Article Text

FRI0204 Patient Global Estimates of Status Vary More than Physician Global Estimates, and Estimates Become Increasingly Discordant from ONE Another at Both High and Low Extremes of Disease Activity
  1. K.A. Gibson1,
  2. G. Luta2,
  3. T. Pincus3
  1. 1Liverpool Hospital, Liverpool, NSW, Australia
  2. 2Georgetown University, Washington, DC
  3. 3NYU School of Medicine, New York, NY, United States


Background Global estimates by both patient (PATGL) and physician (DOCGL) are concordant within 2 units on a 0-10 visual analog scale (VAS) in about 40-65% of encounters, while discordance of PATGL>DOCGL occurs in about 25-60% and DOCGL>PATGL in about 10-20%.1 DOCGL is expected to be higher if DOCGL>PATGL than if DOCGL=PATGL, and might be expected to be lower if PATGL>DOCGL. However, this matter has not been examined extensively.

Objectives To analyze DOCGL and PATGL according to whether DOCGL>PATGL, DOCGL=PATGL or PATGL>DOCGL in 225 patients with various rheumatic diseases seen in a general rheumatology clinic in Australia.

Methods All patients completed a multidimensional health assessment questionnaire (MDHAQ), which includes PATGL, pain (PN), and physical function (FN), each scored 0-10. RAPID3 (routine assessment of patient index data) is calculated as a 0-30 scale from 3 0-10 scores for FN, PN and PATGL. The rheumatologist assigned a DOCGL for each patient on a 0-10 VAS identical to the PATGL VAS. PATGL and DOCGL were compared for concordance or discordance in 225 individual patients, classified in 3 groups: PATGL=DOCGL, i.e., both scales within 2 units of each other; PATGL>DOCGL by ≥2 units; or DOCGL>PATGL by ≥2 units. DOCGL, PATGL and RAPID3 were compared in these 3 groups, using Kruskal-Wallis tests to analyze statistical significance.

Results Among 225 encounters, PATGL>DOCGL was seen in 80 (36%), PATGL=DOCGL in 118 (52%) and DOCGL>PATGL in 27 (12%), similar to published reports.1 Median DOCGL was 3.5 in concordant encounters (PATGL=DOCGL), compared to 4.0 when DOCGL>PATGL or when PATGL>DOCGL (p=0.078). By contrast, median PATGL was 3.8 when PATGL=DOCGL, compared to 1.0 when DOCGL>PATGL and 7.0 when PATGL>DOCGL (p<0.001). The range of medians in all categories was 3.0–5.5 for DOCGL, 1.0–7.5 for PATGL and 5.8–18.3 for RAPID3.

Median (IQR) values for DOCGL, PATGL and RAPID3 according to whether PATGL>DOCGL, PATGL=DOCGL or DOCGL>PATGL in all patients, RA patients, and other patients.

Conclusions Discordance between PATGL and DOCGL is driven largely by variation in PATGL rather than DOCGL. DOCGL was higher in patients in whom DOCGL>PATGL vs DOCGL=PATGL, as expected, but also was higher in encounters in which PATGL>DOCGL, which was unexpected. Efforts to understand and reduce discordance may result in better care and outcomes.


  1. Barton JL, Imboden J, Graf J, et al. Arthritis Care Res 2010;62:857-64

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.3945

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.