The practices and strategies for implementing team-based learning in pre-clinical medical education: a systematic review

Abstract

Introduction

Team-Based Learning (TBL) is increasingly adopted in medical schools to promote active learning, particularly in the foundational pre-clinical years. Despite its structured framework, variations in implementation across institutions have led to limited clarity on consistent best practices. This systematic review aims to identify and synthesise how TBL is applied globally during the first two years of pre-clinical medical education.

Methods

A comprehensive search of SCOPUS, MEDLINE, PUBMED, WEB OF SCIENCE, and ERIC databases was conducted for English-language articles published between January 2000 and October 2024. Studies were included if they described the implementation, structure, and outcomes of TBL in the first two years of pre-registration medical programs. Comparative studies were excluded. Data were extracted and synthesised across TBL’s core phases (preparation, readiness assurance, and application) and reported using PRISMA guidelines, MERSQI for quality assessment, and Kirkpatrick’s model for learning outcomes.

Results

Of 1136 articles screened, 37 studies from 15 countries met the inclusion criteria. Most focused on first-year students, with 24 using classical TBL methods and 10 employing modified approaches. TBL was primarily delivered in person and applied across individual subjects, such as anatomy, physiology, and biochemistry, as well as in integrated curricula.

Conclusion

The findings demonstrate strong adherence to TBL’s core structure, with students valuing its active, student-centred approach and educators noting its efficiency. However, TBL remains largely supplementary, with only a few institutions (primarily in Australia) using it as a primary teaching method. Most studies use TBL in conjunction with other pedagogical approaches, and TBL sessions, when delivered, were reported to be, on average, 3 h in length, with distinct phases of readiness assurance and application. Limited reporting of the use of the ‘4S’ framework in application design and peer review was noted. This gradual adoption may reflect resource constraints inherent in transitioning fully to a TBL-based curriculum, or it may indicate a preference for retaining conventional teaching methods to align with country-specific requirements. Our review identifies a need for medical schools to align curriculum design with core TBL strategies supported by curriculum governance frameworks and faculty development programs,—to enable effective and sustainable scaling of TBL across pre-clinical programs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12909-026-08713-x.

Introduction

TBL is an active learning instructional strategy for small groups increasingly adopted globally in medical education [1, 2]. However, the principles for successful TBL design and delivery in the context of pre-clinical medical education are not consistently or universally adopted. Despite the defined, structured sequence of specific activities in the TBL process, the implementation of each TBL stage varies across medical programs. There is a need to determine which TBL design and implementation features are associated with improved learning outcomes in pre-clinical medical education and to formulate evidence-based recommendations for best practice.

TBL was developed as an alternative to traditional lecture-based formats and is designed to enhance active learning, higher-order thinking, immediate feedback, and teamwork, while being scalable and efficient [1, 2]. The TBL process is characterised by three key components: individual student advance preparation; individual and team readiness assurance tests with feedback; in-class team decision-based application and problem-solving activities with feedback [2–4]. Initially developed for business education, TBL has since been adapted for medical education to fit the scientific integration and clinical reasoning requirements of medical students [4–6]. Several studies have already published empirical evidence of TBL's comparative benefits over other pedagogies and improved learning outcomes for medical students learning [2, 7–9]. The individual components of the TBL process have been adapted by medical schools in diverse ways to suit different contexts, learning objectives, and curriculum structures [10–12]; however, the optimal design and delivery in medical education remain unknown.

TBL is frequently utilised during pre-clinical medical education. Pre-clinical medical education broadly refers to the initial phase of medical training, typically encompassing the first 1–2 years of a medical program. In this initial phase, there is a particular focus on foundational and biomedical sciences, such as anatomy, physiology, biochemistry, pathology, pharmacology, microbiology, immunology, etc., to provide a strong foundation for clinical education and practice [13–16].In addition, there is an increasing focus on developing early professional attributes, such as communication, ethical conduct, teamwork and advocacy [17–19] to shape students' clinical readiness and future professional identities. Therefore, implementing TBL in the pre-clinical years presents unique benefits but also pedagogical and logistical challenges, such as aligning within dense curricula, managing student and staff workloads, and ensuring suitable facilities and resources [19, 20].

Previous studies have reported that TBL enhances knowledge retention, student engagement and experience, professional skills development, and collaborative problem-solving when compared with other pedagogies used for pre-clinical medical education [3, 13, 14, 19]. However, given the disparate design and implementation of TBL across medical programs, it is not evident which TBL features are consistently associated with improved learning outcomes for medical students and greater student and faculty satisfaction. A systematic synthesis of the core design principles, implementation strategies, and contextual adaptations that support successful TBL delivery in this formative phase of medical education is necessary to enable medical programs to consistently and optimally implement TBL design principles into their educational context. This systematic review aims to address this gap by examining empirical studies of TBL implementation within the pre-clinical years of medical education. Specifically, the review synthesises evidence on TBL practices targeting first and second-year medical students in pre-registration medical programs. Through this review, we seek to generate practical, evidence-based recommendations to support educators, curriculum developers, and institutions in optimising TBL for foundational science education within medical schools. Therefore, the objectives of this research were to systematically review TBL interventions implemented in pre-registration medical programs (Years 1–2) globally to:

1. Identify key design elements and instructional strategies that characterise effective TBL use of TBL design principles (e.g., team formation, readiness assurance testing, application exercises, peer evaluation) in pre-clinical contexts

2. To evaluate the educational outcomes associated with TBL, including learner performance, team-based competencies, and student perceptions, mapped to the core stages of TBL.

Methods

This systematic review was reported (see Annexure Supplementary Table 1) in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [21]. The study protocol is registered with the PROSPERO registry CRD42024613102 [22].

Search strategy

An independent reviewer (VK) performed a systematic search in consultation with an information library technologist at the Australian National University (ANU). A search string of keywords was developed to systematically review titles, abstracts, and keywords. The authors inductively derived the keywords after completing an initial literature search on TBL in medical education. In our search, we defined pre-registration or pre-clinical medical students as those enrolled in the initial pre-clinical years of a medical course that ultimately awards a medical degree, thereby qualifying them as graduate doctors. As medical programs globally vary in course level (undergraduate or postgraduate) and duration, we included any studies with first- and/or second-year medical students enrolled in a pre-registration medical degree.

The medical and educational electronic databases SCOPUS, MEDLINE, PUBMED, WEB OF Science, and ERIC were searched until 24 October 2024. Additional records were identified by scanning the reference list of published reviews and retrieved studies. As TBL in health professions education commenced in the 2000 s, we limited our search to original articles published from 1 st January 2000 to 24th October 2024. The search terms included the following: ‘Team Based Learning’, ‘TBL’, ‘Medical Education’, ‘Medical Schools’, ‘Medical Students’, ‘Medical education undergraduate’, and ‘Medical education postgraduate’. The details of the electronic search strategy are provided in Annexure (Supplementary Table 2).

Study selection and inclusion/exclusion criteria

Eligible studies that met the following pre-defined PICOS inclusion and exclusion criteria generated from the review protocol were included.

		Inclusion	Exclusion
Population	Medical students enrolled in pre-registration medicine programs or courses	Medical(allopathic) Students	Non-allopathic medicine students
		First and/or second year of undergraduate or postgraduate medical schools	Beyond the second year of medical school, students' studies are based on clerkship years
		Pre-registration medicine students	Post-registration medical students
Intervention	Medical curriculum using TBL	At least one step of TBL, as defined conventionally, is provided	Study uses non-TBL-based pedagogy
Comparison	No comparison. As TBL's pedagogical effectiveness is already well established in the medical education literature, this review aimed to synthesise TBL implementation practices in pre-clinical medicine, therefore, comparative effectiveness studies evaluating TBL against other pedagogical approaches (e.g., PBL or lecture-based designs) were excluded	TBL only	Exclude if TBL is compared to any other pedagogy, e.g. traditional lectures, problem-based learning etc
Outcomes	Medical student (primary) and/or medical school faculty (secondary) perceptions of TBL effectiveness	Medical student outcomes primarily focus on knowledge retention, student perception, well-being, engagement, and academic performance	Exclude studies that did not report any student or staff outcomes related to the implementation of TBL
Study	International primary research only	Primary research	Non-primary research
		Research published in English between 2000 and 2024	Any non-English research published before 2000 or after 2024
		Peer-reviewed articles in indexed journals	Non-peer-reviewed articles, including preprints, published in a non-indexed journal

Literature screening

Two investigators (DA, VK) independently screened the titles and abstracts of all retrieved articles using the predefined inclusion and exclusion criteria. While screening titles and abstracts, we followed a two-step process to identify articles implementing the TBL pedagogical approach in medical education. First, we searched for the phrase ‘team-based learning’ as a single term, rather than the individual words ‘team’, ‘based’, and ‘learning’ in separate sentences, in the abstract. Secondly, we also scanned for articles that described the core elements of TBL pedagogy related to the structured steps of pre-learning or advanced preparation, readiness assurance tests/phase (RAT), application exercises, and peer review, even if TBL was not explicitly mentioned as a pedagogy. All reviewers (AW, KV, DA, VK) jointly resolved any conflicts at the title and abstract review by consensus. Studies then progressed to the full-text review stage, which was independently conducted by two reviewers (DA, VK). Conflicts at this stage were similarly resolved by consensus among all reviewers (AW, KV, DA, VK).

Data extraction

Two investigators (DA and VK) independently extracted the data from all included studies. The two investigators then cross-checked the data extraction results to identify any discrepancies. In the event of a conflict, a consensus was reached through a conference among all reviewers (AW, KV, DA, VK). The following information was extracted from all eligible studies:

• First author’s name, study title, publication year, host institution name, location, study design, medicine course length, medicine course description, medicine course cohort size, course main pedagogy,

• TBL module topic, TBL module details, number of TBL participants, TBL delivery mode (in-person/online/hybrid), TBL duration, TBL team allocation method, TBL team size, Pre-learning format(s), individual RAT format (i-RAT), team RAT format (t-RAT), TBL application exercises format, use of ‘4S’ framework(i.e. significant problem, same problem, specific choice, simultaneous reporting), TBL group output format, TBL peer evaluation format.

• TBL key outcomes across quantitative and qualitative headings, final exam score, the study author's main conclusion, and study limitations.

Critical appraisal and quality assessment

Study quality was independently assessed by two reviewers (DA, VK) using the Medical Education Research Study Quality Instrument (MERSQI), a validated tool originally developed and widely used to appraise methodological quality in medical education [23, 24]. The MERSQI tool is a ten-item scale that appraises studies across six domains: study design, sampling institutions, sampling response rate, data type (subjective or objective), instrument validity, data analysis sophistication and appropriateness and outcome [23]. Each domain is scored from 0 to 3, yielding a maximum MERSQI score of 18 across the six domains, with higher scores indicating greater methodological rigour and quality [23, 24]. While total MERSQI scores are routinely reported in the literature, there are no universally accepted cutoffs to differentiate between low-quality and high-quality studies. Although a few studies suggest cutoff scores of below 9 for low quality and above 12 for high quality, interpretations of scores vary. We therefore report total MERSQI scores, with no studies excluded for low scores.

Additionally, as part of the quality assessment, we separately evaluated each study against the four levels of learning outlined in the Kirkpatrick learning model [25]. Kirkpatrick's model has been extensively used in medical education and elsewhere to assess learners and the impact of teaching pedagogies, such as TBL and curricula, by mapping them against four learning levels [26–28]. In our review, two reviewers (DA, VK) allocated each study to one of four Kirkpatrick learning levels with level I for papers assessing learner reactions to TBL, level II assessing learner’s knowledge attitudes, beliefs to TBL, while level III was used for papers describing behaviour and practises adopted by TBL participants, and finally level IV for papers where results or specific learning outcomes related to TBL participation were provided. Disagreements were addressed by initial discussion among the two investigators (DA, VK) followed by consensus among all authors.

Data synthesis

Our review synthesised results for key TBL practices, including student and staff feedback where available, categorised by TBL stage (pre-learning, RAT, application activities) and reported in narrative format. We also extracted pre-TBL coordination data across TBL duration, group allocation method, and team size to provide necessary context for TBL administrative preparation. Furthermore, we collected information on the main teaching pedagogy used in medical schools to provide context for how TBL fits within it. Although we initially planned a pooled meta-analysis of TBL effectiveness in knowledge, skill and attitude acquisition using assessment performance as the primary endpoint, a formal meta-analysis could not be conducted due to extensive heterogeneity across studies. Studies varied considerably in their pre-learning, RATs, and application-stage designs, as well as in reported outcomes (including summative exam scores, readiness test results, motivation, and satisfaction surveys) and the timing of those assessments.

Results

Search results

Of the 1,136 screened studies, 37 were finally included in this analysis, as shown in the PRISMA flow diagram (Fig. 1). These studies were conducted across 15 different countries, with the United States of America (USA) accounting for approximately half of the publications. All studies, except for two from Australia, were conducted in medical schools offering pre-registration undergraduate medical degrees. Of the 37 studies, 24 focused exclusively on student data, while 3 included staff and student data. Most studies (24) captured the experiences of first-year students in TBL, seven focused on second-year students, and six included experiences from both years. Regarding the TBL pedagogy, 24 studies reported the use of classical TBL methods, while 10 presented modified TBL approaches. The mode of TBL delivery was predominantly in-person (32 studies), with four studies conducting hybrid (online/in-person) sessions, and one study entirely online.

Fig. 1.

PRISMA flow diagram

Our narrative review results are presented first for the included studies, in which TBL is contextualised as a pedagogy or instructional modality against the main pedagogy already in use in medical schools for pre-clinical teaching. A summary of TBL modules by subject and discipline follows. We then summarise key teaching practices reported across core TBL steps in pre-clinical education globally.

Medical school pedagogy and TBL context

As medical schools globally vary in their pedagogical design for the pre-clinical years, our review captured the main pedagogical approaches used by the included studies (when reported). It contextualised whether TBL was the primary teaching pedagogy or used in conjunction with other instructional methods. Most USA-based medical school studies have predominantly indicated the use of a regional block structure, combining lectures with active, team-based practical exercises [18, 29–31]. Additionally, USA medical schools employ a variety of teaching modalities, including traditional and online lectures, dissection labs, and computer-aided instruction, with a strong emphasis on TBL throughout the first year [29, 32–34] In contrast, three Australian-based studies, all in one medical school, reported that they often used a hybrid Problem-Based Learning (PBL) curriculum, transitioning students gradually to TBL after initial PBL sessions [4, 15, 35]. A Colombian study noted that the curriculum was characterised by extensive lectures, gross anatomy laboratories, and clinically oriented case discussions [36]. Meanwhile, TBL is a core instructional strategy in studies conducted in Saudi Arabia [37] and Korea [38]. While a Dutch study [39] reported a hybrid structure combining traditional lectures with small-group instruction, two Singapore-based medical schools [40, 41] primarily used TBL as the main pedagogy, which involves team-taught sessions led by facilitators and content experts, emphasising pedagogical management and subject matter expertise.

Summary of TBL delivery outlines across subjects, country and course year

Our analysis revealed that TBL has been implemented across various individualpre-clinicalsubjects, including anatomy, physiology, and biochemistry, in pre-clinical medical education, with studies also utilising TBL for integrated learning combining subjects/disciplines. We provide an overview in Table 1. of the TBL delivery outline reported by subject/topic, country, and course year.

Table 1.

Provides a summarised pre-clinical TBL delivery outline reported by subject, medical school location, and course year (n=37)

Subject/Topic	Location	Institution Author, (Year)	Course year	TBL Delivery Outline
Anatomy	Colombia	Universidad del Norte (Martinez et al.,2014) [36]	1	Two TBLs were implemented as a pilot project in the human musculoskeletal anatomy course by combining TBL with iPad applications. TBL replaced two traditional lectures with pre-readings on osteology (skull) and the thigh muscles. One iPad given to each pair of students within each of the 18 teams, with two iPad applications (The Skeleton System Pro III (3D4Medical) and The Muscle System Pro III (3D4Medical) selected based on visual image quality and anatomical content. The TBL aimed to identify anatomical landmarks and use quiz modules. Graded i-RAT &t-RAT contributed 20% (1 point) to the mid-semester exam. The use of the 4S framework and peer evaluation was not reported
	Oman	Sultan Qaboos University (Inuwa,2012) [42]	1	A series of 11 modified two-hour TBL sessions replaced lectures in the ‘Introduction to Anatomy’ course over two semesters of the academic year. The graded TBL accounted for 20% of the overall course grade, with TBL grades distributed as follows: 16% (i-RAT), 3% (t-RAT), and 1% (peer evaluation). Separate gender TBL groups were provided with pre-reading materials, which included recommended textbook readings, prosected specimens in the anatomy laboratory, and an image bank created on the learning management platform. Peer evaluation was conducted at the end of the course to assess post-TBL final reflections; however, the use of the 4S framework was not reported
	UAE	University of Sharjah (Eladl et al.,2020) [43]	2	A modified practical-based team-based Learning (PTBL) activity created to support and prepare for formative anatomy assessments. PBTL simulated an objective structured practical examination (OSPE) with students rotating through practical stations: individually for i-RAT and in groups for t-RAT, where students discussed their answers together and collectively arrived at a single answer. The module concluded with the facilitator discussing the answers through photographic projections of the classroom stations. The use of the 4S framework and peer evaluation was not reported
	USA	Wright State University School of Medicine (Nieder et al.,2005) [29]	1	TBL was implemented in the ‘Human Structure’ course, which comprised 140 h of instruction (live embryology lectures, online gross anatomy lectures, dissection lab, and TBL). Twelve 2-h TBL sessions, including pre-reading, RAT, and application activities, are scheduled within a nine-week anatomy course. The course was divided into thorax, back, and upper extremity units, covering topics such as embryology, anatomy, and radiology. One week before each TBL, specific assignments were posted on the course website, including live and online lectures, textbook readings, computer exercises in sectional anatomy and imaging and dissection lab sessions. TBL i-RAT and t-RAT followed by clinical application cases. Peer evaluation is conducted three times during the course, followed by an end-of-course student-faculty TBL evaluation; however, the use of the 4S framework is not reported
	USA	New Jersey Medical School (Vasan et al., 2009) [30]	1	TBL is used for a 19-week human anatomy course, which is divided into three units: thorax, back, and upper extremity (6 weeks); head and neck (5 weeks); and abdomen, pelvis, perineum, and lower extremity (8 weeks). Over 60% of the course time is spent on dissections, while embryology and clinical correlates are presented as lectures. Students were provided with textbook readings and faculty-developed notes on clinical conditions that require knowledge of anatomy, one week before the TBL. Peer evaluation was conducted after the unit examination for internal use only by academics and low-scoring peers, who provided counselling. Use of the 4S framework was not reported in the study
	USA	New Jersey Medical School (Vasan et al.,2011) [32]	1	19-week Human Anatomy course taught using TBL, with the course divided across three anatomical units and 55–60% of the course time spent on the cadaveric dissection module, while 30–33% of the course time is TBL, with the remaining on embryology and clinical correlates lectures. TBL pre-reading comprised anatomy textbook chapters (Clinically Oriented Anatomy or Grey's Anatomy for Students). Additionally, faculty developed ‘content-specific discussion notes’ for pre-reading, made using other didactic material and on the assigned cadaver dissection. Graded peer review (worth 5% of the final course score) was collected after the unit examination, and low-scoring students were identified proactively and received counselling from staff. The study did not report using the 4S framework
	USA	University of Alabama School of Medicine (UASOM) (Bass et al.,2018) [14]	1	A six-week GI module integrating basic science disciplines that covers gross anatomy and radiology of the abdominal cavity through lectures, vodcasts, three cadaveric dissections, two TBL exercises, a computed tomography (CT) radiology small-group exercise, and a hands-on ultrasound session where students perform an ultrasound of the gallbladder and a FAST (focused assessment with sonography in trauma) exam. TBL supported students in applying knowledge of the GI tract, posterior abdominal wall, and abdominal neurovasculature to diagnose and treat abdominal disease. Textbook (Clinically Oriented Anatomy, seventh edition) pre-readings were provided four days before TBL, followed by i-RAT/t-RAT and TBL application exercises based on a single patient diagnosed with pancreatic cancer, allowing students to follow the patient through clinical presentation, radiologic examination, surgical intervention, and palliative care. Use of the 4S framework and peer evaluation was not reported
	USA	University of Texas Southwestern (UTSW) (Prange-Kiel et al.,2016) [33]	1	Two TBL sessions were developed for the Human Structure course, which integrates human anatomy, embryology, and basic radiology. The course uses traditional lectures, video lectures, cadaver-based dissections, osteology exercises, online learning modules, and two TBL sessions. Anatomy head and neck lectures and dissection sessions are coupled with an introductory lecture on basic radiographic images, covering the fundamental concepts of radiographs, CTs, and MRIs. Of the two TBL sessions, this study’s TBL covers the embryological development of the head and neck region, anatomy, and radiographic presentation. The TBL provided three cases of malformations of the head and neck regions as application exercises, using i-RAT and t-RAT, which were held online via a computer-based testing tool (ExamSoft). Use of the 4S framework and peer evaluation was not reported
	USA	Tulane University (TU) School of Medicine (Williams et al.,2019) [34]	1	An 11-week human gross, developmental, and radiological-based anatomy course taught in four regional blocks consisting of weekly lectures (9 h) and team-based dissection (10 h) Five clinically oriented radiology TBL sessions incorporating common critical findings in medical practice were conducted during the 11-week course. Before each TBL, TU radiologists trained teaching assistants to read radiographs and interpret the clinical implications. Students were assigned pre-TBL work, including videos, reading assignments, and practice questions, and worked on four application exercises during the TBL. Use of the 4S framework and peer evaluation was not reported
Biochemistry	India	PSG Institute of Medical Sciences and Research (Govindarajan et al., 2022) [44]	1	Three online TBL sessions are planned on nucleotide chemistry and metabolism, regulation of gene expression, and liver function tests. Pre-readings were shared one week in advance via Google Classroom. Ten MCQ-based i-RAT& t-RAT questions followed by a team application exercise framed using the 4S principles. Following the final TBL, students provided anonymous feedback on their TBL experience, peer review, and facilitator discussion using Likert-scale questions
	Japan	Hyogo College of Medicine (Eguchi et al., 2021) [45]	1	TBL focused on nutritional metabolism of fasting and satiation with students' prior topic knowledge covering carbohydrate structure, glucose metabolism, and beta oxidation. TBL pre-learning provided via LMS on glucose metabolism one week before TBL, and included advanced questions related to fasting, glucose metabolism, and lipid consumption during TBL. TBL included 15 MCQs for i-RAT and t-RAT, with a peer evaluation conducted after the final TBL to assess attitude, knowledge (preparation), coordination, and active participation. Use of the 4S framework was not reported
	Malaysia	Universiti Kebangsaan Malaysia (Ismail, 2016) [17]		A pilot biochemistry study using TBL to teach mutation and mutational analysis involved students providing readings and compulsory YouTube lectures a week before TBL sessions. Subsequently, students submitted a mind map via email. Students then used individual mind maps and other online resources during the t-RAT. Use of the 4S framework and peer evaluation was not reported
Dyslipidemia	USA	Oakland University William Beaumont (OUWB) (Megee et al.,2024) [46]	1	Dyslipidaemia TBL on themes of hyperlipidaemia, nutrition, and the pathogenesis of atherosclerosis was presented in week 5 of a 6-week cardiovascular system course, the third of four organ system courses of an 18-week semester. TBL was developed using a backwards design to devise preparatory materials and application exercises that promote experiential transfer of knowledge and skills from foundational science disciplines crucial to dyslipidaemia patient care. Pre-reading was provided 1 week prior via the LMS, including consolidated PowerPoint lecture slides across 4 weeks and clinical care recommendations. TBL sessions were presented online (2021–2022) via video conference, with i-RAT/t-RAT held using InteDashboard, and later in person (2023) without web-based software. Team application simulated patient interaction activity and was developed using 4S framework. Peer evaluation was not reported
Hematology	USA	Brigham and Women's Hospital Learning (Langer et al.,2020) [47]	2	TBL module on haemolytic anaemia and haemoglobin structure and function created, emphasising clinical practice and underlying pathophysiology Pre-readings (3 h) included textbook chapters provided one week before TBL and developed by the facilitators (two haematology/oncology fellows and two haematology/oncology faculty members). Use of the 4S framework and peer evaluation was not reported
Infectious Diseases (ID and Immunology)	USA	Cooper Medical School of Rowan University (Carrasco et al., 2019) [18]	1	4 Week ID course integrating microbiology and virology, with weekly graded TBL comprising RAT, and two application questions. Each graded TBL received an i-RAT (33.33%) and a t-RAT (66.33%), contributing 3% to the overall course grade. TBL was scheduled at the start of the week, with the final TBL held days before the final multiple-choice examination. Use of the 4S framework and peer evaluation was not reported
	USA	University of Texas Health Science Center (Wiley et al., 2019) [48]	1	TBL module (week-long) on Human Papillomavirus (HPV), integrating foundational virology and immunology with clinical application. Module included lectures (day one,1 h) on HPV virology and immunology, followed by a clinical scenario video on HPV vaccination and a thematic patient case on HPV-related cancers On day 5, i-RAT and t-RAT were held, followed by a 40-min group application exercise using various resources. The session ended with an HPV immunisation and a discussion of communication strategies. The total module time was approximately three hours. Use of the 4S framework and peer evaluation was not reported
Integrated Basic Sciences	Australia	University of Sydney medical program (Burgess et al., 2019) [35]	1	Three TBL sessions were held as part of the Musculoskeletal, Respiratory, or Cardiovascular block, once per week for 2 h. Pre-reading consisted of essential online readings and/or pre-recorded lectures, and the TBL utilised MCQ-based i-RAT/t-RAT with a single best answer, concluding with a clinical application case. TBL facilitator teams remained constant, with nine academic clinicians participating, depending on their speciality: three Rheumatologists, three Respiratory physicians, and three Cardiologists. Use of the 4S framework and peer evaluation was not reported
	Australia	University of Sydney medical program (Burgess et al., 2020) [4]	1 and 2	27 TBL classes held across years one (11 TBL) and two (16) teaching blocks respectively comprising ‘Orientation and Foundations (3 TBLs), Musculoskeletal sciences (4 TBLs), and Respiratory sciences (4 TBLs)’ and ‘Neurosciences (6 TBLs), Endocrine, Nutrition, Sexual Health, HIV (6 TBLs), and Renal-Urology (4 TBLs)’. Weekly TBL classes across each teaching block with the TBL facilitation team comprising one clinician, one basic scientist, and one medical registrar. TBL comprised 1–2 h of pre-reading and/or prerecorded lectures followed by online i-RAT via Blackboard, t-RAT via Kuracloud and clinical application, all done in 2.5 h. Use of the 4S framework and peer evaluation was not reported
	Australia	University of Sydney medical program (Burgess et al.,2021) [15]	2	Five TBL classes (2.5 h in duration) held simultaneously with the TBL team comprising three facilitators (medical consultant, registrar and basic scientist). Students provided peer review on professional learning behaviours within TBL teams twice yearly using the online tool ‘Sparkplus’. Peer review/evaluation was undertaken after the first two teaching blocks (Neurosciences and Endocrinology),comprising12 TBL sessions, and again after the completion of another two teaching blocks (Renal/Urology and Gastroenterology), comprising 8 TBL sessions. Use of the 4S framework was not reported
	Kingdom of Saudi Arabia	Alfaisal University College of Medicine (Obad et.al 2016) [37]	1	Anatomy and Physiology are taught across a five-year course and integrated with function, abnormality and pathophysiology. Year 1 content was mainly taught using 2-h weekly TBL sessions, practical sessions and didactic lectures. Separate gender TBL teams were provided pre-reading one week in advance via LMS(Moodle) and included recommended textbooks, a prosected lab. Specimens, videos, and image bank. Graded i-RAT (50%), t-RAT (30%), and clinical application (20%) contribute 15% to the overall course/block grade. RATs were conducted using an audience response system(clickers) from Turning Technologies, with a single clicker provided for t-RAT. Application activity included clinical cases. Use of the 4S framework and peer evaluation was not reported
	Singapore	Lee Kong Chian School of Medicine (Rotgans et al.,2019) [40]	1 and 2	TBL sessions were held across six teaching blocks in Year 1, covering the introduction to medical sciences, the renal and endocrine systems, and the cardiorespiratory systems. In year two, blocks included neurology, eye and ENT, reproductive medicine, and child health. Each TBL is facilitated by a two-person team of a content expert (scientist or clinician) and a facilitator. Use of the 4S framework and peer evaluation was not reported
	USA	Uniformed Services University of the Health Sciences (Knollmann-Ritschel et al., 2015) [49]	1	A modified TBL approach used in the ‘Integrated Fundamentals’ course comprises 14 individual courses encompassing basic and clinical sciences, with TBLs modified to use concept maps instead of MCQs in RAT. i-RAT required an individual concept map on ‘inflammation’ in 20 min, while t-RAT activity allocated 75 min for a group concept map using 30 pre-identified concepts. In-session feedback was provided, and independent graders evaluated individual and group maps on a pass/fail basis, focusing on content accuracy, concept interlinkages, and completeness. Use of the 4S framework and peer evaluation was not reported
	USA	Nova Southeastern University Dr. Kiran C. Patel College of Allopathic Medicine (Beckler et al., 2021) [31]		The immunology TBL module focuses on immune mechanisms and hypersensitivity to renal disease, developed as part of a 12-week cardiovascular, pulmonary, and renal (CPR) block taught using a multi-modal instructional approach, including TBL with weekly formative quizzes. Two TBL approaches were described: one involving in-person delivery supplemented by an LMS, and a second approach during COVID-19, which utilised online TBL delivery via web software (Zoom and InteDashboard). Students were provided textbook chapter readings one week in advance. During TBL, MCQS were given for i-RAT/t-RAT, culminating in two renal injury-based clinical application cases on haemolytic uremic syndrome and the immunopathogenesis of renal diseases. Use of the 4S framework and peer evaluation was not reported
Medical Ethics	Republic of Korea	Chonnam National University Medical School (Chung et al.,2009) [50]	1	Four two-hour-long TBL sessions were used for teaching medical ethics education. First TBL was an instructional session with the remaining three TBLs on principles of medical ethics (autonomy, non-maleficence, beneficence and justice), how to make ethical decisions and the patient-doctor relationship. Pre-reading, provided 2–3 days before TBL, comprised study guides and textbook readings. During the TBL, i-RAT and t-RAT were held using MCQ, followed by two MCQ-based application exercises on clinical medical ethics. Use of the 4S framework and peer evaluation was not reported
Pharma-cology	Korea	Eulji University School of Medicine(Kim et al., 2020) [38]		1 week-long integrated pharmacology course (ICP) with morning lectures and afternoon TBLs for the following topics: antibiotics and anti-cancer drugs; pharmacokinetics and pharmacodynamics; endocrine systems; circulatory systems; and central nervous systems. Faculty members facilitate pharmacology sessions for clinicians. ICP TBL sessions followed preparation, 10-min MCQ-based i-RAT, open-book t-RAT, non-4S-based application activity and graded peer evaluation (worth 10%) where students selected best and worst team members based on communication skills, problem-solving abilities, and participation
	Lebanon	American University of Beirut Faculty of Medicine (Zgheib et al., 2010) [51]	2	TBL was used in two laboratory sessions: the first on drug metabolism and pharmacogenetics, which included three cases (18 questions), and the second session on pharmacokinetics/pharmacodynamics (PK/PD), which included five cases (6 questions). TBL sessions were held 2–3 days after the lecturer delivered two didactic PowerPoint presentations about each topic. Use of the 4S framework and peer evaluation was not reported
	USA	Cooper Medical School of Rowan University (Carrasco et al., 2021) [52]	1	TBL exercise is reported for block 3 (of 4) of a 16-week basic science foundational course [4 weeks/block]. While the third block overall comprised 22 h of lecture and 20 h of application sessions, a 2-h TBL exercise on pharmacology was held covering 5 lecture hours and 2 h of application on the autonomic nervous system. Use of the 4S framework and peer evaluation was not reported
Physiology	India	Manipal Academy of Higher Education Nayak et al.,2020) [53]	1	Modified TBL approach developed using a clinical case-based readiness assurance phase (CBRAP) for physiology revision. Physiology was taught using a system-based approach, for which four separate CBRAPS were made, including clinical case-based MCQs across four systems (haematology, muscle physiology, cardiovascular physiology, and renal physiology). Students received textbook prereading, PowerPoint presentations, and recorded video lectures three to four weeks before the CBRAP. Each TBL with CBRAP required students to choose appropriate clinical features, pathophysiology, and lab findings to diagnose a clinical problem. Student scores are determined by an equally weighted average of the i-RAT and t-RAT scores, with underperforming students identified and assigned for academic mentoring. Use of the 4S framework and peer evaluation was not reported
	Japan	Gunma University Graduate School of Medicine, (Fujiwara et al., 2023.) [16]	2	A pilot online TBL module on renal physiology was developed during the COVID-19 pandemic, with students conducting onsite practical exercises related to renal function tests while the class participated online. The online t-RAT discussions were held in break-out rooms. Subsequently, a clinical case (polyuria caused by pituitary diabetes insipidus) with laboratory test results, including a hypertonic saline stress test, was presented, followed by teacher feedback. Following the discussion, students were assigned additional homework. Use of the 4S framework and peer evaluation was not reported
	USA	Duke University School of Medicine, (Carbrey et al.,2015) [54]	1	Year 1 included an integrated basic science course in which students participated in three TBL sessions, spaced 4 weeks apart, with in-class i-RAT and t-RAT assessments. Subsequently, the current study was conducted during the physiology segment of the second integrated basic science course, 'Normal Body’. Use of the 4S framework and peer evaluation was not reported
Obesity and Nutrition	USA	Case Western Reserve University School of Medicine (Olson et al., 2023) [55]	1	During Block 3 of the obesity pathogenesis and treatment curriculum, a two-part TBL seminar was implemented on the gastrointestinal system, biochemistry, metabolism, and nutrition. Students were familiar with the TBL format, having participated in 13 other TBLs throughout the first year of medical school. Use of the 4S framework and peer evaluation was not reported
Other Subjects/topics	Austria	Medical University of Vienna (Wiener et al.,2009) [56]	1	TBL was first offered to all first-year students in the first semester, block 3, ‘From Molecule to Cell,’ as a single, non-obligatory 2-h exercise to illustrate the value and dynamics of a learning team. In semester two, TBL is an optional, intensive course covering the material of block 3, ‘Functional Systems and Biological Regulation.’ Eight parallel courses, each with six intensive 2-h sessions over three days, are held two weeks before the final summative integrative exam at the end of the first year. Use of the 4S framework and peer evaluation was not reported
	Netherlands	Amsterdam University Medical Centre (Roossien et al.,2022) [39]	2	TBL was used as part of a four-week hybrid curriculum (including lectures and small group teaching) for the course ‘Observing, Thinking, Acting 2’. The TBL study topic was ‘Post-Traumatic Stress Disorder’ (PTSD), to measure student knowledge before, during, and after TBL. Pre- and post-tests were held one week (before/after TBL) and comprised an essay assignment on PTSD. At the same time, student knowledge development during each TBL phase was measured using a concept recall assignment (CRA), in which students were asked to write down all keywords related to PTSD. Use of the 4S framework and peer evaluation was not reported
	Singapore	Nanyang Technological University, (Koh et al.,2019) [41]	1 & 2	67 TBL sessions in year 1 and 75 in year 2, covering four teaching blocks per year. Year 1 blocks included ‘Introduction to Medical Sciences’, ‘Cardiorespiratory’, ‘Renal and Endocrine’, and ‘Musculoskeletal and Skin’.While year 2 included ‘Gastrointestinal, Blood and Infection’, ‘Neuro Ear-Nose-Throat and Eyes’, ‘Reproduction Medicine and Child Health’, and ‘Mental Health, Ageing and Family Medicine’. Two TBLs were held per week, with students receiving prior core pre-reading as recorded lectures (e.g., voice-over PowerPoint files), while supplementary materials were mostly texts (e.g., PDFfiles of journal articles) or book chapters. Use of the 4S framework and peer evaluation was not reported
	USA	Creighton University (Smith et al.,2023) [57]	1	Year one orientation week activity developed using a four-hour Case-Based Learning (CBL) -TBL session, utilising a fictional narrative of a new medical student (non-traditional female of colour) encountering the medical school's academic environment. Narrative included challenges related to medical education and practice across peer classroom interaction and references to resident and attending physician interactions in clinical environments. General maladaptive responses to stressors, including suicidal ideation, depression, anxiety, and substance use and abuse. CBL activity held on day two and TBL on day three of orientation week. Use of the 4S framework and peer evaluation was not reported
	USA	Wright State University Boonshoft School of Medicine Deardorff et al., 2014) [58]	1 & 2	Pre-clinical TBLs (> 25/year) with 2–3 h-long sessions, of which the group application activity lasts for 80–120 min. Group Activity comprising clinically based case scenarios linked with MCQs created by a content expert and reviewed by a faculty member is conducted in TBL sessions during which reference material is permitted. Graded i-RAT,t-RAT, and peer evaluation contributed to the final course grade. Use of the 4S framework was not reported
	USA	Oakland University William Beaumont School of Medicine Lerchenfeldt et al.2023) [20]	1 & 2	Orientation week TBL activity designed for year one students to learn about TBL pedagogy and how to provide appropriate/valuable peer feedback using Likert and narrative questions. Students are required to state the single most valuable contribution of a team member and the most important area for improvement. Faculty feedback on the quality of peer feedback provided. Similar peer evaluation/feedback is repeated at term end during the first two years. Use of the 4S framework was not reported

Summary of key practices across TBL implementation

Summary results follow in subsections for each TBL stage (Fig. 2).

Fig. 2.

Shows the sequencing of team-based learning activities in pre-clinical medical education

I. TBL coordination: group allocation, TBL duration and team size

The coordination of TBL varied widely in terms of duration, group allocation methods, and team sizes; however, common themes were identified, as shown in Table 2. Additionally, we also analysed individual studies, and present preliminary TBL coordination details organised by subject, location, TBL duration, team allocation and team size (see Annexure Supplementary Table 3).

Table 2.

Shows key practices around TBL duration, group allocation and team size(n = 37)

	Common Theme identified	Description of theme	Studies
TBL Duration	Short Programs	Sessions last around 55 min to 2 h, held weekly or over a few days	17,29,31,34,55
	Moderate Programs	Multiple TBL sessions within an organ system block, ranging from 2 to 3 h per TBL session	4,15,40,43,54
	Long Programs	Extended durations, such as 300 min	16,42,50
Group Allocation	Random Allocation	Teams formed randomly to ensure diversity and prevent preexisting subgroups	31,40,48,54,55
	Stratified Allocation	Students were stratified based on factors like previous exposure to science courses, gender, and academic performance	34,35,36,53
	Diverse Backgrounds	Teams formed to ensure a mix of backgrounds, including gender, ethnicity, and academic experience	4,43,47,58
Team Size	Recommended size (5–7)	Teams with 5 to 7 students	4,15,17,29,34,35,36,43 50,53,54,55,58 59
		Teams with 6 to 7 students	4,31,36 38 43,48
	Above-average team size	8- 10 students	33, 37 33

II. TBL Pre-learning

Pre-learning was almost universally implemented in the included studies, although various modalities were employed. Most studies sought to provide structured and cognitively engaging pre-learning to prepare students for collaborative work, incorporating a mix of videos, reading assignments, practice questions, and other preparatory materials, as shown in Table 3 below.

Table 3.

Pre-learning approaches: consolidated summary Table(n = 37)

Pre-Learning Type	Preparation Time	Released	Key Pre learning Characteristics	Studies
Textbook Readings	1–2 h avg	1 week prior	Specified chapters or subject topics linked to learning objectives	4,14, 15,30, 38,43,44, 47,59
Pre-recorded Lectures and videos	1–2 h average(avg)	1 week prior	Prerecorded lectures primarily consisted of voice-over PowerPoint files, recordings of didactic lectures, YouTube lectures, Vodcast, and other pre-reading materials, including pre-TBL practice questions	14, 17,36,42,44,52 57
Digital resources	1–3 h avg	1 week prior	PDFs containing journal articles, objectives, quizzes, textbook chapter readings, facilitator compiled content-specific notes, and anatomy image bank posted digitally (e.g. using LMS or Google Classroom)	16,31,33, 37,54,57
Formative Tests			Pre-TBL practice quizzes and knowledge tests are provided along with pre-recorded TBL lectures Pre-TBL individual preparatory material required that students use later in group application activities	17, 29,36, 38,56 38,
Application/ CBL-style Preparation	2–4 h	1–2 days to 1 week before TBL	CBL fictional patient narratives (10 pages), mind maps, and conceptual frameworks provided	17,29,58
Blended (Mixed Format)	2–4 h on average	2 days to 1 week	Combines textbook chapter readings, faculty-developed notes, lab prep, or dissection recorded lectures, radiographic images and pre-TBL tasks	33,34,35,40,52
Others	N/A	N/A	NA	32, 41,53,

III. TBL readiness assurance phase

One of TBL’s main pedagogical aspects that influences learner behaviour and outcomes is the use of readiness assurance tests, which are first administered individually (i-RAT and then in a team setting (t-RAT). Across the reviewed studies, i-RATs were almost universally implemented using multiple-choice questions (MCQs) with a single best answer to assess foundational knowledge before team collaboration (Fig. 3) [33, 35, 36, 57]. Most i-RATs were conducted in a closed-book format [29, 37, 54, 57], typically comprising 5 to 30 items, and 10 to 30 min duration [14, 38, 50]. These assessments were administered in person or via online platforms such as the LMS or examination management system [4, 33, 44].

Fig. 3.

Summarises key i-RAT practices in pre-clinical education (n = 37 studies)

The t-RATs, as shown in Fig. 4, almost always used the same MCQs as the i-RAT and followed it immediately [15, 33, 57]. The team discussion phase fostered consensus building and often incorporated tools such as the Immediate Feedback Assessment Technique (IF-AT) delivered via online platforms or paper-based scratch cards [14, 33, 36, 57], which allowed teams to receive instant confirmation of responses. Across studies, immediate feedback was most often provided in-session, during the t-RAT through answer-reveal mechanisms and peer discussion. In addition, many studies reported brief facilitator-led clarification provided immediately following the t-RAT, often in the form of targeted discussion or short mini-lectures to address common misconceptions [35, 38, 47, 54]. In contrast, fewer studies described written or delayed faculty feedback provided after completion of the TBL session. Appeal processes were allowed in some programs, enabling students to contest group answers or unclear items [33, 36, 46]. In several cases, grading weights were assigned to both i-RAT and t-RAT scores, underscoring their importance in summative assessments [37, 38, 58].

Fig. 4.

Summarises key t-RAT practices in pre-clinical education (n = 37 studies)

IV. TBL group activity and peer evaluation phase (Application, Review & Consolidation)

In our analysis, we provide a summary of practices observed in the application phase of TBL subdividing it into ‘application exercise format’, ‘group output format’, ‘peer review’ and insights into whether studies reported any quantitative or qualitative ‘outcomes’ to assess the TBL modules.

Application exercises format

Across the 37 TBL studies analysed, most (33 out of 37) implemented structured application exercises, often grounded in real-world clinical scenarios or case-based MCQs. Advanced formats included sequential clinical scenarios [31], iPad-enhanced simulations [36], and group concept maps [49]. Only a minority (3/37) of studies [14, 44, 46] reported developing application activities using the ‘4S’ principles/framework, viz., significant problem, same problem, specific choice, simultaneous reporting.

Group output format

Over half of the studies (21 of 37) clearly described how team outputs were structured or shared. Common formats included simultaneous reporting via letter cards or whiteboards [14, 33, 51], diagrammatic or visual outputs such as flowcharts or concept maps [31, 35], and verbal or digital group discussions [44, 47]. Where not explicitly described, the group output mechanism may have been integrated with reporting or application scoring.

Peer evaluation format

Approximately one-third of the studies (10/37) reported incorporating peer evaluation, though methods varied considerably. Some used structured rating tools [15, 45], point allocation systems [29], or anonymous narrative feedback [20]. Evaluation frequency ranged from once per course to post-session. Cultural considerations or course logistics were cited by some studies (e.g., 30,56) as reasons for omission.

V. TBL module student outcomes: quantitative results- exam scores & qualitative insights

Most studies (34/37) included an outcome measure, whether qualitative (e.g., student perceptions, engagement) or quantitative (e.g., exam performance, score correlations). Surveys using Likert scales were common, as seen in [14, 31, 58]. A subset also conducted statistical comparisons between TBL and traditional methods [32, 37]. Our analysis also drew on studies reporting measurable academic impacts and learner perceptions associated with TBL modules. We reviewed the study results and synthesised quantitative outcomes, such as improved assessment scores and the predictive value of readiness assessments, along with qualitative feedback on student engagement and satisfaction. We also report key insights into how structured collaborative learning can enhance educational performance.

Quantitative results—exam scores & TBL participation

Numerous studies have demonstrated a positive correlation between TBL participation and improved assessment performance. For example, Bass et al. [14] and Beckler et al. [31] showed a relationship between high post-TBL t-RAT scores and strong summative exam scores, indicating better comprehension and retention. Similarly, Carrasco et al. [18, 52] and Nayak et al. [53] found that i-RAT scores predict final examination performance, particularly among lower-performing students. Studies like Vasan et al. [32] and Langer et al. [47] reported upward trends in course and National Board of Medical Examiners(NBME) scores after switching to a TBL model. Interestingly, Wiley et al. [48] found that students in their TBL intervention scored comparably or better than graduating seniors on an HPV knowledge module, suggesting TBL's potential to close educational gaps. A few studies, such as those by Ismail [17] and Knollmann-Ritschel et al. [49], have shown mixed or modest gains, often depending on how closely the TBL tasks were aligned with assessments.

TBL Participation Qualitative Results (where provided)

Of 37 studies, 19 included qualitative insights on student outcomes related to TBL participation. Studies on student perceptions of TBL were generally favourable, emphasising enhanced engagement, collaboration, and critical thinking. Recurring themes included increased confidence, motivation, and appreciation for peer learning. Burgess et al. [4, 35] found major benefits of social learning, joint enterprise, and mutual engagement. Eguchi et al. [45] and Martínez et al. [36] noted that students found TBL more interactive and retained knowledge better through collaborative problem-solving. However, peer feedback was sometimes met with hesitation. Studies by Burgess et al. [15] and Lerchenfeldt et al. [20] have highlighted students’ difficulty in offering constructive criticism, indicating a need for more training in delivering professional feedback. Qualitative responses also emphasised the importance of facilitator involvement and the structure of application exercises as key factors in perceived success.

Key author conclusion of included studies

In our analysis, we also summarised key conclusions presented by authors of included studies, who broadly stated that TBL is an effective pedagogical strategy for integrating clinical and foundational sciences, improving engagement, and fostering teamwork. Studies such as Deardorff et al. [58] and Nieder et al. [29] emphasised how TBL supports accountability and deeper understanding. Multiple papers, including those by Bass et al. [14] and Megee et al. [46], have stressed the value of co-teaching between clinicians and basic scientists. Kim et al. [38] and Vasan et al. [30, 32] emphasised the importance of challenge and structure in preparatory materials. A few authors, such as Obad [37] and Rotgans et al. [40], have also explored learner differences, noting that high-performing students often respond more positively, while underperformers may require additional support.

Results for quality assessment (MERSQI score & Kirkpatrick levels)

MERSQI results

The MERSQI was applied to 37 research studies published between 2000 and 2024 (November) in peer-reviewed journals. The mean MERSQI score was 12, with total MERSQI scores ranging from 8 to 14 (out of a maximum score of 18). The mean domain scores were highest for the type of data (2.67), data analysis sophistication (1.25), and outcomes (1.16) domains, while the lowest were for sampling institutions (0.51) and the validity of evaluating instruments (0.75) (Supplementary Table 1). The majority (31 out of 37) employed a cross-sectional survey design, with three studies using a repeated cross-sectional design to collect data over 2–4 years. No comparators were reported because we included only studies that used TBL as a pedagogy.

Three studies [30, 39, 58] reported the highest MERSQI scores of 14. The reason why studies reported the highest score varied. Deardorff et al. [58], for example, conducted their study at Wright State University Boonshoft School of Medicine in the USA, comparing graded versus ungraded Group Application exercises within TBL using a non-randomised, two-group design. The study demonstrated rigorous sampling with a single institution contributing a score of 0.5 and a response rate exceeding 75% (score: 1.5). Data were collected using objective measures (score: 3), and the evaluation instruments provided clear validity evidence across internal structure, content, and relationships to other variables (all reported, score: 1 each). Additionally, the data analysis extended beyond descriptive statistics (score: 2) and was deemed appropriate for the design (score: 1). Although the outcome was based on satisfaction, attitudes, perceptions, and general facts (score: 1), the overall robust methodology contributed to the high MERSQI score. Similarly, Roossien et al. [39] from the Amsterdam University Medical Centre in the Netherlands also scored 14. This study employed a single-group pretest–posttest design (score: 1.5) and met sampling criteria with one institution (0.5) and a response rate above 75% (1.5). It utilised objective data (3 points) and provided full validity evidence for the evaluation instruments (scores of 1 each for internal structure, content, and relationship to other variables). The data analysis was sophisticated, extending beyond descriptive analyses (2 points), and was well-suited to the study design (1 point). The outcome measures focused on knowledge and skills (score: 1.5), underscoring the study’s strength in linking TBL processes with measurable learning gains. Detailed MERSQI domain scores for included studies are provided in Annexure (Supplementary Table 4).

Kirkpatrick levels results

Table 4 shows all studies that reported at least one of the four Kirkpatrick levels of learning. Most studies reported Level II Outcomes-Learning (16/37), Level I Reaction (15/37), Level IV Outcomes-Results (11/37), and finally Level III Behaviour (8/37). Summary details of each level, based on selective studies, are described in Table 4 below.

Table 4.

Shows a summary of the four Kirkpatrick Levels of learning across studies(n = 37)

Kirkpatrick Level	Description & Key Outcomes	Example of Selective Studies
Level 1—Reaction	Captures students’ immediate responses and satisfaction with TBL. Typically assessed via surveys or evaluations of perceptions and attitudes	17, 30,31,33 55
Level 2—Learning	Focuses on the acquisition of knowledge, skills, and attitudes. Measured through objective tests (e.g., i-RAT, t-RAT scores), quizzes, and exam performance, reflecting cognitive gains	7,1450; 57; 59
Level 3—Behaviour	Examines changes in learner behaviour and application of acquired skills (e.g., teamwork, communication, and peer feedback). Usually assessed through observations or self-reported behaviour changes	20,38,42,45
Level 4—Results	Measures the broader impact of TBL, such as improvements in exam performance, patient outcomes, or overall institutional effectiveness, often via long-term follow-up or comparative studies	29,34, 37,41,56

Discussion & conclusion

This systematic review identified the key design elements used to adopt TBL in pre-clinical medical education and evaluated its educational outcomes. Our analysis shows that Western institutions, particularly in the USA, are at the forefront of implementing TBL in pre-clinical curricula. Since the effectiveness of TBL stems from its structured format, studies consistently retain its core stages of pre-learning, readiness assurance, and application exercises, highlighting a broad consensus on the model’s fundamental design. Learners consistently value TBL’s student-centred and active-learning approach, while educators appreciate its relative efficiency compared to traditional lecture-based teaching. Nonetheless, apart from two notable Australian examples, most medical schools utilise TBL as an adjunct rather than a primary pedagogical method, complementing, rather than fully replacing established lecture- or problem-based curricula. The gradual pace at which TBL is being adopted as the primary instructional method may reflect resource constraints inherent in transitioning to a fully TBL-based curriculum, or it may indicate a preference for retaining conventional teaching methods to align with country-specific medical accreditation requirements. Our systematic review of 37 pre-clinical TBL studies reveals several consistent patterns and important insights into how TBL is implemented and measured in foundational medical curricula. Although individual implementations differ in scope, ranging from single-session modules on dyslipidaemia [46] to course-wide applications in anatomy [32, 33], a core set of design features recurs across settings. We discuss key study findings across the consolidated TBL stages, starting with TBL preparation (Fig. 2) and provide theoretical context for each TBL phase.

TBL preparation (Pre-learning Phase)

TBL, in essence, builds on the flipped classroom model, where students are provided pre-learning materials to support knowledge acquisition. Whether TBL modules are designed with the outcome in mind and content developed retrospectively (backward design) or sequentially developed with learning aims guiding the next steps (forward design), providing students with preparatory material in line with learning aims and in a manner that facilitates active learning is key. In this review, most programmes release preparatory material 5–7 days in advance and expect 1–2 h of student engagement. Textbook chapters [31, 47] and narrated slide decks [48], were the dominant media, often supplemented by formative quizzes on a learning-management system [44, 55]. Blended approaches, combining readings, short videos and online MCQs, were associated with higher individual readiness scores [45], echoing the findings of Langer et al. [47] that multimodal preparation improves pre-class knowledge acquisition. Conversely, studies that relied solely on passive readings reported lower mean i-RAT scores (< 70%) and greater variance [18, 39], suggesting that interactive pre-learning may be essential for consistent readiness.

TBL assurance (Readiness Phase)

In TBL, the Readiness Assurance Phase is the basic mechanism for encouraging student learning and ensuring exposure to course content. This phase links the individual test (i-RAT) stage, where students self-assess their knowledge levels, to the team test (t-RAT) stage, where students receive peer input to strengthen and/or modify their understanding of course content. The readiness assurance phase both supports team development and provides a feedback-rich learning environment. All included studies implemented the canonical i-RAT → t-RAT sequence, but the formats varied. Item counts ranged from 5 to 30 MCQs; timing spanned 10–30 min. Closed-book, graded i-RATs predominated [38, 47], whereas a minority used ungraded or open-book versions [41]. Regardless of format, team scores (t-RAT) consistently exceeded individual scores by ≥ 20 percentage points, replicating the well-documented ‘safety-net’ effect of team discussion. Yet only Eguchi et al. [45] provided SDs for both i-RAT (70.1 ± 3.0) and t-RAT (96.2 ± 0.4), presenting a pooled estimate of knowledge gain across readiness stages. Several authors [14, 50] have linked higher i-RAT scores to subsequent examination success, supporting longstanding evidence that initial individual accountability predicts downstream achievement.

TBL application (Group Work and Peer Review)

The application phase is one of the TBL foundational pillars that provides the opportunity to repeat the knowledge-acquisition-knowledge-application cycle within each TBL module. The application phase supports critical thinking, team problem-solving and professionalism. These are developed in the context of the application activity and the linked ‘peer evaluation’ stage, which many TBL practitioners consider part of the TBL application stage.

Application exercises

Most studies used clinical case-based MCQs for application activity with simultaneous reporting via whiteboards or electronic polling [31, 47]. Visual artefacts, concept maps or flowcharts, were less common but correlated with higher post-exercise scores [29, 57]. Only a minority of studies (2/37) reported the use of the 4S principles/framework in application design, suggesting that globally most medical schools design their activities flexibly, which may account for the variation in TBL experiences. Moreover, limited reporting of 4S elements, particularly the ‘specific choice’ requirement (i.e. single best answer), has also been noted in health professions education and may restrict opportunities for discussion and critical thinking [59, 60]. Few studies tracked application-phase performance against summative exams, limiting our ability to attribute final knowledge gains to this stage. The review also found that immediate feedback during the application and readiness assurance phases functioned primarily as in-session, formative feedback, arising from peer discussion, answer-reveal mechanisms, and brief facilitator clarification, rather than from delayed written faculty feedback. This feedback-rich environment supports iterative knowledge refinement and aligns with constructivist learning theory, where understanding is actively negotiated through social interaction and timely correction.

Peer evaluation/review

Structured peer evaluation was reported in nine studies. However, the tools used varied: one employed the Fink method [44], another utilised the Michalsen method [29], and one applied a hybrid approach [45]. Additionally, two employed narrative feedback forms [4, 15], and two provided either ad hoc or anonymous rating scales [52, 58]. Two studies in the USA [30, 32] reported using peer review scores internally to identify low-scoring students for proactive staff counselling and support. Where reported, students rated peer feedback as valuable (mean ≥ 3.5 on 5-point scales); yet only three studies [15, 38, 45] demonstrated a quantitative link between peer-evaluation scores and academic outcomes, echoing the broader literature, which suggests that the instructional impact of peer feedback remains under-examined. The underutilisation of peer evaluation, a TBL core activity, may be due to factors, with studies observing time management challenges within compact TBL modules as well as initial student resistance to engage fully with the process [2, 6, 19, 20]. Improved peer evaluation uptake thus provides an opportunity to strengthen TBL delivery across programs.

Cross-study commonalities

Across studies, consolidated barriers to effective TBL implementation emerged at institutional and faculty levels. These barriers include limited faculty development in TBL pedagogy and facilitation, student resistance to collaborative learning expectations, time constraints within existing curricula, and challenges in aligning TBL with conventional pedagogical structures and assessment practices Yet across institutions, TBL adopters maintained the essential sequence of structured pre-learning, timed readiness assurance, application activity design (without the ‘4S’ framework) and reported favourable student perceptions of engagement and teamwork. Educators endorsed TBL for its efficiency relative to faculty-driven lectures, and most programmes incorporated TBL as a supplemental rather than stand-alone pedagogy, reflecting curricular and accreditation constraints. Despite contextual diversity, these shared elements suggest a convergent ‘minimum specification’ for successful TBL in pre-clinical settings. Recommendations for Implementing TBL in Pre-clinical Education.

Orientation and team formation

Evidence supports maintaining TBL group sizes of five to six students, assigned randomly at the semester’s start. Introducing orientation sessions during the first-year induction helps familiarise students with TBL structure, expectations, and collaborative norms, which have been shown to enhance engagement, confidence, and performance over time [20, 29].

Session design and timing

Most studies implemented 2.5- to 3-h TBL sessions with dedicated time for readiness assurance and application activities led by a facilitation team comprising both basic scientists and clinicians. Including structured closure and brief peer-reflection periods may further consolidate learning and encourage accountability [15, 29, 44].

Pre-class preparation

Adopt multimodal pre-learning strategies that combine readings, short videos, and online quizzes, released at least 1 week in advance. Optimal preparation time of two to three hours per TBL enhances readiness, encourages deeper cognitive engagement, and supports equitable team participation [4, 16, 46].

Readiness assurance phase

Standardise individual and team readiness assurance tests (i-RAT/t-RAT) using 10–15 graded, closed-book MCQs with immediate feedback. Such standardisation promotes accountability, reinforces foundational knowledge, and ensures alignment across course modules [14, 18, 42].

Application phase and 4S framework

Design application exercises around the ‘4S’ principle to strengthen conceptual integration. Incorporating visual and technology-enhanced tools (e.g., concept mapping, anatomical imaging, simulation) deepens understanding and connects basic and clinical sciences [15, 37, 49].

Peer evaluation and reflective feedback

Embed structured peer evaluation and facilitator-guided debriefing at the end of each teaching block. Providing time for reflection enhances professional behaviours, communication, and team accountability, contributing to the development of collaborative competencies [20, 29, 44, 45].

Reporting and research transparency

When publishing TBL outcomes, report sample sizes and mean ± SD for i-RAT, t-RAT, application, and summative scores. Consistent reporting enables future meta-analytic synthesis and contributes to the global evidence base for TBL effectiveness [4, 14, 29, 35].

Strengths and limitations

This review offers several notable strengths. Concentrating exclusively on Team-Based Learning in the first two years of medical school highlights a formative phase of medical education that strongly influences students’ cognitive development and professional identity. Its global scope, drawing on 37 studies from high-, middle- and low-income regions across five continents, provides a broadly applicable picture of how TBL is being used and adapted worldwide. Unlike previous syntheses that focus on single endpoints, our analysis spans every stage of the TBL cycle, detailing the workload and format of pre-learning tasks, the design of i-RAT/t-RAT readiness tests, and the nature of application exercises and peer review. These granular insights give curriculum planners practical examples of what works and why.

At the same time, several limitations temper the strength of our conclusions. Outcome measures, reporting formats and assessment timing varied widely between studies; as a result, a formal meta-analysis could not be performed. The scope of TBL interventions also differed markedly, ranging from single-topic sessions to full-course implementations, making direct comparisons difficult. Moreover, most schools deployed TBL as an adjunct to lectures or PBL rather than as the primary teaching modality, so learner outcomes cannot be attributed solely to TBL. Many studies lacked complete reporting of statistics, e.g. means, standard deviations, and sample sizes, particularly for summative exams, and only a few linked performances across the pre-learning, readiness, and application phases to final assessments. Finally, the predominance of positive student-perception reports raises the possibility of publication bias, as neutral or negative experiences may be less likely to appear in the literature. These factors warrant cautious interpretation of our findings, despite our review providing a synthesis of current TBL practice in pre-clinical medical education. Future studies can mitigate these issues by standardising the reporting of TBL design and outcome measures and by conducting multicentre implementation-evaluation studies that allow comparisons across institutional and cultural contexts.

Conclusion

TBL is gaining global traction in pre-clinical medical education, with consistent evidence of high readiness scores, strong student engagement and favourable faculty perceptions. Nevertheless, heterogeneity in design and outcome reporting currently precludes definitive quantitative conclusions about its overall efficacy. Standardised implementation protocols and rigorous, transparent reporting will be critical to unlock the full evidentiary value of future TBL research and guide curriculum transformation. These findings highlight the need for greater standardisation in TBL design and reporting to support evidence-informed curriculum reform in pre-clinical medical education. Future research priorities should include multicentre implementation studies examining TBL scalability and adaptation across diverse institutional contexts, longitudinal evaluations of TBL's impact on clinical competency development, and comparative analyses of resource allocation models that support sustainable TBL delivery.

Authors’ contributions

D.A., A.W., and K.V. conceptualised the study and supervised the research. V.K. conducted the search strategy and extracted data from the included articles. D.A. and V.K. conducted the abstract and full-text review and data extraction for analysis, with D.A. supervising V.K. All conflicts were resolved during the abstract and full-text screening. D.A. analysed and interpreted the final data with A.W. & K.V supervising. D.A. wrote the original and revised drafts, with A.W., K.V., and V.K. providing major contributions to its finalisation. All authors read, commented on, and approved the final manuscript.

Funding

This study did not receive any specific funding. However, one of the co-authors, Dr Vaishnavi Krishnan Namboothiri, was awarded the Future Research Talent-India scholarship by the Australian National University, which supported her visit to ANU as a visiting scholar.

Data availability

All data supporting the findings of this study are included in this published article. No additional datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethical approval was not required, as the study did not involve any human or animal participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.