Item Analysis, Validity and Reliability of the Patient–Reported Outcome Measure of Pharmaceutical Therapy for Quality of Life (PROMPT-QoL) at Ramathibodi Hospital

Abstract

Objective: To conduct item analysis and investigate validity and reliabilityof the Patient-Reported Outcome Measure of Pharmaceutical Therapy forQuality of Life (PROMPT-QoL) questionnaire by using classical test theory.Method: In this study, data were collected from 400 outpatients who hadbeen using medications for at least 3 months at Ramathibodi hospital.Participants completed PROMPT-QoL, Medication Taking Behaviorquestionnaire (MTB), WHOQOL-BREF-THAI and EQ-5D-5L Thai version bythemselves and retested these questionnaires in one week later. Classicaltest theory was used to analyze item and test construct validity, criterionvalidity, internal consistency reliability and test-retest reliability. Results:Among 9 PROMPT-QoL domains, 7 domains were correlated with overallquality of life domain significantly. Different sex, age, levels of education,number of medicines per day, disease control level and incidence ofadverse drug reaction affected PROMPT-QoL scores in each domaindifferently. For criterion validity, PROMPT-QoL moderately correlated withWHOQOL–BREF-THAI and EQ-5D-5L (r = 0.25 – 0.50). Every domain inPROMPT-QoL showed excellent test-retest reliability (ICCs  0.75). Mostdomains revealed acceptable internal consistency reliability (Cronbach’salpha  0.7) except availability/accessibility domain. However this domainshowed acceptable corrected item-total correlations. Conclusion:PROMPT-QoL consists of 9 domains (43 items) that reveal acceptablevalidity and reliability. Further studies about responsiveness are required tocomplete psychometric properties.Keywords: item analysis, validity, reliability, Patient–Reported OutcomeMeasure of Pharmaceutical Therapy for Quality of Life,PROMPT-QoL