RESEARCH REPORT PLATFORM PRESENTATION

USE OF VIDEOTAPES IN DETERMINING OUTCOME AFTER INTENSIVE REHABILITATION FOR CHRONIC BACK PAIN. Barker K¹, Allen A¹, Frost H²; ¹Physiotherapy Research Unit, Nuffield Orthopaedic Centre NHS Trust, Oxford, UK. ²Department of Public Health, University of Warwick, Warwick, UK

PURPOSE: To investigate using videotaped recordings of patients performing functional tasks to evaluate change following a pain rehabilitation programme. RELEVANCE: There is a need to assess the benefit derived from treatment programmes. Currently a number of self-report questionnaires and timed performance tests are used. These are effective at documenting changes in the quantity of movement and in patient reported function. However, they do not reflect changes in the quality of movement. This study presents a method of recording this information. PARTICIPANTS: 120 patients with chronic back pain attending a rehabilitation programme. METHODS: Videotapes were recorded of patients performing two standard tasks, walking and sitting down on a chair, 3 weeks apart at the beginning and end of a rehabilitation programme. Two experienced physiotherapists scored the videos using observational grids. Walking was scored using the modified Rivermead Mobility Index; sit to stand based on deviation from an ideal normalised posture and observed pain behaviour using the system of Keefe (1). Changes in observed performance were compared with changes in other standard outcome measures, Oswestry Disability Index (ODI), Pain Self Efficacy (PSE) and Tampa Kinesiphobia Scale (TKS); and three timed performance measures: the shuttle walking test, stairs climbed and sit to stand in 60 seconds. ANALYSIS: Inter-rater reliability was calculated using ICC. Changes after treatment by students paired t- test. Associations between the observational and other outcome measures were analysed using Pearson’s Correlation coefficient. RESULTS: The inter-rater agreement for walking was I.C.C. 0.969 (C.I. 0.954–0.979); p < 0.001; sit to stand 0.785 (C.I. 0.686–0.853) p < 0.001 and pain behaviour 0.870 (CI 0.811-0.911), p < 0.001. There were significant improvements in the measures between baseline and end of the programme for observed gait, sit to stand and pain behaviour (p < 0.001). There were only weak associations between the observed function and performance in timed tests (observed gait and shuttle walking test p < 0.01; observed and timed sit to stand p < 0.05). There was no correlation between pain behaviour and performance in timed tests, nor with the ODI, PSE or TKS. CONCLUSIONS: The interrater reliability of the observational grids was high and they were sensitive to changes over time. The percentage improvement measured by observation was similar to those measured using timed tests; indicating that the grids have good criterion-related validity. The lack of any strong association between the observed and timed measures indicates that improved quality of movement did not correlate with improvements in the speed or quantity of physical performance. IMPLICATIONS: The use of observational grids to measure changes in the quality of movement can provide a useful alternative indicator of outcome in the field of low back pain rehabilitation. It may be more relevant to the goals of a rehabilitation programme than some of the timed measures that are currently used. KEYWORDS: Back pain, video, outcome. FUNDING ACKNOWLEDGEMENTS: The study was supported by a project grant from the OHSRC. CONTACT: karen.barker@noc.anglox.nhs.uk

ETHICS COMMITTEE: Oxford LREC