Section Information: Abstracts of Poster and Platform Presentations for the 2004 Combined Sections Meeting: Poster Presentations
RANGE OF MOTION MEASUREMENT IN CEREBRAL PALSY: A COMPARISON OF ASSESSMENT METHOD.
A.M. Glanzman, Physical Therapy, Children’s Seashore House of The Children’s Hospital of Philadelphia, Philadelphia, PA, A.E. Swenson, Physical Therapy, Drexel University, Philadelphia, PA, H. Kim, Physical Medicine and Rehabilitation, Children’s Seashore House of The Children’s Hospital of Philadelphia, Philadelphia, PA.
PURPOSE/HYPOTHESIS: This study was designed to compare intrarater reliability of passive range of motion (PROM) in children with cerebral palsy.
NUMBER OF SUBJECTS: 50 children with cerebral palsy were assessed.
MATERIALS/METHODS: PROM was taken by goniometric measurement by one rater both with and without a holder. 37 of those children were assessed by visual estimation (VE). Measurements were taken of ankle dorsiflexion, popliteal angle, hip abduction, hip extension by Thomas Test (TT), and Staheli’s Test (ST) in a blinded fashion to assess intrarater reliability of the three forms of measurement. VE data was compared to goniometric measurement since blind intrarater visual measurement was not possible.
RESULTS: Intraclass correlations (3,1) were performed to establish reliability for each measure. To compare the ICCs, a Z statistic for the test for equal correlations was used. All ICCs for one-person, two-person, and VE were above 0.90 and in the excellent range. ICCs for one-person ranged from 0.9439 to 0.9956; and for two-persons from 0.9671 to 0.9971. When VE was compared with two-person measurement, the ICC ranged from 0.9137 to 0.9988. Comparison of one-person vs two-person goniometry showed no significant difference between the groups, with the exception of ST (P < 0.05) which was more reliable with a holder. When TT was compared to ST, TT was more reliable with one-person (P < 0.05), but with two there was no difference between the methods. Comparison of two-person goniometry and VE showed that two-person measurements were significantly more reliable than VE in all measures except popliteal angle, where there was no statistically significant difference, and abduction, where VE was more reliable (ICC = 0.9988 vs ICC = 0.9896). Similarly, when compared with VE, one-person goniometry was more reliable for dorsiflexion and hip extension by TT, but inferior for abduction. There was no statistically significant difference between VE and one-person goniometry for ST or popliteal angle.
CONCLUSIONS: Goniometry provides a significant advantage in terms of reliability over VE in children with cerebral palsy. However, the addition of a second assessor to provide stabilization of the patient during measurement did not improve measurement reliability, with the exception of ST for hip extension where a holder improved reliability.
CLINICAL RELEVANCE: Reliability is critical to assessment of change. Understanding the impact of technique in goniometry improves reliability.