PRS59 PSYCHOMETRIC EVALUATION OF ASTHMA-SPECIFIC PATIENT-REPORTED OUTCOME MEASURES USED IN ADULTS

Researchers always strive to find a patient-reported outcome measure (PROM) with the best psychometric properties that suites their research. To help researchers in their decision, this study aims to review and compare the psychometric properties of asthma adults’ specific PROMs. A literature review was conducted; seven English and adult-based PROMs were identified. Specific PROMs were selected for review and comparison based on Classical Test Theory (CTT), Item-Response Theory (IRT), and the number of citations. PROMs were evaluated based on seven criteria: conceptual model, practicality, versatility (race, gender, and severity), responsiveness, reliability (internal consistency and test-retest for CTT PROMs, marginal reliability and reliability plot over the scale range for IRT PROM), validity (face, convergent, and divergent [Streiner et.al, 2015]), and depth. The selected PROMs included Juniper’s Asthma Quality of Life Questionnaire (AQLQ-J) and Mark’s AQLQ (AQLQ-M) – based on CTT, and the RAND negative Impact of Asthma on Quality of Life short form-12 (RAND-IAQL-12) – based on IRT. AQLQ-J and AQLQ-M were cited 1254 and 451 times, respectively. Although the selected PROMs based their item generation on information from asthma patients, only AQLQ-M and RAND-IAQL-12 used factor analysis (FA) in the item reduction process. The RAND-IAQL-12 satisfied the dimensionality criterion using confirmatory FA (CFA) to confirm unidimensionality; neither AQLQ-J nor AQLQ-M used CFA. All reviewed PROMs were brief (completed in <15 minutes), self-administered, versatile (used in different races, genders, and severities), easy to score, and responsive. All PROMs satisfied reliability for group-level decision-making (>=0.7), and RAND-IAQL-12 presented reliability plot over the scale range. All PROMs satisfied the face validity criterion, and the RAND-IAQL-12 fully satisfied and AQLQ-J partially satisfied convergent validity. None of the PROMs satisfied divergent validity or provided depth data. Although none of the reviewed PROMs satisfied all evaluation criteria, the RAND-IAQL-12 development satisfied most of the evaluation criteria.