Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose: Defining flare in rheumatoid arthritis (RA) raises a complex issue related to the difference in perception or concept of flare between patients or health professionals. Altogether, defining a worsening in RA should clearly not depend on physicians’ perspectives only. The Strategy of Treatment in Patients with Rheumatoid arthritis (STPR) group developed a self-administered questionnaire (FLARE) to detect a past or present RA flare to be used to identify patients whose appointment should be anticipated or treatment revised (Berthelot et al, ARD 2011). Purpose: To validate the measurement properties of the FLARE self-administered questionnaire.
Methods:
the validation study was conducted in a prospective trial of RA patients with 3 main objectives to determine : 1) the truth of the tool.The internal validity was measured with a principal component factor analysis, the external validity evaluated by Spearman correlation coefficients with the DAS28, rheumatoid arthritis impact of disease [RAID], routine assessment of patient index data [RAPID], and HAQ scores . The measurement invariance was assessed with Rasch analysis ; 2) the reliability by intra-class correlation coefficients (ICC) and Bland et Altmann) ;and 3) the feasibility of using this score in daily practice. To be included patients should be at least 8 years old, have RA according to ACR criteria 1987 and/or 2012, evolving for more than 6 months. RA treatments (DMARDs, symptomatic treatments including stéroids) should be stable for at least 2 months. Patients were examined at baseline and 3 months and questionnaires were filled in 2 days before and at M3 clinical visit.
Results:
207 patients were recruited from 13 centres: 78.7% women, 57.7 years old, 84.4% RF+, 78.7% ACPA+, DAS28 was 2.9, CRP was 5.7 mg/l and 81.2% had erosive disease. At baseline, mean FLARE score (SD) – arithmetic mean of the 13 subscales – were 2.1 (+/-2.1) and 2.4 (+/-2.3) in groups 2 and 3 respectively.
Flare score internal validity was fair, with scores left-skewed. A substantial floor effect (1.7%), but no ceiling effect (0.7%), were observed.
The principal component factor analysis evidenced 2 dimensions of the FLARE questionnaire , one dealing with physical items and the other with emotional items. Partial credit model analysis further confirmed that one item on steroid treatment was inadequate.
Flare scores were correlated with DAS28 (0.43; p<0.0001) and (0.40; p<0.0001), RAID (0.69; p<0.0001) and (0.69; p<0.0001), RAPID3 (0.70; p<0.0001) and (0.70; p<0.0001, HAQ (0.47; p<0.0001) and (0.39 ; p<0.0001) in physical and emotional dimension respectively.
FLARE questionnaire was reproducible in both physical (ICC=0.91 [95%CI 0.88-0.94[) and emotional (ICC=0.93 [95%CI 0.90-0.95]) dimensions.
Conclusion: FLARE self-administered questionnaire is highly reproducible. Internal validity is fair with a good correlation with other questionnaires assessing disease activity, functional status and quality of life. This self-administered questionnaire may represent a tool to detect flare between visits to the physician.
Disclosure:
J. Morel,
Abbott Laboratories,
2;
J. M. Berthelot,
None;
A. L. Constantin,
None;
M. De Bandt,
Abbott Laboratories,
2;
P. Gaudin,
None;
O. Vittecoq,
None;
J. F. Maillefert,
None;
O. Meyer,
None;
T. Pham,
Abbott Laboratories,
2;
A. Saraux,
None;
E. Solau Gervais,
Abbott Laboratories,
2;
E. Spitz,
Abbott Laboratories,
2;
D. Wendling,
None;
F. Guillemin,
None;
B. Fautrel,
None.
ACR Meeting Abstracts - https://acrabstracts.org/abstract/validation-of-the-flare-self-report-questionnaire-for-assessing-flare-out-of-the-clinical-visit-in-rheumatoid-arthritis/