A Scoping Review of Psychological Distress Outcomes in NICU Healthcare Providers

Highlights

Public health relevance—How does this work relate to a public health issue?

• Advances in neonatal medicine have increased the survival of critically ill and extremely premature infants, resulting in NICU healthcare providers being increasingly exposed to emotionally, psychologically, ethically, and morally challenging clinical situations that may affect their physical and mental health.
• NICU healthcare provider well-being is a critical public health concern because psychological distress can influence workforce functioning, staff retention and engagement, and the quality of care delivered to infants and families, ultimately affecting patient and family health outcomes.

Public health significance—Why is this work of significance to public health?

• This review synthesizes evidence from 76 quantitative publications across 34 countries and provides the first examination of multiple forms of psychological distress among NICU healthcare providers, including burnout, traumatic stress, moral distress, mental health, stress and physiological symptoms.
• Findings suggest that working in the NICU is, on average, associated with low to moderate levels of psychological distress across professions and elevated distress in some providers; however, substantial variability within studies and heterogeneity across studies make it difficult to draw firm conclusions about the prevalence and correlates of psychological distress in NICU healthcare providers. Improved research methodologies are needed.

Public health implications—What are the key implications or messages for practitioners, policy makers and/or researchers in public health?

• Future research should examine the complex interplay of multiple distress constructs as well as individual, biological, environmental, organizational and social determinants of health in NICU providers using longitudinal approaches and standardized measures and methods to identify potential risk and protective factors to guide targeted prevention and intervention efforts.
• Hospitals, healthcare organizations and policy makers should prioritize routine assessment of NICU healthcare provider well-being and implement evidence-informed interventions that address modifiable workplace stressors, foster supportive work environments, improve individual stress management skills and promote workforce sustainability.

Abstract

The Neonatal Intensive Care Unit (NICU) is a demanding work environment that provides medical care for acutely ill and premature infants while supporting parents experiencing significant stress and psychological distress. This scoping review is the first to examine quantitative research across a broad range of psychological constructs, including burnout, traumatic stress, moral distress, and other distress outcomes such as mental health, somatic symptoms and physiological indicators, among NICU healthcare providers. Following PRISMA-ScR guidelines, PsycINFO, Ovid via MEDLINE, Embase, and CINAHL were searched. Two searches (2021 and 2024) identified 1889 unique records, with 76 publications from 34 countries meeting inclusion criteria (i.e., original quantitative research assessing work-related psychological consequences with NICU healthcare providers) and exclusion criteria (i.e., intervention studies; qualitative studies; NICU provider data inseparable from aggregate results). Most participants were White, female NICU nurses. Across studies, psychological distress was most often reported in the low-to-moderate range, with considerable variability within and across studies making it difficult to clearly summarize the average levels of distress and associations that best predict psychological distress outcomes. Consistent differences by healthcare profession or sociodemographic characteristics were not identified, largely because these factors were infrequently assessed, or findings were inconsistent. Environmental factors were more frequently assessed and reported to be associated with psychological distress in NICU providers. The current evidence base is constrained by cross-sectional designs, inconsistent practices, and conceptual overlap across psychological constructs. Future research should prioritize longitudinal designs, standardized methodologies, and more representative samples across professions. In the interim, unit-level assessment and targeted interventions may help address modifiable work-related stressors, individual responses to stressors, and support NICU healthcare providers’ well-being.

1. Introduction

… there is nothing morally neutral about a NICU. NICUs are horrible, wonderful places. They are the best and the worst of pediatrics. They save many lives and they cause much pain and suffering

[1].

The Neonatal Intensive Care Unit (NICU) is a complex and physically and psychologically demanding work environment where healthcare providers care for fragile and acutely ill infants while also providing emotional support for significantly stressed parents/caregivers [2,3]. Providers function in a liminal space that hovers between the beginning of life and possible end-of-life care, where they are often challenged with determining if, when, how, and how much to intervene. At the same time, they must provide information, reassurance, and guidance to parents/caregivers in the face of uncertain outcomes. The continuous complex interplay of clinical and emotional care practices, combined with repeated exposure to trauma, loss, and parental distress as well as ethical dilemmas and conflicts, highlights the significant psychological demands of working in the NICU setting.

Given the notable exposure to occupational stress, the National Academy of Medicine published a discussion paper emphasizing the high prevalence of provider burnout and its impacts on the quality, safety, and performance in healthcare systems [4]. In addition to burnout, other responses to workplace stressors, such as moral distress, traumatic stress, and mental health symptoms, may be present, negatively impacting healthcare providers. While many studies have examined occupational stress in healthcare, relatively few address the broad range of related constructs within NICU settings, and the overall state of the literature across multidisciplinary NICU providers is unclear.

1.1. Occupational Stress and Distress Constructs: Definitions and Challenges

There is a long and extensive literature evaluating stress that emphasizes the distinction between stressors and stress responses. Wheaton and Montazer [5] define stressors as “conditions of threat, challenge, demands, or structural constraints that, by the very fact of their occurrence or existence, call into question the operating integrity of the organism” (p. 173). Stressors can range from single discrete events, such as traumatic incidents (e.g., traumatic birth) or life-changing events (e.g., changing employment) to moderately discrete events (e.g., daily hassles such as commuting or interpersonal problems at work), to more continuous events and circumstances (e.g., chronic medical or mental health challenges) and ongoing traumatic stressors (e.g., child abuse or harassment at work). Some argue that these different types of stressors may have unique effects on psychological distress and mental health outcomes [5] and thus, “proper assessment of the total impact of stress requires consideration of multiple sources of stress.”

Stress, itself, is the potential response to a stressor. Using a concept analysis, stress was defined as “a non-specific biological response to a demand or stressor that is not necessarily harmful to the individual” (p. 539, [6]), while distress was harmful. Furthermore, the term psychological distress was defined as “the unique discomforting, emotional state experienced by an individual in response to a specific stressor or demand that results in harm, either temporary or permanent, to the person” (p. 539, [6]). Most often, these terms are not well defined or differentiated, and many other terms, such as burnout, compassion fatigue, secondary traumatic stress and moral distress, have been developed in attempts to distinguish and measure a variety of biological and psychological responses to stressors in the workplace.

Some of the most systematic work on stress in the context of the workplace was conducted by Maslach and colleagues, who defined burnout as “a psychological syndrome of exhaustion, cynicism, and inefficacy, which is experienced in response to chronic job stressors” (p. 93, [7]) and developed the Maslach Burnout Inventory (MBI) to measure the three components [8,9]. In 2019, the World Health Organization included burnout in the International Classification of Diseases [10] as an “occupational phenomenon” (not a mental health disorder) for which people may seek and benefit from professional assessment and intervention.

Despite the common use of the term burnout in scientific literature and now popular culture, some researchers question whether the symptoms of burnout are distinct from Major Depressive Disorder given the high correlations among symptom measures [11]. A recent literature search identified 88 “unique definitions of burnout” [12] and attempted to reach a consensus, with the majority of experts agreeing on a definition of burnout that involved physical and emotional exhaustion in response to work stressors; however, not all agreed and the construct continues to be redefined and debated [11,13,14].

Compassion fatigue is another common term used to describe a response to work stressors, particularly in healthcare professionals [15]. Stamm [16] considered compassion fatigue largely synonymous with secondary traumatic stress and related to vicarious trauma, which at that time was witnessing traumatic events or hearing others talk about their traumatic events rather than the direct experiencing of the traumatic events themselves (i.e., primary traumatic stress). Since the Diagnostic and Statistical Manual of Mental Disorders [17] expanded the definition of a traumatic event (Criterion A) to include witnessing and repeated exposure to hearing about traumatic events as part of one’s job, there has been a recognition of multiple pathways of exposure to possible traumatic events. It is now clearly recognized that traumatic events that happen at work, including those witnessed happening to others and heard about through our work, can contribute to traumatic stress symptoms and even a diagnosis of posttraumatic stress disorder (PTSD).

Moral distress is another major construct that has been widely studied in healthcare, including pediatric and NICU settings, alongside related concepts such as moral injury and ethical confrontation [18,19]. Originally defined as distress arising when individuals know the right action to take but are constrained from taking it [20], contemporary definitions are broader. Moral distress is now commonly conceptualized as psychological distress causally related to a morally challenging situation or moral stressor/dilemma, including experiences of moral constraint, conflict, uncertainty, or tension [21,22].

Common definitions for stress and burnout constructs integrate the context of the stressor (i.e., workplace) and the stress response (e.g., fatigue), thus making it difficult to distinguish the stressor from the frequency and intensity of the stress response. Another challenge is the lack of clarity in the current literature as to whether the psychological distress responses discussed as unique in the literature are truly distinct constructs. Prior reviews on NICU providers have been published separately for burnout [23] and moral distress [19,24]; however, no reviews have examined the extant literature across multiple constructs. This scoping review is the first to examine the quantitative research across a broad range of psychological distress constructs that have been studied in NICU providers.

1.2. Current Scoping Review

The purpose of the current scoping review is to explore the state of the quantitative research literature regarding the nature and frequency of psychological distress outcomes that have been studied in NICU healthcare providers. We aim to clarify the main concepts and identify gaps in research that are needed to advance our understanding. Specifically, a scoping review of the current knowledge base can help determine the extent of NICU staff psychological distress across constructs, how best to measure and interpret these constructs, what factors may increase or decrease distress, and how to advance future research in hopes that we may be better able to intervene, prevent and mitigate these difficulties in healthcare professionals caring for some of our most vulnerable patients and families.

2. Materials and Methods

We conducted this scoping review in accordance with the PRISMA-ScR guidelines [25] (Supplementary File S1) to identify publications that would address the following question: “What is the state of the literature regarding work-related psychological distress responses in NICU healthcare providers?” To accomplish this goal, all authors worked together to identify a discrete set of relevant search terms. We used the librarian resources at Henry Ford Health Sladen Library to conduct a targeted search across the following databases: PsycInfo, Ovid via MEDLINE, Embase, and CINAHL. Search strings by database are included in Supplementary Table S1. We restricted our search to articles that were peer-reviewed (i.e., no gray literature sources), written in English, and published in the year 1960 or later.

Procedures

For the initial search (2021), the ten co-authors were paired into dyads and assigned a subset of ~200 publications to screen. We used dummy coding (i.e., 0 = no; 1 = yes) to record our assessment of the following eligibility criteria for each publication: (1) Is this original research? (2) Was data collected from healthcare providers in a NICU setting? (3) Were work-related psychological consequences assessed? Each member of the dyad coded their assigned articles independently and then met with their partnered co-author to resolve any disagreements. If a pair was unable to decide based on partnered consensus, the publication was flagged for full group review. Additional data columns were included in the database for reviewers to record any notes, whether full manuscript review was required to determine eligibility, and whether articles not meeting eligibility for inclusion should be considered for the introduction and/or discussion sections. We conducted 16 group consensus meetings to finalize our title and abstract screening phase in the first search. Full-text reviews and data extraction for the first search were completed by all co-authors, with dyads reviewing a subset of articles in a manner similar to the screening procedures. The first authors (PAG, ALN) reviewed all the titles, abstracts and full texts resulting from the second search (in 2024) in a similar manner as described above. Specifically, we independently screened each record, with disagreements resolved through discussion, and then divided the texts by constructs for data extraction (ALN: Burnout, Traumatic Stress; PAG: Moral Distress, Other Distress outcomes). Extraction entries were reviewed and verified by the other author when any concern regarding clarity or accuracy was identified, with coding discrepancies resolved through discussion. No inter-reviewer agreement statistics were calculated.

Across both searches, we excluded 1688 of the 1889 articles in the screening phase, leaving a total of 201 articles eligible for inclusion. We conducted full-text reviews to confirm inclusion criteria and conduct data extraction. Full-text review excluded an additional 125 publications, including those that reported intervention studies, qualitative-only studies, studies in which NICU provider data could not be separated from other populations, studies that did not assess a psychological construct of interest, and studies without primary research. The final review included 76 articles. Once inclusion was confirmed, we extracted the following information: Sample characteristics (e.g., sex, race, and country); number and type of healthcare provider(s) included in the study (e.g., nurses, attending physicians, and nurse practitioners); study setting (e.g., NICU, Cardiac Intensive Care Unit (CICU)); psychological construct(s) assessed (e.g., burnout, trauma, moral distress); measure(s) used to assess these constructs; and results pertaining to constructs (i.e., levels of distress, associations).

All parts of this process were recorded and tracked in our Excel file to ensure accurate completion of our PRISMA flow diagram (Figure 1).

3. Results

Our initial search was conducted on 25 August 2021 and returned 1609 abstracts (1205 after duplicates were removed). We conducted a second search on 15 October 2024 to capture articles published since our initial review, as notable time had passed while determining how to consolidate and organize this large and diverse literature. We ran the identical search terms through the same library service. This second search returned 762 additional abstracts (684 after duplicates were removed). Across the two searches, we screened a total of 1889 unique abstracts for inclusion, with 201 publications retained for full-text review (see Figure 1). Of these, 76 publications met the inclusion criteria and were included in this review. These 76 publications represented 70 independent samples; four sets of publications used or appeared to use the same underlying datasets: (1) [26,27]; (2) [28,29]; (3) [30,31]; and (4) [32,33,34,35]. Because each publication reported distinct findings relevant to the review, findings from all 76 publications are retained and presented in our results table (i.e., Supplementary Table S2).

No relevant publications were identified prior to 1984, and our search revealed a spike in the number of relevant publications beginning in the 2000s (see Figure 2). Samples from seven major locations in the world, representing 34 distinct countries, are represented in this review (see Figure 3). Most studies were conducted in the United States (USA; 39%), followed by European countries (18%).

3.1. Study Samples

The samples from all but one publication included NICU nurses (n = 75; 98.7%), with 45 publications including nurses exclusively (59.2%), and 30 publications including nurses and other NICU professionals (39.5%). One study sampled only neonatologists [36]. The sample size of the included publications ranged from 7 to 2073 NICU providers (see Figure 4). Most publications included samples with 100 or fewer providers (n = 25) or 101–200 providers (n = 21). Although 76 distinct publications are included in this review, four data sets were used to generate multiple publications (see Figure 4 footnotes). For example, one large USA data set that included approximately 2000 NICU providers in California was used to generate four publications [32,33,34,35].

Complete and detailed sociodemographic factors of the samples were not reported in many publications. The 56 that did report on sex indicated that their sample was largely comprising females/women (i.e., 32 publications specified 95–100% female/women, 11 specified 88–92%, 10 specified 71–85%, and three specified 60% or less). Distribution by sex was not reported in 20 publications. Sixty publications (78.9%) did not report the racial distribution of their sample. Of the 16 that did, the majority were White/Caucasian: nine reported that 100% of their sample was White/Caucasian; five reported that between 87 and 95% of their sample was White/Caucasian, and two reported that 65% of their sample was White/Caucasian.

3.2. Study Designs

Of the 76 publications, 74 used a cross-sectional survey design to obtain data on psychological distress. The remaining two used a repeated measures design from a single-item measure of moral distress [37,38]. We identified which psychological distress constructs were evaluated in each publication and classified them into four construct categories: “burnout,” “moral distress,” “traumatic stress”, and “other distress” outcomes. The traumatic stress category includes publications that labeled their psychological distress construct as either “secondary traumatic stress” (n = 8) or “traumatic stress” (n = 3). The “other distress” category includes publications that measured such things as self-reported mental health symptoms (e.g., depression), somatic symptoms (e.g., fatigue), and perceived stress, as well as physiologic indicators (e.g., cortisol level). Figure 5 illustrates the number of publications that reported data within construct categories, including the number of publications that assessed more than one construct.

Most publications reported use of validated measures, although there was variability across studies as to whether the measures were used in their entirety and as intended. Seven publications used unvalidated questionnaires and were included to ensure the complete scope of the literature was represented [36,39,40,41,42,43,44]. Three publications measured cortisol levels as distress indicators [45,46,47]. The specific psychological distress measures that were used to assess constructs as well as the constructs themselves are listed for each publication in

Table 1. General Summary of Results by Construct.

Construct	Number of Publications	Samples a	Measures b	Summary of Distress Levels c,d	Significant Associations	Limitations
Burnout	38 (50%) with 35 independent samples	N = 8481; 21 publications (55.3%) included only nurses (n = 4189) & 17 (44.7%) included nurses & other professions (n = 4283).	BM (n = 2); CBI (n = 1); MBI (n = 25); OBI (n = 1); ProQOL (n = 7); TS (n = 1); unvalidated scale (n = 1).	MBI: Mean EE (scale 0–54) scores ranged from 14.4 (SD = 7.91) to 26.44 (SD = 13.55); mean DP (score 0–30) scores ranged from 3.37 (SD = 3.31) to 7.66 (SD = 7.18); mean PA (scale 0–48) scores ranged from 27.35 (SD = 7.39) to 38.4 (SD not reported); 5% burnout estimate (n = 1). ProQOL: means in the low (n = 4) to moderate (n = 1) ranges.	Adverse events; workload; high job demands; low resources; greater overtime; staffing shortages; high admissions & occupancy levels; poor work environment; considering leaving position; poor teamwork & safety climate; poor communication; lack of support; nurse leadership styles; bereavement training; age (younger & older); less work experience; race; religious affiliation or spirituality	Only nine publications reported the means for the 3 MBI scales; Only one publication reported the rate of burnout as recommended by Maslach & colleagues.
Traumatic Stress	11 (14.5%) with 11 independent samples	N = 1983; 6 publications (55%) included only nurses (n = 835) &5 (45%) included nurses & other professions (n = 1148)	CFS (n = 1); IES (n = 1); IES-R (n = 1); PC-PTSD (n = 1); ProQOL (n = 5); STSS (n = 3).	ProQOL: Means in the low (n = 4) to moderate (n = 1) ranges. PC-PTSD: 13.2% reported clinical levels. STSS: Means in low (n = 2) to moderate (n = 1) ranges.	Adverse events; less vigor; less self-compassion; poorer nurse-infant/family relationships; lower reported nurse–physician collegiality; communication training; bereavement training; pandemic stress; older age; greater experience	Traumatic stress symptoms were not associated with a specific event; no evaluation of trauma history; no diagnostic assessments.
Moral Distress	20 (26.3%) with 20 independent samples	N = 5208; 10 publications (50%) included only nurses (n = 1360) & 10 (50%) included nurses & other professions (n = 3848)	MDS (n = 5), MDS “not in the patient’s best interest” subscale (n = 1); MDS-R (n = 9); MMD-HP (n = 1); MDT (n = 2); unvalidated scale (n = 3).	MDS: Mean frequency & intensity “average” (n = 2), mean frequency “average” & intensity “high” (n = 1), mean frequency & intensity “moderate” (n = 1). MDS-R: Means were in the “low” (n = 2), “low-moderate” (n = 2), “moderate” (n = 2) & “high” ranges (n = 1) MMD-HP: Mean in “low to moderate” range (n = 1).	Higher acuity patients; nurse uncertainty about benefit of care; nurse uncertainty about child’s prognosis; perceptions of futile care; the wish to diminish treatment; end-of-life care & communication; perception of “parents wanting too much”; critical reflective practice in NICU; considering leaving position; lower psychological empowerment; less work engagement; lower perceived hospital supportiveness; less favorable hospital ethical climate; religiosity; increasing years of ICU experience; more work experience; younger age.	Inconsistent reporting of scales & statistics across the publications that used the MDS, MDS-R & MMD-HP.
Other Distress	31 (40.8%) with 29 independent samples	N = 4759; 21 publications (68%) included only nurses (n = 2317), 9 (29%) included nurses & other professions (n = 1721), & 1 (3%) included only neonatologists (n = 721)	Mental health & somatic symptoms: BDI (n = 1); CES-D (n = 1); CIS (n = 1); CFSI (n = 1); EDI (n = 1); GAD-7 (n = 2); GHQ (n = 2); HADS (n = 2); MHI-21 (n = 2); PHQ-2; (n = 1); PHQ-9 (n = 1); PSC (n = 1); STAI (n = 3); unvalidated survey (n = 1) Perceived Stress/Distress: CESS (n = 1); COVID PRSI (n = 1); DAS (n = 1); EOL (n = 1); ENSS (n = 1); NSS (n = 2); SA-M (n = 2); MHPSS-M (n = 1); NPIWBS (n = 1); PSS (n = 1); PSS-14 (n = 1); WSQ (n = 1); unvalidated survey (n = 3). Physiological Indicators of Stress/Distress: cortisol (n = 3); glycol (n = 1).	Mental health & somatic symptoms: BDI: 3% above clinical cutoff. CES-D: Mean below clinical cutoff (M = 10.67, SD = 8.06). GAD-7: 5.45–11.3% severe range; 6.15–20.8% moderate range. GHQ: 27–32% of nurses & 32% of physicians in impaired range. HADS: Anxiety & depression in normal range. MHI-21 (scale 1–6, higher scores = better mental health): Psychological distress (M = 3.08, SD = 0.96); hopelessness (M = 1.73, SD = 0.59); emotional wellbeing (M = 3.52, SD = 0.83). PHQ-2: 5.7% at clinical levels. PHQ-9: 29.62% mild, 8.85% moderate, 2.31% moderate-severe & 1.92% severe ranges. STAI: State anxiety in high (n = 2) & low (n = 1) ranges. Perceived Stress/Distress: CESS: Mean below cutoff for stress affecting health. DAS: Mean in moderate range. ENSS: Mean in moderate range. EOL: Mean above scale midpoint. MHPSS: Mean scores showed workload as greatest source of stress, followed by difficulties with infant/parent & organizational structure/processes. NPIWBS: Mean in mild range. NSS: Means on low end of scale. PSS: Mean in high range. PSS-14: 59.2% moderate & 17.5% severe range. SA-M: Stress caring for dying infants (M = 2.49, SD = 1.20; 0–4 scale); highest-rated NICU stressors: floating (M = 3.38, SD = 1.02), unnecessary prolongation of life (M = 3.10, SD = 1.27), emergencies/arrests (M = 3.07, SD = 1.03) & inadequate staffing (M = 3.04, SD = 1.04). WSQ: 5.5% high, 79.2% moderate, & 15.6% low ranges. Physiologic Indicators of Stress/Distress: Reduced immune function & altered physiological stress regulation in nurses & elevated stress markers in NICU staff.	Unnecessary prolongation of life; emergencies/arrests; distressing patient event; patient death/sudden death; being involved in an error or adverse event; poor patient well-being, suffering & quality of life; close relationships (with child, parents) during end-of-life care; end-of-life care assignments > 4 weeks; inadequate staffing & overwork; floating assignment; work schedule; insufficient remuneration for on-call duty; task, organizational & technology obstacles; lack of clearly defined work duties; inadequate involvement in patient pain care decisions; conflict between job demands & personal life; limited nursing administration support; unsupportive unit atmosphere; lower team effectiveness; overcrowded work environment; ≥4 patients per shift; 11–20 night shifts per month; NICU alarm bells & lights; communication problems/conflict among staff (e.g., physician–nurse, nurse–nurse); weak leadership support; inconsistent assignments; lack of family centered care; lack of training; lack of formal & informal support from co-workers/leadership; not talking with friends & family; lack of rest/”mental health days”; low emotional intelligence; negative outlook; lack of acceptance; pessimism; mood disregulation; non-religious background; <1 year of experience; younger age.	Large number of distinct measures & wide range of specific symptoms assessed; variety of scoring formats used & statistics reported; no diagnostic assessment of mental health.

Footnotes: ^a Duplicate data sets were counted only once in sample totals. ^b Abbreviations: BDI = Beck Depression Inventory; BM = burnout measure; CBI = Copenhagen Burnout Inventory; CESS = Commonly Experienced Stress Symptoms; CES-D = Center for Epidemiological Studies Depression Scale; CFS = Compassion Fatigue Scale; CFSI = Cumulative Fatigue Symptoms Index; CIS = Checklist Individual Strength Questionnaire; COVID PRSI = COVID-19 Pandemic-Related Stress Index; DAS = Death Anxiety Scale; EDI = Global Index of Emotional Distress; EOL = End of Life Care Stress; ENSS = Expanded Nurse Stress Scale; GAD-7 = General Anxiety Disorder-7; GHQ = General Health Questionnaire; HADS = Hospital Anxiety and Depression Scale; IES = Impact of Events Scale; IES-R = Impact of Events Scale Revised; MBI = Maslach Burnout Inventory; MDS = Moral Distress Scale; MDS-R = MDS—Revised; MDT = Moral Distress Thermometer; MHI-21 = Mental Health Inventory-21; MHPSS-M = Mental Health Professional Stress Scale, modified; MMD-HP = Measure of Moral Distress for Healthcare Professionals; NPIWBS = Nurse Perceptions of Infant Well-Being Survey; NSS = Nurse Stress Scale; OBI = Oldenburg Burnout Inventory; PC-PTSD = Primary Care PTSD Screen; PHQ-2 = Patient Health Questionnaire-2; ProQOL = Professional Quality of Life; PSC = Psychosomatic Symptom Checklist; PSS = Perceived Stress Scale; SA-M = Stress Audit, modified; STAI = State-Trait Anxiety Inventory; STSS = Secondary Traumatic Stress Scale; TS = Tedium Scale; WSQ = Work Stress Questionnaire. ^c When the measures did not have validated ranges (e.g., low, moderate, or high), we reported the statistical results and/or reported the authors’ interpretation in quotation marks. ^d See Supplementary Table S2 for detailed data gleaned from specific publications.

and Supplementary Table S2, with selected details elaborated in the text below. Data on psychological distress levels and the significant associations that were extracted from each publication are also summarized in Supplementary Table S2, with a general summary presented in Table 1.

3.3. Construct Summaries

3.3.1. Burnout

Burnout Measures

Burnout results were provided in 38 of the 76 (50%) publications, and the list of measures and frequency of use can be found in Table 1 and Supplementary Table S2. The most commonly used measure across all publications was the MBI [48] which consists of three scales: Emotional Exhaustion (EE; nine items with scores ranging from 0 to 54), Depersonalization (DP; five items with scores ranging from 0 to 30), and Personal Accomplishment (PA; eight items with scores ranging from 0 to 48), indicated by high scores on EE and DP and low scores on PA [47]. The ProQOL [16] was the next most used measure of burnout. This measure consists of three scales, each with 10 items with scores ranging from 10 to 50: Burnout, Secondary Traumatic Stress (previously termed Compassion Fatigue), and Compassion Satisfaction. Each scale score can be categorized as low (22 or less), moderate (between 23 and 41) or high (42 or more).

Burnout Levels

Of the 25 publications that provided results from the MBI, 16 (64%) reported data using the full measure, while two publications used only one scale and seven modified the measure (4–10 items). Moreover, only nine publications reported the means for the three MBI scales. The mean subscale scores are listed in Table 1.

Rochefort and Clarke [49] reported that 35.7% of their sample of NICU nurses had EE scores higher than published norms for medical personnel with a cutoff score of 27. While some publications reported percentages of staff in low, moderate and high ranges, Maslach and colleagues do not support the interpretation of these results; thus, the mean scores are only summarized in our results. Specifically, Maslach and colleagues removed the ranges in the 4th edition of the MBI manual since they are not meaningful [48,50]. Only one publication reported the rate of burnout as recommended by Maslach and colleagues and reported that 4.8% of their sample were high on EE and DP and low on PA [51]. Of the six publications using the ProQOL burnout scale, five reported overall burnout means from 18.3 (SD = 3.80) to 24.61 (SD = 1.72), which were in the low (n = 4) to moderate (n = 1) ranges.

Of the 38 publications assessing burnout, 21 (55.3%) included only nurses in their sample, and 17 (44.7%) included nurses and other professions such as physicians, nurse practitioners (NPs), midwives, recreation therapists (RTs), and other NICU staff. Thirteen (34.2%) of the publications assessing burnout included NICU physicians, and while the majority did not compare professions or found no significant differences in burnout across professions (e.g., [52,53]), the findings that were statistically significant were generally mixed. One publication using a modified burnout measure found that physicians and NPs had greater burnout than nurses and RTs [54]. However, another publication that measured the three components of burnout found that physicians had greater DP only relative to nurses [55]. In contrast, a very large multisite study with approximately 2000 participants in California found that physicians had significantly less burnout (on a four-item measure) than the other professions [32,34]. One study with overall low burnout rates found that physicians rated two of the ten burnout items statistically higher than nurses [56], yet these differences were relatively small. Across all studies, however, there is limited replicated evidence to suggest that specific professions in the NICU report significantly more or less symptoms of burnout and instead suggest that these levels may be more impacted by other factors.

Burnout Associations

NICU staff burnout symptoms have been found to be associated with a range of environmental and individual factors. One publication compared nurses in the NICU to other pediatric environments and found that NICU nurses had a significantly increased risk of moderate burnout (OR = 2.78, p = 0.0001) and traumatic stress (OR = 1.67, p = 0.05) relative to other departments [57]. However, other data suggested that working in the pediatric intensive care unit (PICU) may result in more EE and DP symptoms than working in the NICU [58].

Winning and colleagues [59] found that burnout prevalence varied as a function of having observed or been involved in an error or adverse event in the NICU. These data suggest that those who reported no such event had lower burnout scores (M = 20.67, SD = 5.17) than those who observed an adverse event (M = 21.89, SD = 5.00) or were involved in an adverse event themselves (M = 22.44, SD = 4.91). These scores fall in the low range and the lower end of the moderate range.

More generally, burnout symptoms in the NICU have been found to be associated with heavy workloads [60]; high job demands and low resources (e.g., [61]); greater overtime hours [62]; staffing shortages [49]; high admissions and occupancy levels [34]; ratings of a poor work environment [63]; and reports of considering leaving their position [30]. Furthermore, social factors, such as poor teamwork and safety climate [32], management support [32], and nurse leadership styles [64], can have a negative impact on burnout in NICU healthcare providers.

The most common individual-level demographic variables associated with burnout were age and work experience; however, these results were mixed. In some publications, [56,61] greater work experience in years was associated with fewer symptoms of EE and DP as well as positively associated with PA [61]. In a much larger sample of a range of professions across 44 NICUs in California [35], EE was greater in those with 5 or more years’ experience relative to those with less than 3 years’ experience (29% (2 years) vs. 16% (6 years)). Most publications did not find reliable associations between age or experience and burnout symptoms, as well as any consistent direction of association.

Race and/or ethnicity were most often not measured, and diverse and representative samples were very rare. Only one publication reported a significant effect of race [58] such that white providers reported greater PA than non-white providers. Another study found an effect of religious affiliation [65] on levels of emotional exhaustion; however, these findings did not generalize across studies. Stronger associations were most often found between burnout and other constructs of psychological distress [30,44,66,67], and these associations will be discussed further below.

3.3.2. Traumatic Stress

Traumatic Stress Measures

Eleven (14.5%) publications measured traumatic stress, with eight measuring symptoms of secondary traumatic stress (STS) and three publications measuring symptoms of traumatic stress more generally. Measures and frequency of use are listed in Table 1 and Supplementary Table S2.

Traumatic Stress Levels

The prevalence of traumatic stress symptoms in NICU providers was generally low, with some publications showing moderate to high levels. Using the ProQOL, most publications found the average STS was in the low range [57,59,66,68], with one publication reporting that 78% were in the moderate range and 8.5% in the high range [44]. Using the STSS, two publications found STS symptoms to be low [68,69], while another publication found moderate levels on average, with 8.5% in the high and 19.5% in the severe ranges [51]. The one publication using the CFS showed average to high rates of traumatic stress symptoms [70]. Furthermore, using the IES-R in response to neonatal deaths, 34% reported moderate and 35.3% had high traumatic stress symptoms [53], and similar results were found using the IES [26]. Using the PC-PTSD screen, 13.2% reported clinical levels of traumatic stress symptoms [58]. Of the five publications that included a range of NICU professions, only one [51] found that nurses reported greater traumatic stressors than midwives (11.5% vs. 1.7%; χ ² = 13.28, p < 0.001), and generally, there were no other significant differences in the levels of traumatic stress symptoms among professional groups.

Traumatic Stress Associations

Relatively few environmental and individual factors have been assessed and found to have significant associations with traumatic stress in NICU providers. Higher levels of traumatic stress were associated with less vigor (i.e., emotional energy and cognitive liveliness) [69], less self-compassion, poorer nurse-infant/family relationships, and lower reported nurse–physician collegiality [68]. In addition, Ravaldi and colleagues [53] reported that professionals who had received communication training were less likely to develop hypervigilance, a component of traumatic stress, and that having training in bereavement was the best and only independent predictor of both traumatic stress and burnout. Lastly, Winning and colleagues [59] reported that providers involved in an error or adverse event reported higher levels of STS than those who experienced no such events.

3.3.3. Moral Distress

Moral Distress Measures

Moral distress was measured in 20 (26.3%) of the 76 publications, and a summary of measures and results is in Table 1 and Supplementary Table S2. The Moral Distress Scale (MDS) and MDS-Revised (MDS-R) were used most frequently to assess moral distress. The original MDS was developed by Corley and colleagues [71] to assess moral distress in intensive care nurses. The measure consists of 32 items scored on a 7-point Likert scale grouped into three key factors/subscales reflecting sources of moral distress: (1) Futile Care/Treatment (situations involving life-sustaining treatments or interventions considered unnecessary or burdensome); (2) Ethical Misconduct/Compromise (occurrences of perceived wrongdoing by other healthcare professionals or institutions); and (3) Lack of Autonomy/Communication (issues where healthcare providers feel unable to act on their professional judgment or patients’ wishes due to constraints). Items describe specific, morally loaded clinical situations, such as “carrying out physician’s orders for unnecessary tests” or “working with inadequate staff”. The MDS-R [71] is a shorter 21-item version intended for broader use in healthcare. Items were refined and adapted to be used with varied providers (e.g., nurses and physicians) and settings (e.g., adult/pediatric and critical/acute care). The MDS-R often reveals a multi-dimensional structure across studies, with common subscales including: (1) Futile Care (instances of continuing treatment that does not provide benefit; (2) Ethical Misconduct/Lack of Moral Courage (witnessing or participating in unethical, unsafe, or incompetent care); (3) Deceptive Communication (issues regarding honesty and communication with patients/families); and (4) Poor Teamwork/Organizational Pressure (inadequate collaboration or environmental pressures restricting ethical care). One common iteration of the MDS-R—the Measure of Moral Distress for Healthcare Professionals (MMD-HP) [72] is an updated, more comprehensive version (27-items) that addresses three types of underlying root causes of moral distress, including: (1) patient-level issues (e.g., providing care that is not in the patient’s best interest); (2) team-level issues (e.g., incompetence or lack of respect among colleagues); and (3) system-level issues (e.g., inadequate resources or staffing). Scoring for the MDS, MDS-R and MMD-HP focuses on how often a situation occurs (frequency) and how distressing the situation feels (intensity). None of the moral distress measures provide validated ranges (e.g., low, moderate or high) for interpretation; therefore, we used the language used in each publication.

Moral Distress Levels

Overall, reported levels of moral distress varied greatly across and within studies, but largely were in low to moderate ranges. Of the five publications that used the MDS, “average” levels of moral distress indicated by mean composite scores, mean frequency scores, and mean intensity scores were reported in two publications. “Moderate” levels of moral distress frequency and intensity were reported in one publication [73], while another publication reported mean frequency in the “average” range, and intensity in the “high” range [70]. The one publication that used only the “not in the patient’s best interests” subscale revealed “moderately low” moral distress [74]; Cavaliere and colleagues [75] noted that individual scores in their sample displayed a wide range of moral distress scores, indicating that moral distress may be experienced at the individual level versus at a group level.

Different scales and statistics were reported across the nine publications that used the MDS-R; levels of moral distress ranged. Most reported levels in the “low” range [31,76], “low-moderate” range [77,78], and “moderate” range [79,80], with only one publication reporting moral distress levels in the “high” range [81]. Nurses generally reported significantly greater moral distress on the MDS-R than physicians [52,81,82]; however, only one of these publications reported high distress levels in their sample [81]. Three publications reported that the clinical situation that produced the most moral distress (i.e., highest-scored MDS/MDS-R item) was “Following the family’s wishes to continue life support even though I believe it is not in the best interest of the child” [75,80,81]. One publication [78] reported that the clinical situation most often endorsed on the MDS-R was “Initiate extensive life-saving actions when I think it only prolongs dying.” This publication also reported that high moral distress (i.e., MDS-R composite score > 64) was experienced more by male nurses than by female nurses.

The one publication that used the MMD-HP reported a mean composite moral distress score in the “low to moderate” range [83]. Mean frequency and intensity scores were not presented. In one of the two studies that used the Moral Distress Thermometer, moral distress was reported in 687 of 4589 of survey occasions (15%) for 525 healthcare providers, and the mean intensity of moral distress was 5.0 (SD = 2.2) on a 0–10 scale [38]. In the other study that used a prospective cohort design, the overall average increase in moral distress was 0.84 (SD = 3.15, p = 0.002) for the total sample, with physicians reporting a larger average increase in moral distress relative to nurses (1.1, SD = 3.52, p = 0.01; versus 0.55, SD = 2.66, p = 0.07) [37].

Moral Distress Associations

Across publications, significant associations with various individual and environmental factors were reported. Factors that were positively associated with moral distress include: working with higher acuity patients [42]; nurse uncertainty about whether their care was of benefit [81]; nurse uncertainty about a child’s prognosis [81]; perceptions of futile care [79]; the wish to diminish treatment [82]; end-of-life care and communication [81]; perception of “parents wanting too much” [38]; critical reflective practice in the NICU [74] and nurse considering leaving their position [37,75,78,83] with 9% to 33% reporting that they were considering leaving their positions in the NICU [37,76,77,78].

Moral distress was inversely associated with psychological empowerment (i.e., meaning, competence, and self-determination); work engagement; perceived hospital supportiveness [81]; and favorable hospital ethical climate [73]. In a prospective cohort study [37], providers who were not satisfied with the goals of care discussion and those who perceived disagreement among the participating clinicians had a greater increase in moral distress relative to providers who were satisfied. In this sample, moral distress was more likely to decrease when providers respected parents’ views about their child’s quality of life. In the four publications that reported significant associations with sociodemographic factors, religiosity [82] and increasing years of ICU experience [81] were positively associated with moral distress, whereas age and working experience were inversely associated [79,84].

3.3.4. Other Distress Outcomes

Other Distress Measures

Of the 76 publications, 31 (40.8%) assessed an aspect of psychological distress that we classified as “other distress” and further subcategorized based on the type of outcome measured, including (1) self-reported mental health and somatic symptoms (e.g., anxiety, depression, fatigue); (2) self-reported perceptions of stress or distress; and (3) physiological indicators of stress/distress, such as cortisol level. A large variety of measures were used. The number of publications that used each measure within each of these subcategories is listed in alphabetical order in Table 1.

Other Distress Levels

The levels of psychological stress/distress outcomes reported by the publications that we classified as Other Distress are summarized below, grouped by our designated subcategories. Given the variety of measures, scoring formats used, and statistics reported, the summary of stress/distress levels is reported across publications. Levels of stress/distress among NICU healthcare provider types are reported when available.

• Mental Health and Somatic Symptoms.

Within the publications that assessed mental health responses, results were mixed depending on the measures used and the specific symptoms assessed, as well as the level of symptoms distributed within the sample. Although some publications did not show elevated symptoms above clinical thresholds for the respective measure used [51,59,61,66,85], across publications, moderate levels of specific emotional symptoms were generally endorsed as a reflection of occupational stress. Specific symptoms that were identified included: feelings of helplessness [85,86,87], hopelessness [28], intense sorrow [85,86], sadness [85], irritability [86], depression [58], anxiety/state anxiety [26,58,87,88], and defensiveness [85]. NICU nurses reported higher state anxiety than physicians (p = 0.041), but not other NICU professionals [26]. A higher prevalence of psychological impairment (i.e., symptoms of anxiety, depression, poor well-being) was found in both NICU nurses and neonatologists than in the general population [89,90].

Mild to moderate somatic/physical symptoms and behaviors were also endorsed with notable frequency, such as lack of exercise [86], chronic fatigue [45,86,91]; headaches [85,86]; and work-related sleep problems [36].

• Perceived Stress/Distress.

A subset of the publications that measured self-reported stress and distress associated with working in the NICU and caring for critically ill and dying infants showed nurses generally endorsed mild to moderate levels of overall distress [29,43,58,60,85,86,87,92,93,94,95]. One publication showed “higher than medium” level of end-of-life care stress for NICU nurses [96]. The majority of NICU nurses in another sample reported moderate to severe stress [44], while NICU nurses, physicians, nurse practitioners, and physician assistants in another study reported levels of perceived stress in the high range [67]. Forty-four percent of a large sample of neonatologists reported that they feel overworked often/always [36]. Another publication reported that females and physicians had significantly higher stress symptoms than males and nurses, respectively [93].

• Physiologic Indicators.

Although participants in studies classified within the other distress category did not consistently report high levels of distress on self-report measures, several studies incorporating physiological indices suggest evidence of elevated stress responses among NICU healthcare providers. Specifically, Fujimaru and colleagues [45] found that NICU nurses exhibited lower salivary immunoglobulin A concentrations—indicative of reduced immune function—alongside higher anxiety, depression, and fatigue scores compared to general ward nurses. Similarly, Fumagalli and colleagues [46] reported that nurses endorsing higher perceived stress demonstrated a smaller decline in cortisol levels across their work shifts (p < 0.001), suggesting altered physiological stress regulation. In addition, findings from Mörelius and colleagues [47] indicate that physiological stress markers may be elevated in NICU staff, although associations with self-reported distress were variable across studies.

Other Distress Associations

Various factors were found to be associated with healthcare provider stress/distress in the NICU, and a general summary list can be found in Table 1, with more details provided in Supplementary Table S2. The following environmental factors, including system-level and work-related interpersonal factors, were reported as associated with elevated stress/distress in nurses: unnecessary prolongation of life; emergencies/arrests [97]; distressing patient event [91]; patient death/sudden death [43,87,96]; being involved in an error or adverse event [59]; poor patient well-being, suffering, and QOL [43,92]; close relationships (with the child and parents) during end-of-life care [40]; inadequate staffing and overwork [36,40,60,89], floating assignment [97]; end-of-life care assignments greater than 4 weeks [40]; work schedule [94]; insufficient remuneration for on-call duty [36]; task, organizational and technology obstacles [94]; lack of clearly defined work duties [90]; inadequate involvement in patient pain care decisions [89]; conflict between job demands and personal life [90]; limited nursing administration support [90]; supportive atmosphere of unit [90]; team effectiveness [94]; overcrowded work environment [89]; having four or more patients per shift, 11–20 night shifts per month, NICU alarm bells and lights [44]; and communication problems/conflict among staff (e.g., physician–nurse and nurse–nurse) [43,89,90] In one publication, neonatologists who reported greater distress reported greater frequency of conflict between job demands and personal life [90].

Environmental factors associated with reductions in psychological stress/distress included: strong leadership support [91,94]; consistent assignments [94,97]; family-centered care [94]; training [94,97]; formal and informal support from co-workers/leadership [59,87,91,94,97]; “discussion with person respected for wisdom” [44]; talking with friends and family [87,97]; and rest/”mental health days” [87,97]. Very few studies reported notable associations between individual or sociodemographic factors and stress/distress. Those factors that were positively associated with stress/distress included age [43] and having less than 1 year of experience [44]. Individual-level factors that were negatively associated with stress/distress included: religious background [97]; emotional intelligence [29]; positive outlook and acceptance [87]; and optimism and mood regulation [29].

3.3.5. Associations Across Constructs

Of the 76 publications, only 18 (23.7%) included measures for two or more constructs of psychological distress (see Figure 5). Of these publications, relatively few reported the zero-order correlations or assessed more complex relationships among constructs.

Burnout was associated with most other psychological distress constructs when reported. Of the nine publications that included measures of both burnout and traumatic stress, three reported correlations in numerical format with values ranging from r = 0.44 to r = 0.55 [26,66,68]. One publication [53] reported results in scatter plots without numerical values and noted that all scales on the MBI (burnout) were significantly associated with the IES (traumatic stress) scales. In addition, one publication [66] found that support-seeking strategies moderated the association between burnout and traumatic stress symptoms.

Moreover, two publications reported small, yet significant associations between a subscale of burnout and moral distress [52,81]. Similarly, Barr [28] found that the moral distress dimensions explained 20%, 13%, and 7% of the variance of burnout demoralization, exhaustion, and loss of motive, respectively. Burnout (or EE) was also found to be associated with depression, with correlations from r = 0.29 [66] to r = 0.50 [26], and to a lesser degree with anxiety [26,61] and self-reported stress [60,67]. Burnout was not significantly associated with cortisol levels [47].

Similar to burnout, traumatic stress was also associated with moral distress and other types of psychological distress overall. One publication found that traumatic stress was correlated with the intensity (but not frequency) of moral distress (r = 0.44 [70]). Of the six publications to include other measures of distress, two reported correlations and found that traumatic stress was associated with psychosomatic symptoms (r = 0.8 [26]), depression (r = 0.6 [26]), anxiety (r = 0.5 [26]) and, to a lesser degree, with self-reported perceived stress [44].

4. Discussion

This scoping review is the first to synthesize 76 quantitative publications (70 independent samples) examining psychological stress and distress among NICU healthcare providers. Studies varied widely in the measures and variables used, how measures were applied and scored, and how results were reported and interpreted. This heterogeneity made it challenging to draw firm conclusions about the prevalence and correlates of psychological distress among NICU providers. Overall, we found that this research literature has focused on three primary constructs: burnout, traumatic stress, and moral distress. In addition, a range of other psychological distress outcomes were examined, including mental health and somatic symptoms, self-reported stress/distress, and physiological outcomes. Together, these findings suggest that the NICU workplace can be a stressful environment across professions, with most of the research being conducted with female nurses. Moreover, working in the NICU may contribute to problematic and even clinical levels of psychological distress for some individuals that warrant assessment and intervention. Overall, a wide range of environmental factors seem to be associated with distress with minimal replication, whereas findings for specific individual and sociodemographic factors were less frequent and consistent. These results limit our ability to identify potential risk and protective factors to inform organizational assessment, prevention, and intervention of psychological distress in NICU providers.

4.1. Psychological Distress Levels

Half of the publications in this scoping review used some measure of burnout, and a quarter measured moral distress; fewer evaluated traumatic stress. Nearly 40% of the publications assessed other mental health and distress outcomes using multiple distinct measures. On average, despite the substantial variability, the levels of psychological distress in NICU providers fell in the low to moderate ranges. Generally speaking, these results are consistent with the literature on healthcare providers more broadly for burnout [98] and with findings of moral distress in ICUs [99], with levels of moral distress found to be greater in intensive care settings like the NICU than in other areas of nursing [99,100]. Overall, mental health symptoms measured by self-report are prevalent, yet most publications reported that the means of their sample were below clinical levels, suggesting that mental health problems, including traumatic stress, may be less common than symptoms of exhaustion, burnout and moral distress [101,102].

Importantly, the included studies spanned pre-COVID-19 pandemic and post-COVID-19 pandemic periods, which may have contributed to variability in distress levels. The broader healthcare-worker literature indicates substantial mental health burden during the pandemic [103], with meta-analytic evidence showing higher probable post-traumatic stress disorder during COVID-19 than before the pandemic and a subsequent return toward pre-pandemic levels by 2022 [104]. A nationally representative USA study found that differences in anxiety and depression between direct- and non-direct-patient-care healthcare workers were specific to the pandemic period [105]. However, differences in study populations, measures, clinical settings, and timing limit direct comparisons of distress levels across the pre-pandemic, pandemic, and post-pandemic studies included in this review.

Nonetheless, mean levels of distress do not tell the full story of NICU healthcare providers’ experience, as the distribution of data and variability are also important. A subset of providers reported high or even clinical levels of psychological distress that are of concern and may warrant intervention. For example, while only one study reported the overall burnout level as recommended by Maslach and colleagues [51,106], they found that almost 5% met the full definition of burnout. This result is comparable with other studies of healthcare professionals when the data are scored the same way [107], and Maslach and Leiter [106] estimate that, on average, 10% to 15% of employees are likely to fit their full burnout profile. Furthermore, another publication found that 13.2% of the NICU sample screened positive for possible clinical levels of traumatic stress [58], and this prevalence rate is comparable with a meta-analysis reporting a PTSD prevalence rate of 13.5% in healthcare workers [108]. The general population has lifetime PTSD estimates of 7.8% and past-year estimates of 3.6% in the USA when measured through diagnostic interviews [109]. However, none of the included studies used a diagnostic interview for PTSD or other mental health conditions, and thus the prevalence of mental health conditions in NICU providers is unknown.

Furthermore, there were insufficient data comparing NICUs with other healthcare environments to determine whether NICU providers are at greater, similar, or lower risk of psychological distress than other healthcare providers. Research on other medical settings has found that healthcare providers show burnout prevalence rates between 35% and 52% depending on the measures and cutoff scores used (which vary greatly) and that intensive care nurses are among the highest estimates with 58%, only second to emergency physicians at 67% [110]. However, these prevalence rates are often only based on the EE subscale rather than the full MBI and often include those in moderate and high ranges together and thus are not estimated as directed by Maslach and colleagues [106]. Therefore, most often, these results are likely overestimating the clinical levels of distress and contribute to the inability to adequately quantify the levels of burnout and other constructs of psychological distress among NICU and healthcare providers more broadly.

4.2. Psychological Distress Associations

Across studies, NICU providers’ psychological distress appeared to be associated with a wide range of modifiable workplace factors, whereas individual and sociodemographic characteristics were assessed less frequently and yielded mixed results. These results are consistent with broader healthcare literature (see [111] for a review) across a range of professions, including medical professionals, in which job factors (e.g., workload, job satisfaction, and hours) accounted for 23.1% of the variance in occupational stress, followed by work environment factors (e.g., noise and conflict) at 20.8%, economic factors (e.g., inflation and rights) at 19.1% and organizational factors (e.g., climate, management, and pay) at 19.1%. Individual factors (e.g., age, work experience, and gender/sex) accounted for only 7.9%, with highly mixed findings like those observed in the current review.

The strongest and most consistent findings were associations among psychological distress constructs. However, there are very few conceptual models that explain the nature of the associations among burnout, traumatic stress, and moral distress. A commonly used model for understanding occupational stress, the Job Demands-Resources (JD-R) Model [112], primarily focuses on the health outcome of burnout and does not explicitly include other forms of psychological distress. The Figley’s Compassion Fatigue Model [113,114] and Stamm Professional Quality of Life Model [16] incorporate the combined contributions of burnout and traumatic stress, conceptualizing them as independent yet interconnected processes contributing to compassion fatigue, or the reduced capacity to provide compassionate care. These models would suggest that increases in one form of distress may negatively affect the other. For example, a provider who develops traumatic stress following a workplace event may be more likely to experience burnout symptoms and vice versa. However, empirical support for this model is limited to date, and its validity and structure have been questioned (e.g., [115]). Furthermore, these models do not explicitly incorporate the contribution of moral distress, suggesting that expanded models may be needed to account for the observed associations among these constructs.

More recently, the Alternative Response Program Moral Injury and Burnout Model [116], based on the Moral Injury Continuum Model [117], attempts to incorporate moral distress, burnout, and traumatic stress within a continuum in which moral dilemmas and persistent moral distress may contribute to moral injury. Moral injury and burnout are conceptualized as distinct but interacting constructs, with workplace conditions potentially contributing to or protecting against both constructs. This model identifies traumatic stress as a distinct construct and one of several potential adverse outcomes, along with symptoms of anxiety and depression, associated with moral injury and/or burnout. Importantly, the model identifies multiple potential pathways for intervention, including individual, relational, workplace, and systems-level approaches to reducing or preventing moral distress, moral injury, and burnout. Nevertheless, in the current NICU literature, there was a lack of theoretical frameworks and multilevel modeling approaches that could clarify associations among the broad range of psychological distress constructs, highlighting another important direction for future research.

Moreover, in Stamm’s Professional Quality of Life Model [16], job satisfaction is believed to balance out the negative consequences of burnout and traumatic stress. In the current review, there was some evidence of moderate to high job satisfaction across NICU providers. While only two publications [44,57] reported results from the ProQOL satisfaction scale, a range of other publications used some scale of satisfaction [27,32,43,63,64,77]. Similarly, results from the PA scale of the MBI were also generally in the moderate to high ranges (M = 27.35 to 38.40) [62,118] and may be considered an indirect indication or correlate, but not a direct measure of job satisfaction. Together, these results suggest that healthcare providers in the NICU are overall satisfied with their work and likely continue to work in this environment, despite many stressors, challenges, and psychological distress [119]. Although NICU work is highly demanding, it can foster resilience among providers [120] and is often experienced as deeply meaningful and rewarding work [121]. Future research would benefit from directly measuring what contributes to and sustains job satisfaction and professional fulfillment.

Lastly, across publications, 9 to 33% of participants indicated that they were considering leaving their position in the NICU [37,76,77,78], and one publication reported that almost 17% of nurses left their position in an 11-month period [122]. Reports of leaving have been associated with burnout ([76], but not [122]), and moral distress ([83], but not [76]). However, the factors that likely contribute to a provider’s staying in the NICU or leaving their position (not just considering leaving) are likely multifactorial and complex [123]. Turnover rates in nursing worldwide are estimated to be 16% [124], with greater turnover rates in ICUs at 23%; attempts to reduce distress and turnover are likely to result in many workforce, economic, and medical benefits [125].

4.3. Limitations and Gaps in the Literature

Making clear summative conclusions about the current state of psychological distress in NICU healthcare providers is challenging and incomplete given substantial conceptual, methodological, and sampling limitations. A primary concern is the reliance on cross-sectional survey designs. As such, this literature provides little insight into how distress develops, fluctuates, or resolves over time. This is a critical gap given the dynamic nature of NICU work environments and the likelihood that provider distress is episodic and context-sensitive. Longitudinal studies using validated measures are necessary to clarify temporal relationships among stressors, potential mediators, and outcomes, and to identify times of increased vulnerability or resilience across providers’ careers [52,126].

Conceptual ambiguity across psychological constructs further complicates the extant literature. Many studies measured single constructs in isolation, despite high correlations reported when multiple constructs were included. For example, burnout symptoms overlap with depressive symptoms, with the current literature defining burnout as “occupational depression” [11,127]. Moreover, the term “burnout” is often misunderstood and used imprecisely as a catch-all for diverse forms of work-related distress, without adherence to its established conceptual parameters [106]. Secondary traumatic stress reflects symptom patterns similar to primary traumatic stress [128]. Moral distress also differs conceptually from these other constructs because it arises from exposure to ethically challenging situations rather than from the broader stressors associated with working in the NICU [24,129]. From a physiological perspective, stress responses may not meaningfully differentiate between sources of stress. Future research would benefit from distinguishing between workplace stressors (e.g., ethical conflict, workload) and stress responses (e.g., emotional exhaustion and trauma symptoms) and examining multiple constructs within integrated models.

Inconsistencies in measurement further limit cross-study comparisons and interpretation of the results. Several studies did not apply measures according to established guidelines, particularly the MBI, MDS, and MDS-R, where it is unfortunately common for the results to lack standardization [130]. In some cases, subscales were used independently, or scoring methods were modified, resulting in partial assessment of the distress constructs [8,106]. Furthermore, many publications report results attempting to estimate clinical levels of distress when measures, such as the MBI, were neither designed nor found to be effective clinical diagnostic tools [131]. Maslach has emphasized that the MBI was designed for research only [106]. Similarly, studies using the MDS and MDS-R reported findings using different statistics and different subsets of scales. Greater consistency in the use and reporting of validated measures would strengthen the evidence base and allow more meaningful comparisons across studies.

In terms of sampling, when reported, most of the publications included in this review relied primarily on predominantly White, female nursing samples. Other NICU professionals, including physicians, respiratory therapists, and social workers, were underrepresented. More diverse and representative samples are needed to capture the experiences of the broader multidisciplinary NICU workforce. Further, it is possible that the publications included were at risk of sampling bias such that highly distressed healthcare providers may be more or less likely to provide data; these bias effects are unknown. In fact, providers who are most distressed may be more likely to leave their positions and may not be represented, thus underestimating distress levels. In addition, the data reported within and across publications were highly variable, with wide confidence intervals suggesting large heterogeneity and limited precision in estimating levels of distress.

4.4. Limitations of Scoping Review

We note several potential methodological limitations within our scoping review. First, our search terms were intentionally broad so that we could include different types of psychological distress outcomes; however, it was not an exhaustive list, and there may be additional aspects of distress to explore (e.g., guilt and spiritual distress) that other search teams may have captured. Second, some publications may have been excluded incorrectly based on review of the titles and/or abstracts, and it is possible that having ten reviewers for the initial round of selection could have led to some inconsistencies in selection. The first two authors (PAG and ALN) reviewed the second set of search results to increase accuracy. Third, although our focus on quantitative studies provides a broad overview of patterns of distress in NICU providers, our review does not capture insights from qualitative research that may offer a deeper and more personal understanding of provider experiences. Fourth, we acknowledge the lack of a registered review protocol prior to conducting the review. Finally, the most recent literature search was completed on 15 October 2024; therefore, publications appearing after this date are not represented in this review, which is an important limitation given the continued growth of research on psychological distress among NICU healthcare providers.

4.5. Research and Practice Implications

In summary, future research should extend beyond single-construct prevalence studies and examine multiple psychological constructs simultaneously. Longitudinal and multilevel models are needed to clarify relationships between workplace stressors, sociodemographic, individual, and organizational factors, and psychological outcomes in NICU providers. More consistent use of theoretical and conceptual models, standardized definitions, validated measures used as intended, and consistent methodologies can improve scientific rigor and facilitate cross-study comparisons. In addition, future research should also include larger and more diverse samples representing the full range of NICU professionals beyond white, female nurses that comprised most participants in the current literature.

The moderate to high levels of distress among some NICU providers highlight the need for evidence-informed assessment and intervention. One systematic review published in 2020 found only six publications documenting interventions to address psychological distress within NICUs and reported some efficacy in reducing work-related stress across a small range of approaches (e.g., mindfulness, stress management, and consultation rounds); however, the evidence base also remains limited by the small number and quality of studies and substantial heterogeneity in study designs, populations, interventions, and measures. In the larger healthcare literature, a systematic review found that 30 of 33 studies focused on individual interventions (e.g., mindfulness, yoga, and journaling), with only three interventions targeting organizational factors (e.g., workload reduction and peer support) [132], with some evidence of efficacy for both approaches. However, another review found that while individual interventions such as cognitive–behavioral therapy and relaxation approaches were effective in reducing stress, burnout, and anxiety symptoms in healthcare providers, there were very few studies (n = 2) and no clear evidence to support the organizational interventions [133]. Furthermore, Critical Incident Stress Management (see [134] for a review), in theory, has the potential to reduce traumatic, moral, and general psychological distress with group and/or individual debriefing after stressful events, yet this organizational approach currently lacks rigorous evaluation and is not commonly used in the NICU [135]. Together, these results suggest that while some interventions may reduce psychological distress among NICU providers (also see [136] for a Cochrane review), the evidence remains limited, and interventions have generally been narrow in scope and focused on treating the individual’s response to stressors rather than addressing workplace stressors and processes more directly.

This pattern is also evident in the literature specifically targeting moral distress, although more recent work has begun to emphasize workplace-based approaches. A systematic review of ICU interventions identified moral empowerment and end-of-life education, reflective writing and debriefing, multidisciplinary case discussions, and moral resilience training, with insufficient evidence to identify a clearly effective approach [137]. More recent NICU-specific interventions have emphasized interdisciplinary and multifaceted strategies, including trained support clinicians who provide peer support, facilitate reflection and communication, and connect clinicians with ethics, palliative care, and other resources [138]. Brief, case-specific interprofessional meetings have demonstrated small but significant short-term reductions in moral distress among NICU clinicians [139], whereas a pilot intervention combining education with regularly scheduled reflective emotional debriefing resulted in a small but nonsignificant reduction in moral distress [140]. More recently, a randomized controlled trial of NICU and PICU nurses found that an interactive ethics education intervention significantly reduced moral distress while increasing moral sensitivity, with effects maintained at two-week follow-up [141]. The recent literature has emphasized that moral distress may reflect ethical engagement and prompt reflection and discussion around complex clinical decisions [24,38,42] and therefore should not necessarily be viewed as something interventions should seek to eliminate. Indeed, the simultaneous reduction in moral distress and increase in moral sensitivity in the recent trial suggests that interventions may help mitigate harmful distress without diminishing clinicians’ ethical awareness. Instead, this literature suggests that interventions may be most useful when they support decision-making in the face of morally challenging situations and aim to reduce the harmful consequences of unresolved moral distress [138]. Together, these findings suggest that timely, interdisciplinary interventions that facilitate reflection, communication, and action around challenging cases may be particularly promising, but evidence remains preliminary, and more rigorous research examining sustained effects is needed.

Given the wide range of psychological distress constructs and associated factors in the current review, there is likely no single individual or organizational intervention that will address all sources or types of distress responses. Rather, the broader literature suggests that multicomponent organizational strategies are more promising than individual resilience training alone [142,143]. Consistent with Maslach’s approach to burnout, NICU providers may benefit from assessing sources and levels of distress within their units across professions on a repeated basis over time to target and customize interventions to identified areas of need, rather than a one-size-fits-all approach. The results of the current review further suggest that NICUs should assess multiple constructs beyond burnout, using the most widely used and validated measures. NICUs would also benefit from engaging their healthcare providers in collaborative consultation and problem-solving to develop solutions and interventions at individual, team, systems, and organizational levels [144], supported by investing in strong NICU leadership [145].

5. Conclusions

This review is the first to synthesize quantitative research across multiple psychological constructs relevant to NICU provider distress. By examining these constructs together, this review highlights important conceptual overlap, variability in measurement, and methodological inconsistencies that limit clear interpretation of distress levels and associated factors reported in the current literature.

As reflected in Lantos’ observation that “there is nothing morally neutral about a NICU…they are horrible, wonderful places” [1], the findings of this review underscore the fundamentally complex and paradoxical nature of NICU work. While average levels of distress among providers are often low to moderate, a meaningful subset experience elevated or clinically significant symptoms, reflecting the emotional and ethical intensity inherent in this setting. Distress appears to be associated with a wide range of potentially modifiable workplace and moral stressors and less consistently with individual sociodemographic characteristics, pointing to the critical role of organizational context that varied greatly across publications. At the same time, relatively moderate to high levels of job satisfaction suggest that NICU work remains deeply meaningful, reinforcing the coexistence of burden and reward described by Lantos.

Advancing the field will require greater conceptual clarity, consistent and appropriate use of validated measures, and more rigorous longitudinal and multilevel research designs. Efforts to support NICU providers should prioritize research evaluating multicomponent interventions tailored to specific unit needs and address both individual and system-level factors. Ultimately, acknowledging and addressing both the “horrible” and “wonderful” aspects of NICU care is essential for supporting provider wellbeing, sustaining the workforce, and ensuring high-quality care for infants and families.