Policy and attitude-related reasons for gender disparity in post allocation for graduate medical education in Austria

Abstract

Objective: A previous study found that in Austria 50.3% physicians (m: 43.2%, f: 58.6%) have not attained their chosen specialty. We aimed to explore the policy – and attitude-related reasons for gender disparity in training post allocation.

Methods: This cross-sectional study used a quantitative and a qualitative method. A self-administered 12-item questionnaire was sent twice to all 8127 licensed Viennese physicians. Physicians’ opinions regarding why the chosen medical specialty was not attained were analyzed. To estimate the responder bias respondents from the first and second mailing were compared.

Results: A total of 2736 questionnaires (34%) were returned. When a specialty is favored by men, the chance for women to achieve that specialty decreases. According to the qualitative results, men were more often ready to accept training in a specialty different from the one originally desired. Female physicians were put at a disadvantage by consultants due to organizational considerations and sex-stereotyping.

Conclusions: According to physicians’ self-reported opinions, consultants do not place female candidates at disadvantage as a result of an unconscious process but mainly based on reasoning about organizational aspects and sex-stereotyping. Several explanations for the phenomenon that men are more often ready to accept training in a specialty different from the one originally desired were identified.

Introduction

We have established previously that in Austria the allocation procedures control specialty choice and are, in addition, ineffective, uneconomical and do not ensure equal rights and fair opportunities for male and female applicants (Spiegel et al. 2004). An average of 50.3% of physicians – 43.2% of male physicians but 58.6% of female physicians – did not become specialists in their chosen medical specialty (CMS). For specialists’ attainment of the CMS was found to be a predictor of relative job satisfaction (Spiegel et al. 2008).

In 1976, the European Union issued a directive on the implementation of the principle of equal treatment for men and women with regard to access to employment, vocational training (VT), promotion, and working conditions (The Council of the European Communities 1976). Although, the European Union defined standards for the duration and content of postgraduate medical training (The Council of the European Communities 1993), there exist none for the allocation of training posts ensuring equal rights for men and women in attaining the CMS. In the UK, a first screening of applicants has to be done by reviewing written applications, which should not disclose name, sex, marital status, age, religion, and place of study or ethnic origin (Career Progress of Doctors Committee 2000). However, there are specific needs of female graduates which must be taken into account to ensure equal chances for men and women in attaining their CMS. It is, for instance, important that allocation procedures allow more flexible training and work practices, particularly through the years of child raising (Lawrence et al. 2003).

With regard to the need to choose the right language in discussing equal treatment for men and women, it must be remembered that the terms ‘sex’ and ‘gender’ are not synonymous. As Mayer et al. (2008) put it, biology determines the sex of humans and society determines the gender. They also remind us that socialized differences lead women to a place of greater priority on interpersonal satisfaction and integration than men. Since women are more often motivated by encouragement and men tend to respond more to challenges, Mayer et al. (2008) conclude that socialized gender differences should have a role in mentoring.

Two theoretical frames of reference have been identified to explain gender differences in physicians’ specialty choices (Gjerberg 2002): structural-oriented explanations attempt to explain gender differences by looking at the possibilities and limitations given in a certain context (Crompton & Sanderson 1990; Reskin & Ross 1990), whereas individual-oriented explanations view gender differences in education and career as a result of gender specific interests and values (Polachek 1976) with a basis in socialization (Marini & Brinton 1984).

The aim of this study is to analyze the interpretations of those female and male physicians who have not attained their CMS in respect of whether discrimination of female applicants in the allocation of medical training posts is mainly an unconscious process or based on reasoning and on sex-stereotyping by consultants. It also examines whether female and male applicants are equally ready to change their CMS when given a suitable chance.

Table

Practice points

Female physicians are in many countries disadvantaged in reaching their chosen medical specialty. It has been hypothesized that this is mainly due to an “unconscious process” on the part of consultants. Our study does not support this hypothesis. Most of physicians’ self-reported opinions for not achieving their chosen medical specialty have to do with the Austrian training post allocation procedures, which are based mainly on the subjective preferences and considerations of male and female consultants. According to those physicians who have not attained their chosen medical specialty female physicians were put at a disadvantage by consultants due to a conscious process based on reasoning about organizational aspects and on sex-stereotyping. Although, considerably more men than women achieve their chosen medical specialty men in our population are more often ready to change their chosen medical specialty when given a suitable chance.

Methods

This analytical cross-sectional study, which uses both a quantitative and a qualitative method, re-examines the data of a previous survey on training post allocation in Austria (Spiegel et al. 2004) in order to answer the research question.

Instruments and procedures

Participating physicians were explored with the aid of a self-administered 12-item questionnaire. In the questionnaire, seven questions concerned the course of postgraduate medical education and five were aimed at assessing the demographic data of the respondents. In order to ensure clarity and answerability, the questionnaire was given a field trial before completion of the final version. It consisted of both probe questions and free answer questions. It contained questions relying on the memory and/or honesty of participants; e.g. individuals’ original specialty choice (e.g. ‘Which specialty did you aspire to on finishing medical school?’), questions referring to facts (‘Which specialty are you now practicing?’, ‘Did you train (A) for a certain period or (B) complete training in any other specialty?’, ‘How many months did you spend in a recognized training post over and above the prescribed training period for the specialty currently practiced?’), and free-answer questions asking participants to reflect on their course of VT (‘If you were not able to attain your specialty of choice what, in your opinion, was the reason?’). We have correlated the educational outcomes with demographic data such as sex. For more details about data collection please refer to the parent paper of the present study (Spiegel et al. 2004).

In addition to the quantitative aspect of the survey, we looked at the self-perceived reasons given by the physicians for their not attaining the CMS using a qualitative approach (Hayhow et al. 2006). It is a goal of our qualitative research to present a coherent and illuminating description of physicians’ perception and perspective on their experiences.

Study population

The cohort consisted of all licensed physicians practicing in Vienna (n = 8127; all general practitioners, all specialists except dentists). The city of Vienna is the most populated of Austria's nine provinces and has the highest number of physicians. Twenty-seven percent of the 30,509 Austrian physicians practice there. The questionnaire was mailed to all 8127 physicians together with a letter explaining the purpose of the study and asking them to complete and return the questionnaire by telefax or mail. In order to reach those who had failed to reply the questionnaire was mailed a second time to all physicians, asking only those to complete the questionnaire who had not done so before. To identify those who might have already answered the first mailing, a new question referring to the first mailing was added to the questionnaire of the second mailing. A small sample of the non-respondents was identified by personal contact and interviewed in order to estimate any bias due to a failure to reply to either of the two mailings.

Statistical analysis

The data were analyzed using frequency tables, univariant tables or contingency tables. When appropriate, the Chi-square and Mantel–Haenszel test was used and statistical significance was set at 0.05. To evaluate the data we used a computerized program based on an Oracle databank.

Results

Of the 2736 questionnaires returned by respondents (34%) 166 were excluded from the analysis due to invalid replies or for technical reasons. A total of 1595 (about 20%) questionnaires were validly returned after the first mailing; 722 from female, 836 from male respondents, and 37 did not state their sex. Nine-hundred and seventy-five (about 12%) questionnaires were validly returned after the second mailing; 416 from female, 550 from male respondents, and 9 did not state their sex.

Respondents’ socio-demographic data

In the cohorts examined the ratio of female to male physicians was 1 : 1.48. In relation to the study population, more questionnaires were returned by female than male physicians. The Chi-square test showed no significant difference, however, in the response rate between female and male doctors (p = 0.184).

Whereas 51.6% of physicians between 35 and 55 years of age (66% of the study group at the time of the study) failed to attain their CMS, only 45.6% in the age group ‘older than 55 years’ (25% of the study group) did so. Twenty-five percent of the respondents had no child, 21% had one child, and 54% had more than one child. 9.3% of respondents had not possessed Austrian nationality at their time of birth. The educational level of respondents was defined by their profession and therefore not assessed separately.

We have published the details about the respondents before (Spiegel et al. 2004). In short, there were no significant differences between the data of the first and second mailing except for the answers stating the self-perceived reasons why the CMS was not attained (Chi-square p < 0.0001) and which specialty was being practiced (p = 0.006). Of the 50 non-respondents who were questioned orally, 26 (52%) stated that they were not able to attain training in their specialty of preference. Physicians gave various reasons for not responding to the first and second mailing: 22 were not interested or had no inducement to respond; 17 have had no time to answer the questionnaire; six could not remember having received the questionnaire; five declared that the fax transfer of the questionnaire had failed. The following shows the combined results from the first and second mailing.

Figure 1 illustrates the percentages of female and male doctors who did attain a training post in their CMS. The considerable gender disparity, which exists with regard to achieving the CMS is most impressive for the specialties ‘traumatic surgery’ (41% of male and 11% of female physicians achieved this CMS) and ‘internal medicine’ (56% of male and 28% of female physicians).

Figure 1. This figure shows how many female and male physicians reached their initially CMS depending on the specialty of choice.

Table 1 shows the answers of those physicians who did not achieve their CMS (50.3%) to the question ‘If you did not have the possibility to be trained in the specialty originally desired, what do you think the reason was’. In answering, they could choose between five reasons (one or more reasons could be selected) or state any other reason (free answer question). On average, every person who did not attain his/her CMS gave 1.25 self-perceived reasons.

Table 1.  Physicians’ self-reported reasons for failing attainment of the specialty originally chosen^a

Self-perceived reasons	Female physicians (%)	Male physicians (%)
Was offered a different specialty. I grasped the opportunity	25	45
The waiting period for the preferred specialty was too long or too uncertain	31	26
Was not offered the preferred specialty	18	16
At a disadvantage due to having a child/children	20	3
My contract was not renewed after my GP training period	11	10
Other reasons (free answer questions consolidated)
Changed preference during postgraduate training	6	13
Enjoyed no favoritism	6	8
Was of the wrong sex	3	0
Wanted more time for my family	4	1

^aRespondents were allowed to state more than one reason.

A total of 7.9% of respondents were not of Austrian nationality at the time of birth. Fifty-four percent of these foreign-born physicians achieved their CMS, as compared with 49.7% of all respondents (including foreign-born).

Qualitative data

In addition to selecting one or more of the five possible reasons offered by the questionnaire in answer to the question ‘If you did not have the possibility to be trained in the specialty originally desired, what do you think the reason was?’ (Table 1) physicians could give any other reason (free answer text). In order to present a coherent and illuminating description of physicians’ perception and perspective on their experiences, we are including the examples of their original statements as to why the CMS was not attained, in as far as they are relevant to the research question of this study.

Data sets are presented in the following format: questionnaire (Q no.), sex (f/m), specialty practiced, specialty originally aspired to (desired: ….), children (no child, one child, and more than one) statement (‘text’).

Statements category A: Family orientation of physicians

Q26, f, GP, desired: paediatrics, more than one: ‘Wanted more time for my family and my two children.’; Q494, f, GP, desired: dermatology, more than one: ‘Decided of my own accord for family-friendly activity.’; Q611, f, anaesthetist, desired: traumatic surgery, one child: ‘Renounced desired specialty for my child’s benefit, because at the time I had to do 16 night shifts per month.’; Q808, f, GP, desired: paediatrics, one child: ‘I didn’t accept the training post offered me because my son was eight weeks old at the time.’; Q852, f, specialist for transfusion medicine, desired: anaesthesiology, more than one: ‘didn’t want to do night shifts because of the children.’; Q1158, m, GP, desired: ENT, more than one: ‘My family was more important to me than specialty training (VT for GP was 3 years, for ENT six).’; Q1307, f, GP, desired: internal medicine, one child: ‘After the birth of my child my priorities changed – wanted more time for family!’; Q2134, f, GP, desired: internal medicine, more than one: ‘There are too few part-time specialty training posts, essential for female physicians with children.’.

Statements category B: Related to sex-stereotyping of consultants

Q82, f, orthopaedic surgeon, desired: traumatic surgery, no child: ‘Was of the wrong sex.’; Q199, f, anaesthetist, desired: surgery, more than one child: ‘Because women today still are disadvantaged with regard to the allocation of training posts in surgical specialties.’; Q298, f, paediatrician, desired: GP, more than one: ‘The surgical consultant said: in an emergency women always do things wrongly!’; Q342, f, GP, desired: obstetrics-gyn., one child: ‘Consultants allocate the specialty training post only to men.’; Q790, f, pathologist, desired: forensic medicine, no child: ‘The consultant told me to go home and have children!’; Q1319, f, internist, desired: obstetrics-gyn., one child: ‘sexism!’; Q1506, f, GP, desired: paediatrics, more than one: ‘More than one consultant told me: ‘As long as I’m in charge here there’ll be no training posts for women!’’; Q1704, f, GP, desired: paediatrics, one child: ‘Because I was pregnant I was disadvantaged when I asked to be added to the unofficial waiting-list for specialty training.’; Q1814, f, GP, desired: internal medicine: ‘Because as a woman I could drop out due to pregnancy the post was given to a man.’; Q1959, f, GP, desired: psychiatry: ‘I am female, I belong to no political party and have no connections.’; Q2169, f, GP, desired: dermatology, one child: ‘Was disadvantaged because of my child. But I think that should not be a problem because there are such things as babysitters.’; Q2180, f, GP (occupational medicine), desired: forensic medicine, children not known: ‘A male applicant was given preference because women are not wanted in forensic medicine.’; Q2269, f, GP, desired: internal medicine: ‘Consultant wanted a man for the training post.’

Discussion

Our study shows that according to those physicians who have not attained their CMS female physicians were put at a disadvantage by consultants due to a conscious process based on reasoning about organizational aspects and on sex-stereotyping. The gender disparity – female physicians are disadvantaged in achieving the CMS compared to their male colleagues (f: 41.4% vs m: 56.8%) – can not be explained by different levels of qualification of male and female graduates, because VT is only the first step of specific education in a CMS. We see from our quantitative and qualitative data that there is an overall tendency for consultants to give preference to male trainees in their decisions about post allocation. Since, in Austria, consultants are allowed to choose trainees arbitrarily (Schuetz et al. 1998) from among numerous, more or less equally qualified candidates, training post allocation allows selection based on factors other than suitability for a chosen specialty.

Attitude-related aspects of gender disparity

According to our qualitative data (‘Statements category A’) we conclude that many female physicians in our study who did not achieve their CMS were put at a disadvantage in their VT mainly, not as the result of an unconscious process on the part of male and female consultants, but due to reasoning based on organizational aspects, which is a cognitive process. The question arises whether such reasoning can be the basis of a just allocation system. We do not think it can, because, first, it is in breach of the European Union's Directive on the Implementation of the Principle of Equal Treatment for Men and Women and, second, it is based on prejudices. Our qualitative data also demonstrate that, in addition to the so-called ‘reasonable’ but unfair considerations of consultants, who take into account such issues as pay-off probability of the educational investment in a particular trainee, drop out rates, availability and flexibility of physicians – e.g. when putting together an operation team – and the willingness to accept inconvenient night duties (e.g. questionnaire no. 494, 611, 852), sex-stereotyping against women as doctors in general is still prevalent (‘Statements category B: sex-stereotyping; such sex-stereotyping against women as doctors has been described by various researches) (Kato et al. 2004; Woodrow et al. 2006). Therefore, we think, that both structural-oriented and individual-oriented explanations (Gjerberg 2002) need to be taken into account when trying to understand why female physicians are unfairly put at a disadvantage in reaching their CMS in Austria.

Zuber argues that this discriminatory attitude against women is manifested by consultants of both sexes and that this attitude is, primarily, an unconscious one (Zuber 2001). The opinion that gender stereotypes are held by men and women and that they are ‘largely unconscious’ has as well been expressed in a critical review of gender issues in neurosurgery (Woodrow et al. 2006). However, this position is not supported by our data because, according to physicians’ opinions and experiences, the gender disparity in post allocation in our study is mainly based on reasoning.

Wang-Cheng et al. (1995) found that female students received the highest mean clinical marks in the ambulatory care clerkship when the preceptor was male and male students the lowest when the preceptor was female. One study found that male assessors displayed no gender bias when assessing female medical authors while women assessors upgraded female authors in the particular study (Johansson et al. 2002). Another study, which investigated whether there exists a fundamental gender bias in the evaluation of residency candidates applying in different specialties, found that physicians consistently rated the women candidates more favorably (Smith et al. 1991). None of these studies support the argument (Zuber 2001; Woodrow et al. 2006) that a discriminatory attitude against women is an unconscious one. On the contrary, some data even suggest that male senior physicians tend to display a soft spot for female students/trainees by discriminating in their favor.

Respondents’ answers to the question ‘How many children do you have?’ (no children; one child; more than one) were used to ascertain the number of children according to the specialty practiced (data not shown). Although, these data could be of principle interest due to a methodological shortcoming it was not possible to draw any conclusions as to whether physicians with children had been put at a disadvantage with regard to achieving their CMS. To explore this question the questionnaire would have had to ask about the number of children at the time when physicians were aspiring to their CMS.

Specialty preferences of male and female doctors

There were clear gender differences regarding preferences for certain specialties. The ‘preference list’ shows a tendency on the part of women to choose a primary care specialty (for details, see Spiegel et al. 2004). This tendency has also been described in other studies (Gorenflo et al. 1994; Diehl et al. 2006; Soethout et al. 2008). It is known that a stronger feeling of comfort in caring for patients with psychological problems and a person-oriented approach to problem solving in medicine (Vaglum et al. 1999) is positively correlated with choosing a primary care career. In addition, as has also been described for Norway, general practice possibly offers conditions that make it simpler to combine work and family (Gjerberg 2001). In the Netherlands, Soethout et al. (2008) have shown that students had a higher preference for a career in general practice if one of their parents was a general practitioner or if they were female.

Considering the ratio of GPs to specialists in Austria (0.44 : 1), the number of female and male physicians who wanted to become GPs was low (f: 14%; m: 16%), which is in line with the international trend for specialization (Barondess 2000). As can be seen from our data there is a general trend that, when a specialty is strongly favored by men, the chance for women to achieve that specialty decreases considerably. For example, when the CMS was internal medicine, which ranks first among male doctors, only 28% of the female doctors achieved that specialty, as compared with 56% of male doctors (male : female 2 : 1). Our results correspondent with the ones of Gjeberg who found in her survey on gender differences in final specialization in Norway that a significantly lower proportion of women than men completed specialist training in surgery (Gjerberg 2002). Forty-eight percent of men but 20% of men who had worked in surgical specialty in their first or second job later completed specialist training in a surgical specialty. Similarly, more men than women (31.6% vs 23.8%) who had worked in internal medicine in their first or second job completed specialist training in internal medicine (statistically not significant).

An exception to the general trend, which we have found in our study, that when a specialty is strongly favored by men, the chance for women to achieve that specialty decreases considerably, is general practice, where even more women than men who want to become GPs finally do so. This would confirm our hypothesis that men tend to accept chances offered them more readily than women.

Consequences of these allocation procedures

Most of the reasons stated (physicians’ self-reported reasons) for not achieving the CMS have to do with the Austrian training post allocation procedures, which are based mainly on (male and female) consultants’ subjective preferences and considerations. A female doctor could, for instance, become pregnant and the consultant would then lose an employee or have to assign the post temporarily to a locum not familiar with the department. Female physicians with children would more often be hindered from coming to work than their male colleagues or would have to take child care leave, it being still not the rule for men to share such responsibilities. In this connection we should not forget the twenty-first century trend for women to raise their children on their own – as single mothers (Potee et al. 1999). As Bindal et al. (2007) point out, it is vital that trainees are clear and realistic in their career plans in order to take the right career pathway after the foundation year. Currently, there are plans in Austria to introduce an ‘approbation year’, which should be in between graduation from the medical school and VT of specialist and general practitioners (Fischer 2008). We think that such an approbation year would be advisable, since it would allow informed specialty choice by trainees, to take place within or at the end of the approbation year, and ensure fair post allocation for both male and female applicants. For the latter goal in Austria, a change from hospital to a central training post allocation procedure would be helpful; it would also allow the implementation of objective selection procedures. Austrian policy makers could learn from the positive experiences made in England, where trainee recruitment has been shifted from hospital Trusts to central Deaneries, as for example has been done by the West Midlands Deanery (Bindal et al. 2007).

From the point of view of the consultants, their bias in favor of male doctors may be understandable – decisions being made in the interest of their departments. But it is clearly both ethically and legally unacceptable. But even more than individual consultants the hospital operator, who, in Austria, in the majority of cases is a state organization, is obliged to guarantee equal treatment for men and women (The Council of the European Communities 1976). Ideologies support sex stereotypes, which limit the opportunities of those being stereotyped (van der Reis 2004). In any discussion of this ethical issue the normative framework used (e.g. egalitarianism – justice as fairness; feminist justice – genderless society) must be taken into account, so as to arrive at meaningful conclusions (Mansdotter et al. 2004).

Limitations of this study

Some of the limitations of this survey have already been reported (Spiegel et al. 2004). One limitation of the present study is that it does not examine consultants’ decision-making by exploring consultants themselves. Our conclusions are based on the experiences and the opinions of those physicians who had not reached their CMS. Furthermore, it does not examine the motives behind the choice of specialty, which could be purely professional but could also include such factors as expected income, prestige, interest in primary care or hospital-based practice or other reasons (Rubeck et al. 1995; Wright et al. 1997; Vaglum et al. 1999). Also, we could not objectively ascertain whether the number of children really put female doctors at a disadvantage, since we did not ask about the number of children at the time when physicians tried for training in their desired specialty.

We asked ourselves whether our statistical analysis must be based on the sex ratios of applicants for different specialties, as has been described in the important publication on sex bias in graduate admissions, where examination of aggregate data on different graduate studies at the University of California, Berkeley, showed a clear but misleading pattern of bias against female applicants (Bickel et al. 1975). In contrast to Berkeley, where the number of available places per year for each graduate study was known, in our retrospective study on the Austrian post allocation system it was not possible to determine the number of available training posts per year per specialty. This was because only those posts were relevant which were up for allocation, the number of which changes depending on various conditions, e.g. physicians finishing VT, physicians taking parental leave, variation in number of training posts due to creation or cancellation of same over time, non-allocation of posts, etc. Our study, therefore, examined an open system. The replies to our two mailings were not significantly different as regards the main criteria. The random sample of non-respondents contacted showed no major difference. We, therefore, believe that our results are valid.

Conclusions

There is a more subtle bias against women than merely ‘rank’ (career) and this bias exists already at the initial level of postgraduate medical education and is mainly based on cognitive factors (reasoning of consultants). Female physicians are considerably more disadvantaged in attaining the CMS than their male counterparts. According to those physicians who have not attained their CMS male and female consultants discriminate not unconsciously against female applicants but mainly due to a conscious process based on reasoning. In order to ensure gender balance in achieving desired specialty, objective post allocation criteria for postgraduate medical training must be developed and implemented in Austria. Although the principle of equal treatment for men and women was defined three decades ago (The Council of the European Communities 1976) in Austria male physicians, in general, enjoy an advantage over women in the allocation of specialty training posts – a finding which could well be prevalent in other European health systems.

Acknowledgment

The authors wish to thank Professor Philip Lupton, MA cantab, PhD Vienna, for reviewing the English version of this article.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Additional information

Notes on contributors

Wolfgang Spiegel

WOLFGANG SPIEGEL, MD, is a general practitioner and a senior research fellow at the Department of General Practice, Center for Public Health of the Medical University of Vienna. As a clinical teacher he has a focus on research on medical education and on policy-related research.

Gustav Kamenski

GUSTAV KAMENSKI, MD, is a lecturer at the Department of General Practice, Center for Public Health of the Medical University of Vienna. He has been working as a GP in his single-handed medical office in a rural area in Austria since 1981.

Ingrid Sibitz

INGRID SIBITZ, MD, is a psychiatrist. She has been the leading psychiatrist at the day clinic of the Department of Psychiatry and Psychotherapy, Medical University of Vienna since 2003. She has a focus on research on gender aspects in patients with psychotic disorders and their carers.

Barbara Schneider

BARBARA SCHNEIDER, PhD, is a statistician and an assistant professor of the Institute of Medical Statistics of the Medical University of Vienna.

Manfred Maier

MANFRED MAIER, MD, is a professor of general practice, head of the Department of General Practice and head of the Center for Public Health of the Medical University of Vienna. He has spent a postdoctoral fellowship at Harvard Medical School.