Methods
Study design and data collection
Online survey among female doctors
Our online quantitative survey took place between May 9 and July 15, 2013 and focused on physicians and dentists who worked in Hungary. A link to an anonymous online self-administered questionnaire was sent to all potential participants, followed by four reminder e-mails (sent every two weeks on average). Response rate increased continuously: 4000 physicians and dentists responded in the first fortnight of the research period, and a further 2607 responders were inspired by the reminder letters.
Among the population, we considered all those who were registered members of the Hungarian Medical Chamber and had a valid e-mail address (N = 42.342).
Response rate was 16.18% (n = 5.607), which is considered acceptable in comparison with the average response rate of online surveys [
23].
Our data were weighted by gender, age and profession (physicians, dentists). After three-dimensional weighting, the data with regard to the region and type of workplace were compared with the same type of data from the Hungarian Central Statistical Offices. We considered our survey to be representative, because we did not find significant difference according to regions (by counties), nor according to type of workplace (general practice, inpatient and outpatient care).
The survey was performed with the permission of the Hungarian Medical Chamber. Ethical permission was obtained from the Ethical Committee of Semmelweis University, Budapest (No: 60/2013).
The cross-sectional representative study
The Hungarostudy 2013 survey, a large cross-sectional representative study, was conducted among the Hungarian adult population.
A clustered, stratified sampling procedure was used. From the 150 sub-regions of Hungary, all settlements with more than 10,000 inhabitants, as well as representative samples of smaller settlements, were included. Samples of 2,000 individuals each, which were representative according to gender, age and the sub-regions for the adult population of Hungary, were selected from the national Population Register. Structured interviews concerning demographic data, physical and mental health characteristics, psychosocial risk and protective factors and participants’ attitudes towards the health system were conducted [
24].
Measurement of reproductive health variables in both samples
The following five reproductive health factors were assessed:
(1)
Participants were asked whether they had had any miscarriages in their lives (Yes/No).
(2)
Participants were asked whether time-to-pregnancy had been longer than one year in case of any of their pregnancies.
(3)
Participants were asked whether any of their pregnancies were high risk (Yes/No).
(4)
Participants were asked whether they had ever been treated for infertility (Yes/No).
(5)
Participants were asked about how many terminations of pregnancy they had had in their lives (To be answered with a number).
In case of miscarriages, high risk pregnancies and terminations of pregnancy, the year of the last event was also asked.
Ascertainment of covariates in both samples
The following age groups were applied: 24 to 35, 36 to 45, 46 to 55, 56 to 65 years, and older than 65 years. Marital status was coded into a two-category variable as ‘living with partner’ (married or cohabitating) or ‘single’ (single, living separately from partner, divorced, or widow) to increase statistical power. Answers to the item regarding the number of children were dichotomised into having or not having children, to obtain a proxy measure for parity. The variables of current workplace were categorised into four groups: working in inpatient care or outpatient care, working as a general practitioner or other (non-governmental organisation, private practice, etc.)
In the multi-adjusted binary regression analysis, we used the following variables: tobacco habits were categorised as ‘never smoked’ or ‘current’/‘former’ smoker. Self-reported height and weight were recorded, and body-mass index (BMI) was calculated. A question with regard to any ongoing gynaecological treatment and any induced abortions in history was also included. The variable regarding shift work was dichotomised.
Assessment of burnout
Burnout was measured by the Hungarian version of Maslach Burnout Inventory-Human Services Survey (MBI) [
19,
25]. The 22-item questionnaire is aimed to measure each of the three components of burnout (emotional exhaustion or depersonalisation, and lack of personal accomplishment) by three different subscales. Responses are rated on a 7-point Likert scale (0 meaning ‘never’ and 6 meaning ‘every day’).
The 22 items put up three subscales:
Emotional exhaustion subscale (EE) consists of nine questions and assesses the feelings of being overextended and exhausted in one’s workplace (e.g., “I feel depressed at work”).
Depersonalisation (DP) subscale has five questions and measures an unfeeling and impersonal response toward recipients of one’s service, care treatment, or instruction (e.g., “I don't really care what happens to some recipients”).
Personal accomplishment (PA) subscale consists of eight questions assessing feelings of competence and success in one’s work (e.g., “I have accomplished many worthwhile things in this job”) [
19].
Using cut-off scores for the means based on data from a normative sample of 1104 health professionals in the United States, respondents are classified as high, moderate or low burnout cases on the respective subscales [
19]. Among physicians, the cut-off points of high scores are as follows: EE: >27 score; DP: >10 score; and PA: <33 score. In our study, the Cronbach’s alpha coefficients for the subscales proved to be 0.909 for EE, 0.767 for DP and 0.818 for PA.
Statistical analyses
Descriptive statistics were used to determine the prevalence of reproductive disorders among Hungarian female physicians. Characteristics of physicians and the control group were compared by χ2 tests. Correlations between certain factors of reproductive health and the three dimensions of burnout were detected by Pearson correlations and X
2 test. All tests were 2-sided, with a type I error level of 0.05.
Unadjusted and multi-adjusted binary logistic models were performed to examine the association between burnout and the five reproductive health factors. In the multi-adjusted binary regression analyses, we adjusted for age, parity, tobacco habits, BMI, gynecological treatment, induced abortion, and shift work in the past few years. A threshold of p <0.05 was established in order to consider an association significant. Statistical analyses were conducted using the SPSS 15.0 for Windows.
Ethical approval
The study was approved by the Ethics Comittee of the Semmelweis University, Budapest.(ref. number: 60/2013).
Discussion
The reproductive health of Hungarian female medical doctors was mapped through a representative survey of doctors in 2013. Data were compared with the results of a representative Hungarian sample.
We found that compared with the general female population, a larger proportion of female medical doctors were characterised by time-to-pregnancy intervals longer than one year, and more of them had undergone infertility treatment or miscarriage(s). On the other hand, the same prevalence of terminations of pregnancy was found in both groups. Comparing data of reproductive morbidity of physicians and the professional control group of Hungarostudy 2013, the same trends can be found, with the exception of miscarriages.
Since we compared our data with those of the general female population, the negative trend of our data is more apparent: in general, educational level and health status correlate positively; therefore, one might expect highly educated women as less likely to suffer from reproductive problems [
26]. We also assume that physicians are more familiar with methods of prevention and preservation of health, with options of early recognition and treatment of pathological conditions.
According to our original hypothesis, reproductive disorders and burnout, which is a crucial symptom of work stress, correlate with each other. Therefore, the first step of our study was to examine burnout among female physicians. Moderate and severe emotional burnout was found in cases of more than half of our respondents (51.4%). More than one-third (37.7%) of the participants reached moderate to high scores on the depersonalisation scale. Moderate or low levels of personal accomplishment were found in 66.3% of the cases. When examining socio-demographic characteristics of burnout, we found that high scores reached on all three subscales of burnout were significantly more frequent in cases of the youngest group of female medical doctors (24 to 35 years), in cases of childless women and among those who worked in in-patient care.
The next step was to explore the potential association between reproductive disorders and burnout. We applied methodological restriction to be able to precisely analyse the coincidence of reproductive disorders and burnout: only reproductive disorders (high-risk pregnancies and miscarriages) of the last five years were taken into account. It was found that high-risk pregnancies and miscarriages were in correlation with both depersonalisation and low personal accomplishment. Results of the multivariate analysis showed that, beside traditional risk factors, depersonalisation acted as an important explanatory factor in case of high-risk pregnancies. While data from the literature confirmed [
27,
28], that of the three dimensions of burnout, the first one to show up was emotional burnout, our results showed that it only depersonalisation and low personal accomplishment were associated with high-risk pregnancies and miscarriages.
We found rather conflicting international results with regard to reproductive disorders of female medical doctors. On the other hand, according to our previous results, Hungarian female physicians seemed to be overrepresented compared with the general population, with respect to reproductive disorders [
14]. We might assume the effect of more than just one factor when studying the background of reproductive disorders. Physical and emotional workload placed considerably heavy stress on young female doctors, who lack enough experience [
28,
29]. The question of simultaneous harmonisation might also be an important issue. Furthermore, the harmonisation of different fields of life might be another difficulty: starting their professional lives and establishing a family simultaneously often leads to irresolvable conflicts [
30,
31]. It is one of the Hungarian peculiarities that young female physicians are left almost unaided with the difficulties of starting their professional life. The effective mentor system, featured in international publications, is almost entirely missing in Hungary [
32]. Another characteristic is the entire lack of part-time jobs for entrants in health care [
14]. Furthermore, according to the results of a previous Hungarian survey, female medical students bear more altruistic, family-centred, idealistic ideas and show earlier commitment during their career socialisation process, compared with their male colleagues and female law students [
33]. Another difference was highlighted by foreign studies reporting about the importance of ‘controllable lifestyle’: the role of the amount, characteristics and controllability of working hours is getting more and more relevant when launching one’s career [
34]. On the contrary, Hungarian female doctors are basically driven by altruistic motivations, in spite of all the difficulties of their would-be profession, when choosing their medical profession and specialty [
35].
Numerous surveys reported that changing shifts and the realignment of the circadian rhythm led to somatic disorders. It is possible that the heavy work load related to working in shifts, especially night shifts, has a significant effect on female reproductive functioning [
36,
37]. International publications also emphasised that it was the younger generation of entrants (residents) who were most affected by burnout syndrome. [
38].The first years of their professional life were characterised by enormous physical and emotional strain and very slight control over their work. Both factors have the potential to contribute significantly [
39,
40].
Our survey highlights several new aspects contributing to research on burnout. While international surveys call attention to emotional exhaustion playing the key role, this dimension of burnout did not correlate with reproductive disorders in the case of our sample of female physicians. On the other hand, reduced personal accomplishment seems to be a more important factor, compared with the international trend. [
41,
42] Since the importance of this dimension of burnout was emphasised by other Hungarians previously, it is possible that in the background there are special circumstances related to the local society and health politics [
43]. According to certain domestic studies, the personal accomplishment dimension of burnout can be detected independently of the other two dimensions of burnout [
44]. Nevertheless, background factors of this issue need to be studied further.
Burnout is strongly associated with heavy workload, as well as physical and emotional strain. It is also associated with stressful situations and one’s ability to cope with them [
45]. The key factors in this process are emotional exhaustion, chronic fatigue and the depletion of cognitive stores. In our analysis, we handled the three dimensions of burnout as a certain indicator of complex work stress. We examined the potential association between burnout and reproductive disorders on the basis of this theoretical framework. Numerous studies revealed the link between burnout and physical and mental health, and many studies aimed to unveil the biological mechanisms in the background of burnout [
45‐
50]. On the other hand, the correlation between burnout and reproductive health has only been examined in male samples; therefore, its analysis among women might supply a deficiency [
51]. Based on the abovementioned results, we can assume that the effect of burnout on somatic and mental health extends to the field of reproductive health. However, the association between burnout and reproductive health can be explained by an approach in evolutionary psychology, namely when there are inadequate environmental conditions (such as more work stress and a greater workload), the adaptive response can be the postponement of pregnancy or reproductivity [
52].
The most important factors that potentially jeopardise the reproductive health of female doctors are well documented in the literature: physical and chemical risks such as anaesthetic gas, X-ray, working in changing shifts and other psycho-social stress factors, such as long and unforeseeable working hours, and the lack of control over work or collegial support. A further important task is to create a balance between work and family, which can act as a significant source of stress among female physicians [
53‐
55]. All the above mentioned factors mutually interact: work-related stress factors can directly and indirectly (by mediating burnout) influence reproductive health. On the other hand, incidental reproductive failure might result in compensatory hyperactivity in the workplace. This escape into work can lead to burnout in a one-track, uncompensated and abusive way.
One of the most important questions of our study pertains to the mechanisms in the background of the interaction between reproductive disorders and burnout. It is possible that burnout acts as a mediator in the development of reproductive disorders: we assume that in the background of burnout, the potential stress factors of the female medical profession can be found, such as emotional burden, urge to perform, and being pushed for time [
56,
57]. It must be emphasised that it is not easy to differentiate between cause and effect in cases of burnout and its correlates. It is possible that there is a circulatory causality between burnout and the development of reproductive disorders. Burnout is an important risk factor of high-risk pregnancies and miscarriages, and it has a negative effect on the outcome of pregnancies. At the same time, reproductive disorders are more likely to cause burnout. This latter assumption was supported by the finding that there were more childless women who suffered from burnout in our survey. The balance between work and family plays a crucial role in the physical and mental health of female doctors, according to the international literature [
58,
59]. These publications emphasise that reaching equilibrium is only possible when resources from different fields of life are mutually converted. According to our results, it is possible that healthy pregnancies and babies who are born play a substantial role in preserving the health of female physicians. Based on our survey, it seems likely that burnout is not a cause but rather a consequence of reproductive morbidity. Female doctors who face difficulties in the field of reproduction are potentially less successful in coping with burnout.
This study’s aim was to give a comprehensive overview of the association between Hungarian female physicians’ reproductive health and burnout. One of the strengths of our survey is that it was possible to compare our representative sample of medical doctors with a representative sample of the general population. Up to our present knowledge, there have not been any studies that surveyed burnout and reproductive disorders simultaneously, or compared the results with the data of a population-based survey. We used validated international tools to describe the dimensions of burnout. However, one of the limitations of our analysis was that although the association between the different variables was well-described, the cross-sectional character of the study prevented us from determining the causality precisely. Furthermore, it was not easy to compare our results with those of certain other international studies, since many of them used different methodology and tools. A further limitation of our study is that the group of female dentists was not examined separately. This was due to the low number of cases with reproductive disorders. Further research needs to explore the potential association between burnout and reproductive disorders among dentists.
Another limitation of our study was that a precise medical definition of high-risk pregnancy was not used; rather, it was based on self-report. At the same time, our survey raised very important questions in the field of physical and mental health of female physicians. Our results further stressed the crucial role of prevention: it is essential to help these women when beginning their careers and when establishing families by enhancing the mentor system, stress coping strategies, flexible working hours, etc. However, female doctors’ wellbeing is not the only field where prevention plays an essential role. Numerous publications affirm that a healer with burnout is not effective in his/her work. [
60‐
62]. Psychic distress can lead to complications and malpractice, and, from a broader perspective, burnout in doctors can have a significant effect on the morbidity of the entire society. Therefore, physicians’ physical and mental conditions are important indicators of the functioning and effectiveness of the health care system [
63,
64].
Competing interests
All authors have completed the Unified Competing Interest form at
http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.
Authors’ contributions
ZSGY participated in data analysis and interpretation, wrote the manuscript, read, and approved the final version. DD participated in data interpretation, read, and approved the final version of the manuscript. EG participated in planning and designing of the study and approved the final version of the manuscript.