Moral distress, emotional impact and coping in intensive care unit staff during the outbreak of COVID-19

Background From the beginning, the COVID-19 pandemic increased ICU workloads and created exceptionally difficult ethical dilemmas. ICU staff around the world have been subject to high levels of moral stress, potentially leading to mental health problems. There is only limited evidence on moral distress levels and coping styles among Spanish ICU staff, and how they influenced health professionals’ mental health during the pandemic. Objectives To assess moral distress, related mental health problems (anxiety and depression), and coping styles among ICU staff during the first wave of the COVID-19 pandemic in Spain. Design Cross-sectional. Settings and participants The study setting consisted of intensive care unit and areas converted into intensive care units in public and private hospitals. A total of 434 permanent and temporary intensive care staff (reassigned due to the pandemic from other departments to units) answered an online questionnaire between March and June 2020. Methods Sociodemographic and job variables, moral distress, anxiety, depression, and coping mechanisms were anonymously evaluated through a self-reported questionnaire. Descriptive and correlation analyses were conducted and multivariate linear regression models were developed to explore the predictive ability of moral distress and coping on anxiety and depression. Results Moral distress during the pandemic is determined by situations related to the patient and family, the intensive care unit, and resource management of the organisations themselves. intensive care unit staff already reached moderate levels of moral distress, anxiety, and depression during the first wave of the pandemic. Temporary staff (redeployed from other units) obtained higher scores in these variables (p = 0.04, p = 0.038, and p = 0.009, respectively) than permanent staff, as well as in greater intention to leave their current position (p = 0.03). This intention was also stronger in health staff working in areas converted into intensive care units (45.2%) than in normal intensive care units (40.2%) (p = 0.02). Moral distress, coupled with primarily avoidance-oriented coping styles, explains 37% (AdR2) of the variance in anxiety and 38% (AdR2) of the variance in depression. Conclusions Our study reveals that the emotional well-being of intensive care unit staff was already at risk during the first wave of the pandemic. The moral distress they experienced was related to anxiety and depression issues, as well as the desire to leave the profession, and should be addressed, not only in permanent staff, but also in temporary staff, redeployed to these units as reinforcement workers.


Moral distress
Moral distress was first described in nurses as a process of pain or anguish occurring when health professionals know the ethically correct action they should perform in their work, but there are real or perceived limitations preventing them from doing the right thing (Jameton, 1984). The study was later extended to other professionals such as physicians, psychologists, occupational therapists, social workers, and other healthcare professionals (Austin et al., 2005;Schwenzer and Wang, 2006;Brazil et al., 2010;Epstein et al., 2019). The factors have been divided into three main categories related to 1) the patient and family (e. g., patient-centred care conflicts, participating in non-beneficial treatments), 2) the intensive care unit (ICU) and work staff (e.g., ethical conflicts with colleagues, lack of shared decision-making, team discordance) and 3) the healthcare system itself (e.g., poor communication, institutional policies) (Hamric and Epstein, 2017). When a health professional is repeatedly unable over time to perform actions they consider ethically correct, a "moral residue" is created (Webster and Bayliss, 2000;Epstein and Hamric, 2009) and the negative feelings remain after the situation has passed, leading to a "crescendo" effect, with increasingly intense responses to new experiences (Epstein and Hamric, 2009).
Intensive care units are one of the departments where moral distress has been studied most extensively (Bruce et al., 2015;Henrich et al., 2016;Meltzer and Huckabay, 2004). In these units, situations that facilitate this emotional state include ethical dilemmas with critically ill patients, the use of technology and life support, patients who require withdrawing or withholding treatment, and clinical actions that could imply futile medical care (Hiler, et al., 2018;Mealer and Moss, 2016;Piers et al., 2011). A qualitative phenomenology study by Henrich et al. (2016) also identified additional factors: excess or lack of quality and quantity of the care provided, non-existent or inappropriate care plans, communication issues, need to make end-of-life care decisions, interaction and conflict between ICU staff and the family, recommendations given to patients by other professionals, and problems related to ICU resources and support received. In 2016, various scientific critical care societies appealed for action, highlighting moral distress as one of the essential areas for investigation and intervention in ICUs, together with burnout, compassion fatigue, and the perception of inappropriate care . Health professionals may perceive that the care they offer to patients is "inappropriate" when their knowledge or beliefs about its quality or quantity are not aligned (Kon et al., 2016).

COVID-19: emotional impact and coping strategies
The COVID-19 pandemic has been associated with an increase in the psychosocial risks of healthcare work, implying significant emotional consequences for these professionals (Blanco-Donoso et al., 2020). In ICUs in particular, moral distress levels have risen (Sheather and Fidler, 2021), and various studies mention decisive aggravating factors in the units, such as rationing and triage due to the scarcity of resources e.g., lack of ventilatory support, duplication of ICU beds in departments other than the usual ICU, and a lack of sufficient personal protective equipment (PPE), the need to prioritise COVID-19 patients over others, difficulty with team collaboration due to members affected by the disease, and patient and family solitude and isolation (Cacchione, 2020;Kanaris, 2021;Morley et al., 2020;Sheather and Fidler, 2021).
The distress of healthcare work during the pandemic has affected the emotional health of professionals, who have shown symptoms of anxiety, depression, peritraumatic dissociation, and burnout (Shanafelt et al., 2020;Azoulay et al., 2020a;Azoulay et al., 2020b). In Spain, some studies report that 45.7% of health professionals were at high risk for a mental disorder during the first phase of the pandemic, with reports also of symptoms related to general anxiety, panic attacks, depression, substance abuse, and suicidal ideation (Alonso et al., 2020;Mortier et al., 2020). Although there is research on this emotional impact and its prevalence, only a few studies have investigated how professionals cope with the resulting stress.
According to Lazarus and Folkman's traditional model (1984), coping strategies are understood to be any cognitive and behavioural efforts developed to handle specific external and internal demands considered overwhelming for the person's resources. These strategies are aimed at reducing the pressure of stressful situations. This objective is achieved through coping directly (thus reducing stress), but also indirectly, by mediating between stressors and emotional consequences for health (Folkman and Lazarus, 1988). During the pandemic, Babore et al. (2020) in Italy reported on an example of direct coping, finding

Implications for Clinical Practice
• Intensive care unit staff were clearly at risk of suffering moral distress during the first wave of the pandemic.
• The mental health of intensive care unit staff is largely related to moral distress and avoidance-oriented coping styles; moral distress is particularly prominent in health professionals from other departments redeployed to unit and in professionals working in hospital units converted into intensive care units. Emotional support should be offered to intensive care unit workers during the pandemic, particularly to staff brought in from other departments. • Preventive and emotional support interventions for health professionals should consider including psychoeducation related to active coping strategies, training in dealing with ethical dilemmas, and specialised training to care for critically ill patients. • This research provides key concepts for the evaluation of mental health that can be improved and used by the Occupational Health and Mental Health departments of hospitals, during a pandemic, for the assessment, follow-up and, if necessary, intervention, in intensive care unit staff.
that health staff's main functional strategy for coping was a positive attitude that enabled them to reinterpret the negative situations they encountered, thus making it easier to deal with them. The mediating effect of various coping strategies on emotional health was investigated by Savitsky et al. (2020) in nursing students and by Huang et al. (2020) in nurses, who found that the use of humour reduced anxiety levels, whereas dysfunctional strategies, such as the use of alcohol or antianxiety agents and eating in excess, increased these levels. However, efficient active coping strategies, flexibility in shifting focus or perspective during work hours, and acceptance have been related to greater emotional well-being in frontline health professionals (Cai et al., 2020). To date, there are no published studies in Spain on the extent to which moral distress and coping strategies affected ICU staff emotionally during the first wave of the pandemic. Unlike Asian countries, Western countries such as Spain have had little prior experience with pandemics of this magnitude, and the situation faced was unfamiliar to health professionals.

Aim of the study
The aim of this study was to assess moral distress, possibly associated emotional health issues (anxiety and depression), and coping styles in ICU staff during the first wave of the COVID-19 pandemic.

Study design and participants
This study was based on a multicentre, cross-sectional, descriptive and correlational design.
The study setting consisted of ICUs and areas converted into ICUs in public and private hospitals in Spain during the first wave of the COVID-19 epidemic. The study population was made up of permanent ICU staff and temporary ICU staff, redeployed from other departments due to the pandemic, i.e., ICU workers who staffed these units between February and May 2020. The inclusion criteria were age ≥18 years and intensive care unit work during the pandemic. The exclusion criterion established was work in paediatric intensive care units or intermediate care units.
A consecutive, non-probabilistic sampling method was used until the sample size had been achieved. A sample size of 427 individuals was calculated to estimate a population percentage of around 50% (with a 95% confidence interval and precision of ±5 percentage points). A replacement level of 10% was estimated.
All responders were recruited through a request posted on social media pages of the International Research Project for the Humanization of Intensive Care Units (HU-CI).

Data collection
All data were collected from March to June 2020 through an online form that included the following scales: -Demographic and occupational characteristics: Questionnaire to collect information on sex, age, number of children, profession, management duties, type of hospital, usual work department, workplace during the pandemic, type of ICU cubicle, patients per day under their responsibility, work hours per week, and sick leave during the pandemic. (2021): Self-administered questionnaire to evaluate the level of moral distress, with 27 items on a Likert-type scale with five frequency scores, from 0 (never) to 4 (frequently) and five intensity scores, from 0 (none) to 4 (high). The two scores for each item are multiplied (range, 0-16) and all results are then added, yielding an overall score of moral distress (range, 0-432) where higher scores indicate greater levels of moral distress. The questionnaire also included an open-ended question where the professional could describe other situations associated with moral distress but not included in the instrument, and two additional items on any past or present thoughts of leaving the profession due to moral distress. In this study, Cronbach's alpha for this scale was 0.93. -Carver's Brief COPE (1997), adapted into Spanish by Crespo and Cruzado (1997): Self-reported questionnaire comprising 28 items to assess coping strategies according to level of use. A Likert-type scale with four choices, from 0 (not at all) to 3 (a lot). Fourteen coping styles were obtained: active coping, planning, use of instrumental support, self-distraction, venting, behavioural disengagement, positive reframing, denial, acceptance, religion, substance use (alcohol, medication), humour, and self-blame. Cronbach's alpha for this scale was 0.78. -Generalized Anxiety Disorder (GAD-7) by Spitzer, et al. (2006), Spanish version by Garcia-Campayo et al. (2010): Seven items on symptoms and disability associated with generalised anxiety with four Likert-type response options, from 0 (not at all) to 3 (nearly every day). The total score was obtained by adding all items, yielding a score between 0 and 21. According to the original authors, the total score may be categorised into four severity groups: minimal (0-4), mild (5-9), moderate (10-14), and serious (15-21). Cronbach's alpha was 0.92. -Patient Health Questionnaire (PHQ-9) by Kroenke et al., 2001;Kroenke and Spitzer, 2002;Spitzer et al., 1999), adapted into Spanish by Diez-Quevedo et al. (2001): Assessment of depressive symptoms, composed of nine items with four Likert-type response options, from 0 (not at all) to 3 (nearly every day), referring to the past two weeks. The total PHQ-9 score is between 0 and 27 (5-9 is classified as mild depression, 10-14 as moderate depression, 15-19 as moderately severe depression, and ≥20 as severe depression). Cronbach's alpha for PHQ-9 was 0.892 (Sun et al., 2020).

Data analysis
A descriptive analysis was performed on all variables. Quantitative variables are expressed as measures of central tendency and dispersion (mean, median, standard deviation [SD], and dispersion), and qualitative variables are reported as percentages and frequencies. To determine the relationship between the main variables and the demographic or occupational characteristics, Pearson's correlation coefficient and the Mann-Whitney U test were used for variables with two categories and Kruskal-Wallis test for variables with more than two categories; Pearson's chi-squared test was also used. Multivariate linear regression models were developed to explore the predictive ability of moral distress and coping on anxiety and depression. To construct the reduced model, variables of low statistical significance were discarded, controlling the beta coefficients (confounding factor). The models were tested using Snedecor's F-distribution (p < 0.05) and the adjusted R-squared.
The information gathered from the open-ended question in the Measure of Moral Distress for Health Care Professionals (MMD-HP) instrument was analysed with code-based content analysis, using NVivo-12 software.

Ethical approval
The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Clinical Research Ethics Committee of the hospital (Code 20/432-E_COVID). Prior to completing the questionnaire, all participants gave informed consent, confirming that they understood that participation was voluntary and anonymous and that they could withdraw at any time, and that their data would be completely confidential, stored, and analysed on a secure computer, and used only for the study.

Demographic characteristics
A total of 434 individuals took part in the study. The mean age was 41.33 years (SD, 9.80), and 81.8% (n = 355) were women. The occupational characteristics are listed in Table 1.

Moral distress, anxiety, and depression in the sample
Average levels of moral distress were obtained in the sample, with a mean score of 153.89 (SD, 80.34) and moderate levels of generalised anxiety and depression, with mean scores of 11.62 (SD, 5.50) and 10.71 (SD, 5.99), respectively.
Moral distress was significantly higher in women (p = 0.01), at younger ages (p = 0.04), and in staff with no management duties (p = 0.005). Higher levels of moral distress were also observed in people who had higher levels of anxiety (p < 0.05) and depression (p < 0.05), as well as those who required sick leave during the pandemic due to workrelated stress, anxiety, and/or depression (p = 0.01).
As regards professional groups, the score for moral distress was higher in nurses, although the difference between the various professions was not statistically significant. However, differences were observed according to the unit where a person was working, with higher levels of moral distress expressed by professionals working in areas converted into ICUs during the pandemic than by individuals working in normal ICUs (p = 0.04).
Anxiety and depression levels were higher among temporary ICU personnel than among permanent ICU staff, with these differences being statistically significant (anxiety, p = 0.038; depression, p = 0.009).

Results of the open-ended question about "Other situations in which you have experienced moral distress" (MMD-HP)
A total of 79 people identified "other situations", as listed below. The analysis of the response content ( Table 2) described four overall situations that could lead to moral distress: 1. Absence of family and a patient dying alone, i.e., absence of family members requiring professional interventions not considered to be part of their professional role. For instance, participant 25 (P-25) mentioned as an example: "Every time I had to accompany a patient as they took their last breath, so they would not die alone because their family couldn't come to the hospital." 2. Excessive workload, leading to situations related to (1) providing infrequent, low-quality care and (2) difficulty supervising health professionals who were recently qualified or inexperienced in critical care. P-198: "Insufficient time to provide proper care or teach inexperienced colleagues." 3. Restricted resources, categorised as: (1) Lack of PPE, increasing personal risk and lowering health care quality. P-309: "Lack of material, personal protection equipment to be able to enter as often as necessary." (2) Lack of material resources, mainly beds and ventilators. P-195: "(…) having to decide which patients can be admitted to the ICU and which ones can't (…) as it isn't possible to care for everyone." (3) Lack of sufficient human resources with adequate ICU experience. P-57: "Working in a unit transformed into an ICU for adults (…) with no ICU specialist." 4. Clinical uncertainty, i.e., a perception of providing non-evidencebased healthcare. P-147: "(…) uncertainty due to insufficient knowledge of the disease and what we were dealing with, leading to improvised healthcare." Some participants even relate this to the behaviour mentioned by P-64: "therapeutic obstinacy."

Intention to leave
In response to two MMD-HP items about leaving the profession (first item "considered leaving prior to the pandemic" and second item "considered leaving current position during the pandemic and in relation to moral distress"), 58.5% (n = 254) reported they had not considered leaving the profession in the past (before the pandemic), 35.9% (n = 156) had considered it, and 5.5% (n = 24) had actually left at some point. In these health professionals, there was a significant difference (p = 0.03) in the usual workplace: 38.4% (n = 124) of permanent ICU staff had considered leaving whereas 50.4% (n = 56) of temporary staff, redeployed from other departments had considered it.
A significant difference (p = 0.02) was also observed in the workplace of staff who considered leaving their current position during the pandemic due to moral distress (second item): more professionals working in units converted into ICUs considered leaving their position (45.2%, n = 52) than staff in established ICUs (40.2%, n = 128).
Professionals who thought about leaving the profession and those who did not also showed significant differences in the levels of anxiety

Coping styles, moral distress, and emotional health
Active coping, acceptance, emotional and instrumental support, and planning were the styles used most often by health professionals. The levels of the different coping styles are listed in Table 3.
Health professionals with the highest moral distress scores made greater use of instrumental support (p < 0.05), self-distraction (p < 0.05), venting (p < 0.05), behavioural disengagement (p < 0.05), denial (p = 0.004), substance use (p < 0.05), and self-blame (p < 0.05) than professionals with low levels of moral distress. Significant differences were also seen in coping by acceptance, such that lower levels of moral distress were associated with higher levels of acceptance (p < 0.05).

Prediction of emotional health variables
A multiple linear regression analysis was used to identify potential predictors of mental health issues in the sample. Anxiety and depression were taken as dependent variables, and variables exhibiting a significant relationship were considered independent. Table 4 lists the predictors identified by the models.
The presence of moral distress, together with the coping styles of selfdistraction, substance use, low level of acceptance, self-blame, denial, use of instrumental support, and behavioural disengagement, were the variables identified by the model as predictors of anxiety, altogether explaining 37% (AdR 2 ) of the variance.
In the case of depression, the presence of the coping styles of moral distress, such as self-blame, self-distraction, absence of positive reframing, denial, substance use, and behavioural disengagement, altogether explain 38% (AdR 2 ) of the variance.

Discussion
The pandemic experienced by health professionals is a completely new and unfamiliar situation in their professional lives for most of them. Their reactions could be considered "normal" responses to an "abnormal" situation. This in-depth study on moral distress among ICU staff explored an area that has been insufficiently researched in the Spanish population, as the main instrument has only recently been validated in the Spanish language (Rodriguez-Ruiz et al., 2021). Consistent with international findings during the pandemic (Cacchione, 2020;Kanaris, 2021;Rodríguez-Sheather and Fidler, 2021), our results report a real, albeit moderate, risk of moral distress in Spanish ICUs during the first few months of the pandemic. The risk profile was unfavourable for women, for young health professionals who did not hold management posts, and for nurses, but profession-related differences were not statistically significant. Several studies report that nurses tend to be more severely affected by moral distress Neumann et al., 2018;Whitehead et al., 2015), whereas other research finds some inconsistencies in the differences between health professionals (Epstein et al. 2019). In ICUs, interdisciplinary teamwork is absolutely essential for daily routines (Donovan et al, 2018), but even more so during a pandemic. This approach gives staff shared exposure to morally challenging situations, a common ethical climate known to be key in explaining moral distress (Atabay et al., 2015;Corley et al., 2005;Pauly et al., 2009). It also enables them to experience the stress as a group (Peiró, 2001), providing common and shared responses to the process.
Data obtained from the open-ended items of this research exemplify pandemic-related situations conducive to the development of moral distress, consistent with other qualitative studies (Dolgin et al., 2020). These situations fall into the three theoretical groups described by Hamric and Epstein (2017). The first group is designated "patient and family," and in our study, professionals found it extremely difficult to alleviate their emotional suffering through the usual work standards and practices, especially when coping with watching a patient dying alone. The second group, "unit and staff," refers to high-pressure ICU situations where actions perceived as ethically correct became difficult due to the need to coordinate with professionals who were recently qualified or who lacked experience in critical care. Third, "the system or organisation itself", covers the lack of material, human, and PPE resources that led to situations where professionals were aware of the ethically correct actions but were unable to perform them.
These situations influenced the moral distress and mental health of ICU staff, with our results showing a clear bidirectionality between the two. Consequently, professionals experiencing anxiety, depression, and/  or sick leave during the pandemic had higher levels of moral distress. Likewise, moral distress plus avoidance-oriented coping styles (e.g., selfdistraction, denial, substance use, behavioural disengagement, low selfacceptance, and high self-blame) were strong predictors of these emotional symptoms.
A study by Lazarus and Folkman (1986) shed light on how active coping helps ICU staff to deal with the problem directly, minimise the impact of the situation, regulate emotions, and reinterpret and seek an adaptive sense, whereas passive or avoidance-oriented coping can aggravate it. Furthermore, when the professional's experiences lead to moral injury as a result of performing or witnessing actions that go against the moral code, then avoidance-oriented strategies and maladaptive behaviours are often seen (Litz et al., 2009). Among our sample, moral distress was also linked with a stronger intention to leave the job, consistent with findings from other countries, such as Iran (Naboureh et al., 2020), Lithuania (Laurs et al., 2020), and the United States (Dyo, et al., 2016). These results could be relevant for the health of healthcare professionals and organisations, as the risk of an intention to leave was already present before the pandemic, especially among nurses (Heinen et al., 2012). The pandemic has even adversely affected the professional identity of student nurses (Nie et al., 2021), with this being associated with mental health issues (Jiang et al., 2019), as in our study.
These results were especially significant among health professionals redeployed to ICUs from other departments and among professionals who worked in hospital areas converted into ICUs. In addition to experiencing higher levels of moral distress, anxiety, and depression, these professionals were more likely to have considered leaving their job. Other research in Europe has not reported significant differences in anxiety levels between permanent staff and temporary workers; although the latter did show higher levels of depressive symptoms than permanent ICU staff (Altmayer et al., 2021). In this pandemic, health professionals were redeployed to ICUs, compelling them to work in an unfamiliar environment with patients of different characteristics and clinical course who required very specific critical decisions. Urgent incorporation into the ICU setting sometimes means health professionals miss training and coaching in tackling ethical dilemmas, a variable related to moral distress (Greenberg et al., 2020). Conversely, suitable airway management training is associated with lower levels of moral distress in ICUs (Golitaleb et al., 2021). From the workplace perspective, the need to improvise ICUs in previously unprepared areas also appears to have introduced new ethical and emotional challenges, confirming that the setting can influence people's well-being and that a healing and humanised environment should be created (Bosia et al., 2016;Huisman et al., 2012;Rubert et al., 2007), thus promoting the humanisation of ICUs (Velasco and Heras, 2020).

Limitations
This study has several limitations. First, due to the cross-sectional design, causal relationships could not be established between the variables. Longitudinal and prospective studies are needed to further investigate the interaction between moral distress and the variables of anxiety, depression, intention to leave, and coping. This should be done considering the population with possible mental health problems prior to the pandemic, a variable that was not controlled in this study. Second, voluntary participation via an online form could have created a biased response due to the self-selection of the sample. Last, this is an early study on moral distress in the Spanish population and used a recently validated scale; therefore, additional research is required to confirm the results.

Conclusions
This study reveals the real risk of moral distress among ICU staff during the COVID-19 pandemic. This level of distress, coupled with avoidance-oriented and passive coping styles, provides significant predictors of anxiety and depression and is related to a stronger desire to leave the job. Training in functional coping styles, ethical dilemmas, and critical care can help alleviate moral distress and reduce related harm. Mental health care of frontline health professionals during a pandemic should focus not only on ICU staff, but also on reinforcement workers who have since returned to their original roles and who may have experienced significant harm in terms of emotional well-being, as reported by our study. Further research should be undertaken to identify useful criteria for selecting these reinforcement professionals and for providing them with adequate training, follow-up, and emotional support.

Funding
This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.

Ethical statement
The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Clinical Research Ethics Committee of the hospital (Code 20/432-E_COVID). Prior to completing the questionnaire, all participants gave informed consent, confirming that they understood that participation was voluntary and anonymous and that they could withdraw at any time, and that their data would be completely confidential, stored and analyzed on a secure computer, and used only for the study.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.