Empathy and Perspective Taking in Understanding Burnout
Empathy and Perspective Taking in Understanding Burnout
Of the 81 general practitioners approached during the annual French National Congress of General Practice, 75% completed the questionnaires (n = 61) and of the 680 members of the 'Société de Formation Thérapeutique du Généraliste', 36% completed the questionnaires (n = 247). In total, 308 questionnaires were completed. The total response rate was 39%. We excluded 14 questionnaires due to incomplete data. Therefore, the final sample was composed of 294 participants. No statistically significant differences were found between the two sub-samples on any of the variables, and thus the two sub-samples were merged.
The characteristics of the sample are described in Table 1. The description of burnout and empathy scores is reported in Table 2. In preliminary analyses we observed that women had higher levels of emotional exhaustion (N 143, M 18.4, Standard Deviation (SD) 9.6) than men (N 151, M 15.4, SD 10.3) (p < .01 d = .30). They had lower levels of depersonalization (M 6.0, SD 4.8 in comparison to M 7.1, SD 5.1) but higher levels of empathic concern (M = 48.8, SD 5.3 in comparison to M 44.9, SD 5.3) than men (respectively p < .05 d = .23 and p < .001 d = .40). The other dimensions did not differ according to gender. When examining bivariate correlation with burnout, burnout was found to be significantly associated with marital status (living alone) (r = 0.14, p < 0.05), but was not related to other demographics or practice variables. Therefore, following the recommendations of Cohen (1988) we controlled for marital status in subsequent analyses. Burnout was also related with lower scores in perspective taking (r = −0.24, p < 0.01) and empathic concern (r = −0.17, p < 0.01). When looking into components of burnout, higher emotional exhaustion was associated with being a woman (r = 0.15, p < 0.05), living alone (r = 0.12, p < 0.05), and lower JSPE Standing in the Patient's Shoes (r = −0.14, p < 0.05). There were no significant associations between emotional exhaustion and other empathy measures. Depersonalization was negatively associated with being a woman (r = −0.12, p < 0.05) and all empathy measures (r = −0.18 to −0.32, p < 0.01). Higher personal accomplishment was associated with higher empathy on all empathy measures (r = 0.18 to 0.40, p < 0.01). When predicting burnout in a linear regression model using empathic concern, perspective taking and their interaction as predictors, we observed that each of the predictors was uniquely associated with the outcome, with lower empathy scores predicting higher burnout (model adjusted for marital status). Higher empathic concern (ß = −0.17, p < 0.05) and higher perspective taking (ß = −0.21, p < 0.001) predicted lower burnout. Interestingly, the interaction term was also significant (ß = 0.11, p < 0.05) suggesting a moderating role of empathic concern in the perspective taking – burnout relationship. The total model explained 10% of the variance of burnout (R = 0.08 without controls). In order to interpret this interaction we plotted burnout levels as a function of empathy patterns determined by median splits and observed that burnout was even lower in participants with high perspective taking scores when empathic concern scores were also lower. To investigate this interaction in a clinically meaningful manner, we explored factors associated with the odds of extreme burnout using a logistic regression model (odds adjusted for marital status, Figure 1). This interaction is well illustrated when comparing the proportion of burnout according to levels in PT and EC (Figure 2). Proportions of participants showing burnout differed according to the combined effects of PT and EC (χ = 22.74, p < 0.001, Cramer's V = 0.28). Higher PT could be even more beneficial when participants showed higher EC and higher EC could be beneficial only when participants showed higher PT.
(Enlarge Image)
Figure 1.
Perspective Taking, Empathic Concern and their Interaction as Predicting Burnout in 294 French General Practitioners (Odds Ratio Adjusted for Marital Status). We dichotomized Perspective Taking and Empathic Concern by the median; low Perspective Taking = score ≤ 53, high Perspective Taking = score > 53; low Empathic Concern = score ≤ 46 (n = 151), high Empathic Concern = score > 46 (n = 143). The odds for burnout were significantly lower in participants with high PT (OR 0.37, 95% CI 0.21–0.65, p < 0.001), high EC (OR 0.57, 95% CI 0.33–0.98), p < 0.05) and high on both PT and EC (OR 0.31, 95% CI 0.15–0.63, p < 0.001).
(Enlarge Image)
Figure 2.
Interaction Between Perspective Taking and Empathic Concern to Explain Burnout Frequencies. PT = Perspective Taking; EC = Empathic Concern. Perspective Taking and Empathic Concern were dichotomized at the median. Low Perspective Taking = score ≤ 53, n = 153; High Perspective Taking = score > 53, n = 141; Low Empathic Concern = score ≤ 46, n = 151; High Empathic Concern = score < 46, n = 143. We divided the participants into four groups based on the distribution of their empathy scores: (1) low Empathic Concern–low Perspective Taking (n = 99), (2) low Empathic Concern–high Perspective Taking (n = 52), (3) high Empathic Concern–low Perspective Taking (n = 54) and (4) high Empathic Concern–high Perspective Taking (n = 89). Percentage of extreme burnout for low Empathic Concern–low Perspective Taking, low Empathic Concern–high Perspective Taking, high Empathic Concern–low Perspective Taking and high Empathic Concern–high Perspective Taking was 33%, 22%, 32% and 11% respectively.
Results
Of the 81 general practitioners approached during the annual French National Congress of General Practice, 75% completed the questionnaires (n = 61) and of the 680 members of the 'Société de Formation Thérapeutique du Généraliste', 36% completed the questionnaires (n = 247). In total, 308 questionnaires were completed. The total response rate was 39%. We excluded 14 questionnaires due to incomplete data. Therefore, the final sample was composed of 294 participants. No statistically significant differences were found between the two sub-samples on any of the variables, and thus the two sub-samples were merged.
The characteristics of the sample are described in Table 1. The description of burnout and empathy scores is reported in Table 2. In preliminary analyses we observed that women had higher levels of emotional exhaustion (N 143, M 18.4, Standard Deviation (SD) 9.6) than men (N 151, M 15.4, SD 10.3) (p < .01 d = .30). They had lower levels of depersonalization (M 6.0, SD 4.8 in comparison to M 7.1, SD 5.1) but higher levels of empathic concern (M = 48.8, SD 5.3 in comparison to M 44.9, SD 5.3) than men (respectively p < .05 d = .23 and p < .001 d = .40). The other dimensions did not differ according to gender. When examining bivariate correlation with burnout, burnout was found to be significantly associated with marital status (living alone) (r = 0.14, p < 0.05), but was not related to other demographics or practice variables. Therefore, following the recommendations of Cohen (1988) we controlled for marital status in subsequent analyses. Burnout was also related with lower scores in perspective taking (r = −0.24, p < 0.01) and empathic concern (r = −0.17, p < 0.01). When looking into components of burnout, higher emotional exhaustion was associated with being a woman (r = 0.15, p < 0.05), living alone (r = 0.12, p < 0.05), and lower JSPE Standing in the Patient's Shoes (r = −0.14, p < 0.05). There were no significant associations between emotional exhaustion and other empathy measures. Depersonalization was negatively associated with being a woman (r = −0.12, p < 0.05) and all empathy measures (r = −0.18 to −0.32, p < 0.01). Higher personal accomplishment was associated with higher empathy on all empathy measures (r = 0.18 to 0.40, p < 0.01). When predicting burnout in a linear regression model using empathic concern, perspective taking and their interaction as predictors, we observed that each of the predictors was uniquely associated with the outcome, with lower empathy scores predicting higher burnout (model adjusted for marital status). Higher empathic concern (ß = −0.17, p < 0.05) and higher perspective taking (ß = −0.21, p < 0.001) predicted lower burnout. Interestingly, the interaction term was also significant (ß = 0.11, p < 0.05) suggesting a moderating role of empathic concern in the perspective taking – burnout relationship. The total model explained 10% of the variance of burnout (R = 0.08 without controls). In order to interpret this interaction we plotted burnout levels as a function of empathy patterns determined by median splits and observed that burnout was even lower in participants with high perspective taking scores when empathic concern scores were also lower. To investigate this interaction in a clinically meaningful manner, we explored factors associated with the odds of extreme burnout using a logistic regression model (odds adjusted for marital status, Figure 1). This interaction is well illustrated when comparing the proportion of burnout according to levels in PT and EC (Figure 2). Proportions of participants showing burnout differed according to the combined effects of PT and EC (χ = 22.74, p < 0.001, Cramer's V = 0.28). Higher PT could be even more beneficial when participants showed higher EC and higher EC could be beneficial only when participants showed higher PT.
(Enlarge Image)
Figure 1.
Perspective Taking, Empathic Concern and their Interaction as Predicting Burnout in 294 French General Practitioners (Odds Ratio Adjusted for Marital Status). We dichotomized Perspective Taking and Empathic Concern by the median; low Perspective Taking = score ≤ 53, high Perspective Taking = score > 53; low Empathic Concern = score ≤ 46 (n = 151), high Empathic Concern = score > 46 (n = 143). The odds for burnout were significantly lower in participants with high PT (OR 0.37, 95% CI 0.21–0.65, p < 0.001), high EC (OR 0.57, 95% CI 0.33–0.98), p < 0.05) and high on both PT and EC (OR 0.31, 95% CI 0.15–0.63, p < 0.001).
(Enlarge Image)
Figure 2.
Interaction Between Perspective Taking and Empathic Concern to Explain Burnout Frequencies. PT = Perspective Taking; EC = Empathic Concern. Perspective Taking and Empathic Concern were dichotomized at the median. Low Perspective Taking = score ≤ 53, n = 153; High Perspective Taking = score > 53, n = 141; Low Empathic Concern = score ≤ 46, n = 151; High Empathic Concern = score < 46, n = 143. We divided the participants into four groups based on the distribution of their empathy scores: (1) low Empathic Concern–low Perspective Taking (n = 99), (2) low Empathic Concern–high Perspective Taking (n = 52), (3) high Empathic Concern–low Perspective Taking (n = 54) and (4) high Empathic Concern–high Perspective Taking (n = 89). Percentage of extreme burnout for low Empathic Concern–low Perspective Taking, low Empathic Concern–high Perspective Taking, high Empathic Concern–low Perspective Taking and high Empathic Concern–high Perspective Taking was 33%, 22%, 32% and 11% respectively.
Source...