PT - JOURNAL ARTICLE AU - Liesbeth Heuft-Dorenbosch AU - Debby Vosse AU - Robert Landewé AU - Anneke Spoorenberg AU - Maxime Dougados AU - Herman Mielants AU - Hille van der Tempel AU - Sjef van der Linden AU - Désirée van der Heijde TI - Measurement of spinal mobility in ankylosing spondylitis: comparison of occiput-to-wall and tragus-to-wall distance. DP - 2004 Sep 01 TA - The Journal of Rheumatology PG - 1779--1784 VI - 31 IP - 9 4099 - http://www.jrheum.org/content/31/9/1779.short 4100 - http://www.jrheum.org/content/31/9/1779.full SO - J Rheumatol2004 Sep 01; 31 AB - OBJECTIVE: To investigate if the tragus-to-wall distance (TWD) is more reliable compared to the occiput-to-wall distance (OWD) as a measurement for thoracic spine extension in patients with ankylosing spondylitis (AS). METHODS: Data from the OASIS cohort, an international longitudinal observational study on outcome in AS, were used. Measurements of OWD and TWD were performed at baseline and at 6, 12, 18, and 24 months. Paired data of Tx and Tx+6 months were used to perform test-retest measurements (intraclass correlations, limits of agreement, and interperiod correlation matrix). Bland and Altman plots were constructed to investigate the agreement between both observations, assuming that there was no true change between 0 and 6 months. To investigate whether a change in disease activity would have influenced the results, limits of agreement were calculated in a subgroup of patients with a stable Bath Ankylosing Spondylitis Disease Activity Index (BASDAI; defined as a maximum BASDAI change of +/- 1) between T0 and T6 and compared with the results of the whole group. Limits of agreement were also calculated for kyphosed patients only. RESULTS: The test-retest intraclass correlations were between 0.94 and 0.96 for OWD and between 0.93 and 0.95 for TWD. The direct measurement-remeasurement correlation calculated by extrapolation of the interperiod correlation regression line was 0.92 for OWD and 0.90 for TWD. OWD and TWD showed comparable reliability on the entire value of scores. The lower 95% limit of agreement was between -3.4 cm and -2.5 cm for OWD and between -3.4 cm and -3.1 cm for TWD. The upper limit of agreement was between 3.1 cm and 4.2 cm for OWD and between 2.9 cm and 3.9 cm for TWD. In all patients as well as in kyphosed patients only, limits of agreement were comparable between OWD and TWD. The patterns of the scatterplots according to Bland and Altman were similar for OWD and TWD. Measurement error was more pronounced in kyphosed patients compared to patients with a normal thoracic extension. However, over the entire range of kyphosis, measurement error was similar. CONCLUSION: OWD and TWD are equally reliable in assessing thoracic spine extension. Although the TWD is in general easier to perform in AS patients compared to OWD, we recommend the OWD measurement over TWD: in OWD measurement a value of zero easily distinguishes patients with normal thoracic spine extension from kyphosed patients.