ABSTRACT
Background: The consequences of moral challenges among healthcare workers (HCWs) have been increasingly recognized. To date, however, there is limited knowledge about the predictors of and associations between moral distress, moral residue, and other psychological consequences among HCWs working in a pandemic context.
Objective: We aimed to assess the levels of, and the associations between HCWs’ experiences of moral distress and moral residue with traumatic stress, burnout, and general psychological distress, and evaluate the role of empathic and compassion aspects for these outcomes, while adjusting for demographic and professional factors.
Methods: This is a cross-sectional study of 6551 Swedish HCWs participating in a survey during the COVID-19 pandemic, in the autumn of 2020. The survey included questions related to moral challenges, moral distress and residue, and self-report scales for general psychological distress (GHQ-5), traumatic stress (ITQ), compassion satisfaction and fatigue (ProQoL), empathic ability (B-IRI), and burnout (SMBQ).
Results: There were strong associations between moral challenges and moral distress and residue. There were also clear associations between moral distress and residue and psychological distress scales. Empathic ability, compassion fatigue, and compassion satisfaction were associated with moral residue whereas compassion satisfaction was not associated with moral distress. Demographic and professional predictors included gender, age, and occupational role. No interactions between moral challenges and empathic ability were found for moral distress or residue.
Discussion/conclusion: Moral distress and moral residue share some variance with both general and stress-related psychological distress; however, moral distress and residue are both theoretically and empirically distinct outcomes of moral challenges as compared to other established psychological consequences of occupational stress. Empathic ability was associated with moral distress, but not with moral residue. Moving forward, longitudinal research is needed to better understand the interplay among moral aspects relevant to HCWs.
KEYWORDS: Moral stress, moral distress, moral residue, moral challenges, health care workers, COVID-19 pandemic, psychological distress, burnout, traumatic stress, empathic ability
HIGHLIGHTS
Moral challenges, moral distress and moral residue were clearly associated with general and stress-related psychological distress.
Weak associations were found between moral distress, empathic ability, and compassion satisfaction.
Moral distress and moral residue are theoretically and empirically separate from other types of consequences of occupational stress but share variance with both general and stress-related psychological distress.
Abstract
Antecedentes: Las consecuencias de los desafíos morales entre los trabajadores de la salud (HCWs, por sus siglas en inglés) se reconocen cada vez más. Sin embargo, hasta la fecha, existe un conocimiento limitado sobre los factores predictores y las asociaciones entre el distrés moral, las huellas morales y otras consecuencias psicológicas entre los HCWs que trabajan en un contexto de pandemia.
Objetivo: Nuestro objetivo era evaluar los niveles y las asociaciones entre las experiencias del distrés moral y las huellas morales de los HCWs con el estrés traumático, el burnout y el malestar psicológico general, y evaluar el rol de los aspectos empáticos y de compasión para estos resultados, ajustando los resultados por los factores demográficos y profesionales.
Métodos: Este es un estudio transversal de 6.551 HCWs suecos que participaron en un cuestionario durante la pandemia de la COVID-19, durante el otoño de 2020. El cuestionario incluyó preguntas relacionadas con los desafíos morales, el distrés y las huellas morales, y cuestionarios de auto reporte para medir el malestar psicológico general (GHQ-5), el estrés traumático (ITQ), la satisfacción y la fatiga por compasión (ProQoL), la capacidad empática (B-IRI) y el burnout (SMBQ).
Resultados: Se observaron fuertes asociaciones entre los desafíos morales y el distrés y las huellas morales. También se observaron asociaciones claras entre el distrés moral, las huellas y las escalas de malestar psicológico. La capacidad empática, la fatiga por compasión y la satisfacción por compasión se asociaron con las huellas morales, mientras que la satisfacción por compasión no se asoció con el distrés moral. Los predictores demográficos y profesionales incluyeron el género, la edad y el rol ocupacional. No se observaron interacciones entre los desafíos morales y la capacidad empática en lo que respecta al distrés o las huellas morales.
Discusión/conclusión: El distrés moral y las huellas morales presentan cierta variación con el malestar psicológico general y la relacionada con el estrés; sin embargo, el distrés moral y las huellas morales son resultados, tanto teórica como empíricamente, distintos de los desafíos morales en comparación con otras consecuencias psicológicas establecidas del estrés laboral. La capacidad empática se asoció con el distrés moral, pero no con las huellas morales. En el futuro, se necesitarán investigaciones longitudinales para comprender mejor la interacción entre los aspectos morales relevantes para los trabajadores de la salud.
PALABRAS CLAVE: Estrés moral, distrés moral, huellas morales, COVID-19, desafíos morales, trabajadores de la salud, malestar psicológico, burnout, estrés traumático, capacidad empática
1. Introduction
In health crises, health care workers (HCWs) strive hard to provide good care to patients, despite elevated needs and a lack of time and resources, which make them prone to both moral stress and occupational stress (Anzaldua & Halpern, 2021; Sheather et al., 2022). For example, HCWs might be confronted with moral challenges such as having to choose which of two critically ill patients to attend to due to a lack of trained staff or material resources. Negative experiences of being confronted with morally stressful situations, called moral distress, have come to attention as contributing to negative consequences such as lower professional quality of life and staff turnover (Lamiani et al., 2017). Moral distress can be manifested in pondering over one’s behaviour, blaming self or others, degrading oneself, and/or somatic problems. Perceived violations of one’s own moral values have been suggested to lead to negative outcomes that are described as different from general and stress-related psychological distress; for example, moral residue and moral injury (Epstein & Hamric, 2009). However, it is unclear to what extent the consequences of moral challenges and the resulting moral distress and moral residue relate to known forms of occupational stress that are commonly observed among HCWs. Furthermore, there is a conceptual lack of clarity among the different phenomena that are used to describe the consequences of moral challenges.
A moral challenge arises when HCWs are confronted with situations that make it difficult to act according to their moral values. The HCW may experience moral stress as a response to this challenge (Grimell & Nilsson, 2020). Depending on the frequency, intensity, and duration of moral challenges and the availability of adequate support, moral stress may develop into a negative stress reaction herein called moral distress (Gustavsson et al., 2020). The distinction between moral stress, moral distress and moral residue is made in order to separate the initial, common response to moral challenges from the lingering distressing consequences that may exert a negative influence on health and functioning (see Figure 1). This distinction could be compared to a similar distinction: the one between an acute stress reaction and post-traumatic stress disorder (PTSD), in which the acute reaction is a result of exposure to a potentially traumatic event and could be classified as PTSD if the stress reaction persists over a longer time period. Moral distress can be viewed as a short-term reaction and may dissipate over time, but can also lead to long-lasting consequences called moral residue, which can be expressed as, for example, withdrawing, cynicism, and avoidant behaviour. Also, we suggest that moral distress is related to secondary psychological consequences such as burnout and compassion fatigue (Gustavsson et al., 2020). Moral residue is described by Epstein and Hamric (2009) as the accumulation of persisting moral distress from repeated exposure to morally stressful situations. As a result, moral residue might threaten moral and personal integrity and could therefore be associated with burnout. There are also several related concepts used to describe consequences of moral challenges, such as moral injury. Moral injury relates to acting, or failing to prevent, an act which contravenes own moral values and emanates from a military context related to posttraumatic stress among veterans, but has gained attention also in civilian settings during the COVID-19 pandemic, (Griffin et al., 2019; Williamson et al., 2018). According to Williamson et al. (2018), moral distress is a short-term reversible response to potentially morally injurious events, (with similarities to traumatic stress reactions), whereas moral injury is the more severe outcome with linkages to posttraumatic stress. Furthermore, the term negative/secondary consequences include both moral residue and other psychological consequences (e.g. burnout and psychological distress). The theoretical separation between moral versus psychological consequences is used in order to highlight the different sources of the (di)stress; moral challenges or occupational challenges even though the concept ‘distress’ in itself can be perceived as a psychological phenomenon.
Figure 1.
Illustration of the theoretical concepts related to moral issues and the potential timeline.
1.1. Relationships between moral distress and psychological distress
Moral distress leading to moral residue is argued to constitute a vulnerability factor for psychological distress (Lamiani et al., 2017). Although we believe that there may be an association between psychological distress and moral distress and residue, we suggest a bidirectional association. Burnout is a work-related syndrome with three related components: exhaustion, depersonalization (cynicism), and reduced sense of professional efficacy (Maslach et al., 2001). A systematic review of moral distress and its correlates found inconclusive evidence for the association between moral distress and burnout, and slightly stronger links between moral distress and decreased professional satisfaction as well as intentions to leave or resign (Lamiani et al., 2017). Furthermore, high frequency and intensity of moral distress were in one study found to be clearly related to depersonalization and emotional exhaustion, which are components of burnout syndrome (Meltzer & Huckabay, 2004). Also, as traumatic stress reactions have been linked to potentially morally injurious events during the COVID-19 pandemic (Williamson et al., 2023), it is important to evaluate the overlap between traumatic stress and moral distress and moral residue.
The professional objective for HCWs is to provide care to patients in need, where empathy plays a crucial role, but moral stress might arise when good care is difficult to provide. Empathy-based stress is a related concept, described as physical and psychological reactions resulting from empathic engagement to trauma exposure, which includes compassion fatigue, secondary traumatic stress, and vicarious trauma (Rauvola et al., 2019). Compassion fatigue stems from interactions with patients that leave the health professional preoccupied with the trauma experienced by the patient, and enhanced emotional arousal or avoidance behaviours such as withdrawing, cynicism, and/or emotional exhaustion associated with the patient (Ray et al., 2013). Emotional distress has been suggested to link difficult working conditions to compassion fatigue and to moral distress, although its role is as of yet unclear (van Mol et al., 2015). Nonetheless, empathy-based stress could indeed be induced in situations where HCWs face difficulties acting upon their empathy and could therefore be similar to moral stress and moral distress. Furthermore, other types of stress reactions (e.g. burnout) than moral stress stem from an imbalance between increased job demands and available individual resources (Bakker & Demerouti, 2007).
Positive consequences related to moral distress have also been reported, such as personal and professional growth/professional fulfillment, if moral distress is successfully dealt with (Gustavsson et al., 2022; Hanna, 2004). In turn, professional fulfillment has been reported to result in lower possibility of burnout and decreased moral distress (Dalmolin et al., 2014). Compassion satisfaction is another concept included in life quality within caring professions reflecting the positive aspect of caring for others, e.g. being proud of the achievements as a helper (Heritage et al., 2018).
Empathic ability could potentially increase the experience of moral stress, as empathy could be considered involved in moral behaviour, and possibly relate to the development of moral distress and moral residue. However, it is unclear if empathic ability constitute a risk – or protective factor in this context (Ter Heide, 2020). Empathic ability is often described as a multidimensional concept involving an affective and a cognitive dimension (Ingoglia et al., 2016). According to Morse et al. (1992), empathic ability within caring professions involves four dimensions: emotive, cognitive, behavioural and moral. Empathic ability differs from sympathy as it relates to the ‘other’ without losing the borders of self and maintaining the ability to act upon the feeling of empathy (Lamiani et al., 2020). An Italian cross-sectional study found that empathic ability increased clinicians job satisfaction and did not find empathy as a risk factor for moral distress or intention to leave the workplace (Lamiani et al., 2020).
The increased interest in consequences of moral challenges within healthcare settings has given rise to a developing field. However, less is known about moral residue as compared to moral distress. As above noted, there are several partially overlapping phenomena that describe negative reactions to moral and work-related stress. There is a lack of clarity regarding the relationship between moral distress, moral residue, and burnout, traumatic stress reactions, and general psychological distress, and there is a possibility that variations in moral distress and residue after moral challenges are equally explained by variations in these established forms of psychological distress. Hence, the underlying hypothesis for this study is to examine if and to what extent variance in moral distress and residue could be explained by: burnout, traumatic stress reactions, or general psychological distress. This knowledge would be of value to develop support tools to decrease negative consequences deriving from moral issues among HCWs, as the consequences of facing morally stressful situations might only be visible when they have also resulted in negative psychosocial outcomes, such as burnout, depression, and/or sick leave. Also, there is limited knowledge on how empathic ability and compassion satisfaction, or fatigue are related to moral distress and moral residue. Furthermore, there is a lack of clarity regarding the extent of which demographic and professional factors among HCWs are related to moral distress and residue during an ongoing pandemic.
1.2. Aims
This study has three aims: first the study aims to (1) assess the levels of moral distress, moral residue, and other forms of psychological distress among HCWs during the COVID-19 pandemic. This study also aims to (2) examine the associations between the frequency of moral challenges, moral distress and moral residue. Furthermore, this study aims to (3) assess the associations between moral distress and moral residue with different forms of psychological distress, empathic ability and compassion satisfaction/fatigue, while adjusting for demographic and professional factors during the COVID-19 pandemic. In addition, we examined whether the level of empathic ability was related to the association between moral challenges and moral distress and moral residue.
2. Methods
This study used a subset of data from a cross-sectional survey of Swedish HCWs conducted between the first and the second wave of the COVID-19 pandemic. In Sweden, the second wave begun in the end of October 2020, a time when HCWs were strained due to elevated needs and limited resources which caused situations of moral challenges and exhaustion due to longer working hours. We have previously published experiences of moral stress (Gustavsson et al., 2023) while this study focuses on moral distress, moral residue and other forms of negative consequences.
2.1. Procedure
At the request of the Swedish National Board of Health and Welfare, the Karolinska institutet developed a web course specially designed for Swedish HCWs, administrative and support staff, related to COVID-19 (The Centre for Research on Healthcare in Disasters, 2020). Between 23 September and 13 October 2020, an email invitation was extended to HCWs among the 153,300 individuals who had participated in the course, inviting them to take part in the survey. Information about the study aim and that the survey was intended for HCWs in patient care during the pandemic was included in the invitation. The survey comprised three sections: demographic and occupational questions; questions about moral challenges and their consequences; and questions about other related psychological phenomena. In all, the survey took approximately 30–45 minutes to complete. Two email reminders were sent out, and the survey was closed on November 3, 2020. For an overview of the survey questions, see (Gustavsson et al., 2023).
The survey was administered through a web-based, secure platform hosted by the Karolinska institutet (Research Electronic Data Capture) (Harris et al., 2009; Harris et al., 2019).
2.2. Participants
The participating HCWs came from all regions in Sweden. There were 23,425 HCWs who started replying to the survey, which constituted 15% of those invited. There were 6551 who dropped out before finishing the first survey section with demographic questions and there were 832 duplicate records. Of the 16,044 participants who provided complete responses to the first survey section, 11,580 responded that they had been exposed to moral challenges and were eligible for participation in this study. Those who endorsed having been exposed to moral challenges were included in this study if they had complete data on the second and third survey sections (n = 6551), representing 4.3% of the course sample, see Table 1. The included participants represent 57% of those who fulfilled the exposure criterion.
Table 1.
Characteristics of included participants exposed to moral stress and participants with incomplete survey responses.
| Characteristics | Included respondentsa | Excluded due to missing data | ||
|---|---|---|---|---|
| Exposedb | Not exposedc | Missing exposured | ||
| Total N | 6551 | 5029 | 3698 | 766 |
| Age, M (SD) | 46.2 (11.6) | 45.0 (12.1)*** | 48.0 (11.7)*** | 45.5 (12.6) |
| Sex, n (%) | ||||
| Women | 5593 (85.4) | 4290 (85.3)* | 3043 (82.3)** | 637 (83.2) |
| Men | 934 (14.3) | 628 (12.5)* | 607 (16.4)** | 107 (14.0) |
| COVID care, n (%) | ||||
| Yes | 3457 (52.8) | 2236 (44.5)*** | 983 (26.6)*** | 310 (40.5)*** |
| No | 3094 (47.2) | 2793 (55.5)*** | 2715 (73.4)*** | 456 (59.5)*** |
| Occupational group, n (%) | ||||
| Direct patient care | 5313 (81.1) | 4037 (80.3) | 2694 (72.9)*** | 596 (77.8) |
| Indirect patient caree | 455 (6.9) | 126 (2.5) | 88 (2.4)*** | 52 (6.8) |
| Coordinators | 37 (0.6) | 22 (0.4) | 16 (0.4)*** | 7 (0.9) |
| Managers | 230 (3.5) | 329 (6.5) | 383 (10.4)*** | 26 (3.4) |
| Otherf | 516 (7.9) | 463 (9.2) | 498 (13.5)*** | 77 (10.1) |
| Moral challenges, n (%) | ||||
| Never | NA | NA | 3698 (100) | NA |
| Seldom | 2526 (38.6) | 2107 (41.9)*** | NA | NA |
| Sometimes | 2714 (41.4) | 2097 (41.7)*** | NA | NA |
| Often | 974 (14.9) | 641 (12.7)*** | NA | NA |
| Very Often | 337 (5.1) | 184 (3.7)*** | NA | NA |
| Self-report scales, M (SD) | ||||
| Moral distress | 20.1 (11.3) | 20.6 (11.4) | NA | NA |
| Moral residue | 17.8 (11.0) | 18.6 (11.5)* | NA | NA |
| General psych. distress | 3.9 (3.4) | 3.9 (3.3) | 2.5 (2.7)*** | 3.4 (3.4) |
| Burnout | 71.5 (29.7) | 70.6 (27.7) | 58.0 (25.4)*** | 73.1 (25.3) |
| Traumatic stress | 4.1 (4.7) | 4.1 (4.6) | 2.4 (3.7)*** | 6.3 (4.8)* |
| Empathic ability | 33.6 (9.6) | 33.1 (9.9) | 33.4 (9.1) | 36.4 (10.4) |
| Compassion satisfaction | 25.7 (6.1) | 25.4 (6.4) | 27.2 (5.7)*** | 24.9 (6.2) |
| Compassion fatigue | 16.9 (7.0) | 17.5 (7.3)* | 13.2 (4.9)*** | 17.8 (7.2) |
Note: Each excluded group was compared against the included participants with Dunnett’s post-hoc test for continuous variables and χ2-test for categorical variables. Frequency of moral challenges was compared only for exposed excluded. P-values were adjusted for simultaneous testing within each variable with the R multcomp package v. 1.4-25. aRespondents with complete data on all included survey questions and included in analysis. bRespondents who endorsed ‘seldom’ to ‘very often’ on the question about exposure. Number of respondents for the self-report scales range from 1759 (moral distress) to 121 (traumatic stress).cRespondents who stated ‘never’ on exposure and had missing data. Number of respondents for the self-report scales range from 3501 (general psychological distress) to 2745 (traumatic stress).dRespondents with missing data on exposure. Respondents for the self-report scales range from 55 (general psychological distress) to 28 (traumatic stress).epsychologists, dentists/dental nurses, physiotherapists/occupational therapists/dieticians, audiologists/speech therapist, pharmacists, radiology nurses/medical physicist, biomedical scientists/laboratory assistants, and chiropractors.fother, which is not specified in the above.
*p ≤ .05. **p ≤ .01. ***p ≤ .001.
Table 1 shows detailed characteristics of included participants exposed to moral challenges and participants excluded due to missing data or not having been exposed to moral stress. Included respondents were involved in COVID-19 care to a higher extent than excluded respondents. Also, those included differed from the excluded non-exposed and those with missing exposure data, where those excluded were not involved in COVID-19 care, and less proportion involved in direct care professions to the same extent, while those were more involved in ‘other’ professions than those included. Otherwise, the differences between excluded and included participants were minimal. The only notable difference between the groups who reported exposure to moral challenges or not with regard to demographic and professional characteristics, was a higher mean age among those who reported no exposure to moral challenges.
The majority of the participants were employed in professions related to direct patient care, including healthcare assistants, assisting nurses, nurses, specialized nurses, and physicians. Coordinators/staffing assistants and managers were professions that were in the minority. Professions related to patient care but not directly involved in COVID-19 care included professions such as dentists/dental nurses and radiologists. Most of the respondents were women and around half had worked with direct COVID-19 care. As for the self-report scales, those exposed to moral challenges reported higher mean levels of psychological distress than those unexposed, but there were no clear indications that the included participants differed from the total eligible sample.
2.3. Survey
2.3.1. Moral challenges, moral distress, and moral residue
The survey gave participants a brief description of moral challenges, and asked them to rate whether they had experienced situations that evoked moral stress related directly to a moral challenge (see Appendix 1 for the survey questions). The responses were on a five-point scale (i.e. never, seldom, sometimes, often, or very often). Then followed questions related to moral distress and moral residue that were developed based on the results from a scoping review (Gustavsson et al., 2020) and a qualitative study (Gustavsson et al., 2022) that explored characteristics of moral stress, moral distress, and their consequences among Swedish disaster healthcare responders. The questions were based on the participants’ descriptions of perceived main moral challenges, the management of these challenges and their perceived consequences during the disaster-response deployment. The survey was piloted among four disaster-oriented healthcare responders and refined based on their comments, e.g. questions about support when coming home after deployment was removed. Thereafter, the survey questions were adapted to fit the pandemic circumstances, the survey questions were tested a second time with four Swedish HCWs with healthcare experience during the pandemic, and the questions were again adjusted based on their comments. The final set of questions on moral distress included eight questions including mulling over ones’ behaviour; blaming oneself, or others; degrading oneself; problems with sleep or somatic problems; sadness; loneliness; and using substances/alcohol. The respondents were asked to rate, on a scale from 1 (disagree) to 7 (fully agree), to what extent they felt any such lingering reactions that were related to situations that evoked moral stress. Cronbach’s alpha for the moral distress items was 0.898, indicating excellent internal consistency. The final set of questions on moral residue included nine questions including increased difficulties with empathizing with others or experiencing emotions; being more reclusive; avoiding similar situations; feeling more insecure/dependent, cynical, anxious; and relying less on one’s intuition. The respondents were asked to rate, on a scale from 1 (disagree) to 7 (fully agree), to what extent they experienced any such reactions during a prolonged time after situations that evoked moral stress. The items for each scale were summed to provide a total score. Cronbach’s alpha for the moral residue items was 0.913, indicating excellent internal consistency. See Appendix 1, for the specific survey questions. Psychometric analysis of the survey questions, including factor analysis and item-response analysis, suggests an overall good performance of the questions (unpublished manuscript).
2.3.2. Psychological distress, compassion, and empathy
The General Health Questionnaire, 5-item version (GHQ-5) was used to assess general psychological distress during the past weeks (Goldberg & Williams, 1988). The GHQ-5 included five questions about having in the past weeks: felt constantly under strain; not been able to handle one’s problems; felt constantly unhappy and depressed; lost faith in oneself; and felt worthless. The responses to each item were on a four-point scale from not at all (0) to much more than usual (3) and a total score was achieved by summing all items. GHQ-5 has been used by the Public Health agency (Public Health Agency of Sweden, 2018). Cronbach’s alpha for GHQ-5 was 0.888.
The International Trauma Questionnaire (ITQ) is a measure of symptoms of post-traumatic stress disorder (PTSD) and complex post-traumatic stress disorder (CPTSD) according to the 11th edition of the International Classification of Disorders (ICD-11) (Cloitre et al., 2018; Cloitre et al., 2021). In this study, we only included the six items assessing symptoms of PTSD: two items each for symptoms of intrusion, avoidance, and hyperarousal, and in the instructions, we asked the respondents to rate the items with reference to work-related stressful events. Respondents rated how distressing each symptom has been in the past month on a five-point scale from not at all (0) to extremely (4), and a total score was achieved by summing all six items. Cronbach’s alpha for ITQ was 0.883.
The Shirom-Melamed Burnout Questionnaire (SMBQ) included 22 items assessing burnout: physical and emotional fatigue, cognitive exhaustion, tension and energy depletion/listlessness (Almén & Jansson, 2021; Lundgren-Nilsson et al., 2012). Respondents rated how often they usually experience each sign of burnout on a seven-point scale from almost never (0) to almost always (6). A mean score was calculated to indicate the level of burnout. Cronbach’s alpha for SMBQ was 0.862.
The Professional Quality of Life (ProQOL) aims to capture the positive and negative aspects of helping those who experience trauma and suffering (Ängeby et al., 2022; Stamm, 2010). A revised version of the Professional Quality of Life (ProQOL) was used to assess the constructs compassion satisfaction and compassion fatigue (Heritage et al., 2018). Responses were given on a five-point scale from never to very often. For this revised version, the two subscale total scores were calculated according to different scoring procedures described in detail in Heritage et al. (2018). Cronbach’s alpha for compassion satisfaction was 0.901 and Cronbach’s alpha was 0.880 for compassion fatigue.
The Brief Interpersonal Reactivity Index (B-IRI) is a widely used self-report measure of disposition to empathic responsiveness (Cliffordson, 2001; Ingoglia et al., 2016). The B-IRI included 16 items answered on a 5-point Likert scale ranging from Does not describe me at all to Describes me very well and a total score was achieved by summing all items. Cronbach’s alpha was 0.845.
2.3.3. Demographic and occupational variables
The participants’ professions were categorized into five groups: professions with direct patient care, managers, coordinators, professions with indirect patient care (e.g. psychologists and dentists), and ‘other’. Age was grouped into six categories (≤19, 20–29, 30–39, 40–49, 50–64, ≥ 65 years of age). One survey question was related to professional involvement in COVID-19 patient care, with three response options: (1) worked only at workplaces with direct COVID-19 care; (2) provided care to patients but did not work with direct COVID-19 care; and (3) worked with both COVID-19 patient care and ordinary patient care. The responses were dichotomized into those who did or did not work with direct COVID-19 care.
2.3.4. Statistical analysis
Descriptive statistics were performed to summarize the data distributed on age, gender, professional group, and involvement in COVID-19 care. Frequency of moral stress, moral distress, moral residue, ITQ, SMBQ, GHQ-5, B-IRI, CS, and CF scores were then transformed to z-scores to reduce risk of multicollinearity and aid in the interpretation of regression coefficients. Bivariate correlations were presented. Linear regression models were used to assess to what extent variation in moral challenges explained variation in moral distress and moral residue and according to age, gender, professional group, and involvement in COVID-19 care. Four and five models were used for moral distress and moral residue, respectively. In these models, we added sets of predictors for each model to assess their independent main effects and present all parameter estimates to increase the informational value of these analyses. For this study, compassion fatigue and compassion satisfaction were conceptualized as socio-emotional constructs together with empathic ability rather than as psychological distress, and were thus introduced as a set of predictors into the models after the set of psychological distress variables. In the last model for each outcome, interaction effects regarding empathic ability were evaluated. The level of statistical significance was specified at p < .05. Dunnett’s post-hoc test for continuous variables and χ2-test for categorical variables were performed to compare each group of excluded participants to the included participants, see Table 1. For all statistical analyses, R version 4.2.1 was used, and the regressions were run with the package drgee version 1.1.10.
3. Results
3.1. Exposure to moral challenges affecting moral distress and moral residue
Table 1 shows the HCW endorsement of moral challenges, moral distress, and residue, as well as mean scores for the self-report scales included in this study. Experiences of moral challenges were related to moral distress and moral residue: Figure 2 displays the bivariate associations between moral stress exposure and moral distress and moral residue. The figure indicates a clear linear association between exposure to moral challenges and the outcomes of moral distress and residue, indicating that the moral distress and residue scales can capture phenomena related to such events. This is also noted in the regression analyses reported below.
Figure 2.
Distribution of moral distress and moral residue across exposure levels of morally stressful situations.
3.2. Associations of moral distress, moral residue, and psychological distress
Overall, the bivariate correlations among the moral and psychological distress measures were strong, and similar for individuals involved in and those not involved in COVID-19 care (Figure 3). Moral distress and moral residue were strongly associated with burnout, diminished general psychological health, traumatic stress, and compassion fatigue, whereas their associations with empathic ability and compassion satisfaction were small to moderate.
Figure 3.
Correlations between moral distress/residue, psychological distress, empathic ability, and compassion satisfaction/fatigue. The left panel shows correlations for participants working in COVID-19 care, and the right panel shows not working in COVID-19 care. BIRI = Brief Interpersonal Reactivity Index, CF = Compassion fatigue, CS = Compassion satisfaction, GHQ5 = General Health Questionnaire, 5 item version, ITQ = International Trauma Questionnaire, MD = Moral distress, MFRQ = Frequency of moral challenges, MR = Moral residue, SMBQ = Shilom-Melamed Burnout Questionnaire.
3.3. Moral distress and its association with psychological distress
Table 2 shows the regression analyses in which moral distress was the outcome. Across the models, frequency of morally stressful events was a significant predictor for moral distress, although the addition of psychological distress measures in Model 3 slightly attenuated its predictive power. Higher age was associated with lesser distress, although the pattern was somewhat inconsistent. Women had higher levels of moral distress than men, and the association was attenuated when adjusting for other forms of psychological distress.
Table 2.
Predictors of moral distress among 6551 Swedish health care workers during the COVID-19 pandemic.
| Predictor | Model 1 | Model 2 | Model 3 | Model 4 | ||||
|---|---|---|---|---|---|---|---|---|
| β | CI | β | CI | β | CI | β | CI | |
| Intercept | −0.000 | −0.022; 0.022 | −0.191 | −0.291; – 0.092 | −0.140 | −0.218; – 0.061 | −0.103 | −0.181; – 0.025 |
| Moral challenges | 0.439 | 0.416; 0.462 | 0.417 | 0.394; 0.440 | 0.232 | 0.212; 0.253 | 0.201 | 0.180; 0.221 |
| Age 17–19 years | 0.096 | −0.226; 0.418 | −0.204 | −0.465; 0.056 | −0.147 | −0.391; 0.096 | ||
| Age 20–29 years | 0.281 | 0.196; 0.365 | 0.119 | 0.050; 0.188 | 0.127 | 0.060; 0.194 | ||
| Age 30–39 years | 0.146 | 0.079; 0.212 | 0.025 | −0.029; 0.080 | 0.027 | −0.026; 0.079 | ||
| Age 40–49 years (ref.) | ||||||||
| Age 50–64 years | −0.142 | −0.196; – 0.089 | −0.043 | −0.086; – 0.000 | −0.081 | −0.123; – 0.039 | ||
| Age 65–90 years | −0.305 | −0.435; – 0.174 | −0.067 | −0.173; 0.039 | −0.105 | −0.208; – 0.001 | ||
| Women | 0.098 | 0.039; 0.157 | 0.046 | −0.001; 0.093 | 0.051 | 0.004; 0.098 | ||
| Other gender | 0.001 | −0.300; 0.301 | −0.118 | −0.453; 0.217 | −0.072 | −0.369; 0.226 | ||
| Men (ref.) | ||||||||
| Direct patient care | 0.098 | 0.019; 0.176 | 0.061 | −0.001; 0.123 | 0.050 | −0.011; 0.111 | ||
| Manager | 0.008 | −0.117; 0.133 | 0.065 | −0.029; 0.160 | 0.078 | −0.014; 0.169 | ||
| Coordinator | 0.151 | −0.127; 0.429 | 0.023 | −0.164; 0.210 | 0.064 | −0.139; 0.266 | ||
| Other profession | 0.027 | −0.080; 0.134 | −0.030 | −0.116; 0.057 | −0.016 | −0.101; 0.070 | ||
| Indirect patient care (ref.) | ||||||||
| COVID-19 care | 0.069 | 0.025; 0.113 | 0.105 | 0.069; 0.141 | 0.071 | 0.036; 0.106 | ||
| non COVID-19 care (ref.) | ||||||||
| General psych. distress | 0.255 | 0.225; 0.285 | 0.211 | 0.181; 0.240 | ||||
| Burnout | 0.178 | 0.149; 0.207 | 0.075 | 0.043; 0.107 | ||||
| Traumatic stress | 0.231 | 0.206; 0.255 | 0.140 | 0.114; 0.166 | ||||
| Empathic ability | 0.034 | 0.014; 0.054 | ||||||
| Compassion satisfaction | 0.015 | −0.007; 0.036 | ||||||
| Compassion fatigue | 0.270 | 0.239; 0.302 | ||||||
| Adj. R2 | 0.192 | 0.218 | 0.489 | 0.520 | ||||
Note: All continuous predictors were transformed to z-scores. Statistically significant associations are denoted in boldface.
Professions working with direct patient care reported a higher degree of moral distress, and participants working with COVID-19 care had higher levels as compared to those not working with COVID-19 care. However, although all models in Table 2 had statistically significant additions to overall variance explained (ps < .01) the variables in Model 2 added only slightly to the variance explained. When adding the psychological measures in Model 3, overall variance explained increased markedly. All measures were significantly associated with moral distress, although burnout seemed to show a lower association with moral distress as compared with general psychological distress and trauma-related distress. Empathic ability and compassion fatigue, but not compassion satisfaction, were each associated with moral distress and provided slight improvement of explained variance (Table 2, Model 4). We also evaluated an interaction term, Empathic ability × Moral challenges, but this interaction was not statistically significant, β = 0.014, 95% CI -0.004, 0.033, p = .13 and the other model parameters were essentially unchanged from Model 4.
3.4. Moral residue and its association with psychological distress
Table 3 shows the regression analyses in which moral residue was the outcome. Here, moral distress was included as a predictor in all models as it is seen as a precursor to moral residue. Thus, the models show associations adjusted for moral distress. For completeness, Models 1–4 in Table 3 were replicated while excluding moral distress in Table A1 in Appendix 2. The pattern of associations suggests that the link between exposure to moral challenges and moral residue in large parts was explained by moral distress. The results suggest that younger age was associated with lower levels of moral residue. Female gender was associated with lower levels of moral residue, however, the association between gender and moral residue was clearly attenuated in models where empathic ability and compassion satisfaction/fatigue were included. Working with COVID-19 care and profession was associated with moral residue: direct patient care was consistently associated with higher levels of moral residue across models, and the group of other professions showed some association with moral residue, while slightly lower levels were reported among managers. Overall, however, the demographic and professional predictors in Model 2 added very little to the amount of explained variance in Model 1, although the additional variance explained were statistically significant for all models in Table 3 (ps < .05).
Table 3.
Predictors of moral residue among 6551 Swedish health care workers during the COVID-19 pandemic.
| Predictor | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | CI | β | CI | β | CI | β | CI | β | CI | |
| Intercept | 0.000 | −0.017; 0.017 | 0.056 | −0.019; 0.131 | 0.044 | −0.026; 0.113 | 0.002 | −0.066; 0.071 | 0.001 | −0.068; 0.070 |
| Moral challenges | 0.027 | 0.005; 0.049 | 0.031 | 0.009; 0.052 | 0.008 | −0.012; 0.028 | −0.001 | −0.021; 0.018 | −0.000 | −0.020; 0.019 |
| Moral distress | 0.702 | 0.677; 0.726 | 0.706 | 0.681; 0.731 | 0.505 | 0.473; 0.537 | 0.478 | 0.445; 0.511 | 0.477 | 0.444; 0.509 |
| Age 17–19 years | −0.234 | −0.446; −0.021 | −0.389 | −0.593; −0.184 | −0.342 | −0.526; −0.157 | −0.347 | −0.532; −0.161 | ||
| Age 20–29 years | −0.094 | −0.160; −0.029 | −0.133 | −0.194; −0.072 | −0.129 | −0.188; −0.070 | −0.129 | −0.188; −0.070 | ||
| Age 30–39 years | 0.017 | −0.036; 0.070 | −0.025 | −0.074; 0.025 | −0.035 | −0.084; 0.013 | −0.035 | −0.084; 0.013 | ||
| Age 40–49 years (ref.) | ||||||||||
| Age 50–64 years | −0.006 | −0.049; 0.036 | 0.026 | −0.014; 0.067 | −0.001 | −0.041; 0.039 | −0.000 | −0.040; 0.039 | ||
| Age 65–90 years | −0.062 | −0.148; 0.025 | 0.023 | −0.060; 0.105 | −0.012 | −0.092; 0.068 | −0.011 | −0.091; 0.069 | ||
| Women | −0.069 | −0.114; −0.025 | −0.082 | −0.123; −0.040 | −0.027 | −0.068; 0.014 | −0.027 | −0.068; 0.014 | ||
| Other gender | 0.201 | −0.076; 0.479 | 0.136 | −0.153; 0.425 | 0.194 | −0.087; 0.476 | 0.202 | −0.081; 0.485 | ||
| Men (ref.) | ||||||||||
| Direct patient care | 0.050 | −0.008; 0.108 | 0.053 | −0.001; 0.107 | 0.068 | 0.015; 0.121 | 0.068 | 0.015; 0.121 | ||
| Manager | −0.111 | −0.200; −0.021 | −0.077 | −0.158; 0.005 | −0.054 | −0.134; 0.025 | −0.055 | −0.134; 0.025 | ||
| Coordinator | −0.066 | −0.299; 0.167 | −0.101 | −0.318; 0.116 | −0.086 | −0.309; 0.137 | −0.086 | −0.306; 0.135 | ||
| Other profession | 0.082 | 0.002; 0.162 | 0.059 | −0.016; 0.134 | 0.074 | 0.001; 0.148 | 0.073 | −0.001; 0.147 | ||
| Indirect patient care (ref.) | ||||||||||
| COVID-19 care | −0.055 | −0.089; −0.020 | −0.020 | −0.052; 0.013 | −0.032 | −0.064; −0.000 | −0.032 | −0.063; −0.000 | ||
| non COVID-19 care (ref.) | ||||||||||
| General psych. distress | 0.159 | 0.128; 0.189 | 0.133 | 0.103; 0.163 | 0.133 | 0.103; 0.163 | ||||
| Burnout | 0.112 | 0.085; 0.139 | 0.022 | −0.007; 0.051 | 0.022 | −0.007; 0.051 | ||||
| Traumatic stress | 0.112 | 0.088; 0.137 | 0.098 | 0.073; 0.123 | 0.098 | 0.073; 0.123 | ||||
| Empathic ability | −0.046 | −0.065; −0.027 | −0.045 | −0.064; −0.026 | ||||||
| Compassion satisfaction | −0.094 | −0.113; −0.074 | −0.094 | −0.114; −0.075 | ||||||
| Compassion fatigue | 0.150 | 0.118; 0.181 | 0.149 | 0.118; 0.181 | ||||||
| Empathic ability × moral challenges | −0.020 | −0.043; 0.003 | ||||||||
| Empathic ability × moral distress | 0.015 | −0.008; 0.037 | ||||||||
| Adj. R2 | 0.510 | 0.513 | 0.573 | 0.593 | 0.593 | |||||
Note: All continuous predictors were transformed to z-scores. Statistically significant associations are denoted in boldface.
When adding the psychological self-report measures in Model 3, overall variance explained increased markedly. All measures were associated with moral residue, although burnout seemed to show a less consistent association with moral residue as compared with general psychological distress and trauma-related distress in the following models, which might be related to the overlap between burnout and compassion fatigue. Also, the addition of these predictors meant that the association between moral distress and moral residue diminished markedly, although still remaining significant.
Adding measures of empathic ability and compassion satisfaction/fatigue in Model 4 indicated that increased empathic ability and compassion satisfaction were related to lower moral residue and compassion fatigue was associated with higher levels of moral residue. These measures added meaningfully to the overall explained variance. They did not, however, decrease the association between moral distress and residue.
Finally, in Model 5 we added two interaction terms to assess more in detail whether empathic ability interacted with frequency of moral challenges or with levels of moral distress. Here, no statistically significant interaction effects were found.
4. Discussion
This study assessed the associations between negative consequences of moral stress with burnout, traumatic stress and general psychological distress, and how these consequences are related to empathic ability and compassion satisfaction and fatigue in a large sample of HCWs during the COVID-19 pandemic in Sweden. Furthermore, this study assessed the role of demographic and occupational factors related to moral distress, moral residue, and various forms of psychological distress. We found that moral distress and moral residue, as assessed in this study, are consistently linked to frequency of moral challenges, and that this clear association remains in light of a substantial overlap between both moral distress and residue on the one hand, and psychological distress on the other. The pattern of associations suggests that the link between exposure to moral challenges and moral residue in large parts was explained by moral distress, although a full mediation analysis was beyond the scope of this study. Empathic ability and compassion satisfaction and fatigue showed clear associations with moral residue whereas the role of empathic ability as protective for moral residue was unclear.
Moral distress and residue are concepts that theoretically are easy to distinguish from other forms of general and stress-related psychological distress, because of the centrality of moral aspects. Nonetheless, this study found that moral distress and residue overlap to various degrees with other forms of psychological distress: burnout, traumatic stress, and general psychological distress. The associations were substantial, and clearly larger than what has been reported before (Spilg et al., 2022). In this study, traumatic stress reactions seemed to be more strongly correlated with moral distress as compared with moral residue. We speculate that the strength of association between moral outcomes on the one hand and traumatic stress outcomes on the other is context dependent; in this study, the pandemic was ongoing, and we suggest that the healthcare context was characterized by ongoing moral challenges. Another study among HCWs during the pandemic in the UK, which examined moral injury and psychological wellbeing, found that potentially morally injurious events were associated with adverse mental health symptoms (Williamson et al., 2023). However, there is far from a total overlap, and taken together with previous findings, it seems clear that moral distress and moral residue captures something unique, noted also in the association between frequency of exposure to moral challenges and these outcomes. We also note that different types of psychological distress such as burnout, compassion fatigue, depression and anxiety are difficult to clearly distinguish since they often overlap. Indeed, the associations among the various forms of psychological distress were slightly higher as compared with their associations with moral distress and moral residue.
Empathic ability was only weakly associated with moral distress and residue, which is similar to the results from an Italian study (Lamiani et al., 2020). Furthermore, empathic ability has weak correlations with other forms of psychological distress as well. This could be valuable to investigate further, especially within HCWs as those constitute professions that need empathic ability in their daily work with patients. As for the professional quality of life, compassion satisfaction also had weak correlations with psychological distress and moral distress and residue in this study. This is in line with another study which underlined that professional fulfillment could be a protective factor for burnout and moral distress (Dalmolin et al., 2014).
Compassion satisfaction and empathic ability could be phenomena that give value to one’s professional life, and in that sense, may function as a protective factor for moral distress, moral residue, and psychological distress. However, as captured in a French qualitative study, for empathy to be a protective for burnout there needs to be a well-balanced empathy, to avoid overload at work (Picard et al., 2016). The relationships between compassion, empathy and moral and psychological outcomes seem to be complex, and future research may benefit from analytic designs that evaluate nonlinear patterns of association as well as potential interactions with other predictors. Nonetheless, our findings indicate that compassion and empathy to some extents are relevant to moral outcomes.
Even though this study found that moral residue overlaps with psychological distress, the overlap is similar to associations among different constructs of psychological distress. We therefore suggest that moral residue is a separate, but related, concept as compared to other forms of psychological distress. We are cognizant of the fact that moral residue can be seen as a particular form of psychological distress, but we suggest that this distinction is helpful to further the understanding of the consequences of moral challenges. We also suggest that empathic ability and compassion satisfaction and fatigue are relevant as risk and protective factors for moral distress and residue. It is important to note, however, that this cross-sectional study does not provide any support to the causal sequence, and the directionality of some components remains to be determined. Further research is needed in this area, where a conceptual model of the theoretical concepts and its relation to psychological distress-components and risk and protective factors could be beneficial for conceptual clarity.
To conclude, the findings highlight that moral challenges can lead to consequences for HCWs that are either moral as in moral residue, and/or psychological in the form of stress related and general psychological distress. These consequences can be developed depending on available risk and protective factors. Even though moral residue and psychological distress overlap and are difficult to empirically separate, there is a theoretical difference. This knowledge could be of value for both HCWs and for health care organizations when developing support to HCWs; adequate support for managing moral challenges and consequences of moral stress should be provided, which might not be included in current formal support offerings that are focusing merely on mental health issues. Furthermore, it could be of value for health care organizations to promote HCWs professional quality of life since that seems to be protective against negative consequences among HCWs.
4.1. Methodological considerations
This is a cross-sectional study, which provides a picture of self-reported psychological health during a specific period when HCWs were especially strained, which should be considered when interpreting the findings. The questionnaires included in this study reflect participants’ self-reported health during the past weeks, and we lack information about participants’ psychological health from the first wave of, and before, the pandemic. Related to this, we suggest that the measure of empathic ability reflects an ability that is stable over time. However, the context may have influenced the empathic ability ratings and thus limited the possibility to detect interactions with moral challenges. Furthermore, we note that repeated exposure to moral challenges and to other work-related stressors may lead to feedback effects over time, for example, that high levels of work-related stress during the first wave of the pandemic could influence exposure to moral challenges and resulting moral stress. The existence and scope of these feedback effects are uncertain and need longitudinal designs to be detected.
Similarly, as noted above, we conceptualized moral distress and moral residue as temporally distinct, and made efforts to prompt the respondents to consider moral distress as more brief reactions and moral residue as long-term consequences. Nonetheless, we found a strong correlation between these concepts, and although they are conceptually distinct, longitudinal studies are needed to shed light on whether this temporal distinction is borne out empirically. Furthermore, we acknowledge the conceptual ambiguity between several concepts related to moral issues in the literature (e.g. moral distress, moral injury). We believe that this in part reflects a developing field, and we encourage further research to investigate the conceptual overlap.
Although there were no clear indications of bias due to survey dropout, it is difficult to ascertain lack of bias, for example, related to fatigue, which is reflected also in the measures in the study. The recruitment method may have led to selection bias, calling into question whether the sample is representative of HCWs during the pandemic. Also, generalizability might be decreased due to that those who partook in the survey were only those who had participated in a COVID-19 training. HCWs that experienced strong reaction might be the ones who tended to participate; however, they might not have had the energy to participate. The attrition pattern suggests that there is a risk that participants with the highest levels of fatigue are underrepresented. However, the comparison of each group of excluded participants to the included participants indicates that the probable reasons for dropping out of the survey were due to being less exposed – either due to not having been involved in COVID-19 care, or that it was less relevant due to being exposed to less moral challenges. Further, although the large sample size offered statistically significant differences between included and excluded respondents regarding age, gender, moral residue, and compassion fatigue, the differences were small or negligible. The risk of selection bias may or may not introduce systematic errors with regard to the associations that were of primary interest in this study. Furthermore, the response rate was low, however also difficult to ascertain since we did not have information about the true eligible participants that were among the participants that the survey was sent to: we tried to recruit only direct patient care HCWs, and we do not have information about how many from the invited sample fulfilled this criterion. Also, the high proportion of women in this study should be mentioned. However, professions providing direct care represent the largest professional group in this study. Among these professions; nurses, nursing assistants and health care assistants, women constitute the majority in this country’s statistics ranging from 72-90%, where nursing assistants represents the highest proportion (90%) of women.
The high correlation with psychological distress may indicate overlap between moral phenomena and mental health phenomena, but also that more work is needed to evaluate the questions needed to measure moral issues. These questions were chosen due to their fit to the disaster context, which characterized the pandemic setting during this period. The items were the result of a qualitative study in which perception and management of moral issues were described by Swedish disaster-health responders and piloted in two phases as described. Nonetheless, more validation is needed, for example, examining the content, criterion-related, construct, and face validity. Similarly, although Cronbach’s Alpha points to good internal consistency for each subset of items, further reliability testing is advised through test-retest and parallel or alternate forms for these items. As mentioned earlier, factor analysis and item-response analysis point to an overall good performance of the questions, although these analyses are not yet published.
5. Conclusions
This study provides additional knowledge that moral distress and residue are separate constructs from other forms of psychological distress usually related to occupational stress in HCWs, although there is some overlap between moral distress and moral residue on the one hand and general and stress-related psychological distress on the other. Empathic ability and compassion satisfaction and fatigue were to some extent associated with moral distress and moral residue, whereas no buffering effect of empathic ability was found. In summary, the constructs of moral distress and moral residue seem to be useful to capture unique moral issues resulting from moral challenges among HCWs that cannot be fully explained as psychological distress. Moving forward, we suggest longitudinal studies are needed to further our understanding of how moral challenges lead to moral distress and moral residue.
Supplementary Material
Acknowledgements
Special recognition to the participants. Further acknowledgement to Johan Zetterqvist for support with statistical analysis.
Appendix 1. , Survey Questions
Moral Stress
In this first part, the questions are about your experiences of difficult situations in your professional role at work during the COVID-19 pandemic.
Difficult situations and choices at work
On this page, the questions are about your experience of difficult situations in your professional role at work during the COVID-19 pandemic. The following questions are about dealing with difficult situations and their consequences.
Some situations may mean that you cannot follow, and act upon, your moral values. These situations may give rise to moral stress, e.g., feelings such as powerlessness, frustration, helplessness, anger/sadness. The situations may, for example, be that you have needed to make decisions even though the options available to you seemed wrong, or where you have been prevented by circumstances from doing what is in line with your values, or where you have been involved in another’s action or decision that goes against your beliefs.
O Never
O Rarely
O Sometimes
O Often
O Very often
To what extent do you feel that you have been involved in such situations?
Reactions after moral stress
We now turn to the consequences and reactions that can arise from situations of moral stress. Moral stress may remain after the situation(s) and can give rise to various types of reactions. How well do these apply to you?
Think about the big picture, even if you have been in several types of situations and even if you still experience moral stress in your work. Indicate the extent to which you have noticed following reactions/consequences, related to the feelings of moral stress that arose in the situations.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
| I ruminated/agonized over my action(s) | O | O | O | O | O | O | O |
| I blamed myself | O | O | O | O | O | O | O |
| I disparaged myself | O | O | O | O | O | O | O |
| I blamed others (anger, bitterness) | O | O | O | O | O | O | O |
| I had physical symptoms (e.g., difficulty sleeping, stomach pain) | O | O | O | O | O | O | O |
| I was sad and felt depressed | O | O | O | O | O | O | O |
| I felt lonely/isolated/misunderstood and withdrew | O | O | O | O | O | O | O |
| I self-medicated/used alcohol or other substances to numb the feeling |
O |
O |
O |
O |
O |
O |
O |
| Other: | |||||||
| (Other:) |
Rate what you have written under “Other” as regards whether you have noticed this reaction/consequence,
related to the feelings/emotions that arose in the situations.
O 1
O 2
O 3
O 4
O 5
O 6
O 7
Rate what you have written under “Other” as regards whether you have noticed this reaction/ consequence, related to the feelings/emotions that arose in the situations. Rate as not at all to very much (1=not at all) (7=very much)
Consequences of moral stress
These reactions can, in some cases, linger for a longer time afterwards. In what way do you think this has affected you today?
Questions regarding positive impacts will be posed in the next section. Please indicate how well you think the following apply to you, rating each option from 1 to 7.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
| I find it harder to feel empathy | O | O | O | O | O | O | O |
| I find it harder to feel emotions | O | O | O | O | O | O | O |
| I feel more withdrawn | O | O | O | O | O | O | O |
| I avoid similar situations | O | O | O | O | O | O | O |
| I feel more insecure/ independent | O | O | O | O | O | O | O |
| I don’t listen to my gut feeling as much | O | O | O | O | O | O | O |
| I feel more irritable | O | O | O | O | O | O | O |
| I feel more cynical | O | O | O | O | O | O | O |
| I have become more anxious |
O |
O |
O |
O |
O |
O |
O |
| Other: | |||||||
| (Other:) |
Please indicate how much you think what you have written under “Other” applies to you
O 1
O 2
O 3
O 4
O 5
O 6
O 7
Please indicate how much you think what you have written applies to you, rating each option from 1 to 7, not at all to very much (1=not at all) (7=very much)
Appendix 2.
Table A1.
Predictors of moral residue among 6551 Swedish health care workers during the COVID-19 pandemic, without adjusting for moral distress.
| Predictor | Model 1 | Model 2 | Model 3 | Model 4 | ||||
|---|---|---|---|---|---|---|---|---|
| β | CI | β | CI | β | CI | β | CI | |
| Intercept | 0.000 | −0.023; 0.023 | −0.079 | −0.185; 0.027 | −0.027 | −0.108; 0.054 | −0.047 | −0.127; 0.033 |
| Moral challenges | 0.335 | 0.310; 0.361 | 0.325 | 0.299; 0.351 | 0.125 | 0.104; 0.146 | 0.095 | 0.074; 0.115 |
| Age 17–19 years | −0.166 | −0.417; 0.085 | −0.492 | −0.712; – 0.272 | −0.412 | −0.600; – 0.224 | ||
| Age 20–29 years | 0.104 | 0.017; 0.191 | −0.073 | −0.142; – 0.004 | −0.068 | −0.134; – 0.002 | ||
| Age 30–39 years | 0.120 | 0.049; 0.192 | −0.012 | −0.069; 0.045 | −0.023 | −0.078; 0.033 | ||
| Age 40–49 years (ref.) | ||||||||
| Age 50–64 years | −0.107 | −0.163; – 0.050 | 0.004 | −0.042; 0.050 | −0.040 | −0.085; 0.005 | ||
| Age 65–90 years | −0.277 | −0.396; – 0.158 | −0.011 | −0.108; 0.086 | −0.062 | −0.157; 0.033 | ||
| Women | 0.000 | −0.061; 0.062 | −0.058 | −0.106; – 0.010 | −0.003 | −0.049; 0.044 | ||
| Other gender | 0.202 | −0.148; 0.551 | 0.076 | −0.298; 0.451 | 0.160 | −0.179; 0.500 | ||
| Men (ref.) | ||||||||
| Direct patient care | 0.119 | 0.035; 0.202 | 0.084 | 0.020; 0.148 | 0.092 | 0.030; 0.155 | ||
| Manager | −0.105 | −0.233; 0.023 | −0.044 | −0.137; 0.049 | −0.017 | −0.108; 0.073 | ||
| Coordinator | 0.041 | −0.249; 0.331 | −0.089 | −0.303; 0.125 | −0.056 | −0.269; 0.158 | ||
| Other profession | 0.101 | −0.011; 0.214 | 0.044 | −0.045; 0.133 | 0.067 | −0.020; 0.154 | ||
| Indirect patient care (ref.) | ||||||||
| COVID-19 care | −0.006 | −0.053; 0.041 | 0.033 | −0.004; 0.071 | 0.002 | −0.034; 0.038 | ||
| non COVID-19 care (ref.) | ||||||||
| General psych. distress | 0.287 | 0.253; 0.321 | 0.233 | 0.200; 0.267 | ||||
| Burnout | 0.202 | 0.172; 0.232 | 0.058 | 0.025; 0.091 | ||||
| Traumatic stress | 0.229 | 0.202; 0.256 | 0.165 | 0.136; 0.193 | ||||
| Empathic ability | −0.029 | −0.051; – 0.008 | ||||||
| Compassion satisfaction | −0.087 | −0.109; – 0.065 | ||||||
| Compassion fatigue | 0.279 | 0.244; 0.314 | ||||||
| Adj. R2 | 0.112 | 0.124 | 0.442 | 0.483 | ||||
Note: All continuous predictors were transformed to z-scores. Statistically significant associations are denoted in boldface.
Funding Statement
This work was supported by a grant from the National Board of Health and Welfare (Socialstyrelsen), Akademiska University Hospital, and a faculty funding (KID) grant from Karolinska Institutet.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Ethical consideration
Ethical permission was obtained for this study from the Regional Ethics Review Board, Stockholm (2017-12-21, DNR: 2017/2182-31), with approved amendments 2020-06-26, DNR: 2020-03161. Information about the study aim, process, voluntary participation, confidentiality of the participants, and that participants could withdraw from the study at any point in time was provided in the email where the survey-link was included. Thus, accepting the invitation to the survey was an agreement of participation. All methods were carried out in accordance with relevant guidelines and regulations.
Contributor details
All authors (MG, JvS, NJ, FA) were involved in the idea and design of the study. The first author (MG) managed data collection, and development of the manuscript, in close collaboration with the last author (FA). The last author (FA) led the data analysis process, in collaboration with the first author (MG). All authors (MG, JvS, NJ, FA) read and agreed to the final version of the manuscript.
Data availability statement
Additional data is available upon reasonable request to the authors.
Supplemental Material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/20008066.2025.2512677.
References
- Almén, N., & Jansson, B. (2021). The reliability and factorial validity of different versions of the Shirom-Melamed Burnout Measure/Questionnaire and normative data for a general Swedish sample. International Journal of Stress Management, 28, 314–325. 10.1037/str0000235 [DOI] [Google Scholar]
- Anzaldua, A., & Halpern, J. (2021). Can clinical empathy survive? Distress, burnout, and malignant duty in the age of COVID-19. The Hastings Center Report, 51(1), 22–27. 10.1002/hast.1216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ängeby, K., Rubertsson, C., Hildingsson, I., & Edqvist, M. (2022). Self-compassion and professional quality of life among midwives and nurse assistants: A cross-sectional study. European Journal of Midwifery, 6, 47. 10.18332/ejm/149520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bakker, A. B., & Demerouti, E. (2007). The Job demands-resources model: State of the art. Journal of Managerial Psychology, 22(3), 309–328. 10.1108/02683940710733115 [DOI] [Google Scholar]
- The Centre for Research on Healthcare in Disasters, K. I. (2020). Web courses in COVID-19 for Swedish healthcare workers and support staff. https://utbildning.ki.se/e-utbildningar-om-covid-19 [Swedish].
- Cliffordson, C. (2001). Assessing empathy: Measurement characteristics and interviewer effects. (Doctoral thesis), Acta Universitatis Gothoburgensis, Göteborg. http://urn.kb.se/resolve?urn=urn:nbn:se:hv:diva-2114.
- Cloitre, M., Hyland, P., Prins, A., & Shevlin, M. (2021). The International Trauma Questionnaire (ITQ) measures reliable and clinically significant treatment-related change in PTSD and complex PTSD. European Journal of Psychotraumatology, 12(1), 1930961. 10.1080/20008198.2021.1930961 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cloitre, M., Shevlin, M., Brewin, C. R., Bisson, J. I., Roberts, N. P., Maercker, A., Karatzias, T., & Hyland, P. (2018). The International Trauma Questionnaire: development of a self-report measure of ICD-11 PTSD and complex PTSD. Acta Psychiatrica Scandinavica, 138(6), 536–546. 10.1111/acps.12956 [DOI] [PubMed] [Google Scholar]
- Dalmolin, G., Lunardi, V., Lunardi, G., Barlem, E., & Silveira, R. (2014). Moral distress and burnout syndrome: Are there relationships between these phenomena in nursing workers? Revista latino-americana de enfermagem, 22(1), 35–42. 10.1590/0104-1169.3102.2393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein, E. G., & Hamric, A. B. (2009). Moral distress, moral residue, and the crescendo effect. The Journal of Clinical Ethics, 20(4), 330–342. [PubMed] [Google Scholar]
- Goldberg, D., & Williams, P. (1988). A user's guide to the general health. NFER-NELSON. [Google Scholar]
- Griffin, B. J., Purcell, N., Burkman, K., Litz, B. T., Bryan, C. J., Schmitz, M., … Maguen, S. (2019). Moral injury. An Integrative Review. J Trauma Stress, 32(3), 350–362. 10.1002/jts.22362 [DOI] [PubMed] [Google Scholar]
- Grimell, J., & Nilsson, S. (2020). An advanced perspective on moral challenges and their health-related outcomes through an integration of the moral distress and moral injury theories. Military Psychology, 32(6), 380–388. 10.1080/08995605.2020.1794478 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gustavsson, M. E., Arnberg, F. K., Juth, N., & von Schreeb, J. (2020). Moral distress among disaster responders: What is it? Prehospital and Disaster Medicine, 35(2), 1–8. 10.1017/S1049023X20000096 [DOI] [PubMed] [Google Scholar]
- Gustavsson, M. E., Juth, N., Arnberg, F. K., & von Schreeb, J. (2022). Dealing with difficult choices: A qualitative study of experiences and consequences of moral challenges among disaster healthcare responders. Conflict and Health, 16(1), 24. 10.1186/s13031-022-00456-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gustavsson, M. E., Juth, N., von Schreeb, J., & Arnberg, F. K. (2023). Moral stress among Swedish Health Care workers during the COVID-19 pandemic: A cross-sectional study. Scandinavian Journal of Work and Organizational Psychology, 8(1), 1–13. 10.16993/sjwop.170 [DOI] [Google Scholar]
- Hanna, D. R. (2004). Moral distress: The state of the science. Research and Theory for Nursing Practice, 18(1), 73–93. [DOI] [PubMed] [Google Scholar]
- Harris, P. A., Taylor, R., Minor, B. L., Elliott, V., Fernandez, M., O'Neal, L., … Duda, S. N. (2019). The REDCap consortium: Building an international community of software platform partners. Journal of Biomedical Informatics, 95, 103208. doi: 10.1016/j.jbi.2019.103208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Harris, P. A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., & Conde, J. G. (2009). Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heritage, B., Rees, C. S., & Hegney, D. G. (2018). The ProQOL-21: A revised version of the Professional Quality of Life (ProQOL) scale based on Rasch analysis. PLoS One, 13(2), e0193478. 10.1371/journal.pone.0193478 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ingoglia, S., Lo Coco, A., & Albiero, P. (2016). Development of a Brief Form of the Interpersonal Reactivity Index (B-IRI). Journal of Personality Assessment, 98(5), 461–471. 10.1080/00223891.2016.1149858 [DOI] [PubMed] [Google Scholar]
- Lamiani, G., Borghi, L., & Argentero, P. (2017). When healthcare professionals cannot do the right thing: A systematic review of moral distress and its correlates. Journal of Health Psychology, 22(1), 51–67. 10.1177/1359105315595120 [DOI] [PubMed] [Google Scholar]
- Lamiani, G., Dordoni, P., Vegni, E., & Barajon, I. (2020). Caring for critically Ill patients: Clinicians’ empathy promotes Job satisfaction and does not predict moral distress. Frontiers in Psychology, 10, 2902. 10.3389/fpsyg.2019.02902 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lundgren-Nilsson, Å, Jonsdottir, I. H., Pallant, J., & Ahlborg, G. (2012). Internal construct validity of the Shirom-Melamed Burnout Questionnaire (SMBQ). BMC Public Health, 12(1), 1. 10.1186/1471-2458-12-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397–422. 10.1146/annurev.psych.52.1.397 [DOI] [PubMed] [Google Scholar]
- Meltzer, L. S., & Huckabay, L. M. (2004). Critical care nurses’ perceptions of futile care and its effect on burnout. American Journal of Critical Care, 13(3), 202–208. [PubMed] [Google Scholar]
- Morse, J. M., Bottorff, J., Anderson, G., O'Brien, B., & Solberg, S. (1992). Beyond empathy: Expanding expressions of caring. Journal of Advanced Nursing, 17(7), 809–821. 10.1111/j.1365-2648.1992.tb02002.x [DOI] [PubMed] [Google Scholar]
- Picard, J., Catu-Pinault, A., Boujut, E., Botella, M., Jaury, P., & Zenasni, F. (2016). Burnout, empathy and their relationships: A qualitative study with residents in general medicine. Psychology, Health & Medicine, 21(3), 354–361. 10.1080/13548506.2015.1054407 [DOI] [PubMed] [Google Scholar]
- Public Health Agency of Sweden . (2018). Purpose and background to the questions in the national public health survey: Health on equal terms 2018. Report No. 18083 (Swedish). https://www.folkhalsomyndigheten.se/folkhalsorapportering-statistik/om-vara-datainsamlingar/nationella-folkhalsoenkaten/syfte-och-bakgrund/.
- Rauvola, R. S., Vega, D. M., & Lavigne, K. N. (2019). Compassion fatigue, secondary traumatic stress, and vicarious traumatization: A qualitative review and research agenda. Occupational Health Science, 3(3), 297–336. 10.1007/s41542-019-00045-1 [DOI] [Google Scholar]
- Ray, S. L., Wong, C., White, D., & Heaslip, K. (2013). Compassion satisfaction, compassion fatigue, work life conditions, and burnout among frontline mental health care professionals. Traumatology, 19, 255–267. 10.1177/1534765612471144 [DOI] [Google Scholar]
- Sheather, J., Apunyo, R., DuBois, M., Khondaker, R., Noman, A., Sadique, S., & McGowan, C. R. (2022). Ethical guidance or epistemological injustice? The quality and usefulness of ethical guidance for humanitarian workers and agencies. BMJ Glob Health, 7(3), e007707. 10.1136/bmjgh-2021-007707 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spilg, E. G., Rushton, C. H., Phillips, J. L., Kendzerska, T., Saad, M., Gifford, W., … Robillard, R. (2022). The new frontline: Exploring the links between moral distress, moral resilience and mental health in healthcare workers during the COVID-19 pandemic. Bmc Psychiatry, 22(1), 19. 10.1186/s12888-021-03637-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stamm, B. (2010). The concise ProQOL manual (2nd ed.). ProQOL.org. [Google Scholar]
- Ter Heide, F. J. J. (2020). Empathy is key in the development of moral injury. European Journal of Psychotraumatology, 11(1), 1843261. 10.1080/20008198.2020.1843261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- van Mol, M. M. C., Kompanje, E. J. O., Benoit, D. D., Bakker, J., & Nijkamp, M. D. (2015). The prevalence of compassion fatigue and burnout among healthcare professionals in intensive care units: A systematic review. PLoS One, 10(8), e0136955. doi:ARTNe013695510 .1371/journal.pone.0136955 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williamson, V., Lamb, D., Hotopf, M., Raine, R., Stevelink, S., Wessely, S., … Greenberg, N. (2023). Moral injury and psychological wellbeing in UK healthcare staff. Journal of Mental Health, 32(5), 890–898. 10.1080/09638237.2023.2182414 [DOI] [PubMed] [Google Scholar]
- Williamson, V., Stevelink, S. A. M., & Greenberg, N. (2018). Occupational moral injury and mental health: Systematic review and meta-analysis. British Journal of Psychiatry, 212(6), 339–346. 10.1192/bjp.2018.55 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Additional data is available upon reasonable request to the authors.