Sleep, Mindfulness, and Shift-Level Safety Performance in Perianesthesia Nurses

Background and Rationale Sleep deprivation and shift work are established threats to human performance, attention, and decision-making, and nurses working rotating or extended shifts in high-acuity perioperative environments are particularly vulnerable to their effects. Perianesthesia nurses, who care for patients before and immediately after anesthesia and surgery, must perform frequent, time-critical, safety-critical tasks such as airway monitoring, medication verification, hemodynamic surveillance, and handoff communication, often under conditions of high cognitive load and interruption. Impaired sleep and accumulated fatigue have been associated in the broader nursing literature with increased rates of near-miss events, procedural deviations, and medication administration errors, but the specific relationship between sleep, fatigue, and shift-level safety performance has not been well characterized within perianesthesia/PACU nursing.

Mindfulness, defined as the capacity to sustain present-moment, non-judgmental awareness of one's internal state and surroundings, has been studied as a psychological trait that may protect against the performance-degrading effects of stress and fatigue. Nurses with higher trait mindfulness may be better able to notice early signs of fatigue-related lapses in attention and to compensate for them during safety-critical tasks. To date, however, few studies have examined whether trait mindfulness moderates the relationship between sleep/fatigue and observed safety performance at the shift level in perianesthesia nursing.

Conceptual Model This study is guided by a conceptual model in which shift-level and person-level sleep parameters (habitual sleep quality, pre-shift sleep duration, and pre-shift sleepiness) and post-shift acute fatigue are hypothesized to be associated with a higher rate of observed safety-process deviations during the shift. Trait mindfulness, measured once at baseline, is hypothesized to act as a moderator that attenuates (buffers) the strength of the sleep-safety and fatigue-safety associations. Because shifts are nested within nurses, and because the same nurse's shifts are likely to be correlated with one another, the design and analysis explicitly account for this nested (multilevel) structure.

Aims and Research Questions

The study addresses the following research questions:

1. What is the level and variability of habitual sleep quality (B-PSQI), pre-shift sleep duration and sleepiness, and post-shift acute fatigue among perianesthesia/PACU nurses?
2. What is the shift-level rate of observed deviations from safety-critical perianesthesia care processes (the Shift-Level Perianesthesia Safety Deviation Rate), and how does it vary across nurses and shifts?
3. Are pre-shift sleep duration and sleepiness, and post-shift acute fatigue, associated with the shift-level safety deviation rate and with the occurrence of near-miss events?
4. Does trait mindfulness (MAAS) moderate the association between sleep/fatigue variables and shift-level safety performance?
5. What shift-level and nurse-level characteristics (e.g., shift length, workload, time of day, years of experience) are associated with safety performance, independent of sleep and fatigue?

Hypotheses H1: Shifts preceded by shorter sleep duration, higher pre-shift sleepiness, or followed by higher post-shift acute fatigue will show a higher Shift-Level Perianesthesia Safety Deviation Rate than shifts with more favorable sleep/fatigue profiles.

H1a: Higher trait mindfulness (MAAS) will be associated with a weaker (attenuated) association between pre-shift sleepiness and the safety deviation rate (a sleep × mindfulness moderation effect).

H1b: Higher trait mindfulness will similarly attenuate the association between post-shift acute fatigue and the safety deviation rate (a fatigue × mindfulness moderation effect).

Design Overview This is a prospective, observational, repeated-measures (within-subject) cohort study. Each participating nurse is followed across multiple of her or his own regularly scheduled shifts, so that both nurse-level (time-invariant) and shift-level (time-varying) predictors of safety performance can be examined. There is no randomization and no experimental intervention; nurses continue their usual clinical duties, and the study team observes and documents safety-process adherence during regularly occurring shifts.

Setting The study will be conducted in a single center, in the perianesthesia/PACU care areas of Ağrı Training and Research Hospital (Ağrı Eğitim ve Araştırma Hastanesi). This is the same institutional setting used in the investigators' related perioperative nursing research.

Population and Eligibility Two levels of eligibility apply: nurse-level eligibility and shift-level eligibility.

Nurse-level inclusion criteria: registered nurses currently working in the perianesthesia/PACU care areas of the study hospital; willingness and availability to be followed across multiple shifts during the study period; and willingness to provide informed consent.

Nurse-level exclusion criteria: nurses who are not directly involved in bedside perianesthesia/PACU patient care during the observation period (e.g., purely administrative roles), and nurses unwilling or unable to complete the baseline and per-shift study instruments.

Shift-level inclusion criteria: regularly scheduled perianesthesia/PACU shifts of the participating nurse during the study period in which direct patient care involving at least one applicable safety-critical process occurs.

Shift-level exclusion criteria: shifts that are shortened, cancelled, or substantially altered such that the planned pre-shift/post-shift data collection and safety-process observation cannot be completed, and shifts for which the researcher-observer is unavailable to complete the observation form.

Because the population under study is hospital nursing staff rather than patients, and because no patient-level clinical intervention is applied, the study accepts healthy working adults (staff) rather than patients as its unit of observation; patients who are cared for during observed shifts are not themselves enrolled as research subjects, and no patient-identifiable data are collected.

Sample Size Based on the anticipated number of eligible perianesthesia/PACU nursing staff at the study hospital and an expected 8 to 12 observed shifts per nurse, the study targets enrollment of approximately 25 to 40 nurses, corresponding to a target of at least 250, and preferably approximately 300, evaluable shift-level observations. This target was selected to provide adequate power for multilevel models examining shift-level associations while accounting for the expected intraclass correlation of repeated shifts within nurses, and to allow for an anticipated degree of missing or unusable shift-level data (e.g., incomplete diaries or observation forms).

Data Collection Instruments

Data collection uses a five-part toolkit developed and/or adapted for this study:

1. Baseline Form: a demographic and professional-characteristics questionnaire (age range, sex, years of nursing experience, years in perianesthesia/PACU care, education level, employment status, typical shift pattern) administered once at enrollment, together with the B-PSQI (an adapted Pittsburgh Sleep Quality Index instrument assessing habitual sleep quality over the preceding month) and the Mindful Attention Awareness Scale (MAAS), a validated self-report measure of trait mindfulness (dispositional attention to and awareness of present-moment experience), including its validated Turkish-language adaptation.
2. Pre-Shift Sleep and Sleepiness Diary: a brief instrument completed by the nurse before each observed shift, capturing sleep duration and subjective sleep quality for the prior sleep period and a momentary sleepiness rating immediately before shift start.
3. Post-Shift Acute Fatigue and Shift-Characteristics Form: completed by the nurse at the end of each observed shift, capturing a momentary/acute fatigue rating together with shift-level characteristics such as shift length, patient volume/acuity, and any self-reported workload or interruption burden during the shift.
4. Perianesthesia Safety Process Observation Form: a researcher-developed structured observation instrument used by trained study personnel during each observed shift to document, for each applicable safety-critical perianesthesia care process (e.g., patient identification and hand-off verification, airway/oxygenation monitoring, medication verification, vital-sign/hemodynamic monitoring at required intervals, and other unit-specific safety-critical checks), whether the process was completed as required or whether a deviation occurred.
5. Near-Miss and Event Log: a structured log used to record the occurrence of any near-miss safety events identified during the observed shift, independent of, but complementary to, the systematic process-deviation observations captured on the Safety Process Observation Form.

Procedures Eligible nurses will be informed about the study, and written informed consent will be obtained prior to any data collection. Consenting nurses will complete the Baseline Form, B-PSQI, and MAAS once, at enrollment. For each of the nurse's shifts selected for observation during the study period, the nurse will complete the Pre-Shift Sleep and Sleepiness Diary shortly before shift start and the Post-Shift Acute Fatigue and Shift-Characteristics Form at the end of the shift. During the same shift, a trained member of the research team, positioned to observe perianesthesia/PACU care without disrupting clinical workflow, will complete the Perianesthesia Safety Process Observation Form and the Near-Miss and Event Log. Data collection will continue until the target number of evaluable shift-level observations (approximately 250-300) has been reached across all participating nurses.

Data Analysis Plan Data will be analyzed using IBM SPSS and R. Descriptive statistics will characterize the nurse sample (demographic and professional characteristics, baseline B-PSQI and MAAS scores) and the shift-level sample (sleep, sleepiness, fatigue, and safety deviation rate distributions). Because shifts are nested within nurses, the primary analyses will use multilevel (hierarchical/mixed-effects) regression models with a random intercept for nurse to account for within-nurse correlation of repeated shift observations. The primary outcome, the Shift-Level Perianesthesia Safety Deviation Rate, will be modeled as a function of pre-shift sleep duration, pre-shift sleepiness, and post-shift acute fatigue, adjusting for relevant shift-level and nurse-level covariates (e.g., shift length, years of experience). Moderation analyses will test the interaction terms of sleep duration × MAAS and acute fatigue × MAAS to evaluate whether trait mindfulness attenuates the association between sleep/fatigue and safety performance (Hypotheses H1a and H1b). Secondary outcomes (post-shift acute fatigue as an outcome in its own right, and occurrence of near-miss events) will be analyzed using appropriate mixed-effects models (linear or generalized linear, as distributionally appropriate) with the same nested random-effects structure. Statistical significance will be set at a two-tailed alpha of 0.05, with effect sizes and confidence intervals reported alongside p-values.

Anticipated Difficulties The investigators anticipate several practical difficulties, including variability in nurses' shift schedules that may complicate the scheduling of repeated observations, the burden of completing pre-shift and post-shift instruments on busy clinical shifts, potential reactivity of nursing behavior to the presence of a researcher-observer (Hawthorne-type effects), and the logistical challenge of achieving observer availability across the full range of shift times (day, evening, night) needed for representative sampling.

Limitations As a single-center, observational, non-randomized study, causal inference regarding the effects of sleep and fatigue on safety performance is limited, and findings may not generalize beyond the study hospital's perianesthesia/PACU setting. The use of self-reported sleep, sleepiness, and fatigue measures is subject to recall and reporting biases, and observer-rated safety-process adherence may be influenced by observer reactivity effects despite efforts to minimize workflow disruption. The relatively modest anticipated sample size, while adequate for the planned multilevel analyses, limits statistical power for detecting smaller moderation effects. Finally, because participating nurses are aware of the study's focus on sleep and safety, social-desirability bias in self-report measures cannot be excluded.

Ethical Considerations This study involves the observation of nursing staff and their shift-level practices; patients cared for during observed shifts are not themselves research subjects, are not approached for consent, and no patient-identifiable information is collected or recorded as part of this study. All participating nurses will provide written informed consent prior to enrollment, and participation is voluntary, with no impact on employment status for nurses who decline to participate or who withdraw at any time. All collected data will be handled in accordance with applicable confidentiality and data-protection principles, with study data coded and stored securely, accessible only to the research team. The study protocol has been submitted to the Ağrı İbrahim Çeçen Üniversitesi Bilimsel Araştırmalar Etik Kurulu (Ağrı İbrahim Çeçen University Scientific Research Ethics Committee) for review.