Airbags to prevent hip injuries in the aging stroke population - PROJECT SUMMARY/ABSTRACT Over 95% of the 300,000 hip fractures that occur annually in the US are caused by falls, and 27% of older adults with hip fracture die due to complications within a year. Moreover, up to 70% of older adults with a history of stroke experience multiple falls every year, resulting in significant injuries and a high mortality rate. In the stroke population, over half of all fall-related bone fractures occur at the hip. Research aiming to address this issue has largely focused on fall prevention strategies, but due to different fall patterns and various risk factors that result from stroke, these strategies have little to no effect for older adults with a stroke, who continue to fall repeatedly and are often left with severe long-term disabilities. This project offers an alternative: stroke-specific fall impact mitigation using wearable airbags in the community. Wearable airbags depend on pre-impact fall detection algorithms to detect falls as they occur to deploy airbags and mitigate the impact, but algorithms developed for non-neurologically impaired individuals (called generalized algorithms) are less accurate for people with a history of stroke, prompting our prior development of a stroke-specific fall detection algorithm. The objective of this proposal is to evaluate the effectiveness of a wearable hip airbag embedded with our stroke-specific pre-impact fall detection algorithm to accurately deploy airbags during falls, reduce hip injuries, decrease fear of falling, and increase community ambulation in older adults with stroke. In Aim 1, 44 older adults aged 65-90 with a history of stroke and falls will be recruited in a community-based crossover trial. Aim 1 will be divided into two sub-aims: Aim 1a will compare the performance rates of the stroke-specific and generalized fall detection algorithms in detecting real-world falls in the community and deploying airbags, while Aim 1b will compare the rates of injury, fear of falling and community engagement during the use of the airbag device with each embedded fall detection algorithm. Aim 2 will develop a clinical tool that uses the baseline clinical data to identify and predict responders to each algorithm, helping us to predict which algorithm provides the most benefit for specific individuals with stroke. The primary outcomes will be the F1 score (a metric that accounts for precision and sensitivity) of the stroke-specific algorithm compared to the generalized algorithm (Aim 1), and the accuracy of the clinical tool for predicting responders (Aim 2). Secondary outcomes include the rates of fall injury, fear of falling, community ambulation, impression of the device, and quality of life. This work will (i) determine the performance rates of the stroke-specific algorithm compared to the generalized algorithm for airbag deployment during a fall in older adults with stroke history in the community, (ii) evaluate the benefits of a stroke-specific airbag system in reducing hip injuries and fear of falling, and increasing community engagement, and (iii) develop a clinical tool to utilize a single visit’s worth of clinical data to predict responsiveness to the stroke-specific or generalized fall detection algorithms and guide future prescription. These results will lay the groundwork for condition-specific fall protection for other high-risk populations, and systems to protect against other injuries, like head impacts.
