This cross-sectional survey was originally designed as pilot study, which was conducted before the COVID-19 pandemic. The survey period was from mid-July 2014 to the beginning of August 2014 (department of trauma surgery) and the beginning of October 2016 (control center for EMS Munich). The total sample size across all groups (ED, EN, and TS) was N = 77, who were invited to voluntarily participate in the study.

Inclusion criteria: Specifically, the survey included all EN (N = 19) and all TS (N = 30) in the department of trauma surgery from a high level I university hospital, representing a complete census of these populations. In contrast, the survey at the control center for EMS Munich was a purposive sample, which was selected by the director and deputy director of the Munich fire brigade control center, resulting in a targeted sample of N = 28 ED. ED were between 35 and 58 years old and by the time of the survey they had at least about 3 years of experience in the control center for EMS Munich. Of note, ED profession in Germany is traditionally a male profession, as they work as firefighters in addition to dispatching.

Exclusion criteria: Non-medical staff such as administrative staff and people not directly employed by the trauma surgery department, such as medical staff from other medical departments, medical students, or ED with less than 3 years of work experience in the control center for EMS Munich were excluded from participation.

The invitation for voluntary participation in the written employee survey was communicated to TS during the daily morning meeting and to EN during a separate nursing team meeting, which was also attended by the head of surgical emergency department and the director of the department of trauma surgery who expressed their positive opinion of the project. EN and TS received the paper survey by internal post including an envelope for anonymous return to the secretariat of the department of trauma surgery.

In relation to ED, preselected ED were invited to voluntarily participate in the anonymous paper-based employee survey via e-mail and were informed about the aim and process of the project in a 10-min PowerPoint presentation followed by a Question-and-Answer session. The study was conducted as a paper-and pencil-survey (didactic study design) in a total of four available time blocks before or after the work shift at the control center for EMS Munich.

The short version of the instrument for stress-related job analysis for hospital physicians (ISAK-K), version 01/2013 (16, 22), from the German social accident insurance (DGUV) for non-state institutions within the health and welfare service sectors (BGW) is a well-established, reliable and valid questionnaire that was originally developed by the BGW for measuring work-related stressors and resources in hospital physicians and is based on the extension of the transactional stress model of Lazarus and Folkman (16, 22). ISAK-K can be downloaded free of charge from the BGW website² and via Leibniz Center for Psychological Information and Documentation (23). Linguistic and contextual modifications were made to accommodate the unique characteristics and requirements of ED (see Supplementary material: Adapted survey questions of the ISAK-K for emergency dispatchers). The questions were closed-ended and could be rated on a 5-point Likert scale from 1 to 5. The survey took approximately 5–10 min to complete. Consent was implied by completing and returning the survey. The 4-page questionnaire ISAK-K consists of 30 items in 14 scales including 7 stressors and 7 resources:

Stressors	Resources
Social stressors with patients and families/callers Uncertainty Time pressure Emotional dissonance Frustration about how work needs to be done Problems in workflow with other professions Problems in workflow (own professional group or supervisors)	Support from colleagues Support from supervisors Professional and skill development at work Possibilities in further education Autonomy at work Justice (fairness in the distribution of tasks and support) Participation at work

The staff councils of both institutions, the control center for emergency medical services (EMS) Munich and the TUM School of Medicine and Health, Department of Trauma Surgery Munich, have approved the employee survey as part of staff development measures. This approval complies with the requirements of the German Occupational Health and Safety Act (Arbeitsschutzgesetz §5 III Nr. 6)³ on psychological burdens in the workplace. Additionally, the department of Human Resources and Organization of the City of Munich approved and accompanied the study at the control center for EMS Munich.

In accordance with the legal regulations applicable in Germany, no formal ethical approval was required for this study. Employee-related topics such as staff surveys as part of staff development initiatives are approved by the Staff Council (see Permission of the Staff Council), which gave its approval for this study. It should be emphasized, that the questionnaire used in this study was a validated instrument provided by the German social accident insurance (DGUV) for non-state institutions within the health and welfare service sectors (BGW), which was developed specifically for staff development measures for hospital physicians (see section 2.3). All participants were given comprehensive verbal and written information about the project prior to their participation. By voluntarily participating in the survey and submitting their questionnaires, all participants gave their informed consent to take part in the study. To maximize anonymity, no socio-demographic data were collected in this study.

Data analysis was performed by downloading the ISAK-K analysis program free of charge from the BGW website.⁴ Surveys with incomplete answers or multiple answers to a question were excluded due to limitations of the ISAK-K software. The data set was extracted using Excel and the descriptive data including median and interquartile ranges were calculated after transferring the data to IBM SPSS for Windows 2000. Differences between professions were analyzed using the Mann–Whitney U test, whereby the significance level α was set at 5%. The standard scatter range of the BGW was based on the range of the ISAK-K reference study (16) and median values were kindly provided by the authors of the original ISAK-K development (22). Box plots were generated with the statistical program R and PowerPoint for Mac OS 2021.

The comparison between professions showed significant differences between ED, EN and TS regarding the evaluation of major stressors (p < 0.05) (Figure 2; Table 1). ED and EN reported significantly higher perceptions of stress compared to TS for social stressors and emotional dissonance (p < 0.05) (Figure 2; Table 1). At least 50% of ED (med = 4.0; IQR = 3) and at least 50% of EN (med = 4.0; IQR = 1) reported having to suppress their feelings ‘almost every day’ to ‘several times a day’ (Figure 2). The comparison between ED and EN further showed that ED rated social stressors significantly higher than nurses (p = 0.18), whereas there was no significant difference in emotional dissonance (p = n.s.; Table 1). TS rated the emotional dissonance factor significantly lower than ED (p = 0.003) and EN (p = 0.046) (Table 1).

ED, EN and TS all indicated a tendency for uncertainty and frustration at high stressor levels, whereas time pressure has been found as major stressor among ED and TS but not EN. ED rated uncertainty in decision-making during work significantly higher than EN (p = 0.000) and TS (p = 0.000; Table 1). The responses of both EN and TS showed an overall wider range (p = n.s.; Figure 2; Table 1). ED and TS reported significantly higher affectedness than EN for time pressure (p < 0.01; Figure 2; Table 1). The comparison between ED and TS further showed that ED rated time pressure significantly higher than TS (p = 0.011). All occupations showed a trend towards a moderately high level of frustration (p = n.s.; Figure 2; Table 1).

Overall, all occupations showed a broadly balanced resource profile, with four out of seven resources in the high scatter or benchmark range (Figure 3). This is an important result, especially with the high stressor profiles found among ED as described in the last section. In addition, ED rated autonomy at work and possibilities in further education within the scatter range of high resources (Figure 3). However, the results clearly showed a trend towards optimizing and strengthening specific work resources, such as work participation for all professional groups, and individual factors, such as strengthening work autonomy for emergency clinical staff (Figure 3).

Notably, ED, EN and TS all reported support from colleagues and superiors as their greatest positive resource (p = n.s.) and participation as their lowest resource (p = n.s., Table 1). The factor social support from colleagues in their own professional group was rated in the high resource range by ED (med = 4.0; IQR = 1), EN (med = 4.0; IQR = 1) and TS (med = 4.0; IQR = 1; Figure 4). At least 50% of each occupational group reported that they could rely ‘fairly’ to ‘completely’ on their team colleagues when things got difficult at work, and at least 50% of TS also reported that they could rely ‘fairly’ to ‘completely’ on their supervisors when things got difficult at work without group differences (med = 4.0; IQR = 2, Figure 4). With regard to autonomy and decision making, ED rated this factor as a higher resource than EN (p = 0.000) and TS (p = 0.002), but it should be noted that this factor is not directly comparable due to linguistic modifications of the survey question in relation to the working conditions among ED (see Methods section). Fairness in workplace task allocation was rated as a higher resource by TS than by ED (p = 0.015) but no significant difference was found between EN and TS or between EN and ED (Table 1).

Pre-pandemic studies identified constant high-complexity incoming calls as a significant stressor among ED with increased salivary cortisol levels (32, 33). The daily total cortisol concentrations among ED correlated with subjective perceptions of emotional distress (34) and other objective data such as elevated heart rate among ED have also been reported (18). Both chronically elevated cortisol levels and elevated heart rate are considered biomarkers of chronic stress with a disrupted hypothalamic–pituitary–adrenal axis and an increase in pro-inflammatory cytokines (35), which affect an individual’s epigenetics and correlate with many chronic inflammatory diseases (36). Subjectively perceived stressors therefore often reflect stress biomarkers (37), which is consistent with Lazarus and Folkmann’s transactional stress model, which describes that the primary and secondary evaluation of a difficult situation is crucial (14). Our data consequently indicate that ED have difficulty of coping with negative emotions (38) reflecting earlier pre-pandemic findings identifying communication issues with difficult callers among ED (39). Similarly, EN appear to face a comparable constant burden when dealing with difficult patients and their relatives in the emergency department, both before (40) and after the COVID-19 pandemic (41), in line with our findings. Pre-pandemic data demonstrate that 46% of EN actively experience physical aggression from intoxicated patients, compared to only about 20% of physicians (42), explaining the differences in social stressors between the nursing and physician professions. A retrospective study of a German high level I university hospital confirmed that violence in emergency departments occurs approximately every 0.7 days (43). Another aspect is that emergency physicians face a lower number of daily patient contacts and a higher proportion of indirect patient work compared to EN who have a higher percentage of direct patient work time (44–46). Moreover, it has been reported that the amount of patient communication correlates with the level of social stressors (22). Based on data of the German Trade Union Confederation (DGB) “Good Work Index,” the level of external interaction work further correlates with the level of emotional dissonance, especially in health care professions⁵ (47). Emotional dissonance indicates the conflict between the emotions experienced and the emotions expressed in order to fulfill the required external representation of the professional role (48). According to Ellis’ concept of Rational Emotive Behavior Therapy (REBT) the ability for rational thinking is a key factor for handling social stressors and emotional dissonance effectively (49). This concept has already been shown as an effective intervention for reducing irrational performance beliefs in firefighters (50). Most physicians working in a high-level I university hospital are typically involved in publishing scientific papers on a regular basis and are therefore trained in scientific thinking, which may partly explain why they experience less emotional dissonance. In view of the fact, that emotional dissonance correlates with irritability and emotional exhaustion (22) and has been shown to be a stronger indicator of burnout than time pressure (51), the high levels of emotional dissonance among ED and EN suggest a specific call to action, such as implementing training modules to manage negative emotions, including concepts such as REBT (50, 52) or the Nonviolent Communication model of Marshall Rosenberg (53). Appropriate training modules in emotion regulation, resilience, and health education were already shown to reduce nurses’ workplace stress levels (52, 54, 55).

Time pressure has been reported as major stressor among German hospital physicians (15, 22, 56), independent from specialties (22) and age (56), indicating to a general problem such as the high documentation burden (57). Time pressure positively correlates with frustration about how work needs to be done including too much documentation (22), which is in line with our findings. Interestingly, surgeons have been found with the highest work ability index and the lowest depression scores compared to other specialties, even though they are exposed to a high level of stress, such as time pressure (58), indicating high psychological resilience. EN did not rank time pressure as a major stressor, which is consistent with other research data (41). Nurses in health care systems like Germany have less responsibility and decision-making authority, which might correlate with less time pressure (59). An approach to reduce time pressure is to focus on reducing frustration first through specific training of key qualifications for handling administrative work and also on organization level including a reduction in documentation and clear instructions such as standard operating procedures (22). However, the most useful intervention to reduce physician stress appears to be the further development of the framework and range of non-patient tasks, such as documentation, using technological tools such as artificial intelligence (22, 60).

In terms of uncertainty at work, data during the COVID-19 pandemic showed that nurses had significantly higher levels of anxiety, depression and general fear compared to physicians and paramedics, combined with lower levels of subjective information, suggesting that regular educational training has the potential to improve confidence and counteract uncertainty by providing sufficient background knowledge, particularly for EN (61). Insufficient information from physicians about the patient’s medical condition has been reported of being the root cause of nurses’ uncertainty (41). A high level of uncertainty among ED with regard to emergency calls has been shown to correlate with lack of guidance, feedback and training (39), supporting the valuable factor of encouraging and productive feedback as leadership tool (31) which was recently reported in EMS staff associated with improved clinical performance (62). Internal managers and policy makers should be aware that if stress profiles remain high, intentions to leave will increase (20).

According to Johnson and Hall’s extension of Karasek’s job-demand-control model (63), a high level of social support acts as a positive modulating factor for occupational stress and chronic diseases (64). Indeed, social support at work and mindfulness correlated with stress levels among ED (65). Data from the German Stress Report 2019 have shown that 80% of employees in healthcare professions rate social support from colleagues as frequent and social support from supervisors as more than 50% frequent, which has remained consistent over the past few years (66), in line with our findings. This fact is particularly remarkable as other studies have found low team support among emergency medicine staff (67, 68). Moreover, since our data indicate a positively strong resource in terms of both colleagues and supervisors compared to other data (16, 56), it is reasonable to assume a positive, exemplary leadership culture correlates with improved team support, which is an essential factor particularly in HRO such as the emergency department and EMS control center (5).