Does medical students’ personality have an impact on their intention to show empathic behavior?

Empathy is a crucial skill for every physician. Recent literature has shown that physician’s level of empathy correlates with patient satisfaction (Winefield and Chur-Hansen 2000; Schmid Mast et al. 2004; Hojat et al. 2011; Derksen et al. 2013), patient compliance (Winefield and Chur-Hansen 2000; Schmid Mast et al. 2004; Hojat et al. 2011), and clinical outcome (Winefield and Chur-Hansen 2000; Derksen et al. 2013). It was additionally demonstrated that empathy is both learnable and trainable (Drdla and Löffler-Stastka 2016), meaning its teaching is an essential duty of every medical university. However, it was often found that the level of empathy was not only the result of the quality of the training, but heavily contingent on the student’s personality as well. An effective way to measure the personality structure is the Big Five personality model (NEO Five-Factor Inventory-3 (NEO-FFI)) (McCrae and Costa 2012), which include the five personality traits of Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. We found seven studies investigating the relationship between the NEO-FFI and empathy among Japanese, German, Spanish, Portuguese, and American university students. The level of empathy was evaluated by different questionnaires: Del Barrio et al. (2004), Nettle (2007), and Wakabayashi and Kawashima (2015), and Melchers et al. (2016) measured the level of empathy by the Empathy Quotient (Baron-Cohen and Wheelwright 2004). All four studies showed a significant association between Agreeableness (Del Barrio et al. 2004; Nettle 2007; Wakabayashi and Kawashima 2015; Melchers et al. 2016), two between Extraversion (Wakabayashi and Kawashima 2015; Nettle, 2007), two between Conscientiousness (Del Barrio et al. 2004; Melchers et al. 2016), and one between Openness (Del Barrio et al. 2004) and Empathy Quotient score. Other authors (Magalhães et al. 2012; Costa et al. 2014) measured the level of empathy using the Jefferson Scale of Physician Empathy (JSPE) (Kane et al. 2007), and both showed a significant correlation between Agreeableness and Openness and the Jefferson Scale of Physician Empathy score. Others (Lourinho and Severo 2013; Melchers et al. 2016) used the Interpersonal Reactivity Index (Davis 1983), demonstrating a correlation between Agreeableness (Lourinho and Severo 2013; Melchers et al. 2016), Conscientiousness (Melchers et al. 2016) and Openness (Lourinho and Severo 2013) and Interpersonal Reactivity Index score. Inconsistent findings may reflect the use of different questionnaires; regardless they emphasize the need for further conceptual reflection. In the subject-specific literature, much is being investigated and discussed on the subject of “empathy” (Pedersen 2009, p. 307). The researchers in this area are confronted with a number of problems, as various authors describe (see Pedersen 2009; Stepien and Baernstein 2006). A single, uniform definition of empathy is still a contested point (see Stepien and Baernstein 2006; Preusche 2013; Preusche and Wagner-Menghin 2013). Finally, the numerous survey methods do not clearly refer to a concise definition of empathy. Moreover, the interventions to increase empathy often seem to have no solid theoretical foundation (see Stepien and Baernstein 2006). The distinction between attitudes towards empathic behavior and empathic behavior per se is often neglected (see Preusche 2013).

Empathy consists of different aspects and processes (e.g., Batson 2011; Decety and Jackson 2004; Ickes 1993; Levenson and Ruef 1992; Zaki et al. 2008): cognitive, emotional, and unconscious influences (Knaus et al. 2016), self-other distinction, empathic behavior, and also some form of willingness—we are not constantly and unwillingly empathizing with everyone we see, a point stressed by de Vignemont (2006). Although it seems quite common to speak of empathy as an automatic reaction, it has to be assumed that it cannot be so (de Vignemont 2006; Lamm and Silani 2014).

The response rate was 65.5%. One hundred thirty-one questionnaires were returned: 54 from the second-year students and 78 from the fourth year. Not all questionnaires were filled out completely. Therefore, N = 100 could be taken for calculation, and the mean age of the participants was 23–24 years (range 19–34), and the gender distribution was quite equal (male/female = 50.8:48.5%) (Table 1).

Of the students, 35.6% did not list which specialization they planned on pursuing after graduation. Of the rest, 29.4% reported a desire to specialize in Internal Medicine, 24.7% in Surgery, 12.9% in Pediatrics, 8.2% in Psychiatry, 4.7% in Neurology, 3.5% in General Medicine, 3.5% in Dentistry, 3.5% in Radiology, 2.4% in Anesthesia, 2.4% in Ophthalmology, 2.4% in Dermatology, one person in Rehabilitation, and another in General Research. Compared to the second year, significantly more students of the fourth year preferred to specialize in Internal Medicine (32 versus 25%) and fewer in Pediatrics (0.07 versus 21.8%). Gender-specific differences could be shown, too. Significantly more men wished to pursue surgery or psychiatry while significantly more women wished to pursue pediatrics.

Our study showed only a few small correlations between the personality traits Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness (NEO-FFI) and the student’s GISEB. Although extraverted persons seemed to have a greater intention to show empathic behavior, the personality trait itself was shown to be no predictor for General Intention (GISEB). Measured with these questionnaires, this means that the intention to show empathic behavior is not influenced by personality traits. Our results are not necessarily in conflict with the literature (Del Barrio et al. 2004; Nettle 2007; Magalhães et al. 2012; Lourinho and Severo 2013; Costa et al. 2014; Wakabayashi and Kawashima 2015; Melchers et al. 2016), which showed an association between personality and empathy. It might be that persons with specific personality traits vary in showing empathy even though the intention might be equal. In light of this distinction, these are encouraging results. With comprehensive and regular communication training of empathy, certain disadvantages due to personality may be ameliorated.

Another important finding is the decrease of general intention to show empathic behavior as students’ progress through their curriculum. These findings are in keeping with the literature which shows a decrease in empathy over the course of the curriculum, especially after the first practical experiences (Newton et al. 2008; Hojat et al. 2009; Colliver et al. 2010; Neumann et al. 2011).

Possible reasons might be the lack of models (Seitz et al. 2017) or the increasing stress (Neumann et al. 2011; Seitz and Löffler-Stastka 2016) that accumulates over the course of students’ studies. Other reasons might be students’ insecurity, feelings of being overwhelmed in working with clients, or the overload of processing client feelings (Pololi et al. 2001; Seitz and Löffler-Stastka 2016). This might lead to social withdrawal and the halting of prosocial behavior (Pololi et al. 2001; Lamm et al. 2007).

In our study, no gender-specific difference regarding General Intention to Show Empathic Behavior could be shown. However, literature shows differences in actual behavior. Female students and physicians on average have longer conversations with patients (Roter et al. 1997; Bylund et al. 2008; Löffler-Stastka et al. 2016) and include the psychosocial situation of the patient more often (Roter et al. 1997; Löffler-Stastka et al. 2016) compared to male colleagues. Additionally, women show a stronger non-verbal communication style (Roter et al. 1997; Bylund et al. 2008). However, it should be considered that the female students in our study showed higher scores in Neuroticism, suggesting the tendency to feel emotionally involved more easily. Some authors (Gleichgerrcht et al. 2013) support this thesis. As a result, women and men may have the same intention to show empathy, but due to gender-specific differences in the personality, women show more empathy towards patients.

The fourth interesting finding in our study is the influence of gender and personality in the students’ wish for specialization after graduation. This topic has already been discussed in several publications (McGrath and Zimet 1977; Wallick et al. 1999; Buddeberg-Fischer et al. 2003; Buddeberg-Fischer et al. 2006; Pawełczyk et al. 2007; Hojat and Zuckerman 2008; Malhi et al. 2011; Rotge et al. 2015). It is also interesting that the intended specialization typically changes after practical training in the fourth year. More students were interested in Internal Medicine than Pediatrics, showing that a lot of students may have had an inaccurate picture of their specific desired specialization in their first years of university. The influence of age regarding the general intention is most likely explained by the higher study year with increasing age.

Our study revealed that although literature shows an association between empathy and personality, the intention to show empathic behavior is not influenced by personality traits measured by the NEO-FFI. This emphasizes the importance of comprehensive and regular communication and empathy training at Medical University Vienna. In the framework of this training, gender-specific and personality-related differences should be discussed. For future studies, it would be useful to evaluate how to successfully include such topics in the training.