Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jul 18;49(5):735–743. doi: 10.1002/nur.70106

Sleep Behaviors and Quality of Care Among Public‐Sector Healthcare Workers: Cross‐Sectional and Prospective Associations

Yuan Zhang 1,✉, Alicia Kurowski 2, Suzanne Nobrega 2, Mazen El Ghaziri 1, Rebecca Gore 2, Laura Punnett 2
PMCID: PMC13539641  PMID: 42470295

ABSTRACT

Previous research has shown the effect of healthcare workers' sleep on their safety and health outcomes. The association between healthcare workers' sleep and the quality of care (QOC) has not been systematically investigated. A large survey was collected from a prospective cohort of healthcare workers at five public‐sector facilities in the northeastern U.S. in 2018 and again in 2021. Sleep characteristics assessed were short sleep duration (≤ 6 h/day), sleep disturbances, and sleep at two or more episodes. Among 1553 healthcare workers (mean age 46.4 years, 61.5% female) who completed the 2021 survey, sleep risk factors (range 0–3, sleep disturbances or short duration or at ≥ 2 episodes) were cross‐sectionally associated with a linearly increasing prevalence of poor QOC on the unit and in the last shift, and deteriorated QOC over the past year. Two or more sleep risk factors were associated with all poor QOC outcomes after adjusting for covariates. Among 428 healthcare workers who completed both surveys, two or more sleep risk factors at baseline were associated with a 133%–196% increase in reported deterioration of QOC at follow‐up, but not in poor QOC on the unit or in the last shift, after adjusting for covariates. Healthcare workers' sleep behaviors were significantly associated with QOC outcomes cross‐sectionally and prospectively. Evidence‐based interventions at individual and organizational levels to improve healthcare workers' sleep would likely benefit QOC.

Keywords: quality of care, sleep disturbances, sleep duration, sleep episodes, sleep risk factors

1. Introduction

Healthcare is one of the largest and fastest‐growing sectors in the U.S., employing 22 million workers, or 14% of the U.S. workforce (Laughlin et al. 2021; National Institute of Mental Health 2022). Overall employment in healthcare is projected to grow much faster than all other occupations in the next decade (Bureau of Labor Statistics 2025). Nursing is the largest healthcare profession, with approximately 4.7 million registered nurses (RNs) in the U.S. (American Association of Colleges of Nursing 2024), along with other healthcare workers such as physicians, mental health workers, physical and occupational therapists, dietitians, et al. These workers work collaboratively to provide care to the 342 million people in the U.S., with a projected increase to 383 million by 2054 (Congressional Budget Office 2024).

Due to shift work, long work hours, and high occupational stress, healthcare workers are at high risk for short and disturbed sleep (Hulsegge et al. 2019; Zhang et al. 2021). A recent systematic review of 36 studies reported a link between shift work/night shift and sleep disturbances or quality among healthcare workers (Czyż‐Szypenbejl and Mędrzycka‐Dąbrowska 2024). On the other hand, healthcare workers' sleep could be associated with their safety and health. A systematic review and meta‐analysis of 27 observational studies with a general worker population (n = 268,332) has concluded that workers with sleep problems had a 1.62 times higher risk of being injured than those without sleep problems (Uehli et al. 2014).

Sufficient, high‐quality sleep is crucial for human health and performance. Sleep deprivation may increase fatigue, slow reaction time, reduce attention, and impair work memory and cognitive function (García et al. 2021). Fatigue resulting from short or disturbed sleep may reduce healthcare workers' ability to concentrate and make appropriate clinical judgments, increasing the possibility of errors and reducing the safety of care delivery (Bell et al. 2023). Several studies have shown that sleep deprivation in physicians or nurses contributes to medical errors and near misses (Trockel et al. 2020). One study reported that sleep‐deprived resident physicians reported instances of falling asleep while talking to patients, writing patient notes, or reviewing the investigation (Papp et al. 2004). One meta‐analysis of 60 studies with 959 physicians and 1028 nonphysicians reported sleep loss of 24–30 h (continuous duty without sleep) reduced physicians' overall performance by nearly 1 standard deviation and clinical performance by more than 1.5 standard deviations; in addition, the effect of sleep deprivation on performance was larger in nonphysicians than in physicians (Philibert 2005). All of these are largely related to patient safety.

On the other hand, due to night and rotating shift work, and the resulting circadian misalignment, many healthcare workers develop Shift Work Sleep Disorder (Pallesen et al. 2021), characterized by insomnia and daytime sleepiness. To accommodate daytime sleep and better performance at night work, many workers choose to sleep at two or more episodes (also called segmented or split sleep), for example, sleep immediately after the night shift, then nap before the next night shift. When shift workers sleep in multiple shorter episodes rather than a single consolidated episode, not only may total sleep time be compromised, but sleep may also become less restorative (Lammers‐van der Holst et al. 2024). Shorter episodes, especially during the day or when interrupted, can reduce deep and REM sleep, which support physical recovery, memory, and emotional regulation, and may lead to poor performance. Harrison and colleagues found that roughly 40% of night workers prefer to sleep at more than one episode (Harrison et al. 2021). A literature review of 22 prospective studies suggested that multiple sleep episodes and the related sleep deficiency are associated with a variety of negative performance outcomes (Weaver et al. 2018).

Although studies have shown the effect of healthcare workers' sleep on their safety and health outcomes as well as patient safety, the association between healthcare workers' sleep and direct quality of care (QOC) measures has been understudied, with limited work examining specific indicators of care delivery. A systematic review of 30 studies with healthcare workers (mostly cross‐sectional) concluded that insufficient sleep duration and low sleep quality were related to frequent medical errors and medication errors, low job efficiency, reduced motivation and performance, loss of concentration or attention, and failure to take emergency actions, many of which are patient safety or QOC indicators (Fox et al. 2025). One study (Stimpfel et al. 2020) conducted a cross‐sectional analysis of 1568 surveys with nurses and reported that short sleep duration was significantly associated with lower ratings of QOC and patient safety. Although this study used a one‐item direct QOC measure, the cross‐sectional design and the only measure of sleep duration limited the determination of causality between healthcare workers' sleep behaviors and QOC.

Therefore, to fulfill the knowledge gap in the literature, the objective of this study was to examine the cross‐sectional and prospective associations between sleep behaviors (duration, quality, and pattern) and QOC among healthcare workers.

2. Methods

This study is part of a larger research project. The cross‐sectional analysis used the survey data collected from a sample of 1553 healthcare workers from June to August 2021. The prospective analysis used surveys collected with an overlap sample of 428 healthcare workers who answered both the baseline survey (January to June 2018) as well as the follow‐up survey (June to August 2021).

2.1. Study Population and Sampling of Participants

The larger research project invited participation of healthcare workers from five unionized, state or federal government‐managed public sector facilities in the northeastern U.S., including two mental health hospitals (with a workforce size of 211 and 1183 employees) and three veterans' facilities (with a workforce size of 315, 330, and 1050 employees). A convenience sampling method was used, and all full‐time, part‐time, and per‐diem employees over 18 years old and hired directly by these facilities were eligible to participate.

2.2. Measurements

2.2.1. Dependent Variables

Quality of care (QOC) in the 2021 survey: Perceived QOC was measured with three individual items developed by the research team. These items asked healthcare workers to rate the QOC delivered to patients on the unit, in the last shift, and compared to the past year. QOC on the unit or in the last shift was rated on a 4‐point Likert scale (1 = excellent; 2 = good; 3 = fair; and 4 = poor). Answers to these two items were dichotomized into two categories (excellent/good vs. fair/poor). QOC, compared to the past year, was rated on a 3‐point Likert scale (1 = improved; 2 = remained the same; and 3 = deteriorated) and was dichotomized into two categories (improved/remained vs. deteriorated).

2.2.2. Independent Variables

For cross‐sectional analysis, we used the variables below in the 2021 survey. For prospective analysis, we used the variables below in the 2018 survey.

Sleep duration: Sleep duration was assessed with one item on the typical amount of sleep per 24 h during the work week (5 h or less; 6 h; 7 h; 8 h; 9 h; and 10 h or more). Sleep duration was dichotomized as > 6 h per day, versus ≤ 6 h per day (short sleep duration) (Qiu et al. 2012).

Sleep disturbances: The severity of sleep disturbances was assessed with 8 items from the PROMIS Sleep Disturbance Short Form (SD‐SF) (Yu et al. 2012). Each item was rated on a 5‐point Likert scale with the sum ranging in total raw score from 8 to 40, and a higher score indicates greater severity. We calculated a T‐score from the total raw score, with a range from 28.9 to 76.5 based on the instrument scoring manual. The T‐score was categorized as mild (55.0–59.9), moderate (60.0–69.9), or severe (70.0 and over) sleep disturbances (Yu et al. 2012). A descriptive analysis of this T‐score showed a mean of 50, a median of 51, and a 75% quartile of 56; therefore, sleep disturbances were dichotomized as no (< 55) or yes (≥ 55) for these analyses. The scale demonstrated very good reliability in this study sample (Cronbach's alpha = 0.89).

Sleep episode: The number of sleep episodes was assessed with one item with three categories (1 = all at one time; 2 = at two different times during the day; and 3 = at three or more time periods), which was dichotomized into one episode or ≥ 2 episodes.

2.3. Covariates

For cross‐sectional analysis, we used the sociodemographic covariates below in the 2021 survey. For prospective analysis, we used the sociodemographic covariates below in the 2018 survey.

Sociodemographics: The questionnaire collected information on age, gender, race, education, and job title. Age was grouped as < 40 years or ≥ 40 years. Race was grouped as white or non‐white. Education was grouped as high school or lower compared to college or higher levels. Job titles were grouped as direct care (doctors, nurses, nursing assistants, mental health workers, social workers, first responders, orderlies, and rehabilitation staff) or non‐direct care jobs (lab, housekeeping, dietary, facility, and office/administrative staff).

Work schedules: Usual work shift (days, evenings, nights, and rotating), weekly work hours (≤ 40 or > 40), overtime (yes or no), and mandatory overtime (yes or no) were collected. Weekly work hours, overtime, and mandatory overtime were combined into one variable as any overtime (yes or no).

2.4. Data Collection

Paper surveys were distributed and collected at each facility over a two to 5‐day period to accommodate employees working in different shifts and units. The team members explained the study purpose and procedure, potential benefits and risks, and protection of confidentiality to participants. Employees were given the option to take home the questionnaires to complete in private. Online surveys were also distributed as an option for employees who were not able to complete the paper surveys. Compensation of $10 was offered in 2018, and $25 was offered in 2021 for each completed questionnaire with a paper or online consent. The study was approved by the University of Massachusetts Lowell Institutional Review Board (No. 16‐131‐PUN‐XPD).

2.5. Data Analysis

We used the SPSS software 26.0 to complete all statistical analyses. The associations of each sleep variable (dichotomous) with each QOC outcome (dichotomous) and all sociodemographic and work schedule variables (dichotomous or categorical) were first examined using cross‐tabulation analyses (for associations between two categorical variables) (Field 2019).

Based on an analytic strategy used by Barger et al. (2017), the three sleep variables (dichotomized as yes [1] or no [0]) were combined to generate a sleep‐risk index, ranging from 0 to 3, with the number indicating the number of sleep risk factors (short sleep duration, sleep disturbances, or sleep at ≥ 2 episodes). The association between the sleep‐risk index and each sociodemographic and work schedule variable in the 2021 survey was examined using cross‐tabulation analyses. The association of sleep‐risk index (categorical 0–3) in the 2021 survey (cross‐sectional analysis) and then in the 2018 survey (prospective analysis) with each quality of care outcome (dichotomous) in the 2021 survey was first explored using cross‐tabulation analyses, then examined using the multivariable regression modeling, with adjustment for potential covariates (Field 2019).

Multivariable Poisson regression modeling was used to calculate coefficients and 95% confidence intervals (CI) for each QOC outcome. The prevalence of poor QOC on the unit, in the last shift, and compared to the past year was all over 10%, so Poisson regression modeling with robust variance estimates was more conservative and accurate than logistic regression modeling for calculating prevalence ratios (PR) and 95% CI (Barros and Hirakata 2003). Sociodemographic and work schedule variables were adjusted as covariates in the multivariable regression models.

3. Results

3.1. Descriptive Analyses

A total of 1553 healthcare workers completed the 2021 survey (response rate of 50.3%), with 62.1% females and an average age of 46.6 years (Table 1). Over one‐half (56.3%) were involved in direct care, about one fourth (26.9%) routinely worked evenings, nights, or rotating shifts, and over a third (38.9%) reported working overtime. Short sleep duration (51.1%), sleep disturbances (29.4%), and sleep at ≥ 2 episodes (19.9%) were relatively common. One tenth of these workers reported poor QOC on the unit (10.6%), in the last shift (9.5%), and deteriorated QOC over the past year (12.3%). A total of 37.2% of workers reported no sleep risk factors, while 33.1% reported one, 21.7% reported two, and 8% reported three sleep risk factors.

Table 1.

Sociodemographic characteristics of study participants in the 2021 survey.

Variables Follow‐up participants (n = 1553) Overlap participants (n = 428) Missed participants (n = 1125)
Mean or Percentage SD Mean or Percentage SD Mean or Percentage SD
Age 46.6 12.2 50.1*** 10.9 45.3 12.4
40 years or older 68.1% 78.8%*** 64.1%
Gender (female) 62.1% 65.2% 60.8%
Race (non‐White) 32.8% 32.0% 33.1%
Education (≤ high school) 27.0% 29.8% 25.9%
Job title (direct care) 56.3% 52.8% 57.7%
Shift work
Permanent days 73.1% 77.7% 71.3%
Permanent evenings 14.1% 10.8% 15.3%
Permanent nights 7.6% 5.6% 8.3%
Rotate and others 5.3% 5.9% 5.1%
Overtime (yes) 38.9% 34.8% 40.4%
Sleep‐risk index
0 37.2% 34.5% 38.2%
1 33.1% 33.3% 33.0%
2 21.7% 22.9% 21.3%
3 8.8% 9.3% 7.5%
Open in a new tab

Abbreviation: QOC, quality of care.

***

p < 0.001.

A sub‐sample of 428 healthcare workers completed both the 2018 and 2021 surveys, with similar socio‐demographic characteristics (except older) as reported in 2021 (Table 1). Similar prevalence of sleep at ≥ 2 episodes (20.7%) and short sleep duration (52.7%), and slightly higher prevalence of sleep disturbances (33.5%) were reported in this sub‐sample. This sub‐sample reported similar poor QOC on the unit (9.8%) and in the last shift (10.9%), and slightly more deteriorated QOC over the past year (16.9%) compared to the 2021 survey participants above (n = 1553). Similar distributions of the number of sleep risk factors were reported (Table 1). A comparison between this sub‐sample and those missed in the follow‐up sample revealed similar socio‐demographic characteristics, sleep behaviors, and QOC outcomes, except that this sub‐sample was older (p < 0.001, Table 1).

3.2. Bivariate Analyses

3.2.1. Cross‐Sectional Associations

In the cross‐sectional analysis with 1553 healthcare workers, sleep disturbances and short sleep duration were associated with poor QOC on the unit (p < 0.01) and deteriorated QOC over the past year (p < 0.01). Sleep disturbances (p < 0.01) and sleep at ≥ 2 episodes (p < 0.05) were associated with poor QOC in the last shift.

Sleep‐risk index (range 0–3) was associated with a linear‐increased prevalence of poor QOC on the unit (p < 0.01) and in the last shift (p < 0.01), and deteriorated QOC over the past year (p < 0.01) (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Sleep‐risk index and poor QOC outcomes at follow‐up. **p < 0.01; *p < 0.05. QOC = quality of care. Sleep‐risk index was associated with a linear‐increasing prevalence of poor QOC on the unit (trend p < 0.001), in the last shift (trend p < 0.01), and deteriorated QOC over the past year (trend p < 0.001).

Healthcare workers' short sleep duration was associated with older age (p < 0.05), non‐white (p < 0.05), direct care job (p < 0.01), non‐day shift (including evening, night or rotating shifts, p < 0.01), and overtime (p < 0.01). Sleep disturbances were associated with female (p < 0.01) and white (p < 0.01). Sleep at ≥ 2 episodes were associated with older age (p < 0.01), male (p < 0.05), non‐white (p < 0.01), high school or lower education (p < 0.05), and non‐day shift (p < 0.05). Those who work non‐day shifts or overtime reported more prevalence of at least one sleep risk factor (p < 0.01) than day shift workers or those without overtime.

3.2.2. Prospective Associations

In the prospective analysis with the sub‐sample of 428 healthcare workers, sleep disturbances at baseline (the 2018 survey) were associated with reported deteriorated QOC over the past year at follow‐up (the 2021 survey, p < 0.01). No significant associations were observed between other sleep variables and other QOC outcomes. Sleep‐risk index at baseline had a linear‐by‐linear association with deteriorated QOC over the past year at follow‐up (p < 0.05, Figure 2). Sleep‐risk index was not associated with any other QOC outcomes.

Figure 2.

Figure 2

Open in a new tab

Baseline sleep‐risk index and follow‐up deteriorated QOC. **p < 0.01; *p < 0.05. QOC = quality of care. Sleep‐risk index at baseline had a linear‐by‐linear association with deteriorated QOC over the past year at follow‐up (trend p < 0.05).

3.3. Multivariable Analyses

3.3.1. Cross‐Sectional Associations

Multivariable Poisson regression models of sleep‐risk index and each QOC outcome were built for the 1553 participants (Table 2). Two or more sleep‐risk factors were associated with all poor QOC outcomes after adjusting for sociodemographic and work schedule covariates. Compared to no sleep‐risk factors, healthcare workers with two sleep‐risk factors reported an 85%–126% increase, and those with three sleep‐risk factors reported a 125%–175% increase in each poor QOC outcome, after adjusting for age, gender, race, education, direct care job, shift work, and overtime (Table 2). Older workers reported less poor QOC on the unit and in the last shift, while non‐White workers reported less deteriorated QOC over the past year (Table 2).

Table 2.

Multivariable Poisson regressions for cross‐sectional associations between sleep‐risk index and poor QOC.

Poor QOC on unit (n = 987) Poor QOC in the last shift (n = 966) Deteriorated QOC over the past year (n = 1233)
PR 95% CI PR 95% CI PR 95% CI
Sleep‐risk index
0 1 1 1
1 1.17 0.70–1.94 0.91 0.52–1.61 0.90 0.58–1.40
2 2.15 ** 1.33–3.50 1.85 * 1.11–3.08 2.26 ** 1.59–3.22
3 2.75 ** 1.57–4.80 2.28 ** 1.24–4.21 2.25 ** 1.38–3.67
Age
< 40 years 1 1 1
≥ 40 years 0.45 ** 0.31–0.66 0.49 ** 0.33–0.73 0.79 0.59–1.06
Race
White 1 1 1
Non‐White 1.08 0.72–1.63 1.06 0.67–1.69 0.65 * 0.43–0.99
Open in a new tab

Note: All models were also adjusted for gender, education, direct care job, shift work, and overtime, which were not associated with either QOC outcome.

Abbreviations: CI, confidence intervals; PR, prevalence ratio; QOC, quality of care.

**

p < 0.01

*

p < 0.05.

3.3.2. Prospective Associations

Multivariable Poisson regression models using sleep‐risk index at baseline (the 2018 survey) to link to each QOC outcome at follow‐up (the 2021 survey) were built for the 428 sub‐sample (Table 3). Two or more sleep‐risk factors at baseline were associated with a significantly increased prevalence of deteriorated QOC at follow‐up, but not with poor QOC on the unit or in the last shift, after adjusting for sociodemographic and work schedule covariates. Compared to no sleep‐risk factors, healthcare workers with two sleep‐risk factors reported a 133% increase, and those with three sleep‐risk factors reported a 196% increase in deteriorated QOC at follow‐up (Table 3). Older workers reported less poor QOC on the unit, while workers with high school or less education reported less deteriorated QOC over the past year (Table 3).

Table 3.

Multivariable Poisson regressions for prospective associations between sleep‐risk index and poor QOC.

Poor QOC on unit (n = 241) Poor QOC in the last shift (n = 236) Deteriorated QOC over the past year (n = 326)
PR 95% CI PR 95% CI PR 95% CI
Sleep‐risk index
0 1 1 1
1 0.86 0.31–2.41 0.94 0.34–2.60 1.37 0.70–2.69
2 0.78 0.26–2.35 1.20 0.43–3.37 2.33 * 1.23–4.45
3 1.61 0.59–4.43 1.04 0.33–3.31 2.96 ** 1.36–6.47
Age
< 40 years 1 0.13–0.73 1 0.18–1.01 1 0.43–1.08
≥ 40 years 0.31 ** 0.42a 0.68a
Education
≥ college 1 0.39–2.70 1 0.37–2.41 1 0.26–0.97
≤ high school 1.02 0.94 0.50 *
Open in a new tab

Note: All models were also adjusted for gender, race, direct care job, shift work, and overtime, which were not associated with either QOC outcome. Baseline QOC was not adjusted in these models.

Abbreviations: CI, confidence intervals; PR, prevalence ratio; QOC, quality of care.

**

p < 0.01

*

p < 0.05.

a

p < 0.1.

4. Discussion

In this cross‐sectional and prospective cohort study of mixed‐occupation healthcare workers, over one‐half of the study participants reported short sleep duration (≤ 6 h per day), nearly a third had sleep disturbances, and one‐fifth reported sleep at two or more episodes. The high prevalence of poor sleep behaviors among direct and non‐direct healthcare workers was similar to those in a sample of hospital nurses (Zhang et al. 2018). Roughly ten percent of these workers reported poor quality of care in three different categories, a little lower prevalence than that (16.3%) reported in a previous study of 635 nurses (Gómez‐García et al. 2016).

We selected healthcare workers' sleep behaviors as the primary risk factor, in terms of sleep duration, quality, and pattern, considering the moderately high prevalence of these risks in healthcare workers (Zhang et al. 2021) and their potential direct effects on patient safety (Trockel et al. 2020). Not surprisingly, we found that a greater number of healthcare workers' sleep risk factors (0–3) was associated with a linear‐increasing prevalence of one or more poor QOC outcomes cross‐sectionally and prospectively. Specifically, two or more sleep risk factors were associated with 85%–175% increase in the prevalence of each QOC outcome cross‐sectionally; while two or more sleep risk factors at baseline were associated with a 133%–196% increase in the prevalence of later reported deteriorated QOC over the past year. To our knowledge, this is the first study evaluating the effect of multiple sleep behaviors among healthcare workers on their direct perceived QOC outcomes, not only cross‐sectionally but also prospectively. Consistently, Stimpfel et al. reported in a cross‐sectional analysis that more minutes of sleep before work were significantly associated with higher ratings of patient safety and perceived QOC in over 1500 nurses (Stimpfel et al. 2020). These results are important to healthcare organizations and employers because sleep‐deprived healthcare workers could represent a threat to the safety and health of both workers themselves and patients (Shaik et al. 2022).

In this study, older age, non‐White, and lower education level were protective factors to poor perceived QOC. It is beyond our expectation that shift work and overtime were not associated with QOC in this study. Shift work or overtime usually results in fatigue among healthcare workers, impairing performance and safety and increasing the possibility of errors (Trockel et al. 2020). A scoping review by Bell et al. reported that 82% of the 38 analyzed studies identified fatigue as a contributing factor in medication administration errors and near misses (Bell et al. 2023). Rogers et al. reported that nurses working 12.5 h or longer are three times more likely to make an error compared to those working 8.5 h or less (Rogers et al. 2004). Also reported in a previous study (Olds and Clarke 2010), working more than 40 h in an average week was associated with increased risks of adverse events and errors, including needlestick injuries, work‐related injuries, patient falls with injury, nosocomial infections, and medication errors in a survey of 11,516 nurses. On the other hand, Stimpfel et al. did not find an association of nurses' age, education level, gender, marital status, children responsibility, unit and hospital type with perceived QOC, but found that hospital Magnet status was a protective factor for QOC (Stimpfel et al. 2020).

Given the essential role of healthcare workers in caring for the health of the public, it is an important priority to address deficits in their sleep health, especially in specific sociodemographic and work schedule groups who report higher sleep risks. In this study, we found that healthcare workers who work non‐day shifts or overtime reported a higher prevalence of one or more sleep risk factors. This is expected since shift workers, especially night and rotating shift workers, usually report short sleep duration, sleep disturbances (Min and Hong 2022), and/or multiple sleep episodes (Harrison et al. 2021). Generally speaking, older age, female, non‐white, lower education, direct care job, longer weekly work hours, and overtime were associated with higher sleep risks. This is broadly consistent with previous study findings (Min and Hong 2022; Shigaki et al. 2023) and indicates the need for targeted interventions in these specific sociodemographic or work schedule groups.

4.1. Strengths and Limitations

The strengths of this study include a large sample of mixed‐occupation healthcare workers, both cross‐sectional and prospective analyses, and adjustment for multiple sociodemographic and work schedule covariates.

This study has several limitations. The workforce of these five public sector facilities had union representation, which improved job security and hence supported data quality and response rate. At the same time, it may limit the generalizability of the study findings. The prospective analysis was limited to baseline sleep characteristics linked to perceived QOC outcomes at 3‐year follow‐up, although it strengthens the interpretation that early sleep disruptions may have a long‐term impact on how healthcare workers experience and recall their quality of care, this might not be sufficient to capture the negative effects of persistent sleep problems. Also, the deteriorated QOC over the past year reflected a retrospective perception reported at follow‐up; the temporal association between baseline sleep variables and this retrospective outcome needs to be interpreted with caution.

The survey response was 50.3% of all employees at follow‐up, with an even smaller sample who completed both baseline and follow‐up surveys, which implies potential selection bias. Finally, because both sleep behaviors and the QOC outcomes were self‐reported by the same study participants, this might inflate or deflate the observed relationships due to common method variance. Future studies addressing these limitations and using a larger, more representative sample are needed to further examine the prospective relationships.

5. Conclusion and Implications

In this study, we found that healthcare workers' sleep behaviors were associated with the QOC they provided to patients cross‐sectionally and prospectively. The study findings have great implications for clinical practice. To ensure high quality of care to patients, we need to consider improving the sleep of healthcare workers.

Although sleep is an important self‐care domain, it is greatly affected by extrinsic factors. As we have found in this study, certain demographic groups reported more sleep risks, and several work schedule factors, such as shift work and overtime work, significantly affect healthcare workers' sleep. Healthcare workers' sleep could be improved at the individual level through promoting healthy sleep practices and use of non‐pharmacological strategies, such as aroma therapy, dietary supplements, cognitive behavioral therapy, mind‐body therapy and exercise (Zhang et al. 2023). Moreover, in light of the high prevalence of sleep problems in this critical workforce, organizations and employers may consider offering workplace resources and/or evidence‐based workplace programs to promote the sleep health of this workforce. Effective workplace programs might include shift schedule modification such as policies limiting long work hours and overtime, and workplace‐based sleep promotion programs such as light therapy and multicomponent interventions (Zhang et al. 2023). By addressing both individual behaviors and system work features, healthcare organizations can improve healthcare workers' sleep and well‐being while also enhancing QOC and patient safety and health.

Author Contributions

All authors have made significant contributions to the manuscript and are qualified for authorship. Yuan Zhang, Alicia Kurowski, Suzanne Nobrega, Mazen El Ghaziri, and Laura Punnett had substantial contributions to the design of the work and interpretation of the findings; Yuan Zhang and Rebecca Gore have contributed to the acquisition and analysis of the data; Yuan Zhang drafted the manuscript; and all authors have revised it critically, approved the final version of the manuscript, and agreed to be accountable for all aspects of the work.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

National Institute for Occupational Safety and Health (NIOSH). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NIOSH. This study was supported by Grant Nos. U19 OH008857 and U19 OH012299 from the U.S.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

  1. American Association of Colleges of Nursing . 2024. Nursing Workforce Fact Sheet. https://www.aacnnursing.org/news-data/fact-sheets/nursing-workforce-fact-sheet.
  2. Barger, L. K. , Rajaratnam S. M. W., and Cannon C. P., et al. 2017. “Short Sleep Duration, Obstructive Sleep Apnea, Shiftwork, and the Risk of Adverse Cardiovascular Events in Patients After an Acute Coronary Syndrome.” Journal of the American Heart Association 6, no. 10: e006959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barros, A. J. , and Hirakata V. N.. 2003. “Alternatives for Logistic Regression in Cross‐Sectional Studies: An Empirical Comparison of Models That Directly Estimate the Prevalence Ratio.” BMC Medical Research Methodology 3, no. 1: 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bell, T. , Sprajcer M., Flenady T., and Sahay A.. 2023. “Fatigue in Nurses and Medication Administration Errors: A Scoping Review.” Journal of Clinical Nursing 32, no. 17–18: 5445–5460. 10.1111/jocn.16620. [DOI] [PubMed] [Google Scholar]
  5. Bureau of Labor Statistics . 2025. Healthcare Occupations. https://www.bls.gov/ooh/healthcare/.
  6. Congressional Budget Office . 2024. The Demographic Outlook: 2024‐2054. https://www.cbo.gov/publication/59697.
  7. Czyż‐Szypenbejl, K. , and Mędrzycka‐Dąbrowska W.. 2024. “The Impact of Night Work on the Sleep and Health of Medical Staff—A Review of the Latest Scientific Reports.” Journal of Clinical Medicine 13, no. 15: 4505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Field, A. P. 2019. Discovering Statistics Using SPSS. SAGE. [Google Scholar]
  9. Fox, J. , McGrail M., Cha Y. J., et al. 2025. “A Mixed‐Methods Systematic Review of Sleep Duration and Quality in Healthcare Workers: Impacts on Patient Safety and Quality of Care.” Behavioral Sleep Medicine 23: 698–714. 10.1080/15402002.2025.2522682. [DOI] [PubMed] [Google Scholar]
  10. García, A. , Angel J. D., Borrani J., Ramirez C., and Valdez P.. 2021. “Sleep Deprivation Effects on Basic Cognitive Processes: Which Components of Attention, Working Memory, and Executive Functions are More Susceptible to the Lack of Sleep?” Sleep Science 14, no. 2: 107–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gómez‐García, T. , Ruzafa‐Martínez M., and Fuentelsaz‐Gallego C., et al. 2016. “Nurses' Sleep Quality, Work Environment and Quality of Care in the Spanish National Health System: Observational Study Among Different Shifts.” BMJ Open 6, no. 8: e012073. 10.1136/bmjopen-2016-012073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Harrison, E. M. , Easterling A. P., Yablonsky A. M., and Glickman G. L.. 2021. “Sleep‐Scheduling Strategies in Hospital Shiftworkers.” Nature and Science of Sleep 13: 1593–1609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hulsegge, G. , Loef B., van Kerkhof L. W., Roenneberg T., van der Beek A. J., and Proper K. I.. 2019. “Shift Work, Sleep Disturbances and Social Jetlag in Healthcare Workers.” Journal of Sleep Research 28, no. 4: e12802. 10.1111/jsr.12802. [DOI] [PubMed] [Google Scholar]
  14. Lammers‐van der Holst, H. M. , Quadri S., Murphy A., Ronda J., Barger L. Z., and Duffy J.. 2024. “Evaluation of Sleep Strategies Between Night Shifts in Actual Shift Workers.” Supplement, Sleep Health: Journal of the National Sleep Foundation 10, no. 1S: S108–S111. 10.1015/j.sleh.2023.08.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Laughlin, l. , Anderson A., Martinez A., and Gayfield A.. 2021. “Who are our Healthcare Workers?” https://www.census.gov/library/stories/2021/04/who-are-our-health-care-workers.html.
  16. Min, A. , and Hong H. C.. 2022. “Work Schedule Characteristics Associated With Sleep Disturbance Among Healthcare Professionals in Europe and South Korea: A Report From Two Cross‐Sectional Surveys.” BMC Nursing 21, no. 1: 189. 10.1186/s12912-022-00974-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. National Institute of Mental Health . 2022. Major Depression. https://www.nimh.nih.gov/health/statistics/major-depression.
  18. Olds, D. M. , and Clarke S. P.. 2010. “The Effect of Work Hours on Adverse Events and Errors in Health Care.” Journal of Safety Research 41, no. 2: 153–162. 10.1016/j.jsr.2010.02.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Pallesen, S. , Bjorvatn B., Waage S., Harris A., and Sagoe D.. 2021. “Prevalence of Shift Work Disorder: A Systematic Review and Meta‐Analysis.” Frontiers in Psychology 12: 638252. 10.3389/fpsyg.2021.638252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Papp, K. K. , Stoller E. P., and Sage P., et al. 2004. “The Effects of Sleep Loss and Fatigue on Resident‐Physicians: A Multi‐Institutional, Mixed‐Method Study.” Academic Medicine 79, no. 5: 394–406. 10.1097/00001888-200405000-00007. [DOI] [PubMed] [Google Scholar]
  21. Philibert, I. 2005. “Sleep Loss and Performance in Residents and Nonphysicians: A Meta‐Analytic Examination.” Sleep 28, no. 11: 1392–1402. 10.1093/sleep/28.11.1392. [DOI] [PubMed] [Google Scholar]
  22. Qiu, C. , Gelaye B., Fida N., and Williams M. A.. 2012. “Short Sleep Duration, Complaints of Vital Exhaustion and Perceived Stress are Prevalent Among Pregnant Women With Mood and Anxiety Disorders.” BMC Pregnancy and Childbirth 12, no. 1: 104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Rogers, A. E. , Hwang W. T., Scott L. D., Aiken L. H., and Dinges D. F.. 2004. “The Working Hours of Hospital Staff Nurses and Patient Safety.” Health Affairs 23, no. 4: 202–212. 10.1377/hlthaff.23.4.202. [DOI] [PubMed] [Google Scholar]
  24. Shaik, L. , Cheema M. S., Subramanian S., Kashyap R., and Surani S. R.. 2022. “Sleep and Safety Among Healthcare Workers: The Effect of Obstructive Sleep Apnea and Sleep Deprivation on Safety.” Medicina 58, no. 12: 1723. 10.3390/medicina58121723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Shigaki, L. , Cardoso L. O., and Silva‐Costa A., et al. 2023. “Association Between Sleep Problems and Sociodemographic Characteristics Among ELSA‐Brasil Participants: Results of Multiple Correspondence Analysis.” Sleep Epidemiology 3: 100067. [Google Scholar]
  26. Stimpfel, A. W. , Fatehi F., and Kovner C.. 2020. “Nurses' Sleep, Work Hours, and Patient Care Quality, and Safety.” Sleep Health 6, no. 3: 314–320. 10.1016/j.sleh.2019.11.001. [DOI] [PubMed] [Google Scholar]
  27. Trockel, M. T. , Menon N. K., and Rowe S. G., et al. 2020. “Assessment of Physician Sleep and Wellness, Burnout, and Clinically Significant Medical Errors.” JAMA Network Open 3, no. 12: e2028111. 10.1001/jamanetworkopen.2020.28111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Uehli, K. , Mehta A. J., Miedinger D., et al. 2014. “Sleep Problems and Work Injuries: A Systematic Review and Meta‐Analysis.” Sleep Medicine Reviews 18, no. 1: 61–73. 10.1016/j.smrv.2013.01.004. [DOI] [PubMed] [Google Scholar]
  29. Weaver, M. D. , Vetter C., and Rajaratnam S. M. W., et al. 2018. “Sleep Disorders, Depression and Anxiety are Associated With Adverse Safety Outcomes in Healthcare Workers: A Prospective Cohort Study.” Journal of Sleep Research 27, no. 6: e12722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Yu, L. , Buysse D. J., Germain A., et al. 2012. “Development of Short Forms From the PROMIS™ Sleep Disturbance and Sleep‐Related Impairment Item Banks.” Behavioral Sleep Medicine 10, no. 1: 6–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Zhang, Y. , Duffy J. F., de Castillero E. R., and Wang K.. 2018. “Chronotype, Sleep Characteristics, and Musculoskeletal Disorders Among Hospital Nurses.” Workplace Health & Safety 66, no. 1: 8–15. 10.1177/2165079917704671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Zhang, Y. , ElGhaziri M., Siddique S., et al. 2021. “Emotional Labor and Depressive Symptoms Among Healthcare Workers: The Role of Sleep.” Workplace Health & Safety 69, no. 8: 383–393. 10.1177/21650799211014768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Zhang, Y. , Murphy J., Lammers‐van der Holst H. M., Barger L. K., Lai Y. J., and Duffy J. F.. 2023. “Interventions to Improve the Sleep of Nurses: A Systematic Review.” Research in Nursing & Health 46, no. 5: 462–484. 10.1002/nur.22337. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Articles from Research in Nursing & Health are provided here courtesy of Wiley

RESOURCES