Navigating the night: evaluating and accessing wearable sleep trackers for clinical use

The global wearable sleep trackers market is experiencing a significant upsurge, reflecting a burgeoning interest in personalized sleep health monitoring and management. The market is projected to expand at a compound annual growth rate of 7.92% from USD 2.06 billion in 2023 and reach an estimated USD 3.02 billion by the end of 2028 [1]. Among this growth, smart and AI-based wearable sleep trackers, offering functions akin to polysomnographic (PSG) studies, have gained popularity in the digital health revolution [2–5]. Such devices, designed in various form-factors ranging from smartwatches, wristbands, and rings to chest bands, headbands, and armbands [6], are continuously innovated to create more accurate and personalized sleep-tracking experiences [7, 8]. Furthermore, integrating cutting-edge biometric sensors and AI-driven analytics into these devices enhances their capability to measure total sleep duration and cycles and parameters such as heart rate variability, sleep positions, and brainwave patterns. The interoperability of diverse sensors is key to customizing wearable sleep trackers. By enabling data sharing between sensors, these devices can adapt to each user’s unique sleep patterns and health conditions, providing a personalized monitoring experience. The increasing awareness of the critical role of sleep in overall health, exacerbated by the COVID-19 pandemic’s impact on mental health and sleep-related difficulties, has further fueled the demand for these devices [9, 10]. With the nonstop innovations of wearable sleep trackers and the rising demands for personalized sleep health trackers, the standardization and rigorous validation of these devices have become paramount [11].

Standardized evaluation and assessment are crucial in elevating wearable sleep trackers to medical-grade devices, yet they are still challenging [12]. The process begins with the refining of the algorithms, bio-sensing system designs, and firmware, aligning these procedures more closely with those in research protocols and clinical standards [13]. A significant part of this standardization involves benchmarking post-processed data from sleep trackers against PSG annotations. Three main groups of approaches have emerged to address these challenges. The first method group is based on Tryon’s approach [14] to evaluate the sensitivity, specificity, and accuracy of sleep staging compared to PSG standards. Additionally, statistical tests like paired sample t-tests, two-way repeated measure analyses of variance, and Bland–Altman plotting tests are employed to evaluate differences in sleep stages across various devices [15, 16]. Furthermore, methodologies assessing epoch-by-epoch correlations using tools such as confusion matrices and Pearson correlations [16], Cohen and Fleiss’s kappa [17], and localized mismatch index [3] offer a comprehensive evaluation of these trackers. However, constraints in accessing raw data and the inconsistencies in classifying sleep stages among sleep-tracker devices add complexities to the validation process, necessitating diverse and in-depth comparative methods.

In advancing the utility of wearable sleep trackers, it is essential to transcend ordinary technical assessments and analyze the complexities of individual sleep patterns and health conditions [18]. The heterogeneity of sleep endotypes—the distinct behavioral or physiological patterns in sleep—presents a significant challenge [5, 19]. These variations are often intertwined with diverse symptoms and comorbidities, encompassing conditions like sleep-related breathing disorders, neurological disorders, mental health issues, and chronic pain conditions, all of which can significantly influence the precision of sleep tracking [20]. Moreover, the effectiveness of these devices is also influenced by individual responses to different treatment modalities. For instance, a device that accurately tracks sleep patterns in a healthy individual might not perform as well in someone with a chronic sleep disorder under treatments [21]. This necessitates customizing devices to accommodate device users with varying characteristics. Additionally, user interference factors [22], such as how a device is worn or interacts with other electronic devices, also play a critical role. These factors can skew the data, leading to sleep tracking and analysis inaccuracies. The ultimate goal is to refine system performance and sensing system design to deploy these wearable devices for specific medical applications. This involves improving the wearable devices’ accuracy and reliability and ensuring they are user-friendly and adaptable to clinical scenarios.

The study by Reifman et al. [23], featured in this issue of Sleep, represents a quantitative method to evaluate the accuracy of sleep-tracker devices in fatigue management. The approach entailed a comparison of sleep-measurement errors across 18 different sleep tracker devices against PSG data. It utilized these variances as the input to the unified model of performance [24]—a quantitative model to predict the alertness impairment using sleep schedule history. By simulating the alertness fluctuations over 30 consecutive days, based on sleep schedules of 5, 8, and 9 hours/night and irregular sleep schedule, the study found that nearly 80% of the time, the predicted differences in alertness were under their within-participant variability of 30 milliseconds threshold. The method provides the means to determine the operationally acceptable sleep tracker device for fatigue management given its sleep-measurement error characteristic. This work presents a significant step to establishing a method to quantitatively assess sleep trackers for predicting alertness impairment of a fatigue-management system. The proposed framework facilitates the generation of new ways to assess the validity of sleep-tracker devices for other clinical or operational applications.

Building on the study by Reifman et al. to evaluate wearable sleep trackers for fatigue management, the integration of AI, mainly through edge computing, in wearable devices is a promising development for sleep health monitoring and prediction [25]. This includes utilizing generative AI components and edge computing technologies, which go beyond traditional PSG features and sensors for sleep tracking. These AI-based algorithms leverage diverse features, including movement patterns, physiological signals, and environmental factors to analyze and score sleep, offering a more holistic understanding of sleep conditions and leading to new standards for evaluating wearable devices. Moreover, the edge computing technology processes and analyzes data directly on the device [26], enabling real-time insights into various clinical conditions of cardiac, neurodegenerative, and respiratory diseases [27, 28]. Utilizing such edge computing systems for sleep health monitoring results in faster, more efficient data processing, decreasing latency, and lessening dependence on cloud connectivity. This fusion of advanced sleep tracking with AI technologies necessitates rigorous quality control in AI algorithms and sensor system in wearable sleep trackers to ensure their accuracy and reliability in healthcare applications. The evolution of these technologies, emphasizing data integrity and algorithmic accuracy, is vigorous for their successful deployment in healthcare, especially in fatigue management and other clinical applications.

In summary, wearable sleep trackers with evolving digital health technologies stand at the forefront of a revolution in personalized healthcare. The rapid growth of this market toward clinical applications, its expansion into AI integration, and the advent of edge computing herald a new era in patient-centric medical monitoring and diagnostics. The pivotal study by Reifman et al. underscores this progress, showcasing how these devices can effectively be employed in critical areas like fatigue management. As technology progresses, its application extends beyond sleep to potentially transformative roles in managing cardiac, neurodegenerative, and respiratory conditions. This expansion necessitates a dual focus on technological sophistication and inclusivity in design and testing, particularly for underserved populations. The future of wearable sleep trackers, thus, lies not only in their technical prowess but also in their capacity to adapt to and address a broad spectrum of health needs, paving the way for a more accessible, responsive, and personalized healthcare system.

Disclosure Statement

The authors have no financial interests to disclose. The authors have no potential conflicts of interest to disclose.