The impact of theory of mind, stress and professional experience on empathy in Romanian community nurses—a cross-sectional study

The nurse-patient relationship is modulated by emotional competences. To be empathetic to patients and to have the capacity to read emotions is a fundamental aspect of care. These capacities can be regulated by other emotional abilities as well as technical skills in order to acknowledge the patient’s needs and offer therapeutically appropriate response [10]. To the best of our knowledge, empathy was not studied in a Romania national sample of community nurses. Our study improves the understanding of the emotional competences of this professional group.

As required by roles played by these professionals, the mean levels of empathy are increased, with 74.7% of the sample being over the threshold of high empathy level (scoring > = 45), while the mean (SD) reported score was 49.1 (6.7). A recent study investigating the empathy among senior medical students from Romania has found a similar mean (SD) level of empathy of 48.76 (5.65) [8]. Using the same instrument, when studying nurses that work in psychiatric wards, Alhadidi et al. [11] have identified lower levels of empathy, reporting a mean (SD) of 46.07 (6.7). Conversely, Cosper [12] found similar levels of empathy 49.8 (4.3) in nurses working in the Intensive Care Unit (ICU).

Our results indicate that as experience as a community nurse accumulates, the empathy level increases: nurses with more than 6 years of experience had 56% more chances to be in the higher quartile of empathy, suggesting that job or field retention is a key factor for increased empathy. Observing the tertiles of years of experience, the empathy level increases steadily. A publication on the 10 year trend of empathy in intensive care units from China [13] found a reduction in empathy levels. Cosper et al. [12] did not identify a significant association between empathy level and years of experience as a nurse, however found a negative correlation with the time spent as a ICU nurse. For nurses working in oncology, increasing experience was related to reduced capacity to verbally respond and listen to the patients, perhaps as their roles and the need to learn technical skills were expanding [14]. A longitudinal study [15] that included nursing students involved as part of their curricula in different aspects of training, found that nursing students that had more clinical encounters with patients had a decline in empathy levels after 1 year of follow-up, while students with less clinical encounters did not significantly modify the empathy during the evaluation. A cross-sectional study [16] has similar results, reporting a drop in empathy in 4th year students, as compared to students from the first year. This erosion of empathy in the clinical setting was attributed to patients’ negativity and time restricted interaction between nurses and patients [17, 18]. Others [19] have found that among emergency department nurses, low level of manager support was a significant predictor of higher levels of burnout and compassion fatigue. Using a subgroup analysis among master nursing students, Hakansson Eklund et al. found different empathy levels, with public-health nursing, midwifery and psychiatry students scoring higher than anesthesia and intensive care master students [20].

Similarly, studies performed in medical doctors have shown that empathy levels decrease over time due to a coping mechanism for burnout, in association with depressive symptoms [21]. In medical students, several studies have discovered a decline in empathy at the beginning of the clinical phase of medical education [18, 22, 23]. Several studies looked into the reasons for this decline and found that its coping mechanism for distress [24] and was caused by school related factors such as lack of time, physical and psychological fatigue, competitiveness, performance demands, cramming, stress, high workload [25, 26], some of these factors belonging to the hidden curriculum to become a medical professional [27].

Possible explanations on the differences between our study and studies on nurses working in high demanding positions like in ICU, oncology or psychiatry or while in training, lie in the job description of community nurses, which are engaged in a diverse and challenging health landscape, providing care is different settings from patient’s own houses, nursing homes, family doctor practices, or other types of clinics.

High levels of self-perceived stress are a good predictor of low levels of empathy, both in univariate and multivariate models. Others have also reported negative associations between stress/anxiety and empathy, with stress/anxiety management having a positive effect on empathy levels [2, 28, 29]. Extensive prior research [30] has shown how anxiety reduces empathy, suggesting that experiencing emotions associated with uncertainty, increases egocentric perspectives when perceiving the mental states of others.

Cognitive empathy—the capacity to understand other people by ascertaining their mental state, also known as the theory of mind, quantified in our study by the RMET, shows a positive association with the TEQ score, sharing 8% of variance. In the multivariate model, by each one point increase in the RMET score, the chances to be in the high quartile of TEQ as opposed to the low quartile grow by 16.5%. In a recent study [2] on nurses, cognitive empathy was a mediator between emotional intelligence and humanization of care, with 18% shared variance between cognitive empathy and humanization of care. In another study [31], performed in general population, using a prediction model for the RMET score by components of the emotional intelligence, have found that the highest contributor is the understanding subscale (β = 0.28, p < 0.001), followed by perceiving (β = 0.09, p < 0.01), and managing subscales (β = 0.08, p < 0.05). The variance shared between emotional and cognitive intelligence was 8.4%, similar to our results. Using subgroup analysis from the general population, Baron-Cohen et al. [32] have found lower shared variance, at about 2%, between cognitive empathy and empathy quotient.

Our study did not find significant differences in empathy levels between genders in this population, similar to several studies on nursing students [13] or psychiatric ward nurses [11] or ICU nurses [12]. Also, education level, familial background and income levels did not influence the empathy level in our sample. Others have found that empathy was higher in nurses with higher education, albeit those nurses were working in ICU [33] or were students with a previous degree before enrolling in studying nursing [34]. Studies performed to capture empathy in family dynamics have shown that positive family environments are positively related to empathy levels in adolescents [35]. In a longitudinal study [36], high compassion for others in adulthood predicted, after a 10-year follow-up, higher adulthood social-economic status, although not vice versa.

Considering that high levels of empathy are critically needed for nursing, it is important to acknowledge that empathy can be learned and taught. Making a review of European training programs for emotional competencies in vocational training, Sauli et al. [1] have found four types of training available in Europe, however they lacked a rigorous scientific approach, proposing a set of guidelines for program development. Using an educational program designed by patient and family advisers, Cosper [12] demonstrated an increase in empathy levels of nurses, higher in those younger than 30 years, as compared to their counterparts. Thus recognizing the benefit of early career educational programs.

Two of the strengths of this study are the large number of participants and the high response rate of 85.8%, leading to the assessment of a representative national sample of community nurses from Romania. Looking into the response rate of different organizational research based on questionnaires, Baruch et al. [37] found that the response rate in public/state sector varied from a minimum of 27.0% to a maximum of 82.8% with a mean (SD) of 54.5 (16.7) and in Health care from a minimum of 17.4% to a maximum of 94.0% with a mean (SD) of 53.8 (20.0). By comparison, our results sit in the highest quartile. If this high response rate, which was obtained without reminders is used as a proxy for empathy, the result is concordant to high scores in TEQ, where 74.7% of responders had a high empathy score. If we would use this proxy, it might be that the non-responders would have lower empathy levels, inducing a selection bias, because people interested in participation usually have a genuine interest in the subject of the research.