Video-based Performance Assessment following Stroke: Design of the Critical Elements Video Assessment Instrument

H.F. Machiel Van der Loos, PhD; Deborah E. Kenney, MS OTR; Ellie E. Buckley, MS PT; Eric L. Topp MS; Carol Ann Davidson, PT

Objectives: The goal of this project is to test the efficacy of personalized video-based training following stroke for wheelchair transfers and to develop a web-based clinical tool for assessing the change in a person’s functional ability over time. The goals of the first phase of the study are to 1) develop a set of criteria for the assessment of transfer safety and test the reliability of this instrument; 2) complete a generalized videotape to show proper transfer techniques for persons with hemiplegia; and 3) develop digital video authoring techniques and test this tool with one stroke subject before the full 45-subject, 2-year study.

Methods: A ten-criteria Critical Elements Video Analysis (CEVA) was developed in collaboration with the VA Palo Alto Comprehensive Rehabilitation Center and the Stanford Comprehensive Inpatient Rehabilitation Unit. The CEVA describes the necessary elements for a safe wheelchair transfer, including wheelchair preparation, positioning, body mechanics, weight-shift and stability through the transfer, and final position on the target surface.

While watching a digital video of a transfer on-line, the clinician scores each of the ten criteria as "good" (1 point) or "poor" (potentially unsafe, 0 points). An able-bodied subject performed six mock transfers demonstrating correct and incorrect techniques (Figure 1). CEVA wordings were adjusted after results from 4 raters were tallied.

In the full study, 45 subjects will be drawn from the population of sub-acute VA and Stanford stroke patients. The subjects will be randomly assigned to one of 3 groups, receiving take-home written material, a generalized videotape, or a personalized videotape instructing wheelchair transfers. After the 6-week follow-up visit, the videotape of a transfer is compared, by 2 blinded raters using the CEVA, with a taped transfer from 6 weeks earlier. Data logging instruments were developed to chronicle falls and pain-causing incidents in the home.

Incorrect (on chair) and correct (in lap) positioning of right hand.

Results: Goal 1: The 4 preliminary CEVA raters, all stroke clinicians, correctly scored 89.6% of all elements in the 6 transfers. Rater feedback revealed that wordings in 2 categories were ambiguous, and changes were made to reach consensus. Reliability results from an additional 20 raters are pending. Goal 2: A 20-minute generalized transfer instruction videotape was scripted, filmed and digitally edited to be used by the 2nd subject group in the full study. Goal 3: As a pilot test, a personalized video was made for one subject, who viewed it at home 3 times in 5 weeks. He reported 6 transfer-related falls or incidents in all, with 2 in the first week, and one in each subsequent week. CEVA scoring is pending his return for the follow-up visit.

Conclusions: The CEVA could be a first step to standardizing the evaluation of wheelchair transfers, and, if web-based, could offer a powerful new clinical assessment tool. The clinical relevance of this study is to improve patient function and safety while reducing the incidence of falls through the development of better and personalized training and assessment tools.

Funding: This study is being funded by VA Rehabilitation R&D Service Project E2169TC.