Johanna Darrah1, Lesley Wiart2, Jan Willem Gorter3, Mary Law4. 1. J. Darrah, PT, PhD, Department of Physical Therapy, Faculty of Rehabilitation Medicine, University of Alberta, 2-50 Corbett Hall, Edmonton, Alberta, Canada T6E 2S7. johanna.darrah@ualberta.ca. 2. L. Wiart, PT, PhD, Clinical Support, Glenrose Rehabilitation Hospital, Edmonton, Alberta, Canada, and Department of Physical Therapy, Faculty of Rehabilitation Medicine, University of Alberta. 3. J.W. Gorter, MD, PhD, FRCPC, Department of Pediatrics, CanChild Centre for Childhood Disability Research, McMaster University, Hamilton, Ontario, Canada. 4. M. Law, PhD, OTReg(ON), School of Rehabilitation Science and CanChild Centre for Childhood Disability Research, McMaster University.
Abstract
BACKGROUND: Serial joint range-of-motion (ROM) measurements are an important component of assessments for children with cerebral palsy. Most research has studied ROM stability using group data. Examination of longitudinal intraindividual measures may provide more clinically relevant information about measurement variability. OBJECTIVE: The aim of this study was to examine the stability of intraindividual longitudinal measurements of hip abduction (ABD), popliteal angle (POP), and ankle dorsiflexion (ADF) ROM measures of children with cerebral palsy. DESIGN: Secondary data analyses were performed. METHODS: The stability patterns of individual serial measurements of ABD, POP, and ADF from 85 children (mean age=3.8 years, SD=1.4) collected at baseline (T1), 3 months (T2), 6 months (T3), and 9 months (T4) were examined using T1 as the anchor and bandwidths of ±15 degrees (ABD and POP) and ±10 degrees (ADF) as acceptable variability. Frequencies of stability categories (0°-5°, 5.1°-10°, 10.1°-15°, and >15°) were calculated. Patterns of stability across the 4 time periods also were examined. Group means (T1-T4) were compared using repeated-measures analysis of variance. RESULTS: No significant differences in group means were found except for ABD. Stability patterns revealed that 43.3% to 69.5% of joint measurements were stable with T1 measurements across all 3 subsequent measurements. Stability category frequencies showed that many measurements (ABD=17%, POP=29.9%, and ADF=37.1%) went outside the variability bandwidths even though 39% or more of joint measurements had a change of 5 degrees or less over time. LIMITATIONS: Measurement error and true measurement variability cannot be disentangled. The results cannot be extrapolated to other joint ROMs. CONCLUSIONS: Individual ROM serial measurement exhibits more variability than group data. Range-of-motion data must be interpreted with caution clinically and efforts made to ensure standardization of data collection methods.
BACKGROUND: Serial joint range-of-motion (ROM) measurements are an important component of assessments for children with cerebral palsy. Most research has studied ROM stability using group data. Examination of longitudinal intraindividual measures may provide more clinically relevant information about measurement variability. OBJECTIVE: The aim of this study was to examine the stability of intraindividual longitudinal measurements of hip abduction (ABD), popliteal angle (POP), and ankle dorsiflexion (ADF) ROM measures of children with cerebral palsy. DESIGN: Secondary data analyses were performed. METHODS: The stability patterns of individual serial measurements of ABD, POP, and ADF from 85 children (mean age=3.8 years, SD=1.4) collected at baseline (T1), 3 months (T2), 6 months (T3), and 9 months (T4) were examined using T1 as the anchor and bandwidths of ±15 degrees (ABD and POP) and ±10 degrees (ADF) as acceptable variability. Frequencies of stability categories (0°-5°, 5.1°-10°, 10.1°-15°, and >15°) were calculated. Patterns of stability across the 4 time periods also were examined. Group means (T1-T4) were compared using repeated-measures analysis of variance. RESULTS: No significant differences in group means were found except for ABD. Stability patterns revealed that 43.3% to 69.5% of joint measurements were stable with T1 measurements across all 3 subsequent measurements. Stability category frequencies showed that many measurements (ABD=17%, POP=29.9%, and ADF=37.1%) went outside the variability bandwidths even though 39% or more of joint measurements had a change of 5 degrees or less over time. LIMITATIONS: Measurement error and true measurement variability cannot be disentangled. The results cannot be extrapolated to other joint ROMs. CONCLUSIONS: Individual ROM serial measurement exhibits more variability than group data. Range-of-motion data must be interpreted with caution clinically and efforts made to ensure standardization of data collection methods.