Acceptability and Content Validity of Patient-Reported Outcome Measures Considered From the Perspective of Patients With Rheumatoid Arthritis.

Objective: To consider the acceptability and content validity of patient-reported outcome measures commonly used in rheumatoid arthritis by describing patients' perceptions of patient-reported outcome measures and comparing patients' responses on patient-reported outcome measures with their verbal accounts of disease impacts.

Methods: We used a sequential mixed methods approach, combining analysis of interviews and data from patient-reported outcome measures (from the Health Assessment Questionnaire, the Functional Assessment of Chronic Illness Therapy-Fatigue subscale, the EuroQol 5-domain instrument, the Short Form 36 health survey, and a visual analog scale [VAS] for pain, fatigue, sleep, and patient global assessment of disease activity). Qualitative analysis of patients' perceptions of patient-reported outcome measures informed a subsequent comparison between data from patient-reported outcome measures and verbal accounts of pain, fatigue, sleep, and functional limitations to assess the effectiveness of patient-reported outcome measures in communicating disease impact.

Results: The study included 18 patients. Although a few patients offered positive comments about patient-reported outcome measures, most doubted that patient-reported outcome measures could accurately convey their experience of symptoms and functional limitations. Patients discussed the ease of responding to questions, capturing and conveying symptoms, and concerns about the underreporting of symptoms and interpretation of responses. Compared with verbal accounts, patient-reported outcome measures often did not convey the personal significance of limitations; however, patient-reported outcome measures captured limitations that patients omitted or described with insufficient detail during interviews. Although verbal accounts of pain could be categorized into 3 levels of severity (pain without interference in activities, pain is not the worst ever experienced but interferes with activities, and pain is omnipresent), the pain VAS was more effective at conveying finer gradations in pain severity.

Conclusion: Although patients may feel that patient-reported outcome measures have certain shortcomings, patient-reported outcome measures also have advantages relative to verbal discussion for communicating symptoms and disease impact.