cronbach's α coefficients and responsiveness both in each subdomain and in the whole scale were also acceptable, as was feasibility. CONCLUSIONS: Mixed methods were beneficial for developing this scale and could be applied to PROs in other health areas. The instrument provides a means for comprehensive assessment of the impact of hypertension, and for quantification of benefits of hypertension interventions from the patients' perspective.
OBJECTIVE: To validate a patient-reported outcomes (PROs) measure for patients with hypertension. DESIGN: This was a validation study that employed mixed methods including classical test theory, item response theory, and reliability and validity evaluation. SETTING: The community health centers in two Chinese cities. PARTICIPANTS: Fifteen patients participated in the pilot survey, and 641 patients completed the PRO survey. MAIN OUTCOME MEASURES: Classical test theory and item response theory were applied to the item selection. Evaluations and adjustments were based on results of reliability, validity, responsiveness and feasibility analyses. RESULTS: The item selection process produced a final scale with 4 domains (physiology, psychology, society and treatment), 12 subdomains and 56 items. According to confirmatory factor analysis, the construct validity was adequate. The split-half reliability, Cronbach's α coefficients and responsiveness both in each subdomain and in the whole scale were also acceptable, as was feasibility. CONCLUSIONS: Mixed methods were beneficial for developing this scale and could be applied to PROs in other health areas. The instrument provides a means for comprehensive assessment of the impact of hypertension, and for quantification of benefits of hypertension interventions from the patients' perspective.