Validation and predictive capacity of a Dutch version of the FLARE-RA questionnaire within the context of a TNFi-tapering trial

Flare Assessment in Rheumatoid Arthritis (FLARE-RA) is a self-administered tool aiming to capture current or recent flares in rheumatoid arthritis (RA). We aimed to externally validate the FLARE-RA instrument and its existing cutoffs for flare detection within a bDMARD-tapering context in established RA. Data were analyzed from the Tapering Etanercept in Rheumatoid Arthritis (TapERA) trial, which studied the feasibility of tapering etanercept in patients with established RA in sustained remission. The English FLARE-RA was translated and cross-culturally adapted into Dutch, and internal consistency and test-retest reliability were evaluated with Crohnbach's alpha and intraclass correlation coefficient (ICC). Participants completed the FLARE-RA 3-monthly for 12 months. Accuracy and optimal cutoffs of FLARE-RA to detect DAS28-defined flares were assessed using area under the receiver operating characteristic curves (AUC). Association of FLARE-RA scores with current and future flares was studied using logistic generalized estimating equations (GEE). The Dutch FLARE-RA showed excellent internal consistency and test-retest reliability (Cronbach's alpha 0.96; ICC 0.96 [95% CI 0.70-1.00]). Discriminatory capacity of FLARE-RA to detect flares was acceptable (AUC 0.77, 0.80, and 0.72 for global, arthritis, and general symptoms subscales, respectively), with an optimal global score cutoff of 4/10. In GEE-models, higher FLARE-RA scores were associated with increased odds of both current and future flares. We successfully translated and cross-culturally adapted the FLARE-RA into a Dutch version and validated its capacity to detect flares in a bDMARD-tapering context in established RA. Finally, higher FLARE-RA scores might indicate an increased risk of future flares.