Article Text


SAT0236 A longitudinal evaluation of the effect of a decision-support system on outcome in rheumatoid arthritis
  1. J Fransen,
  2. G Stucki,
  3. T Langenegger,
  4. M Seitz,
  5. BA Michel
  1. SCQM C/o. Department of Rheumatology, University Hospital, Zurich, Switzerland


Background Management of patients with RA is challenging and poses specific problems. With the help of a decision-support system (Swiss Clinical Quality Management in RA) that gives the rheumatologist feedback on treatment outcomes, the treatment strategy can be adjusted. It is assumed that control of disease activity avoids or reduces signs and symptoms in the short term, and reduces the development of joint damage and disability in the long term.

Objectives To study if the decision-support system is effective in a) reducing disease activity, and b) maintaining the level of joint damage and disability in RA patients.

Methods 264 RA patients from 63 rheumatologists were included in a longitudinal evaluation, starting with a 3 month control period P0 (no feedback) followed by a 12 month intervention period P1 (with provision of feedback). Patients are nested in physicians, crossed with time nested in period. The main outcome variable is the Rheumatoid Arthritis Disease Activity Index (RADAI), a questionnaire on signs and symptoms in RA. The RADAI ranges from 0–10, higher scores indicate more disease activity. Disease activity (DAS28), disability (HAQ), and joint damage (X-ray score) were additionally measured in P1. To compare the development of the RADAI in P0 with P1, multilevel analysis (linear regression with random coefficients) was used; the unit of time is 30 days. For analysing differences in P1 for DAS28, HAQ and X-ray score, the paired t-test was used.

Results All patients had at least 1 feedback report at start of P1, 40% had 2, 23% had 3–6 reports. The regression model shows no time effect for RADAI in P0: -0.02 (p = 0.75) and a small time effect in P1: -0.04 (p < 0.0001). The latter means ca. -0.5 RADAI point in 12 months. The difference between P0 and P1 is significant (p = 0.0011). The time effects were not influenced by: age, gender, disease duration, baseline level of disease activity, number of consultations, or changes in DMARD therapy. Of the 202 patients on DMARD therapy during P1, n = 114 had the same therapy at begin and end. Over P1, the DAS28 was reduced with -0.4 (1.2) points (p < 0.0001), the HAQ remained stable with -0.1 (0.4) points (p = 0.03), and the X-ray score increased with +6.4 (10.4) points (p < 0.0001).

Conclusion These results point to a significant but small effect of a decision-support system in reducing disease activity in RA. Over 12 months, the medication strategy was changed in only a minority of patients. The SCQM may increase its efforts to stimulate physicans to adjust treatment, e.g. by the implementation of explicit treatment guidelines. Other possible reasons for suboptimal effectiveness need to be explored, including limited treatment options and medication compliance.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.