Abstract
Background
Skills in electrocardiogram (ECG) interpretation are critical in clinical decision making but remain poor among Internal Medicine residents. There remains no standardized method for teaching ECG interpretation.
Methods
This study aimed to leverage the benefits of microlearning by assessing an asynchronous microlearning ECG curriculum for Internal Medicine residents at a single academic residency program between September 2023 and June 2024. The ECG curriculum was comprised of 30 mandatory weekly ECGs which included a clinical vignette, a 12-lead ECG, and a single multiple-choice or free response question posted on Microsoft Teams. Residents were required to submit interpretations within 1 week, after which the answer, a short description of ECG findings, and clinical pearls were posted. Data collection included pre-post surveys with a 1- to 5-Likert scale measuring self-reported confidence and a test containing 15 unique ECGs.
Results
Of 151 total categorical and preliminary Internal Medicine residents, 38 (25.2%) completed pre-post surveys and were analyzed as matched pairs. Residents demonstrated increases in Likert ratings of confidence (mean post-pre score difference 0.44 of 5, P = <.01), overall performance (mean post-pre raw score difference 0.97 of 15, P = <.01), and performance in identifying normal variants (mean post-pre score subset difference 14.21%, P = .017). Residents had no significant change in performance in identifying tachyarrhythmias (mean post-pre score subset difference 5.9%, P = .24), ST-segment changes (0.88%, P = 0.89), or conduction disease (−0.65%, P = .92).
Conclusions
Overall, this curriculum was effective in improving resident confidence in and knowledge of ECG interpretation, with a particular strength in identifying normal/normal variants. Next steps include targeting specific ECG pathologies in which residents show particular knowledge gaps and creation of a 3-year longitudinal curriculum to reinforce knowledge over the course of a 3-year residency.
Background
The American College of Physicians recommends that Internal Medicine (IM) Residency programs provide adequate training in electrocardiogram (ECG) interpretation, but developing curricula to encourage trainee growth in ECG interpretation remains challenging. 1 Proper ECG interpretation is critical in dictating care and identifying medical emergencies. 2 Lee et al 3 found that 49% of patients with missed acute myocardial infarctions who were discharged home from the Emergency Department could have been appropriately diagnosed with more proficient ECG reading skills. Reliance upon computer interpretations for clinical decision making is detrimental to patient care, as trainees tend to agree with the computer interpretation, leading to greater inaccuracy, especially when the computer interpretation is incorrect. 4
Professional guidelines report a minimum list of abnormal findings for practicing internists to recognize, with the belief that internists with “adequate” ECG interpretation skills should have no difficulty diagnosing common abnormal findings.1,5 Despite marked improvement with clinical experience, ECG interpretation skills among IM residents remain poor, with a lack of confidence and proficiency in ECG interpretation across graduates of IM programs nationally.2,6 Several residency programs have developed various ECG interpretation curricula ranging in structure from monthly didactics 7 to a dedicated “boot camp,” 2 with varying results.7–9 However, there remains no standardized method for teaching ECG interpretation to residents. 2
Microlearning has emerged as an educational strategy characterized by accessible, multimodal, asynchronous, and “bite-sized” delivery of teaching points. 8 Microlearning has been positively received among trainees in graduate medical education, with benefits in improving learner interest and morale. 9 Our study describes a novel microlearning spaced asynchronous curriculum intended to improve confidence and competency of IM residents in ECG interpretation at a single academic IM residency program.
Methods
Study design
At a single academic IM residency program during the 2023 to 2024 academic year, we designed and implemented a mandatory microlearning spaced asynchronous ECG curriculum for all preliminary and categorical residents (N = 151). The curriculum was comprised of 30 weekly ECGs administered over a 30-week period. Each weekly ECG included a clinical vignette, a single standard 12-lead ECG, and a single multiple-choice or free response question asking the resident to identify the correct ECG finding within the clinical context.
All residents were included in the curriculum, irrespective of their completion of the pre- and postcurricular surveys. Although this precluded the use of a separate control group, we employed matched pre- and postintervention assessments among residents who completed both surveys, allowing each participant to serve as their own internal control. Weekly ECGs and questions were posted on the official program Microsoft Teams group at the start of the week. Discussion throughout the week and follow-up questions were encouraged. Residents were required to submit interpretations by the end of the week, after which the answer, a short description of the ECG findings, and pertinent clinical pearls were posted. 75% completion was required for residents to be considered compliant. Since the weekly ECG microlearning assignments were not anonymous, we tracked only participation to maintain a nonpunitive, supportive learning environment and to encourage engagement. To assess the program's effectiveness, we used de-identified pre- and post-intervention ECG survey assessments. A pre- and post-curriculum survey was administered to assess knowledge and comfort level. These assessments each contained 15 ECGs, which combined with weekly ECGs, totaled a sum of 60 ECGs for the entire curriculum. All survey ECG questions included multiple-choice options with a single correct answer.
All ECGs were from adult patients. Weekly ECGs and pre-post curriculum survey ECGs were drawn from Wave Maven, a free online self-assessment tool containing hundreds of ECGs obtained at Beth Israel Deaconess Medical Center. 10 Maven ECGs are interpreted by cardiologists and include a clinical vignette and detailed explanation. ECGs selected from Wave Maven are assigned a difficulty of 1 to 5, 1 being the simplest and 5 being the most difficult. An even number of simple and complex ECGs were incorporated into the curriculum to ensure that the average difficulty of our pre- and post-curriculum survey were similar (Table 1). Each ECG included in weekly ECGs as well as pre- and post-curriculum surveys either contained normal findings, a normal variant, or only one primary abnormality. Residents received credit for weekly ECGs as long as they provided an answer. In the pre- and post-curriculum surveys credit was given if they correctly identified normal findings, the specific normal variant, or the primary abnormality.
ECG pre- and post-quizzes: ECG categories and number of/average difficulty of ECGs included.
Inclusion/exclusion criteria
All 151 IM and Preliminary residents (Postgraduate Year [PGY] 1-3) in the residency program were included. Participation in the educational sessions with weekly ECG microlearning assignments was mandatory as part of the IM residency curriculum, but survey completion was voluntary and had no bearing on resident evaluations. No exclusion criteria were applied.
We used the SQUIRE reporting guideline to draft this manuscript, and the SQUIRE reporting checklist when editing, included in Supplement A. 11
Data collection
All residents received a pre- and post-curriculum survey which included a 1- to 5-Likert scale question measuring self-reported ECG-interpretation confidence and a test containing 15 unique ECGs. ECGs were drawn from Wave Maven with an average difficulty of 3 and had 1 associated multiple-choice or short-answer question. Clinical training data including PGY year and number of prior cardiology rotations, defined as the cardiology consult service or cardiac intensive care unit, were also collected. The number of prior cardiology rotations was represented as categories of 0, 1, 2, 3, or greater than 4 total experiences.
Statistical analysis
A nonparametric approach was utilized that does not assume normality in a matched analysis comparing pre-post surveys, with residents serving as their own control. Two-tailed resident's t-tests were utilized to compare pre-post curriculum general performance and performance in identifying specific pathologies organized into 4 knowledge subsets within the test: tachyarrhythmias, ST-segment changes, conduction disease, and normal variants and other miscellaneous pathologies. Analysis of variance (ANOVA) was performed to stratify test performance by PGY year, Likert scale rating of confidence, and number of prior cardiology rotations. This study has received an exemption from the IRB as it does not meet the definition of human subject research (IRB Request number: HSRD24-0126-LIJ).
Results
Of 151 total PGY-1, PGY-2, and PGY-3 categorical and PGY-1 preliminary year residents, 117 (77.4%) residents completed the presurvey and 57 (37.7%) completed the postsurvey. 38 completed both pre-post surveys (response rate of 25.2%) and were analyzed as matched pairs. Data is summarized in Table 2. From pre- to post-curriculum surveys, residents demonstrated a significant increase in Likert ratings of confidence (mean post-pre score difference 0.44 of 5, P = <.01) and overall test performance (mean post-pre raw score difference 0.97 of 15, P = <.01, d = 0.913).
ECG general and subdomain test performance, Likert scale ratings, and test performance stratified by secondary measures.
Within knowledge subsets, residents also demonstrated a significant increase in test performance in identifying normal variants and miscellaneous pathologies (mean post-pre score subset difference 14.21%, P = .017, d = 0.820). Residents, however, did not demonstrate a significant change in test performance in identifying tachyarrhythmias (mean post-pre score subset difference 5.9%, P = .24), ST-segment changes (mean post-pre score subset difference 0.88%, P = .89), or conduction disease (mean post-pre score subset difference −0.65%, P = .92). Stratifying by PGY year demonstrated no difference in pre- (F = 1.00, P = .40) or post-test (F = 0.50, P = .69) performance. When stratified by prior number of cardiology rotations and Likert scale ratings of confidence, there similarly was also no difference in pre- and post-test performance.
Discussion
This novel microlearning spaced asynchronous ECG curriculum was designed to improve residents’ confidence and competence in ECG interpretation through longitudinal exposure using a publicly available ECG resource database. A microlearning curricular format encompasses several aspects including flexibility in learning, an asynchronous structure, use of “bite-sized” content, multimodal forms of learning, and peer-to-peer collaboration, elements which are recommended when constructing an effective ECG curriculum.8,9,12 All of these elements were included in our curriculum.
ECGs were distributed weekly for asynchronous review, allowing residents to engage individually or in groups. This flexible format enabled those on inpatient rotations to integrate ECG interpretation with patient care and fostered deeper discussions among residents, students, and faculty. Collaborative interpretation was encouraged, with answers and key clinical insights provided at the end of each week.
Our curricular structure was similar to that of Saba et al, 9 who utilized a resident-led email platform where neurology residents shared clinical pearls among their peers, promoting a culture of “mutual microlearning.” Peer-to-peer exchange of clinical pearls was frequently how residents learned most during their clinical experiences, and the “mutual microlearning” structure was well received, with benefits seen in increasing resident morale, collaboration, and interest in clinical learning. 9 This ECG curriculum was structured with a goal of fostering a culture of teaching and learning to make rounds a more academic experience.
The EDUCATion Curriculum Assessment for Teaching Electrocardiography (EDUCATE) trial, a large international randomized controlled study, assessed the effectiveness of various web-based educational strategies for improving ECG interpretation among medical professionals. The study demonstrated significant improvements across disciplines using question-based, lecture-based, or hybrid formats compared to no intervention, underscoring the potential of asynchronous learning to enhance diagnostic skills. 13 These findings align with our own study approach, which explores the integration of an online asynchronous microlearning-based curriculum to optimize educational outcomes for residents. It also informs that there may be added benefits to additional augmentation of our curriculum with more robust lecture-based teaching either synchronously or asynchronously.
Within our study, postcurricular data showed that the curriculum significantly improved resident knowledge and confidence in ECG interpretation, with both statistically and educationally meaningful gains. On average, a test score improvement of 0.97 was achieved, which translated to correctly interpreting one additional ECG per 15 ECGs. The greatest gains were in recognizing normal variants and miscellaneous findings, while no significant improvement was seen in identifying tachyarrhythmias, ST-segment changes, or conduction disease. A ceiling effect where high baseline proficiency limits measurable improvement is unlikely to explain the observed differences in ECG category performance, as pre- and post-test scores remained below 80% across all categories, indicating meaningful opportunity for learning. Variability in academic discussions across inpatient teams may have contributed to inconsistent improvement in recognizing specific pathologies.
It is notable that pre- and post-test performance was not different when stratified by PGY year, number of prior cardiology rotations, or self-rated confidence in ECG interpretation. Our findings suggest that residents generally have difficulty estimating their ability to interpret ECGs. Additionally, our findings support the idea that prior cardiology experience did not predispose residents to greater baseline or postcurricular ECG interpretation proficiency. Klein et al 14 reported greater improvement in proficiency in only PGY-1 residents with no prior cardiology rotations. 14 In our study, the majority of residents already had prior cardiology rotations. We found that 8 of 38 (21.1%) residents during the pretest reported no prior cardiology rotations, but by the post-test only 1 of 38 (2.6%) residents reported no prior cardiology rotations. This may be the reason why we did not see a similar effect. Of note, Klein et al 14 found no relationship between level of interest in cardiology and postcurricular improvement in ECG interpretation proficiency. There is a key distinction between prior experience and interest in cardiology. We did not stratify test performance by interest in cardiology, which is worth further investigation.
Our study has multiple strengths including that it was conducted prospectively at a large academic medical center, allowing for real-time assessment of an educational intervention within a diverse internal medicine residency program. We utilized a low resource microlearning approach, fostering ongoing engagement through brief, accessible weekly ECG assignments. By leveraging the free, publicly available Harvard Wave-Maven ECG database, our curriculum can be easily replicated by other institutions without financial or logistical barriers improving generalizability.
Limitations
This study was limited by multiple factors including a limited sample size due to a low response rate of 25.2% and by the single center design. The limited number of participants who completed both the pre- and postintervention assessments represents a potential source of selection bias as those completing both surveys may be inherently different than those who did not. Additionally, a repeat post-test at a sufficient length of time from the end of our curriculum to assess long-term retention of knowledge was not conducted. Future avenues of study include measuring long-term data and other metrics such as resident engagement.
Next steps
The curriculum warrants adjustment to target specific ECG pathologies and measure improvement in these knowledge subsets. One consideration is to build on the curriculum by adding 2 more years’ worth of questions and explanations, thereby creating a 3-year longitudinal curriculum that can be recycled every 3 years going forward. This will allow residents to see upwards of 180 ECGs, reinforcing their knowledge and building on their ECG interpretation skills over the course of their training without viewing the same ECG twice. Another alternative strategy would be to include more didactic engagement and to have more frequent survey assessment tools on individual high yield ECGs topics, although this would require more institutional resources and faculty time.
Conclusion
Assessment of our microlearning asynchronous ECG curriculum at an academic internal medicine residency program demonstrated promising results suggesting improved resident knowledge and confidence in ECG interpretation. We hope that our curriculum will encourage residents to independently interpret their patients’ ECGs and promote academic discussion and teaching within their clinical teams. Additionally, our curriculum can potentially serve as a model for other institutions looking to build an ECG curriculum for their residents.
Supplemental Material
sj-docx-1-mde-10.1177_23821205251358038 - Supplemental material for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs
Supplemental material, sj-docx-1-mde-10.1177_23821205251358038 for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs by Matthew Capustin, Spencer Weintraub, Ji-Cheng Hsieh, Jack Alboucai, Farzana Antara and Karen Friedman in Journal of Medical Education and Curricular Development
Supplemental Material
sj-docx-2-mde-10.1177_23821205251358038 - Supplemental material for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs
Supplemental material, sj-docx-2-mde-10.1177_23821205251358038 for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs by Matthew Capustin, Spencer Weintraub, Ji-Cheng Hsieh, Jack Alboucai, Farzana Antara and Karen Friedman in Journal of Medical Education and Curricular Development
Supplemental Material
sj-docx-3-mde-10.1177_23821205251358038 - Supplemental material for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs
Supplemental material, sj-docx-3-mde-10.1177_23821205251358038 for A Novel Microlearning Asynchronous Curriculum to Improve Internal Medicine Trainees’ Confidence and Skill in Interpreting ECGs by Matthew Capustin, Spencer Weintraub, Ji-Cheng Hsieh, Jack Alboucai, Farzana Antara and Karen Friedman in Journal of Medical Education and Curricular Development
Footnotes
Acknowledgments
The authors would like to thank their primary faculty advisor, Dr Karen Friedman, who provided significant guidance in our experimental design, data collection, and writing of the manuscript.
Ethical Considerations
This study was reviewed by the local Institutional Review Board (IRB) and determined to be exempt from full review as an educational quality improvement initiative. Participation in the assessments was integrated into the residency training program, and all data were de-identified prior to analysis. No financial incentives or penalties were associated with participation.
Author Contributions
MC, SW assisted with curriculum development, curriculum implementation, survey tool development/design and manuscript writing/editing. JH and JA assisted with curriculum implementation and manuscript writing/editing. FA assisted with data analysis and manuscript writing. KF assisted with manuscript writing/editing and educational curriculum development.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Consent
Verbal consent was sufficient, as all surveys were anonymous, and residents were given the opportunity to opt out of performing pre- and post-curriculum surveys. They were instructed that this would have no bearing on their evaluations.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
