In-home sensing technologies hold enormous potential for early detection of health changes that can dramatically affect the experiences of aging: enabling functional independence, improving self-management of chronic or acute conditions, and improving quality of life. Chronic diseases especially affect older adults. Problems in chronic disease management are often the cause of losing independence for aging Americans. In 2012, 1 in 2 American adults (117 million) had at least one chronic condition, and 26% of the population had multiple chronic conditions, accounting for 84% of US health care costs. Early illness recognition and early treatment is key to improving health status with rapid recovery after an exacerbation of a chronic illness or acute illness, and also key to reducing morbidity and mortality in older adults and controlling health care costs. In previous work, the team developed a health alert system that captures and analyzes data from sensors embedded in the home. Sensor data are captured passively and continuously in the home. In a pilot NIH R21 study, significant differences in health outcomes were shown with health alerts from motion and bed sensor data, based on bed restlessness and low, normal, and high pulse and respiration rates. The system actually detected changes in chronic diseases or acute illnesses on average 10 days to 2 weeks before usual assessment methods or self-reports of illness. For this project, the team will expand from the clinician-focused system to a consumer-focused system by incorporating more finely grained sensing (gait and quantitative pulse and respiration), with new improved algorithms that integrate individual health status and medication use, and track trajectories of health changes, for more sensitive, and more personalized health alerts with fewer false alarms. A recently developed bed sensor will be incorporated to passively capture quantitative pulse, respiration, and restlessness while the subject is resting. Gait parameters (e.g., in-home walking speed, stride time and stride length) will also be captured using depth images that show shadowy silhouettes. In addition, the team will solicit the consumer perspective on customized health alerts and a user interface for displaying sensor and alert information. The views of seniors and their family members will be used to inform the development of the new customized alert algorithms and drive the development of a consumer-focused interface that will provide empowering tools for self-management of chronic illnesses. In addition, the use of commercially available wrist-worn sensors will be explored for the purpose of recognizing health changes. The study will include a retrospective analysis of sensor data collected in 13 senior housing sites in Missouri. New participants will be recruited in 5 senior housing sites in Columbia, MO to investigate the consumer perspective. The important process of engaging consumers in this work is the next step in translating these systems into clinical practice for self-managing chronic health conditions, supporting seniors living independently.
We will build on our current work using intelligent, in-home sensor systems with automated health alerts to investigate new health alert algorithms that are more sensitive and more customized to the individual. These will be tested with data from 13 senior housing sites in Missouri using clinician feedback and actual health trajectories for evaluation. We will also recruit subjects in independent living housing to solicit the views of consumers on wearable sensors, health alerts, and user interfaces that allow seniors and their family members to view the health alert information in a way that empowers them to better self-manage their own health.
