To provide high-quality care in response to increasingly complex medical needs, healthcare professionals must acquire the competency to collaborate across professional boundaries. Interprofessional education (IPE) has become a critical foundation for developing this competency.¹ IPE is promoted as an integral part of undergraduate education, and there has been a rapid increase in the number of universities integrating IPE into their curricula.² As future physicians who will be part of interprofessional healthcare teams, medical students are expected to develop positive attitudes toward interprofessional collaboration during undergraduate education. Studies show that medical students exposed to IPE before graduation tend to adopt a more positive attitude toward interprofessional work.^3,4 Therefore, IPE plays a crucial role in preparing medical students for collaborative practice.

Evidence supporting the educational effectiveness of undergraduate IPE is increasing. A systematic review⁵ found that about half of the study results showed a significant improvement in positive attitudes toward IPE when introduced during pre-clinical and clinical years following the intervention. Guraya's review⁶ highlighted improvements in students' knowledge, skills, and attitudes when comparing pre- and post-program evaluations. However, many studies that analyze the effects of IPE rely on pre-post or post-only designs, which limit understanding of how these attitudes change throughout the entire medical curriculum.

The spiral curriculum model, which introduces IPE early and progressively builds on it according to learners' levels, is recommended to optimize its effectiveness.² The spiral curriculum model emphasizes repeated exposure to important learning themes across different stages of education.⁷ This progressive sequencing is particularly relevant to IPE, as readiness for collaboration is expected to develop cumulatively through repeated learning experiences across educational stages. However, longitudinal studies examining changes in readiness for interprofessional learning have reported inconsistent findings. For example, previous studies have reported both declines in readiness over time and improvements following IPE interventions.^8,9 In addition, the number of such longitudinal studies remains limited.¹⁰

Early classroom-based IPE during pre-clinical years can be reinforced by later clinical experiences, such as community-based medical education (CBME), where IPE is embedded in real clinical settings. CBME provides an important opportunity for students to observe and participate in interprofessional collaboration and has been reported to enhance learning related to collaborative practice.¹¹ Given that the goal of undergraduate IPE is to prepare students for collaborative work in interprofessional teams upon graduation, it is crucial to understand whether improvements in readiness are sustained during the final stage of undergraduate medical education. However, it remains unclear whether changes in readiness for interprofessional learning following CBME are maintained during this final stage.

The Readiness for Interprofessional Learning Scale (RIPLS) is commonly used to assess students’ attitudes toward collaborative practice. This scale was developed to evaluate readiness for interprofessional learning, as changes in attitudes are considered essential for effective interprofessional learning among students participating in IPE.¹² RIPLS assesses learners’ readiness to acquire the knowledge, skills, and behaviors required for collaboration within interprofessional teams.

While many studies have demonstrated the immediate impact of IPE programs on students’ attitudes, evidence regarding the long-term sustainability of these effects remains limited. This gap is particularly important at the final stage of undergraduate medical education, which represents the transition to collaborative clinical practice.  Therefore, this study aimed to examine changes in medical students' readiness for interprofessional learning before and after CBME and whether these changes are sustained until graduation, while also providing a descriptive overview of longitudinal changes across the medical curriculum.

Participant flow is shown in Figure 2.

In the CBME 2018–2019 group, 118 students participated in CBME. One student was lost to follow-up at graduation due to repeating a year. Among the remaining students, 9 had missing data at either T5 or T6 (or both), and 4 had missing data at T7; these students were excluded from the analysis. As a result, 104 students with complete data for T5–T7 were included in the analysis.

In the CBME 2019–2020 group, 136 students were scheduled to participate in CBME; however, 35 students could not participate in the CBME from April to May 2020 owing to COVID-19–related cancellations, and one student did not provide consent to participate in the study. These students were excluded prior to analysis. One additional student was lost to follow-up at graduation due to repeating a year. Among the remaining students, 8 had missing data at either T5 or T6 (or both), and 3 had missing data at T7; these students were excluded from the analysis. As a result, 88 students with complete data for T5–T7 were included in the analysis.

In total, 192 students (104 in the CBME 2018–2019 group and 88 in the CBME 2019–2020 group) were included in the primary analysis.

Table 1 presents the baseline characteristics of participants at the start of CBME (T5). No significant differences were observed between the CBME 2018–2019 and 2019–2020 groups in age or sex.

Given the lack of statistically significant differences between groups in baseline characteristics (age and sex) or RIPLS total scores at any time point (all independent samples t-tests were non-significant, p > .05; for example, T5: t₍₁₉₀₎ = 1.85, p = .066), and the primary aim of examining overall changes associated with CBME, the two groups were combined for subsequent analyses.

Figure 3 illustrates the changes in mean RIPLS total scores across all time points. Descriptive statistics, including means, standard deviations, and sample sizes at each time point, are presented in Table 2. The mean RIPLS total score increased from T1 (73.8) to T2 (78.2), decreased at T3 (74.4), increased again at T4 (78.0), remained relatively stable at T5 (78.6), increased to its peak at T6 (82.2), and decreased at T7 (76.8). From T1 to T4, these changes represent descriptive trends. Due to missing data and academic year repetition at earlier time points, the number of observations varied across time points. Analyses for T5–T7 were based on participants with complete data (n = 192).

The primary analysis focused on changes from T5 to T7. Mauchly’s test indicated that the assumption of sphericity was violated (p < .05); therefore, Greenhouse–Geisser corrections were applied. A one-way repeated-measures analysis of variance (ANOVA) conducted among participants with complete data for T5–T7 showed a significant effect of time on mean RIPLS scores, F (1.60, 305.28) = 31.50, p < .001, ηp² = 0.14. Post hoc comparisons with Bonferroni correction indicated that scores at T6 were significantly higher than those at T5 (mean difference = 3.51, 95% CI [2.35, 4.67], p < .001), whereas scores at T7 were significantly lower than those at T6 (mean difference = −5.31, 95% CI [−7.12, −3.49], p < .001). There was no significant difference between T5 and T7 (mean difference = −1.80, 95% CI [−3.66, 0.06], p = .062).

This study demonstrated that medical students' readiness for interprofessional learning increased after CBME and reached its peak at the completion of CBME. However, the increase in readiness following CBME was not sustained until graduation. To our knowledge, this study is among the few to longitudinally examine changes in medical students' readiness for interprofessional learning and follow students through to graduation.

The increase observed between before and after CBME suggests that CBME provides an effective learning environment for enhancing readiness for interprofessional collaboration. Clinical training offers opportunities for authentic engagement in team-based care, which may facilitate the integration of interprofessional competencies beyond classroom-based learning.

Notably, readiness for interprofessional collaboration decreases at graduation, despite reaching its peak following CBME. This finding contrasts with the findings of previous research, which suggest that students in their final year are well prepared for increased interprofessional collaboration.¹⁵ In Japan, the final year of medical school is often dedicated to preparing for the National Medical Practitioners Qualifying Examination, which limits opportunities for collaborative learning. This may explain the observed decrease in readiness.

Descriptive trends observed during the classroom-based IPE periods suggest short-term increases in readiness following each intervention. Previous research has shown that about half of pre-graduate IPE programs significantly increase positive attitudes toward IPE before and after the program.⁵ Similarly, this study finds an increase in medical students' readiness for IPE both before and after the classroom-based IPE. However, readiness decreases between the second and third year of the IPE program, possibly owing to a lack of opportunities to engage with other faculties or professions during this period. However, these findings should be interpreted with caution, as no formal statistical analyses were conducted for these time points due to missing data.

The revised Model Core Curriculum for Medical Education in Japan, 2022 Revision, emphasizes interprofessional collaboration as a core competency16, and includes learning objectives based on the interprofessional competency framework for collaborative practice in Japan.¹⁷ To ensure seamless continuity between undergraduate and postgraduate clinical training, IPE in medical school should connect to interprofessional collaboration in clinical practice. A spiral curriculum, which introduces early IPE and incorporates repeated learning at appropriate levels, is recommended in the literature.^2. Frequent repetition in earlier years may help sustain readiness for interprofessional collaboration, while clinical experience in later years can reinforce this readiness. Additionally, to counter the decrease in readiness observed at graduation, introducing collaboration-focused programs in the early stages of postgraduate clinical training might prove effective.

However, implementing IPE programs faces challenges, including schedule adjustments and insufficient staff numbers.^18,19 Although an increasing number of universities inJapan are adopting IPE,²⁰ many still lack continuous programs, with most offering only a single program for lower-year students.^21,22 Online IPE, which has gained prominence since the COVID-19 pandemic, may help address these challenges by overcoming geographical and logistical barriers.²³

Several limitations of this study warrant mention. First, because this was a single-institution study conducted in Japan, we need to generalize the results cautiously. Second, the primary analysis was conducted using complete cases, which may have introduced selection bias, as students with complete data may differ systematically from those with missing data. For example, students with complete data may have been more consistently engaged in the curriculum and progressed without interruption, which could have led to an overestimation of readiness for interprofessional learning. In addition, the CBME 2019–2020 group experienced potential disruptions related to the COVID-19 pandemic, which may have influenced learning opportunities and students' readiness for interprofessional collaboration. However, as clinical rotations were scheduled across different time periods, the impact of these disruptions may have been distributed across students, potentially limiting systematic bias. Third, students participated in interprofessional learning in small groups during classroom-based IPE and interacted with various professionals across multiple training sites during CBME. Because group composition and training environments varied across time points, potential clustering effects could not be fully accounted for in the analysis. Fourth, although no substantial differences were observed between the CBME 2018–2019 and 2019–2020 groups in baseline characteristics or total RIPLS scores, the observed changes, particularly the increase from T5 to T6, cannot be definitively attributed to the effects of CBME, as maturation or other unmeasured factors may have influenced the results. Finally, concerns have been raised about the reliability and validity of the questionnaires used to evaluate IPE,²⁴ including the RIPLS instrument used in this study.²⁵ Although no perfect tool exists for measuring the educational effect of interprofessional learning, the RIPLS has demonstrated internal validity in numerous studies and remains a widely used evaluation instrument for pre-graduate IPE.²⁶ For consistency, we follow previous studies and use the total RIPLS score.^27,28

This study examined whether improvements in medical students' readiness for interprofessional learning following CBME are sustained until graduation. Readiness peaked after the CBME but was not maintained until graduation.

Introducing collaboration-focused training during the early stages of postgraduate clinical training and creating a seamless connection between pre-graduate IPE and postgraduate interprofessional collaboration may help sustain readiness and enhance the transition to clinical practice.