Empathy Levels Among Medical Students at a Romanian University: A Cross-Sectional Study Using the Jefferson Scale of Empathy

Abstract

Background/Objectives: Empathy is a fundamental component of patient-centered care and plays a critical role in the physician–patient relationship. This study aimed to evaluate empathy levels among undergraduate medical students at a medical university in Romania, examined from a Medical Humanities perspective, and to analyze how these scores are distributed according to gender, academic year, clinical stage, and direct experience with patient contact. Methods: A descriptive, cross-sectional study was conducted among 209 undergraduate medical students across all six academic years. Empathy levels were assessed using the Romanian-language version of the Jefferson Scale of Empathy—Student Version (JSE-S). Differences in empathy scores were analyzed using independent samples t-tests and a one-way analysis of variance (ANOVA) with Tukey Honestly Significant Difference (HSD) post hoc comparisons. Statistical significance was set at an alpha level of less than 0.05. Results: Internal consistency of the JSE-S total scale was good (Cronbach’s α = 0.829). The overall mean total JSE-S score was 107.34 ± 14.34. Female students demonstrated significantly higher empathy scores compared to male students (109.92 ± 13.10 versus 99.37 ± 15.20; t = 4.450; p < 0.001; Cohen’s d = 0.77). No statistically significant differences were found between preclinical and clinical stage students (106.84 ± 14.75 versus 108.98 ± 12.90; p = 0.363; Cohen’s d = −0.15), between students with and without direct patient contact (108.20 ± 14.14 versus 104.48 ± 14.78; p = 0.115; Cohen’s d = 0.26), or across the six academic years (F = 1.614; p = 0.158; partial η² = 0.038). Conclusions: Medical students demonstrated moderate-to-high empathy levels, with gender representing the only statistically significant demographic factor. The absence of major differences across academic-year cohorts and clinical exposure suggests that empathy levels were comparably represented across the different stages of training within this cross-sectional sample. These findings contribute descriptive data on empathy levels among Romanian medical students and highlight gender as the main factor associated with variation in empathy scores in this sample.

1. Introduction

Empathy is defined as the ability to interpret, understand, and respond appropriately to another person’s cognitive and emotional experiences, while distinguishing between one’s own feelings and those of others [1]. Known as a complex concept that incorporates cognitive and affective dimensions, empathy is important in human relationships and plays a significant role in healthcare environments [1,2].

Empathy in the clinical setting is a key element of the physician–patient relationship and an essential part of patient-centered care. Understanding patients’ emotions, as well as their psychological and social experiences, is facilitated by empathic communication, resulting in higher-quality and effective care. Research shows that higher levels of physician empathy lead to greater patient satisfaction, improved clinical outcomes, and better compliance with treatment [2,3].

Within medical education, empathy is conceptualized as a dynamic, trainable capacity rather than a fixed personality trait. Being continuously exposed to academic stress and escalating professional responsibilities, medical students may develop changes in their empathic attitudes over time, a pattern well documented in the existing literature. A significant decline in empathy, attributed to emotional exhaustion and increased workload, has been reported in some studies, particularly during the transition from preclinical to clinical years [2,4]. By contrast, other authors suggest that empathy can be preserved or even enhanced throughout academic training via reflective learning, communication-oriented curricula, and supportive educational environments [4,5]. These contradictory findings suggest that the development of empathy is complex and highlight the importance of contextual and pedagogical factors.

In recent years, Medical Humanities has become a relevant field within medical education, integrating literature, ethics, and reflective practices into the curriculum. Medical Humanities aims to complement scientific and technical training while fostering empathy and communication skills in future physicians. This integrative approach can facilitate the cultivation of humanistic competencies alongside clinical expertise, while preparing students to become better physicians and make a greater impact on patient care [6,7].

Despite the growing recognition of Medical Humanities internationally, the existing literature on empathy development during medical training remains limited, particularly across diverse educational and sociocultural contexts. Romanian medical education research on empathy remains insufficiently developed, with few studies examining potential variation across academic years and levels of clinical exposure [8,9].

The aim of this study was to evaluate empathy levels among medical students and to analyze their distribution according to academic years, demographic characteristics, and clinical exposure. The results of the study contribute to the existing literature on empathy in medical education and may inform the design of empathy-enhancing educational interventions.

2. Materials and Methods

2.1. Study Design

The study was designed as a descriptive, cross-sectional study with a quantitative approach. Data were collected through a self-administered, anonymous online questionnaire distributed to undergraduate medical students across all academic years, examining empathy scores in relation to demographic and academic variables.

2.2. Participants and Eligibility Criteria

The study population included students enrolled at the George Emil Palade University of Medicine, Pharmacy, Science and Technology of Târgu Mureș, Romania. Students from all six academic years were eligible to participate. Inclusion criteria were active enrollment in the undergraduate medical program and voluntary consent to participate. No exclusion criteria based on gender, age, or academic performance were applied. Participation was voluntary and uncompensated. The final sample comprised 209 respondents.

2.3. Recruitment and Data Collection

The questionnaire was distributed online through official student communication channels during the study period. Prior to accessing the questionnaire, all participants were informed of the academic purpose of the study, its title, the anonymous nature of participation, and the confidential handling of collected data. No personally identifiable information was collected. All data were used exclusively for academic and scientific purposes, following ethical principles for anonymous research. The response collection lasted five weeks, from 1 April to 6 May 2026.

2.4. Instruments

The questionnaire was structured into two sections. The first section collected demographic and academic variables, including gender, age, academic year, and self-reported patient interaction experience—variables considered relevant to the analysis of empathy levels across different stages of medical training. Academic stage was divided into two categories, with years I–III classified as the preclinical stage and years IV–VI as the clinical stage.

The second section consisted of the Romanian-language version of the Jefferson Scale of Empathy—Student Version (JSE-S), one of the most widely used and validated instruments for assessing empathy in medical education [10,11]. The JSE-S was developed at Thomas Jefferson University and was used in the present study with institutional permission, including the Romanian-language adaptation validated for use in this study. The JSE was used in this study with permission from Thomas Jefferson University. Written permission for the use of the Romanian-language version of the JSE-S was obtained from the scale developers prior to data collection. The instrument comprises 20 items rated on a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Ten negatively phrased items require reverse scoring in accordance with the original scoring methodology. Total scores range from 20 to 140, such that higher total scores indicate greater levels of empathy. As presented in Table 1, the items are organized into three subscales reflecting distinct dimensions of clinical empathy: Perspective Taking (10 items), which represents the cognitive ability to understand the patient’s viewpoint and constitutes the main dimension of the scale; Compassionate Care (8 items), which covers the emotional and relational aspects of the physician–patient interaction; Walking in Patient’s Shoes (2 items), which captures the ability to adopt the patient’s perspective in a more immediate sense. This three-factor structure is consistent with the known factorial structure of the instrument as described in the psychometric literature and confirmed in multiple international validation studies.

Table 1. Items of the Jefferson Scale of Empathy—Student Version (JSE-S) organized by subscale.

No. Crt.	Order in the Questionnaire	Question Statement
Perspective Taking: understanding the patient’s viewpoint.
1	2	Patients feel better when their physicians understand their feelings.
2	4	Understanding body language is as important as verbal communication in physician patient relationship.
3	5	A physician’s sense of humor contributes to a better clinical outcome.
4	9	Physicians should try to stand in their patients’ shoes when providing care to them.
5	10	Patients value a physician’s understanding of their feelings which is therapeutic.
6	13	Physicians should try to understand what is going on in their patients’ mind by paying attention to their non-verbal cues and body language.
7	15	Empathy is a therapeutic skill without which the physician’s success is limited.
8	16	Physicians’ understanding of the emotional status of their patients, as well as their families is one important component of the physician–patient relationship.
9	17	Physicians should try to think like their patients in order to render better care.
10	20	I believe that empathy is an important therapeutic factor in medical treatment.
Compassionate Care: emotional and relational dimensions of care.
11	1	Physicians’ understanding of their patients’ feelings and the feelings of their patients’ families does not influence medical or surgical treatment.
12	7	Attention to patients’ emotions is not important in history taking.
13	8	Attentiveness to patients’ personal experiences does not influence treatment outcomes.
14	11	Patients’ illnesses can be cured only by medical or surgical treatment; therefore, physicians’ emotional ties with their patients do not have significant influence in medical or surgical treatment.
15	12	Asking patients about what is happening in their personal life is not helpful in understanding their physical complaints.
16	14	I believe that emotion has no place in the treatment of medical illness.
17	18	Physicians should not allow themselves to be influenced by strong personal bonds between their patients and their family members.
18	19	I do not enjoy reading non-medical literature or the arts.
Walking in Patient’s Shoes
19	3	It is difficult for a physician to view things from patients’ perspectives.
20	6	Because people are different, it is difficult to see things from patients’ perspectives.

2.5. Statistical Analysis

Collected data were organized in table format using Microsoft Excel (Microsoft 365 MSO, Version 2607, Build 16.0.20228.20124; Microsoft Corporation, Redmond, WA, USA) and then analyzed using IBM Statistical Package for the Social Sciences (SPSS) Statistics (Version 26.0.0.0; International Business Machines (IBM) Corporation, Armonk, NY, USA). Participants’ characteristics and empathy scores were described using descriptive statistics. Categorical variables were expressed as frequencies and percentages, while continuous variables were reported as mean values with standard deviations. Differences in empathy scores according to gender, academic stage (preclinical vs. clinical years), and patient contact (yes vs. no) were assessed using the independent samples t-test, with Levene’s test applied to evaluate the homogeneity of variances. When the assumption of equal variances was violated, Welch’s t-test was applied and reported. Differences in total JSE-S scores across the six academic years were examined using one-way analysis of variance (ANOVA), with the Tukey HSD post hoc test applied for pairwise comparisons. Statistical significance was established at a 95% confidence level, with p-values below 0.05 considered statistically significant. Internal consistency reliability of the JSE-S total scale was assessed using Cronbach’s alpha. Effect sizes were calculated for all group comparisons, using Cohen’s d for independent samples t-tests and partial eta squared (η²p) for the one-way ANOVA.

2.6. Ethical Considerations

The study was conducted in accordance with the Declaration of Helsinki and with ethical principles for research involving human participants. Ethical approval was granted by the Ethics Committee for Scientific Research of the George Emil Palade University of Medicine, Pharmacy, Science and Technology of Târgu Mureș (Decision No. 4138, dated 30 March 2026). No personally identifiable information was collected at any stage. Given the cross-sectional design and voluntary participation, findings reflect a single point in time and selection bias cannot be excluded.

2.7. Use of Artificial Intelligence Tools

During the preparation of this manuscript, the authors used Claude Sonnet 5 (Anthropic) to assist with language editing and structural consistency of the text. Artificial intelligence tools were not involved in study design, data collection, statistical analysis, or interpretation of results. All AI-assisted content was reviewed and validated by the authors, who take full responsibility for the published work.

3. Results

3.1. Sociodemographic Characteristics of the Study Sample

A total of 209 undergraduate medical students completed the questionnaire. The sociodemographic characteristics of the participants, distributed by gender, year of study, age group, and clinical experience with patients, are presented in Table 2.

Table 2. Sociodemographic data of the study participants.

Parameter	Respondents (n)	Respondents (%)
Gender
Female	158	75.6%
Male	51	24.4%
Year of study
Year I	65	31.1%
Year II	43	20.6%
Year III	52	24.9%
Year IV	20	9.6%
Year V	14	6.7%
Year VI	15	7.2%
Age groups
18–20 years old	72	34.4%
21–23 years old	113	54.1%
24–26 years old	18	8.6%
27–35 years old	6	2.9%

Most respondents were female (n = 158, 75.6%), while male participants accounted for 24.4% of the sample (n = 51).

Regarding the distribution by academic year, the largest group was represented by first-year students (n = 65, 31.1%), followed by third-year-of-study (n = 52, 24.9%) and second-year students (n = 43, 20.6%). Clinical-year students were less represented, with 20 participants in the fourth year of study (9.6%), 14 in the fifth year of study (6.7%), and 15 in the sixth year of study (7.2%).

In terms of age, most participants were between 21 and 23 years old (n = 113, 54.1%), followed by the 18–20 age group (n = 72, 34.4%). Older age groups were less represented, with 18 participants aged 24–26 years (8.6%) and 6 participants aged 27–35 years (2.9%).

The distribution of gender across the six academic years is presented in Table 3. The proportion of female students ranged from 69.2% in first year (n = 45/65) to 92.9% in fifth year (n = 13/14), with the male proportion correspondingly highest in first year (n = 20/65, 30.8%) and lowest in fifth year (n = 1/14, 7.1%). A chi-square test of independence showed no significant association between gender and year of study (χ²(5) = 5.355, p = 0.374). As three cells had expected counts below 5, Fisher’s exact test was also computed, confirming the result (p = 0.403).

Table 3. Distribution of gender across academic years.

Academic Year	Female (n)	Female (%)	Male (n)	Male (%)
Year I	45	69.2%	20	30.8%
Year II	32	74.4%	11	25.6%
Year III	41	78.8%	11	21.2%
Year IV	14	70.0%	6	30.0%
Year V	13	92.9%	1	7.1%
Year VI	13	86.7%	2	13.3%

3.2. Empathy Scores on the Jefferson Scale of Empathy—Student Version

The participants’ responses to the Jefferson Scale of Empathy—Student Version (JSE-S) are presented in Table 4, expressed as mean scores with standard deviations for each individual item and for each subscale. Internal consistency of the JSE-S total scale in this sample was good, with a Cronbach’s alpha of 0.829 (20 items). The overall mean total JSE-S score across all participants was 107.34 ± 14.34. The highest mean score at item level was recorded for item 2, belonging to the Perspective Taking subscale (Mean = 6.28 ± 0.99), while the lowest mean score was observed for item 18, belonging to the Compassionate Care subscale (Mean = 3.28 ± 1.61). At subscale level, the Perspective Taking subscale recorded the highest mean score (Mean = 56.53 ± 8.49), followed by Compassionate Care (Mean = 40.48 ± 8.87), and Walking in Patient’s Shoes (Mean = 10.33 ± 2.76).

Table 4. Mean score and standard deviation for each item of the Jefferson Scale of Empathy—Student Version (JSE-S), organized by subscale.

No. Questions	Mean ± SD
	Perspective Taking	56.53 ± 8.49
2		6.28 ± 0.99
4		6.02 ± 1.12
5		5.07 ± 1.26
9		5.32 ± 1.38
10		5.73 ± 1.22
13		5.61 ± 1.30
15		5.58 ± 1.46
16		5.52 ± 1.35
17		5.11 ± 1.49
20		6.16 ± 1.13
	Compassionate Care	40.48 ± 8.87
1		4.58 ± 2.04
7		5.46 ± 1.64
8		5.11 ± 1.70
11		5.36 ± 1.54
12		5.31 ± 1.64
14		5.60 ± 1.65
18		3.28 ± 1.61
19		5.54 ± 1.89
	Walking in Patient’s Shoes	10.33 ± 2.76
3		5.16 ± 1.48
6		5.16 ± 1.57
Total JSE-S		107.34 ± 14.34

3.3. Empathy Score Comparisons Across Participant Groups

The comparison of JSE-S total scores across the major participant groups is presented in Table 5. A statistically significant difference in empathy scores was identified between female and male students (109.92 ± 13.10 vs. 99.37 ± 15.20; t = 4.450; df = 75.46; p < 0.001; Cohen’s d = 0.77), with female participants recording a higher mean score. No statistically significant differences were observed between preclinical and clinical stage students (106.84 ± 14.75 vs. 108.98 ± 12.90; t = −0.912; df = 207; p = 0.363; Cohen’s d = −0.15) or between students with and without direct patient contact (108.20 ± 14.14 vs. 104.48 ± 14.78; t = 1.583; df = 207; p = 0.115; Cohen’s d = 0.26), both reflecting small effect sizes.

Table 5. Comparison of Jefferson Scale of Empathy—Student Version (JSE-S) scores across major participant groups.

Variable	Group	Participants	Mean ± SD	p-Value	Cohen’s d
Sex	Female	158	109.92 ± 13.10	<0.001	0.77
	Male	51	99.37 ± 15.20
Stage of training	Preclinical	160	106.84 ± 14.75	0.363	−0.15
	Clinical	49	108.98 ± 12.90
Patient contact	Yes	161	108.20 ± 14.14	0.115	0.26
	No	48	104.48 ± 14.78

3.4. Empathy Scores Distribution by Academic Year

The distribution of JSE-S total scores across the six academic years is presented in Table 6. Mean empathy scores were higher in the Year VI cohort (111.93 ± 8.80) compared to the Year I cohort (104.14 ± 15.18), with a lower mean observed in the Year IV cohort (105.50 ± 16.06). One-way ANOVA revealed no statistically significant differences in empathy scores across academic years, F (5, 203) = 1.614, p = 0.158, partial η² = 0.038, indicating a small effect size.

Table 6. JSE-S mean score and standard deviation by academic year.

Academic Year	Mean ±SD	p-Value
Year I	104.14 ± 15.18	0.158
Year II	106.93 ± 16.89
Year III	110.15 ± 11.57
Year IV	105.50 ± 16.06
Year V	110.79 ± 11.06
Year VI	111.93 ± 8.80

4. Discussion

This study aimed to evaluate empathy levels among undergraduate medical students, at George Emil Palade University of Medicine, Pharmacy, Science and Technology of Târgu Mureș, Romania, using the Jefferson Scale of Empathy—Student Version (JSE-S).

Empathy plays an important role in clinical practice, influencing how physicians relate to their patients, how satisfied patients feel, whether they follow treatment recommendations, and the overall outcomes of medical care [1,2,3]. Understanding how empathy develops—and which factors maintain or diminish it—throughout undergraduate medical training is therefore of direct relevance to curriculum design and to the overall aim of educating physicians with strong humanistic skills [4,5]. The findings of the present study are discussed below in relation to the four main areas of analysis: overall empathy scores, gender differences, clinical stage and patient contact, and distribution across academic years.

4.1. Overall Empathy Scores

The present study found a mean total JSE-S score placing the cohort above the theoretical midpoint of the scale and reflecting a moderate-to-high empathic orientation toward patient care (Table 4).

The individual score distribution shows clear variation among students, consistent with the idea that empathy is not a uniform trait even within the same medical school and training program. Instead, it is a dynamic construct shaped by personal, contextual, and experiential factors [1,2]. These differences align well with national norms and the psychometric properties of the JSE-S established in large, multi-institutional studies [10].

When looking at the three subscales of the JSE-S, the Perspective Taking dimension recorded the highest mean score, which is expected given its central role in the instrument (Table 4). This subscale measures the ability to understand a patient’s thoughts, feelings, and experiences from their point of view. Because it contains the largest number of questions, it serves as the dominant component of the total score. Compassionate Care recorded the second highest mean score, while Walking in Patient’s Shoes had the lowest subscale average (Table 4). This pattern, where cognitive empathy takes precedence over its emotional and relational counterparts, may be consistent with the highly academic nature of undergraduate medical training. Medical school typically trains students to analyze patients’ experiences logically rather than systematically developing their emotional responsiveness [5]. This three-factor structure and the relative importance of each subscale match cross-national evaluations of the JSE-S, confirming that the structure of the test remains stable across different educational and cultural settings [11].

At the individual question level, the highest score was achieved for item 2—“Patients feel better when their physicians understand their feelings” (Table 4). The near-universal agreement here shows that even at this stage of training, students firmly believe in the therapeutic value of emotional understanding, which is the cornerstone of patient-centered medicine [3]. By contrast, the lowest score was recorded for item 18—“Physicians should not allow themselves to be influenced by strong personal bonds between their patients and their family members” (Table 4). Because this is a negatively phrased item that requires reverse scoring, the low value indicates that students largely disagreed with the idea of excluding emotional bonds from clinical practice. Rather than showing a lack of professional boundaries, this suggests students intuitively recognize that emotional engagement does not compromise clinical judgment, but actually enhances it— an idea central to Medical Humanities education [6,7].

Placed within the broader international literature, the average total score of this cohort is comparable to recent findings using the same tool. A study across four medical universities in Denmark reported a mean JSE-S score slightly higher than the one found in this study [12], while a study among Saudi medical students reported a somewhat lower average [8]. Comparative data collected over 11 years at Sidney Kimmel Medical College showed higher average scores for both female and male students [13]. On the other hand, a Spanish study found a mean score that closely matches our current findings [14]. This gap may reflect differences in educational, cultural or sampling characteristics. Overall, these comparisons place this Romanian cohort within the moderate-to-high range internationally, while highlighting how much local context and sociocultural factors shape empathy scores [2,3,9].

4.2. Gender Differences in Empathy Scores

The most significant finding of this study was the clear difference in empathy scores between male and female students (Table 5), with female students scoring substantially higher overall. The size of this gap is notable, suggesting that gender is a meaningful factor when it comes to empathy levels within this group.

This result aligns closely with international research on gender and empathy in medical education. For instance, a study using the JSE-S among Greek medical students showed that women scored significantly higher than men, a pattern that directly mirrors our findings [15]. Similarly, a Danish study across four medical schools found that female students had higher empathy scores, a difference that remained true even after adjusting for other variables in a more complex analysis [12]. Furthermore, data collected over 11 years at Sidney Kimmel Medical College confirmed that women consistently outscored men across most incoming classes [13]. An Italian study focusing on the specific subscales of the test also found that the biggest gaps between men and women occurred in the Compassionate Care and Walking in Patient’s Shoes sections, while their scores for general perspective-taking were much closer, a pattern that matches the subscale results in our own study [16].

Researchers have suggested several reasons for these consistent differences between genders. One common explanation points to how men and women are socialized, noting that cultural norms often encourage women to develop and express emotional sensitivity more than men [3,15]. Another study tracking empathy longitudinally found that gender differences in empathy remain stable over time, with research suggesting these differences may reflect social or biological factors [17]. This idea is supported by the fact that the gender gap is usually wider on questions dealing with emotional empathy rather than logical, cognitive empathy, suggesting that social norms might lean into how students self-report their feelings [17]. Regardless of the exact cause, the fact that this gender difference appears so consistently across different cultures and schools shows how important it is for medical education. Developing empathy training programs that recognize these differences and help all students strengthen their skills may represent an area worth exploring in curriculum design [4,7].

The predominance of female respondents in this sample is consistent with demographic patterns increasingly reported among medical student cohorts, both in Romania and internationally, where female enrollment has come to represent a substantial majority in many institutions [18,19,20,21]. Such proportions are consistent with broader shifts in the gender composition of medical education rather than a sampling artifact specific to this study. A further consideration relates to the distribution of participants across academic years, which was itself uneven (Table 2) and accompanied by variation in gender representation from one year to another (Table 3). Since training stage and gender are not fully independent in this dataset, differences attributed to academic year should be interpreted with some caution, as they may partly reflect the underlying gender composition of each cohort rather than training stage alone. The exact number of students who received or had access to the questionnaire in each academic year was not centrally tracked, which limits the ability to characterize the representativeness of the responding sample with precision. Given the pronounced gender imbalance in this sample, the possibility that the female-skewed distribution contributed to a higher overall mean empathy score also warrants consideration; however, the direction of the gender effect observed here—with female students scoring higher than male students—is itself consistent with the broader international literature, independent of this sample’s particular composition.

4.3. Comparison of Empathy Scores Across Participant Groups

Comparing overall JSE-S scores across the main participant groups showed a statistically significant difference only for gender. On the other hand, a student’s clinical stage and their direct contact with patients did not emerge as significant factors driving empathy within this group (Table 5).

Preclinical and clinical students displayed comparable empathy levels, as did those with and without direct patient contact (Table 5). The fact that empathy scores did not differ meaningfully between these groups suggests that, within this sample, empathetic attitudes toward patient care were not strongly associated with training stage or with prior patient contact as measured in this study. This pattern aligns well with international research. For example, a cross-sectional study at Leicester Medical School found no statistically significant difference in empathy scores between preclinical and clinical phase students [22]. Regarding patient contact, a study examining empathy levels before and after early clinical exposure found that the mere fact of interacting with patients matters less than the quality and educational framing of that interaction—including reflective supervision and communication-focused teaching [23]. This is consistent with our current results, suggesting that structured, humanistically oriented training may be a more decisive factor in shaping empathy than clinical exposure alone [5,7]. This interpretation is further supported by a recent Romanian study among practicing pathologists, which found that empathy remains professionally relevant even in specialties without direct patient contact, underscoring that empathic engagement need not depend on face-to-face clinical interaction [24]. A plausible contributing factor to the comparable empathy levels seen between preclinical and clinical cohorts may be the curriculum’s ongoing integration of humanistic content across all six years of study [7]. This remains, at present, an interpretive possibility rather than an established finding, as the present analysis was not designed to capture individual-level curricular engagement or to compare outcomes against a cohort trained under a different pedagogical model. Establishing this relationship with confidence would require a study specifically designed to test it directly.

4.4. Empathy Scores by Academic Year

The distribution of overall JSE-S scores across the six academic years showed higher mean scores in later cohorts compared to earlier ones, with a lower mean observed in Year IV (Table 6). While the statistical tests, including a one-way ANOVA and post hoc Tukey HSD comparisons, did not show statistically significant differences between the individual years, mean empathy scores varied numerically across cohorts. As this is a cross-sectional comparison of six distinct student cohorts rather than a longitudinal follow-up of the same individuals, the pattern observed cannot be interpreted as evidence of individual developmental change over time; it may instead reflect cohort-level differences. Although the present findings cannot be interpreted longitudinally, studies using longitudinal designs have reported changes in empathy over time. For instance, a longitudinal study following medical students over six years found a significant increase in total JSE-S scores from the first year to the final year, which the authors linked to a structured professionalism program featuring built-in empathy workshops during the clinical years [25]. Similarly, a German longitudinal study reported a small but statistically significant increase in empathy scores from the start to the end of medical school [26]. Furthermore, a study involving preclinical students taking an art-based humanities course found no significant drop in empathy scores as training progressed, attributing this stability to the steady integration of humanistic content throughout the curriculum [27]. A systematic review and meta-analysis examining the association between empathy and burnout found that higher empathy is associated with lower burnout across both preclinical and clinical years [28], suggesting that maintaining empathetic attitudes may also carry protective implications for student wellbeing. In particular, narrative medicine has shown potential to improve empathy levels and prevent emotional detachment among trainees [29].

The comparatively lower mean score observed in Year IV is particularly noteworthy, as this cohort corresponds to the transition from preclinical to clinical training. The literature frequently describes this period as emotionally demanding, marked by heavier workloads, exposure to difficult clinical realities, and a natural tendency for students to emotionally protect themselves [2,4,22]. A systematic review looking at how empathy fluctuates during medical school identified heavy workloads, emotional exhaustion, and the “hidden curriculum” as recurring factors linked to shifts in empathy during clinical years [30]. Additionally, a meta-analysis involving large numbers of medical students confirmed a strong negative link between empathy and burnout [28], suggesting that this pattern in Year IV might partly reflect a heightened vulnerability to burnout during this transitional stage of training. Rather than indicating an erosion of empathetic attitudes at the individual level, this cross-sectional difference may instead reflect the particular circumstances of this cohort at the time of data collection [30]. Notably, the Year V and Year VI cohorts showed higher mean scores, with the highest values observed among final-year students. This pattern may reflect characteristics of these cohorts, although longitudinal research is required to determine whether it corresponds to changes occurring during medical training [7,25].

A key methodological point to consider is the limited sample size in the clinical years (Table 2), which likely reduced the statistical power of the ANOVA test. To further examine this pattern within the context of Romanian medical education, future studies will need to look at larger and more balanced groups of students across all academic years.

4.5. Implications for Medical Education and the Romanian Context

The findings of this study offer several important insights for medical education, both in Romania and internationally. The moderate-to-high empathy levels observed across all academic-year cohorts, along with the absence of a significant difference between preclinical and clinical-stage students, suggest that empathetic attitudes were comparably represented across the different stages of training in this sample. This is an encouraging finding, especially since empathy erosion during medical education is widely reported and remains a major concern for medical educators worldwide [2,3,4].

Medical Humanities, as an educational approach, emphasizes reflective thinking, ethical reasoning, and humanistic engagement alongside scientific and clinical training, and has been proposed as a means of creating an environment where empathy can grow rather than diminish [6,7]. A systematic review and meta-analysis found that Medical Humanities interventions were associated with measurable improvements in empathy outcomes for both medical students and healthcare professionals [7]. In particular, narrative medicine has been highlighted as a specific component of Medical Humanities, with evidence suggesting it can improve empathy levels and prevent emotional detachment among trainees through a high-impact, low-cost teaching strategy [29].

Within the Romanian context specifically, research on medical student empathy is still quite limited. The validation of the Romanian version of the Toronto Empathy Questionnaire by Ursoniu is one of the few existing contributions to this field, making this study among the first to apply the JSE-S in a Romanian medical school setting [9]. While the lack of baseline Romanian data makes direct benchmarking difficult, the results suggest that students at George Emil Palade University display empathy levels that align well with European and international standards, which is a positive sign for the country’s medical education system.

The significant gender difference identified in this study also points to practical next steps. Because female students consistently score higher on empathy scales, it may be beneficial to design targeted educational strategies that help male students strengthen their empathetic skills. These strategies could include structured reflection exercises, communication skills training, and narrative medicine approaches—all areas where Medical Humanities education has shown promising results [6,7,29].

Finally, the pattern of higher mean scores among later-year cohorts compared to earlier ones is compatible with, though does not confirm, the possibility of empathy differences between academic-year cohorts. Genuine longitudinal studies, following the same students across academic years with larger and more balanced samples, will be needed to determine whether this cross-sectional pattern reflects true developmental change.

4.6. Limitations

Several limitations of this study warrant consideration. Given its cross-sectional design and reliance on voluntary, anonymous participation, the findings capture a single point in time, and some degree of self-selection cannot be entirely excluded. It follows that the differences observed across academic-year cohorts are best understood as comparisons between distinct groups of students rather than as evidence of change within individuals over time; confirming a genuine developmental trajectory would call for a longitudinal design able to follow the same students throughout their training.

A further consideration relates to the composition of the sample. As the number of students invited or eligible to participate in each academic year was not centrally recorded, precise response rates could not be established. The sample also showed some imbalance across academic years, with fewer participants represented in the later, clinical years, which may have limited the statistical power available for certain by-year comparisons. In addition, gender representation varied somewhat across academic years, meaning that gender and training stage were not entirely independent within the dataset; the between-year and preclinical–clinical findings are best interpreted with this in mind.

The measurement of patient contact also merits mention. This variable was captured through a single self-reported yes/no item, without a standardized definition of what constituted meaningful contact or any indication of its frequency, duration, or clinical context. Such a measure may not fully reflect the range of patient exposure experienced by participants, and the absence of an association between patient contact and empathy scores is best understood as pertaining to this specific self-reported measure, rather than to patient exposure more broadly.

Similarly, although the study was situated within a Medical Humanities framework, it did not assess individual-level exposure to, or engagement with, this component of the curriculum, nor did it include a comparison group trained under an alternative pedagogical model. Any interpretation linking the institution’s Medical Humanities framework to the empathy levels observed is therefore best regarded as a hypothesis meriting future investigation, rather than a conclusion drawn directly from the present data.

Finally, this study is among the first in Romania to use the JSE-S instrument, so there is no prior Romanian data to compare these results against directly. Future studies could address these limitations by using larger, more balanced samples across academic years, reporting response rates clearly, accounting for gender when comparing results by year, and using a longitudinal design to properly test how empathy changes over time.

5. Conclusions

Measuring empathy in medical students is far more than just an academic exercise; it reflects a deep commitment to understanding how future doctors connect with their patients and handle the human side of medicine. This study, conducted at George Emil Palade University of Medicine, Pharmacy, Science and Technology of Târgu Mureș, provides one of the first JSE-S-based assessments of empathy in Romanian undergraduate medical education, examined from a Medical Humanities perspective.

The overall picture that emerges from this data is one of relative strength across the sample. Students showed moderate-to-high empathy levels across all academic-year cohorts, with no statistically significant differences observed between clinical stages or by patient contact. This finding contrasts with the empathy decline frequently reported in some longitudinal studies internationally. However, the present study was not designed to identify the curricular or educational factors underlying this pattern. A comparatively lower mean score was observed in the Year IV cohort, coinciding with the transition to clinical training, while the Year V and VI cohorts showed higher mean scores, with the highest values recorded among final-year students. This pattern is compatible with, but does not confirm, the possibility of higher empathic orientation among later-year cohorts. Gender remained the only factor with a significant difference, as female students consistently scored higher than their male peers—a well-established global pattern that warrants continued attention in curriculum design.

Looking at individual questions, the analysis revealed that students firmly believe in the therapeutic value of emotional understanding. However, they showed more mixed feelings regarding the boundaries of professional emotional involvement. This tension sits at the very heart of Medical Humanities education and highlights the need for structured guidance within the curriculum.

Taken together, these findings suggest that empathy levels can remain comparably strong across different stages of medical training, although this cross-sectional design cannot establish causality or individual-level change. They also underscore the clear need for future longitudinal, multi-institution research in Romania to track how empathy develops across different programs and to determine the specific contribution of Medical Humanities to this process.