Background Differences in physician evaluations, heterogeneity of disease and within patient variability contribute to the lack of a uniform definition of RA flare. As a result, a standardised approach to treatment of RA flare is not likely to exist. However, when presented with the same case scenario, rheumatologists’ choice of treatment should establish a standard of care.

Objectives To identify physician medication preferences and the role of steroids in the treatment of rheumatoid arthritis (RA) flare.

Methods A questionnaire describing two versions of the same case of a previously stable RA patient maintained on 17.5 mg/week of methotrexate (MTX), hydroxycholoroquine and NSAIDs, and presenting with signs and symptoms suggesting an increase in disease activity was developed; the second version included the addition of a standing dose of 5 mg of prednisone. Rheumatologists were asked to identify their choice (s) of first-line therapy for each version of the case. The questionnaire was mailed to rheumatologists (adult practice) (n = 995) who were identified through the ACR membership directory. Only physicians currently in practice were included.

Results 375 (37.7%) questionnaires were returned between March 12, 2000 and April 25, 2000. 280 qualified for analysis. The results (A: No baseline steroid use, B: 5 mg/d prednisone) below illustrates the frequency with which rheumatologists cited each drug or drug combination as their first choice of treatment for each scenario. The most common change in medication for both cases was addition/adjustment of prednisone (A: n = 248; B: n = 229, C2 = 4.58, p = 0.03). Of those given oral steroids for a limited amount of time, mean increase in daily dose of prednisone was A: 16.4 ± 12.5 mg, B: 18.3 ± 12.5 mg. Duration of increased dose was highly variable ranging from 1 day (bolus) to 3 months. Some rheumatologists chose to continue increased prednisone indefinitely, (A: n = 19; B: n = 14). There appeared to be no consistent relationship between steroid dose and duration. For both cases there was no difference in the number of rheumatologists who increased MTX dose (A: n = 84; B: n = 85). The number of rheumatologists who added another DMARD to the patient’s regimen were more in group A and differed significantly (x2 = 3.97, p = 0.046) from group B. Similarly, rates of initiation of anti-TNF therapy was higher in group A (x2 = 4.68, p = 0.03). However, rates of addition of leflunomide showed no statistically significant difference.

Conclusion Though the cases presented were accepted by rheumatologists as RA flare requiring intervention; no consistency of practice other than an increase in steroids could be identified. Further studies determining optimal therapy for increases in RA activity are necessary.