A Narrative Review and Novel Framework for Application of Team-Based Learning in Graduate Medical Education
ABSTRACT
Background
Team-based learning (TBL) promotes problem solving and teamwork, and has been applied as an instructional method in undergraduate medical education with purported benefits. Although TBL curricula have been implemented for residents, no published systematic reviews or guidelines exist for the development and use of TBL in graduate medical education (GME).
Objective
To review TBL curricula in GME, identify gaps in the literature, and synthesize a framework to guide the development of TBL curricula at the GME level.
Methods
We searched PubMed, MEDLINE, and ERIC databases from 1990 to 2014 for relevant articles. References were reviewed to identify additional studies. The inclusion criteria were peer-reviewed publications in English that described TBL curriculum implementation in GME. Data were systematically abstracted and reviewed for consensus. Based on included publications, a 4-element framework—system, residents, significance, and scaffolding—was developed to serve as a step-wise guide to planning a TBL curriculum in GME.
Results
Nine publications describing 7 unique TBL curricula in residency met inclusion criteria. Outcomes included feasibility, satisfaction, clinical behavior, teamwork, and knowledge application.
Conclusions
TBL appears feasible in the GME environment, with learner reactions ranging from positive to neutral. Gaps in the literature occur within each of the 4 elements of the suggested framework, including: system, faculty preparation time and minimum length of effective TBL sessions; residents, impact of team heterogeneity and inconsistent attendance; significance, comparison to other instructional methods and outcomes measuring knowledge retention, knowledge application, and skill development; and scaffolding, factors that influence the completion of preparatory work.
Introduction
The adoption of a competency-based system of evaluation places a greater focus on skill acquisition than prior methods. Graduate medical education (GME) educators must be equipped with efficient and effective instructional methods to develop the necessary knowledge and skills in resident physicians to achieve these competencies.
Team-based learning (TBL) is an instructional method that promotes problem solving and teamwork. TBL exemplifies a flipped classroom: learners acquire knowledge through assigned readings or videos before class, and class time is used to integrate and apply knowledge, building a deeper understanding of concepts.1 This method has the potential to foster the development of several Accreditation Council for Graduate Medical Education competencies, such as medical knowledge, professionalism, and interpersonal and communication skills.
A TBL session consists of 3 key components: (1) preclass preparation; (2) readiness assurance; and (3) application (figure 1). In preclass preparation (prework), learners review assigned materials before the session to acquire knowledge on a particular topic. The readiness assurance phase begins with a short individual readiness assurance test (IRAT) of multiple-choice questions. Learners then work in teams to complete the same questions, called the group readiness assurance test (GRAT), which is followed by a large group discussion of answers. The application exercise phase presents a real world, complex problem for teams to solve. This exercise mimics the realities of clinical practice, where there is often no single right answer, promoting further discussion within and among teams. These discussions create constructive controversy and encourage peer learning.1,2 Across multiple TBL sessions, team members remain consistent to foster teamwork and collaboration.1 Due to the focus on knowledge application, TBL covers a smaller breadth of material in greater depth.
Citation: Journal of Graduate Medical Education 8, 4; 10.4300/JGME-D-15-00516.1
Most published accounts of TBL in medical education pertain to undergraduate medical education (UME). In comparison with lectures, learning outcomes for UME learners in TBL have been shown to be equivalent or superior.2 Additionally, UME learners have been reported to prefer TBL over lectures.4 It has not been established whether these findings in UME will extend to GME.
Descriptions of TBL in GME are beginning to be published. Despite potential benefits of TBL, unique challenges arise in its application in GME. A residency program's protected educational time may lack the flexibility to support the 2- to 3-hour length of traditional TBL sessions. Attendance may be more inconsistent due to off-site rotations, duty hour limits, and unpredictable clinical demands. Instructors also may not be consistent, with different faculty facilitating each session. Additionally, TBL requires preparation by faculty and residents, which may be challenging in the context of clinical demands. Given competing demands on residents' time, specific motivators may be necessary to encourage completion of prework.
To date, no guidelines or reviews regarding the implementation of TBL in GME have been published. This narrative review aims to summarize the existing literature describing the application of TBL in GME and to identify gaps. Using these studies as a foundation, we propose a framework to guide development of TBL curricula for GME.
Methods
We searched PubMed, MEDLINE, and ERIC databases from January 1990 to December 2014 using the keywords team-based learning, graduate medical education, residency, and residents. The terms fellowship and fellows did not identify additional articles. Each article's references were reviewed to identify additional publications.
Studies were included if they were published in English in a peer-reviewed publication and described TBL in GME. Studies were excluded if they did not describe a specific TBL curriculum, were limited to an abstract that lacked the detail necessary to be evaluated using our framework, or described a team-based curriculum that did not follow the TBL model. The full text of articles were reviewed to confirm inclusion.
Two authors (R.S.P. and A.V.) completed a preliminary review of all articles for inclusion criteria. Three authors (C.A.L., R.S.P., A.V.) independently reviewed the full text of included publications to extract data, including details of curriculum planning, curricular structure, learning outcomes, and learner and faculty satisfaction (table). Corresponding authors of the publications were contacted, as needed, to clarify details. Following independent data extraction, the 3 authors reviewed the data for comprehensiveness and consistency. Consensus was reached by discussion.
Development of a TBL Framework for GME
Based on the publications reviewed, we developed a framework for GME educators to provide a road map for the development of future TBL curricula. The framework also provides a stepwise approach to identify and address the unique challenges presented with applying TBL in GME. Based on our review of the included publications, the framework for TBL in GME includes 4 elements: system, residents, significance, and scaffolding.
System is the organizational context in which TBL curriculum is developed, and thus directly affects the other steps in the framework. It encompasses faculty leaders, curricular schedule, and institutional support for TBL. Practical considerations in GME include the frequency and length of sessions within clinical obligations, faculty development for TBL, faculty time investment, and facilitator consistency across sessions.
Residents encompasses the learner characteristics and team composition for the planned TBL curriculum. Compared to UME, GME learner groups may be more heterogeneous due to different training levels (postgraduate year [PGY]), distinct medical school experiences, and the variable order of residency rotations and experiences. Although this heterogeneity can improve learning in TBL through knowledge exchange between learners, it needs to be addressed in the planning stage. Learning goals may differ for a group of PGY-1 residents compared with a multilevel group. In addition, when there is disparity among learners, attention must be given to balancing knowledge and experience among teams. Expected attendance may vary and should also be considered in planning the size and number of teams. Once learner characteristics and team composition are defined, the curriculum can be tailored to match learners' needs.
Significance includes the development of goals, objectives, and assessments for a TBL curriculum. Based on formulated goals and objectives, curricular considerations include the type of application exercise and the facilitation of peer feedback. Assessment methods should be considered in parallel with learning goals, with measures aiming to capture the critical thinking and concept application skills central to TBL, rather than simple recall.
Scaffolding refers to structural components essential for a successful TBL curriculum, such as orientation to TBL, prework, and learner motivators. Orientation reduces learner resistance to a new educational approach. Prework provides learners the core knowledge necessary for knowledge application. Because residents have limited free time, the amount of assigned prework is an important consideration, and specific motivators may be necessary to encourage completion of prework, given the usual absence of grades in residency, to motivate learners.
Results
Search Results
The initial literature search yielded 15 articles, of which 8 (53%) met inclusion criteria. A manual search of the reference sections identified 1 book chapter.12 This review includes 9 publications describing 7 unique TBL curricula (figure 2; table).5–13
Citation: Journal of Graduate Medical Education 8, 4; 10.4300/JGME-D-15-00516.1
Study Design and Data Presented
Of the 9 publications included, none employed an experimental design or compared TBL with another instructional method. Only 1 study10 compared outcomes to a control group, with a nonrandomized waitlist control. Four studies6–9 reported pretest and posttest data, and 1 study5 reported posttest data only.
Specialties represented include family medicine,7 internal medicine,10 pathology,5 psychiatry,8,11,13 and physical medicine and rehabilitation (PM&R).12 Topics included alcohol screening and brief intervention,7,10 anatomic and clinical pathology,5 addictions,8,9 psychodynamic psychotherapy,11 evidence-based psychiatry,11 journal club,13 and an entire PM&R didactic curriculum.12 Resident learners ranged from PGY-1 to PGY-4. A subset of curricula also included faculty, nurses, and staff as participants.7,12,13
Outcomes measured included performance on a standardized patient assessment,10 Value of Teams survey,8,9 team performance survey,5 self-reported clinical behavior,6,7 and comparison of IRAT to GRAT scores.5,7 Three studies11–13 presented descriptive reports of TBL curricula, with 1 study12 describing plans to compare outcomes with historical controls in the future.
All curricula assessed learner satisfaction with the curriculum and/or TBL methodology.5–13 Feasibility data regarding faculty time5–7,12 and faculty development5–9,12 were also included in several reports. No studies addressed costs beyond faculty/staff time and training.
Examining Articles Within the TBL Framework for GME
We examined the 7 curricula using the framework we developed for TBL application in GME, summarized findings (table), and used this to identify unanswered questions about TBL in GME.
System
Most curricula consisted of 1 to 8 TBL sessions; however, 1 study12 described the replacement of an entire 2-year PM&R residency curriculum (100 sessions) with TBL. Experts recommend multiple sessions with consistent teams to yield the communication and team-building benefits of TBL, although no minimum number of sessions has been established.1 TBL sessions lasted between 1 and 3 hours and replaced didactics or morning report.
In the majority of TBL curricula, faculty development was led by a local “champion,”9,12 with some faculty also attending national TBL conferences.8,9 In addition to training facilitators, faculty development increased enthusiasm and buy-in among faculty.12
Facilitators varied from a single faculty member to multiple faculty members planning and leading sessions. Measures to ensure consistency and quality among TBL sessions included peer review of materials,9,12 feedback during practice sessions,7 and presence of 1 consistent faculty member at all sessions.9,12
Notably, articles5,7,12 described that preparation for TBL sessions required a greater time investment than lectures, with 1 study7 estimating that 3 TBL sessions required 34 hours of preparation. Faculty subjectively reported that the improved resident enthusiasm was worth this time investment.7 Objective data describing faculty preparation time are lacking.
Residents
Most curricula included heterogeneous teams of residents at multiple training levels; 1 included only PGY-2 psychiatry residents.8 Some curricula also included faculty7,13 and interdisciplinary clinic staff7 as learners.
Team size ranged from 4 to 8 learners. When specified, team composition was balanced by training level5,11 and prior knowledge.8,12 The impact of inconsistent attendance, potentially creating unbalanced teams, on learning and teamwork outcomes is not discussed in the literature.
Significance
TBL curricula primarily used a traditional case-based format for the application exercise.7,8,11,12 Alternative formats included role play10 and a research design exercise.13 These formats align with TBL principles, as they require knowledge application without a single “right” answer.
Peer evaluation included a peer evaluation tool,5 online peer evaluation,8 and in-person feedback.10 The exchange of constructive peer feedback is an integral skill in medical training, and the benefits of including this process in TBL warrant further investigation.
Outcomes of studies largely focused on changes in knowledge, rather than higher-level outcomes of behavior or skill. The sole exception10 showed significantly higher standardized patient assessment scores for the TBL group compared with a control group. Knowledge gains were demonstrated by significant increases in GRAT scores compared with IRAT scores and resident self-assessment.7,10,13 No significant change in teamwork outcomes was seen with pre– and post–Value of Teams surveys8 and team performance surveys.5 Additionally, most curricula (with 2 exceptions6,7,12) measured outcomes only immediately following TBL sessions with no assessment of retention.
Scaffolding
Approaches to orienting learners to TBL included in-person orientations9,12 and electronic information distribution.9,13 Studies did not examine the impact of orientation characteristics on learner acceptance of TBL. Varied approaches were used to foster completion of prework by busy residents, often limiting prework to a few articles or online videos. One study8 considered allotting protected time for prework, but this proved not to be feasible; when the IRAT was included as prework to most efficiently use class time, the completion rate was a disappointing 33%.
Although a range of performance incentives were provided, including food,7,13 textbook money,9,11–13 or an iPad,9 competition was subjectively reported as the most-effective motivator.7,12
Learner Reaction
Most learners had positive reactions to TBL as an instructional method and requested more TBL sessions7,8; however, a subset of learners felt TBL was less efficient than lectures.8 One study10 found residents were neutral about recommending that specific TBL session to colleagues. Informal faculty feedback identified several benefits of TBL: increased learner engagement,5,8 enthusiasm,11 and interactivity.9 This observation is supported by significantly higher scores on a Classroom Engagement Survey for TBL versus didactics.8
Discussion
The 7 TBL curricula identified from the literature review provide a foundation for the future study of TBL in GME and are most instructive in regard to feasibility. Learner feedback was largely positive or neutral. Several authors reported increased investment of faculty time in curriculum development in comparison with other instructional methods. The time investment needed to develop TBL curricula is an important feasibility consideration. Most publications included in this review incorporated limited learner assessment and curricular evaluation. Further research with more relevant outcome measures is needed to measure the impact of TBL in GME, in comparison with traditional instructional methods. The studies reviewed do not provide clear evidence of TBL's efficacy in GME. Although some studies showed significant knowledge gains with TBL curricula, future studies should utilize comparison groups or comparison with other instructional methods to study TBL. Given the time investment needed to prepare TBL sessions, it would be useful to identify what topics and learning objectives lead to superior learning outcomes with TBL, rather than with lecture- or small group–based methods, to efficiently utilize program resources.
The majority of significant knowledge outcomes presented showed increases from IRAT to GRAT scores. These readiness assurance tests are completed early in TBL sessions and test the acquisition of background knowledge, not knowledge application. Future work should aim to capture higher-level outcomes than knowledge acquisition, in line with TBL's aim to promote knowledge application. In addition, the GME learning environment presents unique challenges to the use of TBL, and the impact of these challenges on learning outcomes should be explored; these include variable team attendance, heterogeneous teams, session length, variable facilitators, and prework completion motivators.
Limitations of this review include the limited number of included publications, which varied in the level of detail provided, with faculty development, teams, orientation, incentives, and prework often not specified. Most studies had methodological limitations, consisting of observational approaches or a simple posttest design. There do not appear to be published studies of TBL in several specialties, including surgery, emergency medicine, anesthesiology, and pediatrics, and the exclusion of TBL curricula not reported in the literature or missed by our search strategy may have affected the findings. Our framework was developed based on the available descriptions of TBL in GME and may need to be adapted as more is learned about TBL in GME.
Using the framework, we identified areas in need for future study. These include research to determine the minimum length for an effective TBL session in order to address time constraints in GME. Future research should also explore the impact of team heterogeneity in training level and discipline on session development, learning outcomes, and teamwork; develop improved ways for measuring knowledge application and skill development; and measure knowledge retention at longer intervals after TBL. A final area for study is the impact of different incentives on prework completion and the effect of prework on learning outcomes.
Conclusion
This review identified 7 unique TBL curricula in GME, and proposes a framework for TBL in GME—system, residents, significance, and scaffolding—to guide the development of future curricula. While TBL aims to promote problem solving and deliberate practice of skill-based competencies not easily addressed by didactics—namely communication and teamwork—a limited number of TBL curricula have been described in GME with none specifically designed to address and measure these competencies. The literature shows that TBL can be implemented with residents, although the additional time investment may be considerable and the benefit of this approach for GME learners requires further study.
Structure of Team-Based Learning Sessions
Note: Figure 1 is adapted from Parmelee and Michaelsen.3 Reproduced with permission.
Flow Diagram of Selection and Screening Process for Literature Review
Author Notes
Funding: The authors report no external funding source for this study.
Conflict of interest: The authors declare they have no competing interests.
The authors would like to thank Ara Tekian, PhD, MHPE, and Janet Riddle, MD, Department of Medical Education, University of Illinois at Chicago, for their support and editorial feedback on this manuscript.
