This study used an explanatory sequential mixed methods design following a pragmatic approach [25-27], capitalizing on quantitative and qualitative designs’ strengths while minimizing their shortcomings. Furthermore, it allowed researchers to better understand experimental results while incorporating participants’ perspectives. The National Institutes of Health advises a mixed methods approach “to improve the quality and scientific power of data” and to better address the complex issues facing health sciences today, including health professional education [28,29]. In the qualitative analysis, medical students’ reflection papers were text-mined to analyze the word frequency in community-oriented PBL. Next, we conducted individual interviews with medical students during their clinical clerkship.

This study was conducted as part of the undergraduate medical curriculum at Chiba University, Japan. Community-oriented PBL was conducted from October 24 to November 2, 2023, as part of a preclerkship/clerkship transition course [30], a 5-week preparatory education period before clinical clerkship. Additionally, PBL is integrated into the Chiba University medical school’s curriculum in years 1-4, and all students experience it. Participants were 113 fourth-year medical students who had attended lectures and received simulation training in basic and clinical medicine. To minimize potential biases and ensure an even distribution of student characteristics, participants were randomly divided into 16 groups of 7-8 each using the random number generation function in Microsoft Excel. The 16 groups received community-oriented PBL of a patient case using a community-based integrated care system with real-patient videos.

Quantitative data were gathered using the comprehension test, a tutor assessment with rubrics during the core time, and pre- and postintervention questionnaires to assess community health care perceptions (Figure 1). Additionally, a qualitative evaluation assessed community health care perceptions using a free-text reflection paper after PBL and follow-up interviews during clinical clerkships in community hospitals or clinics. The study used a mixed methods, sequential explanatory design to integrate the results [25,27,31].

Real-patient videos were meticulously prepared to enhance the learning experience authenticity, simulating real-life scenarios in community health care settings (Multimedia Appendix 1). Prepared in collaboration with medical education experts, community health care professionals, and audiovisual production specialists, these videos aimed to accurately depict home health care characteristics, including medical interviews, physical examinations, and in-home patient interactions. Real patients participated in the production process under strict ethical guidelines to ensure authenticity and respect for patient privacy, emphasizing community health care’s unique challenges and dynamics.

Each real-patient video was carefully scripted and filmed to represent common community health care situations, encompassing multiple scenes depicting different patient care stages and interactions, ranging from initial patient assessments in a community hospital setting to follow-up in-home visits. The video durations ranged from 3 to 5 minutes. Along with the videos, patient information sheets and tasks were presented to the students (Multimedia Appendix 2). One of the patient’s primary conditions was underlying diabetes, and they presented with severe lower leg edema. The case involved transitioning from acute care to a chronic care hospital, followed by the introduction of home visit medical services. Patient consent was obtained for video use, and students were instructed to adhere to confidentiality guidelines.

During the community-oriented PBL sessions, students collectively viewed the real-patient videos in designated classrooms equipped with audiovisual facilities, allowing for simultaneous viewing on shared screens. Before watching the videos, students were divided into small groups and assigned a tutor to facilitate discussions and learning activities. The tutors observed whether the students could achieve the learning objectives and facilitated discussions. The tutors, randomly selected faculty members in medical education and community-oriented medical education, were given standardized instructions and materials before the sessions to ensure consistency and effectiveness [32].

Community-oriented PBL sessions were divided into 2 sessions per case, each lasting approximately 3 hours. In the first session (core time 1), students were presented with the patient’s history and physical examination findings. In the second session (core time 2), the investigation findings and treatment plans were discussed. The same case scenario was used for all students.

This study also served as an educational program for fourth-year medical students in a basic clinical clerkship course. Altogether, 113 medical students from 12 groups were recruited. For quantitative data, the sample size required a 2-tailed t test of the difference between the pre- and post-PBL means, assuming a significance level of .05, a power of 0.8, and an effect size of 0.5. When the Mann-Whitney U test was conducted with those values, the required sample size was 54 in each group, totaling 108.

All statistical analyses of quantitative data were conducted using SPSS Statistics for Microsoft Windows version 29.0 (IBM Corp), with a significance level under 5% for each analysis. The comprehension test results, including total scores and individual question responses (Q1-Q5), were analyzed using the Wilcoxon signed rank test for paired total scores. Additionally, the McNemar test was used to compare pre- and post-PBL correct response rates for individual questions. For rubric-based performance assessment, the Wilcoxon signed rank test was used to compare scores from core time 1 and core time 2 for total scores and individual rubric items. Effect sizes were calculated for all analyses: r values were derived from z scores for the Wilcoxon signed rank test, and the Cohen w value was calculated for the McNemar test.

This study was approved by the Ethics Review Committee of the Graduate School of Medicine, Chiba University (approval number 3425). The procedures for obtaining informed consent were explained to the medical students, who were also informed that this study would not affect their grades. All data collected in this study were anonymized to ensure privacy and confidentiality. Participants did not receive any compensation for their participation in this study.

In total, 113 medical students participated in PBL. Of the 113 participants, 71 (62.8%) were male and 42 (37.2%) female. In addition, 2 (1.8%) students were excluded from the survey because they were absent for a core time session; 111 (98.2%) students participated in the quantitative evaluation. The study flowchart is shown in Figure 2.

Table 2 presents the results of the rubric-based assessment of the students’ performance in core time 1 and core time 2. The students’ total scores significantly improved from core time 1 (median 38, IQR 33-43) to core time 2 (median 40, IQR 35-44), with P<.001 and an effect size (r) of 0.516. Regarding individual rubric items, significant improvements were observed in domains such as acquiring knowledge applicable to community health care, developing comprehensive care processes, and fostering self-directed learning. In contrast, no significant improvement was observed in the motivation to learn, whereas interpersonal skills showed a small but statistically significant enhancement.

Results indicated significant changes in students’ perceptions of community health care after participating in community-oriented PBL. For the statement “I am interested in community health care,” the preintervention median score was 3 (IQR 3-4), which increased to 4 (IQR 3-4) post intervention, with a statistical significance of P<.001 (U=4446.5). Similarly, the median score for the statement “I can envision a community health care setting” improved from 3 (IQR 3-4) preintervention to 4 (IQR 3-4) postintervention, also showing a significant difference, with P<.001 (U=2589.5).

This study demonstrated that integrating real-patient videos into community-oriented PBL improves medical students’ knowledge, skills, and attitudes toward community health care. Comprehension test results showed significant improvements in students’ understanding of core concepts, including community-based integrated care systems (Q1), the ICF framework (Q2 and Q3), and holistic, patient-centered care (Q5). These findings highlight students’ enhanced theoretical knowledge essential for community health care practice.

Rubric-based performance assessments revealed notable improvements in 3 key domains:

Although interpersonal skills (item 5) improved slightly, no significant changes were observed in the motivation to learn (item 4), indicating areas for potential curriculum enhancement.

Self-assessment questionnaires revealed increased interest in community health care and an improved ability to envision a community health care setting. These results suggest that the intervention can positively influence students’ attitudes and readiness for community health care practice, potentially guiding their career interests toward rural areas.

Qualitative analysis of students’ reflections underscored themes such as readiness for home care visits, professional and personal growth, and community engagement. Students reported a deeper understanding of the complexities of community health care, fostering empathy and patient-centered approaches essential for effective practice in underserved areas.

This study provided valuable insights into how community-oriented PBL, enhanced by real-patient videos, fosters medical students’ ability to conceptualize their future professional roles. The qualitative data indicated that students develop a deeper understanding of the principles and complexities of rural care, including holistic approaches, patient-centered decision-making, and the importance of interprofessional collaboration. These findings suggest that the intervention successfully prepares students to engage with the challenges and rewards of rural and community-based practice.

Importantly, the qualitative analysis revealed that many students began to envision themselves as contributors to rural health care systems. Themes such as “professional and personal growth” and “community engagement” highlighted students’ recognition of their potential roles in underserved areas. This shift was supported by their increased interest in community health care, as measured by the “questionnaire for perceptions of community health care self-assessment.” Postintervention, students reported greater interest and confidence in envisioning community health care settings.

However, although the data indicated a significant attitudinal shift, we lack sufficient evidence to confirm a direct impact on medical students’ long-term career intentions to pursue rural care roles after graduation. Future studies should include longitudinal tracking to assess whether the observed changes in perceptions and interests translate into tangible career decisions. Additionally, research is needed to validate the predictive validity of the self-assessment questionnaire in forecasting students’ career trajectories.

This intervention lays a strong foundation for addressing the global challenge of physician maldistribution by bridging theoretical knowledge with practical applications. However, further investigation is required to understand its long-term influence on medical workforce trends and rural health care outcomes.

Our findings will contribute to a growing body of evidence supporting the efficacy of PBL in medical education. Previous studies have demonstrated PBL’s ability to enhance clinical reasoning and decision-making skills [19,21,41]. However, our research added a unique dimension by integrating real-patient videos, which provide authentic learning experiences and contextualize medical knowledge within the framework of community-oriented care. Similar studies have reported that experiential learning approaches, such as case-based learning with audiovisual materials, improve students’ engagement and retention of knowledge [41].

The significant gains in self-directed learning observed in this study echo findings from Matsuyama et al [35], who emphasized the role of contextual attributes in promoting self-regulated learning. Additionally, the qualitative themes identified in our analysis, such as the importance of interprofessional collaboration and navigating the home care environment, align with the existing literature on the competencies required for effective community health care practice [42-48].

Our study has some limitations. First, it was conducted as part of the curriculum of a single medical school, potentially limiting the findings’ applicability to other institutions and geographical settings. Future research should involve multiple institutions to enhance the results’ generalizability. Additionally, the demographic and cultural context of the participants may not fully represent broader populations, especially in countries with different health care challenges and educational frameworks. Second, the absence of a control group makes it difficult to attribute observed improvements solely to the intervention of incorporating real-patient videos into PBL. Our study evaluated a “bundle” of educational strategies, including real-patient videos, the PBL framework, faculty interventions, and testing conditions. We could not measure these components’ differential effects or potential synergistic interactions. Although a randomized controlled trial (RCT) could offer stronger evidence, implementing RCTs in educational settings poses ethical and logistical challenges, such as withholding valuable learning resources from a control group or ensuring equivalent baseline characteristics. To address these limitations, we recommend that future research consider alternative designs to balance rigor and feasibility. Third, the scalability of real-patient videos poses a significant challenge. Producing high-quality real-patient videos requires substantial time, resources, and collaboration among medical educators, health care professionals, and audiovisual specialists. These demands may limit the feasibility of widespread adoption. We suggest collaborative efforts, such as interschool partnerships and the development of shared digital repositories, to distribute production costs and enhance scalability. Fourth, the mixed methods design relied on self-reported measures and reflections, which could introduce a response bias. Future studies could benefit from incorporating objective measures of clinical performance and patient care outcomes. Fifth, the effectiveness of clinical reasoning education via hybrid PBL may vary depending on instructors’ teaching skills. Despite standardized training for tutors, differences in tutor effectiveness may have influenced the consistency of outcomes. We propose further methods to ensure uniformity in future implementations, such as advanced tutor workshops and peer evaluations. Sixth, because the study participants were fourth-year medical students at a single Japanese institution, the results may not be directly generalizable to other populations, such as residents or general physicians, or contexts outside Japan, underscoring the need for further validation. Seventh, the comprehension test used in this study, although developed specifically for the program’s educational objectives, was not formally piloted with a separate cohort. Instead, the test underwent cognitive debriefing with faculty members to ensure clarity, relevance, and alignment with the intended learning objectives. Although this process enhanced content validity, the absence of a formal pilot test may limit the ability to fully validate the test’s reliability and generalizability. Similarly, the 2 items in the self-assessment questionnaire—designed to evaluate medical students’ interest in and ability to envision community health care—were developed based on previous studies and focus group discussions but have not been validated using established scales. This lack of formal validation for the comprehension test and self-assessment questionnaire limits the generalizability and robustness of the findings. Finally, although the qualitative data provided valuable insights into students’ conceptualization of professional roles and their preparedness for community health care settings, the lack of longitudinal data limits our ability to assess the long-term impact of these interventions on career trajectories or practice in rural or underserved areas. Future research should include follow-up assessments to evaluate the sustained influence of such educational interventions on students’ career decisions and professional development.

Integrating real-patient videos into a community-oriented PBL curriculum shows significant promise in fostering medical students’ interest and competencies in community and rural medicine. Our study demonstrated improvements in knowledge acquisition and application, as indicated by enhanced rubric and comprehension test scores. Moreover, qualitative analysis revealed PBL’s effectiveness in developing essential skills and shaping medical students’ perceptions toward community health care. Although these changes may not directly translate to career decisions, they represent an essential step toward fostering awareness of rural health care needs and aligning medical students’ competencies with the demands of underserved areas. This approach highlights the potential of combining real-patient videos with PBL as an innovative educational strategy to address physician maldistribution and support rural health care systems.