Predictors of Moral Distress Among Nurses: A Cross-Sectional Study

Reviewer 1

The rationale is generic and weakly theorised. Moral distress is introduced descriptively without anchoring in a clear conceptual framework.

Thank you for this valuable comment. In response, we have substantially revised the theoretical background and explicitly anchored the concept of moral distress within a clear conceptual framework. We added the justification to the text.

Moral distress is currently understood as a complex, multilevel, and dynamic phenomenon that arises in situations where healthcare professionals recognize the ethically appropriate course of action but are unable to act in accordance with this judgment due to various constraints (Andrew Jameton, 1984). This foundational concept has been further expanded to include additional dimensions reflecting its complexity and contextual nature (Mary C. Corley, 2001).

From a conceptual perspective, moral distress can be interpreted through the interaction of three key domains: (1) organizational, (2) individual, and (3) ethical. The organizational domain includes structural and systemic factors influencing care delivery (Kälvemark et al., 2004), while the individual domain encompasses personal characteristics such as personality traits (Burston & Tuckett, 2013).

An important component of this conceptual framework is its dynamic and cumulative nature. Repeated exposure to morally distressing situations leads to the accumulation of so-called moral residue and may result in the “crescendo effect,” characterized by a progressive intensification of psychological burden (Epstein and Hamric, 2009).

The manuscript mixes individual (personality, coping) and organisational variables (shift work) without an explicit explanatory model.

Thank you for the valuable comment. The investigated variables were explained explicitly. The sources of moral distress may vary and can stem from clinical conditions, the work environment, and both external and internal factors.

External factors primarily include inadequate staffing, lack of time, insufficient administrative support, and institutional policies and priorities that may be in conflict with patient care. Internal factors relate, for instance, to personal aspects that influence the perceived ability of providers to deliver optimal care. These internal factors include fear of job loss, anxiety, feelings of helplessness, a lack of assertiveness, insufficient understanding of the situation, and low self-esteem. Risk factors for moral distress include not only nursing staff shortages but also insufficient experience and education, poor teamwork between physicians and nurses, high workload, and low quality of care. Previous research suggests that moral distress among nurses may be more strongly associated with older age and longer work experience, and less associated with gender or religion.

No explicit theoretical model is guiding variable selection, which makes the analysis largely exploratory rather than hypothesis-driven.

Thank you for your comment regarding the absence of an explicit theoretical model in the submitted manuscript. Based on your suggestion, we have revised the manuscript to include and explicitly formulate a theoretical framework that integrates the examined variables into a coherent model of moral distress.

The proposed model assumes that moral distress arises from the interaction of organizational factors (e.g., work environment and shift work), individual characteristics (age, years of practice), personality traits (Big Five), coping strategies, and burnout. Within this framework, coping strategies play a moderating role, while burnout represents a key psychological mechanism linking work-related stressors to the experience of moral distress.

By incorporating this framework, we aimed to systematize the selection of variables and to demonstrate that the analysis is not purely exploratory, but rather grounded in an integrated, theoretically informed approach to moral distress.

The claimed gap (“limited evidence in Slovakia”) is contextual but not scientific.

We thank the reviewer for this comment. We agree that the statement (“limited evidence in Slovakia”) is primarily contextual rather than scientific in nature. Therefore, we have removed this part from the manuscript in order to improve its scientific rigor and focus.

The international literature on predictors of moral distress is already extensive.

Thank you for your valuable comment. Yes, we note that the international literaturer on predictors of moral distress is already extensive, but the gap of this phenomenon in the conditions of Slovakia is missing. The results bring our sociocultural context closer. Specifically, the Slovak context extends existing knowledge on moral distress primarily by emphasizing the importance of organizational factors, identifying deficits in ethics education, and confirming the interaction between individual and systemic determinants. The findengs also support integrative models of moral distress as a context-dependent and dynamic phenomenon. We also added this text to the implications for practice.

The manuscript does not explain how the Slovak context modifies or extends existing knowledge.

The Slovak context extends existing knowledge on moral distress primarily by emphasizing the importance of organizational factors, identifying deficits in ethics education, and confirming the interaction between individual and systemic determinants. The findings also support integrative models of moral distress as a context-dependent and dynamic phenomenon.

Translation process described, but no psychometric validation results beyond alpha

Thank you for this helpful comment. We appreciate the suggestion to include additional psychometric information. Together with the reliability analysis, the Exploratory factor analysis results were added to the Methods section:

Exploratory factor analysis (principal component extraction) supported a one‑factor solution explaining 31.5% of the variance for the frequency domain and 37.8% of the variance for the intensity domain. All items loaded positively on the single component, with factor loadings ranging from 0.41 to 0.64, resp. 0.48 to 0.69, indicating a coherent unidimensional structure of both domains of moral distress scale.

No sample size justification or power analysis.

The results of power analysis were included in the Methods section and as the final sample (N = 412) exceeded the threshold, it is resulting in increased statistical power and improved stability of the regression estimates.

Text in the Methods section:

The power analysis using G*Power software (ver. 3.1.9.7) suggested a minimum sample size of N=249 participants to reliably detect a medium effect in a multiple regression model with 27 predictors at α = 0.05 and power = 0.95. Given the number of predictors included in the model, a medium effect size was assumed, reflecting the expectation that individual predictors would contribute modestly to the explained variance. The relatively high number of predictors suggests that a larger sample may further improve model stability. The final sample (N = 412) exceeded this threshold, resulting in increased statistical power and improved stability of the regression estimates.

No assessment of assumptions is provided, such as linearity, Multicollinearity, homoscedasticity, or Normality of residuals.

We thank the reviewer and the missing information were added to the Methods section:

Assessment of assumptions revealed these results. Residual diagnostics indicated that the residuals were centred around zero (M = 0.00), suggesting no systematic bias. Standardized residuals ranged from -2.46 to 4.02, indicating the presence of a small number of potential outliers. However, given the large sample size, their impact is unlikely to substantially affect the results. Multicollinearity was not a concern (VIFs ≈ 1.00–1.12). Visual inspection of the scatterplot of standardized residuals against standardized predicted values indicated no evidence of heteroscedasticity.

No control for confounders has been performed, such as for age, experience or unit type.

Following justification was added to the manuscript in the Methods, Results and Discussion sections:

Methods:

All available demographic and professional variables (age, gender, education, years of practice, and unit type) were entered into the initial stepwise regression model. However, as stepwise procedures retain variables based on statistical criteria, this approach does not guarantee full control of confounding. This is now acknowledged as a limitation in the manuscript.

Results:

The initial regression models included demographic and professional variables (age, gender, education, years of practice, and unit type), personality traits (5 variables), coping strategies (14 variables), and burnout dimensions (3 variables). All variables were entered into the models, and a stepwise procedure was used to identify predictors retained in the final model.

Discussion (limitations):

Although relevant demographic and professional variables were considered, the use of stepwise regression limits the ability to ensure full control of potential confounding factors.

Several statements imply causal direction despite a cross-sectional design.

We thank reviewers for this important issue. As the appropriate statements were not changed, important explanations were added to Methods section and to Study limitations.

Methods section:

Although regression models were used, the cross‑sectional design precludes causal interpretation. The findings should therefore be understood as associations between variables measured at a single point in time.

Limitations section:

The data were cross-sectional, which restricts the ability to draw causal conclusions. Although regression analyses were conducted to examine associations between the variables, the design does not allow for causal inferences. The relationships identified should be interpreted as correlational rather than causal.

Methodology must be organised using a subparagraph.

Thank you for the valuable comment. The methodology was revised according to your requirements.

The number and characteristics of the participants are results and not methods. In the methods, you must describe the inclusion and exclusion criteria and the sample size estimation.

Thank you for the relevant comment. First, the sample characteristics were moved to the Results section (beginning of the section).

Second, the results of power analysis were included in the Methods section and as the final sample (N = 412) exceeded the threshold, it is resulting in increased statistical power and improved stability of the regression estimates. (described in detail above)

The instruments' description must be similar, e.g. items, dimensions, response, scoring, interpretation, validity in the validation study and in this research.

The section was revised in order that the descriptions follow the same manner. For all instruments, the reliability coefficients were added both, for original measures and computed from our data sample.

Ethical considerations are missing in the methods and in the data collection process.

The research was approved by the ethics committees of the participating healthcare institutions.

Before the conclusions, I suggest a paragraph with "Implications for clinical and organisational practice".

Thank you for this valuable comment. The study findings highlight the need for a systematic approach to addressing moral distress at multiple levels of healthcare practice. In clinical practice, it is essential to strengthen nurses’ ability to manage ethically challenging situations, particularly through the development of adaptive coping strategies and the promotion of psychological resilience. The identification of maladaptive strategies, such as distraction, underscores the need for targeted interventions aimed at more effective stress management. At the same time, the results emphasize the importance of burnout prevention, especially in the dimension of personal accomplishment, which was identified as a significant predictor of moral distress.

At the organizational level, the findings have important implications for the management of working conditions, particularly with regard to shift work, which was identified as a significant factor influencing the frequency of moral distress. Optimizing work schedules, reducing workload, and stabilizing teams may help decrease exposure to ethically challenging situations. Another important finding is the insufficient presence of ethics education, highlighting the need for its systematic integration into continuing professional development. The Slovak context extends existing knowledge on moral distress primarily by emphasizing the importance of organizational factors, identifying deficits in ethics education, and confirming the interaction between individual and systemic determinants. The findings also support integrative models of moral distress as a context-dependent and dynamic phenomenonT

Reviewer 2

Line 1: if the sample was nation wide- please include a phrase national study in the title.

The study sample had a multicenter character and included nurses from multiple healthcare facilities in two regions of Slovakia; however, we cannot claim that it was nationally representative.

Line 62: provide a citation for this definition and provide a more comprehensive explanation of what moral distress is truely is.

Thank you for this valuable comment. The paraphrase has been replaced with Jameton’s original concept: “ Moral distress is currently understood as a complex, multilevel, and dynamic phenomenon that arises in situations where healthcare professionals recognize the ethically appropriate course of action but are unable to act in accordance with this judgment due to various constraints.” (JAMETON, A. 1984. Nursing Practice. Prentice-Hall: New Jersey. 1984. ISBN 0-13-627448-X)

Line 74: Consider adding to the discussion of outcomes of moral distress by giving example related to quality of care, patient satisfaction, Missed care, retention, resignation and others.

Thank you for your valuable comment. We consider your suggestion highly beneficial and will incorporate it into the discussion section, where we will expand the interpretation of the consequences of moral distress to include the mentioned areas. Moral distress has significant consequences not only for nurses themselves but also for the quality of care provided. Factors contributing to missed nursing care are also predictors of the intensity and frequency of moral distress. Therefore, the reasons for omitted nursing care may increase both the frequency and the level of moral distress. A similar pattern can be observed in patient and family satisfaction, where an inverse relationship exists between the frequency of moral distress and family satisfaction, as well as adherence to patients’ rights.

In terms of work-related outcomes, a consistent association has been demonstrated between moral distress, burnout, and the intention to leave the workplace or the profession, which in turn reduces staff retention. These findings highlight the need for systematic interventions aimed at reducing moral distress, not only from the perspective of protecting nurses’ mental health, but also in ensuring the quality and continuity of healthcare delivery. We added the text to the discussion section.

Line 94: Provide a rationale for excluding nurses working children. these are more likely to be even more distressed when they experience death and complications associated with wrong medical decisions.

Nurses providing care to pediatric patients were not included in the study in order to ensure sample homogeneity and enhance the internal validity of the findings. Pediatric care represents a specific clinical context that differs significantly from adult care, particularly in terms of ethical dilemmas, decision-making processes, and interactions with patients’ families. In pediatrics, legal guardians play a key role, which modifies patient autonomy and alters the nature of ethical conflicts that may lead to moral distress.

Including nurses from pediatric settings could increase the variability of the studied phenomenon, as the sources, intensity, and frequency of moral distress may differ depending on the patient population. In line with methodological recommendations for moral distress research, it is therefore appropriate to analyze relatively homogeneous groups of healthcare professionals in order to more accurately identify relationships between variables and minimize the influence of contextual factors.

For these reasons, the study focused exclusively on nurses providing care to adult patients, which allowed for a more consistent interpretation of the relationships between organizational and individual determinants of moral distress.

Line 147: How were the assumptions for linear regression tested? Was the data normally distributed?

Missing information about tested assumptions were added to the Methods section:

Assessment of assumptions revealed these results. Residual diagnostics indicated that the residuals were centred around zero (M = 0.00), suggesting no systematic bias. Standardized residuals ranged from -2.46 to 4.02, indicating the presence of a small number of potential outliers. However, given the large sample size, their impact is unlikely to substantially affect the results. Multicollinearity was not a concern (VIFs ≈ 1.00–1.12). Visual inspection of the scatterplot of standardized residuals against standardized predicted values indicated no evidence of heteroscedasticity.

Line 148: provide a rationale for choosing stepwise regression analysis.

Linear regression analysis was used to evaluate predictors (personality factors, burnout level, and coping strategies) of moral distress. Stepwise regression was employed as an exploratory approach to identify the most significant and parsimonious set of predictors of moral distress from multiple potentially interrelated variables, particularly in the absence of strong theoretical guidance on their relative importance.

Although regression models were used, the cross‑sectional design precludes causal interpretation. The findings should therefore be understood as associations between variables measured at a single point in time.

Additionally, the limitations section was strengthened according to this issue:

The data were cross-sectional, which restricts the ability to draw causal conclusions. Although regression analyses were conducted to examine associations between the variables, the design does not allow for causal inferences. The relationships identified should be interpreted as correlational rather than causal.

Table 1: provide data about age (categories, Mean and standard deviation).

Age categories were added to the Table 1 and Mean and standard deviation age characteristics were added to the results section.

Table 1: provide data about mean working hours and SD.

We thank the reviewer for this comment. However, the calculation of mean and standard deviation was not possible, as working hours were assessed using categorical response options rather than exact numerical values. The distribution of these categories is already presented in Table 1.

Line 155: provide a table of the factors that were significantly correlated using correlations.

Thank you for the suggestion. Because the study uses multivariable regression to examine adjusted associations, we did not include bivariate correlation analyses. The regression models already identify which predictors are significantly associated with the outcome while controlling for other variables. We clarified this in the Methods and Results sections.

The appropriate statement was rewritten to strengthen this issue: The results examine the associations between the frequency of moral distress and demographic characteristics, personality traits, burnout level, and coping strategies using multivariable regression models.

Discussion: Please suggest feasible evidence based interventions that can be used to address moral distress.

Thank you for the suggestion.  We have added this to the text. Moral distress can be mitigated through organizational support and education, organizational guidelines, unit-specific guidelines, formal forums for discussing ethically challenging situations, ethics-focused education, peer support, watching films, communicating with family and friends, taking walks, maintaining a sense of humor, yoga, meditation, listening to music, adequate nutrition, and sufficient sleep (Koivisto, 2024; Ghazanfari, 2022)

Provide and explanation of why more than 80% of nurses did not attend any training related to ethics.

The shortage of nurses, retirement, and insufficient interest among young people in the profession have become significant problems, including in Slovakia. In 2022, Slovakia had an average of 5.4 nurses per 1,000 inhabitants, while the European Union average was 8.1 nurses per 1,000 inhabitants, indicating a shortage of approximately 15,000 nurses in the Slovak healthcare system (National Health Information Center, 2022). Maghsoud et al. (2022) confirmed that the physical demands caused by nurse shortages, lack of work-supporting tools, time pressure, and challenging working conditions lead to occupational stress and reduced performance.

An explanation of the nursing education system and qualifications in the country may help.

Nursing education in Slovakia is part of the tertiary education system and is harmonized with European Union requirements, particularly Directive 2005/36/EC on the recognition of professional qualifications. (Legal and policy framework - Employment, Social Affairs and Inclusion)

Despite the existence of a formal system of lifelong learning, its implementation in practice is uneven. Educational activities are often primarily focused on clinical and technical aspects of care, whereas areas such as ethics, communication, and coping with psychological burden are not systematically addressed. This gap may affect nurses’ preparedness to manage ethically challenging situations and is also associated with the occurrence of moral distress in nursing practice. .  We have added this to the text.

Line 301; limitations should include the factor that most participants were working high emotion intensity settings (ICU), and lack of data about social factors such as income, culture, all of which can affect moral distress.

Thank you for your valuable comment. We will incorporate the mentioned limitations into the limitations section of the study and appropriately reflect them in the interpretation of the results.

The reference need to re-written without capitalizing all the letters in the authors names.

Thank you for your relevant comment. The reference has been rewritten.

The manuscript is improved and no longer appears fatally flawed. The authors addressed several earlier methodological transparency issues and strengthened the theoretical framing. However, important concerns remain present.

We thank the reviewer for positive evaluation of the manuscript and for appreciating the changes made during the revision process. We greatly appreciate that the reviewer perceives an improvement in the methodological transparency and theoretical anchoring of the study. At the same time, we respect that despite the changes made, some questions remain open, and in the next revision we tried to address these comments as precisely and transparently as possible.

The framework proposed in the introduction suggests that the most appropriate analysis is mediation, moderation, or hierarchical relationships. Instead, the authors use exploratory stepwise regression. Consequently, the theoretical model and the statistical analysis are misaligned.

Thank you for this thought-provoking comment. We agree that the theoretical framework may suggest potential mediating or moderating relationships between the variables under investigation. However, our aim was not to test a formal causal or hierarchical model, but rather to explore which individual and work factors are associated with moral distress in the so far under-researched Central European post-socialist context.

For this reason, we have chosen an exploratory stepwise regression approach as a preliminary analytical tool to identify relevant predictors, while recognizing that this approach is not equivalent to testing mediation or moderating mechanisms. In the revised manuscript, we therefore more explicitly emphasized the exploratory nature of the analysis, softened the formulations implying causal or hierarchical relationships, and listed this methodological discrepancy among the study limitations.

The rationale has improved, but the knowledge gap remains insufficiently precise. The authors repeatedly emphasize the Slovak sociocultural context, but they do not clearly explain what is theoretically or clinically unique about this context compared with existing international evidence.

Thank you for the suggestion. Based on the comment, we have clarified in what way we consider the Slovak socio-cultural context relevant. In the revised text, we emphasize that most of the knowledge on moral distress comes from countries with different healthcare organization, degree of professional autonomy of nurses and working conditions. Slovak healthcare is characterized by a long-term personnel deficit, high workload and a more hierarchical model of labor relations, which can influence the emergence and consequences of moral distress. At the same time, we have formulated more explicitly that the contribution of the research lies in the comprehensive examination of moral distress in relation to personality factors, working conditions, burnout syndrome and the burden of nurses in the conditions of Slovak healthcare.

The importance of investigating moral distress in the Slovak sociocultural context is based on the assumption that although moral distress in nurses is a globally described phenomenon, its determinants and consequences are significantly modified by the cultural, organizational and systemic environment of healthcare. Most of the research to date comes mainly from non-European countries such as North America, the Middle East, and Asia. From European countries, only Western European and Northern European countries are listed (Xue et al., 2025). These are countries with different working conditions, professional autonomy of nurses, staffing and models of interdisciplinary cooperation compared to the Central European countries, within which we present new findings. In the Slovak healthcare system, moral distress can be specifically influenced by factors such as long-term staff shortages, high workload, lower participation of nurses in decision-making processes, hierarchical model of work relationships, limited personnel and material resources, as well as culturally conditioned expectations regarding the professional role of nurses. These characteristics can change not only the frequency of morally stressful situations, but also the way they are psychologically processed and their relationship to burnout or workload.

The contribution of our research also lies in the fact that moral distress is not examined in isolation, but in relation to personality characteristics, working conditions, burnout syndrome and subjectively perceived workload of nurses. This complex model allows for a better understanding of which individual and organizational factors can increase the vulnerability of nurses to moral distress in the conditions of the Slovak healthcare system. At the same time, it can contribute to the design of more targeted prevention and intervention strategies aimed at supporting the mental health of nurses and the sustainability of the nursing profession.

The response rate of only 37.45% is substantially a limit of representativeness and increases the risk of self-selection bias. Nurses experiencing higher distress may have been either more motivated or less motivated to participate. This issue is insufficiently discussed.

Thank you to the reviewer for this valuable comment. We agree that a response rate of 37.45% may represent a limitation regarding the representativeness of the sample and may increase the risk of self-selection bias. Given the nature of the investigated topic, it cannot be excluded that nurses experiencing higher levels of moral distress were either more motivated to participate in the study or, conversely, less willing to participate due to psychological burden or work-related exhaustion.

We have therefore expanded the Limitations section to explicitly address the potential risk of selection bias in the interpretation of the findings. At the same time, we consider it important to emphasize that the achieved sample size (N = 412) exceeded the minimum required sample size to ensure adequate statistical power, and the study included nurses from multiple healthcare institutions and various clinical departments across two regions of Slovakia. Nevertheless, the findings should be interpreted with appropriate caution, and future longitudinal or multicenter studies with higher participation rates are recommended.

Second, the regression approach remains the major methodological weakness. The authors justify stepwise regression as exploratory due to the lack of strong theoretical guidance. However, this justification is internally inconsistent because the manuscript simultaneously proposes a detailed theoretical model inserted as a response to the previous round of revisions.

Thank you for this important comment. We agree that the regression approach represents a limitation of the study, and in the revised manuscript we have attempted to explain more clearly the relationship between the theoretical framework and the exploratory analytical approach.

Our intention was not to present the detailed theoretical model as a fully established or empirically stabilized causal model. The expanded theoretical framework was included in order to better anchor the investigated variables in the existing literature and to clarify the presumed relationships between moral distress, personality characteristics, working conditions, burnout syndrome, and nurses’ burden. At the same time, however, we maintain that the Central European post-socialist context remains relatively underexplored in this research area and that the available international findings cannot be automatically generalized to the conditions of the Slovak healthcare system.

For this reason, we approach the proposed model rather as a less theoretically grounded framework that requires further empirical exploration. We therefore consider the choice of an exploratory regression approach to be appropriate given the limited number of studies conducted in comparable sociocultural and healthcare settings. In the revised text, we explained this position more explicitly and also more clearly identified the use of stepwise regression as one of the methodological limitations of the study.

Another limitation is the exploratory use of stepwise regression analysis, which was applied due to the still limited theoretical and empirical evidence concerning moral distress in the Central European post-socialist healthcare context. Therefore, the proposed model should be interpreted cautiously as a preliminary exploratory framework requiring further validation in future research.

Brief COPE subscale reliability reaches α = 0.32 for some subscales. This is a major issue. Although two-item scales often produce lower alpha values, reliability this low raises concerns about measurement error and interpretation of regression coefficients. At a minimum, the authors should specify which subscales showed unacceptable reliability and interpret findings cautiously.

Thank you for this important comment. We agree that the lower reliability values observed for some Brief COPE subscales represent a methodological limitation of the study and was not described in more needed detail. In the revised manuscript, we therefore explicitly report the Cronbach’s alpha coefficients for subscales with α < .50.

We also note that the Brief COPE consists of two-item subscales, for which lower Cronbach’s alpha values are relatively common in the literature. Nevertheless, we acknowledge that the reliability coefficients observed in several subscales with α < .50, including Active coping, Denial, Behavioral disengagement, Venting, and Self-blame, and this may increase the risk of measurement error and affect the stability of the regression coefficients. Therefore, findings related to these subscales were interpreted with greater caution, and this issue was explicitly included among the study limitations.

However, lower reliability coefficients (α < .50) were observed for several subscales, particularly Active coping, Denial, Behavioral disengagement, Venting, and Self-blame. Therefore, findings related to these subscales should be interpreted with caution due to the increased risk of measurement error and reduced stability of regression estimates.

A limitation of this study is the lower internal consistency observed for several Brief COPE subscales, particularly Active coping, Denial, Behavioral disengagement, Venting, and Self-blame. Although lower reliability is relatively common in two-item Brief COPE subscales, these findings should be interpreted cautiously due to the potential impact of measurement error on the stability of regression estimates.

“validity” in keywords is inappropriate and misleading because this is not primarily a validation study.

Thank you for this comment. We have removed the term “validity” from the keywords.

The abbreviation section contains obvious template artefacts (“TLA,” “LD,” etc.) unrelated to the manuscript and must be removed.

Thank you very much for this comment. The unrelated template artefacts have been removed from the manuscript.