Clinical significance testing in rehabilitation research: what, why, and how?

Abstract

Rehabilitation research often uses statistical significance to demonstrate the effectiveness of an intervention. This practice may produce misleading results as statistical significance does not imply clinical significance. An intervention that impacts the well-being and function of clients is clinically significant. How is clinical significance evaluated, and how can it be incorporated into rehabilitation research? This paper addresses these questions by using a combination of distribution- and anchor-based approaches to determine the clinical impact of treadmill training on comfortable walking speed in individuals with incomplete spinal cord injuries. Effect size, standard error of measurement, and the Walking Index for Spinal Cord Injury are used to estimate clinically significant change in walking speed. The results from these three methods are combined into a robust and interpretable conclusion regarding clinical significance. Such conclusions will facilitate knowledge transfer between research and practice.

Keywords:

• CLINICAL SIGNIFICANCE
• SPINAL CORD INJURY
• WALKING SPEED