Acceso abierto

Empathy and burnout in Slovenian family medicine doctors: The first presentation of jefferson scale of empathy results


Cite

Introduction

Empathy has been recognised as an important and powerful part of communication in general practice, strengthening the physician-patient relationship (1, 2). Burnout in family medicine doctors (FMDs) at the front line of health care, disrupts this relationship (3, 4, 5, 6). An empathic attitude is described as the capacity to understand what another person is experiencing from within their frame of reference (7). Physicians with empathic attitudes experience greater patient satisfaction and better patient compliance and adherence to treatment (8, 9). In addition, they tend to face fewer medical errors (10), have improved health outcomes (11), report fewer symptoms of burnout and have better well-being (12). The concept of empathy consists of cognitive and affective components (13), with a known moderate correlation between the concepts of sharing understanding and sharing emotion in patient-care (14).

Almost 65% of European FMDs exhibit signs of burnout (3) with various and non-specific symptoms (6). As the response to chronic emotional and interpersonal stressors at work, burnout leads to reduced job performance (15); the physicians’ behaviour can have a detrimental effect on the health of patients and lead to more malpractice suits (16) and patient dissatisfaction (17). In physicians, the heavy workload and the lack of financial and organisational resources are important risk factors for burnout (15, 18), with an intense empathic attitude leading to emotional exhaustion and causing burnout syndrome (19). Some studies determined depersonalisation to be the main reason for a decrease in empathic attitude (20).

In Slovenia, burnout has only been evaluated in family medicine trainees, who scored highly (71%) in at least one burnout dimension (21). This is the first study in Slovenian family medicine focusing on the relationship between empathy and burnout and aiming to assess the extent of burnout and the level of empathic attitude in FMDs, and also to explore their associations with sociodemographic factors, working conditions and health. The Jefferson Scale of Empathy (JSE) for physicians was used as a validated self-assessment tool for the first time in Slovenia (see Additional File).

Methods
Participants and Procedure

This was a cross-sectional survey of Slovenian FMDs; 565 out of 1139 FMDs in Slovenia (22) were invited twice by e-mail to complete an online survey. The invitation was sent through the e-mail distribution lists of the Slovenian syndicate of FMDs (396 specialists’ e-mail addresses) and family medicine trainees (169 trainees’ e-mail addresses).

The questionnaire, which had been validated previously, comprised socio-demographic and other questions concerning working conditions, health and well-being, the Slovenian version of the Maslach Burnout Inventory (MBI) (4), and the Jefferson Scale of Empathy – Health Professional version (JSE-HP) (23, 24). The data were collected from April to June 2016.

The response rate was 56%; of 316 respondents, 123 (38.9%) were family medicine trainees and 193 (61.1%) were specialists (who had finished a four-year period of specialised training), aged 40±10.2 years.

Instruments
Socio-Demographic Characteristics Questionnaire

The participants answered demographic questions assessing gender (male/female), age (years), time working in family medicine (years), marital status (single/married/in a relationship/widowed), and children (yes/no). Further questions regarded working conditions and working environment (urban/rural) and workload (the number of patients per day (<40/40-60/>60)), emergency care duty during the regular workday (yes/no), the number of nightshifts per month (0/1-3/≥4)). At the end, there were some questions that concerned the self-reported health of the physician (the number of sick leave days per year (0/1-5/≥6), having a chronic illness (yes/no)) and their self-assessment of their general health, mood and emotional state on a five-point Likert-type scale (1=poor, 5=excellent).

Self-Assessment of Empathic Attitude

The JSE-HP (JSE in further text) was developed by Hojat et al. to evaluate the empathic capacity of practitioners in health professions, including physicians (23, 24). It consists of 20 items, which use a 7-point Likert-type scale (1=strongly disagree, 7=strongly agree) to elicit responses, with a score range of 20-140 (24). Previous studies have suggested a three-factor structure, with the components being Perspective Taking, Standing in the Patient’s Shoes and Compassionate Care (24, 25). The first two subscales address the cognitive aspect of empathic behaviour/attitude (23). The validity and reliability of the JSE were evaluated (11, 23, 24); it has been translated into 53 languages and used in more than 80 countries worldwide (26). The only Slovenian study so far used the JSE in medical students (JSE-S) and confirmed the three-factor structure of the 18-item scale (13). For our study, the JSE was translated into Slovenian and the authorisation for its implementation was obtained. Cronbach’s α coefficients for JSE subscales were 0.865 for Perspective Taking, 0.722 for Standing in the Patient’s Shoes, 0.784 for Compassionate Care and 0.798 for the total of 20 items.

Self-Assessment of Burnout

The MBI is the gold standard for assessing burnout (4), using 22 items scored on a 7-point Likert-type scale (0=Never, 6=Every day). The MBI consists of 3 subscales: Emotional Exhaustion ((EE), 9 items, score range from 0 to 54); Depersonalisation ((D), 5 items, score range from 0 to 30); and Personal Accomplishment ((PA), 8 items, score range from 0 to 48). High scores on the EE and D subscales, combined with low scores on the PA subscale, indicate high levels of burnout. Cut-offs for high burnout were determined by the upper quartile for each dimension, and were for EE>37, for D>15 and <30 for PA. Slovenian version of MBI (27) was used and Cronbach’s α for EE subscale was 0.929, 0.765 for D and 0.801 for PA. The MBI has previously been tested and used several times on groups of healthcare staff, including FMDs (3, 6, 21).

Results of Confirmatory Factor Analysis for MBI and JSE

The confirmatory factor analysis was used to address psychometric properties and measurement invariance of MBI and JSE (28). We calculated the average variance extracted (AVE), maximum shared variance (MSV) and average shared squared variance (ASV) for the items loading on a construct. Conducted AVEs in Table 1 were all above 0.5, which represents a good conversion of MBI and JSE items (29). The resulting 3-factor structure for both MBI and JSE were confirmed as adequately fitting the data. For measurement invariance, the comparative fit index (CFI) and the root mean square error of approximation (RMSEA) were measured. Given that that CFI values were above 0.9 and RMSEA values were below 0.08, this was considered as acceptable (30, 31).

Psychometric properties and measurement invariance by gender and age of the MBI and JSE scales.

Psychometric propertiesAVEMSVASV
MBI: Emotional Exhaustion0.5720.0280.014
MBI: Depersonalisation0.5160.0310.016
MBI: Personal Accomplishment0.5520.0300.015
JSE: Perspective Taking0.5840.0270.014
JSE: Standing in the Patient’s Shoes0.5120.0320.017
JSE: Compassionate Care0.5500.0300.015
MBI scaleJSE scale
Measurement invariance
CFIRMSEACFIRMSEA
Gender: males (n=57) versus females (n=259)
Configural invariance0.9320.0260.9440.020
Metric invariance0.9290.0260.9300.026
Scalar invariance0.9100.0300.9110.031
Age: 27-39 years (n=182) versus 40 years or above (n=134)
Configural invariance0.9480.0190.9340.025
Metric invariance0.9400.0210.9300.026
Scalar invariance0.9200.0270.9180.027

AVE: average variance extracted, MSV: maximum shared variance,

ASV: average shared squared variance

CFI: comparative fit index, RMSEA: root mean square error of approximation

Data analysis

The sample was presented by the frequency and percentage distribution or by the values of the mean and standard deviations. Univariate statistical analysis included various statistical tests: the t-test for independent samples, the one-way analysis of variance, Pearson’s correlation coefficient and Spearman’s correlation coefficient. G*Power software (version 3.1.9) was used to calculate the achieved statistical power (32). Firstly, the confirmatory factor analysis was used to address psychometric properties and measurement invariance. In multivariate analysis, linear regression was used to calculate the associations between demographic variables, factors of empathy and burnout dimensions (EE, D, PA). The results of linear regression were presented by β coefficient, t value and p-value. All the analyses were performed using the SPSS version 22.0 for MS Windows (IBM Corp., Armonk, NY), with the significance set at p<0.05. Additionally, multiple comparisons bias was addressed by changing the level of statistical significance (p<0.001).

Results

Most of the responding FMDs were females (78.9% trainees and 83.9% specialists), working in this speciality for 11.2±10.4 years. The majority were married or in an intimate relationship (272 (86.1%)), most of them had children (218 (69%)). At the time of the survey, more FMDs worked in urban (199 (63%)) than in rural (117 (37%)) settings. Approximately half of the participants had to provide emergency care during regular worktime (152 (48.1%)) vs. 164 (51.9%)). A quarter did not work night shifts (75 (23.7%)), the same proportion reported night shifts at least 4 times per month, while one to three night shifts were reported by 166 (52.5%) respondents. Most of the participants examined at least from 40 to 60 patients per a working day (230 (72.8%)), 53 (18.7%) FMDs dealt with 60 or more patients per day and only a minority worked with less than 40 patients per day (33 (10.4%)).

Participants mainly reported no chronic illnesses (230 (72.8%)) and used either none (136 (43%)) or less than six sick leave days in a year (119 (37.7%)). Only 61 (19.3%) needed 6 or more sick leave days per year. Self-assessment of their general health was well (3.5±0.9), the same for mood and emotional state (3.2±0.9).

Empathic Attitude and Burnout Level in the FMDs

A mean total JSE score (JSEtot) was 112.8±10.2; mean scores of burnout dimensions were 27.8±11.6 for EE, 10.8±5.5 for D and 33.5±6.0 for PA (Table 2). The highest proportion of physicians reported high burnout in one dimension (24.8%), 17.2% reported it in two and 6% in all three dimensions. Altogether, almost half (48%) reported high burnout in at least one dimension, while a quarter (25.2%) scored high in EE and even more in D (25.8%) and PA (26.2%).

Univariate analysis of JSE and MBI scores in FMDs by demographic characteristics, working conditions and health status (sick leave days per year, presence of chronic illness).

EEDPAJSEtot
Categorical variablesnMSDt (p)MSDt (p)MSDt (p)MSDt (p)
Score:31627.811.610.85.533.56.0112.810.2
Gender:ES=0.3P=0.51.686ES=0.0P=0.10.259ES=0.3P=0.51.819ES=0.0P=0.10.279
male5725.412.4(0.093)11.05.3(0.796)34.86.2(0.070)112.49.9(0.780)
female25928.411.410.85.533.25.9112.910.3
Patients/day:ES=0.5P=0.72.481ES=0.3P=0.41.641ES=0.1P=0.10.613ES=0.2P=0.20.782
<403323.111.0(0.014)9.45.0(0.102)34.15.3(0.540)114.28.4(0.435)
≥4028328.411.611.05.533.46.1112.710.5
Night shifts/month:ES=0.2P=0.31.341ES=0.2P=0.31.404ES=0.2P=0.31.274ES=0.0P=0.10.292
<424127.311.8(0.181)10.65.6(0.161)33.36.2(0.204)112.810.5(0.771)
≥47529.411.111.64.934.35.3113.29.6
Emergency care duty:ES=0.2P=0.41.742ES=0.1P=0.10.785ES=0.2P=0.41.812ES=0.1P=0.11.098
yes15229.011.1(0.083)11.15.4(0.433)34.15.4(0.069)113.510.2(0.273)
no16426.712.010.65.632.96.4112.210.3
Chronic illness:3.1460.1000.1412.582
yes8631.311.4(0.002)10.85.9(0.921)33.45.8(0.888)115.49.9(0.010)
no23026.611.510.95.333.56.1112.010.2
Continuous variablesrprprprp
Age (years):0.213 (P=0.9)<0.001-0.036 (P=0.1)0.534-0.009 (P=0.1)0.8720.080 (P=0.3)0.161
Years in current speciality:0.185 (P=0.9)0.001-0.038 (P=0.1)0.5110.005 (P=0.1)0.9310.087 (P=0.3)0.130
Sick leave days/year:

Spearman’s correlation coefficient was calculated for ordinal scale of sick leave

0.081 (P=0.3)0.1590.013 P=0.1)0.820-0.054 (P=0.2)0.3480.037 (P=0.1)0.525

M: mean value, SD: standard deviation, t: Student’s t-test, r: Pearson’s correlation coefficient, df: degrees of freedom=314, ES: effect size (Cohen’s d), P: achieved power

EE – Emotional Exhaustion, D – Depersonalisation, PA: Personal accomplishment, JSEtot– total JSE score

Correlation between Socio-Demographic Factors, Working Conditions and JSE and MBI Scores

There were no statistically significant differences in the EE, D and PA scores according to gender, while older physicians and those with more work experience turned out to be more emotionally exhausted; however, the correlation was weak (Table 2). In regard to working conditions (patients per day, night shifts per month, emergency care duty), physicians who examined more than 40 patients per day had a higher EE. More characteristics are presented in Table 2.

Multivariate analysis revealed a higher EE, higher D and lower PA in specialists, compared to trainees. Physicians working in a rural environment were more emotionally exhausted than those in an urban area. A lower PA was associated with female gender, while physicians with children reported a higher PA, yet no significant correlations between marital status and burnout dimensions were identified. See Table 3.

Multivariate analysis of associations between demographic factors, factors of empathy and individual burnout dimensions (EE, D, PA).

EEDPA
Categorical variablesBtpBtpBtp
Gender (female/male)0.091.620.106-0.03-0.550.583-0.12-2.160.032
Working experience (specialist/trainee)0.283.94<0.0010.172.240.026-0.20-2.920.004
Marital status (in a relationship/single)-0.03-0.550.582-0.09-1.510.133-0.01-0.120.907
Children (yes/no)0.050.750.454-0.03-0.370.7110.172.580.010
Working environment (rural/urban)0.122.030.0430.040.630.532-0.06-1.190.236
Years in current speciality0.00-0.010.995-0.12-1.640.1030.000.020.988
Perspective Taking0.050.820.413-0.10-1.530.1270.355.80<0.001
Standing in the Patient’s Shoes-0.04-0.620.536-0.11-1.700.0910.081.280.203
Compassionate Care-0.07-1.130.259-0.10-1.570.1180.081.350.178
R2=0.129, df=9, p<0.001R2=0.078, df=9, p=0.006R2=0.224, df=9, p<0.001

R2: Coefficient of determination

E – Emotional Exhaustion, D – Depersonalisation, PA – Personal Accomplishment

Correlation between Physicians’ Health, Empathic Attitudes (JSEtot) and Burnout

Physicians with a chronic disease scored a higher JSEtot and they were more emotionally exhausted. Subjective evaluations of physicians’ general health, mood and emotional state were not significantly correlated with their empathic attitude, nor were the number of sick leave days. Details are in Table 2.

Correlation between Empathic Attitude and Burnout Dimensions

JSEtot was weakly negatively correlated with D (r=-0.224, p<0.001), yet the correlation between JSEtot and PA was positive and of moderate strength (r=0.372, p<0.001). However, EE was not correlated to JSEtot. Additional multivariate analysis positively associated Perspective Taking with PA (Table 3).

Discussion

This study assessed the level of empathic attitude and extent of burnout in Slovenian FMDs, and tested associations between JSEtot and individual burnout dimensions with socio-demographic factors, working conditions and physician health (Tables 2, 3). Psychometric properties of Slovenian MBI and JSE scales are also presented (Table 1), introducing the Slovenian version of JSE as validated and highly recommended instrument.

The participants reported a higher level of burnout when comparing mean values for each dimension with European FMDs (Table 2) (3, 33) and Slovenian family medicine trainees (21). A large cross-national burnout study (34) included Slovenian psychiatry trainees, but used a 16-item MBI-GS scale, which made the comparison very difficult. Another hurdle is different speciality, given that psychiatrists could exhibit even higher burnout and have different work process than FMDs. The positive association between emotional exhaustion and age and work experience (Table 2) was identified, concordant with a Slovenian study of family medicine trainees results (21). Maslach (15) discovered greater burnout in those at the beginning of their careers, but warns of survival bias. With regard to work experience, specialists reported higher burnout scores in all three dimensions, in comparison to trainees (Table 3). This is plausible, considering the situation in the Slovenian healthcare system, with specialists taking on an even higher workload, i.e., an excessive number of patients and extensive bureaucracy (35). The overall higher burnout in specialists (Table 3) could be attributed also to the sampling method.

Slovenian female FMDs reported lower personal accomplishment than males (Table 3), while in an European burnout study (3), a strong association between male gender and a high score in all three burnout dimensions was found. Parenting was associated with higher personal accomplishment scores, whereas living in an intimate partnership was not associated with any dimension of MBI (Table 3). Parenting probably shapes personality more than having a partner, in terms of hardiness, self-esteem, non-avoidant coping style, which are found to protect against burnout (15). An increased tendency to experience burnout in those who are single or not married was reported previously (15), while Park et al. reported of no correlation between MBI and marital status or parenting (36).

The correlation between workload and emotional exhaustion (Table 2), supported by the findings highlighting difficult working conditions in Slovenian primary health care, with 90% of physicians dealing with at least 40 patients and 20% with at least 60 patients per day (Table 2), is concordant with a previous Slovenian study of family medicine trainees and Croatian FMDs (21, 37). An additional burden was demonstrated in those working in rural family medicine clinics (Table 3), which is concordant with the study in which those working in rural settings scored higher on the MBI (28).

Unlike burnout, there are few studies that deal with physicians’ empathy, yet research into this topic is on the rise (38). It is difficult to compare scores for empathic attitude, as different instruments and study populations were in use (39, 40). The FMDs in this study reached a JSEtot (Table 2) comparable to that of medical students (40, 41), but the scores were lower than in doctors in other studies (11, 24). These findings could be attributed to cultural, educational and organisational differences of studied populations. Some other researchers also shortened the scale to improve internal consistency (13, 36). Literature often describes greater empathy in female physicians and medical students (1, 13, 19, 24, 36, 40, 41, 42), but no correlation between gender and empathic attitude was discovered in this study (Table 2).

FMDs with a chronic illness reported higher emotional exhaustion and showed a higher empathic attitude (Table 2). Physical illness affecting burnout has not yet been adequately researched and there is a high possibility of burnout in the presence of a persistent stressor, such as chronic disease (43). The impact of chronic illness on empathic capacity was examined in some qualitative studies, showing greater empathy in those who had experienced illness themselves (44).

More empathic participants felt greater personal accomplishment (Table 2), similarly to studies of Spanish (42) and French (45) FMDs and to family medicine trainees (20); physicians with greater empathic attitudes were reported to be less emotionally exhausted (36). Cognitive component of empathy (Perspective Taking) was associated with greater personal accomplishment (Table 2), similarly to Paro et al. (46), who reported personal accomplishment to be significantly associated with decreasing personal distress and which was found to be a reliable predictor for perspective taking. Some research shows that deficits in perspective taking alone might be a risk for burnout, whereas higher perspective taking and empathic concern might be protective (19, 47).

Given that multiple comparisons bias was additionally addressed by holding alpha error rate at 5% and changing the level of statistical significance (p<0.001), this emphasised the most important results of this study, i.e., a positive association between emotional exhaustion and age and working experience ((r=0.213, p<0.001) and (β=0.28, p<0.001), respectively), and between personal accomplishment and cognitive component of empathy (β=0.35, p<0.001). That is concordant with several previous studies and strengthens the reliability of association between empathy and burnout. Contrary to previous studies, the correlations between age and burnout in FMDs could be explained by the Slovenian health care system organisation.

In this study, organisational issues are shown to be a high risk factor for burnout (Table 2), high workload being an issue in Slovenia as well as in other developed countries (3, 15, 37). The situation in the country, with an aging population, the lack of physicians working in primary care, a decrease in interest in family medicine residencies, and shortage of time for the patient (48), suggests an urgent need for psychological help and support for FMDs, taking into consideration empathic attitude, which is associated with the feeling of personal accomplishment.

This study, being the first in Slovenia to assess burnout and empathic attitude in trainees and specialists in family medicine, and with a satisfactory response rate (56%), opens a new dimension of the physician-patient relationship and emphasises the empathic approach to family medicine. The results from this study could be the starting point for a discussion on the restructuring of curricula for both medical students and family medicine trainees, and support for the work of specialists, since FMDs work in relative isolation in their outpatient clinics and are deprived of peer support, as capacity building was shown to be empowering (49, 50). Additionally, the Slovenian version of JSE is presented and could be applied in further research.

Limitations of the Study

One of the main limitations of our research is sampling, being of convenience, with only a part of FMDs (members of syndicate and trainees) and not all FMDs were approached. Secondly, as the study was cross-sectional, the responses were analysed at a certain moment. Furthermore, self-assessment scales, such as JSE and MBI, are subjective and biased, as they are based on the respondent’s self-observation. Finally, there was also a slightly larger proportion of women included in the study (82%), compared to female FMDs in Slovenia (73.5%, (51)). Therefore, further research is needed on representative, random samples that would confirm our findings, especially prospective longitudinal studies to explore the association between empathic attitude and burnout.

Conclusions

The associations between emotional exhaustion, workload, age and work experience, depict important issues of Slovenian family medicine. Given that the association between burnout and empathy was confirmed, the importance of learning empathic communication and peer-support-based capacity building in FMDs for the prevention of burnout have been shown. There is a need for further intervention studies in medical students and FMDs, in order to evaluate these findings, elaborate peer support interventions and enhance the quality of patient care.

eISSN:
1854-2476
Idioma:
Inglés
Calendario de la edición:
4 veces al año
Temas de la revista:
Medicine, Clinical Medicine, Hygiene and Environmental Medicine