Barrett A, DiBenedetti D, Phatak H, lloeje U. Structured review of patient-reported outcome instruments for assessing atrial fibrillation. Poster presented at the 2013 ISPOR 18th Annual International Meeting; April 23, 2013. [abstract] Value Health. 2013 May; 16(3):A293.

OBJECTIVES: To evaluate the characteristics of existing patient-reported outcome (PRO) instruments assessing atrial fibrillation (AF)-specific healthrelated quality of life (HRQOL) or symptoms, focusing on Food and Drug Administration (FDA) PRO guidance.

METHODS: This review was conducted in two phases. In phase 1, multiple sources (PubMed, Embase, Patient-Reported Outcome and Quality of Life Instruments Database [PROQOLID], and ClinicalTrials.org) were reviewed to identify instruments. In phase 2, data related to the development and measurement properties (internal consistency and testretest reliability, convergent and divergent construct validity, known-groups validity, and responsiveness) of these instruments were compared.

RESULTS: The literature search identified 625 studies. Six AF-specific measures were identified: four HRQOL instruments (Atrial Fibrillation Effect on Quality-of-Life [AFEQT], Atrial Fibrillation Quality of Life [AF-QoL], AF6 [a six-item AF questionnaire], Questionnaire for Quality of Life in AF [QLAF]) and two symptom assessments (Symptom Checklist—Frequency and Severity [SCL], University of Toronto Atrial Fibrillation Severity Scale [AFSS]). The AFEQT extensively involved AF patients in its development, establishing content validity. The AFEQT demonstrated robust measurement properties including internal consistency Cronbach’s alphas of 0.88-0.95 for domains, test-retest intraclass correlation coefficients of 0.7-0.8 for three of four domains and overall score, evidence of construct and known-groups validity, and responsiveness to treatment-related changes in a longitudinal study. The SCL and AFSS have demonstrated responsiveness in randomized controlled trial (RCT) settings but neither involved AF patients in the development process, and both have limited published psychometric validation data.

CONCLUSIONS: No existing PRO measure appears likely to support an FDA PRO label claim in AF. The AFEQT is the strongest instrument for measuring AF-related HRQOL; additional studies are needed to assess responsiveness in RCT settings. The SCL and AFSS have significant limitations related to development and validation. Qualitative research with AF patients may provide support for the content validity of the symptom measures.