How healthy are Dutch general practitioners? Self-reported (mental) health among Dutch general practitioners

Abstract

Objective: To investigate the level of burnout and health status of male and female Dutch general practitioners (GPs), and to compare this with former samples of GPs and with the Dutch general population. Methods: A postal survey of 350 male and 350 female practising GPs in the Netherlands. Results: Although levels of emotional exhaustion of Dutch GPs were lower than those of national samples of GPs in the 1990s, the prevalence of burnout was still almost twice that of the general population. In contradiction with this, GPs reported better general health and fewer diseases than their fellow countrymen. Another remarkable finding was that female GPs were as healthy as their male colleagues, while in the general population, males report better health than females.

Conclusion: The positive self-reported health status of general practitioners might reflect the high standards of the medical profession, which make physicians reluctant to show their own vulnerability. This might result in fewer, but more serious cases of (mental) illness among GPs as compared to the general population.

• Burnout
• gender differences
• general practitioners
• health status
• the Netherlands

Introduction

Generally, physicians are regarded as caretakers of the health of others, their patients. The fact that the doctor her/himself can suffer from illness is hardly considered in the medical literature, or in professional practice itself. At the end of the 20th century, however, there has been a growing awareness that physicians are not invincible. Despite inconclusive evidence, suicide, depression, burnout, and other emotional problems seem quite common among physicians [1].

Among Dutch general practitioners (GPs), the percentage of policyholders for income protection with an illness of 3 months or longer increased from 2.0% in 1998 to 3.5% in 2002. In that very same period, the percentage of long-term illness (more than 1 year) among GPs almost doubled from 14% to 26% of all sickness notices. Almost one-fifth of all sickness notices are cases of psychological problems or burnout [2]. Considering these data, it seems interesting to investigate the self-perceived health and burnout level of Dutch general practitioners.

When addressing health issues among physicians, it is important to note that the medical profession is rapidly changing from being traditionally male dominated to being more gender neutral: the percentage of female Dutch general practitioners rose from 4% in 1980 to 26% in 2004 [3], and continues to rise. In the Netherlands, women in the general population report lower general health status, more sick leave, and more work disablement than men [4]. Thus, the increasing number of female physicians in the medical profession might require a gender-specific notion of the health status of general practitioners.

In the present study, we intended to answer the following research questions:

1. What is the level of burnout of male and female Dutch general practitioners? Does this level differ from earlier Dutch studies on burnout among GPs and the Dutch general population?

2. What is the self-reported general health of male and female Dutch general practitioners? Does this health status differ from the Dutch general population?

Methods

Study population

A cross-sectional survey was performed among a sample of 700 general practitioners randomly selected from the Dutch GP population. Because we were interested in gender differences, we stratified according to sex: 50% of our study population was female, while in the total Dutch GP population 25% is female. Data were collected using a postal survey sent out in May 2002. GPs who did not respond within 3 weeks were encouraged by telephone to fill out the questionnaire.

Measures

To measure burnout, the Dutch version of the Maslach Burnout Inventory (MBI) was used [5]. This established instrument measures three burnout dimensions: emotional exhaustion (eight items), which measures feelings of being emotionally overexerted by work; depersonalization (five items), which refers to cynical and impersonal attitudes towards patients; and lack of personal accomplishment (seven items), which assesses feelings of diminished professional competence and achievement. The items were scored on a seven-point scale ranging from “never” (=0) to “daily” (=6). High scores on emotional exhaustion and depersonalization, and low scores on personal accomplishment indicate burnout. To assess the prevalence of burnout, the percentage of problematic scores on each dimension was computed; scores in the upper 25% on emotional exhaustion and depersonalization, and in the lower 25% of personal accomplishment on normative national data (first-line healthcare workers) are considered problematic [5]. A problematic score on emotional exhaustion, combined with a problematic score on depersonalization or a problematic score on personal accomplishment, determined cases of burnout [5].

The following general health indicators were assessed: perceived health (one item: “How do you assess your general state of health?” Possible responses were “very good”, “good”, “not good/not bad”, “bad”, and “very bad”) [4], [6]; the presence of a (chronic) disease (yes/no); and sick leave (the self-reported number of days the physician did not work because of illness during the past year).

The following demographics were assessed: sex, age, work hours, marital status, and number and age of children living at home.

Analyses

All analyses were performed with the software package SPSS 11.0. First, the means and standard deviation of the demographics were computed. Gender differences were then analysed using chi-square tests (categorical values) or t tests (continuous values). One-sample t tests were performed to compare burnout scores from the present study with former studies.

Results

Response and demographics

Three hundred and forty-nine completed surveys were returned and 15 proved undeliverable, resulting in a response rate of 51% (48% women).

Demographic data of the study group (Table I) showed certain differences between male and female GPs that ought to be kept in mind when comparing health differences between genders. First, female GPs were almost 6 years younger (and had younger children). This is best explained by the late but rapid feminization of general practice, resulting in an under-representation of women among older GPs and an over-representation among younger GPs. Because the age differences might interfere with health status, the study group was divided in two age groups for comparisons between sexes and between GPs and the general population.

Table I.  Demographics of the study population.

	Total group (n=349)	Male GPs (n=180)	Female GPs (n=169)	P (male–female)
Age, years	45.9 (SD 7.0)	48.7 (SD 6.7)	43.0 (SD 6.1)	0.000
Work hours	42 (SD 13)	49 (SD 13)	37 (SD 11)	0.000
Marital status (with spouse)	89%	93%	86%	0.034
Number of children	1.7 (SD 1.2)	1.6 (SD 1.3)	1.7 (SD 1.2)	0.458
Age of youngest child, years	9.5 (SD 6.5)	11.3 (SD 6.6)	7.9 (SD 6.0)	0.000
Work hours, spouse	29 (SD 18)	20 (SD 14)	39 (SD 16)	0.000

A second gender difference concerns the work hours of GPs and their spouses. This reflects the typically Dutch “one-and-a-half earner” model, with men working full-time and women working part-time and taking the largest share of responsibility for the household and children. Although we did not assess hours spent on unpaid work (family and household tasks), we might assume that the total workload (paid and unpaid work) is higher for women than for men [7], [8].

Burnout

Burnout was found among 19.4% (95% confidence interval [CI] 15.0–23.4%) of the Dutch general practitioners: 20.7% of the males and 17.6% of the females. The gender difference, however, was not significant (P = 0.47).

More specifically, scores on emotional exhaustion and personal accomplishment did not differ between male and female GPs, or between younger (25–44 years) and older (45–64 years) GPs. Depersonalization, however, was significantly higher among male GPs in both age groups (P = 0.02 and 0.01).

Table II shows the scores on the MBI of GPs in the present study in comparison with former Dutch studies among GPs [5], [9], [10] as well as the general working population in the Netherlands [11]. Comparison of burnout scores with other studies is hindered by the fact that, in our study, the percentage of women was two to three times higher than in comparable studies without gender-disaggregated data [5], [9–11]. Although emotional exhaustion and personal accomplishment in our study did not differ between male and female GPs, depersonalization showed a clear gender difference. To make comparison possible, a subsample of our study group was created, by randomly removing female GPs, to equalize the percentage of females to the percentage in each of the studies used for comparison, which is in line with the suggestions of Van Deurssen [12].

Table II.  Comparison of mean MBI scores of the present study with studies among Dutch GPs in the past and with the Dutch general working population.

	Year	Emotional exhaustion	Depersonalization	Personal accomplishment
Present study (n = 349; 48% female)	2002	2.06	1.71	5.08
Dutch GPs
van Dierendonck et al. (10) (n=567, 15% female)^a	1994	2.29*	1.91	3.81*
Schaufeli et al. (5) (n=562, 17% female)	1994	2.29*	1.90	3.81*
Bakker et al. (9) (n=507, 16% female)^a	2001	2.10	1.82	3.82*
General working population
Kant et al. [11] (n=12 140, 27% female)^b	2004	1.85*	1.54*	4.07*

^aSum scores are converted to mean scores to make comparison possible.

^bScores from MBI-GS.

*Scores are significantly different from a sample of the present study that is adjusted for the percentage of female GPs (see Methods) (one-sample t test; P < 0.01).

General health indicators

Figure 1 presents the perceived general health and Figure 2 the presence of a chronic disease for male and female Dutch general practitioners. For comparison, data are given for a sample of the Dutch general population (in which these variables were assessed in the same way) in both figures [4]. To refine the comparison, the study group was divided into two age groups, in line with the reference of the general population. Female GPs reported good or very good health more often than male GPs, but the difference was not significant (P = 0.19 for age 25–44, and P = 0.16 for age 45–64). Compared to the Dutch general population, the self-reported health of young male GPs was similar (P = 0.87). However, older male GPs (P = 0.01), younger female GPs (P = 0.00), and especially older female GPs (P = 0.00) reported better general health than their fellow countrymen.

Figure 1.  Percentage of Dutch general population and general practitioners who assessed their health as good or very good.

Figure 2.  Percentage of Dutch general population and general practitioners who reported one or more chronic diseases.

No gender differences were found in the prevalence of chronic diseases; however, older GPs suffered significantly more often from (chronic) diseases than younger GPs (males: P = 0.01; females: P = 0.02). The difference between the reported chronic diseases of Dutch GPs and the general population is striking: GPs suffered two (older age group) to three (younger age group) times less often from one or more chronic diseases (P = 0.00).

Most GPs (69%) reported no sickness absenteeism during the past year, and 22% reported only a few days of sick leave (5 days or fewer), males and females equally often. No difference was found between younger (25–44 years) and older (45–64 years) GPs. The percentage of sickness absence (days on sick leave per year / 365 days×100%) was calculated to compare GPs with the Dutch general population. The sickness absence percentage of Dutch GPs was 2.5%. In the same year, this percentage was 5.3% among the general population [4].

Discussion

This study investigated the (mental) health status of Dutch general practitioners, in comparison with former samples of GPs and the general population and with special attention to gender-specific patterns.

Burnout

Burnout was found among 19.4% of our sample of Dutch practising GPs. This percentage is approximately twice as high as the prevalence of burnout in the Dutch general population as well as in a subsample of highly educated professionals (vocational college and university), i.e., 9% vs 8% in 2002 [4]. No data were gathered about GPs not practising at the time of the study. Hence, it can be assumed that the actual percentage is even higher, as GPs suffering from severe burnout symptoms are likely to be on sick leave (i.e., healthy-worker effect).

No gender difference was found in burnout prevalence. This is consistent with the findings of other studies and of the Dutch general population [4], [13], [14]. Nevertheless, we did find that, consistent with previous research, female GPs report less depersonalization than their male colleagues [14–16]. This phenomenon is generally explained by feminine gender role socialization, which emphasizes caring and concern for other people. Thus, women are less likely to respond insensitively to patients and their problems.

Comparison of scores on the three burnout dimensions, after adjustment for the percentage of female subjects, indicates more burnout problems in GPs than in the general population: GPs suffer more from emotional exhaustion and depersonalisation [11]. On personal accomplishment, however, Dutch GPs score much better than the general population: they have more confidence in their own professional capacities.

Although burnout appeared to be quite common among GPs, comparison with studies of Dutch GPs in the past [5], [9], [10] revealed that emotional exhaustion has decreased during the last 10 years, while personal accomplishment has increased and depersonalization has remained stable. Several changes in the work situation of Dutch GPs may explain this improvement. Traditionally, GPs covered hours on call with a rota system in small local groups. During the 1990s, large-scale out-of-hours cooperatives were developed across the country. In 2002, 80–90% of GPs had joined these out-of-hours services, resulting in a large reduction of hours on call, lower levels of work overload, and better working conditions while on call [17]. Also, house calls have been reduced, while the rate of telephonic contact has increased during the past 15 years [18].

General health indicators

Data on perceived general health and the presence of chronic diseases indicate that Dutch GPs feel healthier than their fellow countrymen. Physicians are highly educated, and thus belong to the higher social class, which is known to be related to better health [19]. However, even compared to a highly educated subsample of the general population, GPs report better general health and fewer chronic diseases [4], although the differences are less prominent. This is surprising, considering the high burnout prevalence as well as objective data from insurance companies that show that, around the turn of the century in the Netherlands, long-term mental illness among health workers increased dramatically and health insurance premiums for physicians rose accordingly [2].

Some characteristics of physicians in general could explain this discrepancy. The fact that GPs as physicians have extensive medical knowledge and encounter a great deal of illness and suffering among their patients may make them experience their own health problems as minor and insignificant. Moreover, socialization into the medical profession could make physicians reluctant to report their own health problems. After all, physicians are expected to be healers, constantly attending to their patients’ needs, with compassion, expertise, and competence. High standards are set by patients and society, but also by the medical profession itself. During the professional socialization of medical students, these idealized expectations become part of a physician's self-identity [1], [20]. Personal (health) problems, impeding physicians to meet these expectations, may be perceived as professional failure.

In accordance with these high standards, the threshold for taking sick leave is high for physicians. We found a very low sickness percentage of 2.5%. McKevitt et al. [21] reported that a high percentage of general practitioners (85%) have at some time continued to work when it might have been better to take sick leave. This phenomenon is known as “sickness presenteeism” and is thought to be detrimental to health in the long run, because recovery from illness and distress is hampered [22], [23].

Although the general sickness absence percentage is low, 6% were absent because of sickness for more than 1 month. In combination with the fact that almost one in five physicians suffers from serious signs of burnout, this might indicate that GPs deny their “minor” health problems, allowing them to grow more serious over time and thus increasing the risk of long-lasting mental or physical illnesses.

Another remarkable finding in this study is that female GPs are as healthy as their male colleagues, while in the general population, men report better general health than women [24–26]. Ladwig et al. state that women may adopt the sick role more easily because it is more socially acceptable for them [26]. In gender role socialization, men are taught to be strong and tough, while women receive the message that showing weakness and dependency is acceptable. Female physicians, however, are also socialized into the medical profession, a profession with a “macho mentality” and masculine standards [1]. Thus, female physicians might adopt a masculine self-image that makes the sick role less acceptable to them.

Conclusion

It may be concluded that Dutch GPs (and especially female physicians) are in apparently better health than the general population: they report better general health, fewer chronic diseases, and less sick leave. On the other hand, the percentage of burnout is almost twice as high as in the general population, and 6% reported long-term sick leave. This paradox indicates a high level of sickness presenteeism and underestimation of health problems among GPs. This might be explained by the professional socialization of physicians, which sets high standards and prohibits them from showing vulnerability. This effect is especially strong in female GPs, because it is contradictory to female gender role socialization, creating a larger gap between the illness behaviour of women in the general population and female physicians.

In order to reduce burnout and improve health among physicians, attention should not only be paid to possible causes of (mental) illness such as stressors in the work and private domain. It may be even more important to convince physicians to take their own health as seriously as they take the health of their patients, and thus to admit their own vulnerability. This might not be easy, as the current implicit definition of professional commitment and competence will be challenged [1]. As Miller & McGowen state, “the machismo that currently characterises the medical profession has to be replaced by a professional self-image that incorporates the essential elements of professional practice without requiring unhealthy self-neglect” [1]. Future research should focus on the impact of the professional self-image of both male and female physicians on the experience of their (mental) health problems.

Acknowledgements

This work was funded by MOVIR, Nieuwegein, the Netherlands. The authors would like to thank Dr P. Verdonk for her valuable comments on earlier drafts of this paper.