INTRODUCTION

Conceptually, empathy is considered in phenomenological philosophy as the union or fusion with other beings or objects. In psychology, it is considered as an indirect experience of an emotion close to the emotion experienced by another person. This complicity with the other’s situation culminates in the potential to think and develop social or emotional support¹. According to Sampaio et al.², the relationship between affectivity and cognition are important for the internalization and construction of principles that govern the individual’s behavior in society. Empathy is, therefore, a multidimensional construct that involves different levels of affectivity, cognition and behavior³. Through the affective component, the emotional states of others are shared, while the cognitive and behavioral components are responsible for the ability to reason about the mental states of other people and consider communication and help actions⁴.

Thus, empathy is a skill that allows you to perceive and understand the feelings and perspective of others, encompassing cognitive, affective and behavioral components^5)-(7. The cognitive component, called perspective taking, correlates with the capacity to put yourself in someone else’s place and deduce their feelings, without necessarily experiencing them. This component is considered vital for social interactions, as it allows understanding and predicting behaviors^5),(8. The affective component is based on sharing and understanding the emotional states of others (mirror neurons), in which the actions observed in others are represented internally in the observer’s brain^6),(7),(9. The behavioral component (empathic concern) is related to the motivation to care for vulnerable individuals^6),(8),(10.

Several strategies have been tested for developing empathy in medical students^7),(11)-(17; however, measuring this skill remains a challenge. Scales have been developed to measure empathy in different scenarios, such as the Empathy Inventory (EI), which assesses empathy in the context of social interaction situations⁶, the CARE scale, which assesses the patient’s perception of empathy in the context of clinical care¹⁸ and the Interpersonal Reactivity Index, which assesses the affective and cognitive dimensions of empathy in the context of social relationships⁵. However, in medical education, the most often used is the Jefferson Scale of Physician Empathy - JSPE, which assesses empathy in the context of the doctor-patient relationship¹⁹, predominantly considering the cognitive attribute of empathy, not addressing its affective component. Despite being commonly used to measure empathy in healthcare environments, a recent systematic review found inconsistencies in some psychometric properties of the JSPE, which despite demonstrating structural validity, internal consistency and convergent validity, shows that the evidence is limited regarding the properties of reliability, measurement error and cross-cultural validity²⁰.

Thus, considering the importance of empathy in the academic development, a scale is necessary that addresses the affective and cognitive component of the construct and that is culturally and linguistically closer to the scenario of Brazilian students. This study aims to develop a scale that would evaluate the attitudes of Brazilian medical students regarding empathy in healthcare environments. This is based on the assumption that a person’s attitude can be accessed through their communication or demonstration, with the attitude being essentially a mental disposition of a potential action²¹.

METHOD

RESULTS

The sample included 207 students, from the 4^th to the 6^th year of the medical course, from private and public Brazilian institutions. In this group, the age distribution was: 76.8% aged 20 to 25 years old, 14.5% aged 26 to 30 years old and 8.7% aged over 30 years old. The majority of students were female (73.4%), single (94.7%), with a family income above R$5,000.00 (84.1%), intending to pursue a clinical specialty (56.5%), with serious illness in the family (74.9%) and without a chronic illness (86.5%). Moreover, the largest proportion of students who participated in this research were in the 4^th year (51.7%) of the medical course (Table 1).

Statistical analysis of the proposed scale

The version of the scale that was sent to the students contained 34 items. The analysis of student responses to each of the 34 items on the scale showed there was polarization in the extremes of the response scale in six of the items, where more than 85% of students indicated the answer 1 or 5 (items 1, 2, 9 , 21, 26, and 29). This result indicates that these questions are not discriminative, compromising the assessment of the reliability of the domains, and were, therefore, discarded.

An Exploratory Factor Analysis (EFA) was carried out with the remaining 28 items, which indicated the need to remove 6 more items that showed low rotational factor loading (PROMIN method). Therefore, a second exploratory factor analysis was carried out with the remaining 22 items, which indicated the need to remove one more item due to low factor loading. A new exploratory factor analysis was performed and the remaining 21 items showed satisfactory factor loadings. A Parallel Analysis was then carried out, which suggested the presence of two factors (latent variables) that were called: (1) empathic understanding - which involves a cognitive-affective bond and (2) empathic action - which involves a cognitive-behavioral component. (Table 2). It is important to highlight that the Unidimensionality indicators “Unidimensional Congruence (UniCo = 0.867)”, “Explained Common Variance (ECV = 0.774)” and Mean of Item Residual Absolute Loadings (MIREAL = 0.267)²⁷ did not support the unidimensionality of the scale. The suitability parameters of this model are shown in Table 3.

Table 2 Factor structure (Factor loadings) of the items that constitute the Brazilian Clinical Empathy Scale. 

Items	FACTOR 1	FACTOR 2
Item 5	-0.044	0.343
Item 11i	0.129	0.648
Item 12	0.460	-0.022
Item 14	0.748	0.018
Item 15	0.253	0.448
Item 16i	-0.041	0.507
Item 17	0.175	0.440
Item 18	0.654	0.206
Item 19	0.951	-0.321
Item 20i	0.197	0.450
Item 22i	-0.276	0.545
Item 23	0.891	-0.233
Item 24i	0.026	0.525
Item 25i	-0.269	0.926
Item 27i	0.201	0.462
Item 28i	-0.057	0.626
Item 30	0.402	0.275
Item 31	0.136	0.423
Item 32i	-0.230	0.347
Item 33	0.714	-0.043
Item 34i	0.014	0.607
Composite Reliability	0.871	0.843
H - Latent	0.902	0.880
H - Observed	0.891	0.865

Source: Study data.

Table 3 Evaluation of factor analysis parameters for the final model of the clinical empathy scale. 

Suitability parameters	Ideal values	Values observed in the study
KMO (Total MSA)	≥ 0.60	0.59
Bartlett sphericity test (p)	< 0.05	< 0.0001
χ2 / gl	< 3	1.5
RMSEA	≥ 0.50	0.048
Composite reliability	≥ 0.50	≥ 0.843
TLI	> 0.95	0.967
UniCo	< 0.95	0.867
ECV	< 0.85	0.774
MIREAL	> 0.30	0.267
Total Variance Explained	≥ 50%	44.95%
Factorial load of the questions	> 0.35	0.343 to 0.951

Source: Study data.

The final version of the proposed scale consists of 21 items distributed across two domains (Chart 1). The Empathic Understanding domain consists of items 3, 4, 8, 9, 12, 17, 20 and the Empathic Action domain consists of items 1*, 2*, 5, 6*, 7, 10, 11*, 13* , 14*, 15, 16*, 18, 19, 21*. The sum of the scores obtained in each item must be considered to analyze empathy. The questions marked with an asterisk (*) had the scale direction reversed. There is no cutoff value, but the higher the score, the greater the student’s empathetic attitude.

Chart 1 Brazilian Clinical Empathy Scale. 

INSTRUCTIONS This questionnaire contains 21 statements related to the context of clinical care. Please indicate your level of agreement or disagreement for each of the statements by marking the appropriate number on a 5-point scale, where the higher number indicates the highest level of agreement.	I fully disagree	I partially disagree	I neither agree nor disagree	I partially agree	I fully agree
The patient’s feelings about their clinical condition should not interfere with the therapeutic decision.
The patient’s personal problems unrelated to their health status should not be considered.
To understand the patient, I must imagine how I would feel if I were in the same situation as them.
I often become emotionally involved with the patient’s story.
Reflecting on the patient’s desires and needs at the time of the consultation helps in the therapeutic conduct.
The formal record of the anamnesis is more important than eye contact.
The patient’s life history is as important as their disease.
Showing concern for the patient’s feelings is an important component of the consultation.
I often feel distress when the patient has a serious illness.
The patient’s emotions should not interfere with the recording of the anamnesis.
I must always indicate the best treatment, regardless of its financial impact on the patient’s life.
If the patient has a serious illness, I am worried even after the end of the consultation.
The patient’s personal beliefs and convictions should not interfere with the therapeutic conduct.
I believe that complaints of emotional origin should not be considered in the creation of the therapeutic plan.
I should only ask questions related to the patient’s health status in the consultation.
I should avoid talking about the patient’s family issues at the time of the consultation.
I believe that my assistance in the consultation is better when I consider the patient’s wishes.
The needs of the patient’s family members should be considered in the care plan.
I must consider the patient’s social context as secondary to their health problems.
Reflecting on my feelings when I know the patient’s story increases my desire to help them.
The patient’s family relationships should not interfere with the care plan.

Source: Study data

DISCUSSION

The results of the study showed that EBEC has a model comprising two factors that were called empathic understanding (cognitive-affective) and empathic action (cognitive-behavioral). In the EMPATHIC UNDERSTANDING domain, the items are related to perspective taking, conceptualized as the health professional’s capacity to understand what the patient experiences, thinks and feels from their perspective¹⁹, as well as emotional sharing, which is the capacity to understand and share the emotional states of others (mirror neurons).

The EMPATHIC ACTION domain covers items related to perspective-taking, which allows predicting behaviors and empathic concern, which is related to actions to be taken to solve the encountered problems⁷.

In the proposed scale, it was not possible to highlight the three components of the construct reported in the literature, but rather two factors. Given these results, we can infer that the concepts of the components of empathy are closely linked to each other, which makes their discrimination difficult. Paro et al.³⁰, when validating Jefferson’s empathy scale, found a change in the domain of some items, which was justified by the hypothesis of a different view of Brazilian students in relation to the construct.

According to Davis³¹, empathy is related to a series of factors that come into play whenever there is someone’s emotional experience, therefore proposing an integrated approach that identifies the joint role of cognition and affection. Thus, sharing emotions (affective component) without perspective taking and the regulatory processes involved (cognitive component), takes the form of emotional contagion or sympathy. Similarly, accurately perceiving someone’s thoughts and feelings, without experiencing compassion and interest in their well-being, does not translate into an empathetic manifestation³². According to Sampaio et al.², the relationship between affectivity and cognition is important for the internalization and construction of principles that govern the individual’s behavior in society.

The scale developed in this study showed satisfactory Composite Reliability measures and replicability estimates (H > 0.80), that is, the generated factors are replicable and appropriate according to the factorial load measurements found. According to Rogers³³, the observed H index indicates how much the set of items represents the common factor. Its value ranges from zero to one and values ​​above 0.80 suggest a good definition of the latent variable, which will potentially be more stable in future studies. The latent H index reflects the estimated replicability when the items are interpreted as continuous variables and the observed H index reflects the estimated replicability when the items are interpreted as ordinal variables, such as Likert-type measures³³.

It was observed that the female sex showed higher scores in the two factors of the scale. According to Batchelder et al.³⁴, the female advantage is more evident in the affective components, including affective reactivity, and less evident in the cognitive components, revealing that women are more naturally in tune with emotional states and are more likely to react and respond to emotions and feelings of others.

Students who intend to pursue the clinical specialty, those who have experienced serious illness in the family or who have a chronic illness had higher scores in the Empathic Understanding component of the scale (cognitive-affective bond). Bailey³⁵ showed that medical students who intended to work in a clinical area had a higher empathy score than those who planned to work in surgical areas. Studies carried out with doctors showed the same pattern as that observed in this study carried out with medical students. In a study carried out with 704 doctors, using the Jefferson Scale of Empathy (JSE), those involved in the clinical area obtained higher scores than those in the surgical and imaging areas³⁶. It was also observed by Batenburg et al.³⁷ that final year interns who prefer general practice as a specialty showed more empathetic attitudes than interns in the surgical area.

Another important point is the relationship between empathy levels and the presence of one’s own illness or that of family members. A study conducted by Esquerda et al.³⁸ with 191 medical students revealed greater empathy among those who had already experienced illness among family members, friends or personally. These data are in line with our study, which suggests experience as a relevant factor for the development of empathic ability.

These results reveal a positive point of this study by demonstrating that the proposed empathy scale was sensitive in detecting, as mentioned in the literature, the influence of some sociodemographic variables on empathy, which could suggest an external validity of the instrument. The intention to promote comparison with external validity criteria is highlighted, aiming to test items that were not consistent in the studied sample, as well as the discriminatory power of the scale.

Thus, we can infer that the new empathy scale developed in this study showed preliminary evidence of validity, having the distinction of differentiating the cognitive-affective (Empathic Understanding) and cognitive-behavioral (Empathic Action) components of the construct, allowing the identification of factors that affect each component.

It is worth noting that the EBEC, like several instruments used to assess clinical empathy, is a self-report scale, which may have limited conclusions as it depends on self-knowledge and presents the possibility of response bias, with the respondent potentially tending to socially accepted answers. Some studies have identified the lack of correlation between empathy levels obtained by self-assessment instruments and patients’ perceptions and suggest that the patients should be included in the empathy assessment process³⁹.

More studies should be carried out using the scale proposed in this study to demonstrate its reliability, validity and stability.

FINAL CONSIDERATIONS

The model proposed for the Brazilian Clinical Empathy Scale, consisting of 21 items distributed into two factors: Empathic Understanding and Empathic Action, met the criteria of semantic and cultural adequacy, in addition to demonstrating preliminary evidence of validity. Using this instrument, it was possible to identify that the female sex, choosing a clinical specialty, having a chronic illness or having cases of serious illness in the family are predictive factors for empathy.