The psychosocial work environment among physicians employed at Danish oncology departments in 2009. A nationwide cross-sectional study

Abstract

Working as a physician at an oncology department has some distinctive characteristics that may lead to a stressful work environment. The present study was conducted to provide a nationwide description of the work conditions of all oncologists in Denmark. By comparing the results of the present study with those of a similar study carried out in 2006, the aim was furthermore to elucidate changes in the psychosocial work environment over time. Material and methods. From May to September 2009, 330 physicians employed at six oncology centres and seven community based oncology departments were invited to participate in a survey based on the short version of the COPSOQ II questionnaire. The results were compared with data from a representative section of Danish employees and with data from the 2006 survey. Results. Two hundred and twenty of the 330 invited physicians returned the questionnaire (response rate 67%). Concerning the aspects quantitative demands, work pace, emotional demands, influence, burnout and stress, the oncologists reported worse work conditions than the average Danish employee. However, with regard to possibilities for development, meaning of work and commitment to workplace, the oncologists reported better work conditions. Between 2006 and 2009, substantial improvement was seen concerning several of the assessed work environment aspects within the group of young physicians at the oncology centres. Conclusion. Though substantial improvement of the work conditions has been achieved between 2006 and 2009, certain aspects of the psychosocial work environment at Danish oncology departments still require attention.

Working as a physician at an oncology department has some distinctive characteristics that may lead to a stressful work environment. Dealing with chronically ill, incurable or dying patients is emotionally demanding [1]. Due to demographic changes, the cancer incidence is rising in the western world. Furthermore, the number of accessible treatment options and the complexity of therapeutic strategies are rapidly increasing. These changes may result in an increased workload and high quantitative demands for health professionals at oncology departments. It has previously been demonstrated that the prevalence of burnout and stress is relatively high among oncologists [2–7]. Apart from having a detrimental influence on the physicians’ quality of life, burnout and stress may result in sickness absence, early retirement or reduced quality of patient care [1]. Thus, the work conditions at hospitals may affect the overall performance of the healthcare system. Consequently, the work environment of oncologists is a subject that clearly deserves attention.

In 2006, the Association of Young Oncologists in Denmark (FYO) conducted a survey of the psychosocial work environment among junior doctors (in this paper defined as residents, senior residents and senior registrars) employed at the oncology centres in Denmark [8]. This survey demonstrated that young physicians in oncology experienced a psychosocial work environment that was substantially worse compared to the average Danish employee concerning several different aspects. Severe problems were found with regard to quantitative demands, work pace, emotional demands, influence, predictability, work-family conflict, burnout and stress. Since 2006, a number of initiatives have been taken to improve the work conditions and educational environment of young oncologists. In the same period, Danish oncology departments generally experienced increased funding, improved staffing and substantial investment in new equipment. Both factors have the potential to affect the work conditions.

The present survey was conducted to provide an updated nationwide description of the psychosocial work environment at Danish oncology departments and to elucidate changes in the work environment between 2006 and 2009. Opposed to the 2006 survey, the present survey also included senior doctors (in this paper defined as consultants) and physicians employed at the decentralised community based departments (for description of these departments see discussion).

Material and methods

Participants

Three hundred and thirty physicians employed at the oncology centres in Herlev, Odense, Rigshospitalet, Vejle, Aalborg and Aarhus as well as at the community based oncology departments in Esbjerg, Herning, Hillerød, Næstved, Roskilde, Sønderborg and Viborg were invited to participate in the survey. The physicians were identified by asking the departments for lists of all doctors with clinical duty employed as of May 1st 2009. Since virtually all non-surgical cancer treatment in Denmark is undertaken by these departments, the survey can be regarded as nationwide providing a practically complete coverage of all clinically active oncologists. In Denmark, medical oncology and radiation oncology is combined into a single specialty designated ‘clinical oncology’.

Methods

The doctors were invited to participate in the survey by e-mail at the end of May 2009. By the middle of June, a reminder was sent to non-responders and at the end of August a second reminder was sent. Data collection was finalised at the beginning of September. Participation in the survey was voluntary and anonymous. The participating centres and departments accepted publication of the results in a non-blinded format with regard to the name of the centre/department. Submission of the replies was made electronically through a dedicated website and data were exported directly to a spreadsheet. Like the survey conducted in 2006, the present study was based on the short edition of the ‘second version of the Copenhagen Psychosocial Questionnaire’ (COPSOQ II) from the National Research Centre for Working Environment (NRCWE) [9,10]. This questionnaire has been applied to a representative sample of 3517 Danish wage earners from 2004 to 2005 [10,11] and data from this survey [12] was used as standard reference in our study. The short version of the COPSOQII questionnaire includes 40 questions covering 23 different aspects of the psychosocial work environment (English translation of the questionnaire is provided in Supplementary Table I available online at http://informahealthcare.com/doi/abs/10.3109/0284186X.2012.702926). The questions were answered by ticking off on a 4- or 5-point scale. Mean scores for each aspect was calculated and compared with the responses of the reference cohort and with data from the 2006 survey. In addition to the COPSOQ II questionnaire, the survey included background questions about place of employment, age, gender and position.

Table I. Demographic characteristics of the study populations.

	Men (%)	Women (%)	Total (%)
A	Entire 2009 survey
Position
Res.	16 (17.0)	60 (47.6)	76 (34.5)
S. res.	12 (12.8)	16 (12.7)	28 (12.7)
S. reg.	9 (9.6)	20 (16.9)	29 (13.2)
Cons.	57 (60.6)	30 (23.8)	87 (39.6)
Dept. type
Cent.	79 (84.0)	112 (88.9)	191 (86.8)
Com. dept.	15 (16.0)	14 (11.1)	29 (13.2)
Age
< 30	2 (2.1)	18 (14.3)	20 (9.1)
30–39	20 (21.3)	50 (39.7)	70 (31.8)
40–49	31 (33.0)	37 (29.4)	68 (30.9)
50–59	33 (35.1)	11 (8.7)	44 (20.0)
≥ 60	8 (8.5)	10 (7.9)	18 (8.2)
Total	94 (42.7)	126 (57.3)	220
B	Young physicians at oncology centres 2009
Position
Res.	13 (40.6)	52 (59.8)	65 (54.6)
S. res. + s. reg.	19 (59.4)	35 (40.2)	54 (45.3)
Total	32 (26.9)	87 (73.1)	119
C	Young physicians at oncology centres 2006
Position
Res.	25 (67.6)	36 (54.5)	61 (59.2)
S. res. + s.reg	12 (32.4)	30 (45.5)	42 (40.8)
Total	37 (35.9)	66 (64.1)	103

Abbreviations identical to those used in Table 2.

Boxes indicate which group of numbers the percentages refer to.

Statistical analysis

The collected data were primarily analysed using descriptive statistical methods. Differences between individual departments and demographic groups were elucidated using the work environment categories (much better than average, better than average, on average, worse than average and much worse than average) set by NRCWEs reference population [12]. Due to a limited number of employees and a relatively low response rate at the community based departments, it was not possible to conduct a meaningful analysis of the data at department level from this subgroup. Furthermore, such analysis would potentially jeopardise the anonymity of the participants. Therefore, the data from the community based departments were analysed together. As mentioned previously, our survey from 2006 only covered junior doctors at the oncology centres (n = 103) [8]. In order to investigate the changes in the psychosocial work environment between 2006 and 2009, the data from 2006 were compared with the scores from the subset of junior doctors employed at the oncology centres in 2009 (n = 119). Q-Q plots of the raw data demonstrated that the COPSOQII scores exhibited a reasonable approximation to the normal distribution. A two-sided unpaired t-test with a significance level of 0.05 was used to assess the statistical significance with regard to selected comparisons as indicated in Tables II and III. The statistical analysis was conducted using SPSS and the basic statistics applets in Excel. As we did not have access to raw data from the ‘Danish work force reference population’, statistical testing involving this group was made under the assumption that that scores from the reference population had the same standard deviation as that observed in our dataset. Statistical testing was not conducted with regard to individual centres nor for different professional charges as the number of observations in each subgroup was too small to allow a meaningful statistical analysis.

Table II. COPSOQ II scores for the entire 2009 study population.

Table III. a and b. COPSOQ II scores for young physicians at oncology centres and changes between 2006 and 2009.

Results

Demography

A total of 220 out of 330 invited physicians responded to the questionnaire corresponding to a response rate of 67%. The proportion of responders at each department is shown in Table II. Demographic characteristics of the surveyed population are shown in Table I.

Quantitative measures of the psychosocial work environment

Table II provides the mean scores for the assessed psychosocial work environment aspects with regard to the reference population and the physicians at oncology departments in 2009. Scores according to place of employment, gender and position are shown. The results are colour coded according to NRCWEs categories [12]. It appears that the oncologists were much more congested than the average employee in terms of quantitative demands, emotional demands and work-family conflict. Oncologists were more strained in terms of work pace, lack of influence, lack of social support from supervisors, burnout and stress. However, the surveyed oncologists reported a much higher degree of commitment to their workplace than the average employee. For all other categories, physicians in oncology performed better than the reference group. All differences except for social support, justice and general health perception were statistically significant. Employees at the oncology centres in Aalborg and Vejle had the highest number of work environment aspects that were categorised as worse or much worse compared to the reference cohort. For these departments, a very high extent of work-family conflict, stress and burnout was noted. Apart from a significantly higher degree of influence among employees at the community departments, the study did not provide evidence of any major differences in the psychosocial work environment between oncology centres and community based departments. Regarding position, we observed a clear trend that the degree of influence as well as the extent of work-family conflict increases with increasing professional charge. With regard to gender, we noticed that women reported more problems concerning lack of influence, lack of predictability, burnout and stress. These differences were statistically significant. In contrast, there was no obvious difference regarding work-family conflict.

Changes between 2006 and 2009 for young oncologist at the oncology centres

Table Ib and c provides the demographic characteristics of the young physicians at the oncology centres in 2006 and 2009, respectively. As shown, the proportion of women had increased substantially between 2006 and 2009, particularly among the youngest physicians. Among male employees, a shift was observed from the majority being residents in 2006 to the majority being senior residents or senior registrars in 2009. Table IIIa shows data from the 2009 survey whereas Table IIIb shows the changes in the psychosocial work environment calculated by simple subtraction between the scores obtained in 2009 and 2006. Table IIIa is colour coded according to the same principle as in Table II and in Table IIIb colour codes indicate the direction and magnitude of the observed changes. As indicated, an overall improvement of the psychosocial work environment has taken place concerning most of the assessed aspects. The improvements were most pronounced for aspects such as quality of leadership, predictability, commitment to the workplace, trust regarding management as well as justice and recognition. With regard to quantitative demands, emotional demands, influence and work-family conflict, the conditions were still classified as much worse compared to the reference population. Female employees in particular appear to have experienced extensive improvements in the psychosocial work environment. Opposed to the 2006 survey, the female participants from the oncology centres no longer reported psychosocial work conditions that were obviously worse than those for the male participants. It is noteworthy, that the male participants in the 2009 survey experienced the highest degree of work-family conflict compared to female participants though this difference did not reach statistical significance with a p-value of 0.08.

Discussion

Methodological considerations

Like our 2006 survey [8], the present study was based on the short version of the COPSOQII questionnaire [9]. This questionnaire has been developed to provide a research tool for work environment professionals as well as for assessment of the work environment at individual workplaces [10]. In 2004–2005, the questionnaire was subjected to a thorough validation procedure based on a cohort of 3517 randomly selected Danish wage earners [10,11]. As part of this, the reliability of the questionnaire was investigated using different psychometric approaches including tests for internal consistency [10], construct validity [11], predictive validity [13] and test–retest reproducibility [14]. The results support the validity of questionnaire [10]. COPSOQII has become a cornerstone of the mandatory workplace risk assessments that are carried out by Danish companies every third year [10]. Together with its predecessor COPSOQI, the questionnaire has been translated into several languages [10,15]. We therefore consider the questionnaire well suited for the purpose of the present survey.

In the present study, we used the average scores from a representative section of the Danish work force (n =3517) [10,11] as our standard reference. This cohort was also used as reference cohort in our 2006 survey [8]. Whether or not this cohort constitutes a reasonable comparator to oncologists is of course debateable. We are well aware that a substantial proportion of NRCWE's reference population have jobs that have little in common with that of oncologists. Nevertheless, we still believe that this comparison can be useful in the sense that it puts the work conditions of oncologists into perspective and point out aspects that deserve particular attention. Furthermore, we have not been able to find suitable COPSOQII reference data from a population (e.g. people with a university degree or healthcare workers) with greater resemblance to oncologists. A small fraction of NRCWE's reference population was made up by 35 ‘doctors and dentists’. COPSOQ II scores from 2005 are available from this subset of the reference population [16]. Nevertheless, these data are not in a format that allows a direct numerical comparison with our findings. Furthermore, the results are hampered by a very limited sample size.

The present study was the second nationwide survey addressing the psychosocial work environment among oncologists in Denmark. In contrast to the previous survey [8], the present also included senior oncologists (consultants) and oncologists employed at the community based departments. Opposed to the 2006 survey, the questionnaire was web-based. Otherwise, the two studies were very similar. Intentionally, the surveys were carried out during the same seasonal period (spring-summer). The response rate in the present survey was 67%. Similar response rates have been reported in other studies addressing the psychosocial work environment at hospitals [3–7] and we therefore consider it acceptable. Nevertheless, the response rate was somewhat higher in our 2006-survey (87%) [8]. A possible reason for this could be that the 2006 survey aimed at a smaller and more homogeneous group that was highly dedicated to the survey due to a general experience of a stressful work environment at that time. The difference in response rates may potentially introduce bias. However, it is hardly possible to estimate the magnitude or direction of such bias.

Interpretation of the results

Our survey demonstrated that the oncologists reported work conditions classified as ‘much worse’ than those of the reference cohort concerning quantitative demands, emotional demands and work-family conflict. Also with regard to work pace, influence, stress and burnout, substantial strains were found on the oncologists (Table II). A recently published paper provides a comprehensive and thorough review of the scientific literature addressing the work environment of physicians [1]. According to this review, the problems demonstrated by our survey seem to be quite common among physicians. Several surveys have addressed the psychosocial work environment at oncology or radiotherapy departments [3–7]. In agreement with our survey, these studies show that burnout and stress is relatively frequent among health professionals in cancer care. A similar conclusion was reached in a systematic review and meta-analysis from 2007 [2].

In our study, women reported significantly more problems than men concerning recognition, influence, predictability, justice, quality of leadership, burnout and stress. In contrast, there was no obvious difference regarding work-family conflict (Table II). In the interpretation of these results, it should be taken into account that the proportion of young physicians was higher among women while the proportion of senior doctors was higher among men (Table Ia) and that work-family conflict as well as influence seems to increase with increasing professional charge (Table II).

Non-surgical cancer treatment in Denmark is undertaken by two different types of oncology departments; oncology centres and community based departments. Both department categories are publicly run and tax financed. The oncology centres typically provide treatment for a wide range of malignant diseases and offer radiotherapy as well as various systemic cancer therapies. With exception to Vejle and Aalborg, the oncology centres were located in the vicinity to a university with a faculty of medicine. The community based departments were located at medium sized district hospitals and typically provided systemic treatment for a limited number of common malignancies (e.g. breast, lung and colon cancer). Though, the oncology centres and community based departments differ substantially from each other, the COPSOQ II scores were surprisingly alike. The only significant difference was seen for ‘influence’. This may reflect that the medical staff at the community based departments (n = 6–15) was substantially smaller than at the oncology centres (n = 28–64).

Relatively large differences were observed between individual departments (note that these differences have not been subjected to a formal statistical testing). Especially, at the centres in Vejle and Aalborg widespread and quite severe problems were reported. Particularly with regard to work-family conflict, burnout and stress, very high scores were seen. This does probably to some extent reflect that these two centres (that did not have an affiliated medical school in 2009) have faced particular difficulties recruiting and retaining doctors for a longer period. Consequently, these departments have frequently experienced vacancies and understaffing. Unfortunately, our survey did not include information about the average number of working hours per week. Nevertheless, a number of studies have reported that long working hours has a deleterious influence on the work environment [1,17] and we consider it likely that some of the problems seen in Vejle and Aalborg can be attributed to long working hours combined with high quantitative demands. In our view, the differences between departments furthermore underline that a stressful work environment is not a given fact at oncology departments and poor work conditions should not be accepted. A similar conclusion was reached by the authors of the aforementioned meta-analysis of studies addressing the work conditions of oncologists [2].

In contrast to many other studies addressing the work environment of oncologists, we used a comprehensive work environment questionnaire rather than an inventory with a specific focus on burnout or stress. Thus, our survey provides the opportunity to elucidate the positive aspect of working as an oncologist. In general, the surveyed oncologists reported excellent working conditions concerning meaning of work, commitment to the workplace and possibilities for development (Table II). We think that this should be seen as a resource and that the organisation of the work should be used to support these factors for instance by providing protected time for research and educational activities. Such initiatives would furthermore be in accordance with the recommendations of the recently updated European core curricula for clinicians in radiation oncology [18].

The comparison between the 2006 survey and the subset of junior doctors at the oncology centres in the 2009 survey clearly indicates that substantial improvements in the work conditions have been achieved (Table IIIa and b). The present study per se does not provide an exact explanation for these changes. Nevertheless, our 2006 survey [8] was spurred by extensive complaints from the members of FYO about an increasing workload and a widespread lack of job satisfaction. At the turn of the millennium, the cancer incidence was rapidly rising and public expectations for efficient high quality cancer treatment were increasing. However, these expectations were difficult to meet due to a considerable shortage of specialists in oncology. In 2006, 34% of the specialist positions in Denmark were vacant [19]. Dropout rates from the specialty training programmes were high and most departments experienced problems recruiting and retaining young physicians. Thus, the oncology departments were under substantial pressure. In 2007, FYO arranged a ‘work environment conference’ with participation of the department heads and the clinical lecturers from the participating centres. The departments were strongly encouraged to address the problems revealed by the 2006 survey [8]. The improvements observed in the present survey may very well be the result of local initiatives to optimise organisation and management at the departments including increased emphasis on educational activities. This assumption is supported by the fact that the most pronounced improvements were seen for aspects that are closely related to management issues (i.e. trust regarding management, quality of leadership, recognition, justice and predictability) (Table IIIb). During the same period, the ‘Danish Cancer Treatment Plan II’ [19] was implemented. In most departments, the financial situation was substantially improved resulting in purchase of new equipment and an expansion of the staff. Furthermore, an increased interest in the speciality was seen among young physicians. These factors may also have contributed to the improvements seen between 2006 and 2009. In our view, the observed changes between 2006 and 2009 strongly indicate that a concerted effort to improve the work environment is worthwhile and that the work conditions of physicians are profoundly influenced by socioeconomic factors. The latter assumption is supported by a recent longitudinal study addressing the work environment of Swedish doctors in 2002 and 2009. This study demonstrated substantial improvements in the psychosocial work environment after the resolution of certain structural and economical issues [20].

Opposed to the 2006 survey, the female participants from the oncology centres no longer reported psychosocial work conditions that were significantly worse than those for the male participants. Actually, the male participants in the 2009 survey experienced the highest degree of work-family conflict compared to female participants (though this difference did not reach statistical significance). Again, this should be interpreted in the context that the residents generally experienced a lower degree of work-family conflicts compared to their older colleagues (Table II and IIIa) and that the proportion of residents had increased among women, while the proportion of senior residents and senior registrars had increased among men (Table Ib and c).

Despite the achieved improvements, a number of unsolved work environment issues still remain at the surveyed oncology departments (i.e. high quantitative demands, high emotional demands, lack of influence, work-family conflict, burnout and stress). In our view, this should certainly be taken seriously. It has been demonstrated that the intention of physicians to quit their job is strongly associated with emotional exhaustion [1,4]. Furthermore, substantial evidence indicates that stressful work conditions can lead to substance abuse, suicide, health related problems and sickness absence [1,13,21]. Consequently, work environment problems may have severe consequences not only for the individual physician but also for the entire healthcare system [1]. Thus, strategies to improve the work conditions of oncologist should be given a high priority by managers at oncology departments. A recently published French study that shows some resemblance to our 2006 survey addressed the prevalence and possible sources of burnout and stress among oncology residents in 2009 [4]. By means of factorial analysis, the study demonstrated that workload and emotional demands were among the major stressors. Other studies have reached similar conclusions [1,6,7]. This emphasises the importance of achieving a reasonable balance between the available resources and the workload at oncology departments. Furthermore, it is probably also of immense importance to provide oncologists with efficient coping strategies to deal with the inevitable emotional strain that originates from taking care of seriously ill patients [1]. In that regard, possible interventions could be mentoring programmes, support groups and communication training [4,6]. It is well documented that physicians are relatively reluctant to seek help when they are in mental distress or experience symptoms of burnout [1]. Therefore initiatives for prevention and early intervention against burnout and stress are certainly warranted. A recent intervention study demonstrated that short-term counselling for burnout and stress had a significant and sustained impact on physicians’ mental wellbeing [22]. Furthermore, the formation of Balint groups has been demonstrated to improve the job satisfaction of physicians [23]. In our view, similar interventions should be considered at oncology departments.

The 2009 survey was conducted before the recent economic recession had any major impact on the Danish healthcare system. Due to the economical crisis, many departments are now faced with budget cuts and staff reductions. This may have a negative impact on the psychosocial work environment and the commitment of the employees [24]. In order to monitor the consequences of the altered socioeconomic situation, FYO, DSKO and DKOO intend to repeat the work environment survey among Danish oncologists within the coming years. In our experience, the short version of the COPSOQ II questionnaire constitutes a robust, sensitive and feasible tool to monitor and analyse the work conditions among healthcare professionals employed at hospitals. Our 2006 survey provided a detailed quantitative measure of the work environment that could be used to address the problems experienced by junior doctors at the oncology departments. We therefore strongly encourage colleges in other specialties and countries to carry out similar surveys. Hopefully, this will provide data from which the work environment of healthcare professionals will benefit.

Conclusion

The present nationwide survey of the psychosocial work environment at Danish oncology departments has shown that the oncologists experienced substantial quantitative and emotional demands compared to the average Danish employee. Relatively high scores for burnout and stress were observed. These findings are in agreement with the results of a number of recent studies and underline the importance of strategies to counteract a stressful work environment at oncology departments. Our survey indicates that an increased focus on work environment and the allocation of resources seem to improve the work conditions of oncologists.

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

This survey was conducted with financial support from DSKO and DKOO. DSKO and DKOO each provided DKK 10,000 to cover expenses to computer assistance.

Both authors work at Danish oncology centres. CN Andreassen is board member of the Association of Young Oncologists in Denmark.