Objective Assessment of Sleep Patterns among Night-Shift Workers: A Scoping Review

In this scoping review of the literature, we identified the types and the parameters of objective measurements to assess sleep patterns among night-shift workers. We conducted a literature search using electronic databases for studies published from 1991 to 2020 and charted and summarized key information. We included 32 studies in the review. Polysomnography was used in 6 studies and wearable sleep detection devices were utilized in 26 studies. The duration of sleep assessment using the wearable devices ranged from 1 day to ≥4 weeks, and more than half of the studies collected data for >2 weeks. The majority of the studies used subjective questionnaires, such as the Karolinska Sleepiness Scale, Epworth Sleepiness Scale, and Pittsburgh Sleep Quality Index, in addition to objective sleep measurements. Total sleep time was the most common parameter, followed by sleep efficiency, sleep onset latency, and time or frequency of being awake. As the utilization of wearable devices to assess the sleep patterns of night-shift workers is expected to increase, further evaluation of device accuracy and precision, optimal data collection period, and key parameters is warranted.


Introduction
Shift work has been recognized as having a negative influence on health and safety [1]. Working on the night shift, usually from 11 p.m. to 7 a.m. or 12 midnight to 8 a.m. [2], typically causes changes in the sleep-wake patterns, which can lead to transient periods of misalignment between circadian rhythms that are detrimental to health [3]. This includes effects on sleep patterns. Furthermore, sleeping at times that are not optimal may also affect health, with optimal timings being determined by circadian rhythms. With repeated night shifts, sleep loss accumulation and chronic sleep deprivation can occur [4]. The literature suggests that working on the night shift is associated with increased risks of health problems, such as cardiovascular and metabolic diseases, depression, reproductive problems, and cancer [2].
The assessment of sleep patterns among night-shift workers is an essential step in identifying individuals with sleep disturbance and conducting interventions for them [4]. The tools frequently used to measure sleep patterns among night-shift workers are selfreport questionnaires (e.g., Pittsburg Sleep Quality Index); however, these can have some drawbacks in assessing night-shift workers' sleep patterns that might change day by day according to their shift work schedule [3]. Moreover, the data obtained by self-report questionnaires are often confounded with the reporter's emotional and physical status, which can lead to the underestimation or overestimation of sleep patterns [5]. Therefore, an objective assessment of sleep patterns is important for assessing the sleep disturbances that stem from night-shift work.
Recent advancements in sleep detection technologies have led to the development of measurement tools that involve the use of sensing or signal equipment and recording or data processing [4]. Polysomnography (PSG) has been the "gold standard" for sleep measurement to date, but many wearable sleep detection devices can assess sleep parameters derived from the quantifications of physical activity, heart rate, and electrodermal activity [6].
Given the benefits of collecting objective sleep data, it is advisable that such tools are utilized to assess the sleep patterns of night-shift workers. However, little research exists on this topic. Therefore, in this study, we conducted a scoping review of the literature on the types and the parameters of the objective measurements used to assess sleep patterns among night-shift workers. We conducted this review using the framework of Arksey and O'Malley [7]. This study aimed to review several types of objective assessment tools for measuring the sleep patterns of night-shift workers.

Study Selection
The inclusion criteria were studies involving night-shift workers, demonstrating quantitative sleep data obtained using devices, and having been published in peer-reviewed English-or Korean-language journals. The exclusion criteria were qualitative studies, secondary data analyses, studies not including sleep parameters, and unpublished theses or dissertations. We selected the studies using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [8] (Figure 1). All investigators searched and reviewed the literature independently and then reached an agreement through discussions about the final list of studies.

Charting the Data and Summarizing the Results
We charted key information items obtained from the studies, including purpose, methodology, sleep measurement tools, and results [6]. We then summarized the types and the parameters of the sleep measurements (Tables 1 and 2).   Nurses working night shifts demonstrated lower SE and higher levels of fatigue and turnover intention than non-shift-working nurses.  To assess the relationships between the sleepiness and incidence of adverse driving events in nurses commuting to and from night and rotating shifts.
Cross-sectional study Health care workers (N =27)

Sleep Vigilance Performance
For the shift-working group, self-reported sleepiness, drowsiness, and driving events were higher during commutes following night shifts compared to commutes before night shifts.

Sasseville et al., 2010; Canada [37] †
To investigate the possibility of adaptation in shift workers who are exposed to blue-green light at night, combined with blue blockers during the day.

Discussion
In this scoping review, most of the study participants were health care providers working the night shift. Given that shift work is conducted in diverse types of occupations, the investigations about objective sleep quality associated with night-shift work must be expanded to include other types of jobs. Of the 32 studies in this review, 26 utilized wearable sleep detection devices and 6 used PSG to assess sleep patterns. This review demonstrated that, with the advancement of sleep detection technology using sensors, wearable devices have been gaining popularity. These devices have the advantage of being simple, inexpensive, and easy to use in daily life in comparison to the high cost and inconvenience of PSG, which has been the "gold standard" for sleep measurement [41]. The objective assessment of sleep quality using wearable devices would enhance sleep disturbance detection among night-shift workers. By this, they may gain benefits such as the management of sleep disturbances that might be associated with job performance, accidents, and increased risk of health problems [4].
Most of the wearable devices used to measure sleep patterns were worn on the wrist; Actigraphy and Actiwatch were the most common. In some studies, the Fitbit series was used on the nondominant wrist to check sleep patterns. Actigraphy and Actiwatch, which include a microelectromechanical systems accelerometer, have shown high consistency in the sleep parameters when compared to PSG [42], indicating high reliability and validity for use in clinical practice. One study has reported that wearable devices based on light sensors, such as the Fitbit HR, have a higher sensitivity than Actigraphy for measuring sleep parameters such as TST, SE, and SOL [41]. New types of wearable devices, such as glasses, watches, and bands, offer a wide range of user choices, and also provide real-time data by synchronizing the device with a mobile phone or personal computer [43]. Owing to the development of information technology, various forms of wearable devices are being developed and their use is expected to increase in the future because they are inexpensive, convenient to access, and can be worn for a long period regardless of time or place.
However, in this review, only two studies reported the sensitivity, accuracy, or specificity of the wearable devices for sleep assessment. Most devices are classified as "wellness" products, except for Actigraphy, and have not yet obtained approval from the United States Food and Drug Administration [44]. The findings suggest the need for standardization through various repeated studies and clinical use, so that the sleep parameters measured by wearable devices can be validated [43]. Additionally, the sensitivity, specificity, or accuracy of wearable devices have only been tested in participants who were healthy or had insomnia or other sleep disorders [41,45,46]; only a few studies have demonstrated the sensitivity, specificity, or accuracy of these devices in night-shift workers. Considering the high prevalence of sleep disorders among these workers [3], the investigation of the validity and usefulness of wearable sleep detection devices is needed.
A previous study reported that the main clinical characteristics of people with sleep disturbance from night-shift work are shorter sleep duration and sleepiness [4]. In this review, we categorized 11 objective sleep parameters, including TST, SE, SOL, and WASO. Among them, TST was the most common parameter. TST, as measured by a wearable device, was minimally different from the values assessed using PSG in adults with insomnia. Since sleep duration was demonstrated to be a mediator of metabolic syndrome among female hospital employees working in an alternating day-and night-shift work schedule [47], TST can be used as one of the reliable indicators of sleep quality in night-shift workers [42,45]. Moreover, a previous study suggested the inclusion of sleep duration, sleep and wake times, time required to fall asleep, and number and duration of awakenings during the sleep episode to assess sleep quality in shift workers. It is believed that these sleep quality parameters could be efficiently assessed using wearable devices [4].
In addition to the objective measures of sleep, self-report questionnaires are often used to assess the subjective sleep quality or degree of sleepiness that can occur in a specific situation. Most of the studies used KSS, which evaluates sleepiness at a specific time of the day [48], and ESS, which evaluates the degree of sleepiness in a specific situation [49]. Sleepiness is a common complaint among shift workers and can influence job performance or cause accidents [4]; thus, KSS and ESS are useful measures for assessing the consequences of shift work. Sleep disturbance among shift workers is characterized by insomnia or excessive sleepiness that is associated with a recurring work schedule [4]. Thus, the combined use of self-report questionnaires regarding subjective sleep quality or sleepiness in addition to objective sleep measurement would be desirable to assess the sleep patterns of shift workers.
The duration of sleep pattern assessment in night-shift workers was diverse, from 1 day to >4 weeks. To identify the degree and pattern of sleep disturbance caused by shift work, studies have suggested assessing a minimum of 1 week and ideally >2 weeks of continuous recordings of objective measurement by a sleep detection device along with a sleep diary [4]. Additionally, the assessment should include episodes of both the shift work schedule and days off [49]. Overall, 16 studies collected data on sleep patterns for >2 weeks [10,12,[16][17][18][19][20]22,24,25,28,31,34,36,37,40], reporting changes in sleep patterns that could appear in accordance with changes in the work schedule of two-or three-shift workers. In this review, the majority of the studies used a sleep diary in addition to objective measurement, which has been reported to be useful for documenting the factors that can affect sleep over time [42].
This scoping review has a few limitations. First, the studies retrieved for the analysis were limited to English-and Korean-language publications. Second, we have focused on the objective measurements obtained from devices but did not include biophysiological measurements associated with sleep, such as core body temperature, melatonin, and cortisol [50].

Conclusions
With the recent advancements in sleep detection technology that uses sensors, the use of wearable devices in daily living conditions has been gaining popularity for measuring sleep. Wearable devices, which are convenient and reliable, can be used to assess the need for and progress of interventions for sleep disorders in night-shift workers, who may experience changes in their sleep patterns due to disrupted circadian rhythms. The key objective parameters for assessing sleep quality among night-shift workers, such as sleep duration, the timing of sleep and wake times, time required to fall asleep, and number and duration of awakening during the sleep episode, can be efficiently measured using wearable sleep devices. To investigate the consequences of shift work, it is desirable to assess the sleep patterns of night-shift workers for >2 weeks with the combined use of a sleep diary and subjective questionnaire. More studies are needed to test the validity of and identify optimal duration and key parameters for wearable sleep detection devices.