Student Learning Outcomes and Program Evaluation in Medical Humanities Education in Korea: A Systematic Review

1. Research background and the importance of learning outcomes

During the 2024 Korean medical crisis, medical students emerged as a major stakeholder group in the national controversy surrounding healthcare reforms. Their collective leave of absence led to administrative sanctions and widespread public criticism, framing them as a privileged group perceived as neglecting civic duty [1]. This moment marked a shift in societal expectations, as medical students have increasingly come to be regarded not only as learners but as emerging professionals accountable to society.

A similar event was evident in 2000, when conflicts over the Korean Health Care System Reform Act prompted medical students and residents to engage in public advocacy for the first time. This episode broadened the understanding of medicine as a discipline bearing societal responsibilities in addition to its traditional foundations in basic and clinical medicine [2]. In response, courses in ethics, sociology, communication, and related fields were further expanded in Korean medical schools, not merely to complement clinical training but also to redefine the physician’s role within an increasingly complex social environment [3,4].

These developments aimed to cultivate physicians who are both clinically competent and socially responsive. The 2024 crisis renewed public attention to these longstanding educational goals and raised questions about whether current medical education effectively fosters the humanistic and civic competencies required of future doctors.

Aligned with this Korean context, evaluating the effectiveness of medical humanities education requires careful attention to student learning outcomes, which are commonly defined as “statements of what a learner is expected to know, understand, and/or be able to do at the end of a period of learning” [5].

2. Previous literature review

In Korea, research on medical humanities education has largely emphasized conceptual discussions related to curriculum design, instructional aims, and faculty roles. Studies in areas such as medical law education focus primarily on defining educational objectives and proposing strategies for curricular integration [6]. Similarly, recent work on medical artificial intelligence ethics offers theoretical frameworks and recommendations without evaluating learner development or educational outcomes [7]. Even structured analyses, such as the systematic review of patient–physician interaction curricula, describe communication-focused models but do not examine student-level outcomes [8]. Overall, Korean scholarship demonstrates strong interest in expanding humanities curricula, yet outcome-based program evaluations remain relatively scarce.

A similar pattern is observed internationally. Arts-based interventions, such as Theatre of the Oppressed workshops, generate rich student reflections but typically rely on subjective impressions rather than systematic outcome measures [9]. Studies examining humanities integration in the United Kingdom, as well as comparative analyses of medical humanities programs in the United States and China, map curricular offerings but remain largely descriptive and do not track changes in learner competencies or clinical behavior over time [10,11].

Across countries, therefore, the literature emphasizes curricular content and pedagogical philosophy, while systematic evaluations of learners’ actual development remain limited. This gap highlights the need for evaluative frameworks capable of assessing not only curricular intentions but also the measurable learning outcomes achieved by students participating in medical humanities programs.

3. Using the Kirkpatrick model in medical education evaluation

In health professions education, the Kirkpatrick model is widely used as a framework for evaluating the effectiveness of educational programs. Originally developed for workplace training, the four-level model has since been broadly adopted in medical education to examine how educational interventions influence participants’ knowledge, attitudes, behaviors, and downstream outcomes [12]. Importantly, the model conceptualizes learner responses as indicators of program impact, thereby positioning student outcomes as central to educational evaluation.

As the use of the Kirkpatrick model expanded within medical education, several refinements were introduced to enhance its relevance to this field. The Best Evidence Medical Education (BEME) Collaboration proposed modifications that subdivide Level 2 (learning) into Level 2A (attitudinal or perceptual change) and Level 2B (knowledge or skill acquisition), as well as Level 4 (results) into Level 4A (changes in professional practice) and Level 4B (outcomes for patients or systems) [13]. These distinctions enable more precise categorization of program effects, which is particularly advantageous when synthesizing heterogeneous educational studies.

Alternative models, such as the context–input–process–product (CIPP) framework, are also useful for evaluating curriculum design, implementation, and institutional processes. While powerful for guiding program development and organizational decision-making, CIPP places greater emphasis on contextual and process-oriented evaluation rather than on classifying learner-level cognitive, attitudinal, or behavioral outcomes [14,15].

For these reasons, the present review employs the BEME-modified, expanded Kirkpatrick model. This framework foregrounds learner-level outcomes, including attitudinal shifts, knowledge gains, skill development, and early behavioral changes, as evidence of program effectiveness. It also aligns directly with the study’s aim of evaluating how medical humanities curricula contribute to student development.

4. Research questions

This study addresses four research questions (RQs):

RQ1. What types of medical humanities programs and teaching methods have been implemented in Korean medical schools since 2000?

RQ2. What student learning outcomes have been used to evaluate these programs, and how have they been measured?

RQ3. What forms and levels of evidence have been used to assess the educational impact of medical humanities programs in Korean medical schools?

RQ4. To what extent have medical humanities programs led to measurable changes in student behavior?

1. Study design

This study was conducted as a systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 guidelines. The review process included structured database searches, predefined inclusion and exclusion criteria, and thematic synthesis conducted using qualitative coding software (ATLAS.ti; ATLAS.ti Scientific Software Development GmbH, Berlin, Germany).

2. Context

This review focused exclusively on undergraduate medical students in South Korea. Korea has a unique dual system of medical education comprising Western medicine and Korean traditional (oriental) medicine. Because these systems lead to separate licensure pathways and distinct professional roles, studies involving Korean Medicine students were excluded.

Between 2005 and 2015, South Korea operated a dual-track system that included 4-year graduate-entry medical schools alongside traditional 6-year undergraduate programs [16]. As the current and future direction of the Korean medical education system centers on the 6-year undergraduate format, particularly with a concentrated premedical phase in the first 2 years, this review excludes studies based on the 4-year graduate medical school model [17]. Consistent with this educational structure, we refer to Korean medical students using the labels MS1–MS6, corresponding to progression through the integrated 6-year medical education program. This labeling reflects the continuous and cohesive nature of Korean medical curricula, in which premedical education (years 1–2) and medical education (years 3–6) are designed as a unified sequence supporting long-term competency development across basic and clinical domains [18].

The Korean context is also distinct in the historical development of medical humanities education. In the early 1990s, such courses were often framed primarily as tools for individual character development [19]. Following the struggles over the Korean Health Care System Reform Act of 2000, medical humanities education expanded rapidly in response to public demands and growing recognition of the need to better prepare medical students to communicate with society. These programs emerged under pressure to address sociopolitical concerns while preserving the core aims of humanities education [2-4]. Accordingly, we limited our inclusion to studies published after this pivotal turning point in 2000, which catalyzed sustained academic discussion and curricular reform in this field.

In this study, we define medical humanities as an interdisciplinary field that integrates perspectives from the humanities and social sciences, including bioethics, anthropology, philosophy, literature, sociology, psychology, law, and the arts, into the education of future physicians. This definition aligns with both Korean academic traditions and internationally recognized frameworks [2,20].

3. Search strategy

We searched four databases: PubMed, KoreaMed, the Korean Medical Database (KMBASE), and RISS (Research Information Sharing Service). KoreaMed and KMBASE provide access to articles published specifically in Korean healthcare journals and are operated by the Korean Association of Medical Journal Editors (KAMJE) and the Medical Research Information Center, respectively. RISS is a service of the Korea Education and Research Information Service (KERIS) that primarily provides academic sources published in Korean journals. Across all databases, searches were limited to peer-reviewed journal articles.

Although the broader term “humanistic societal medicine” is occasionally used in Korea to describe an interdisciplinary domain encompassing medical humanities and social sciences, we employed “medical humanities” as the primary search term because of its wider international usage and consistent presence in both Korean and global literature [21]. To ensure comprehensive coverage, the search strategy also included “ethics” as an additional term, reflecting disciplinary priorities during the early 2000s [22].

Search terms were adapted to match the technical capabilities of each database, particularly because the Korean databases offered limited Boolean functionality. The complete search strategy is presented in Supplement 1. All searches were conducted on May 8, 2025.

After importing all identified records into EndNote ver. 21.0 (Clarivate, Philadelphia, PA, USA) and removing 114 duplicates, including cross-language duplicates, 121 articles remained.

4. Article selection

Articles were selected according to predefined inclusion and exclusion criteria. Eligible studies were those published between January 1, 2000, and May 8, 2025, that examined undergraduate medical students in South Korea and implemented curricula described as “medical humanities,” “medical ethics,” or “humanities in medicine.” Studies were required to report empirical measures of educational outcomes or student-level development, be written in Korean or English, and present original empirical data collected within South Korea. Studies were excluded if they involved participants outside the target population, reported only satisfaction or usability outcomes without educational impact, evaluated programs at the institutional rather than student level, were conducted outside South Korea, were written in languages other than Korean or English, or did not include empirical data. A detailed summary of all inclusion and exclusion criteria is presented in Table 1.

Table 1.

Inclusion and exclusion criteria for study selection

According to these criteria, article selection was carried out in two phases.

2) Phase 2: full-text review

Fifty-one full-text articles (n=51) were assessed for eligibility. Thirty-seven articles (n=37) were excluded because they lacked a clearly defined medical humanities educational intervention (n=15), did not include empirical analysis (n=13), failed to assess educational outcomes beyond classroom settings (n=7), or targeted populations other than undergraduate medical students (n=2).

During the full-text review, studies were excluded if they presented theoretical reviews or exploratory analyses of medical humanities curriculum development without providing original quantitative or qualitative data on educational outcomes. For example, several studies discussed the prospects, implications, or goals of narrative-based medicine, medical law, or general medical humanities courses. However, these articles were excluded because they did not report program evaluation data.

Ultimately, 14 studies were included in the final synthesis. Figure 1 presents a PRISMA flow diagram illustrating the article search and selection processes.

Figure 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) flow diagram of literature search.

5. Data extraction

Data were extracted using ATLAS.ti ver. 8.0 (ATLAS.ti Scientific Software Development GmbH). After importing all 14 articles into the project library, the first author (P.M.) conducted a secondary screening to ensure consistent application of the inclusion and exclusion criteria. To maintain alignment with the four research questions, the data extraction process followed a structured codebook comprising 10 primary codes.

For RQ1, which focused on the types of medical humanities interventions and teaching methods, the following codes were applied: (1) medical humanities (MH) intervention, (2) intention/learning goals, (3) teaching methods, and (4) curriculum integration. Interventions were categorized into predefined domains, including ethics, moral reasoning, professionalism, communication, and empathy. Additional domains, such as self-reflection and community engagement, were identified when programs specifically targeted these areas. Curriculum integration was determined based on program duration and placement within the medical curriculum.

For RQ2, which primarily addressed assessment methods, the applied codes included (5) assessment tool(s), (6) assessor and assessment timing, and (7) measured outcome. Assessment tool(s) referred to specific evaluation formats, such as questionnaires, reflective writing assignments, and mixed-method instruments. Measured outcome captured the learning domains or behavioral changes assessed, including knowledge, attitudes, empathy, and moral reasoning. These elements were verified through descriptions in the main text, tables, and supplementary materials.

To address RQ3, which focused on the strength of evidence, codes (8) data type and (9) evidence strength/credibility were applied. Data type was classified as quantitative, qualitative, or mixed, based on the measured outcomes identified under RQ2. Evidence strength/credibility focused on internal consistency, methodological rigor, and the presence or absence of comparison groups.

Finally, for RQ4, which examined educational outcomes and perceived effectiveness, code (10) reported educational outcomes was applied. Reported educational outcomes encompassed both quantitative findings and qualitative data, including narrative responses and open-ended comments.

This structured coding framework ensured consistent and systematic analysis of all included studies across the four inquiry domains, thereby enhancing comparability and overall analytical rigor.

6. Organizing educational outcomes using the BEME-modified Kirkpatrick model

To classify and interpret the learning outcomes reported in the included studies, we applied the BEME-modified Kirkpatrick model as the primary analytic framework. With respect to RQ2, the distinction between Level 2A (attitudinal changes) and Level 2B (cognitive or skill-based outcomes) provided a structured approach for identifying the types of learning outcomes assessed in Korean medical humanities programs. Levels 3 and 4A offered criteria for determining whether studies reported evidence beyond self-reported learning, including observable behavioral change or early changes in clinical practice, thereby addressing RQ4. However, studies reporting only Level 1 outcomes were excluded. Although student satisfaction may correlate with educational effectiveness, it does not constitute a learning outcome and cannot demonstrate attitudinal, cognitive, or behavioral change [23]. Accordingly, studies presenting Level 1 outcomes were included only when accompanied by higher-level learning or behavioral evidence.

To categorize study outcomes, we followed level definitions adapted from Hammick et al. [12] (Supplement 2), and each extracted data item was mapped to its corresponding Kirkpatrick level. By organizing outcomes according to the BEME-modified model, this review clarifies the extent to which medical humanities programs produce attitudinal, cognitive, or behavioral change.

7. Risk of bias assessment

To ensure the credibility and transparency of this systematic review, we assessed the methodological quality of all included studies using the Mixed Methods Appraisal Tool (MMAT), version 2018 [24]. The MMAT is specifically designed for systematic reviews that include qualitative, quantitative, and mixed-methods studies. Each study was first screened for empirical eligibility and subsequently appraised according to methodological criteria appropriate to its study design. Detailed item-level assessment results for each study are presented in Supplement 3.

1. RQ1: What types of medical humanities programs and teaching methods have been implemented in Korean medical schools since 2000?

The 14 included studies revealed substantial thematic diversity and shifting educational priorities in medical humanities education in Korea since 2000 (Table 2). Medical ethics education consistently appeared as a fundamental component, serving as the primary focus in five studies [25-29] and as part of broader medical humanities curricula in two additional studies [30,31]. The persistence of ethics-centered programs across the study period underscores a sustained emphasis on moral reasoning and professional identity formation within Korean medical education.

Table 2.

Program characteristics and included MH interventions

More recently, program content has expanded toward community engagement and innovation-oriented approaches. One study introduced a community-based volunteer program designed to foster student motivation and social responsibility [32]. Another study employed a design-thinking framework to guide students in developing creative solutions to everyday societal problems, reflecting a shift toward problem-oriented and socially embedded learning [33].

Additional programs explicitly targeted self-reflection and professional growth. One course emphasized introspective leadership development [34], while another linked career exploration with humanities-based reflection, encouraging students to critically consider their future roles as physicians [35].

Regarding teaching methods, none of the reviewed programs relied solely on traditional lecture-based instruction. Instead, instructors employed a range of interactive and student-centered methods, including debate-based sessions [25,26,28,30,31,36-38], group projects [27,29,32-34], and student presentations accompanied by instructor feedback [35]. This pedagogical shift reflects a broader movement away from didactic ethics instruction toward experiential, dialogic, and reflective approaches to professional identity formation in Korean medical education.

In terms of curricular integration, most medical humanities courses were implemented within relatively short time frames, typically lasting a single semester [25,31,34,36-38] or delivered in condensed formats spanning two 9-week periods [26-29,33]. One motivation induction program was uniquely offered during the first month of the freshman year [32]. Two studies reported longitudinal designs: Park et al. [30] introduced the Patient–Doctor–Society (PDS) program integrated over 4 years (MS3–MS6), while Yoo and Kang [35] evaluated a year-long career development program bridging the final two academic years. These varied formats indicate that, although most medical humanities programs remain semester-based, some initiatives demonstrate efforts toward longitudinal integration within Korean medical schools.

Across studies, curricular duration was predominantly short-term. Only one program followed students for more than one academic year, and no study implemented multi-year, sequential medical humanities experiences beyond the PDS curriculum [30,35].

2. RQ2: What student learning outcomes have been used to evaluate these programs, and how have they been measured?

All 14 studies evaluated student outcomes exclusively through student self-reports, without the involvement of external evaluators. Most relied on post-course surveys using Likert-type scales to capture students’ perceptions and attitudinal shifts [25-28,30-32,35-38]. Seven studies additionally incorporated reflective writing or open-ended narrative responses, offering qualitative insights into students’ internalization of course content and reasoning processes [25,26,29,33-35,38].

Across all studies, evaluations were conducted immediately after course completion; no study incorporated delayed or longitudinal follow-up assessments.

Using the expanded Kirkpatrick framework, Level 1 outcomes were identified in every study except one [29]. All studies demonstrated Level 2A (attitudes and perceptions) outcomes, and six studies additionally reported Level 2B (knowledge or skills) outcomes [26-31,37].

Despite frequent descriptions of learning experiences as “transformative,” none of the studies incorporated triangulated or behaviorally anchored assessment methods, and all relied exclusively on student self-evaluation. Park et al. [30] reported gains in communication and professionalism; however, 68% of measured items showed no statistically significant change, and no external observers assessed performance. Consequently, these outcomes could not be classified as Level 3. These examples illustrate why most studies were categorized as Level 1 or Level 2A.

Overall, outcome assessments primarily reflected short-term, subjective gains. Future studies may benefit from incorporating longitudinal tracking, external evaluation, and objective indicators of behavioral or professional change to more robustly assess the educational impact of medical humanities programs.

3. RQ3: What forms and levels of evidence have been used to assess the educational impact of medical humanities programs in Korean medical schools?

Eight studies used quantitative designs incorporating post-course surveys and structured pre-/post-assessments [27,28,30-32,34,36,37], two relied primarily on qualitative methods [25,33], and four adopted mixed-methods approaches that combined survey data with narrative analysis [26,29,35,38]. These methodological characteristics, together with the corresponding outcome measures and Kirkpatrick classifications, are summarized in Table 3.

Table 3.

Evaluation and outcomes

Among the quantitative and mixed-methods studies, a small number applied basic statistical techniques, such as the t-test or regression analysis, to examine differences across groups or changes over time [26,27,35]. However, none of the included studies incorporated multi-institutional comparisons, longitudinal modeling, or external validation strategies.

Across all quantitative studies, internal consistency was routinely reported, with Cronbach’s alpha values exceeding 0.80 in most cases, except for one study using the Defining Issues Test (DIT; α=0.64), suggesting potential reliability challenges in measuring moral reasoning constructs [26]. Two studies employed custom-developed instruments and described their development processes, including expert consultation and pilot testing, to enhance content validity [32,33].

4. RQ4: To what extent have medical humanities programs led to measurable changes in students’ behavior?

Despite the range of reported outcomes, few studies provided direct evidence of measurable behavioral change. Park et al. [30] presented the strongest case, reporting students’ self-perceived gains in communication and professionalism following participation in a longer curriculum. However, although perceived growth was documented, 68% of measured items did not reach statistical significance, suggesting limited observable change overall [30]. Moreover, because no external assessments were conducted, these outcomes could not be classified as demonstrating Level 3 (observable behavioral change) according to the Kirkpatrick framework.

Other studies reported shifts in ethical reasoning and decision-making, most commonly through reflective essays or student-reported opinion changes following debate-based or case-based learning interventions [25,28,31,36]. One study used the DIT to assess moral reasoning and found no overall change in composite scores; however, it observed a narrowing of score distribution, with improvement among lower-scoring students [26]. Similarly, Chung et al. [27] reported that in a team-based learning (TBL) ethics course, cooperative group scores (group readiness assurance test, GRAT) were significantly higher than individual readiness assurance test scores (IRAT; p<0.001), and that final examination performance improved most markedly among students in the lower-to-middle grade point average (GPA) quartiles.

One study further identified peer norms and institutional culture as stabilizing forces that may limit measurable behavioral change, even with extended exposure to medical humanities content [30].

Beyond ethical development, two studies reported positive associations between participation in medical humanities programs and academic outcomes. One study found significant correlations between participation in a humanities-based leadership course and improved performance in preclinical subjects (r=0.40–0.74, p<0.001) [34]. Another study reported that students in the second and third GPA quartiles demonstrated greater gains in TBL assessments within a medical ethics course [27].

Overall, although several studies reported indicators suggestive of behavioral growth or spillover effects, none provided longitudinal outcome tracking, delayed follow-up assessments, or evidence of behavioral transfer into clinical practice settings.

1. Challenges in achieving higher-level educational outcomes

Across the included studies, outcome evaluations were predominantly limited to short-term, self-reported attitudinal measures corresponding to Kirkpatrick Levels 1 and 2A. Although such indicators provide useful information regarding learners’ immediate perceptions and subjective gains, they offer limited evidence of sustained learning or behavioral change. This pattern mirrors long-recognized trends in the broader medical education literature. Yardley and Dornan [23] demonstrated that the vast majority of educational evaluations report only Level 1 or Level 2 outcomes, with Level 3 or Level 4 outcomes accounting for approximately 12% of published findings. More recent analyses of medical humanities curricula in North America similarly concluded that nearly all interventions report outcomes confined to these lower levels, with limited evidence of observable behavioral change or patient-level impact [39,40]. A comparable pattern was identified in a Taiwanese national systematic review, which reported that none of the 17 included studies provided evidence of Level 3 behavioral change or Level 4 organizational or patient outcomes, despite the broad implementation of medical humanities programs across institutions [20].

Taken together, these findings suggest that the limited emergence of higher-level outcomes is not unique to the Korean context but instead reflects a broader methodological challenge in evaluating medical humanities education. Domains such as empathy, ethical reasoning, and professional identity develop gradually and are strongly shaped by learning environments, interpersonal dynamics, and clinical immersion, which makes them difficult to assess using short-term or self-reported measures alone. To generate more rigorous evidence of educational effectiveness, future research will require longitudinal study designs, multi-source evaluation strategies, and behavioral assessments situated within authentic clinical settings.

2. Beyond curriculum design: the role of peer culture and hidden curriculum

One study highlighted the influence of peer culture on learning outcomes [30]. Although students participating in an extended curriculum reported personal growth, 68% of the measured items showed no statistically significant change. The authors suggested that peer norms and school culture may have exerted stabilizing effects on student behavior. This interpretation aligns with the concept of informal learning or the hidden curriculum, which can either reinforce or undermine the intended goals of formal instruction [41]. This perspective is supported by global research demonstrating that peer interactions and shared expectations strongly shape how students respond to educational interventions [42,43]. In collectivist contexts such as Korea, peer dynamics may dampen the observable effects of individual courses but also present a distinctive opportunity. When medical humanities programs are implemented across entire cohorts and aligned with institutional values, they may influence not only individual learners but also collective student norms. Over time, such alignment has the potential to contribute to broader cultural shifts within medical education.

Taken together, these findings indicate that the effects of medical humanities education are shaped not only by curriculum design but also by peer-level and institutional contexts. This underscores the importance of accounting for these underlying factors when evaluating and further developing medical humanities curricula.

3. Advancing beyond prior reviews: scope, outcomes, and evaluation

Prior systematic reviews provide important foundations for understanding how medical humanities education has been implemented, yet they differ meaningfully from the scope and evaluative approach of the present study. Lee et al. [8] synthesized patient–physician interaction programs with an emphasis on communication but did not examine learning outcomes using a structured analytic framework. Zhang et al. [44] focused exclusively on empathy-related interventions and provided meta-analytic evidence of their effects; however, their review was limited to a single competency domain rather than the broader landscape of medical humanities education. Hoang et al. [20] identified a wide range of curricular approaches, including studies that reported only Level 1 reaction outcomes.

In contrast, the present review examines multiple domains of student development across medical humanities curricula within a unified national educational context. By applying inclusion criteria that exclude studies reporting only Level 1 outcomes, this review is able to more directly evaluate student learning outcomes and the educational effectiveness of medical humanities programs. Collectively, these methodological decisions enable a clearer depiction of how medical humanities education shapes learners’ attitudes, knowledge, and emerging professional capacities, while also identifying specific outcome areas in which Korean programs demonstrate strengths or require further development. As a result, this review offers a more outcome-focused and system-level understanding of medical humanities education than has been available to date.

Despite these strengths, several methodological limitations should be acknowledged. First, the review was not prospectively registered, which may limit procedural transparency. Second, gray literature was not included, raising the possibility of publication bias. In addition, the methodological quality of the included studies varied, with many relying on single-institution samples and short-term, self-reported outcomes, as identified through MMAT appraisal. These limitations align with well-recognized standards for systematic review rigor and should be considered when interpreting the findings.