File
Palmer S.1, Manns S.1, Cramp F.1, Lewis R.2, Clark E.3
1University of the West of England, Allied Health Professions, Bristol, United Kingdom, 2North Bristol NHS Trust, Bristol, United Kingdom, 3University of Bristol, Musculoskeletal Research Unit, Bristol, United Kingdom
Background: Joint Hypermobility syndrome (JHS) is characterised by multiple joint laxity and pain. It is associated with a range of sequelae such as fatigue, proprioceptive deficits, repeated cycles of injury (Terry et al 2015) and psychological conditions (Smith et al 2014). JHS is likely to be very common in musculoskeletal practice, with 30% of patients referred to a musculoskeletal triage service in the United Kingdom (UK) meeting the diagnostic criteria (Connelly et al 2015). The Bristol Impact of Hypermobility (BIoH) questionnaire is a condition-specific patient-reported outcome measure developed for adults with JHS (Palmer et al 2006). It gives a maximum score of 360, with higher scores representing more severe impact. It has strong concurrent validity (r=-0.725, n=615) with the physical component score of the Short Form 36 (SF36) questionnaire (Palmer et al 2006) however other important psychometric properties have yet to be established.
Purpose: The primary aim was to establish the test-retest reliability of the BIoH questionnaire across a two-week period. A secondary aim was to calculate 95% limits of agreement as an initial estimate of the likely minimum clinically important difference (MCID).
Methods: Adults with JHS were recruited through the Hypermobility Syndromes Association (HMSA), a UK patient organisation. Recruitment packs were distributed to all HMSA members who had given permission to be contacted about research (n=1 080). Packs included a participant information sheet, consent form, demographic questionnaire and the BIoH and SF36 questionnaires. Those who returned a consent form and questionnaires were sent repeat questionnaires two weeks later. A global rating of change (GRC) scale asked patients "With respect to your joint hypermobility, how would you describe yourself now compared with two weeks ago when you last completed this questionnaire (please circle a number)?" Response options ranged from -5 ("Very much worse") to 0 ("Unchanged") to +5 ("Very much better"). Analysis was only conducted on ´stable´ patients (-1 to +1). Intraclass Correlation Coefficient (ICC) (one-way random, single measure) and 95% limits of agreement were calculated.
Results: n=462 responses were received (43% response rate). n=98 were excluded, mainly due to non-return of week 2 questionnaires or incomplete data. n=234 were ´stable´ across the two weeks (GRC -1 to +1) and were included in test-retest analysis (95% women, mean ± SD age 44.4 ±13.9 years, BIoH score 223.4 ± 54.1). The ICC demonstrated excellent test-retest reliability (ICC 0.922; 95% CI 0.901, 0.940). The 95% limits of agreement was calculated to be 46 points.
Conclusion(s): The BIoH questionnaire demonstrated excellent test-retest reliability. An initial estimate of the MCID suggests that a change of 46 points on the 360 point scale might be considered clinically significant (equivalent to 13% of the maximum score or 21% of the mean score observed in the current sample). However this MCID estimate is on the basis of data variability in ´stable´ patients and would need to be verified through analysis of sensitivity to change with interventions of known efficacy.
Implications: The results provide further evidence of the psychometric properties of the BIoH questionnaire, enhancing confidence in its use for clinical research and practice.
Funding acknowledgements: Faculty of Health & Applied Sciences, University of the West of England, Bristol, UK.
Topic: Musculoskeletal
Ethics approval: Faculty of Health & Applied Sciences Ethics Sub-Committee, University of the West of England, Bristol (HAS/15/01/99).
All authors, affiliations and abstracts have been published as submitted.
