Zero-interaction technology and its potential for long-term monitoring of older adults

Population aging poses a challenge to any healthcare system due to increased disease burden and related disability. This translates to increased expenditure on the healthcare system and reduced quality of life for individuals. One approach to tackle this challenge is transforming the healthcare system from reactive to predictive, preventive, personalized, and participatory (P4) medicine. This is achieved by approaches like multi-omics profiling and remote health monitoring. The latter utilizes smartphones, wearables, the internet of things, and sensing units to objectively and continuously measure relevant health information daily.

Many examples demonstrate how the use of digital technology could influence health. One study used sensor-derived gait speed and physical activity variations to detect mild cognitive impairment (MCI). In another study, loneliness was assessed using connected sensing technologies.

Much of the available literature involves short-term investigation of digital technologies, like smartphones, smartwatches, and activity trackers. The technologies might not be ideal for long-term monitoring of the old population due to the stigma associated with wearable devices, lack of access to such technologies, older adults’ tendency to be more cautious of novel technologies, compliance issues, and potential memory problems when wearing and maintaining these devices. Therefore, it is best to utilize zero-interaction tools that could help with long-term monitoring. The latter has proved its success, like using infrared technology to capture motion, contact door sensors, electronic pill boxes, and pressure sensors attached to mattresses.

Schütz et al. aimed to evaluate a system's potential for long-term health monitoring in older adults (pooled age of study participants was 87 ±7). The researchers introduced 1268 digital measures (daily activity, behavior, and physiology monitoring) obtainable from sensing technologies backed by extensive, real-world evidence. The included measures were based on cost-effective, contactless sensors that involve no interaction. This makes such sensors suitable to be deployed for long-term monitoring in community-dwelling older individuals. The researchers investigated five datasets: fall risk, frailty, late-life depression, and MCI.

Results revealed that using machine learning-based tools for digital clinical outcome assessments improved predictive capacity across multiple measures. Fall-risk assessments using digital exhaust (multiple measures) produced similar results to wearable devices. However, the results are more robust due to utilizing several parameters, and the technique is more suitable for long-term monitoring. Frailty assessment was based on physical activity and gait, and the assessment results were acceptable and comparable to wearable sensors. In addition, new potential digital biomarkers, like room-transition count, sleep duration, and fridge usage, were found to offer value in assessing frailty. Encouraging results were found with MCI measures and late-life depression as well.

The authors concluded that a contactless, system-based comprehensive approach to remote health monitoring offers potential benefits in long-term clinical care and research. Moreover, utilizing machine learning that assesses multiple parameters represents a potential alternative to commonly employed wearables.

Source: Schütz, N., Knobel, S.E.J., Botros, A. et al. A systems approach towards remote health-monitoring in older adults: Introducing a zero-interaction digital exhaust. npj Digit. Med. 5, 116 (2022). https://doi.org/10.1038/s41746-022-00657-y

‍