A primer in resilience status for German medical graduates - a necessary step in building an emotionally equipped healthcare workforce


 BackgroundResilience is a widely-used catchword in the last couple of years to describe the resistance to psychological strains of life, especially for the healthcare work force. The promises of resilience to burnout sound great and what we all would want: less health impairment despite stress, higher work satisfaction and last but not least higher work performance. However, little is known scientifically regarding the resilience status of the upcoming work-force. With our study we would like to investigate the resilience status of medical graduates from five medical schools within their first year after graduation. MethodsFor the identification of the resilience status we included the 5-point Likert 10-Item Connor-Davidson Resilience Scale, German Version in a graduate survey posted to 5 medical schools and over 1610 eligible participants of whom 610 (60% female) filled out at least parts of the survey. ResultsThe resilience status showed a mean resilience score of M = 37.1 (SD = 6.30). The score ranges from 3.22 (I am not easily discouraged by failure) to 4.26 (I am able to adapt to change). The item “I am able to handle unpleasant feeling” is interesting as one third of the participants did choose not to answer. Relationships to other constructs are presented in the article. ConclusionsThe study shows that the overall resilience status of medical graduates one year after their graduation is rather high, but subjectively they do not feel equivalently resilient for the different aspects they face in their job. Further research needs to see how trainings can provide medical students and professionals with the emotional coping skills that they lack in the moment.


Abstract Background
Resilience is a widely-used catchword in the last couple of years to describe the resistance to psychological strains of life, especially for the healthcare work force. The promises of resilience to burnout sound great and what we all would want: less health impairment despite stress, higher work satisfaction and last but not least higher work performance. However, little is known scienti cally regarding the resilience status of the upcoming work-force. With our study we would like to investigate the resilience status of medical graduates from ve medical schools within their rst year after graduation.

Methods
For the identi cation of the resilience status we included the 5-point Likert 10-Item Connor-Davidson Resilience Scale, German Version in a graduate survey posted to 5 medical schools and over 1610 eligible participants of whom 610 (60% female) lled out at least parts of the survey.

Results
The resilience status showed a mean resilience score of M = 37.1 (SD = 6.30). The score ranges from 3.22 (I am not easily discouraged by failure) to 4.26 (I am able to adapt to change). The item "I am able to handle unpleasant feeling" is interesting as one third of the participants did choose not to answer.
Relationships to other constructs are presented in the article.

Conclusions
The study shows that the overall resilience status of medical graduates one year after their graduation is rather high, but subjectively they do not feel equivalently resilient for the different aspects they face in their job. Further research needs to see how trainings can provide medical students and professionals with the emotional coping skills that they lack in the moment.

Background
Resilience is a widely-used catchword in the last couple of years to describe the resistance to psychological strains of life. The promises of resilience to burnout sound great and what we all would want: less health impairment despite stress (Grossman, Niemann, Schmidt, & Walach, 2004), higher work satisfaction (Shanafelt et al., 2015) and last but not least higher work performance (Loehr, Loehr, & Schwartz, 2005). However, over many professions little is known scienti cally regarding the resilience status of the work-force and if and how resilience can be improved. With our article we would like to investigate the resilience status of medical graduates from ve medical schools in Bavaria in their rst year after graduation. Before we do so, we will take a look at the scienti c constructs related to resilience that have been under investigation a lot longer: burnout and stress among medical students, physicians and healthcare staff.
Physician and nursing staff have been found to be prone to burnout (Buser, Schneller, & Wildgrube, 2007;Schüler & Dietz, 2004). By employing non-functioning coping mechanisms, physicians have been shown to change their social behavior and their attitudes towards their job and patients (Möller, Laux, Deister, & Schulte-Körne, 2013;Schüler & Dietz, 2004). In studies with nurses, a general quality of care impairment could be observed as well as medication errors (Maiden, Georges, & Connelly, 2011). The reason, again, for this can also be attributed to non-functioning coping mechanisms for di cult situations and emotional blunting. The consequence for burned-out staff and employers are grave and include decreased job performance (Parker & Kulik, 1995), reduced job commitment (Cherniss, 2016) and stressrelated health problems (Maslach, Schaufeli, & Leiter, 2001). For medical students a recent review reports burnout rates between 27% and 75%, a wide range with a lot of regional variance. For Germany no single study was included. A German study with data from 2008 medical students in their nal year reported a clinically relevant burnout rate of 20%, a medium burnout score over the whole sample (Koehl-Hackert et al., 2012). The sample has been drawn from one city and since the samples' nal year the medical education system in Germany has undergone some changes, so it is hard to say how representative the data is.
In the eld of medical education, little is known regarding the coping mechanisms for stress of students.
It is known that stress among medical students is rather high but dangerously high stress levels vary between 21% and 90% (Fares, Al Tabosh, Saadeddin, El Mouhayyar, & Aridi, 2016). However, when diagnosing di cult patients showing disruptive behavior, students' diagnostic accuracy was lower than when di cult patients showed friendly behavior (Hege et al., 2018). The reason, again, might lie in nonfunctioning coping mechanisms through which their working memory capacity is impaired; this is an effect that has been long known in cognitive psychological experiments (Eysenck, 1985) To summarize the results for medical students, physicians and healthcare staff for burnout and stress we have found evidence for what we call the burnout-workload circle. If physicians, medical students and healthcare staff suffer from dangerously high rates of burnout and stress the personnel suffers from higher leave of absence and health related problems. This in itself might re ect back and increases overall workload of the remaining staff and the negative effect continues. Due to high workload, the health care (diagnostic accuracy, time for organizational tasks, time for patients and patient satisfaction) is decreased.
Therefore scienti c and public attention has been shifting towards resilience to nd ways to avoid burnout and stress before the burnout-workload circle.
Resilience is de ned in a multitude of ways; some researchers think of resilience as a sense of exibility or adaptability (Jackson, Firtko, & Edenborough, 2007). Others de ne resilience as an individual's set of skills and attributes that grant them 'the ability to succeed, live, and develop in a positive way despite stress or adversity' (Cyrulnik, 2009). Because of resilience's dynamic nature, it differs from related traits such as 'mental toughness' or 'hardiness' (Coulter, Mallett, & Gucciardi, 2010).
As research shows, resilience can be learned and nurtured through training and practice (Zautra, Hall, & Murray, 2010). In the realm of healthcare and healthcare providers education, it is especially necessary to obtain and maintain a level of resilience. How can one offer their best services if they are not operating as their best selves? Building a sense of resilience is clearly appealing; higher resilience can lead to a decreased occurrence of burnout (Howe, Smajdor, & Stöckl, 2012). Resilience also helps to circumvent issues such as impaired concentration, increased cynicism, undermined professional development, and jeopardized care for patients (Dunn, Iglewicz, & Moutier, 2008).
In German and international medical education little research on resilience in medical students has been done. We have limited knowledge regarding how students' themselves rate their personal resilience in the end of medical school and what factors relate to resilience in medical school. Further, in order to provide trainings on a broader range we need to identify whether students' resilience can indeed be positively in uenced.
Therefore, in our study, we aimed to identify medical school graduates' resilience status at the end of their medical education and its relationship to other constructs.

Methods
Assessment of the resilience status of medical students A survey in the course of the so called "The survey of Bavarian medical graduates (Bayerische Absolventenstudie Medizin, MediBAS)" is a multi-centric survey comprised of several validated questionnaires (listed below) conducted in ve Bavarian medical schools (Friedrich-Alexander-Universität Erlangen-Nürnberg, Julius-Maximilians-Universität Würzburg, Ludwig-Maximilians-Universität München, Technische Universität München, Universität Regensburg) and one Bavarian veterinary medical school (Ludwig-Maximilians-Universität München). The study was done from October 2018 to January 2019 in collaboration of the Kompetenznetz Medizinlehre Bayern (KMB) and the Bavarian institute for higher education research and higher education planning (Bayerisches Staatsinstitut für Hochschulforschung und Hochschulplanung, IHF). Target persons were those who had nished their nal state license examination between 01. October 2016 and 30. September 2017 at one of the participating faculties. The questionnaire was sent out in an online version (via Questbeck, Globalpark inc.) and a paper-based version. We will present data of the human medical graduates only. 1610 persons were asked to participate, 613 did participate, 60% were female.
For the identi cation of the resilience status we included the 10-Item Connor-Davidson Resilience Scale, German Version (Sarubin et al., 2015) in the survey. The 10 items measure resilience with a 5-point Likert type response scale form 0 (never) to 4 (almost always). The questionnaire represents the following aspects of resilience (Connor & Davidson, 2003 Participants' answers should be given based on re ection from the previous four weeks of their life. The level of agreement within each item allows for the calculation of a score for each participant. Possible overall score ranges vary from a minimum of 0 (lowest level of resilience) to a maximum of 40 (highest level of resilience). We used the German version validated by Sarubin et al. (2015).
To get a more holistic view of what participants' resilience level entails we also included other measures. To self assess professional competence we included the professional competence scale (13 items) of the Freiburger competence questionnaire with overall 45 items (Giesler, Forster, Biller, & Fabry, 2011). To measure the scienti c and communicative competence we included two self-developed scales, whereof scienti c competence is based on the learning goals of the national competency based learning objective catalogue medicine (Nationaler Kompetenzbasierter Lernzielkatalog Medizin; wwww.nklm.de). We further asked questions regarding their demography, income and working time, discrimination and mobbing experiences as well as their perception of how they deal with excessive demands, how they handle feedback and workplace learning. The latter two measures are adapted from a questionnaire for social workers (Kyndt, Govaerts, Verbeek, & Dochy, 2013).

Resilience of medical students and its relationship to other constructs
The Connor-Davidson resilience scale resulted in a mean resilience score of M = 37.1 (SD = 0,63). Resilience was measured highly reliable (Cronbachs α = .88). We have depicted the individual scores in table 1. Overall score sum 37.10 6.3 The score of the last item is interesting as many of the participants did choose not to answer. In general the scores seem relatively high compared to other nonclinical samples (Sarubin et al., 2015).
To further advance the knowledge on the resilience status we calculated the relationship of the scale to other measurements of the survey with Pearson correlations. Resilience status did positively relate to estimated self-assessment of communication competence (12 items, correlation of r = .18; p<.05) and professional competence (13 items, correlation of r = .19; p<.05) and scienti c competence (10 items, correlation r = .19; p <.05). Interestingly, a signi cant negative relation between resiliency and the question how satis ed participants were with their rst job (r = -.18; p < .01) was found. Conversely, participants' satisfaction with their medical education did relate to their resiliency scores positively (r = .12, p < .01). There was a negative signi cant correlation regarding participants' subjective excessive demands (How does the following attribute apply to you: professionally overwhelmed; r = -.20; p < .01; workload r = -.10; p < .05; too many stand-by duties r = -.09; p < .05). Further, we found that participants scores' regarding discrimination and mobbing experiences regarding their ethnicity was signi cantly related to resilience scores (r = .08, p < .05) Interestingly, the resiliency score did not relate signi cantly to scores of professional workplace learning (r =.08, n.s.) and feedback (r = .01, n.s.). The resiliency score did not relate signi cantly to participants' income (r = .04, n.s.), nor their contractual (r = .04, n.s.) or real working time (r = .05, n.s.). The results are summarized in table 2.

Discussion
The study is a primer on resilience after medical school with a large study population from ve medical schools. The resilience status of medical students after their graduation has been investigated. Our results indicate that medical graduates in their rst year after graduation do estimate their resilience to be rather high. Further, the scores indicate that participants' resilience scores vary. Especially, their estimate regarding discouragement by failure is rather low and over 150 of the participants who answered all other questions chose not to answer whether they are able to handle unpleasant feelings. Medicine does entail negative feelings for physicians and patients (Ofri, 2013) and it is one of the core components of a resilient healthcare workforce to deal with failure (Wears & Wu, 2002). As medical education we need to address this issue in the realm of a broader resilience program. We will now discuss the further ndings before we sketch a plan which parts of a resilience curriculum we deem important.
The resilience status we assessed did provide some insight regarding the relationship of resilience in a broader context. As could be expected, a higher level of professional, communicative and scienti c competences relates to higher levels of resilience and vice versa. Medical education needs to assess uncertainty and address uncertainty-tolerance in order to counteract these tendencies (Luther & Crandell, 2011, Cooke, Doust & Steele, 2013. Further, participants who felt their job demand to be more excessive did have lower resiliency scores or vice versa less resilient participants felt the job demand to be higher. Interestingly, no relation to the actual and contractual weekly working hours was found. This provides evidence that indeed it is not the workload that is important, but the way how physicians cope with the workload (Portoghese et al., 2014).
The negative relationship of job satisfaction and resilience can point towards the idea that resilience is built from a necessity to develop it, same goes for the negative relationship to discrimination against ethnicity. Of course, this is cause for concern for medical schools and medical educators should strive towards a more common and standardized implementation of resilience in their medical school and not wait for students or physicians to have negative experiences in order to develop resilience. As there was no relationship to feedback at the workplace, workplace-based learning and the resilience score the continuing medical education put in place does not seem to su ce to advance physicians' resilience.
Specialized resiliency education might help to build resiliency in undergraduate and continuing medical education (Tempski, Martins, & Paro, 2012).
We will now sketch resilience training aspects drawn from psychotherapy education, which provides a threefold proposal for struggling with one´s own emotions while dealing with patients: 1) Self-worth training, 2) normalizing emotions, 3) dealing with strong emotions. 1) Self-worth/self-care training: To deal with unpleasant feelings self-worth trainings incorporate practices to learn to focus, instead of avoiding, unpleasant feelings (Nadjiri, 2018). Paradoxically focusing on these feelings make them disappear. A recent study found that incorporating practices of self-care signi cantly reduces feelings of exhaustion and disengagement (Rich et al., 2020).
2) Normalizing emotions in medical education: The Dutch psychologist Kristin Neff proposed to embrace those parts that are most hurt (Neff, 2011). For example: mindfulness can provide a tool embracing emotions without judgment and altering the cognitive and emotional experience (Orzech et al., 2011). Further, there are preliminary training programs that incorporate a group activity of emotional speeddating to normalize emotions during medical practice (Kiesewetter & Dimke, 2020).
3) Dealing with strong emotions: Strong emotions are the ones that come rapidly in situations out of our control. An example can be the anger a physician experiences after a patient is frustrated with the physician because they disagree on the treatment plan. Physicians should be educated to seek external help to deal with these strong emotions -in form of Balint-group intervention, supervision or therapy. In psychiatry or psychosomatics this has been standard practice for a long time. Yet, to counteract burnout in the healthcare workforce, it is of major importance to teach that in order to be medical professionals physicians who strictly work somatically also need emotional after-care (Meier, Back & Morrison, 2001). In our manualized training program we focus on the basics of energy management, but go beyond these to provide students with a sense of how to deal with emotions and re ect upon their achievement emotions (Kiesewetter & Dimke, 2018, Kiesewetter, in press). More research is needed to directly assess coping mechanisms in medical education and short-and long-term effects of resilience training.
With a growing need and possibility to train medical students' resilience there is also a growing need for providing educators with the tools how to facilitate resilience trainings. We know little of resilience of staff in higher education and even less regarding their ability to teach courses with content derived from cognitive behavioral group therapy. There is a need to provide faculty with teaching skills to manage and deal with emotionally activated groups, despite being non-clinicians. Emotionally activated groups are those courses explicitly designed to trigger emotional responses necessary to re ect and change one`s behavior. When teaching these courses teachers need to be able to deal with these emotional responses and know if and when to offer further help or guidance where to seek help. Knowledge to other (medical) specialist educators needed could be provided in train the trainer workshops.
Not only in medical education, but also in other elds of higher education there is a growing need to embed resilience trainings. Many of the parts of the trainings can easily be adapted to other specializations. Especially teacher education, social workers, kindergarten teachers or other specializations in the human services industry known to be prone to burnout (Burke & Greenglass, 1988;Chang, 2009) could bene t from training strategies developed in medical education. Medical education could take the lead in resilience training through setting an example of courageously addressing a major inner-professional burnout crisis.

Limitations
Regarding the resilience status of medical graduates, we utilized a sample from graduates from Bavarian medical schools and we can't guarantee for their representativeness, neither for medical students nor for medical graduates worldwide. However, with a large response rate of over 600 participants it is the largest sample for Germany and the largest sample for assessing resilience after medical school worldwide that we know of. We have provided correlations for resilience with various other measurements from the survey and of course correlation does not answer questions regarding causation. Nonetheless, we are only beginning to understand resilience in medical school and the resilience status presented here has provided us with an idea where to look at in further research. More research is needed in order to investigate which coping mechanisms graduates and medical students have in order to create evidence regarding the external validity of their estimate.

Conclusions
For medicine, by gaining increased insight into resilience status, we need to take the advancement of individualized emotional coping mechanisms of medical students and healthcare professionals seriously (Howe Smajdor & Stöckl, 2012). With one-year graduates having medium to high resilience scores, but