Moral distress among physicians and nursing staff in surgical settings: a comparative study Authors: Jordana Lopes Carvalho, Luis Felipe Dias Lopes, Flavia Regina Souza Ramos, Graziele de Lima Dalmolin Published: 2026 ABSTRACT AbstractBackgroundMoral distress is a relevant occupational health issue in hospital settings, particularly in high-complexity environments such as surgical units, where organizational constraints, ethical dilemmas, and interprofessional dynamics shape professionals’ experiences. Comparative evidence is needed to understand how moral distress manifests across professional groups working within the same clinical context.MethodsThis cross-sectional comparative study examined moral distress among physicians and nursing staff working in surgical settings at a public university hospital in southern Brazil. Data were collected from 245 healthcare professionals (167 physicians and 78 nursing staff) using the Brazilian Moral Distress Scale – Short Version (EDME-BR-VR). Group differences in moral distress dimensions were assessed using descriptive statistics and Mann–Whitney U tests. Comparative structural analyses were conducted using Partial Least Squares Structural Equation Modeling (PLS-SEM), with separate models estimated for each professional group to examine similarities and differences in structural relationships.ResultsSignificant differences between physicians and nursing staff were observed in selected moral distress dimensions, particularly Safe and Qualified Care, Defense of Values and Rights, and Working Conditions. Across both groups, Recognition, Power, and Professional Identity emerged as a central antecedent of moral distress, showing significant associations with teamwork, working conditions, safe and qualified care, and ethical violations. Comparative PLS-SEM results indicated similar overall relational patterns across groups, with differences in the magnitude of specific structural paths reflecting profession-specific dynamics.ConclusionsMoral distress among physicians and nursing staff in surgical settings is shaped by shared organizational factors as well as profession-specific ethical pressures. Comparative analysis highlights the central role of professional recognition and identity while underscoring the need for differentiated organizational strategies to address moral distress across professional groups. These findings provide actionable insights for healthcare managers seeking to promote ethical practice, psychological safety, and sustainable work environments in surgical care.Keywords: Moral distress, Ethical dilemmas, Physicians, Psychometric validation, Invariance analysis Introduction The hospital environment is a multifaceted space that plays a central role in healthcare delivery. It is considered complex and challenging due to its organizational characteristics, in which moral problems frequently arise, exposing professionals to difficult and morally demanding decisions. Moral problems can trigger Moral Distress (MD) when professionals encounter barriers that hinder deliberative processes, preventing ethical decision-making and the expression of their ethical–moral positioning [1]. MD is understood as a process in which moral experience is uniquely manifested, as individuals interpret, evaluate, and judge moral problems differently according to their professional background, values, and work context [2]. Over recent years, the concept of MD has expanded beyond nursing, encompassing different professional fields and multidisciplinary healthcare teams [3–5]. International scholarship increasingly recognizes that moral distress extends beyond an individual psychological response, being shaped by organizational ethical climate, leadership practices, and psychological safety [6–11]. This perspective highlights moral distress as a systemic and organizationally embedded phenomenon rather than solely an individual experience. Different team configurations and forms of work organization within hospitals raise important ethical challenges, particularly in closed and highly specialized sectors such as the operating room. Surgical settings are characterized by complex activities that require the use of advanced technologies and coordinated action among multidisciplinary teams. Despite the importance of individual expertise, teamwork is indispensable in a context frequently permeated by time pressure, stress, and high clinical risk, involving surgical procedures, anesthesia, continuous monitoring of vital signs, and strict safety protocols [12]. Within this environment, healthcare professionals may experience situations that generate ethical conflicts related to work organization, scarcity of human and material resources, work overload, and fragmentation of care processes, all of which are closely associated with moral distress [3, 5, 7]. Scholarly analyses further indicate that hierarchical structures, interprofessional power asymmetries, and limited psychological safety constrain moral agency and ethical voice behaviors, particularly in highly specialized environments [8, 9, 11, 16]. These conditions may affect professional roles differently, depending on the nature of responsibilities, decision-making autonomy, and interprofessional relationships within surgical teams. These dynamics suggest that moral distress is shaped not only by exposure to ethical dilemmas but also by professional roles and hierarchical positioning within healthcare teams. The literature has predominantly focused on moral distress among nurses [4, 5, 14–16], using both qualitative and quantitative approaches and validated measurement instruments [4, 5, 17, 18]. Although studies addressing MD in other professional categories exist [18–22], they have often employed different instruments or focused on specific professional groups, limiting direct comparisons across healthcare roles. Evidence comparing moral distress experiences between physicians and nursing staff within the same clinical context, particularly surgical settings, remains limited. As a result, there is a lack of comparative evidence using a common measurement framework across professional groups working within the same clinical context. Internationally, most empirical studies on moral distress have focused on single professional categories—predominantly nurses—or have examined moral distress as an individual-level psychological outcome linked to burnout, turnover intention, or mental health indicators. Recent empirical studies demonstrate strong associations between moral distress, burnout, psychological safety, and organizational ethical climate, highlighting the systemic nature of the phenomenon [7, 23–26]. However, few studies have simultaneously examined both the intensity and the structural relationships among moral distress dimensions across different professional groups within the same organizational setting. In addition, there is limited evidence on whether the relational structure among moral distress dimensions varies across professional groups. Contemporary research in nursing leadership and organizational ethics indicates that ethical leadership and psychological safety mediate the relationship between professional recognition, autonomy, and ethical action, suggesting that moral distress may emerge differently according to hierarchical positioning and perceived institutional support [7–10]. This suggests that the configuration of relationships among moral distress dimensions may differ according to professional roles, but this assumption remains underexplored in empirical research. The Brazilian Moral Distress Scale – Short Version (EDME-BR-VR) [27] was developed to assess moral distress across six dimensions: recognition, power, and professional identity; safe and qualified care; defense of values and rights; working conditions; ethical violations; and work teams. These dimensions represent complementary organizational, relational, and ethical mechanisms through which moral distress is experienced, capturing both structural conditions (e.g., working conditions and safety culture) and relational dynamics (e.g., teamwork and professional identity). Recent literature reinforces the theoretical coherence of these dimensions, particularly the central role of recognition, empowerment, and safety culture in shaping ethical agency [7–9, 26]. Assessing the intensity and distribution of moral distress dimensions among different professional groups working in surgical settings can broaden the understanding of how organizational and ethical challenges are perceived and managed in daily practice. More importantly, examining both differences in levels and differences in structural relationships allows a more comprehensive understanding of moral distress as a multidimensional phenomenon. Thus, this study aims to compare moral distress dimensions between physicians and nursing staff working in surgical settings, using the EDME-BR-VR scale, by simultaneously analyzing group differences and the structural relationships among dimensions, in order to identify similarities and differences in moral distress experiences across professional groups. Hypotheses development The following section presents the theoretical foundation and rationale for each proposed hypothesis, based on the relationships between the factors of the Moral Distress scale and professional performance. The proposed associations are synthesized later in Fig. 1. Conceptual model of the relationships between moral distress factors and professional performance. Source: SmartPLS® software (version 4.1.1.7) [41]. DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care; RPI = Recognition, Power, and Identity Moral distress arises when healthcare professionals understand the ethically appropriate course of action but are unable to act upon it due to institutional or hierarchical barriers [28]. In such contexts, professional identity and recognition function as key psychological resources that strengthen moral agency [29, 30]. When nurses and physicians perceive autonomy and respect within their organizations, they demonstrate greater ethical sensitivity and advocacy behaviors aligned with patients’ rights and human dignity [31, 32]. Professional identity, therefore, serves as a motivational driver that fosters moral courage and ethical action, particularly when healthcare professionals face structural or organizational constraints [33]. Empirical evidence indicates that ethical leadership and supportive ethical climates enhance moral agency and ethical voice behaviors, reinforcing professionals’ willingness to defend values and report ethical concerns [6–8]. These findings support the assumption that recognition, empowerment, and professional identity strengthen the defense of ethical principles within healthcare organizations by enhancing moral agency and ethical voice behaviors. Hence, the first hypothesis of this study is: H1 Recognition, Power, and Identity (RPI) positively influence the Defense of Values and Rights (DVR). Effective team functioning in healthcare settings is grounded in mutual trust, shared professional identity, and respect among members. Recognition and empowerment significantly strengthen teamwork by fostering inclusion, collaboration, and a sense of mutual accountability among healthcare professionals [34]. When healthcare workers perceive recognition of their contributions and professional identity, they are more likely to engage in interprofessional collaboration and open communication [35]. A shared sense of purpose, supported by professional identity, enhances coordination and collective ethical awareness within teams [33]. Contemporary research demonstrates that authentic and ethical leadership styles promote psychological safety and improve teamwork dynamics, reducing care omissions and strengthening collaborative performance [9, 13, 26]. These studies suggest that empowerment and professional recognition are central mechanisms through which team cohesion and interprofessional collaboration are enhanced. These findings support the assumption that RPI strengthens interpersonal dynamics and cooperative behavior through increased trust, shared identity, and psychological safety within teams. Therefore, the following hypothesis is proposed: H2 Recognition, Power, and Identity (RPI) positively influence Work Teams (WT). Professional identity and empowerment not only shape ethical behavior but also influence how individuals perceive their working conditions. Recent evidence suggests that healthcare professionals who feel valued and recognized are more satisfied with their work environment, experience lower stress, and report better well-being [29, 33]. The perception of fairness, respect, and institutional support fosters stronger professional engagement and decreases turnover intention, thereby reinforcing the association between Recognition, Power, and Identity (RPI) and positive work experiences [30]. Multiple investigations have demonstrated that moral distress is strongly associated with burnout and diminished well-being, particularly in contexts where professionals perceive low institutional recognition and limited autonomy (Locke et al., 2025; Popucza & Eriksson, 2025; Alhassan et al., 2024). These findings reinforce the relevance of recognition and empowerment as protective factors influencing professionals’ perceptions of working conditions. Moreover, when recognition and identity are integrated into organizational culture, they help create resilient and ethically sound workplaces [35]. These findings suggest that professional recognition and identity not only influence individual well-being but also shape how working conditions are perceived and evaluated within organizational contexts. Thus, the third hypothesis is formulated as follows: H3 Recognition, Power, and Identity (RPI) positively influence Working Conditions (WC). A robust safety culture encourages transparency and continuous learning from ethical incidents rather than concealment. Recent evidence indicates that when healthcare organizations invest in safety and quality systems, professionals become more willing to report errors and ethical breaches, reinforcing a culture of accountability and trust [28, 36]. Studies examining ethical leadership and psychological safety further demonstrate that environments characterized by strong safety cultures are associated with increased reporting behaviors and constructive responses to ethical violations [7, 8]. In such contexts, the identification of ethical violations reflects organizational maturity and monitoring capacity rather than an increase in unethical conduct itself. This perspective aligns with the notion that stronger safety cultures promote transparency and reporting behaviors, increasing the visibility of ethical violations rather than their actual occurrence. Thus, safe and qualified care operates by increasing reporting behaviors and transparency, which in turn enhances the visibility and management of ethical violations. Consequently, stronger Safe and Qualified Care (SQC) cultures are theoretically expected to be associated with greater visibility, reporting, and proactive management of ethical violations. Hence, the next hypothesis is: H4 Safe and Qualified Care (SQC) positively influences Ethical Violations (EV). Healthcare professionals with high levels of recognition and autonomy are more engaged in implementing practices that ensure patient safety and quality outcomes. Recent studies show that leadership support, professional identity, and recognition contribute to better adherence to safety protocols and ethical standards [30, 31]. Empirical findings indicate that ethical climate, leadership authenticity, and psychological safety significantly predict adherence to safety protocols and ethical competence [7, 9]. When professionals perceive institutional recognition and moral support, they are more likely to act as ethical agents committed to safe and qualified care, translating professional identity into concrete safety-oriented practices. Accordingly, the fifth hypothesis is proposed: H5 Recognition, Power, and Identity (RPI) positively influence Safe and Qualified Care (SQC). Safe and qualified care depends on effective teamwork, open communication, and shared accountability. Recent studies have shown that the presence of a strong safety culture enhances collaboration among multidisciplinary teams and reinforces trust, leading to improved ethical performance and collective responsibility [37, 38]. Evidence from interprofessional research demonstrates that psychological safety strengthens teamwork effectiveness and patient safety competence, particularly in high-risk settings [9, 26]. Thus, safety culture operates as a structural mechanism that strengthens teamwork dynamics by fostering trust, communication, and shared responsibility. H6 Safe and Qualified Care (SQC) positively influences Work Teams (WT). Safe work environments foster professional satisfaction and positive perceptions of working conditions. Empirical evidence indicates that strong safety and quality cultures are associated with lower burnout, greater job satisfaction, and improved psychological safety, while ethical climate further contributes to well-being and retention outcomes among healthcare professionals [23–25, 38, 39]. Therefore, safety culture contributes not only to patient outcomes but also to healthier organizational environments by shaping how professionals perceive support, stability, and fairness in their working conditions. H7 Safe and Qualified Care (SQC) positively influences Working Conditions (WC). Taken together, these hypotheses propose an integrated framework in which recognition and professional identity act as primary drivers, while safety culture functions as a structural mechanism influencing teamwork dynamics, working conditions, and the visibility of ethical violations. Comparative analyses across professional groups are essential to understand how moral distress manifests in complex healthcare environments. Differences in professional roles, levels of autonomy, ethical responsibility, and decision-making authority may influence not only the intensity of moral distress dimensions but also the way these dimensions relate to one another within organizational contexts [29]. Studies examining interprofessional hierarchy and emotional dynamics in clinical environments highlight how hierarchical positioning and authority structures shape ethical tensions and moral agency differently across professional groups [8, 11, 13]. In surgical settings, physicians and nursing staff operate under shared organizational constraints but experience distinct ethical pressures linked to their professional scope, hierarchical positioning, and interprofessional interactions. These differences may lead to variations in how recognition, professional identity, safety culture, teamwork, and working conditions interact in shaping moral distress. Therefore, examining whether the structural relationships among moral distress dimensions differ between professional groups can provide insights into profession-specific ethical dynamics within surgical teams. H8 The structural relationships among the dimensions of moral distress differ between physicians and nursing staff working in surgical settings. Furthermore, beyond relational patterns, it is equally relevant to assess whether physicians and nursing staff differ in the distribution and intensity of moral distress dimensions. Comparative statistical analyses allow the identification of profession-specific vulnerabilities and ethical stressors, supporting targeted managerial and organizational interventions. H9 There are statistically significant differences between physicians and nursing staff in the levels of moral distress dimensions measured by the EDME-BR-VR. Following the presentation of the hypotheses, Fig. 1 illustrates the proposed theoretical model, summarizing the conceptual pathways linking the dimensions of the Brazilian Moral Distress Scale (EDME-BR-VR). The model serves as a conceptual framework to support the comparative examination of structural relationships and moral distress patterns between physicians and nursing staff working in surgical settings. Method Study design This is a cross-sectional comparative study conducted with physicians and nursing staff working in surgical settings. Moral distress was assessed using the Brazilian Moral Distress Scale – Short Version (EDME-BR-VR). Comparative analyses were performed at two levels: (i) nonparametric statistical tests were first applied to examine potential internal heterogeneity within the nursing staff (nurses vs. nursing technicians); (ii) subsequent nonparametric comparisons were performed to assess differences in moral distress dimensions between physicians and nursing staff; and (iii) Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to compare the structural relationships among moral distress dimensions across professional groups. Research setting The study was conducted at a public university hospital located in southern Brazil, with approximately 400 beds. The institution provides medium- and high-complexity care exclusively through the Brazilian Unified Health System (SUS) and serves as a regional referral center for specialized healthcare services. The hospital operates with advanced medical technologies and a multidisciplinary workforce, including healthcare professionals and students. Data were collected in the Surgical Center, encompassing the operating rooms and the post-anesthesia care unit, where physicians and nursing staff work collaboratively in perioperative care. Population, selection criteria, and sample definition The study population consisted of healthcare professionals working in the Surgical Center of the hospital. A total of 245 professionals participated in the study, including physicians (n = 167) and nursing staff (nurses; n = 18 and nursing technicians; n = 60). Eligibility criteria required participants to be actively working in the operating room during the data collection period. Professionals who were on vacation, medical leave, or any type of work leave at the time of data collection were excluded. All 281 eligible professionals working in the Surgical Center were invited to participate in the study. A total of 245 professionals agreed to participate, resulting in a response rate of 87.1%. Participation was voluntary and anonymous. Despite the high response rate, the possibility of non-response bias cannot be completely ruled out. To assess sample adequacy for structural equation modeling, an a priori power analysis was conducted using Soper’s SEM sample size calculator [42]. Considering six latent variables, 24 observed variables, an anticipated small effect size (0.10), statistical power of 0.80, and a significance level of 0.05, the minimum required sample size was estimated at 100 participants. The final sample exceeded this threshold, supporting the adequacy of the dataset for Partial Least Squares Structural Equation Modeling (PLS-SEM). Although the sample included a larger proportion of physicians compared to nursing staff, this distribution reflects the composition of the surgical workforce in the studied hospital. Nevertheless, this imbalance should be considered when interpreting the generalizability of the findings. Additionally, the small number of nurses (n = 18) compared to nursing technicians (n = 60) limits the statistical power of internal subgroup comparisons and should be interpreted with caution. Although all eligible professionals were invited to participate, no formal comparison with the overall workforce composition was conducted, which may limit conclusions regarding sample representativeness. Instruments used The data collection instrument consisted of a sociodemographic and work questionnaire and the Brazilian Scale of Moral Distress in Nurses in its reduced version (EDME-BR-VR). The sociodemographic and work questionnaire addressed the following variables: sex, marital status, age group, education, and length of professional training. The assessment instrument was developed and validated in Brazil [1, 2], and its short version was validated [27], called the “Brazilian Scale of Moral Distress in Nurses – Short Version” (EDME-BR-VR), consists of 24 questions distributed across six dimensions, assessed using a seven-point double Likert scale to analyze both the intensity and frequency of moral distress. Although originally developed for nursing populations, the EDME-BR-VR was applied in this study to physicians due to its focus on organizational, relational, and ethical dimensions that are common across healthcare professions working within the same clinical context. Nevertheless, the use of an instrument originally designed for nurses may represent a limitation when applied to other professional groups, and the results should be interpreted with caution regarding cross-professional validity. Statistical analysis The database was constructed using independent double data entry in Microsoft Excel, and all analyses were conducted using SPSS® (version 26) and SmartPLS® (version 4.1.1.7) [42]. Descriptive statistics were used to summarize and compare the sociodemographic characteristics and EDME-BR-VR dimension scores between physicians and nursing staff. Means and standard deviations were calculated for each professional group to support the comparative analysis of moral distress across categories. Comparative analyses were conducted at three complementary levels. First, nonparametric statistical tests were applied to examine potential internal heterogeneity within the nursing staff (nurses vs. nursing technicians). As no statistically significant differences were identified, the nursing subgroups were aggregated for subsequent analyses. Second, nonparametric comparisons were performed between physicians and nursing staff. Third, Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to compare structural relationships among moral distress dimensions across professional groups. To address the comparative objectives of the study, analyses were conducted at two complementary levels. First, nonparametric statistical tests were applied to compare the distribution of moral distress dimensions between physicians and nursing staff. Given the independent groups and the non-normal distribution of the data, the Mann–Whitney U test was used to assess differences in median scores across professional categories, in accordance with Hypothesis H9. Statistical significance was set at p < 0.05. Second, Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed to examine and compare the patterns of relationships among the dimensions of moral distress across professional groups, addressing Hypothesis H8. Separate structural models were estimated for physicians and nursing staff, allowing for the comparison of path coefficients and relational patterns without imposing measurement or structural invariance assumptions Before conducting multigroup comparisons, measurement invariance was assessed using the Measurement Invariance of Composite Models (MICOM) procedure, in order to verify the comparability of the constructs across physicians and nursing staff. After establishing the required level of invariance, separate structural models were estimated for each group and compared using multigroup analysis. Model fit in PLS-SEM was assessed using the Standardized Root Mean Square Residual (SRMR), geodesic distance (GD), Euclidean Square Distance (SED), and Normed Fit Index (NFI). An SRMR threshold of 0.08 and an NFI value of at least 0.80 were adopted as indicators of acceptable model fit [43, 44]. The SRMR value was below the recommended threshold of 0.08, and the NFI exceeded the commonly reported reference value of 0.80. However, in line with current recommendations for PLS-SEM, these indices were interpreted as complementary indicators rather than primary criteria for model evaluation. Accordingly, the assessment of the model focused primarily on the measurement model quality, the significance of structural relationships, and the predictive performance of the model. The evaluation of the measurement model focused on internal consistency, convergent validity, and discriminant validity. Internal consistency was assessed using Cronbach’s alpha (CA > 0.70) and composite reliability (0.70 < CR < 0.95). Convergent validity was evaluated using the Average Variance Extracted (AVE > 0.50), while discriminant validity was examined using the Fornell–Larcker criterion, which compares the square root of AVE for each construct with its correlations with the other constructs ( should exceed the inter-construct correlations) [45–47]. The significance of structural relationships was assessed using a bootstrapping procedure with 5,000 subsamples. Path coefficients (β) were interpreted as indicating the strength and direction of associations between constructs. Path coefficients (β) were considered statistically significant when the associated t-values exceeded 1.96 (p < 0.05). Differences in path coefficients between physicians and nursing staff were examined through comparative analysis of the group-specific models, in line with the study’s comparative framework. Predictive accuracy and relevance of the structural models were evaluated using the coefficient of determination (R²) and predictive relevance (Q²). R² values above 0.19 were interpreted as indicating strong explanatory power, while values between 0.075 and 0.19 reflected moderate explanatory power. Predictive relevance was considered satisfactory when Q² values exceeded 0.075 [47, 48]. R² values were interpreted comparatively across endogenous constructs as indicators of explained variance, rather than using fixed thresholds. Q² values above zero were considered indicative of predictive relevance. Additionally, predictive performance was assessed using the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE), with lower values indicating greater predictive accuracy and model stability, consistent with the predictive-oriented nature of PLS-SEM [49]. Ethical considerations The ethical precepts of research involving human subjects were followed, in accordance with resolution nº. 510/2016 of the Brazilian National Health Council [50], with the project approved by the local Research Ethics Committee (REC) under Opinion nº. 5.852.705. Results Descriptive statistics A total of 245 professionals, the profile of respondents consists mainly of physicians (68.2%), followed by nursing professionals (31.8%). Most have more than 10 years of training (62.9%), with 51.8% having postgraduate degrees and 48.2% having undergraduate or technical degrees. In terms of age, most are between 41 and 50 years old (32.7%), followed by those between 31 and 40 years old (26.9%) (x̅ = 41.2 years, SD = 10.44). The group is predominantly composed of men (55.5%), while 44.5% are women. Most respondents (66.5%) live with a partner, and 28.2% live without a partner. Model fitting tests The PLS-SEM algorithm converged after seven iterations, indicating adequate model stability. To assess the fit of the structural models, multiple fit indices were examined, including the Standardized Root Mean Square Residual (SRMR), the Euclidean Distance Squared (EDS), the Geodesic Distance (GD), and the Normed Fit Index (NFI). The results indicated an acceptable fit of the structural model, supporting its use for comparative analysis of relational patterns among moral distress dimensions. The main fit statistics were SRMR = 0.077, EDS = 2.294, GD = 0.824, and NFI = 0.896. The SRMR value was below the recommended threshold of 0.08, and the NFI exceeded the acceptable cutoff of 0.80, indicating satisfactory model fit for the purposes of comparative structural examination [44]. Internal subgroup comparison of moral distress dimensions within the nursing staff To examine potential internal heterogeneity within the nursing staff group, an additional comparative analysis was conducted between nurses (n = 18) and nursing technicians (n = 60). As presented in Table 1, no statistically significant differences were observed across any of the moral distress dimensions (p > 0.05). Table 1Comparative analysis of moral distress dimensions between nurses and nursing technicians using the Mann–Whitney U testFactorNurses (n = 18)Nursing technician (n = 60)MW(p-value)Mean ± SDMean ± SDRecognition, Power, and Identity2.5 ± 1.792.3 ± 1.680.810Safe and Qualified Care2.6 ± 1.263.1 ± 1.700.510Defense of Values and Rights1.6 ± 1.971.4 ± 1.620.713Working Conditions2.5 ± 1.422.1 ± 1.710.646Ethical Violations3.1 ± 1.692.9 ± 1.740.708Work Teams3.3 ± 1.303.1 ± 1.420.629Note. SD = standard deviation; MW = Mann-Whitney U Test Comparative analysis of moral distress dimensions between nurses and nursing technicians using the Mann–Whitney U test Note. SD = standard deviation; MW = Mann-Whitney U Test Mean scores were comparable between nurses and nursing technicians for Recognition, Power, and Identity; Safe and Qualified Care; Defense of Values and Rights; Working Conditions; Ethical Violations; and Work Teams. These findings indicate a consistent pattern of moral distress experiences within the nursing staff, supporting the decision to treat nurses and nursing technicians as a single analytical group in subsequent comparative structural analyses. Although nurses and nursing technicians differ in scope of practice and hierarchical positioning, the absence of statistically significant differences in the distribution of moral distress dimensions suggests that shared organizational constraints and ethical pressures within the surgical environment may outweigh differences in professional autonomy. This reinforces the collective and context-dependent nature of moral distress within the nursing team. Comparative evaluation of moral distress dimensions between physicians and nursing staff Based on the results presented in Tables 2 and 3, moral distress dimensions were examined through a comparative analytical framework combining descriptive statistics, nonparametric group comparisons, and structural modeling. Mean scores and standard deviations were calculated separately for physicians and nursing staff to support the comparison of moral distress experiences across professional categories. Table 2Comparative analysis of moral distress dimensions between physicians and nursing staff (Hypothesis H9)FactorNurses staff (n = 78)Physicians (n = 167)MW(p-value)Mean ± SDMean ± SDRecognition, Power, and Identity2.3 ± 1.702.3 ± 1.740.888Safe and Qualified Care2.7 ± 1.663.2 ± 1.540.013Defense of Values and Rights1.5 ± 1.752.0 ± 1.740.088Working Conditions3.4 ± 1.644.0 ± 1.500.002Ethical Violations2.2 ± 1.782.5 ± 2.050.421Work Teams3.2 ± 1.393.3 ± 1.480.561Note. SD = standard deviation; MW = Mann-Whitney U Test Comparative analysis of moral distress dimensions between physicians and nursing staff (Hypothesis H9) Note. SD = standard deviation; MW = Mann-Whitney U Test Table 3Measurement model assessment for nursing staff and physiciansFactorIndicatorNursing staff (n = 78)Physicians (n = 167)LoadingCACRAVELoadingCACRAVERPI0.8360.8910.6730.8030.8700.630 RPI.10.7240.683 RPI.20.8090.724 RPI.30.9150.898 RPI.40.8210.850SQC0.8870.9230.7490.7750.8430.561 SQC.10.9140.780 SQC.20.8910.815 SQC.30.8430.767 SQC.40.8100.737DVR0.8660.9110.7200.8190.8820.653 DVR.10.6980.679 DVR.20.9030.814 DVR.30.8980.876 DVR.40.8790.849WC0.8830.9190.7400.8490.8980.687 WC.10.8610.832 WC.20.9050.871 WC.30.8260.803 WC.40.8470.808EV0.8770.9150.7300.9350.9430.837 EV.10.8350.882 EV.20.8990.927 EV.30.8310.910 EV.40.8520.938WT0.7870.8590.6040.7660.8490.585 WT.10.7250.780 WT.20.7820.781 WT.30.8060.756 WT.40.7940.741Note. DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care; RPI = Recognition, Power, and Identity; CA = Cronbach’s alpha; CR = Composite Reliability; AVE = Average Variance Extracted Measurement model assessment for nursing staff and physicians Note. DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care; RPI = Recognition, Power, and Identity; CA = Cronbach’s alpha; CR = Composite Reliability; AVE = Average Variance Extracted In Table 2, regarding the comparison of moral distress dimensions between groups, statistically significant differences were observed in Safe and Qualified Care (SQC), Defense of Values and Rights (DVR), and Working Conditions (WC). Nursing staff reported higher mean scores for SQC ( = 3.2; sd = 1.54) compared to physicians ( = 2.7; sd = 1.66), with this difference reaching statistical significance (Mann–Whitney U, p = 0.013). Similarly, DVR scores were significantly higher among physicians ( = 2.0; sd = 1.74) than among nursing staff ( = 1.5; sd = 1.75; p = 0.008). Differences were also identified for Working Conditions, with physicians reporting higher levels of moral distress ( = 4.0; sd = 1.50) compared to nursing staff ( = 3.4; sd = 1.64; p = 0.002). Statistically significant differences were observed for Safe and Qualified Care and Working Conditions (p < 0.05). In contrast, no statistically significant differences were identified for Recognition, Power, and Identity (p = 0.888), Defense of Values and Rights (p = 0.088), Ethical Violations (p = 0.421), or Work Teams (p = 0.561), suggesting similar perceptions across professional groups for these dimensions. In addition to group-level comparisons, the measurement properties of the EDME-BR-VR dimensions were examined to support the interpretation of comparative and structural analyses (Table 3). All dimensions demonstrated satisfactory internal consistency, with Cronbach’s alpha values exceeding the recommended threshold of 0.70 across both professional groups. Composite reliability values were within acceptable ranges, indicating stability of the constructs for comparative purposes. Convergent validity was supported by Average Variance Extracted (AVE) values above 0.50 for all dimensions, reflecting adequate item contribution to their respective constructs. Overall, these findings indicate that while the EDME-BR-VR dimensions exhibit consistent measurement properties across professional categories, meaningful differences exist in the intensity of specific moral distress dimensions between physicians and nursing staff. These comparative results provide empirical support for subsequent analyses of relational patterns using PLS-SEM and highlight profession-specific ethical stressors within surgical settings. In Table 4, according to the Fornell–Larcker criterion, discriminant validity is supported when the square root of the Average Variance Extracted () for each construct exceeds its correlations with the other constructs [45], the values are presented on the diagonal and compared with the corresponding Pearson correlation coefficients for nursing staff and physicians. Table 4Comparative assessment of discriminant validity using the Fornell–Larcker criterion between nursing staff and physiciansFactors Pearson Correlation Matrix (Nursing staff)DVREVRPISQCWCWTDVR0.8491.000EV0.8550.6361.000RPI0.8200.6950.5341.000SQC0.8660.6460.6460.5861.000WC0.8600.4380.4110.4940.4041.000WT0.7770.5330.6360.6650.5210.6481.000Pearson Correlation Matrix (Physicians)DVR0,8081,000EV0,9150,5601,000RPI0,7940,6280,4481,000SQC0,7490,6130,5330,4311,000WC0,8290,3240,2920,4230,3921,000WT0,7650,4900,4440,6660,4410,5921,000AVE = Average Variance Extracted; DVR = Defense of Values and Rights; EV = Ethical Violations; RPI = Recognition, Power, and Identity; SQC = Safe and Qualified Care; WC = Working Conditions; WT = Work Teams Comparative assessment of discriminant validity using the Fornell–Larcker criterion between nursing staff and physicians AVE = Average Variance Extracted; DVR = Defense of Values and Rights; EV = Ethical Violations; RPI = Recognition, Power, and Identity; SQC = Safe and Qualified Care; WC = Working Conditions; WT = Work Teams For both professional groups, the values of all moral distress dimensions were higher than the correlations with the remaining constructs, indicating that each dimension shared more variance with its own indicators than with other dimensions. This pattern was consistently observed for nursing staff and physicians, supporting adequate discriminant validity of the EDME-BR-VR dimensions for comparative analysis between professional categories. These findings indicate that the constructs assessed by the EDME-BR-VR are empirically distinct within each group, allowing meaningful comparison of moral distress dimensions and subsequent examination of differences in relational patterns across physicians and nursing staff. Comparative analysis of structural relationships between physicians and nursing staff Hypotheses H1 to H8 examined whether the structural relationships among the EDME-BR-VR dimensions were significant within each professional group and whether these relationships differed between physicians and nursing staff. The results of the comparative PLS-SEM analysis are presented in Table 5, including path coefficients (β), p-values, and multigroup comparison tests (Henseler’s MGA and permutation tests). Table 5Structural coefficients and multigroup analysis (H8)Hip.Relationshipβ (p-value)Invariance Test (H8) (p–value)Nurses staffPhysiciansHenseler MGAPermutation TestH1RPI → DVR0.695 (0.000)0.628 (0.000)0.4150.444H2RPI → WT0.548 (0.000)0.584 (0.000)0.7700.733H3RPI → WC0.391 (0.000)0.312 (0.000)0.5600.533H4RPI → EV0.534 (0.000)0.448 (0.000)0.4260.447H5RPI → SQC0.586 (0.000)0.431 (0.000)0.6260.184H6SQC → WT0.200 (0.003)0.189 (0.002)0.1810.952H7SQC → WC0.174 (0.005)0.258 (0.000)0.9280.618Note. RPI = Recognition, Power, and Identity; DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care Structural coefficients and multigroup analysis (H8) Note. RPI = Recognition, Power, and Identity; DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care All proposed direct relationships were statistically significant (p < 0.05), with moderate to strong effect sizes across the proposed paths. The strongest association was observed for RPI → DVR in both nursing staff (β = 0.695, p < 0.001) and physicians (β = 0.628, p < 0.001), confirming the central role of Recognition, Power, and Identity in ethical positioning. Some differences in effect magnitude were observed across groups. The relationship RPI → SQC was descriptively higher among nursing staff (β = 0.586, p < 0.001) than physicians (β = 0.431, p < 0.001), whereas RPI → WT showed a slightly higher coefficient among physicians (β = 0.584, p < 0.001) compared to nursing staff (β = 0.548, p < 0.001). These differences should be interpreted descriptively, as no statistically significant differences were observed between groups. However, multigroup comparisons (Henseler’s MGA and permutation tests) indicated no statistically significant differences between groups for any structural paths (p > 0.05), supporting invariance of the structural model. These results indicate that, despite descriptive variations in effect sizes, the overall pattern of relationships is comparable between physicians and nursing staff within surgical settings. Comparative analysis of explanatory power and predictive relevance between physicians and nursing staff The explanatory power and predictive relevance of the structural models were examined separately for nursing staff and physicians using the coefficient of determination (R²), predictive relevance (Q²), and predictive accuracy indicators (RMSE and MAE), as presented in Table 6. Table 6Comparative analysis of explanatory power and predictive relevance between physicians and nursing staffPredictive factorsCoefficients (Nurses staff)R² (p-value)Q²RMSEMAEDefense of Values and Rights (DVR)0.483 (0.000)0.4580.2530.312Ethical Violations (EV)0.285 (0.003)0.7590.8890.849Safe and Qualified Care (SQC)0.343 (0.001)0.5550.7220.670Working Conditions (WC)0.264 (0.001)0.4580.2530.312Work Teams (WT)0.468 (0.000)0.7590.8890.849Coefficients (Physicians)Defense of Values and Rights (DVR)0.395 (0.000)0.3810.1890.181Ethical Violations (EV)0.201 (0.000)0.7970.9080.907Safe and Qualified Care (SQC)0.186 (0.001)0.6420.7650.828Working Conditions (WC)0.233 (0.000)0.3810.1890.181Work Teams (WT)0.472 (0.000)0.7970.9080.907Note. R² = coefficient of determination; Q² = predictive relevance (Stone–Geisser); RMSE = root mean square error; MAE = mean absolute error Comparative analysis of explanatory power and predictive relevance between physicians and nursing staff Note. R² = coefficient of determination; Q² = predictive relevance (Stone–Geisser); RMSE = root mean square error; MAE = mean absolute error Among nursing staff, the highest explanatory power was observed for Defense of Values and Rights (DVR) (R² = 0.483; p < 0.001) and Work Teams (WT) (R² = 0.468; p < 0.001), indicating strong predictive capacity for ethically oriented and team-related dimensions. Safe and Qualified Care (SQC) (R² = 0.343; p = 0.001) demonstrated moderate explanatory power, while Ethical Violations (EV) (R² = 0.285; p = 0.003) and Working Conditions (WC) (R² = 0.264; p = 0.001) showed lower but still meaningful levels of explained variance. For physicians, the highest explanatory power was also observed for Work Teams (WT) (R² = 0.472; p < 0.001) and Defense of Values and Rights (DVR) (R² = 0.395; p < 0.001), indicating a similar structural emphasis on collective dynamics and ethical positioning. In contrast, Safe and Qualified Care (SQC) presented a lower explanatory capacity among physicians (R² = 0.186; p = 0.001), while Ethical Violations (EV) (R² = 0.201; p < 0.001) and Working Conditions (WC) (R² = 0.233; p < 0.001) showed modest predictive strength. Predictive relevance, assessed using Q² values, indicated strong predictive capability for Ethical Violations (EV) and Work Teams (WT) in both professional groups (Q² ≥ 0.759), confirming the stability of these constructs across models. Defense of Values and Rights (DVR) and Working Conditions (WC) demonstrated moderate predictive relevance in both groups, whereas Safe and Qualified Care (SQC) showed comparatively lower Q² values, particularly among physicians, suggesting greater variability in its predictive structure. Regarding predictive accuracy, RMSE and MAE values indicated better calibration for DVR and WC in both groups, with lower prediction errors observed among physicians. In contrast, higher RMSE and MAE values for EV and WT in both models reflect greater dispersion in predicted values, although remaining within acceptable limits for comparative structural analysis. Overall, the comparative results indicate that while the general pattern of explanatory and predictive performance is similar between physicians and nursing staff, differences emerge in the strength and precision of specific dimensions—particularly Safe and Qualified Care—highlighting profession-specific dynamics in the experience of moral distress within surgical settings. Figure 2 illustrates the structural relationships among the dimensions of the Brazilian Moral Distress Scale – Short Version (EDME-BR-VR), presenting the standardized path coefficients (β) and corresponding p-values estimated separately for nursing staff and physicians. This graphical representation allows a direct comparison of the strength and direction of the relationships across professional groups. Comparative structural relationships among moral distress dimensions between nursing staff and physicians. Source: SmartPLS® software (version 4.1.1.7) [41]. Note: DVR = Defense of Values and Rights; WT = Work Teams; WC = Working Conditions; EV = Ethical Violations; SQC = Safe and Qualified Care; RPI = Recognition, Power, and Identity The model highlights the central role of Recognition, Power, and Identity (RPI) as a key antecedent of multiple moral distress dimensions, as well as the contribution of Safe and Qualified Care (SQC) to work team dynamics and working conditions. Differences in coefficient magnitude between nursing staff and physicians are visually displayed for each structural path, supporting the interpretation of profession-specific patterns in moral distress within surgical settings. The figure synthesizes the comparative PLS-SEM results reported in Table 6 and serves as a visual summary of the structural relationships observed for each professional group. Discussion This study provides a comparative analysis of moral distress among physicians and nursing staff working in surgical settings, combining nonparametric group comparisons and comparative PLS-SEM to examine both the distribution of moral distress dimensions and the relational patterns among them. Overall, the findings highlight important similarities and profession-specific differences in how moral distress is experienced and structured within multidisciplinary surgical teams. However, these differences should be interpreted in light of the statistical results and the specific context of the study. By integrating multigroup structural modeling with distributional comparisons, this study advances prior research that has predominantly examined moral distress within single professional categories or through descriptive approaches alone [19–21]. An additional internal subgroup analysis was conducted to examine potential heterogeneity within the nursing staff. No statistically significant differences were identified between nurses and nursing technicians across any of the moral distress dimensions. These findings should be interpreted with caution due to the small subgroup size, particularly for nurses (n = 18). This result strengthens the robustness of the comparative framework adopted in this study, suggesting that shared organizational constraints and ethical pressures within the surgical environment may outweigh differences in hierarchical positioning and scope of practice within the nursing team. The absence of internal divergence supports the analytical aggregation of nursing professionals and reinforces the interpretation of moral distress as a context-driven and organizationally embedded phenomenon in surgical settings. At the descriptive level, statistically significant differences were observed between physicians and nursing staff in the dimensions of Safe and Qualified Care (SQC), Defense of Values and Rights (DVR), and Working Conditions (WC). Nursing staff reported higher levels of moral distress related to SQC, suggesting greater ethical tension associated with patient safety, care continuity, and quality standards. This finding aligns with previous evidence indicating that nurses often experience heightened moral distress when organizational constraints compromise care quality or patient advocacy [2, 4, 5]. Studies conducted in hospital and pandemic contexts have similarly demonstrated that nurses frequently report elevated moral distress when safety standards, resource adequacy, or advocacy roles are threatened [22, 24, 26]. In contrast, physicians reported higher levels of moral distress related to DVR and WC, reflecting stronger ethical tensions linked to decision-making responsibility, resource allocation, and structural working conditions in surgical practice. This pattern is consistent with evidence suggesting that greater decisional authority and accountability may intensify ethical burden and responsibility-related distress among physicians [25, 27]. These findings indicate that moral distress is structured not only by exposure to ethical dilemmas but also by differentiated authority gradients and professional accountability within hierarchical surgical environments. Beyond differences in levels of moral distress, the comparative structural analysis revealed that the patterns of relationships among moral distress dimensions were statistically significant within both professional groups. Recognition, Power, and Identity (RPI) emerged as a central antecedent across models, exerting consistent positive effects on Defense of Values and Rights, Work Teams, Working Conditions, Safe and Qualified Care, and Ethical Violations. These results reinforce theoretical perspectives that emphasize professional recognition, autonomy, and identity as key organizational resources shaping ethical sensitivity and moral agency in healthcare settings [12, 51]. Contemporary research further indicates that ethical climate and leadership practices operate through recognition and psychological safety mechanisms, strengthening moral resilience and ethical voice behaviors [6–9]. Thus, the centrality of RPI in both models supports the interpretation of moral distress as an organizationally mediated phenomenon rather than solely an individual moral reaction. Importantly, recognition and professional identity appear to function as structural enablers of ethical agency in high-risk surgical settings, influencing whether professionals feel empowered to voice concerns, advocate for patients, and challenge ethically problematic practices. Although the overall pattern of structural relationships was similar between physicians and nursing staff, differences in the magnitude of specific path coefficients were observed at a descriptive level. For physicians, RPI showed relatively stronger associations with Work Teams, underscoring the importance of authority, leadership, and recognition in interprofessional coordination and team functioning. Conversely, among nursing staff, RPI demonstrated a stronger influence on Safe and Qualified Care, indicating that professional recognition and empowerment are more directly translated into practices related to patient safety and care quality. These differences should be interpreted descriptively, as no statistically significant differences were observed between groups in the multigroup analysis. These findings are consistent with the distinct ethical roles and scopes of practice that characterize medical and nursing professions within surgical environments [12, 30]. They further suggest that ethical agency is relational and context-dependent, mediated by hierarchical positioning, decisional authority, and perceived institutional support. Evidence from studies on psychological safety and authentic leadership suggests that empowerment may translate differently across hierarchical positions, influencing teamwork and safety-related behaviors according to professional authority and perceived institutional support [8–10, 26]. The comparative predictive analysis further supports these interpretations. While both models demonstrated satisfactory explanatory power and predictive relevance, differences emerged in the strength and precision of specific dimensions, particularly Safe and Qualified Care. Lower explanatory power and higher prediction error for SQC among physicians suggest greater variability in how safety-related ethical tensions are structured within this group, whereas nursing staff displayed more consistent predictive patterns for this dimension. Such variability may reflect differences in proximity to direct patient care processes and safety monitoring responsibilities, as suggested in studies examining safety culture and adverse event reporting behaviors in hospital environments [29, 30]. These interpretations should be considered within the limits of the cross-sectional design and the specific institutional context. From an organizational perspective, the findings underscore that moral distress in surgical settings is not solely an individual phenomenon but is deeply embedded in structural conditions, professional recognition, and interprofessional dynamics. However, these implications should be interpreted cautiously, as the study design does not allow causal inference. Interventions aimed at reducing moral distress should therefore prioritize institutional recognition practices, transparent decision-making processes, adequate working conditions, and safety-oriented cultures that acknowledge the distinct ethical pressures faced by physicians and nursing staff. Strategies grounded in ethical leadership development, promotion of psychological safety, and strengthening of professional identity may be particularly effective in mitigating moral distress and fostering collective moral resilience within surgical teams [7–9, 31–36]. Such initiatives require alignment between frontline leadership, hospital governance structures, and safety management systems to ensure that ethical voice and professional autonomy are not only encouraged but institutionally protected. Furthermore, promoting a “just culture” approach—where reporting ethical violations and adverse events is interpreted as organizational learning rather than individual failure—may enhance transparency and reduce moral burden associated with safety concerns [27, 29]. This study has some limitations. The cross-sectional design and the inclusion of professionals from a single university hospital limit causal inference and generalizability. Additionally, although comparative structural analysis was performed, the study did not examine indirect effects or longitudinal changes in moral distress. Moreover, the findings should not be generalized beyond similar institutional and organizational contexts. Future research should expand this comparative approach to other clinical contexts, professional categories, and healthcare systems, as well as explore longitudinal designs to better capture the dynamic nature of moral distress and its organizational determinants. Future studies may also investigate mediating mechanisms such as psychological safety or ethical climate using multigroup invariance testing to further clarify profession-specific structural pathways [41–43]. Despite these limitations, the present findings contribute to the literature by providing integrated comparative evidence on both the levels and relational structure of moral distress among physicians and nursing staff. This dual-level approach offers valuable insights for healthcare managers and policymakers seeking to design targeted, profession-sensitive interventions to promote ethical practice, psychological safety, and sustainable working environments in surgical care. These contributions should be understood within the empirical scope of the study. At the policy level, these findings suggest the importance of incorporating professional recognition, psychological safety, and safety-culture indicators into accreditation standards, workforce governance frameworks, and institutional quality-monitoring systems. By situating moral distress within an organizational, leadership, and safety-culture framework, this study advances a more systemic understanding of ethical strain in high-risk clinical settings. Conclusion This study provides a comparative examination of moral distress among physicians and nursing staff working in surgical settings, offering evidence on both the distribution of moral distress dimensions and the structural relationships that underpin this phenomenon across professional groups. The findings indicate that, although physicians and nursing staff experience moral distress within a shared organizational context, differences are primarily observed in the intensity of specific dimensions rather than in the overall structural pattern of relationships. Recognition, power, and professional identity consistently emerged as central elements shaping moral distress across both groups, reinforcing their relevance as key organizational and psychosocial resources in ethically demanding environments. This study contributes to the literature by demonstrating that Recognition, Power, and Identity functions as a cross-professional structural driver of moral distress, supporting a systemic interpretation of the phenomenon grounded in organizational and relational dynamics. However, variations in the strength of structural relationships and predictive patterns highlight profession-specific dynamics, particularly regarding safe and qualified care, teamwork, and working conditions. These variations should be interpreted descriptively, as no statistically significant differences were observed in the multigroup structural comparisons. From a practical perspective, the results suggest that organizational factors such as professional recognition, safety culture, and working conditions are associated with moral distress across professional groups. These implications should be interpreted within the empirical scope of the study, while highlighting the potential relevance of organizational strategies sensitive to the distinct ethical roles and responsibilities of physicians and nursing staff. Although limited to a single hospital context and a cross-sectional design, this study provides an integrated analytical approach combining distributional comparisons and multigroup structural modeling, offering a replicable framework for cross-professional analysis of moral distress. Future research should extend this approach to other clinical contexts, professional categories, and healthcare systems, as well as explore longitudinal designs to better capture changes in moral distress over time and the impact of organizational factors. These directions may support a more context-sensitive and evidence-based understanding of moral distress in healthcare settings. Author contributions J.L.C.: Conceptualization, Data curation. J.L.C., L.F.D.L., F.R.S.R., and G.L.D.: Formal analysis, Investigation. L.F.D.L.: Methodology, Project administration, Resources, Software, Supervision, Validation. J.L.C., L.F.D.L., F.R.S.R., and G.L.D.: Visualization, Writing – original draft, Writing – review & editing. All authors read and approved the final manuscript. Funding The authors declare that no specific funding was received for this study. Data availability The datasets generated and/or analyzed during the current study are not publicly available due to institutional and ethical restrictions but are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The research is part of a matrix project, entitled “Moral suffering, moral resilience, and moral courage of health and nursing professionals at a university hospital”, approved by the Research Ethics Committee (REC) of the Universidade Federal de Santa Maria (UFSM), with the Ethical Appreciation Presentation Certificate (EAPC) number: 66373322.1.0000.5346. All procedures involving human participants were conducted in accordance with the ethical standards of the institutional research committee and in compliance with the principles of the Declaration of Helsinki and its subsequent amendments. Informed consent was obtained from all individual participants included in the study prior to data collection. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets generated and/or analyzed during the current study are not publicly available due to institutional and ethical restrictions but are available from the corresponding author on reasonable request.