Revisiting Perry’s Gait Speed Categories.

Research Report
Population: Adult

George D Fulk, PT, PhD, Associate Professor, Clarkson University

Kari Dunning, PT, PhD, Associate Professor, University of Cincinnati DUNNINKK@UCMAIL.UC.EDU

Pierce Boyne, PT, DPT, NCS, Research Assistant Professor, University of Cincinnati boynepe@UCMAIL.UC.EDU

Michael O’Dell, MD, Professor, Department of Rehabilitation Medicine, New York Presbyterian Hospital/Weill Cornell Medical

Keywords: Participation, Stroke, Walking Activity, Gait Speed

Purpose/Hypothesis: The ability to fully participate by walking independently in one’s home and community is a primary goal for people with stroke. Clinicians and researchers commonly use gait speed to assess walking performance and as a surrogate for walking participation in the real world. Based off of original work by Hoffer et al, Perry et al developed gait speed categories to classify walking participation in people with stroke and these categories are commonly used in research and clinical practice. However, walking participation classification in this classic study was based on clinician opinion rather than objective measurement. Advances in sensor technology now allow us to directly measure home and community walking activity. Therefore, the purpose of this study was to determine an optimal gait speed cut off to distinguish between home and community ambulators based on real world walking activity. We hypothesized that the gait speed cut off would be greater than those proposed by Perry et al.

Subjects: 147 subjects >3 months post stroke with foot drop.

Methods: Cross sectional secondary data analysis from FASTEST trial. Comfortable gait speed (CGS) was measured over a 10m walk. Home and community walking participation was measured using a Stepwatch Activity Monitor (SAM) worn for 7 consecutive days. Based on work by Tudor-Locke the following criteria were used to define categories of walking participation: home ambulator <3,500 steps/day, limited community ambulator 3,500-7,499 steps/day, and unlimited community ambulator >=7,500 steps/day. Receiver Operator Characteristic (ROC) curve analysis was used to identify the most accurate CGS value to distinguish between home and community ambulators.

Results: Only 1 subject was an unlimited community ambulator (7,718 step/day) therefore was included as a limited community ambulator in the analyses, resulting in 115 home ambulators and 30 limited community ambulators. CGS accurately distinguished between home and limited community ambulators with an Area Under the Curve of 0.83 (95% CI 0.76-0.90). A CGS of 0.72 m/s was the most accurate cut off to distinguish between home and limited community walking participation, with a +LR of 4.0 and a –LR of 0.22.

Conclusion: CGS is able to distinguish between home and limited community walking participation when using real world stepping activity as the criterion. The cut off for distinguishing between home and limited community ambulators is greater than that proposed by Perry et al (0.72 vs. 0.40). A limitation of our study is that the SAM does not provide information on where the walking activity occurred or the purpose.

Clinical Relevance: In individuals with foot drop post stroke, CGS can be used to classify home (=0.72 m/s) walking participation.

Fulk, George D, PT, PhD; Dunning, Kari , PT, PhD; Boyne, Pierce , PT, DPT, NCS; O’Dell, Michael W, MD. Revisiting Perry’s Gait Speed Categories.. Poster Presentation. IV STEP Conference, American Physical Therapy Association, Columbus, OH, July 17, 2016. Online.