Exploring Factors Influencing Nurses’

Intershift Recovery and Its Effects on

Progression From Acute Fatigue to
Chronic Fatigue
Amany Farag, PhD, RN; Linda Scott, PhD, RN, FAAN; Yelena Perkhounkova, PhD; Peter James Abad, MSc, RN; and
Maria Hein, MSW

Background: Occupational fatigue is a pressing concern among shift workers, notably nurses, leading to substantial costs
related to lost productivity and sick leave. Intershift recovery (IR) is pivotal in preventing acute fatigue (AF) from evolving
into chronic fatigue (CF). However, few studies have been conducted to evaluate factors associated with IR and examine its
mediating role between AF and CF. Purpose: To evaluate factors associated with nurses’ IR and IR’s mediating effect between
AF and CF. Methods: Findings reported in this article are part of a larger mixed-methods study. Data for the parent study
were collected from 1,137 registered nurses (a 56.1% response rate) working in eight midwestern hospitals’ inpatient and
critical care units. The study variables were measured using previously validated self-reported surveys. Multiple regression
analysis was used to assess correlates of IR, and a path analysis was used to evaluate the mediating effect of IR. Results:
Daytime sleepiness and three work environment variables (staffing and resource adequacy, nurse-physician relationship,
and leadership support) were the strongest predictors of IR. Adequate IR mitigated AF from progressing to CF. Conclusion:
The findings support the mediating role of IR in the progression of AC to CF. Modifiable personal and work environment
variables are essential to enhance IR. Thus, hospital leadership should intervene by addressing the modifiable variables to
develop appropriate policies to enhance their staff’s IR.

Keywords: Nurse fatigue, acute fatigue, chronic fatigue, intershift recovery

F atigue is a complex multicausal and multidimensional phe-

nomenon defined as a general sense of tiredness and wea-
riness. Occupational fatigue is a state of tiredness (physical
and mental) that is exacerbated by excessive work demands and
inadequate recovery time that can result in a decline of a work-
tion, job involvement, and negative emotions (Winwood et al.,
2005; Winwood et al., 2006). These attributes are similar to those
of burnout, which is a related but different construct. Burnout is
an emotional state that is attributed to prolonged job stress and
is classified into three main domains of depersonalization, lack
er’s physical and/or mental functions (Barker Steege & Nussbaum, of productivity, and emotional exhaustion (Reith, 2018; Maslach
2013; Winwood et al., 2005; Winwood et al., 2006; Pasupathy & et al., 2001). Thus, we conceptualize burnout as more of an emo-
Barker Steege, 2012; Ahsberg, 2000). Occupational fatigue is a tional and mental state while fatigue is a broader state that has
prevalent issue among shift workers, including nurses. It imposes both mental and physical aspects, and prolonged CF could escalate
substantial costs on employers, with an estimated $136 billion in to burnout, a notion that is supported by other scholars, including
lost productivity and sick leave annually (Ricci et al., 2007). Watterson et al. (2023).
Fatigue can be classified based on the duration as either It is evident in the literature that both fatigue and burn-
acute or chronic (Winwood et al., 2005; Winwood et al., 2006). out have similar negative nurse, patient, and organizational out-
Acute fatigue (AF) is a transient state of tiredness that is relieved comes. Studies have linked fatigue to multiple negative nurse and
by sufficient rest (Winwood et al., 2005; Winwood et al., 2006). patient safety outcomes such as medication errors, decision regret,
However, if nurses cannot fully recover between shifts, or in other accidents, and near accidents, as well as nurses’ intent to leave the
words, if they have inadequate intershift recovery (IR), acute profession (Barker & Nussbaum, 2011; Geiger-Brown et al., 2012;
fatigue can escalate to chronic fatigue (CF; Winwood et al., 2005; Rogers, 2008; Scott et al., 2007; Scott et al., 2014; Cho & Steege,
Winwood et al., 2006). CF, a maladaptive manifestation of AF, 2021). Because of the negative consequences of nurses’ occupa-
is characterized by a decline in interest, motivation, concentra- tional fatigue, the American Nurses Association (ANA; 2014)

Volume 15/Issue 3 October 2024 www.journalofnursingregulation.com 23

and the American Academy of Nursing (Caruso et al., 2017) have evaluated the mediating effect of IR between acute and chronic
issued position statements calling for measures to mitigate nurse fatigue (Winwood et al., 2005; Winwood et al., 2006; Barker &
fatigue. For example, the ANA’s statement emphasized that assur- Nussbaum, 2011). Therefore, guided by Winwood et al.’s (2006)
ing nurses’ full recovery before reporting to work is an essential occupational fatigue, exhaustion, and recovery (OFER) frame-
shared responsibility between nurses and their employers (ANA, work, the present study aimed to evaluate personal and occupa-
2014). Our work and that of others found that nurses predomi- tional variables associated with nurses’ IR as well as IR’s mediating
nantly report higher levels of acute than chronic fatigue (Barker effect between AF and CF. The research questions addressed in
& Nussbaum, 2011; Farag et al., 2022; Sagherian et al., 2023). this study were as follows: (a) Which of the nurses’ personal and
Adequate IR is one avenue toward mitigating occupational fatigue, occupational variables significantly predict nurses’ IR? and (b)
particularly preventing AF from escalating to CF (Winwood et al., Does nurses’ IR mediate the relationship between AF and CF?
2005; Winwood et al., 2006). Therefore, it is important to identify
factors that contribute to IR. This information could help hospi-
tal administrators evaluate different aspects of their organizational Methods
structure and work environment, develop appropriate policies to Design, Setting, and Sample
enhance nurses’ IR, and provide nurses with self-management areas The findings reported in this manuscript are part of a larger multi-
to help them achieve appropriate IR. site mixed methods study conducted to evaluate predictors of nurse
The results of a scoping review evaluating fatigue and recov- fatigue and the mediating effect of IR between fatigue and patient
ery among nurses showed that occupational variables of quick safety outcomes (medication errors and near misses). All registered
return to work, rotating shifts, high work demands, and work- nurses (RNs; N = 2,026) working in inpatient, critical care, and
ing overtime were associated with poor IR (Gifkins et al., 2020). emergency departments of eight conveniently selected hospitals,
Regarding shift length, surprisingly, nurses working 12-hour shifts all within a 120-mile radius of the principal investigator’s (A.F.)
reported better IR than nurses working 8-hour shifts (Gifkins et academic institution, received the study survey. The selected sites
al., 2020). Control over one’s work schedule and adequate breaks included one academic medical center, two mid-size community
during shift work were associated with adequate IR (Gifkins et al., hospitals, and five hospitals affiliated with one healthcare system.
2020). Yamaguchi et al. (2023) reported that nurses working three- The academic medical center was the only magnet hospital. Nurse
shift rotations experienced significantly lower IR compared with managers, assistant managers, and agency nurses were excluded
their colleagues working two-shift rotations. One study found that from the study. This exclusion criterion is consistent with previous
regardless of shift rotations, working overtime was associated with seminal fatigue studies (Barker & Nussbaum, 2011; Geiger-Brown
poor IR (Yamaguchi et al., 2023), while in another study, shift et al., 2012; Rogers, 2008; Rogers et al., 2004; Scott et al., 2007;
schedule was not associated with IR (Yu et al., 2019). After con- Scott et al., 2014; Cho & Steege, 2021).
trolling for personal factors, Chen et al. (2014) found that hospital
magnet status (magnet, pursuing magnet, and non-magnet) was Instruments
not significantly associated with IR. The absence of statistical sig- Nurses’ acute and chronic fatigue were measured using the OFER
nificance was attributed to the small sample size, but the findings scale. The two fatigue types were measured using 10 items (5 items
were consistent with the researchers’ assumption: data for magnet for each type). Each item is measured on a 7-point Likert scale
hospitals were trending toward better IR when compared with IR ranging from 0 (strongly disagree) to 6 (strongly agree). The total
findings at non-magnet hospitals (Chen et al., 2014). score for each subscale was computed to yield a summative score
In addition to occupational variables, nurses’ personal vari- ranging from 0 to 100, with a high score indicating a high level of
ables are important and could contribute to IR. Good sleep quality fatigue (Winwood et al., 2005; Winwood et al., 2006). In the pres-
and regular exercise are two important modifiable variables that ent study, the reported Cronbach’s alpha was 0.87 for AF and 0.88
are conducive to appropriate IR (Chen et al., 2014; Yamaguchi for CF. Nurses’ IR was measured using five items of the OFER
et al., 2023; Gifkins et al., 2020). Surprisingly, informal caregiv- scale. The scoring procedure is similar to that of the fatigue pro-
ing responsibility was not found to be related to the quality of IR cedure. A high score indicates high IR. In this study, the reported
(Yamaguchi et al., 2023). Although nurses’ age has been frequently Cronbach’s alpha for IR was 0.95.
examined in nurses’ occupational fatigue and recovery literature, Multiple standardized and investigator-developed measures
findings have been inconsistent (Gifkins et al., 2020). assessed nurses’ work and personal variables. Nurses’ personal vari-
To the best of our knowledge, the relationship between IR ables included age, gender, highest nursing educational degree,
and other personal and work variables such as commute time, social marital status, number and age of children, and home caregiv-
support, caffeine consumption, number of children, and work envi- ing responsibilities. Perceived social support was measured by one
ronment have not been adequately examined. Furthermore, most question in which nurses were asked to rate their perceived social
of the studies evaluating AF, CF, and IR predominantly studied support at home using an 11-point Likert scale ranging from 0 (not
nurses working 12-hour shifts or night shifts. Finally, few studies at all) to 10 (a great extent). Additional personal variables that were

24 Journal of Nursing Regulation

measured included one-way commute time (in minutes), exercise for leadership support, 0.91 for nurse-physician relationship, and
duration (in minutes per day) and frequency (times per week), and 0.81 for adequacy of staffing and resources.
caffeine/caffeinated beverage consumption in ounces before, dur-
ing, and after work. Procedures
Two validated and previously used sleep quality and day- After obtaining the human subjects’ approval from the University
time sleepiness measures were used to evaluate the two variables. of Iowa institutional review board (IRB# 201503758), the study’s
The Pittsburg Sleep Quality Index (PSQI) was used to assess principal investigator attended the monthly staff meeting for all
sleep quality. The PSQI consists of 19 items covering seven sleep units at each of the participating sites. During each meeting, the
domains: duration, disturbance, latency, day dysfunction, effi- principal investigator introduced the study, answered any ques-
ciency, overall sleep quality, and use of sleep medication. Scores for tions, and at the end of the meeting distributed the paper-based
each item range from 0 to 3, with total scores ranging from 0 to study survey in all the nurses’ mailboxes. Each participant had 4
21. A low score indicates good sleep quality (Buysse et al., 1989). weeks to complete and mail (using regular mail) the completed
In the present study, the Cronbach’s alpha for the PSQI was 0.70. survey to the research team. In order to increase the response rate,
The Epworth Sleepiness Scale (ESS) was used to measure daytime two reminder flyers and one “last-call” flyer were posted on the
sleepiness. The ESS consists of eight items assessing participants’ announcement boards weekly for 3 weeks. After a completed sur-
tendency to doze while performing eight daily activities. Each item vey was received, that nurse was sent a $20 compensation check.*
is measured on a 4-point Likert scale, ranging from 0 (would never The study’s data collection started in May 2017 and was con-
doze) to 3 (high chance of dozing). The ESS total score ranges from cluded in January 2018. Although the data were collected before
0 to 24, with scores greater than 16 indicating a pathological ten- the COVID-19 pandemic, presenting a potential threat to exter-
dency for daytime sleepiness (Johns, 1991). The study’s reported nal validity, we posit that the reported results closely reflect nurse
Cronbach’s alpha for the ESS was 0.82. fatigue and IR levels and can be used to inform creative solutions
In addition to the personal variables mentioned earlier, mul- and potential regulatory efforts.
tiple work variables that we hypothesized would influence nurse
fatigue and IR were measured. These variables included hospi- Sample Size Estimation
tal site; unit type; years of experience in nursing, in the unit, and Using F statistics for multiple regression, a conservative approach
with the nurse manager; employment status (full-time, part-time, of medium effect size (0.15) (Cohen, 1988), alpha of 0.05, power of
as needed); working hours per week; and shift worked. Although 0.80, and 30 predictors were entered into G-power software to esti-
nurses from eight sites were surveyed for this study, because of the mate the minimum required sample size. A minimum sample size
small sample size in some of the hospitals, the sites were combined of 200 RNs was calculated. To account for potential nonresponse
into three site categories based on hospital size and affiliation. The and missing data, oversampling was performed and resulted in a
two midsize community hospitals were combined into one cate- minimum required sample size of 600 RNs.
gory (Site 1), the five hospitals affiliated with one healthcare sys-
tem were combined into a second category (Site 2), and the magnet Data Analysis
academic medical center was the final site (Site 3). SAS (version 9.4) and Mplus (version 8.8) software were used
The unit work environment is an additional work variable for statistical analysis. Descriptive statistics were calculated and
that was measured. This variable was measured using two sub- checked for all variables in the study. Variable distributions, pat-
scales of leadership support (5 items) and collegial nurse-physician terns of missing data, and statistical test assumptions were exam-
relationship (3 items) from the Practice Environment Scale of the ined. Missing data did not exceed 5% for any single variable.
Nursing Work Index (Lake, 2002), as well as one subscale of staff- Individual mean substitution was used to impute missing values
ing and resource adequacy (7 items) from the Nursing Work Index- in any of the work environment scales, and listwise deletion was
Revised (NWI-R; Aiken & Patrician, 2000). The NWI-R labeled used for demographics, sleep, and fatigue variables. A significance
this subscale as a measure of control over practice; however, after level of alpha = 0.05 was used for statistical tests.
the research team reviewed the wording of each item, the team
reached a consensus to label the subscale as staffing and resource Correlates of IR
adequacy, not control over practice. This labeling is similar to that Bivariate relationships between IR and sample personal and occu-
used by Lake (2002). It is also consistent with the recommenda- pational variables were tested using the independent samples’ t test
tions of a previous study conducted to evaluate the psychometric or analysis of variance for categorical variables and Pearson’s cor-
properties of the NWI-R (Slater & McCormack, 2007). The 15 relation coefficient for continuous variables. A multiple regression
items of the three subscales were measured on a 4-point Likert * The survey was anonymous. A separate sheet containing the partici-
scale ranging from (1) strongly agree to (4) strongly disagree. A pants’ name and address was mailed back with the completed survey.
high score indicates a favorable work environment dimension. In This sheet was immediately separated from the participants’ survey, saved
the present study, the reported Cronbach’s alpha scores were 0.88 in a locked file cabinet, and shredded after processing the compensation.
This procedure was approved by the IRB.

Volume 15/Issue 3 October 2024 www.journalofnursingregulation.com 25

model for IR as the dependent variable and personal and occupa- (α < 0.05). Standardized estimates of coefficients and confidence
tional factors as independent variables was developed using the intervals were calculated to allow the interpretation of results in
same procedure used in our previous work (Farag et al., 2022). terms of standard deviation units for continuous variables (Muthén
Bivariate relationships among the characteristics were checked for et al., 2016). Unstandardized estimates were also reported to allow
potential collinearity. interpretation of results in units of original scales.

Evaluating the Mediating Effect of IR Between Acute and Chronic

Fatigue Results
To examine a potential mediating role of IR on the relationship A total of 1,137 nurses participated in this study for a response rate
between acute and chronic fatigue, a path analysis approach was of 56.1%. The majority of the participants were women (93.9%),
used and implemented in the structural equation modeling frame- and participants had a mean age of 35.1 (SD = 11.9) years. More
work offered by Mplus software. Multiple regression models for than half of the participants were baccalaureate-prepared RNs
IR and chronic fatigue were estimated simultaneously. AF was (63.5%). One-quarter of the participants worked in critical care
included as an independent variable in both models, assuming it units (25.9%), while 22.9% worked in medical-surgical units
explains both IR and CF. IR served as a dependent variable and (Table 1). Nurses in our study had a mean of 9.6 (SD = 10.6) years
mediator (intermediary variable) between AF and CF. According of nursing experience and worked an average of 35.3 (SD = 7.2)
to our hypothesized model, CF can be affected by AF directly and hours weekly (Table 2). Our participants reported moderate to
indirectly via IR. high acute fatigue (M = 67.3, IQR = 53.3–83.3, SD = 20.4) and
To minimize potential mediator-outcome confounding, a moderate chronic fatigue (M = 41.4, IQR = 20–60, SD = 23.7)
series of personal and occupational control variables were included and IR (M = 49.0, IQR = 33.3–63.3, SD = 21.1).
in the models. Personal variables included exercise, daytime sleepi-
ness, amount of coffee consumed during work, age, one-way com- Correlates of IR
mute time, and sleep quality; occupational variables included unit Bivariate relationships of IR with personal and occupational vari-
type, site, work shift, years as RN, years on the unit, work hours ables are described in Table 1 (categorical variables) and Table 2
per week, leadership support, nurse-physician relationship, and (continuous variables). There was a significant difference with
staffing/resource adequacy. These variables were selected because respect to IR between participants based on their exercise dura-
they had significant bivariate associations with AF and/or CF. In tion (p < .001), daytime sleepiness (p < .001), caffeine consumption
addition, we included in the models for IR and CF other types of during work (p = .004), work site (p = .008), employment status
fatigue measured in this study (physical fatigue, mental fatigue, (p = .005), and work shift (p = .018). IR was significantly associ-
and total fatigue) to decrease the potential for epiphenomenal asso- ated with all continuous variables (p = .003 or less), including per-
ciations (i.e., if an indirect effect through IR is observed because sonal factors (positively associated with age and social support and
IR is correlated with another variable through which AF indirectly negatively associated with commute time and poor sleep quality)
affects CF). We decided not to include gender, marital status, edu- and occupational factors (positively associated with nursing expe-
cation, and caregiving, even though they had significant bivari- rience, experience on the unit, experience with the current man-
ate associations with CF, because these variables had at least one ager, and work environment, and negatively associated with hours
category with a small frequency, which could potentially impact worked).
estimation. After discussion, we determined that these factors were The multiple regression model for IR explained 27% of the
unlikely to cause confounding. Also, we did not include years with variance in IR (Table 3). Lower IR was associated (p < .001) with
the nurse manager even though it had a significant bivariate asso- higher daytime sleepiness (b = -4.56), longer one-way commute
ciation with IR to avoid collinearity issues because we were already (b = -0.09), and poorer sleep quality (b = -1.50). With respect to
including years as RN and years on the unit that relate to both occupational variables, lower IR was associated with more work
IR and CF. hours per week (b = -0.19, p = .021). Higher IR (p < .01) was
The models were fit using maximum likelihood estimation associated with work site (i.e., nurses working in other sites had
with bootstrap standard errors and confidence intervals for parame- higher IR than nurses working in the academic medical center;
ter estimates based on 10,000 bootstrap draws. Bootstrap standard b = 4.68 for Site 1 vs. Site 3, and b = 5.66 for Site 2 vs. Site 3), more
errors are robust to violations of model assumptions, and bootstrap years of nursing experience (b = 0.21), and positive work environ-
confidence intervals are more appropriate for indirect effect esti- ment (leadership support [b = 3.61], nurse-physician relationship
mates than conventional, symmetric confidence intervals because [b = 3.02], and staffing and resource adequacy [b = 7.65]).
the distributions of indirect effects are usually not normal (Bai
et al., 2019; Geiser, 2013; Kline, 2016). Inferences about indirect Mediation Analysis
effects and specified paths in the path diagram were made based Path analysis results for chronic fatigue and IR are presented in
on confidence intervals and the significance of coefficient estimates Table 4 and Figure 1. The models explained 57% of the variability

26 Journal of Nursing Regulation

 TABLE 1

Relationships Between Sample Characteristics (Categorical Variables) and Intershift

Recovery

Intershift Intershift
Recovery Recovery
(0–100) (0–100)
Variable n % M ± SD pa Variable n % M ± SD pa
Personal factors Daytime sleepiness 1,122 <.001
Gender 1,124 .074 0–8 607 54.1 53.4 ± 20.7
Male 69 6.1 53.2 ± 20.2 9–24 515 45.9 43.7 ± 20.6
Female 1,055 93.9 48.6 ± 21.1 Caffeine consumption 1,136 .673
Marital status 1,133 .085 Yes 934 82.2 48.9 ± 21.1
Married 599 52.9 50.4 ± 21.1 No 202 17.8 49.6 ± 21.2
Divorced/widowed 90 7.9 49.2 ± 23.3 Caffeine before work, oz 1,130 .051
Single 321 28.3 47.2 ± 19.9 0 494 43.7 50.5 ± 21.0
Lives with a partner 123 10.9 46.6 ± 22.1 >0 to <16 416 36.8 48.7 ± 20.6
Education 1,135 .770 ≥16 220 19.5 46.3 ± 21.9
Diploma 19 1.7 51.4 ± 25.7 Caffeine during work, oz 1,131 .004
Associate 329 29.0 48.0 ± 22.7 0 352 31.1 50.6 ± 21.9
BSN 721 63.5 49.3 ± 20.3 >0 to <16 339 30.0 50.7 ± 20.7
MSN/DNP 66 5.8 49.1 ± 20.4 ≥16 440 38.9 46.4 ± 20.5
Number of children living with 1,137 .204 Occupational factors
0 667 58.7 48.6 ± 20.9 Type of unit 1,133 .075
1 159 14.0 48.9 ± 21.6 Medical/surgical 259 22.9 49.1 ± 21.3
2 187 16.4 47.8 ± 20.4 Critical care 293 25.9 47.7 ± 20.1
≥3 124 10.9 52.7 ± 22.7 Pediatrics 206 18.2 47.4 ± 19.8
Youngest child age 1,137 .619 Mother and baby 122 10.8 52.7 ± 23.6
No children 667 58.7 48.6 ± 20.9 Emergency department 105 9.3 52.9 ± 20.7
<3 years 190 16.7 49.0 ± 20.0 Other 130 11.5 47.3 ± 22.2
4–11 years 165 14.5 50.9 ± 22.4 Float 18 1.6 54.6 ± 21.5
≥12 years 115 10.1 48.1 ± 22.6 Site 1,132 .008
Caregiving responsibility 1,131 .981 Site 1 156 13.8 51.9 ± 22.8
Yes 80 7.1 49.0 ± 21.5 Site 2 326 28.8 50.9 ± 22.3
No 1,051 92.9 49.0 ± 21.0 Site 3 650 57.4 47.3 ± 19.9
Second job 1,133 .426 Employment status 1,131 .005
Yes 149 13.2 47.7 ± 20.0 Full time 799 70.7 47.9 ± 20.6
No 984 86.9 49.2 ± 21.3 Part time/PRN 332 29.4 51.7 ± 22.2
Exercise, min/d 1,129 <.001 Work shiftb 1,137 .018
0 475 42.1 45.5 ± 21.0 Morning-day 416 36.6 51.1 ± 21.7
1–40 224 19.8 51.2 ± 20.5 Day-evening 73 6.4 51.6 ± 21.7
>40–120 430 38.1 51.7 ± 21.2 Night-morning 294 25.9 46.4 ± 21.4
Rotating/other 354 31.1 48.1 ± 19.9
Note. BSN = bachelor of science in nursing; DNP = doctor of nursing practice; MSN = master of science in nursing; PRN = “pro re nata” (as needed).
a p values calculated using analysis of variance or independent samples t tests.
b Morning-day shifts: 7AM–7PM, 7AM–3PM, 8AM–5PM; day-evening shifts 11AM–7PM, 3PM–11PM; night-morning shifts 7PM–7AM, 11PM–7AM; rotating/other:

includes rotating shifts and other shifts.

in CF and 45% of the variability in IR. The direct effect of AF on The total (combined direct and indirect) effect of AF on CF was
CF was statistically significant (b = 0.26, p < .001), controlling for estimated as 0.41; that is, for a one-point increase in AF, CF is
all other variables in the model. IR was also significant (b = -0.34, expected to increase by 0.42.
p < .001). A one-point increase in IR was associated with a 0.34-
point decrease in CF. Model-estimated total, direct, and indirect
effects of AF on CF are shown in Table 5. The mediating (indi-
rect) effect of AF on CF via IR was significant (b = 0.16, p < .001).

Volume 15/Issue 3 October 2024 www.journalofnursingregulation.com 27

 TABLE 2 Discussion
The two-fold purpose of this study was to assess predictors of
Relationships Between Personal and
Occupation Variables (Continuous Variables)
nurses’ IR and to examine IR’s mediating effect between AF and
and Intershift Recovery CF. Whereas the majority of the proposed occupational and per-
sonal variables were significantly associated with IR (Tables 1 and
Intershift Recovery 2), fewer variables were significant in the model for IR and bivari-
(0–100)
ate regression analysis (i.e., sleep quality, daytime sleepiness, com-
Variable n M ± SD r 95% CI p
mute time, work site, years of experience, working hours per week,
Personal factors
and the three work environment variables; Table 3). These findings
Age, y 1,134 35.1 ± 11.9 .10 (.04, .16) <.001
One-way 1,117 24.7 ± 18.1 -.12 (-.17, -.06) <.001
provide support for the hypothesis that adequate IR is a mecha-
commute, min nism that prevents AF from progressing to CF. A detailed dis-
Social support 1,131 5.3 ± 3.3 .10 (.04, .16) <.001 cussion of nurses’ AF and CF is presented elsewhere (Farag et al.,
(0–10) 2022). In this report, we focused on results related to IR and its
Sleep quality 1,086 7.8 ± 3.5 -.34 (-.39, -.28) <.001 mediating effect.
(1–21)
Occupational factors Personal and Occupational Variables Associated IR
Years as registered 1,136 9.6 ± 10.6 .14 (.09, .20) <.001
The study results highlight the value of good sleep (high sleep
nurse
quality and low daytime sleepiness) and exercise in improving
Years with nurse 1,119 3.0 ± 3.9 .09 (.03, .15) .003
manager nurses’ IR. This finding is consistent with findings from previous
Years on the unit 1,132 6.2 ± 7.8 .12 (.06, .18) <.001 studies (Alsayed et al., 2022; Ross et al., 2021; Caboral-Stevens
Working h/wk 1,106 35.3 ± 7.2 -.14 (-.19, -.08) <.001 et al., 2023). Both sleep and exercise are important measures to
Leadership 1,134 2.1 ± 0.7 .28 (.23, .34) <.001 help nurses recharge and recover before resuming their work
support (0–3) and should be included as an integral part of any nurse well-
Nurse-physician 1,134 2.1 ± 0.7 .27 (.21, .32) <.001 ness initiative. Consuming coffee or caffeinated beverages during
relationship (0–3) work was another personal variable that was associated with IR.
Staffing/resource 1,134 2.0 ± 0.6 .31 (.25, .36) <.001
Interestingly, nurses who did not consume caffeine during work or
adequacy (0–3)
a p values are for Pearson correlation coefficients (r).
did not consume more than 16 oz of a caffeinated beverage during
work reported better IR as compared with their colleagues who
consumed more than 16 oz during work. This result adds to the
TABLE 3 mixed evidence on the impact of caffeine on fatigue and recovery,
including findings from our previous study (Farag et al., 2022).
Multiple Regression Model for Intershift
Recovery (N = 1025) While some studies have suggested that caffeine can help reduce
fatigue and improve alertness in the short term (James & Rogers,
Variable b 95% CI p 2005), others have cautioned about potential negative effects on
Intercept 38.38 (29.45, 47.32) <.001 sleep quality and subsequent recovery (Drake et al., 2013). Based
Daytime sleepiness on our results, we could assume that nurses who consumed caf-
9-24 vs. 0-8 -4.56 (-6.92, -2.19) <.001 feine in moderation were able to have better sleep and, therefore,
One-way commute, min -0.09 (-0.16, -0.03) .006 were able to achieve better IR. However, we could also argue that
Sleep quality -1.50 (-1.84, -1.16) <.001
nurses who achieved better IR (e.g., slept better) didn’t need to
consume as much coffee as those nurses who had poor IR. Becuse
Site
this is a cross-sectional study, so we can’t make causal conclusions.
Site 1 vs. Site 3 4.68 (1.16, 8.20) .009
Further longitudial research is needed to better understand this
Site 2 vs. Site 3 5.66 (2.90, 8.43) <.001
relationship. Although other personal variables, such as age and
Years as a registered nurse 0.21 (0.09, 0.32) <.001 social support, were significantly associated with IR, these vari-
Working hours/week -0.19 (-0.36, -0.03) .021 ables were statistically but not clinically significant. Overall, these
Leadership support 3.61 (1.58, 5.63) .001 findings highlight the importance of designing and implementing
Nurse-physician relationship 3.02 (1.13, 4.91) .002 wellness programs that incorporate modifiable personal variables
Staffing/resource adequacy 7.65 (5.22, 10.08) <.001 (Stanulewicz et al., 2019). Furthermore, organizational policies and
Note. b = unstandardized regression coefficient. Model F (10,1014) = 37.23,
regulatory processes should be developed to provide nurses with
R2 = .27. sufficient time to allow them to participate in these wellness pro-
grams. These policies could also include some incentives for nurses
who actively participate in various organizational wellness pro-

28 Journal of Nursing Regulation

 TABLE 4

Path Analysis Results for Chronic Fatigue and Intershift Recovery (N = 993)

Chronic Fatigue Intershift Recovery

Unstandardized Standardized Unstandardized Standardized
Variable b 95% CI β 95% CI b 95% CI β 95% CI
Intercept 42.74*** 29.25, 56.50 1.81*** 1.23, 2.39 84.38*** 71.85, 96.54 3.96*** 3.38, 4.54
Intershift recovery -0.34*** -0.41, -0.27 -0.31*** -0.37, -0.25 NA NA NA NA
Acute fatigue 0.26*** 0.19, 0.33 0.22*** 0.16, 0.28 -0.46*** -0.54, -0.39 -0.44*** -0.50, -0.37
Physical fatigue 0.06 -0.01, 0.13 0.05 -0.01, 0.10 -0.08* -0.16, -0.01 -0.07* -0.13, -0.01
Mental fatigue 0.08* 0.00, 0.15 0.07* 0.00, 0.14 -0.02 -0.09, 0.05 -0.02 -0.09, 0.05
Total fatigue 0.06 -0.01, 0.13 0.06 -0.01, 0.12 -0.05 -0.12, 0.02 -0.05 -0.13, 0.03
Personal Factors
Exercise, min/d
1–40 vs. 0 1.45 -1.25, 4.23 0.06 -0.05, 0.18 1.40 -1.32, 4.12 0.07 -0.06, 0.19
>40–120 vs. 0 0.07 -2.39, 2.50 0.00 -0.10, 0.11 0.67 -1.76, 3.07 0.03 -0.08, 0.14
Daytime sleepiness
9–24 vs. 0–8 -2.25 -4.58, 0.10 -0.10 -0.19, 0.00 0.48 -1.80, 2.78 0.02 -0.09, 0.13
Caffeine during work, oz
>0 to <16 vs. 0 0.31 -2.18, 2.79 0.01 -0.09, 0.12 1.62 -1.03, 4.27 0.08 -0.05, 0.20
≥16 vs. 0 -1.45 -3.98, 1.09 -0.06 -0.17, 0.05 0.27 -2.27, 2.82 0.01 -0.11, 0.13
Age, y -0.12 -0.32, 0.07 -0.06 -0.16, 0.04 0.03 -0.16, 0.22 0.02 -0.09, 0.12
One-way commute, min -0.02 -0.08, 0.04 -0.02 -0.06, 0.03 -0.03 -0.09, 0.03 -0.02 -0.07, 0.03
Sleep quality 0.22 -0.12, 0.56 0.03 -0.02, 0.08 -0.81*** -1.15, -0.47 -0.13*** -0.19, -0.08
Occupational Factors
Unit type
Medical/surgical vs. ED 7.42** 3.10, 11.86 0.31** 0.13, 0.50 -0.72 -4.85, 3.27 -0.03 -0.23, 0.16
Critical care vs. ED 2.62 -1.76, 7.21 0.11 -0.07, 0.30 -1.67 -5.73, 2.36 -0.08 -0.27, 0.11
Pediatrics vs. ED 0.16 -4.54, 5.01 0.01 -0.19, 0.21 -1.10 -5.26, 3.03 -0.05 -0.25, 0.14
Mother and baby vs. ED 3.82 -1.50, 9.18 0.16 -0.06, 0.39 -1.36 -6.00, 3.29 -0.06 -0.28, 0.15
Other vs. ED 5.13* 0.22, 10.11 0.22* 0.01, 0.43 -2.92 -7.84, 2.11 -0.14 -0.37, 0.10
Site
Site 1 vs. Site 3 -3.15 -6.59, 0.22 -0.13 -0.28, 0.01 3.79 -0.06, 7.63 0.18 0.00, 0.36
Site 2 vs. Site 3 3.30* 0.18, 6.38 0.14* 0.01,0.27 4.68** 1.79,7.70 0.22** 0.09,0.36
Work shifta
Day-evening vs. morning-day 4.06 -0.93, 9.00 0.17 -0.04, 0.38 -0.36 -4.51, 3.84 -0.02 -0.21, 0.18
Night-morning vs. morning-day -0.73 -3.50, 1.96 -0.03 -0.15, 0.08 -2.58 -5.42, 0.12 -0.12 -0.25, 0.01
Rotating/other vs. morning-day -1.72 -4.59, 1.03 -0.07 -0.19, 0.04 -0.67 -3.46, 2.06 -0.03 -0.16, 0.10
Years as RN 0.24* 0.00, 0.48 0.10* 0.00, 0.21 0.00 -0.24, 0.22 0.00 -0.12, 0.11
Years on the unit 0.37*** 0.16, 0.56 0.12*** 0.05, 0.19 0.10 -0.11, 0.31 0.04 -0.04, 0.12
Working h/wk 0.29*** 0.13, 0.44 0.08*** 0.04, 0.13 -0.25** -0.41, -0.09 -0.08** -0.13, -0.03
Leadership support -2.07* -4.11, -0.07 -0.06* -0.12, 0.00 2.47* 0.52, 4.40 0.08* 0.02, 0.14
Nurse-physician relationship -0.62 -2.49, 1.25 -0.02 -0.07, 0.04 0.79 -1.17, 2.76 0.03 -0.04, 0.09
Staffing/resource adequacy -9.97*** -12.44, -7.44 -0.24*** -0.30, -0.18 4.31** 1.84, 6.86 0.12** 0.05, 0.18
Model R2 (explained variability) .57 (p < .001) .45 (p < .001)
Note. Intershift recovery was modeled as a mediator (intermediary) variable in the model for chronic fatigue explained by acute fatigue. b = unstandardized re-
gression coefficient; β = standardized regression coefficient; ED = emergency department; NA = not applicable; RN = registered nurse.
a Morning-day shifts: 7AM–7PM, 7AM–3PM, 8AM–5PM; day-evening shifts: 11AM–7PM, 3PM–11PM; night-morning shifts: 7PM–7AM, 11PM–7AM.

*p < .05. **p < .01. ***p < .001.

grams. Special emphasis should be provided to younger nurses who with staffing and resource adequacy being the strongest predic-
might be new to the organization and unfamiliar with the avail- tors. This result is aligned with the results of the study by Aiken
able wellness programs. et al. (2023), in which adequate staffing was associated with nurses’
All three work environment variables were significantly well-being. A similar result was noted in our earlier work eval-
associated with IR in both the bivariate and regression analyses, uating nurse fatigue and sleep (Farag et al., 2022). The signifi-

Volume 15/Issue 3 October 2024 www.journalofnursingregulation.com 29

 FIGURE 1 ing number of patients who require around-the-clock care, having
full-time nurses scheduled on rotating and night shifts will likely
The Study Model
not change.
Intershift Recovery Years of RN experience was another occupational variable
) -0. associated with IR (i.e., nurses with more years of experience had
.37 31
0, -0 (-0
.37 better IR). This result is aligned with earlier studies showing low
.5
(-0 , -0
fatigue among nurses with more years of experience (Farag et al.,
44 .25
-0. )
0.22 (0.16, 0.28)
2022). This result could be because seasoned nurses have devel-
Acute Fatigue Chronic Fatigue oped better fatigue management and IR measures. It might also
be that more experienced nurses are managing fewer patients or
Personal factors Occupational factors taking fewer night shifts; however, further analysis is needed to
⦁ Exercise ⦁ Unit type
support this explanation.
⦁ Daytime sleepiness ⦁ Site
⦁ Coffe during work ⦁ Work shift
New to the literature is the relationship between work site
⦁ Age ⦁ Years as RN and IR. According to our results, nurses working in our only mag-
⦁ Commute time ⦁ Years in unit net academic medical center had lower IR compared with nurses
⦁ Sleep quality ⦁ Working hours working at other sites. This result is not aligned with an earlier
⦁ Leadership support
study in which nurses working in a magnet hospital reported bet-
⦁ Nurse-physician relationship
ter IR compared with nurses working in a non-magnet hospital
⦁ Staffing/resource adequacy
Other types of fatigue
(Chen et al., 2014). Our finding may be attributed to the fact that
⦁ Physical fatigure ⦁ Mental fatigue ⦁ Total fatigue the participating magnet hospital is the state’s only Level 1 trauma
center and thus receives patients with complex medical conditions
Path coefficients and 95% confidence intervals are standardized
from all over the state. As a result, patient acuity and nurse work-
estimates, Path coefficients for the covariates are reported in Ta-
ble 4. load could be higher at the magnet medical center than the patient
acuity and workload at other sites. Nurses working in the mag-
Note: Standardized estimates of path coefficients and 95% confidence inter-
net medical centers could be more fatigued and do not get suffi-
vals are shown on the arrowed lines. RN = registered nurse.
cient IR. The results of prior work (Ismail et al., 2019; Han et al.,
2014) support the relationship between workload/demands, nurse
TABLE 5 fatigue, and IR.
Model-Estimated Effects of Acute Fatigue
The Mediating Effect of IR
on Chronic Fatigue: Total, Direct, and
Indirect Effects via Intershift Recovery Consistent with the guiding framework (Winwood et al., 2005;
(N = 993) Winwood et al., 2006), our results supported the mediating effect
of IR between AF and CF. This result is consistent with earlier
Unstandardized Standardized
work and validates the importance of IR as a mechanism (inter-
Effect b 95% CI β 95% CI
Total effect of acute 0.41*** 0.35, 0.48 0.35*** 0.30, 0.41
mediary variable) to prevent AF from escalating to CF (Winwood
fatigue on chronic fatigue et al., 2005; Winwood et al., 2006; Barker & Nussbaum, 2011).
Direct effect 0.26*** 0.18, 0.32 0.22*** 0.16, 0.28 Additionally, this finding is aligned with the conservation of
Indirect effect 0.16*** 0.12, 0.20 0.14*** 0.10, 0.17 resources theory (Hobfoll, 1989), which posits that individuals
Note. b = unstandardized regression coefficient; β = standardized regression strive to retain, protect, and build resources that they centrally
coefficient. value. In our case, IR can be conceptualized as a resource; with-
***p < .001. out it, AF could escalate to CF. According to the resource cara-
vans, another component of the conservation of resources theory,
cant relationship between work environment and IR is aligned resources tend to be bundled together, and improving resources in
with the plethora of studies underscoring the significant impact of one area could improve resources in another area (Hobfoll, 1989).
work environment on multiple nurse, patient, and organizational From our results, we can argue that other resources such as work
outcomes. Nurses working part-time or as needed on shifts that environment and some personal factors, such as sleep and exercise,
do not involve rotation or night work hours had better IR than could strengthen nurses’ IR. With clinician burnout being at the
their full-time colleagues and those who worked night-morning front and center of healthcare administrators’ agenda, it is vital to
or rotating shifts. This result is expected because part-time work develop integrated and comprehensive workplace policies, such as
and as-needed, nonrotation, non-night scheduling arrangements investing in creative staffing and scheduling solutions that allow
allow adequate time to rest and recover before returning to work. nurses to adequately recover between shifts. Such strategies may
Unfortunately, with the current nursing shortage and an increas- be fruitful to prevent CF and, potentially, burnout.

30 Journal of Nursing Regulation

Limitations References
This study used a cross-sectional design, so we were not able to Ahsberg, E. (2000). Dimensions of fatigue in different working populations.
Scandinavian Journal of Psychology, 41(3), 231–241. https://doi.
track fatigue patterns and changes in IR over time. The data were org/10.1111/1467-9450.00192
collected using standardized self-reported surveys, so response bias Aiken, L. H., Lasater, K. B., Sloane, D. M., Pogue, C. A., Fitzpatrick Rosen-
is a risk associated with this data collection approach. However, baum, K. E., Muir, K. J., McHugh, M. D., & US Clinician Wellbeing
using the same surveys that were used in prior research allowed Study Consortium. (2023). Physician and nurse well-being and pre-
ferred interventions to address burnout in hospital practice: Factors
the research team to compare the results with prior work. The associated with turnover, outcomes, and patient safety. JAMA Health
study was conducted using a convenience sample of nurses from Forum. 4(7), e231809. https://doi.org/10.1001/
eight hospitals. Therefore, the results may not be generalizable to jamahealthforum.2023.1809
other hospitals. Despite these limitations, this study is among the Aiken, L. H., & Patrician, P. A. (2000). Measuring organizational traits of
few studies to test the mediating effect of IR and to evaluate vari- hospitals: The revised nursing work index. Nursing Research, 49(3), 146–
153. https://doi.org/10.1097/00006199-200005000-00006
ables associated with IR. The large sample size is also a strength
Alsayed, S. A., Abou Hashish, E. A., & Alshammari, F. (2022). Occupational
of the study. Finally, we recognize that we conducted our study fatigue and associated factors among Saudi nurses working 8-hour shifts
prior to the pandemic. Given that nurse fatigue has been further at public hospitals. SAGE Open Nursing, 8, Article
exacerbated by the pandemic and continues to be an issue because 23779608221078158. https://doi.org/10.1177/23779608221078158
of the nursing shortage, it remains crucial to persistently inves- American Nurses Association. (2014). Position statement: Addressing nurse
fatigue to promote safety and health: Joint responsibilities of registered nurses and
tigate methods to promote IR in order to mitigate nurse fatigue employers to reduce risks. Washington, DC.
and burnout. Bai, G., Korfage, I. J., Mautner, E., & Raat, H. (2019). Associations between
maternal health-related quality of life during pregnancy and birth out-
comes: The generation R study. International Journal of Environmental
Conclusion Research and Public Health, 16(21), Article 4243. https://doi.org/
10.3390/ijerph16214243
IR is an important resource; it is an intermidary variable that Barker, L. M., & Nussbaum, M. A. (2011). Fatigue, performance and the
mitigates the progression of AF to CF. Guided by our results, it work environment: A survey of registered nurses. Journal of Advanced
is pragmatic for healthcare leaders to implement strategies that Nursing, 67(6), 1370–1382. https://doi.org/10.1111/
focus on modifiable occupational and personal variables to sup- j.1365-2648.2010.05597.x
port adequate IR, with special attention paid to large academic Barker Steege, L. M., & Nussbaum, M. A. (2013). Dimensions of fatigue as
predictors of performance: A structural equation modeling approach
medical centers. Although these centers might have more or better among registered nurses. IIE Transactions on Occupational Ergonomics and
resources than community hospitals, nurses working at these sites Human Factors, 1(1), 16–30. https://doi.org/10.1080/21577323.2011.6
are caring for a sicker and more complex patient population, which 37153
might subject the nurses to more fatigue and require additional Buysse, D. J., Reynolds, C. F. 3rd, Monk, T. H., Berman, S. R., & Kupfer, D.
J. (1989). Pittsburgh sleep quality index: A new instrument for psychi-
measures to ensure adequate IR. With the ongoing nursing short- atric practice and research. Psychiatry Research, 28(2), 193–213. https://
age, maintaining adequate staffing might be a challenge; there- doi.org/10.1016/0165-1781(89)90047-4
fore, decision makers and nurse leaders should focus on improving Caboral-Stevens, M., Raymond, D., 3rd, & Evangelista, L. S. (2023). Well-
nurses’ working conditions and use a supportive leadership style. being, occupational fatigue, and sleep quality among Filipino nurses
working during COVID-19. Asian Journal of Nursing Education and
Nurse leaders could also invest in creative care delivery mod- Research, 13(1), 67–72. https://doi.org/10.52711/2349-
els and work policies that allow for adequate staffing while pro- 2996.2023.00016
viding nurses with adequate time for IR. At an individual level, Caruso, C. C., Baldwin, C. M., Berger, A., Chasens, E. R., Landis, C., Rede-
nurses should follow good sleep hygiene measures and exercise. ker, N. S., Scott, L. D., & Trinkoff, A. (2017). Position statement:
The advancement in wearable technology that is currently avail- Reducing fatigue associated with sleep deficiency and work hours in
nurses. Nursing Outlook, 65(6), 766–768. https://doi.org/10.1016/j.
able can be used to help nurses organize and track their sleep, exer- outlook.2017.10.011
cise, and caffeine consumption. Maintaining nurses’ well-being is Chen, J., Davis, K. G., Daraiseh, N. M., Pan, W., & Davis, L. S. (2014).
a shared responsibility between nurses and their employers (ANA, Fatigue and recovery in 12-hour dayshift hospital nurses. Journal of
2014). Therefore, nurses should take necessary measures to main- Nursing Management, 22(5), 593–603. https://doi.org/10.1111/
jonm.12062
tain adequate IR. Hospital leaders should provide a healthy work
Cho, H., & Steege, L. M. (2021). Nurse fatigue and nurse, patient safety, and
environment for their nurses and should review their policies to organizational outcomes: A systematic review. Western Journal of Nursing
ensure that nurse schedules account for adequate IR. Hospitals Research, 43(12), 1157–1168. https://doi.org/10.1177/
should be incentivized to incorporate wellness programs as part of 0193945921990892
their human resource services. Finally, measures for nurses’ IR and Cohen, J. (1988). Statistical power analysis for the behavioral science (2nd ed).
Lawrence Erlbaum Associates.
well-being should be incorporated as part of nursing and residency
Drake, C., Roehrs, T., Shambrom, J., & Roth, T. (2013). Caffeine effects on
program curricula. sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep
Medicine, 9(11), 1195–1200. https://doi.org/10.5664/jcsm.3170

Volume 15/Issue 3 October 2024 www.journalofnursingregulation.com 31

Farag, A., Scott, L. D., Perkhounkova, Y., Saeidzadeh, S., & Hein, M. (2022). Ross, A., Geiger-Brown, J., Yang, L., Flynn, S., Cox, R., Wehrlen, L., & Lee,
A human factors approach to evaluate predicators of acute care nurse L. J. (2021). Acute and chronic fatigue in nurses providing direct
occupational fatigue. Applied Ergonomics, 100, Article 103647. https:// patient care and in non-direct care roles: A cross-sectional analysis.
doi.org/10.1016/j.apergo.2021.103647 Nursing & Health Sciences, 23(3), 628–638. https://doi.org/10.1111/
Geiger-Brown, J., Rogers, V. E., Trinkoff, A. M., Kane, R. L., Bausell, R. B., nhs.12862
& Scharf, S. M. (2012). Sleep, sleepiness, fatigue, and performance of Sagherian, K., Steege, L. M., Cobb, S. J., & Cho, H. (2023). Insomnia, fatigue
12-hour-shift nurses. Chronobioliology International, 29(2), 211–219. and psychosocial well-being during COVID-19 pandemic: A cross-sec-
https://doi.org/10.3109/07420528.2011.645752 tional survey of hospital nursing staff in the United States. Journal of
Geiser, C. (2013). Methodology in the social sciences: Data analysis with Mplus. Clinical Nursing, 32(15–16), 5382–5395. https://doi.org/10.1111/
Guilford Press. jocn.15566
Gifkins, J., Johnson, A., Loudoun, R., & Troth, A. (2020). Fatigue and recov- Scott, L. D., Arslanian-Engoren, C., & Engoren, M. C. (2014) Association of
ery in shiftworking nurses: A scoping literature review. International sleep and fatigue with decision regret among critical care nurses. Ameri-
Journal of Nursing Studies, 112 (2020), Article 103710. https://doi. can Journal of Critical Care, 23(1), 13–23. https://doi.org/10.4037/
org/10.1016/j.ijnurstu.2020.103710 ajcc2014191
Han, K., Trinkoff, A. M., & Geiger-Brown, J. (2014). Factors associated with Scott, L. D., Hwang, W. T., Rogers, A. E., Nysse, T., Dean, G. E., & Dinges,
work-related fatigue and recovery in hospital nurses working 12-hour D. F. (2007). The relationship between nurse work schedules, sleep
shifts. Workplace Health & Safety, 62(10), 409–414. https://doi. duration, and drowsy driving. Sleep, 30(12), 1801–1807. https://doi.
org/10.3928/21650799-20140826-01 org/10.1093/sleep/30.12.1801
Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptu- Slater, P., & McCormack, B. (2007). An exploration of the factor structure of
alizing stress. The American Psychologist, 44(3), 513–524. https://doi. the nursing work index. Worldviews on Evidence-Based Nursing, 4(1),
org/10.1037//0003-066x.44.3.513 30–39. https://doi.org/10.1111/j.1741-6787.2007.00076.x
Ismail, K. M., Malak, M. Z., & Alamer, R. M. (2019). Psychosocial correlates Stanulewicz, N., Knox, E., Narayanasamy, M., Shivji, N., Khunti, K., &
of work-related fatigue among Jordanian emergency department nurses. Blake, H. (2019). Effectiveness of lifestyle health promotion interven-
Perspectives in Psychiatric Care, 55(3), 486–493. https://doi.org/10.1111/ tions for nurses: A systematic review. International Journal of Environmen-
ppc.12354 tal Research and Public Health, 17(1), 17. https://doi.org/10.3390/
ijerph17010017
James, J. E., & Rogers, P. J. (2005). Effects of caffeine on performance and
mood: Withdrawal reversal is the most plausible explanation. Psycho- Watterson, T. L., Steege, L. M., Mott, D. A., Ford, J. H., 2nd, Portillo, E. C.,
pharmacology, 182(1), 1–8. https://doi.org/10.1007/s00213-005-0084-6 & Chui, M. A. (2023). Sociotechnical work system approach to occupa-
tional fatigue. Joint Commission Journal on Quality and Patient Safety,
Johns, M. W. (1991). A new method for measuring daytime sleepiness: The 49(9), 485–493. https://doi.org/10.1016/j.jcjq.2023.05.007
Epworth sleepiness scale. Sleep, 14(6), 540–545. https://doi.
org/10.1093/sleep/14.6.540 Winwood, P. C., Lushington, K., & Winefield, A. H. (2006). Further devel-
opment and validation of the occupational fatigue exhaustion recovery
Kline, R. B. (2016) Principles and practice of structural equation modeling: Meth- (OFER) scale. Journal of Occupational and Environmental Medicine, 8(4),
odology in the social sciences (4th ed.). Guilford Press. 381–389. https://doi.org/10.1097/01.jom.0000194164.14081.06
Lake, E. T. (2002). Development of the practice environment scale of the Winwood, P. C., Winefield, A. H., Dawson, D., & Lushington, K. (2005).
Nursing Work Index. Research in Nursing and Health, 25(3), 176–188. Development and validation of a scale to measure work-related fatigue
https://doi.org/10.1002/nur.10032 and recovery: The occupational fatigue ehaustion/recovery scale (OFER).
Maslach, C., Schaufeli, W., & Leiter, M. (2001). Job burnout. Annual Review Journal of Occupational and Environmental Medicine, 47(6), 594–606.
of Psychology, 52, 397–422. https://doi.org/10.1097/01.jom.0000161740.71049.c4
Muthén, B. O., Muthén, L. K., & Asparouhov, T. (2016). Regression and media- Yamaguchi, S., Winwood, P. C., & Yano, R. (2023). Chronic fatigue and
tion analysis using Mplus. Muthén & Muthen. recovery among nurses working two-shift and three-shift rotations. Col-
Pasupathy, K., & Barker Steege, L. M. (2012). Impact of fatigue on perfor- legian, 30(6), 786–794. https://doi.org/10.1016/j.colegn.2023.06.001
mance in registered nurses: Data mining and implications for practice. Yu, F., Somerville, D., & King, A. (2019). Exploring the impact of 12-hour
Journal for Healthcare Quality, 34(5), 22–30. https://doi. shifts on nurse fatigue in intensive care units. Applied Nursing Research,
org/10.1111/j.1945-1474.2011.00157.x 50, Article 151191. https://doi.org/10.1016/j.apnr.2019.151191
Reith, T. P. (2018). Burnout in United States healthcare professionals: A nar-
rative review. Cureus, 10(12), Article e3681. https://doi.org/10.7759/
cureus.3681 Amany Farag, PhD, RN, is an Associate Professor, University of
Ricci, J. A., Chee, E., Lorandeau, A. L., & Berger, J. (2007). Fatigue in the Iowa College of Nursing, Iowa City. Linda Scott, PhD, RN, FAAN,
U.S. workforce: Prevalence and implications for lost productive work is a Professor and Dean, University of Wisconsin Madison College of
time. Journal of Occupational and Environmental Medicine, 49(1), 1–10. Nursing. Yelena Perkhounkova, PhD, is a Statistician, University
https://doi.org/10.1097/01.jom.0000249782.60321.2a
of Iowa College of Nursing. Peter James Abad, MSc, RN, is a doc-
Rogers, A. E. (2008). The effects of fatigue and sleepiness on nurse perfor-
mance and patient safety. Patient safety and quality: An evidence-based toral student, University of Iowa College of Nursing, and a faculty
handbook for nurses (R. G. Hughes, Ed.). Agency for Healthcare Research member at the University of the Philippines Manila College of
and Quality. Nursing. Maria Hein, MSW, is a Data Manager, University of Iowa
Rogers, A. E., Hwang, W. T., & Scott, L. D. (2004). The effects of work College of Nursing.
breaks on staff nurse performance. The Journal of Nursing Administration,
34(11), 512–519. https://doi.org/10.1097/00005110-200411000- Corresponding author: Amany Farag, Amany-farag@uiowa.edu.
00007
This study was funded by a grant from the National Council of State
Boards of Nursing (NCSBN), grant number R91013.

32 Journal of Nursing Regulation