Feasibility of a Comprehensive Medical Knowledge Curriculum in Internal Medicine Using Team-Based Learning

The medical knowledge curriculum is an essential part of residency training and is traditionally taught in a lecture setting,¹  fulfilling the Accreditation Council for Graduate Medical Education mandate for regularly scheduled didactic sessions.²  Challenges posed by work hour restrictions, increasing clinical demands, reduced time for teaching, and limited effectiveness of lectures as a means of engaging learners have resulted in calls for active learning approaches.^3,4  Team-based learning (TBL) is an active learning method that only recently gained traction in graduate medical education (GME).⁵  A growing body of scholarship has demonstrated that TBL promotes learning and teamwork,⁶  learner engagement,⁷  clinical skills development,⁸  and learner and faculty satisfaction.⁹ 

Team-based learning fosters application of knowledge through individual work, small group work and teamwork, and immediate feedback. Learners acquire knowledge through self-study and completion of required reading assignments. During TBL sessions learners apply this knowledge in a series of problem-solving exercises led by an instructor. This may include Individual Readiness Assurance Tests (IRATs), Group Readiness Assurance Tests (GRATs), and application exercises.^10,11 

Most reports of TBL in GME have described its use with limited curricular context such as teaching a single skill⁸  or in pilot applications.^6,9,12  Only 1 report described the replacement of an entire lecture-based curriculum in a physical medicine and rehabilitation residency.¹³  Given the challenges of faculty time commitment and fitting TBL sessions within a busy resident and faculty workday, the feasibility of its adaptation to GME remains unclear.⁵ 

We describe the transformation of the curriculum in 1 internal medicine (IM) program to TBL, and discuss the planning, curriculum design, resident orientation, faculty recruitment and development, and lesson structure, and report on program feasibility and sustainability.

Our residency is a medium-sized, university-based program with 48 categorical and 25 preliminary IM residents. We also provide training to 16 medicine-pediatrics residents. Outpatient clinic settings include a Veterans Affairs hospital clinic and a university practice. Prior to 2013 there was a traditional curriculum consisting of noon conference lectures 4 days per week, repeated yearly. Attendance was poor, and residents described the sessions as lacking educational quality. We decided to trial a new TBL curriculum and deliver it within the program's 4+1 block schedule, in which 5 cohorts of 10 residents have 4 weeks of inpatient/elective time followed by an ambulatory week.

The transformation to TBL occurred from 2013 through 2016. We designed 2 separate TBL curricula: (1) a weekly hospital and subspecialty TBL lesson taught in a 90-minute noon conference on Fridays for residents on inpatient/elective and ambulatory rotations (the hospital curriculum), and (2) an ambulatory curriculum with 2 separate, 90-minute TBL lessons taught during a Tuesday morning academic half-day of the ambulatory week, attended by the 10 residents on the ambulatory rotation. The hospital curriculum included 75 topics taught over 18 months repeated once during the 3-year program. The ambulatory curriculum included 60 topics, each taught once during residency. As residents may not take vacation during ambulatory week, attendance was ensured. Topics for both curricula (provided as online supplemental material) were chosen from the table of contents of the Medical Knowledge Self-Assessment Program. We thought this represented the areas of knowledge for competent practice and for residents to pass the American Board of Internal Medicine certification examination. Clinical reasoning, quality and safety, and morbidity and mortality were taught during a traditional 1-hour noon conference Monday to Thursday. One author (G.S.) who had attended the national Team-Based Learning Collaborative meeting assumed the role of TBL “champion.”

The authors reviewed background resources on TBL^10,14–16  and then developed 9 hospital and 2 ambulatory lessons. We chose readings from standard textbooks and journals, and e-mailed them to the residents as attached PDF files. We wrote TBL questions based on content, and modified existing board review questions. Three authors (G.S., J.W., R.S.) served as instructors for the initial sessions.

On July 1 residents received readings for the first hospital topic (pneumonia) and ambulatory topics (gastroesophageal reflux disease and hyperlipidemia). To introduce TBL to the residents, we split the pneumonia lesson into 2 parts. We provided a presentation on TBL methodology, and then administered the pneumonia IRAT and GRAT to 45 residents. The next week we reviewed TBL methodology, and then administered application exercises for pneumonia. Thereafter, we covered an entire lesson in 90 minutes following a standard TBL format (reading, IRAT, GRAT, facilitated discussion, and application exercises with facilitated discussion).¹⁰  We made 8 permanent resident team assignments with an equal distribution of each training year. The ambulatory topics were taught to 10 ambulatory week residents separated into 2 permanent teams according to their clinic assignment. From our initial experience, we determined that covering 5 IRAT/GRAT questions and 5 application exercises in 90 minutes was optimal, and that development of each lesson required 4 to 6 hours.

Faculty recruitment began with support from the department. We invited faculty to attend TBL, and met with interested faculty at division meetings to provide faculty development. Only a few had experience with TBL as learners or instructors, and none had used it in GME. Our faculty development activity was to assign a background TBL reading, and lead the faculty through an IRAT and GRAT with questions pertaining to the TBL method.

We captured approximately 30 faculty members in 5 divisions through this activity, and created a number of resources to assist faculty, including (1) TBL method overview; (2) TBL lesson prep tips; (3) example TBL lesson; and (4) facilitator tips (provided as online supplemental material). We selected all readings and created a question bank of existing review questions on each topic for faculty to use to develop IRAT questions and application exercises. Many faculty members received faculty development solely through reviewing these resources and coaching by the TBL champion during lesson delivery. We asked faculty to forward their lesson (1-page IRAT with 5 questions, and PowerPoint with 5 IRAT questions and 5 application exercises each followed by a teaching point slide) to the TBL champion 1 week in advance of each session for quality control of structure and content. No additional administrative time or extra funding was made available to faculty.

The surveys were declared exempt by the Albany Medical College Institutional Review Board.

We conducted anonymous surveys of faculty in November 2015 and residents in April 2016 on their experience with the TBL program. We designed the faculty survey and adapted the resident survey from 1 previously published.⁷  We did not test the surveys for validity. We tracked use of resources and expenditures to estimate overall costs. Feasibility was assessed through examination of faculty participation, resident preparedness, acceptability as reflected in the surveys, and resources required for the program.

By the end of the first year (July 2014), we had recruited 25 faculty for the hospital curriculum, and by July 2015, we had introduced all 75 hospital topics with a faculty of 64 and a lesson every week. By July 2016 we finished recruiting all 10 ambulatory faculty practice physicians, introduced all 60 ambulatory curriculum topics, and had grown the participating faculty to 74. Hospital faculty were repeating topics every 18 months. Ambulatory faculty taught 2 topics per year for 5 consecutive weeks during the academic half-day of the ambulatory week. We stored all lessons in electronic files, and sent them to faculty to modify and update as needed. This decreased preparation time.

The resident survey (64 of 89, 72% response rate) indicated that more than 50% of residents completed the assigned reading ahead of time (figure), and 78% (50 of 64) reported the amount of reading was manageable. Most residents felt that TBL helped them provide more confident patient care (86%, 55 of 64), and that they learned better from TBL than from lectures (77%, 49 of 64; table 1). The faculty survey (44 of 64, 69% response rate) showed that although TBL required added preparation time, 70% (31 of 44) of faculty felt it was manageable. Most felt the preparation materials were helpful (95%, 42 of 44), the faculty development adequate (73%, 32 of 44), and most preferred TBL to lectures (77%, 34 of 44; table 2).

Resources consisted of the efforts of the associate program director, who developed the TBL curriculum in addition to having other responsibilities, and who was supported by an administrative assistant (0.2 full-time equivalent). The 4-month development phase required approximately 16 weekly hours of effort. Current effort for scheduling, updating readings and the question bank, performing quality control, and assisting lesson delivery is 8 weekly hours. Participating faculty spent 4 to 6 hours initially developing a lesson, with less time to update it. Additional costs totaled $250 for supplies and $1,500 for textbooks and board review materials. There was no external funding or grant support.

We asked faculty to become competent in delivering TBL, although initially most were unfamiliar with this active learning method and had to devote preparation time to lesson development. Their positive reaction (as evidenced by participation and survey results) may come as a surprise, particularly considering that no additional administrative time or funding was provided. Residents had a similarly positive experience, and showed willingness to prepare by completing advance reading assignments.

Our work builds on previous efforts to implement TBL in GME. Balwan and colleagues⁹  described implementation of a 21-topic ambulatory IM curriculum, with lessons developed by 4 faculty champions and taught by 15 faculty physicians after a training session. Our work shows it is possible to train a larger faculty group to develop lessons themselves, and accept the associated time commitment, allowing for a comprehensive curriculum of 135 topics.

Faculty buy-in is important in successful TBL implementation.¹⁷  Key elements to achieve it included well-designed resource materials, flexible faculty development, and strong guidance from a TBL champion. We showed that TBL can fit within demanding resident and faculty schedules by using an extended 90-minute noon conference and an academic half-day. And though questions have been raised about residents' preparation for active learning,¹⁸  we showed that a majority completed readings in advance of lessons.

Limitations include implementation in a single IM program, use of surveys without validity evidence, self-reported outcomes, and use of estimates to determine lesson preparation time. Future studies should assess generalizability of broad TBL adaptation in GME, and identify best practices for faculty development and incentivizing residents to prepare through reading.

It is feasible to engage a sizeable faculty group in implementing a comprehensive medical knowledge curriculum within an IM residency, and achieve high faculty acceptance and learner satisfaction. Departmental support of a champion, flexible faculty development, and well-designed resource materials promoting specific lesson structures were important determinants of success.