Empathy is a foundational element of high-quality healthcare, associated with better patient outcomes, improved satisfaction, and greater provider wellbeing (1). Yet many healthcare professionals struggle to maintain empathy under pressure with nearly half experiencing “compassion fatigue” (2, 3). As a result, medical students often see supervisors whose actions do not consistently model empathy and compassion towards both patients and colleagues (4). While structured empathy training could address these challenges (1, 5–7), efforts to embed such training across healthcare curricula remain uneven and fragmented. This raises a key question: Why are healthcare education and refresher programs still slow to integrate Medical Humanities as a core requirement, despite its strong evidence base for fostering empathy and reflection?

One reason may be uncertainty around what aspects of Medical Humanities training are most impactful. Both teams-based service learning and reflective engagement with readings are commonly used components, but they differ in focus and demands. Service-learning emphasizes active engagement with communities, often confronting students with real-world dissonance that can trigger reflection and perspective-taking (8–11). When paired with written critical reflection, such experiences have been shown to reduce unconscious biases, strengthen empathy, and, thus, eventually improve the quality of patient care (12–18). These gains, however, may fade if reflection is not ongoing (19–22).

In contrast, reading and discussing humanities texts may foster empathy through exposure to others’ lived experiences and structured class dialogue, without requiring the logistical effort of community-based placements. But few studies have directly compared the influence of these two components within the same course or student population.

In our previous studies, we looked at whether improved empathy in aspiring medical students depended on taking part in service-learning, or if simply writing about the challenges of medical training, clinical practice, and maintaining humanism would be enough to foster higher empathy and compassion (11). We found that students’ Reflective Capacity (23) and Jefferson Scale of Empathy (24) scores increased when a Medical Humanities course required them to perform team service-learning and write critical reflections on those experiences (treatment group). Students who did not engage in service-learning (control group) did not experience these gains.

However, in our prior research (11), there were other important differences between our treatment and control groups. Specifically, our experimental and control groups were divided by whether the students attended school at our Colorado or Utah campuses. This meant variations in campus environments and personal factors influenced each student’s choice of location. In addition, the details of students’ reflection assignments differed because the focus of the assignments in Colorado were on service-learning while the Utah group focused more on reflection assignments tied to readings for their Humanities course. Moreover, there were differences in the reading assignments in Colorado and Utah, and the impact of the readings on students’ empathy and professional development have not been examined.

Due to these limitations in the between-groups design of our prior research (25, 26), in the present study we sought to re-examine the impact of required service-learning and readings for the Medical Humanities course on students’ empathy, professional development, and ability to perform critical reflection in a single population of prospective medical students (i.e., within-participants design). Building on our earlier findings that service-learning and reflective activities can enhance empathy and critical reflection (11, 16, 19, 20), we propose several hypotheses regarding the influence of both service-learning and required readings in a Medical Humanities course. We anticipate that a revised survey, which asks students to consider statements such as “owing to this course, I have changed the way I look at myself,” will prove to be a reliable and valid measure of critical reflection (CR).

We tested six hypotheses:

Descriptive statistics for each questionnaire are shown in Table 1. The CR survey, which asks students to consider statements such as “owing to this course, I have changed the way I look at myself,” proved to be a reliable and valid measure of critical reflection (CR). Cronbach’s alpha value for the CR survey was 0.76, thus establishing its reliability. And its validity is shown by survey results that are predicted by students’ attitudes toward required service-learning (SASL) and readings for a Medical Humanities course, as described and discussed below (Table 2).

The reliability and validity of the readings survey has also not been assessed previously. Here we found the survey to have a Cronbach’s alpha value of 0.86, thus supporting the conclusion that it is a reliable measure of students’ attitudes toward the readings. And its validity is shown by reading survey results that are associated with students’ CR and empathy (Table 2). Moreover, every student’s overall opinion of the readings was favorable (i.e., score above 4.0).

In our prior studies, we showed the SASL to be a reliable and valid measure of students’ attitudes toward performing team-based service-learning (33). Here, we found the Cronbach’s alpha value for SASL to be 0.80 for the present cohort of students. And every student’s overall SASL score was favorable (i.e., above 4.0). Finally, students’ scores on SASL were highly correlated with their CR scores (r = 0.63, Table 2), thus serving to verify Hypothesis 1.

In another study, we found the empathy survey to be a reliable and valid measure of students’ empathy for one another (27). For the present cohort of students, the empathy survey had a Cronbach’s alpha value of 0.89. Moreover, students’ SASL and empathy scores were highly correlated (r = 0.52, Table 2) also verifying Hypothesis 2.

Both the readings survey and the CR questionnaire were found to be reliable measures of the pertinent students’ attitudes. (See above.) Students’ scores for their attitudes toward these readings were correlated with their CR scores (0.28, Table 2), thus supporting Hypothesis 3.

Students’ empathy scores were moderately correlated with their scores on attitudes toward required readings (0.40, Table 2), which supports Hypothesis 4.

A multiple regression analysis was conducted to predict student perceived impact of the course on their critical reflection ability from their self-reported attitude toward service-learning (SASL) and appraisal of readings (READINGS). Together the predictors accounted for 40% of the variance in their perceived impact of the course on critical reflection [R² = 0.40, F(2,47) = 15.69, p < 0.001]. Only SASL (standardized beta = 0.67, p < 0.001) remained a significant predictor while READINGS did not (standardized beta = −0.08, p = 0.55). Results suggest that self-reported attitude towards service-learning is a more important predictor of student appraisal of course impact on their critical reflection than attitude towards reflective reading.

A multiple regression analysis was conducted to predict student self-reported empathy towards peers from their self-reported attitude toward service-learning (SASL) and appraisal of readings. Together the predictors accounted for 29% of the variance in their self-reported empathy towards peers [R² = 0.29, F(2,47) = 9.70, p < 0.001]. Once again, only SASL (standardized beta = 0.43, p < 0.001) remained a significant predictor while READINGS did not (standardized beta = 0.17, p = 0.25). Results suggest that self-reported attitude towards service-learning is a more important predictor of student self-reported empathy towards peers than attitude towards reflective reading.

Our findings support all six hypotheses and align with conclusions from earlier research (11, 16, 19, 20). When service-learning is incorporated into a Medical Humanities course, students report both professional growth and higher empathy (11). In the current study, students’ attitudes toward service-learning were positively associated with both their perception of how well the course fostered critical reflection (CR) and their empathy toward peers. Although CR and empathy were moderately correlated (r = 0.33, Table 2), this does not necessarily imply a direct cause-and-effect relationship.

In a prior study, we found that reflective capacity (RC) correlated with Jefferson Scale of Empathy scores (11), and experiences in the Medical Humanities course led to significant increases in both measures. However, there was no significant correlation between changes in RC and changes in empathy (11). Likewise, in the present study, the modest correlation between CR and empathy (r = 0.33) disappeared (r = 0.01) once attitudes toward service-learning (SASL) and reading scores were included in a multiple regression analysis.

Regression analyses on the data also support the conclusion that students’ positive attitudes towards service-learning offer a much stronger explanation of both their critical-reflection growth and empathy than their appraisal of reflective readings. In other words, although reflective reading may be valuable, students’ enthusiasm for and engagement in service-learning activities seems to be the key factor that shapes their perceptions of how much the course impacts their reflective abilities and empathy.

Importantly, this study used a within-participants design, avoiding some confounds that limited our previous research. All students completed the same course with the same structure, allowing us to directly compare the relative influence of two key components (service learning and reflective readings) within the same cohort. The design provides clearer evidence for the utility of both components, with a slight advantage for the service-learning component.

These results align with earlier studies comparing control and experimental groups, where only the experimental group participated in service-learning (11). While randomized controlled trials are ideal in principle, the main ethical concern is that some students might be denied a beneficial educational intervention. Given the existing body of evidence suggesting that reflective exercises and service-learning promote professional development and empathy (11–17, 19–22), withholding such activities from a control group could disadvantage certain students, potentially affecting their professional growth and patient care outcomes. Even a single 40-min empathy workshop has been shown to boost Jefferson Scale of Empathy scores in students, although the effect may be temporary (34). The central challenge is to maintain healthcare professionals’ compassion for patients and empathy for one another throughout their careers. Numerous programs have been developed to support this goal, and the present findings add to a growing body of research indicating that structured, longitudinal engagement through Medical Humanities courses may provide a durable foundation for empathy and reflective practice (35).

For example, a 6-week Medical Humanities module, which increased students’ Jefferson Scale of Empathy scores after only 16 h of instruction (36), and narrative medicine education, which includes extensive reflective writing, also improved academic performance and empathy among female nursing students, with gains persisting for at least two and a half years (37). However, other studies suggest that without a reflective writing component, such benefits may vary, particularly by gender or training stage (38). Our findings reinforce the view that structured service-learning, paired with critical reflection, is a key mechanism for fostering durable professional growth.

Considering all the findings discussed here, one pressing question remains: Why are healthcare training programs and refresher courses still slow to integrate Medical Humanities as a core requirement? A growing body of research shows that sustained engagement with humanities content, especially when paired with critical reflection and service-learning, consistently boosts learners’ empathy (1, 11, 16, 19, 20, 35, 37). Higher empathy, in turn, is linked to better patient outcomes (39–41), greater patient satisfaction (42, 43), and improved practitioner wellbeing (44–47).

Beyond these clinical benefits, Medical Humanities courses prompt students to recognize and challenge the hidden assumptions and structural biases that shape healthcare delivery (11, 16, 19, 20, 35). By embedding empathy and reflection into routine professional thinking, these courses help foster a culture of social justice and equitable care (48–52). Although resistance to “soft-science” content still exists in some programs (53), our results suggest that when Medical Humanities is integrated thoughtfully into the curriculum, students view it not as peripheral, but as central to becoming competent, compassionate professionals. In short, Medical Humanities is more than an academic add-on; it is a proven pathway to empathetic clinicians, healthier patients, and a more just healthcare system.

Positive attitudes toward team-based service-learning predicted students’ perceptions of the Humanities course’s impact on both critical reflection and peer empathy, whereas attitudes toward required readings played a significantly lesser role. These findings suggest that structured service-learning may be a particularly powerful and scalable lever for embedding Medical Humanities more effectively into healthcare training. When thoughtfully implemented, it has the potential to shape not only what students know, but also how they see themselves and others in the practice of care.