Continuing medical education and burnout among Danish GPs

INTRODUCTION Pressure on GPs has risen, patients have become more demanding, and governments want doctors to be more efficient and effective, and to provide better service. Many GPs experience more administrative work in their daily commitments and, as a consequence, may be more stressed, unsatisfied, and burnt-out. Burnout among GPs has been shown to be associated with suicide, impaired job performance, tendency toward substance abuse, less provision of preventive services, and negative effects on their mental health. Continuing medical education (CME) consists of educational activities to maintain, develop, or increase the knowledge, skills, and professional performance and relationships that a physician uses to provide services for patients, the public, or the profession. Studies have found that some CME activities give rise to better care processes and better patient outcomes than others. Maintaining high quality services in general practice therefore requires encouragement of CME activities. It could be hypothesised that there is an association between GPs’ experience of burnout and their participation in CME activities. Such an association would mean that a greater effort should be made to recruit those GPs who usually do not take part in CME. However, this remains a hypothesis as it has only been possible to identify a single study analysing the association between ABSTRACT Background There has been minimal research into continuing medical education (CME) and its association with burnout among GPs. Aim The aim of this study was to investigate the association between participating in CME and experiencing burnout in a sample of Danish GPs. Design of study Cross-sectional questionnaire study. Setting All 458 active GPs in 2004, in the County of Aarhus, Denmark were invited to participate. Method Data on CME activities were obtained for all GPs and linked to burnout which was measured using the Maslach Burnout Inventory – Human Services Survey. The relationship between CME activity and burnout was calculated as prevalence ratios (PR) in a generalised linear model. Results In total, 379 (83.5%) GPs returned the questionnaire. The prevalence of burnout was about 25%, and almost 3% suffered from ‘high burnout’. A total of 344 (92.0%) GPs were members of a CME group or a supervision group. Not being a member of either a CME group or a supervision group was statistically significantly associated with doubled likelihood of burnout (PR = 2.2). Among GPs not making use of a practice facilitator, a seven-fold higher likelihood of high burnout was found. Conclusion GPs who were not members of a CME group and did not take part in outreach visits had a higher likelihood of suffering from burnout and high burnout than those who were members of a CME group or received outreach visits. Therefore, not being a member of a CME group could indicate that the GP is more likely to suffer from burnout. Although the present study does not unequivocally establish causality, it would be interesting to see whether staying active in CME may also prevent burnout among GPs.


INTRODUCTION
Pressure on GPs has risen, patients have become more demanding, and governments want doctors to be more efficient and effective, and to provide better service. 1,2 Many GPs experience more administrative work in their daily commitments and, as a consequence, may be more stressed, unsatisfied, and burnt-out. 3,4 Burnout among GPs has been shown to be associated with suicide, 1 impaired job performance, 5 tendency toward substance abuse, 6 less provision of preventive services, 7 and negative effects on their mental health. 8 Continuing medical education (CME) consists of educational activities to maintain, develop, or increase the knowledge, skills, and professional performance and relationships that a physician uses to provide services for patients, the public, or the profession. 9 Studies have found that some CME activities give rise to better care processes and better patient outcomes than others. 10 Maintaining high quality services in general practice therefore requires encouragement of CME activities.
It could be hypothesised that there is an association between GPs' experience of burnout and their participation in CME activities. Such an association would mean that a greater effort should be made to recruit those GPs who usually do not take part in CME. However, this remains a hypothesis as it has only been possible to identify a single study analysing the association between

ABSTRACT Background
There has been minimal research into continuing medical education (CME) and its association with burnout among GPs.

Aim
The aim of this study was to investigate the association between participating in CME and experiencing burnout in a sample of Danish GPs.

Design of study
Cross-sectional questionnaire study.

Setting
All 458 active GPs in 2004, in the County of Aarhus, Denmark were invited to participate.

Method
Data on CME activities were obtained for all GPs and linked to burnout which was measured using the Maslach Burnout Inventory -Human Services Survey. The relationship between CME activity and burnout was calculated as prevalence ratios (PR) in a generalised linear model.

Results
In total, 379 (83.5%) GPs returned the questionnaire. The prevalence of burnout was about 25%, and almost 3% suffered from 'high burnout'. A total of 344 (92.0%) GPs were members of a CME group or a supervision group. Not being a member of either a CME group or a supervision group was statistically significantly associated with doubled likelihood of burnout (PR = 2.2). Among GPs not making use of a practice facilitator, a seven-fold higher likelihood of high burnout was found.

Conclusion
GPs who were not members of a CME group and did not take part in outreach visits had a higher likelihood of suffering from burnout and high burnout than those who were members of a CME group or received outreach visits. Therefore, not being a member of a CME group could indicate that the GP is more likely to suffer from burnout. Although the present study does not unequivocally establish causality, it would be interesting to see whether staying active in CME may also prevent burnout among GPs.

Keywords
burnout; continuing medical education; general practice; general practitioner general practice CME and burnout. An Israeli study among family physicians and paediatricians found that doctors participating in CME activities were less likely to report burnout. 11 The aim of the current study was to investigate the association between participating in CME and experiencing burnout using a sample of Danish GPs.

METHOD
In Denmark GPs are organisationally and financially encouraged to participate in postgraduate training, CME, and outreach visits. Each GP has a personal account that may pay for at least 3 full days of CME activities a year (including compensation for lost earnings and course fees). Furthermore, GPs have the opportunity to participate in locally financed CME groups, for which they are also paid a small fee.
All GPs in the County of Aarhus, Denmark have the opportunity to use, free of charge, the 'practice facilitator' -a model which advocates outreach visits which aim to facilitate desired changes in the GPs' clinical behaviour. In general, most GPs support this organisation of CME activities and most take part to some extent. 12,13

Study population
In a cross-sectional questionnaire study, all 458 active GPs in May 2004 in the County of Aarhus were invited to participate. Four GPs were excluded: one was a member of the research group, and three were on leave.

Questionnaire
The 10-page self-administered questionnaire included scales on burnout and questions about CME activities and GP and practice characteristics. Burnout was measured by the Maslach Burnout Inventory -Human Services Survey (MBI-HSS), an international scale which is considered to be reliable and valid. [14][15][16] This self-rating scale consists of 22 items forming three subscales. Each item is scored on a 7-point ordinal scale, where 'never' is scored 0 and 'every day' is scored 6. The three subscales are: emotional exhaustion (9 items), which refers to reduced energy and decreased job enthusiasm; depersonalisation (5 items), which refers to cynicism and lack of empathy; and personal accomplishment (8 items), which refers to inefficacy.
Each dimension receives a score which, relative to normative population scores, is categorised as low or high. Burnout was defined by a high score on the emotional exhaustion subscale (>26) and/or on the depersonalisation subscale (>9). High burnout was defined by a high score on the subscales for emotional exhaustion (>26) and depersonalisation (>9), and a low score on the personal accomplishment subscale (<34). 14 CME was assessed by asking GPs if they were members of local CME groups or supervision groups (CME with a psychological approach as in the Balint groups), if they used the practice facilitator (outreach visits), and how many days during the past 2-year period they had spent on conferences, courses, and seminars relevant to general practice.
Non-responders were sent a reminder with a new questionnaire after 4 weeks. GPs were remunerated for responding.

Analysis
For each GP, scores were calculated on the three burnout subscales. Whether the scores reached burnout or high burnout was also recorded. Only GPs who had answered all items on a subscale were included in the analysis of subscale scores, and burnout and high burnout was only assessed for GPs who had answered all necessary items.
Days spent during a 2-year period on conferences, courses, and seminars relevant to practice were categorised into four groups: 0-5, 6-9, 10-17, and more than 17 days. This classification was made because every GP in Denmark is entitled to at least 3 full days of course activity a year. GPs who were members of both a CME group and a supervision group were compared with GPs who were members of only one of the groups, and with GPs who were not members of either a CME group or a supervision group.
The association between feeling emotionally exhausted, depersonalised, having low personal accomplishment, suffering from burnout or high burnout, and CME activities was calculated. Associations were calculated as prevalence ratios (PRs) in a generalised linear model with log link for a binomial outcome.

How this fits in
Minimal research has been carried out into continuing medical education (CME) and its association with burnout among GPs. This study found that not being a member of a CME-group is associated with higher prevalence of burnout. Focus in the future should be on why GPs engage in CME, and if CME can prevent burnout among GPs.
Ninety-five per cent confidence intervals (CI) were calculated for proportions and ratios with Pvalues of 5% or less considered statistically significant. Data were analysed using Stata 9.0.

RESULTS
In total, 379 (83.5%) GPs returned the questionnaire. Characteristics of the study population are shown in Table 1. A total of 344 (92.0%) GPs were members of a CME group or a supervision group. Nearly 75% of the doctors spent 10 or more days during a 2-year period on practicerelevant courses, conferences, and seminars.
GPs who were members of neither a CME group nor a supervision group had a statistically significantly increased likelihood of suffering from burnout ( Table 2). Among GPs who did not make use of the possibility of receiving outreach visits, a seven-fold higher likelihood of high burnout was seen. Regarding days spent on relevant courses, no significant association was found with burnout or scores on the subscales. However, spending more than 9 days on relevant activities over a 2year period, and especially more than 17 days, was associated with a tendency towards a decreased likelihood of burnout and high burnout.

Summary of main findings
The study results show that GPs who were not   members of a CME group or who did not take part in outreach visits were more likely to suffer from burnout and high burnout than GPs who were members of a CME group or took part in outreach visits. It was not possible to show an association between number of days spent on courses and burnout. The prevalence of burnout among Danish GPs was about 25%, and almost 3% suffered from high burnout. One in three GPs were found to be lacking a sense of personal accomplishment, one in five suffered from depersonalisation, and one in 10 GPs was emotionally exhausted.

Strengths and limitations of the study
The strengths of this study were the use of an international validated scale for measuring burnout, a high response rate, and statistical precision. The high response rate indicates that the GPs found the questionnaire relevant. As the study was crosssectional, it is not possible to make any causal inference; for example, whether GPs would suffer from burnout less frequently if they participated in CME activities or, more reasonably, if the GPs who chose not to participate were those who suffered from burnout. The validity of the burnout measure has been shown to be valid in cross-sectional studies. However, if the measure is unstable, for example if it depends too much on job stress on a particular day, the measure may be inaccurate. It has not been possible to find data on reproducibility of the MBI-HSS. However, the MBI-HSS has been shown to be highly associated with mental health and job satisfaction, for example, [17][18][19] which indicates that the measure is stable.
Although the study had a high response rate, a small group of physicians did not respond, which may provide a basis for possible selection bias. It is reasonable to assume that GPs experiencing burnout would have less energy and enthusiasm to complete the questionnaire. Given a higher prevalence of burnout among non-responders, the prevalence of burnout may therefore have been underestimated. The non-responding group would not have been expected to have participated in more CME activities, and a possible selection bias would not change the association found between CME and burnout. CME activities were measured from GPs' selfreported use of outreach visits, membership of CME and supervision groups, and use of days on courses. The use of days on courses during a 2-year period may invite recall bias which may explain why no association was detected here.

Comparison with existing literature
The prevalence of burnout and high burnout among GPs found in this study is supported by the findings of a Swiss study which showed that one-third of primary care physicians experienced burnout, and 2.8% experienced high burnout. 18 Studies of UK and Italian GPs found higher levels of emotional exhaustion and depersonalisation, but lower values on personal accomplishment compared with the present findings. 8,20 Only one study has been identified that examines burnout and CME among physicians. 11 Kushnir et al found a negative association between CME activities and burnout among Israeli GPs. 11 The authors suggested that CME and professional updating may reduce the level of burnout. The current study is not completely comparable to their study, however, as a different approach to measure CME and burnout was used.

Implications for future research and clinical practice
GPs who were not members of a CME group had a significant higher likelihood of suffering from burnout and high burnout than those who were members of a CME group. Therefore, not being a member of a group could indicate that these GPs may be more likely to suffer from burnout. Although the present study does not unequivocally establish causality, it seems reasonable to be aware of possible implications if a GP stops CME activities, or to encourage GPs to be a member of a CME group. However, it remains unanswered whether it is the group, the CME, or the combination of the two which is most important. Future research could also investigate whether CME groups may be more attractive to GPs with a similar suitable learning style and behaviour, which may indicate that the groups may not be inclusive enough.

Funding body
This study was funded by the Regional Committee on Quality Improvement in the County of Aarhus

Ethics committee
According to Danish law no ethical approval was necessary to conduct this study