P107: Use of activity trackers to count steps of older emergency department patients: a feasibility and validity study

Abstract

Introduction: Mobility is an evidence-based non-pharmacologic strategy shown to reduce delirium and functional decline among older patients in the acute care setting. Activity trackers have been used in previous studies to objectively measure mobility in older hospitalized patients. This study aims to compare the feasibility and validate the accuracy of three accelerometer-based activity trackers (Fitbit Zip, Fitbit Charge HR and StepWatch). This is the first step in a program of research to objectively measure as a potential marker of delirium risk. Methods: This is a prospective study of patients 65 years of age and older during their ED visit. We excluded those with critical illness, unable to communicate or provide consent; and any ambulatory impediments. Consenting participants wore the trackers for up to 8 hour, and completed a 6-meter walk test while a research assistant manually counted their steps. Our primary feasibility measure was the proportion of eligible patient for which we were able to recover the tracker and recorded their steps. The primary validation endpoint was the concordance between steps recorded by the tracker compared to a gold standard manual step count over a fixed distance. Sample size was based on the desired precision of the final estimate of feasibility. Intraclass correlation coefficient (ICC) was calculated to assess agreements between devices and manual count. We will report proportions with exact binomial 95% confidence intervals (CI) for feasibility and validity endpoints. Results: 41 participants were enrolled in this study. Mean age was 74.6 years (+/- 5.76) and 59% were females. The total subjects that wore the Fitbit Zip, Fitbit Charge HR and StepWatch during study participation was, 40/41 (97.5%, CI 0.87–0.99), 33/34 (97%, CI 0.84–0.99) and 31/32 (96.8%, CI 0.83–0.99), respectively. Total subjects with completed data extracted from the Fitbit Zip, Fitbit Charge HR and StepWatch was, 38/41 subjects (92.6%, CI 0.80–0.98), 34 (100%, CI 0.89–1.00), and 32 (100%, CI 0.89–1.00), respectively. All devices were recovered after use (100%, 95%CI 0.91–100). Conclusion: Our results suggest: 1) the use of gait-tracking devices in the ED is feasible, 2) consumer and research-grade devices showed good validity against the gold standard, and 3) the use of small, inexpensive, consumer-grade trackers to objectively measure mobility of older adults in the ED.