Background/Purpose: NIH Patient-Reported Outcomes Measurement Information System (PROMIS) has created highly precise and efficient instruments to measure patient-reported outcomes across health domains relevant to chronic medical conditions. PROMIS is available without charge, and many domains offer item banks for computer adaptive testing (CAT) through Assessment Center, a secure online management tool, in addition to paper-based short forms (SF) and single domain items that form the Global Health measure. PROMIS Fatigue item bank assesses both the experience of fatigue from mild tiredness to an overwhelming, debilitating, and sustained sense of exhaustion as well as impact on physical, mental and social activities. However, the feasibility, validity and psychometric performance of the three PROMIS measures has not been reported in persons with rheumatoid arthritis (RA) where fatigue is highly prevalent and disabling.

Methods: Data are from the baseline visit of the first 125 RA patients enrolled in an ongoing study integrating PROs into routine care at an academic RA clinic. Three PROMIS fatigue measures (CAT [average 5-8 items], 7-item SF and Global Health single item of fatigue) were administered in random order along with other legacy measures immediately before a routine clinic visit. Spearman correlations were used to evaluate convergent validity.

Results: Participants were mostly female (79%) and white (86%) with a mean (SD) age of 56 (13) and RA duration of 12 (9) yr; 10% were diagnosed ≤ 2 yr. In general, PROMIS fatigue measures correlated highly with VAS of fatigue, pain and patient global, moderately with mHAQ, morning stiffness, CDAI and MD global and weakly with joint counts (see Table). A dose response relationship across CDAI levels was evident in all measures.

Conclusion: Our study contributes preliminary evidence of the feasibility of the 3 PROMIS fatigue instruments as part of routine RA visits and of construct validity with legacy measures and clinical indicators. All PROMIS fatigue measures correlated highly with the fatigue VAS; while the Fatigue CAT was most strongly associated with legacy measures and offers greater precision, the 7-item short form and single Fatigue item also performed adequately and are useful when paper-based or briefer versions are needed. PROMIS scores can be compared against population norms. Further evaluation of responsiveness and validity across diverse groups of patients and settings is warranted.