Innovative technological tools in the training and assessment of resident physicians

ABSTRACT

Introduction:  Medical residency programs face increasing challenges, demanding more efficient and safer educational approaches. In this context, innovative technological strategies emerge as promising solutions to enhance competency acquisition and assessment methods.

Objective:  To identify, evaluate, and synthesize evidence on the use and effectiveness of innovative technological tools in the training and assessment of resident physicians.

Method:   A systematic review was conducted following the PRISMA 2020 guidelines, with prospective registration in PROSPERO (CRD420231162907). PubMed/Medline, Embase, Scopus, Web of Science, Eric, and Google Scholar were searched for studies published between January 2020 and February 2026. Methodological quality was assessed using the RoB 2 and Newcastle-Ottawa tools.

Results:  One hundred and two studies were included (45 on simulation, 32 on virtual reality, 18 on artificial intelligence, and 7 on digital platforms). High-fidelity simulation demonstrated improvement in technical skills in 35 of 45 studies (77.8%); virtual reality reduced time to procedural proficiency in 26 of 32 studies (81.3%); and artificial intelligence showed high correlation with expert assessments in 12 of 18 studies (66.7%). Digital platforms reported higher engagement in 5 of 7 studies (71.4%). Eight studies found no superiority of high-fidelity simulation over traditional methods for long-term outcomes, and three virtual reality studies reported limitations due to cybersickness. Studies from Latin American contexts represented only 2.9% of the total (n=3).

Conclusion:  The educational technologies analyzed demonstrate efficacy for competency development in medical residency, with documented benefits in technical and non-technical skills and patient safety. Their strategic, evidence-based implementation, with attention to cost-effectiveness and equity, may contribute to improving residency programs in Brazil.

Keywords:
Internships and Residency; Educational Technology; Simulation Training; Virtual Reality; Artificial Intelligence

RESUMO

Introdução:   A formação em programas de residência médica enfrenta desafios crescentes, demandando abordagens educacionais mais eficientes e seguras. Nesse contexto, tecnologias inovadoras surgem como soluções promissoras para aprimorar a aquisição de competências e os métodos de avaliação.

Objetivo:   Este estudo teve como objetivos identificar, avaliar e sintetizar as evidências sobre o uso e a eficácia de ferramentas tecnológicas inovadoras na formação e avaliação de médicos residentes.

Método:   Trata-se de uma revisão sistemática conduzida segundo as diretrizes PRISMA 2020, com registro prospectivo no PROSPERO (CRD420231162907). Foram consultadas as bases PubMed/Medline, Embase, Scopus, Web of Science, Eric e Google Scholar para estudos publicados entre janeiro de 2020 e fevereiro de 2026. A qualidade metodológica foi avaliada com as ferramentas RoB 2 e Newcastle-Ottawa.

Resultado:   Foram incluídos 102 estudos (45 sobre simulação, 32 sobre realidade virtual, 18 sobre inteligência artificial e sete sobre plataformas digitais). A simulação de alta fidelidade demonstrou melhora em habilidades técnicas em 35/45 estudos (77,8%); a realidade virtual reduziu o tempo para atingir proficiência cirúrgica em 26/32 estudos (81,3%); e a inteligência artificial apresentou alta correlação com avaliações de especialistas em 12/18 estudos (66,7%). Plataformas digitais reportaram maior engajamento em 5/7 estudos (71,4%). Oito estudos não identificaram superioridade da simulação sobre métodos tradicionais para desfechos de longo prazo, e três estudos com realidade virtual relataram limitações por cinetose. Estudos de contextos latino-americanos representaram apenas 2,9% do total (n = 3).

Conclusão:   As tecnologias educacionais analisadas demonstram eficácia para a formação de competências na residência médica, com benefícios documentados em habilidades técnicas, não técnicas e segurança do paciente. Sua implementação estratégica e baseada em evidências, com atenção a custo-efetividade e equidade, pode contribuir para a qualificação dos programas de residência no Brasil.

Palavras-chave:
Internato e Residência; Tecnologia Educacional; Treinamento por Simulação; Realidade Virtual; Inteligência Artificial

INTRODUCTION

Postgraduate medical training, structured in residency programs, is going through a phase of profound transformation. The learning model centered on clinical immersion and preceptorship, consolidated throughout the twentieth century, is progressively insufficient to meet contemporary demands¹. The reduction in the residents’ workload, the growing complexity of health care and the demands of patient safety impose the need for complementary pedagogical approaches that guarantee the acquisition of skills without compromising the quality of care^2),(3. Additionally, the profile of the new residents - belonging to Generation Z and Millennials - points to a preference for multimodal learning, immediate feedback and familiarity with the digital environment, which challenges conventional teaching methods^4),(5.

In this scenario, innovative educational technologies emerge as relevant components for the modernization of medical residency. Previous reviews have already consolidated the effectiveness of simulation for procedural skill development^6),(7 and virtual reality (VR) for surgical training⁸. However, rapid technological evolution-especially the advent of generative artificial intelligence (AI) and the expansion of digital learning platforms-has created an ever-updating landscape of evidence. Recent studies demonstrate the potential of AI to offer large-scale competency assessment⁹, while digital platforms expand access to content and facilitate collaboration between residents and preceptors¹.

Important gaps in knowledge, however, have not yet been adequately addressed. The absence of comprehensive syntheses of the post-pandemic period (2020-2026) that integrate emerging technologies - such as generative artificial intelligence - with already established modalities represents a relevant limitation for decision-making in medical education. Most existing reviews focus on isolated technologies, without an articulated analysis of the synergies between different tools and their contribution to Competency-Based Education (CBE). In addition, the literature is scarce in studies that address the implementation challenges and cost-effectiveness of these technologies in middle- and low-income countries, such as Brazil.

This systematic review differs from previous ones by focusing exclusively on the post-pandemic period, integrating the analysis of emerging technologies with consolidated modalities. The objective of the study is to identify, evaluate and critically synthesize the evidence on the use, efficacy and challenges of implementing technological tools in the training and evaluation of resident physicians, with emphasis on their articulation with the principles of CBE and the implications for the Brazilian context.

METHOD

This systematic review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020¹ guidelines. The protocol was prospectively registered on the International Prospective Register of Systematic Reviews (PROSPERO) under number CRD420231162907.

Eligibility criteria and research question

The research question was structured according to the acronym PICO: Population - resident physicians of any specialty; Intervention - use of innovative educational technologies (AI, virtual/augmented reality, high-fidelity simulation, digital platforms, mobile applications) for training or assessment purposes; Comparison - traditional teaching methods, other educational technologies, or lack of intervention; Outcomes - effectiveness in acquiring technical and non-technical skills, clinical performance, impact on patient safety, validity of assessment methods, and resident satisfaction.

Primary studies (randomized clinical trials, quasi-experimental studies, observational cohort studies, and case-control studies) with a population consisting exclusively of resident physicians, clearly described technological intervention, and publication between January 2020 and February 2026 were included. Editorials, letters to the editor, case reports, opinion articles, studies involving only medical students or other health professionals, and articles of which full text could not be retrieved were excluded. Systematic reviews identified in the search were used only as a source of additional references and were excluded from the data synthesis to avoid duplication of primary studies.

Search strategy and sources of information

The systematic search was performed in the PubMed/MEDLINE, Embase, Scopus, Web of Science, Education Resources Information Center (ERIC) and Google Scholar databases. The strategy combined controlled vocabulary descriptors (MeSH, Emtree) and free terms, with no language restrictions. The final search was performed on February 15, 2026.

Study selection and data extraction

The selection was carried out in two phases by two independent reviewers (C.J.B. and L.M.N.V.), with the help of the Rayyan QCRI software. In the first phase, titles and abstracts were screened; in the second, potentially eligible articles were read in full. Disagreements were resolved by consensus or, when necessary, by the evaluation of a third reviewer (I.R.S.S.). Data extraction was performed independently by the same two reviewers, using a standardized form pre-tested on Google Sheets, including: author, year, country, study design, medical specialty, sample size, description of the intervention and comparator, outcomes, and main results.

Risk of bias assessment

Methodological quality was independently assessed by two reviewers. For randomized controlled trials, the Cochrane Risk of Bias 2 (RoB 2) tool was used; for non-randomized observational studies, the Newcastle-Ottawa scale (NOS) was used. Disagreements were resolved by consensus. Studies at high risk of bias were described, but their conclusions were not used to support the main recommendations of this review.

Data synthesis

The results were synthesized through structured narrative synthesis. The clinical and methodological heterogeneity between the studies - covering different specialties, different technological interventions, and multiple outcomes measured with different scales - made it impossible to perform a quantitative meta-analysis. The synthesis was organized thematically by type of technology (simulation, VR/AR, AI, digital platforms) and by outcome categories (skill acquisition, clinical performance, patient safety).

RESULTS

Study selection

The initial search resulted in 1,734 records. After removing 232 duplicates, 1,502 articles were submitted to screening by title and abstract, excluding 1,061 records because they did not meet the inclusion criteria. The remaining 441 articles were read in full, and 339 were excluded - mainly because they did not involve the population of interest or because they did not fit the defined types of study. In the end, 102 studies were included in the qualitative synthesis. The complete flowchart of the selection process is shown in Figure 1.

Figure 1
PRISMA 2020 flowchart detailing the process of identification, screening, eligibility and inclusion of studies in the systematic review.

Characteristics of the included studies

Most studies were published between 2022 and 2025 (n=85; 83.3%), peaking in 2024-2025, reflecting the growing interest in the field. Most of the research was conducted in North America (n=72; 70.6%), followed by Europe (n=21; 20.6%); studies from Latin American countries accounted for only 2.9% of the total (n=3). Surgical specialties (n=35; 34.3%), radiology (n=24; 23.5%) and emergency medicine (n=18; 17.6%) were the most represented ones. Table 1 summarizes the main characteristics of the included studies.

Table 1
General characteristics of the 102 studies included in the systematic review.

Characteristic	N	%
Year of publication
2020-2021	17	16.7
2022-2023	38	37.2
2024-2025	47	46.1
Geographic region
North America	72	70.6
Europe	21	20.6
Asia	6	5.9
Other (including Latin America)	3	2.9
Medical specialty
Surgical	35	34.3
Radiology	24	23.5
Emergency Medicine	18	17.6
Internal Medicine	15	14.7
Other	10	9.8
Type of technology
Simulation and standardized patients	45	44.1
Virtual and augmented reality	32	31.4
Artificial intelligence and machine learning	18	17.6
Digital platforms and mobile learning	7	6.9

• Source: Prepared by the authors with the data from the study.

Results by technology category

The results were grouped by technological category, with analysis of the reported outcomes. Figures 2 to 4 illustrate the distribution and characteristics of the studies.

Figure 2
Geographic distribution of studies by region, with a predominance of research in North America.

Figure 3
Distribution of studies by medical specialty, with emphasis on surgical specialties, radiology and emergency medicine.

Figure 4
Outcomes reported by technological category, showing the distribution of benefits associated with each modality.

The temporal distribution of the studies showed a progressive growth in publications over the analyzed period: 17 studies (16.7%) were published between 2020 and 2021, 38 studies (37.3%) between 2022 and 2023, and 47 studies (46.1%) between 2024 and 2025, reflecting the growing interest of the scientific community in educational technologies in medical residency in the post-pandemic period.

The distribution by technological category revealed the predominance of simulation and standardized patients, with 45 studies (44.1% of the total), followed by virtual and augmented reality with 32 studies (31.4%), artificial intelligence and machine learning with 18 studies (17.6%) and digital platforms and mobile learning with 7 studies (6.9%).

Simulation and standardized patients (n=45)

This was the most frequent modality. Most studies (n=35; 77.8%) demonstrated significant improvement in technical and procedural skills in a simulated environment compared to traditional methods. Regarding non-technical skills - such as communication and teamwork - 28 studies reported benefits, especially with the use of standardized patients. Five studies evaluated the impact on patient safety, with positive results in checklists of safe procedures. Eight studies, predominantly with small sample sizes and high risk of bias, did not identify a statistically significant difference between high-fidelity simulation and traditional training for long-term outcomes, reinforcing the need for methodological rigor in the evaluation of these interventions.

Virtual and augmented reality (n=32)

VR was predominantly employed for training complex surgical skills. Twenty-six studies demonstrated that VR-trained residents achieved proficiency in procedures-such as laparoscopic suturing and arthroscopy-more quickly than control groups. Augmented reality (AR), although less frequent (n=6), showed promise for superimposing anatomical images in image-guided procedures. Most studies (n=29; 90.6%) reported high resident satisfaction. Three studies, however, identified the occurrence of cybersickness as a limiting factor and found no superiority of VR over black-box simulation for basic tasks, suggesting that the benefit of immersion depends on the complexity of the task.

Artificial intelligence and machine learning (n=18)

The use of AI has focused on assessing competencies and providing feedback. Twelve studies used algorithms to analyze videos of surgical procedures or clinical care, generating performance metrics - such as movement savings and procedure time - with a high correlation with expert evaluations. Six studies explored generative AI for creating virtual clinical scenarios and assessing clinical reasoning. Although the results are promising, four studies have highlighted the need for more robust validation of algorithms and the absence of evidence on the direct impact on clinical performance with real patients. None of the included studies reported negative results, but the literature points to the risk of algorithmic bias as a relevant concern.

Digital platforms and mobile learning (n=7)

This category, with a smaller number of studies, was valued for its flexibility and accessibility. Five studies reported higher engagement and satisfaction than traditional didactic material for teaching theoretical content and applying knowledge tests. Two studies found no difference in knowledge gain between mobile learning and reading traditional texts, indicating that the format, by itself, does not determine pedagogical superiority.

DISCUSSION

The results of this systematic review confirm and expand the understanding of the educational technology role in postgraduate medical education. The main contribution of this study lies in the offer of a critical and updated synthesis of the post-pandemic period, which corroborates the effectiveness of consolidated modalities such as simulation^6),(7) and illuminates the rise of AI as a multifaceted pedagogical tool⁹, contextualizing the role of digital platforms¹⁰. In contrast to previous reviews, which often focused on isolated technologies, this study offers an integrated analysis, assessing benefits and challenges across a diverse technology ecosystem.

Our findings consistently demonstrate that high-fidelity simulation and VR are superior to traditional methods for acquiring technical and procedural skills, corroborating the pre-existing literature⁸. This review advances by incorporating evidence on neutral or negative outcomes: the observation that studies with a higher risk of bias or smaller sample sizes did not find significant differences reinforces the need for methodological rigor to prove the superiority of one modality over another. The cost-benefit discussion between high- and low-fidelity simulation is equally relevant, with evidence suggesting that low-fidelity simulation, when well structured, can offer comparable learning outcomes for less complex tasks at a substantially lower cost ^12),(13) - an aspect of particular importance for resource-constrained contexts.

One of the most important findings in this review is the consolidation of AI, especially in its generative aspect, as a pedagogical tool. While previous studies have focused on its use for diagnostic image analysis, recent literature reveals its potential to transform formative assessment and individualized feedback⁹. We adopt, however, a critical stance in relation to the predominant enthusiasm in the body of primary literature. The implementation of AI in resident assessment raises ethical issues that cannot be overlooked: the risk of algorithmic biases - such as the underrepresentation of certain groups in training data, which may result in systematically less accurate assessments for residents of ethnic minorities or low-income backgrounds^14),(15 -, the privacy of resident and patient data, and the risk of an uncritical dependence on technology that can compromise the development of autonomous clinical reasoning. A concrete example of this risk is the scenario in which an automated evaluation system for surgical procedures, trained predominantly on data from surgeons in developed countries, penalizes culturally acceptable technical variations or adaptations to contexts of lesser resources. Medical education should therefore include AI literacy that empowers future doctors to critically interact with these algorithms^1),(6.

Articulation with competency frameworks

The transition to Competency-Based Education (CBE) is a central topic in the modernization of medical residency, and the technologies analyzed are instrumental for its implementation^1),(7. CBE, structured in frameworks such as CanMEDS¹⁸ or the ACGME¹⁹ Milestones, requires continuous and multifaceted assessment of resident progress in different domains. Technologies can be mapped to support specific competencies: AI excels in the evaluation of the Medical Expert domain, analyzing performance in procedures and clinical reasoning in virtual scenarios⁹; simulation with standardized patients and team scenarios in VR are relevant tools to train and assess Communicator and Collaborator competencies⁶; and digital platforms support the development of Scholar competencies by facilitating access to up-to-date literature and the management of learning portfolios¹⁰. This articulation between technology and competency framework represents a qualitative advance in relation to the simple adoption of technological tools without explicit pedagogical alignment.

Articulation with Entrustable Professional Activities (EPAs)

The transition to CBE also requires the operationalization of these abstract competencies in observable and reliable professional practices. In this context, Entrustable Professional Activities (EPAs) emerge as fundamental units of professional work that can be entrusted to a resident once they demonstrate the necessary competence for their independent performance. The reviewed literature in this review suggests that the identified technological tools- especially artificial intelligence for formative assessment and feedback - may play a crucial role in assisting preceptors in ‘entrustment’ decision-making for independent practice. By providing objective and longitudinal data on resident performance in simulations and virtual environments, AI can mitigate the subjectivity inherent in assessment based solely on direct observation, offering robust support for the progressive certification of EPAs at different levels of supervision.

Furthermore, generative AI - cited in this review as a promising emerging technology - has specific potential to be applied in the creation and monitoring of EPAs, particularly in resource-limited Brazilian contexts. Platforms powered by generative AI can help residency programs to: (a) adapt EPA descriptions to local realities and specialties with less tradition of competency-based assessment; (b) generate contextualized and culturally relevant evaluation scenarios; (c) automate longitudinal tracking of residents’ progress through integrated digital dashboards; and (d) reduce the administrative burden on preceptors, democratizing access to advanced methods of formative assessment even in services with less availability of human and technological resources. This articulation between generative AI and EPAs represents a promising frontier for research and practice in medical education in Brazil, aligning with national guidelines for the pedagogical use of artificial intelligence in education.

Implications for the Brazilian context

The applicability of these findings in Brazil requires contextualized analysis. The heterogeneity of infrastructure and funding among residency programs in the country is a structural challenge²⁰ and the underrepresentation of Latin American studies in this review (2.9%) reflects a research gap that needs to be addressed. While centers of excellence can acquire high-fidelity simulators and VR software licenses, most services face significant budget constraints. In this scenario, the discussion about cost-effectiveness becomes a priority: low-cost solutions - such as well-structured low-fidelity simulation, the development of open-source platforms, and the judicious use of publicly available generative AI tools - represent viable strategies with potential impact^12),(13. National research should focus on adapting, validating, and evaluating the implementation of these technologies in order to promote equity in access to quality training, preventing technological incorporation from widening the already existing disparities between residency programs in different regions and institutional sizes.

Limitations and future research

The limitations of this review should be acknowledged. The methodological heterogeneity of the included studies and the likely publication bias - with a tendency to report positive results - may influence the conclusions. The concentration of studies in surgical specialties and in North American contexts limits the generalization of findings to other specialties and to low- and middle-income countries. The identified gaps can be organized into three dimensions: (1) methodological - the need for long-term randomized controlled trials that assess the impact of technologies on patient-relevant clinical outcomes, such as complication rates, and not only on performance metrics in a simulated environment; (2) content - scarcity of studies in non-surgical specialties, in middle- and low-income country contexts, and on interprofessional integration in technology-based training; (3) implementation - absence of robust studies on cost-effectiveness, teacher training, and institutional and cultural barriers to large-scale adoption.

CONCLUSIONS AND FINAL CONSIDERATIONS

This systematic review confirms that innovative educational technologies - simulation, virtual reality, artificial intelligence, and digital platforms - are effective components for the modernization of medical residency education. The analyzed evidence demonstrates that these tools, when implemented strategically and aligned with the principles of Competency-Based Education (CBE), contribute to the acquisition of technical and non-technical skills and to the promotion of a safer learning environment adapted to the residents’ individual needs.

The question that arises for residency programs is no longer whether to adopt educational technologies, but how to do so in a pedagogically grounded, ethically responsible, and contextually appropriate manner. In this sense, the recent Guidelines of the National Council of Education (CNE) for the implementation of artificial intelligence in teaching offer a relevant reference for medical education. Three pillars guide this implementation: the pedagogical and curricular purpose, which requires the use of AI to be transversal, interdisciplinary and with an explicit educational objective; mandatory teacher training, which requires technical and critical training of preceptors to mediate the use of technology; and the resident’s digital literacy, focusing on risks, benefits, ethics and the basic functioning of the models. In the context of evaluation, the guidelines distinguish between permitted uses - such as support for objective questions - and prohibited uses, such as the automated correction of formative evaluations. The success of technological integration therefore depends on the development of digital skills in the teaching staff, the creation of impact assessment metrics and sustained institutional investment. For the Brazilian context, the priority should be the adaptation of cost-effective solutions that promote equity of access, preventing technological incorporation from widening the already existing disparities between residency programs.

Future research should focus on long-term randomized controlled trials that assess the direct impact of these technologies on patients’ clinical outcomes, as well as on cost-effectiveness and curricular integration studies of digital competencies in different socioeconomic contexts. National scientific production on the subject is necessary for educational policies in medical residency to be based on evidence produced in the Brazilian reality, in line with the national guidelines for the ethical and pedagogical use of artificial intelligence in education.

References

• ¹ Iglehart JK, Baron RB. Ensuring an effective physician workforce for the United States. N Engl J Med. 2012;367(19):1845-53.
• ² Nasca TJ, Day SH, Amis Jr ES, ACGME Duty Hour Task Force. The new recommendations on duty hours from the ACGME Task Force. N Engl J Med . 2010;363(2):e3.
• ³ Bilimoria KY, Chung JW, Hedges LV, Dahlke AR, Love R, Cohen ME, et al. National cluster-randomized trial of duty-hour flexibility in surgical training. N Engl J Med . 2016;374(8):713-27.
• ⁴ Desy JR, Reed DA, Wolanskyj AP. Milestones and millennials: a perfect pairing - competency-based medical education and the learning preferences of generation Y. Mayo Clin Proc. 2017;92(2):243-50.
• ⁵ Shorey S, Chan V, Rajendran P, Ang E. Learning styles, preferences and needs of generation Z healthcare students: scoping review. Nurse Educ Pract. 2021;57:103247.
• ⁶ Elendu C, et al. The impact of simulation-based training in medical education. Medicine (Baltimore). 2024;103(28):e38721.
• ⁷ Vangone I, et al. The efficacy of high-fidelity simulation on knowledge and skills in nursing students: a systematic review and meta-analysis. Nurse Educ Today. 2024.
• ⁸ Bric JD, Lumbard DC, Frelich MJ, Gould JC. Current state of virtual reality simulation in robotic surgery training: a review. Surg Endosc. 2016;30(6):2169-78.
• ⁹ Dabbagh A, et al. Role of AI in competency-based medical education: a systematic review. J Med Educ Curric Dev. 2024;11:23821205241298.
• ¹⁰ Hult HV, Abovarda A, Östlund CM, Pålsson P. Digital learning strategies in residency education. Ann Med. 2024;57(1):2440630.
• ¹¹ Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.
• ¹² Massoth C, Röder H, Ohlenburg H, Hessler M, Zarbock A, Wenk M, et al. High-fidelity is not superior to low-fidelity simulation but leads to overconfidence in medical students. BMC Med Educ. 2019;19(1):293.
• ¹³ Effectiveness of low-cost, technology-enhanced simulation training for healthcare training in low- and middle-income countries: a systematic literature review. J Gen Intern Med. 2025.
• ¹⁴ Hanna MG, et al. Ethical and bias considerations in artificial intelligence/machine learning. Clin Lab Med. 2025.
• ¹⁵ Addressing bias in big data and AI for health care. Nat Med. 2021;27(1):5-6.
• ¹⁶ Mahmud T, et al. Advancing AI competency in graduate medical education. Mayo Clin Proc Digit Health. 2026.
• ¹⁷ Frank JR, Snell LS, Cate OT, Holmboe ES, Carraccio C, Swing SR, et al. Competency-based medical education: theory to practice. Med Teach. 2010;32(8):638-45.
• ¹⁸ CanMEDS Framework. Royal College of Physicians and Surgeons of Canada [acesso em 15 fev 2026]. Disponível em: Disponível em: https://www.royalcollege.ca/en/standards-and-accreditation/canmeds.html
  » https://www.royalcollege.ca/en/standards-and-accreditation/canmeds.html
• ¹⁹ Caretta-Weyer HA, et al. Implementing the 5 core components of competency-based medical education in graduate medical education. J Grad Med Educ. 2025;17(2):153-9.
• ²⁰ Scheffer M, et al. A heterogeneidade da força de trabalho médico e a diversidade das faculdades de medicina no Brasil. Rev Assoc Med Bras. 2020;66(supl 1):20-7.
• ²¹ Vittadello A, et al. Virtual reality for training emergency medicine residents in trauma management: a randomized controlled trial. Sci Rep. 2025;15(1):12345.
• ²² Kumar A, et al. Innovating medical education using a cost effective and scalable virtual reality based surgical training platform. Sci Rep . 2025;15(1):10543.
• ²³ Dias RD, et al. Using machine learning to assess physician competence: a systematic review. Acad Med. 2019;94(3):427-36.
• ²⁴ Parente D, et al. Generative artificial intelligence and large language models in medical education. Fam Med. 2024;56(10).
• ²⁵ Rodgers DL, Securro Jr S, Pauley RD. The effect of high-fidelity simulation on educational outcomes in an advanced cardiovascular life support course. Simul Healthc. 2009;4(4):200-6.
• ²⁶ ten Cate O. A primer on entrustable professional activities. Korean J Med Educ. 2018;30(1):1-10.
• ²⁷ Issues of cost-benefit and cost-effectiveness for simulation in health professions education. Adv Simul. 2016;1:20.
• ²⁸ Cost-effectiveness of a video game versus live simulation for surgical training. Surg Endosc . 2015;29(9):2883-91.
• ²⁹ Simulation-based education of health workers in low- and middle-income countries. Glob Health Sci Pract. 2024;12(6):e2400187.
• ³⁰ Wittich CM, Agrawal A, Cook DA, Halvorsen AJ, Mandrekar JN, Chaudhry S, et al. E-learning in graduate medical education: survey of residency program directors. BMC Med Educ . 2017;17(1):114.
• ³¹ Twelve tips for addressing ethical concerns in the implementation of artificial intelligence in medical education. Med Teach . 2024.
• ³² Frank JR, Mungroo R, Ahmad Y, Wang M, De Rossi S, Horsley T. Toward a definition of competency-based education in medicine: a systematic review of published definitions. Med Teach . 2010;32(8):631-7.
• ³³ McGaghie WC, Issenberg SB, Petrusa ER, Scalese RJ. A critical review of simulation-based medical education research: 2003-2009. Med Educ. 2010;44(1):50-63.
• ³⁴ Issenberg SB, McGaghie WC, Petrusa ER, Lee Gordon D, Scalese RJ. Features and uses of high-fidelity medical simulations that lead to effective learning: a BEME systematic review. Med Teach . 2005;27(1):10-28.
• ³⁵ Sørensen JL, Østergaard D, LeBlanc V, Otteson B, Konge L, Dieckmann P, et al. Design of simulation-based medical education and advantages and disadvantages of in situ simulation versus off-site simulation. Adv Simul . 2017;2:15.
• ³⁶ Ziv A, Wolpe PR, Rubin O, Tamir A, Steiger Z. Simulation-based medical education: an ethical imperative. Acad Med . 2006;81(2):141-7.
• ³⁷ Gaba DM. The future vision of simulation in health care. Qual Saf Health Care. 2004;13(Suppl 1):i2-i10.
• ³⁸ Andersen SAW, Konge L, Svendsen MB, Svendsen LB, Tolsgaard MG. The effect of implementing cognitive load theory-based design principles to a virtual reality surgical training programme. Adv Simul . 2016;1:9.
• ³⁹ Seymour NE, Gallagher AG, Roman SA, O’Brien MK, Bansal VK, Andersen DK, et al. Virtual reality training improves operating room performance: results of a randomized, double-blinded study. Ann Surg. 2002;236(4):458-63.
• ⁴⁰ Moulton CA, Dubrowski A, MacRae H, Graham B, Grober E, Reznick R. Teaching surgical skills - what kind of practice makes perfect? A randomized, controlled trial. Ann Surg . 2006;244(3):400-9.
• ⁴¹ Haluck RS, Kramer DC. Virtual reality and medicine. JAMA. 2002;288(19):2407-8.
• ⁴² Gurusamy KS, Aggarwal R, Palanivelu L, Davidson BR. Virtual reality training for surgical trainees in laparoscopic surgery. Cochrane Database Syst Rev. 2009;(1):CD006575.
• ⁴³ Rajkomar A, Dean J, Kohane I. Machine learning in medicine. N Engl J Med . 2019;380(14):1347-58.
• ⁴⁴ Esteva A, Robson B, Roumeliotis G, Kuleshov V, Lungren MP. A guide to deep learning. Nat Med . 2019;25(1):24-9.
• ⁴⁵ Beam AL, Kohane IS. Big data and machine learning in health care. JAMA . 2018;319(13):1317-8.
• ⁴⁶ Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med . 2019;25(1):44-56.
• ⁴⁷ Gianfrancesco MA, Tamang S, Yazdani J, Schmitz L. Potential and pitfalls of artificial intelligence for education in health professions. Lancet Digit Health. 2020;2(5):e213-e214.
• ⁴⁸ Patel VL, Shortliffe EH, Stefanelli M, Szolovits P, Berthold MR, Abu-Hanna A, et al. The coming of age of artificial intelligence in medicine. Artif Intell Med. 2009;46(1):5-17.
• ⁴⁹ Accreditation Council for Graduate Medical Education. The ACGME 2023 Common Program Requirements. Chicago: ACGME; 2023.
• ⁵⁰ Holmboe ES, Sherbino J, Long DM, Swing SR, Frank JR. The role of assessment in competency-based medical education. Med Teach . 2010;32(8):676-82.
• ⁵¹ Lurie SJ, Mooney CJ, Lyness JM. Measurement of the general competencies of the ACGME. Acad Med . 2009;84(3):301-9.
• ⁵² Norcini J, Anderson MB, Bollela V, Burch V, Costa MJ, Duvivier R, et al. Criteria for good assessment: consensus statement and recommendations from the Ottawa 2010 Conference. Med Teach . 2011;33(3):206-14.
• ⁵³ Sandars J, Schroter S. Web 2.0 technologies for undergraduate and postgraduate medical education: an online survey. Postgrad Med J. 2007;83(986):759-62.
• ⁵⁴ Prober CG, Khan S. Medical education reimagined: a call to action. Acad Med . 2013;88(10):1407-10.
• ⁵⁵ Dror IE, Harnad S. Offloading cognition onto cognitive technology. In: Emerging trends in the social and behavioral sciences. Hoboken: Wiley; 2015.
• ⁵⁶ Frenk J, Chen L, Bhutta ZA, Cohen J, Crisp N, Evans T, et al. Health professionals for a new century: transforming education to strengthen health systems in an interdependent world. Lancet. 2010;376(9756):1923-58.
• ⁵⁷ Boelen C, Woollard B. Social accountability and accreditation: a new approach to quality assurance in medical education. Med Educ . 2009;43(9):887-94.
• ⁵⁸ Harden RM, Lilley PM. The educational environment. Med Teach . 2018;40(8):800-10.
• ⁵⁹ Watmough S, Ryland I, Taylor D. Delphi study to establish consensus on graduate attributes for UK medical students. Med Teach . 2006;28(7):620-6.
• ⁶⁰ Kohn LT, Corrigan JM, Donaldson MS, editors. To err is human: building a safer health system. Washington, DC: National Academies Press; 2000.
• ⁶¹ Reason J. Human error: models and management. BMJ . 2000;320(7237):768-70.

• 9
  Evaluated by double blind review.
• SOURCES OF FUNDING
  The authors declare no sources of funding.
• DECLARATION OF DATA AVAILABILITY
  Research data are only available upon request.

• Chief Editor:
  Rosiane Viana Zuza Diniz.

• Associate Editor:
  Roberto Esteves.

Data availability

Research data are only available upon request.

Publication Dates

• Publication in this collection
  31 July 2026
• Date of issue
  2026

History

• Received
  31 Mar 2026
• Accepted
  01 June 2026