Article Text

SAT0573 Validity of The Short Form 6d Utility Measure in Early Axial Spondyloarthritis: Results from The Desir Cohort
  1. C. Gaujoux-Viala1,2,
  2. L. Gossec3,
  3. M. Dougados4,
  4. J.-P. Daurès2
  1. 1Rheumatology, Nîmes University Hospital
  2. 2EA2415, Montpellier University, Nîmes
  3. 3Rheumatology, Pitié-Salpêtrière Hospital;GRC-UPMC 08 – EEMOIS
  4. 4Rheumatology B, Cochin Hospital, Paris, France


Background The quantification of health improvements is essential, notably in the current context of increasingly expensive therapies and more and more limited resources. Preference-based measures of health have become important for estimating health states in calculating quality-adjusted life years (QALYs), an essential component of cost-utility analysis. The SF-6D is an indirect preference-based health-related quality of life instruments increasingly being used for economic evaluation of clinical interventions and health programs. Data on economic as well as QoL outcomes among early axial Spondyloarthtritis (axSpA) patients remain scarce.

Objectives To evaluate the validity of an utility measure, the SF-6D, in early axSpA especially concerning the ability of this measure to reflect the change in patients' condition over time.

Methods DESIR (Devenir des Spondyloarthropathies Indifférenciées Récentes) is a French, multicentre, longitudinal cohort of 708 patients with early inflammatory back pain suggestive of axSpA.

SF-6D utility measures were assessed in 607 patients over 1 year. To investigate whether the change in SF-6D is a valid measure of change in axSpA health status, we used Spearman's product-moment correlation to compare change scores for SF6D with those for external measures of health, the HAQ, SF36 physical component, SF36 mental component and AS-QOL from baseline to 6 and 12 months. Responsiveness was tested by the effect size (ES) at 6 and 12 months for the entire sample and subgroups by disease evolution (increase or decrease-stabilization in BASDAI). Bootstrap methods were used to estimate 95% confidence intervals [95% CI]. Sensitivity to change of the HAQ was calculated as a benchmark.

Results At baseline, mean value of SF-6D was 0.69±0.12 (range 0.30 to 0.95). The distribution was near normal. Few missing values were observed: 2.4%. No floor or ceiling effects were evidenced. Correlations of the SF6D change with change in HAQ and physical component of SF-36 scores were moderate at 6 months (r=-0.42 and 0.44, respectively). Correlations with change in mental component of SF-36 and AS-QOL scores were good (r=0.60 and -0.60, respectively). Correlations were stable over 1 year.

For the entire sample at 6 months, the SF-6D was more sensitive to change than the HAQ: ES 0.36 [95% CI 0.28;0.44] versus -0.22 [-0.28;-0.17]. The SF-6D was more responsive than the HAQ for improved condition: ES 0.45 [0.37;0.54] vs -0.32 [-0.38;-0.26] without difference for patients with deteriorated condition: ES SF-6D -0.23 [-0.41;-0.04] vs 0.33 [0.18;0.50] for HAQ. Results were similar at 12 months.

Conclusions The SF-6D is valid and able to reflect the change in patients' condition over time, especially improvement, in patients with early inflammatory back pain suggestive of axSpA.

Disclosure of Interest None declared

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.