Article Text

FRI0215 Variability Over Time in Patient-Reported Outcome Measures in Patients with Ankylosing Spondylitis
  1. I. Essers1,
  2. M. Busch1,
  3. A. Boonen1,
  4. S. Ramiro2,
  5. D. van der Heijde2,
  6. R. Landewé3,4,
  7. A. van Tubergen1
  1. 1MUMC and CAPHRI, Maastricht
  2. 2LUMC, Leiden
  3. 3ARC, Amsterdam
  4. 4Atrium MC, Heerlen, Netherlands


Background The BASDAI, BAS-G, and a VAS for spinal pain (VAS-pain) are used to assess disease outcomes in AS, both in clinical practice and as well as in clinical studies. The intervals with which these patient-reported outcome measures (PROMs) are currently assessed vary widely, depending on the purpose of the assessment, and as such variability in outcome may occur. However, it is currently unknown what the extent and relevance of this variability is, and what the consequences of applying increasing intervals are on follow-up of individual patients.

Objectives To evaluate in AS 1) the variability of self-reported disease activity, patient global and spinal pain during 2 years of follow up, 2) the clinical relevance of this variability on a patient level, and 3) the effect of increasing intervals between two measurements on this variability on an observational level.

Methods Dutch patients from the Outcome in AS International Study (OASIS) completed the BASDAI, BAS-G, and VAS-pain every 2 months during 2 years. Mixed linear models were used to analyze time trends in the average scores over time. The smallest detectable change (SDC) and minimal clinically important difference (MCID) were used to detect relevant changes (worsening or improvement) between two measurements. On a patient level, the proportion of patients exceeding the MCID/SDC in intervals of 2 months was calculated. Next, the effect of increasing the intervals between two measurements (2- to 24-months) on the variability was investigated using all available observations from individual patients. To investigate whether higher scores at the beginning of an interval have a higher chance of improving than lower scores, and thus a higher chance of exceeding the SDC/MCID (and vice versa for lower scores at the beginning of an interval in the opposite direction), the observational data were stratified into two groups based on scores at the beginning of an interval (<4.0 versus >6.0).

Results 90 patients (mean age 47.3 (SD 11.4) years, 68% male, mean symptom duration 25.2 (SD 11.3) years) completed the PROMs at all-time points. The mean of each measure was not variable over time (p=0.885). On the patient level, however, large variability was found. For example, 80 (92%) and 73 (84%) patients exceeded, respectively, the MCID and SDC of the BASDAI at least once (in either direction) in the entire follow up period. On an observational level, exceeding the MCID/ SDC (in either direction) was frequently seen in all intervals, and this increased with prolongation of the interval (Table 1). For example for the BASDAI, MCID and SDC were exceeded in 39.5% and 27.3% of the 2-month intervals, respectively, and in 55.1% and 40.2% of the 24-month intervals. Exceeding the MCID/SDC in terms of improvement was more frequently observed in scores >6.0 at the beginning of an interval than in scores <4.0, and vice versa. Similar results were found for observations with the BAS-G and VAS-pain.

Conclusions Substantial variability in the BASDAI, BAS-G, and VAS-pain was found in individuals over time. Clinically relevant changes were frequently observed and increased with prolongation of the intervals. This intermediate information between two measurements, which may be clinically important, might be missed when intervals are prolonged.

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.