Smartwatches are defined as wrist-worn, sensor-equipped devices with the potential to significantly impact healthcare by enabling continuous health monitoring in daily life. When integrated with smartphone data and patient health records, they can offer a comprehensive view of an individual’s health.
Whilst healthcare professionals have historically been constrained by the limitations of in-person visits and self-reported data, there is a growing body of evidence to suggest that they may benefit from the enhanced accuracy and continuous nature of data provided by smartwatches. While initially popular among recreational users, their role in clinical decision-making is growing and merits thorough evaluation, as research on their medical applications continues to evolve.
Smartwatches in the Clinic? New Evidence Maps Their Medical Potential
The increasing popularity of smartwatches can be attributed to their functionality in monitoring key health metrics, including stress, movement disorders, sleep, and even heart disease. Five recent studies have investigated the potential of using smartwatches for stress monitoring (1, 2, 3, 4, 5). These studies focused on a variety of physiological parameters, including heart rate variability , cortisol levels, skin conductance, body temperature and blood volume. Machine learning techniques have been applied to detect stress, with some studies achieving optimal accuracy using specific stress recognition programs or algorithms. The ability of smartwatches to assess stress levels in different contexts, including of drivers, passengers, employees and patients in online therapy sessions, is a notable advancement. Stress detection is even more effective if data from wearable devices is combined with data from smartphones and contextual information.
Smartwatches are used to monitor movement disorders in a variety of medical conditions, including multiple long-term conditions (multimorbidity) (6), autism (7, 8), fall detection in older adults (9), and monitoring aggressive behaviour in people with dementia or mental health disorders (10).
In the context of autism, smartwatches equipped with accelerometers are highly effective and accurate in detecting stereotypical behaviours. In terms of fall detection, the algorithms used in smartwatches have shown encouraging levels of sensitivity and specificity. However, it is important to recognise that performance can vary depending on the specific algorithm and smartwatch model used. Furthermore, integrating machine learning with smartwatch data is highly accurate in identifying aggressive movements. Smartwatches also show similar results to seismological sensors in measuring the amplitude and frequency of hand tremors in movement disorders such as Parkinson’s disease.
Sleep tracking was the focus of three different studies. In the first, a smartwatch-based system, ApneaDetector, was developed to detect sleep apnoea using embedded sensors (11). While the system accurately classified normal and sleep apnoea events, it encountered challenges with more specific categorisations.
Another study found that smartwatches that use light-based technology to detect blood volume changes are more effective than in-lab polysomnography or home sleep apnoea testing devices for screening suspected cases of obstructive sleep apnoea (12)
Studies have shown that wearing smartwatches that monitor blood pressure can improve blood pressure and other health parameters, increase awareness of high blood pressure, and help reduce risk factors (14). However, some commercial smartwatches overestimate low blood pressures and underestimate high blood pressures, suggesting they are not yet ready for clinical use (15). While seen as a potential strategy for behaviour modifications, there is currently insufficient evidence to support their use in improving hypertension management.
Furthermore, a study by Mehrabadi and colleagues compared the accuracy of the Oura ring, a smart ring used to track sleep and physical activity, to medically approved devices that measure activity. They found that the Oura ring is more accurate for detecting sleep characteristics than the Samsung Gear Sport smartwatch (13).
Given that they can track heart rate accurately by detecting blood volume changes in microvascular tissue beneath the skin, smartwatches show great promise in early detection of heart disease. Studies have shown that doctors can accurately detect atrial fibrillation, a condition that causes an irregular heartbeat, using smartwatches, with a high level of sensitivity (99%) and specificity (83%) (16, 17).
Smartwatches in Scrubs: Technology, Trust, and the Future of Care
Despite these potential benefits, there is an ongoing debate about the reliability of smartwatch diagnoses within healthcare systems. The accuracy and reliability of algorithms, as well as the accuracy of smartwatches in detecting health changes, are critical. It is important to note that relying solely on smartwatch data for making significant healthcare decisions can be challenging. Smartwatches often focus on aggregating biomedical data without a holistic view of the patient. If doctors over-relied on smartwatch data, there would be fewer face-to-face visits, which could violate ethical principles such as their duty to not harm patients. Most smartwatches are designed for convenience rather than medical diagnosis, leading to concerns about their trustworthiness, particularly due to opaque algorithms and variable results in diagnosing symptoms. For these reasons, further evaluation of the safety and reliability of smartwatches is needed.
Based on these observations, there is a need for the development of medical-grade watches and the creation of AI hospital assistants. These AI assistants could help with patient monitoring, scheduling, medication management, patient education and answering common questions, allowing healthcare providers to focus on more complex tasks. Medical institutions also need to train their students to work effectively with “e-patients”, who are increasingly using technologies such as smartwatches for self-care.
Leading healthcare technology companies should prioritise the development of smartwatches equipped with reliable and efficient monitoring systems. Continued research is crucial to fully understand and harness the potential of smartwatches in healthcare and to advance the creation of more sophisticated and effective AI-powered clinical assistants.
- Laura Avogaro from FRESCI
- No Comments
Bibliography
-
- Fauzi, M. A., Yang, B., & Blobel, B. (2022). Comparative Analysis between Individual, Centralized, and Federated Learning for Smartwatch Based Stress Detection. Journal of personalized medicine, 12(10), 1584.
- Fauzi, M. A., & Yang, B. (2021). Continuous Stress Detection of Hospital Staff Using Smartwatch Sensors and Classifier Ensemble. Studies in health technology and informatics, 285, 245–250.
- Vila, G., Godin, C., Sakri, O., Labyt, E., Vidal, A., Charbonnier, S., Ollander, S., & Campagne, A. (2019). Real-time monitoring of passenger’s psychological stress. Future Internet, 11(5), 102.
- Han, H. J., Labbaf, S., Borelli, J. L., Dutt, N., & Rahmani, A. M. (2020). Objective stress monitoring based on wearable sensors in everyday settings. Journal of medical engineering & technology, 44(4), 177–189.
- de Arriba Pérez, F., Santos-Gago, J. M., Caeiro-Rodríguez, M., & Fernández Iglesias, M. J. (2018). Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students. Journal of visualized experiments : JoVE, (136), 57590.
- Ali, S. M., Selby, D. A., Khalid, K., Dempsey, K., Mackey, E., Small, N., van der Veer, S. N., Mcmillan, B., Bower, P., Brown, B., McBeth, J., & Dixon, W. G. (2021). Engagement with consumer smartwatches for tracking symptoms of individuals living with multiple long-term conditions (multimorbidity): A longitudinal observational study. Journal of multimorbidity and comorbidity, 11, 26335565211062791.
- Amiri, A. M., Peltier, N., Goldberg, C., Sun, Y., Nathan, A., Hiremath, S. V., & Mankodiya, K. (2017). WearSense: Detecting Autism Stereotypic Behaviors through Smartwatches. Healthcare (Basel, Switzerland), 5(1), 11.
- Torrado, J. C., Gomez, J., & Montoro, G. (2017). Emotional Self-Regulation of Individuals with Autism Spectrum Disorders: Smartwatches for Monitoring and Interaction. Sensors (Basel, Switzerland), 17(6), 1359.
- Antos, S. A., Danilovich, M. K., Eisenstein, A. R., Gordon, K. E., & Kording, K. P. (2019). Smartwatches Can Detect Walker and Cane Use in Older Adults. Innovation in aging, 3(1), igz008.
- Tchuente, F., Baddour, N., & Lemaire, E. D. (2020). Classification of Aggressive Movements Using Smartwatches. Sensors (Basel, Switzerland), 20(21), 6377.
- Chen, X., Xiao, Y., Tang, Y., Fernandez-Mendoza, J., & Cao, G. (2021). ApneaDetector: Detecting sleep apnea with smartwatches. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 5(2), Article 59, 1–22.
- Chen, Y., Wang, W., Guo, Y., Zhang, H., Chen, Y., & Xie, L. (2021). A single-center validation of the accuracy of a photoplethysmography-based smartwatch for screening obstructive sleep apnea. Nature and Science of Sleep, 13, 1533–1544.
- Asgari Mehrabadi, M., Azimi, I., Sarhaddi, F., Axelin, A., Niela-Vilén, H., Myllyntausta, S., Stenholm, S., Dutt, N., Liljeberg, P., & Rahmani, A. M. (2020). Sleep Tracking of a Commercially Available Smart Ring and Smartwatch Against Medical-Grade Actigraphy in Everyday Settings: Instrument Validation Study. JMIR mHealth and uHealth, 8(10), e20465.
- Yen, H. Y., & Huang, W. H. (2022). The efficacy of commercial smartwatches with a blood pressure-monitoring feature: A pilot randomized controlled trial. Journal of nursing scholarship : an official publication of Sigma Theta Tau International Honor Society of Nursing, 54(3), 324–331.
- Falter, M., Scherrenberg, M., Driesen, K., Pieters, Z., Kaihara, T., Xu, L., Caiani, E. G., Castiglioni, P., Faini, A., Parati, G., & Dendale, P. (2022). Smartwatch-Based Blood Pressure Measurement Demonstrates Insufficient Accuracy. Frontiers in cardiovascular medicine, 9, 958212.
- Bumgarner, J. M., Lambert, C. T., Hussein, A. A., Cantillon, D. J., Baranowski, B., Wolski, K., Lindsay, B. D., Wazni, O. M., & Tarakji, K. G. (2018). Smartwatch Algorithm for Automated Detection of Atrial Fibrillation. Journal of the American College of Cardiology, 71(21), 2381–2388.
- Liao, M. T., Yu, C. C., Lin, L. Y., Pan, K. H., Tsai, T. H., Wu, Y. C., & Liu, Y. B. (2022). Impact of recording length and other arrhythmias on atrial fibrillation detection from wrist photoplethysmogram using smartwatches. Scientific reports, 12(1), 5364.