Introduction

Physicians are frequently exposed to adverse events, such as medical errors, patient safety incidents, violence, and complaints. Adverse events may lead to mental health problems including posttraumatic stress disorder (PTSD) [1, 2]. Apart from this personal impact, diminished occupational well-being among physicians is linked to decreased professionalism, more medical errors and poorer patient outcomes [3,4,5,6]. The Canadian Medical Education Directives for Specialists 2015 (CanMEDS 2015) specify that physicians should take responsibility for their own health and well-being and that of their colleagues in order to provide optimal patient care [7]. Therefore, it is important to identify which events are risk factors for physicians to develop mental health problems. Furthermore, it is important to know what kind of support is needed to cope with these events.

The 1-year prevalence of mood disorders (major depression, bipolar disorder, and dysthymia) among Dutch citizens with a high income is 3.0%, and the 1-year prevalence of anxiety disorders is 6.0% [8]. In physicians, the prevalence of both depression and anxiety disorders can be as high as 29% and 24%, respectively [9, 10]. Also, physicians have an elevated risk of developing PTSD as they are more frequently exposed to adverse events during their career [11]. Pediatricians experience repetitive stress when dealing with sick children and their emotional and desperate parents, which may be extra stressful compared to other specialties [12]. Physicians’ way of coping and their personal resilience might be important in preventing depression, anxiety, and PTSD. Therefore, the aim of this study was to examine work-related stressors and associated mental health problems in pediatricians as well as their ways of coping and received emotional support in their institute.

Materials and methods

A cross-sectional questionnaire was conducted with members of the Pediatric Association of The Netherlands (NVK).

Among the members of the NVK are residents, attending, non-practicing, and retired pediatricians. At the time of the questionnaire, there were 2160 members in total. All members received an invitation from the NVK for an online questionnaire and three reminders over the course of a 3-month period from October 2016. The questionnaire was sent through SurveyMonkey® using an anonymous (non-traceable) link.

The questionnaire consisted of 56 questions and contained two validated screening instruments, the Trauma Screening Questionnaire (TSQ) and the Hospital Anxiety and Depression Scale (HADS, Appendix) [13,14,15,16]. The first draft of the questionnaire was reviewed by members of the Childbirth and Psychotrauma Research (CAPTURE) group of the hospital OLVG in Amsterdam, The Netherlands, as well as by MdV. Furthermore, this questionnaire was conducted with gynecologists in 2014 and with orthopedic surgeons in 2016 (accepted for publication) [17]. The questionnaires were kept very similar to make it possible to compare the different specialties.

The TSQ is a ten-item screening instrument based on items from the PTSD Symptom Scale-Self Report and has five items concerning re-experiencing and five items concerning arousal. Answers can be “yes” or “no” [18]. The cut-off score indicative of PTSD symptoms is six [14, 15]. Only respondents who answered yes to experiencing a traumatic event at least 4 weeks ago filled out the TSQ. Within 4 weeks after an adverse event, an acute stress reaction might trigger complaints comparable to those found with PTSD. However, if the complaints exist for more than 4 weeks, the complaints are more likely to be due to PTSD.

The HADS is a 14-item screening instrument for depression and anxiety, where both subscales contain seven questions each. Each question contains an answer consisting of a four-point Likert scale. The cut-off value of the Dutch version for clinically relevant depressive symptoms (HADS-D) or anxiety symptoms (HADS-A) is equal to or bigger than eight. The total HADS cut-off value for psychological distress is equal to or bigger than 12.

Further questions were added to the questionnaire to explore demographics (age, employment, working experience, sub-specialism), assess adverse events at work, way of coping with adverse events and how pediatricians learned to cope, satisfaction with current support, and if and how the current support system should change (Appendix).

Statistical analysis was performed with IBM Statistical Package for the Social Sciences (SPSS, version 22). Only completed questionnaires were analyzed. Open questions were categorized by two independent contributors (GY and EJ), the overall inter-rater agreement was calculated with Cohen’s kappa. Demographic data and multiple-choice questions were analyzed using descriptive statistics, and P values were calculated with binomial tests. Differences in categorical outcomes between residents and attending physicians were tested with either chi-square tests (χ2) or Fisher’s exact. In continuous data, independent t or Mann-Whitney U tests were used. A two-sided P value of 0.05 or smaller was considered statistically significant.

This study was exempted from ethical approval by the Medical Research Ethics Committees United (MEC-U) and registered under the number W18.096.

Results

Population characteristics

A total of 456 questionnaires (21.1%) were completed, of which 410 questionnaires (18.9%) were eligible. Figure 1 shows the inclusion diagram.

Fig. 1
figure 1

Inclusion diagram of selecting the questionnaires

Full size image

Table 1 shows the baseline characteristics of the respondents, compared with the members of the NVK, the reference group. Gender and the amount of residents were comparable between respondents and the reference group; however, there were more attending pediatricians, less retired pediatricians, and less pediatricians with a management function in our sample (Table 1). Table 2 shows the baseline characteristics divided in the subgroups resident, attending, retired, and non-practicing.

Table 1 Baseline characteristics respondents and NVK
Full size table
Table 2 Baseline characteristics in subgroups
Full size table

Selected quotes of the respondents are added in Table 3 to visualize the events they experience as adverse, divided in aggression by parents and death of a patient.

Table 3 Quotes of pediatricians*
Full size table

Work-related stressors

The following events were experienced as high emotional impact stressors at work by the respondents (multiple answers possible): missing a diagnosis (71.2%, n = 292), suspicion of child abuse (49.3%, n = 202), doubts about making the right decision (48.3%, n = 198), death of a patient (38.0%, n = 156), and critically ill children (26.3%, n = 108; Fig. 2). Almost 80% (n = 325) of the respondents indicated that they actually perceived an event as an adverse event, of which 277 described this event. Aggressive behavior of parents towards the physician was most commonly named as an adverse event (42.5%, n = 118).

Fig. 2
figure 2

Top 10 events pediatricians describe as an adverse event (n = 410), multiple answers possible

Full size image

Posttraumatic stress disorder

Table 4 shows the outcomes of the TSQ. Among the respondents, 79.3% experienced an adverse event at work, of which 34.9% (n = 143) reported having experienced this event during their work more than 4 weeks ago.

Table 4 HADS and TSQ scores
Full size table

The mean score of the TSQ was significantly lower in the group of participants where a peer support protocol was present for adverse events (0.77 ± 1.06) compared to the group where no protocol was present or where it was not used (1.62 ± 2.10, P = 0.02).

Depression and anxiety

Outcomes of the HADS are shown in Table 4. Attending pediatricians have significantly more depressive symptoms according to the HADS-D compared to residents (P = 0.03).

Coping

The most common coping strategies after adverse events were (multiple answers possible) seeking support from colleagues (86.1%, n = 353), seeking support from friends and family (73.2%, n = 300), seeking some other form of distraction (32.7%, n = 134), or doing sports (22.4%, n = 92). Respondents learned their coping strategies (multiple answers possible) during residency (58.3%, n = 239), as an attending (55.1%, n = 226), during clerkships (20.2%, n = 83), and 21.0% (n = 86) reported to having never learnt to cope with adverse events.

Of the respondents, 41.0% (n = 168) has seriously considered quitting their job at some point in their career. Most common reasons for this were (multiple answers possible) disbalance between work and private life (75.0%, n = 126), high workload (68.5%, n = 115), disutility (working outside of working hours; 44.6%, n = 75), too much stress (38.7%, n = 65), and too much responsibility (37.5%, n = 63; Fig. 3). Furthermore, 6.6% (n = 27) considered quitting because of a complaint to the disciplinary board.

Fig. 3
figure 3

Top 10 possible reasons pediatricians describe to stop working (n = 168). Multiple answers possible

Full size image

Six percent (n = 26) of the respondents admitted that they were dealing with adverse events by drinking more alcohol and 1.2% (n = 5) by taking new medication. Professional psychological help was sought by 9.8% (n = 44) and 16.1% (n = 66) stifled emotions.

After being exposed to a work-related adverse event, 18.5% (n = 76) of the respondents adjusted their work. Most common ways to do this were (multiple answers possible) performing more diagnostic tests (51.9%, n = 40), calling a colleague earlier (36.4%, n = 28), work less (33.8%, n = 26), and starting treatment faster (20.8%, n = 16). Over time, 40.5% (n = 166) of the respondents reported to have become more defensive.

Support

Of the respondents, 26.3% (n = 108) indicated that there was a protocol for support in the case of an adverse event in their current working environment, 34.2% indicated that there was no protocol, and the remainder did not know whether there was a protocol. Furthermore, 50.2% (n = 206) thought that a culture change is necessary concerning coping with adverse events. When asked what the standardized support system involved (multiple answers possible), respondents indicated that discussing the situation with the present team (71.2%, n = 292) and self-initiated peer support with direct colleagues (64.9%, n = 266) were used. Of the respondents, 16.3% (n = 67) indicated that there was no support system at all and 41.5% (n = 170) confided that there is not enough opportunity to discuss adverse events and express emotions.

According to the respondents, the preferred form of support would be (multiple answers possible) discussing the situation with the present team (76.3%, n = 313), peer support from direct colleagues (72.0%, n = 295), and professionally organized peer support (43.2%, n = 178). Even though 28.5% (n = 117) would prefer to get help from a psychologist or coach, only 9.8% (n = 40) actually sought out this kind of help.

Discussion

The aim of this study was to examine work-related stressors and associated mental health problems in pediatricians as well as their ways of coping and received emotional support in their institute. First of all, among pediatricians, work-related stressors during their career were high. Suspicion of child abuse and critically ill children are two topics that distinguish pediatricians from other specialties and why this specific specialty can have high emotional burden. Notable is the high prevalence of aggressive behavior towards pediatricians, as stated in the quotes. Therefore, we think that it is necessary to not only develop a better support system after an adverse event but to also implement ways to teach pediatricians to cope with aggression. For example, certain training programs (conflict management and de-escalation (CMD)) focus on how to cope with aggression [19, 20]. Another stressor may be a complaint to the disciplinary board. This has a high impact on psychological well-being and is associated with defensive practice [21]. More than half of the pediatricians who received a complaint at the disciplinary board seriously considered quitting their job. More work-years corresponded with a higher chance to receive a complaint. Whether the non-practicing group stopped working because of complaints at the disciplinary board cannot be answered because of the small numbers. The amount pediatricians who received a complaint to the disciplinary board is low compared to gynecologists [17].

The point prevalence of PTSD in The Netherlands is 1.3%, making that the point prevalence of work-related PTSD is expected to be even lower [8]. In our questionnaire, when pediatricians experienced a traumatic event, we found a high point prevalence of symptoms indicative of work-related PTSD (2.2%). However, compared to the study of Ruitenburg et al., who found a prevalence of PTSD complaints of 15% in hospital physicians, the percentage we found seems low. Nonetheless, the study of Ruitenburg et al. used a much lower threshold than is normally used to assess PTSD complaints and does not use Criterion A [10]. Furthermore, we found more respondents with depressive and anxiety symptoms as compared to 1-year prevalences found in the general Dutch high-income population [8].

After experiencing an adverse event, a little over a quarter of all respondents indicated that there was a protocol regarding adverse events, but over half of the respondents do think a culture change is needed. Therefore, pediatricians have the need for a better support system. When more than half of the pediatricians perceive that care surrounding adverse events is not sufficient, this could lead to unnecessary stress. Physicians might experience more barriers than non-physicians to seek out professional help for mental health problems due to their fear of losing their license, denial of problems, and embarrassment [22]. When physicians are unfit, this may have a negative impact on their practice [23], whereas occupational well-being can positively contribute to patient satisfaction and the quality of interpersonal aspects of care [6]. In this questionnaire, coping strategies applied by the pediatricians were similar to coping strategies of gynecologists in the study by Baas et al. [17]. Almost 20% of the respondents adjusted their job and 40% seriously considered quitting their job completely. This is consistent with the findings of Hawkins et al. [24], who also found that physicians reduce work hours, retire, or quit medicine altogether because of a high work strain. Reported alcohol use to cope with adverse events seems to be quite low in our respondents with only 6%, especially since Hyman et al. found percentages of 6% (daily use) up to 25% (occasional use) of substances [25].

Thus far, little research has been done on this topic, specifically concerning mental health problems in relation to institutional or peer support. This study, which included validated questionnaires, allowed for detailed data collection. Allowing respondents to fill out examples of their experiences gives further insight in the way they experience their problems.

Limitations of the study are the response rate, with 18.9% lower than the average e-mail response rate of 25–30% [26]. Reasons for the low response rate could not only be the heavy workload of pediatricians but also the fact that the questionnaire was spread through a non-personal email account from the NVK, which people may not always read. With 410 completed questionnaires, however, we have a representative cohort comparable to the reference group. Another limitation is the risk of participation bias, because as with any questionnaire, pediatricians who are involved with this topic are more likely to participate. Concerning depression and anxiety, screening rates are generally an overestimation, especially when using self-report questionnaires. However, these are point prevalences, so when compared to 12-month prevalences, it can be an underestimation. Furthermore, depression and anxiety are not merely work-related.

In conclusion, work-related stressors in Dutch pediatricians are high and can subsequently lead to posttraumatic stress disorder. Parental aggression towards pediatricians seems to be a common problem, something that should be addressed, for example with CMD training programs [19]. In this study, the amount of pediatricians with PTSD complaints was higher compared to prevalences found in the general population and the same applies to depression and anxiety symptoms. There is no national standardized support after adverse events for pediatricians, while other occupations where there is an occupational hazard do have such support (e.g., in the police, military, firefighters). It is advised that evidence-based support (e.g., peer support) after adverse events is protocolled and education on coping strategies is implemented, to improve mental well-being of pediatricians.