Article Text

FRI0432 Well-being, sleep and coping had good psychometric properties in rheumatoid arthritis (RA): A subanalysis of an etanercept trial
  1. L. Gossec1,
  2. Y. Brault2,
  3. I. Logeart2,
  4. M. Dougados1
  1. 1Cochin Hospital, Paris Descartes University
  2. 2Pfizer, Paris, France


Background In RA, there is increasing interest in “unusual” patient-reported outcomes (PROs), over and above pain, function and fatigue; namely, quality of sleep, well-being and ability to cope [1]. Although these outcomes are important from the patient’s point of view, are they useful as outcome measures and in particular, are they reliable, and responsive, i.e., do they improve when disease improves with treatment?

Objectives To assess the reliability and responsiveness after introduction of etanercept, of different PROs.

Methods Study: open-label 12-week trial of etanercept in RA [2]. PROs of interest: numeric rating scales (NRS) for sleep, physical and emotional well-being, and coping [3]. Comparison outcomes: “standard” PROs, namely, pain NRS, patient global assessment of disease activity (PGA), modified-HAQ (mHAQ) and fatigue NRS. Reliability: assessed between screening and baseline visits by intra-class correlation (ICC). Responsiveness: assessed between baseline and 12 weeks, by standardized response means (SRM). Only patients with at least 1 etanercept injection were analysed.

Results In all, 108 of the 120 screened patients received at least one etanercept injection: mean age, 54 (SD 13) yrs, mean disease duration 8 (SD 7), 75% were women, all were biologic naïve. Disease was active: mean DAS28 5.5 (SD 0.8), mean mHAQ 0.9 (SD 0.5), mean PGA, 6.5 (SD 1.9). At 12 weeks, mean DAS28 was 3.4 (SD 1.2).

Reliability was highest for sleep, and lowest for PGA and pain, whereas responsiveness was highest for pain and PGA, and lowest for mHAQ (Table).

Table 1. The 8 outcome measures, ordered by decreasing responsiveness

Conclusions NRS assessing sleep, well-being and coping were found to be generally as reliable and as responsive as “usual” PROs in RA. mHAQ showed poor responsiveness possibly due to floor effects. Sleep and coping were the least responsive “unusual” PROs in this trial, indicating these domains of health may be less accessible to biologic treatment. When assessing the patient’s perspective on treatment, it is feasible and valid to measure “unusual” domains of health by NRS.

  1. Kirwan JR et al, J Rheumatol. 2011;38(8):1711-5.

  2. Dougados M et al, J Rheumatol. 2011;38(11):2326-8.

  3. Gossec L et al, Ann Rheum Dis 2011;70(6):935-42.

Disclosure of Interest L. Gossec Consultant for: Abbott, BMS, Pfizer, Roche, UCB, Y. Brault Employee of: Pfizer, I. Logeart Employee of: Pfizer, M. Dougados Consultant for: Abbott, BMS, Pfizer, Roche, UCB,

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.