Abstract
Introduction:
Despite the critical role of electrocardiograms (ECGs) in patient care, evident gaps exist in ECG interpretation competency among healthcare professionals across various medical disciplines and training levels. Currently, no practical, evidence-based, and easily accessible ECG learning solution is available for healthcare professionals. The aim of this study was to assess the effectiveness of web-based, learner-directed interventions in improving ECG interpretation skills in a diverse group of healthcare professionals.
Methods:
In an international, prospective, randomized controlled trial, 1206 healthcare professionals from various disciplines and training levels were enrolled. They underwent a pre-intervention test featuring 30 12-lead ECGs with common urgent and non-urgent findings. Participants were randomly assigned to four groups: (i) practice ECG interpretation question bank (question bank), (ii) lecture-based learning resource (lectures), (iii) hybrid question- and lecture-based learning resource (hybrid), or (iv) no ECG learning resources (control). After four months, a post-intervention test was administered. The primary outcome was the overall change in ECG interpretation performance, with secondary outcomes including changes in interpretation time, self-reported confidence, and accuracy for specific ECG findings. Both unadjusted and adjusted scores were used for performance assessment.
Results:
Among 1206 participants, 863 (72%) completed the trial. Following the intervention, the question bank, lectures, and hybrid intervention groups each exhibited significant improvements, with average unadjusted score increases of 11.4% (95% CI, 9.1 to 13.7; P<0.01), 9.8% (95% CI, 7.8 to 11.9; P<0.01), and 11.0% (95% CI, 9.2 to 12.9; P<0.01), respectively. In contrast, the control group demonstrated a non-significant improvement of 0.8% (95% CI, −1.2 to 2.8; P=0.54). While no differences were observed among intervention groups, all outperformed the control group significantly (P<0.01). Intervention groups also excelled in adjusted scores, confidence, and proficiency for specific ECG findings.
Conclusion:
Web-based, self-directed interventions markedly enhanced ECG interpretation skills across a diverse range of healthcare professionals, providing an accessible and evidence-based solution.
Keywords: ECG competency, ECG education, healthcare professionals, medical education, online learning
INTRODUCTION
Twelve-lead electrocardiogram (ECG) interpretation is an essential competency and an integral part of medical education across various healthcare disciplines (1–3). However, proficiency is not uniformly developed among health care professionals, which have been attributed to numerous barriers and difficulties in medical training and clinical practice, including time constraints, skill development challenges, and educational obstacles (4–5).
Over the last five decades, numerous reports have identified significant gaps in ECG interpretation skills (6–14). Yet, previous studies have been limited by their lack of participant diversity, narrow topic range, and small sample size, which have constrained a thorough understanding of the issue. Recently, our team conducted a large-scale study (6) with a diverse group of healthcare professionals, demonstrating widespread and pervasive deficits in ECG interpretation skills. These findings provide yet another compelling reason to address pressing need for effective and widely accessible educational interventions for healthcare professionals.
Ideally, ECG interpretation training approaches should not only be efficient, adaptable, and enduring, but also capable of advancing learners from a novice to an expert level of interpretation. As we delve deeper into the digital era, which will continue to disrupt traditional educational paradigms, a progressive strategy for ensuring ECG interpretation skill acquisition should encompass online learning resources (15). Although a broad spectrum of learning methods already exists, ranging from conventional textbook studies to face-to-face teacher-student interaction, there is a noticeable gap in online resources that provide practical, evidence-based, and easily accessible solutions (16–19). Moreover, the effectiveness of emerging technology-based learning approaches remains yet to be conclusively determined (20–22).
To address these challenges and uncertainties, we conducted the EDUcation Curriculum Assessment for Teaching Electrocardiography (EDUCATE) Trial (23). The EDUCATE Trial evaluated the efficacy of various web-based, self-directed learning interventions in enhancing ECG interpretation proficiency. This report presents the outcomes of three different educational interventions, each designed to improve ECG interpretation skills.
METHODS
Overview
The international, prospective, randomized EDUCATE Trial was designed to evaluate the efficacy of web-based, self-directed learning approaches in enhancing ECG interpretation proficiency. The trial followed the CONSORT guidelines (24,25), and its design and rationale have been previously reported (23).
Approval
The educational review committee and institutional review board (IRB) at the Mayo Clinic, along with the IRBs at Washington University School of Medicine in St. Louis, SUNY Upstate Medical University, and Keck School of Medicine at the University of Southern California, and WCG IRB (an external IRB review company), granted the trial approval.
Participants
Healthcare professionals worldwide, aged 18 years or older, who were either in the process of training or had completed their training, were eligible to participate in the study. Practicing cardiologists and emergency medicine physicians were excluded. The study cohort comprised medical students, resident physicians (family medicine, internal medicine, and emergency medicine physicians), cardiology fellows-in-training, primary care physicians (family medicine and internal medicine physicians), nurses, advanced practice providers (physician assistants and nurse practitioners), and allied health professionals (ECG technicians, emergency medicine technicians, and paramedics). Physicians were defined as primary care physicians, cardiology fellows-in-training, and resident physicians, while non-physicians were defined as advanced practice providers, nurses, and allied health professionals. Participants were granted one-year free access to The EKG Guy learning resource upon completion, with no financial or educational incentives.
Trial Protocol
The trial protocol, depicted in Figure 1, began with voluntary participant registration and informed consent. Participants were then given a two-week period to complete a baseline survey and a pre-intervention test. Following this, participants within varying professional groups were randomly assigned to one of four groups: (i) an online, ECG interpretation practice learning resource (question bank [Arm 1]); (ii) an online, lecture-based learning resource (lectures [Arm 2]); (iii) a hybrid of the first two (hybrid [Arm 3]); or (iv) a control group without ECG learning resources (control [Arm 4]). Irrespective of medical discipline, training level, or experience, the educational interventions in Arms 1 (question bank), 2 (lectures), and 3 (hybrid) were identical. The four-month intervention period was followed by a post-intervention test. Further details about the trial protocol and the learning platform are available in previously published work (23).
Figure 1. Study flowchart of the EDUCATE Trial.
Pre- and Post-Intervention Tests
The pre- and post-intervention tests each required participants to interpret 30 12-lead ECGs, each accompanied by a brief statement to provide clinical context (e.g., “35-year-old male with chest pain”). This testing approach intentionally aligns with the methodology used by the Cardiovascular Subspecialty Board of the American Board of Internal Medicine. Participants were instructed to determine ventricular rate, approximate mean QRS axis, and provide a comprehensive interpretation from a pre-determined list of interpretive statements for each ECG, all within a ten-minute timeframe. Participants were required to evaluate one ECG at a time and to complete all tasks prior to moving on to the next ECG. ECG interpretation was made without access to computerized interpretation reports or measurements. Upon submission of responses or time limit expiration, participants were asked to rate their interpretation confidence level as not confident, somewhat confident, or confident. Participants were not permitted to view previously interpreted ECGs, and they were kept uninformed about their performance throughout the assessment. Participants were given a 14-day window to independently complete each test, which could be extended by an additional seven days in the event of unforeseen circumstances.
Both the pre- and post-intervention tests utilized the same set of ECGs, clinical statements, and testing format, encompassing 69 urgent and non-urgent findings that are commonly taught and encountered in clinical practice. The reference standard was established through a consensus among three board-certified cardiologists (P.A.N., N.S.A., and A.M.M.) and a cardiology fellow-in-training (A.H.K.). Further details about the pre- and post-intervention tests can be found in previous studies (23).
Data Processing
All data, de-identified and securely stored, were exported for analysis, maintaining participant and institutional confidentiality.
Performance Evaluation Metrics
To assess ECG interpretation performance, adjusted and unadjusted scores were used.
Unadjusted scores signified the percentage of correctly identified findings. Each correctly identified finding was assigned one point, while incorrect findings received zero points. The total points for each ECG varied depending on the number of findings present. Maximum overall unadjusted score was 100.0%, indicating correctly identifying all 69 interpretative findings.
Adjusted scores represented the percentage of points allocated based on clinical relevance in a scoring system. Each ECG finding was assigned a point value, with higher points assigned to more significant findings (23). The points assigned to individual ECGs varied depending on their clinical relevance and the number of findings present. Incorrect selections resulted in point deductions, with a minimum score of zero for any single ECG. Investigators A.H.K. and A.M.M. determined magnitude of point assignment and deductions. The maximum achievable overall adjusted score was 100.0%, indicating the successful accumulation of all 216 points.
Outcomes
Trial outcomes were determined through the pre- and post-intervention tests. The primary objective was to determine if the educational interventions improved ECG interpretation performance. Secondary objectives included evaluating changes in interpretation confidence, time, and accuracy for specific ECG findings.
Primary outcome.
The primary outcome of interest was the change in the average unadjusted and adjusted scores between pre- and post-intervention tests in each group. We also conducted subanalyses of the primary outcome between professional groups.
Secondary outcomes.
Secondary outcomes included the change in interpretation confidence, time, and accuracy for specific ECG findings between pre- and post-intervention tests in each group. Specific ECG findings evaluated included ventricular rate, mean QRS axis, primary rhythm, sinus rhythm, atrial fibrillation, premature atrial and ventricular complexes, atrioventricular (AV) blocks, bundle branch blocks, ST-elevation myocardial infarctions (STEMIs), emergencies, left ventricular hypertrophy (LVH), and pericarditis. The emergencies category comprised STEMIs, ventricular tachycardia, ventricular fibrillation, and third-degree AV block.
Statistical Analysis
To account for a potential 50% attrition rate, we aimed to enroll 1000 participants, ensuring that each arm had at least 250 participants. This targeted number required a minimum of 87 participants in each arm for the final analysis to achieve a 90% confidence level that the true value would fall within 7% of the primary endpoint. All participants who underwent randomization and completed the post-intervention test were included in primary and secondary analyses.
Descriptive statistics summarized data, with nominal and continuous variables reported as a count (percent of the total) and mean with standard deviation, respectively. We used paired t tests/analysis of variance or χ2 tests or Fisher exact probability tests for pairwise comparisons of survey responses and endpoints, and independent samples t tests for continuous outcomes. One-way ANOVA analysis compared performance metrics between study arms. Statistical significance was based on a two-tailed alpha <0.05. We conducted all statistical analyses using MedCalc for Windows, version 19.4 (MedCalc Software, Ostend, Belgium).
RESULTS
Participants
Among 3500 eligible healthcare professionals, 1206 (34.5%) successfully completed the enrollment process and underwent randomization (Figure 1). Table 1 outlines the baseline characteristics of randomized participants. Participants were evenly distributed across the four groups: 302 (25.0%) in question bank (Arm 1), 301 (25.0%) in lectures (Arm 2), 302 (25.0%) in hybrid (Arm 3), and 301 (25.0%) in the control (Arm 4). No significant differences (P>0.05) existed between the groups in terms of age distribution, sex, weekly ECG interpretation volume, ECG interpretation responsibility or comfort, prior dedicated ECG training, or expert interpreter supervision. Final analysis included 863 (72%) participants who completed the trial: 212 (24.6%) in Arm 1 (question bank), 221 (25.6%) in Arm 2 (lectures), 221 (25.6%) in Arm 3 (hybrid), and 209 (24.2%) in Arm 4 (control).
Table 1.
Baseline characteristics of participants.
| Characteristic – No. (%) | Question Bank (N = 301) | Lectures (N = 302) | Hybrid (N = 302) | Control (N = 301) | P Value |
|---|---|---|---|---|---|
| Age distribution | 0.990 | ||||
| 18-25 years | 45 (15.0) | 50 (16.6) | 45 (14.9) | 38 (12.6) | |
| 26-35 years | 176 (58.5) | 163 (54.0) | 162 (53.6) | 174 (57.8) | |
| 36-45 years | 46 (15.3) | 59 (19.5) | 54 (17.9) | 47 (15.6) | |
| 46-55 years | 23 (7.6) | 21 (7.0) | 29 (9.6) | 28 (9.3) | |
| >55 years | 11 (3.7) | 9 (3.0) | 12 (4.0) | 14 (4.7) | |
| Sex | 0.787 | ||||
| Female | 134 (44.5) | 145 (48.0) | 121 (40.1) | 131 (43.5) | |
| Male | 167 (55.5) | 157 (52.0) | 181 (59.9) | 170 (56.5) | |
| Location | 0.866 | ||||
| United States | 194 (64.5) | 190 (62.9) | 183 (60.6) | 174 (57.8) | |
| Outside United States | 107 (35.5) | 112 (37.1) | 119 (39.4) | 127 (42.2) | |
| Professional group | 1.000 | ||||
| Primary care physicians | 18 (6.0) | 18 (6.0) | 18 (6.0) | 18 (6.0) | |
| Cardiology fellows-in-training | 36 (12.0) | 36 (11.9) | 37 (12.3) | 37 (12.3) | |
| Resident physicians | 88 (29.2) | 89 (29.5) | 88 (29.1) | 88 (29.2) | |
| Medical students | 46 (15.3) | 45 (14.9) | 46 (15.2) | 45 (15.0) | |
| Advanced practice providers | 21 (7.0) | 21 (7.0) | 21 (7.0) | 21 (7.0) | |
| Nurses | 30 (10.0) | 30 (9.9) | 30 (9.9) | 30 (10.0) | |
| Allied health professionals | 62 (20.6) | 63 (20.9) | 62 (20.5) | 62 (20.6) | |
| Average ECG interpretations | 0.976 | ||||
| 0 per week | 57 (18.9) | 48 (15.9) | 62 (20.5) | 52 (17.3) | |
| 1-10 per week | 159 (52.8) | 174 (57.6) | 158 (52.3) | 159 (52.8) | |
| 11-25 per week | 55 (18.3) | 50 (16.6) | 60 (19.9) | 62 (20.6) | |
| >25 per week | 30 (10.0) | 30 (9.9) | 22 (7.3) | 28 (9.3) | |
| ECG interpretation responsibility | 1.000 | ||||
| Directly impacts patient care | 200 (66.4) | 198 (65.6) | 195 (64.6) | 209 (69.4) | |
| Indirectly impacts patient care | 42 (14.0) | 42 (13.9) | 44 (14.6) | 34 (11.3) | |
| No impact on patient care | 13 (4.3) | 17 (5.6) | 17 (5.6) | 13 (4.3) | |
| Not applicable | 46 (15.3) | 45 (14.9) | 46 (15.2) | 45 (15.0) | |
| ECG interpretation comfort | 1.000 | ||||
| Uncomfortable | 132 (43.9) | 137 (45.4) | 134 (44.4) | 132 (43.9) | |
| Somewhat comfortable | 132 (43.9) | 133 (44.0) | 133 (44.0) | 132 (43.9) | |
| Comfortable | 37 (12.3) | 32 (10.6) | 35 (11.6) | 37 (12.3) | |
| Prior dedicated ECG training | 0.668 | ||||
| 0 hours | 131 (43.5) | 142 (47.0) | 153 (50.7) | 126 (41.9) | |
| <5 hours | 90 (29.9) | 87 (28.8) | 68 (22.5) | 85 (28.2) | |
| 5-15 hours | 45 (15.0) | 46 (15.2) | 55 (18.2) | 62 (20.6) | |
| >15 hours | 35 (11.6) | 27 (8.9) | 26 (8.6) | 28 (9.3) | |
| Expert interpreter supervision | 0.999 | ||||
| None | 116 (38.5) | 130 (39.4) | 122 (40.4) | 119 (39.5) | |
| Rarely | 87 (28.9) | 67 (25.8) | 81 (26.8) | 83 (27.6) | |
| Somewhat often | 65 (21.6) | 66 (21.9) | 60 (19.9) | 60 (19.9) | |
| Often | 23 (7.6) | 27 (8.9) | 26 (8.6) | 25 (8.3) | |
| Very often | 10 (3.3) | 12 (4.0) | 13 (4.3) | 14 (4.7) |
Pre-Intervention Performance
Table 2 displays pre-intervention average unadjusted scores for all professional groups. In terms of average unadjusted score, no significant differences existed in pre-intervention performance (P=0.65): Arm 1 (question bank) scored 62.3 ± 15.5%, Arm 2 (lectures) scored 63.1 ± 14.7%, Arm 3 (hybrid) scored 61.4 ± 15.4%, and Arm 4 (control) scored 62.8 ± 13.6%. Cardiology fellows-in-training showed significantly higher pre-intervention unadjusted scores compared to other professional groups across all study arms (P<0.01).
Table 2.
Unadjusted scores by professional group.
| Professional Group | Question Bank (N=212) | Lectures (N=221) | Hybrid (N=221) | Control (N=209) | ||||
|---|---|---|---|---|---|---|---|---|
| Test Score Pre | Post |
Mean Difference P Value |
Test Score Pre | Post |
Mean Difference P Value |
Test Score Pre | Post |
Mean Difference P Value |
Test Score Pre | Post |
Mean Difference P Value |
|
| Primary care physicians | 61.6 | 76.9 | 15.3 ± 20.4 0.001 |
64.0 | 74.5 | 10.5 ± 16.2 0.047 |
71.8 | 74.4 | 2.6 ± 3.7 0.602 |
64.6 | 67.7 | 3.1 ± 10.6 0.635 |
| Cardiology fellows-in-training | 74.8 | 83.2 | 8.4 ± 15.2 0.003 |
73.7 | 80.0 | 6.4 ± 14.1 0.027 |
70.1 | 79.7 | 9.6 ± 8.5 <0.001 |
69.2 | 68.7 | −0.4 ± 12.8 0.851 |
| Resident physicians | 65.4 | 72.6 | 7.2 ± 11.4 <0.001 |
66.6 | 73.8 | 7.2 ± 11.4 <0.001 |
65.7 | 71.4 | 5.7± 11.2 0.008 |
64.9 |62.7 | −2.1 ± 15.3 0.310 |
| Medical students | 59.0 | 71.7 | 12.8 ± 16.4 0.001 |
55.1 | 67.2 | 12.0 ± 17.4 <0.001 |
53.2 | 66.7 | 13.5 ± 15.5 0.002 |
54.9 | 61.2 | 6.3 ± 15.6 0.080 |
| Advanced practice providers | 59.2 | 63.0 | 3.8 ± 16.9 0.592 |
59.0 | 68.5 | 9.5 ± 11.6 0.073 |
49.0 |65.9 | 16.9 ± 12.6 <0.001 |
57.2 | 62.4 | 5.2 ± 12.8 0.350 |
| Nurses | 50.7 | 67.2 | 16.6 ± 23.1 0.008 |
55.9 | 68.3 | 12.4 ± 22.8 0.029 |
.1.0 | 70.8 | 19.8 ± 16.8 <0.001 |
62.7 | 59.9 | −2.7 ± 12.9 0.573 |
| Allied health professionals | 57.8 | 75.1 | 17.3 ± 18.6 <0.001 |
61.4 | 74.3 | 12.9 ± 15.9 <0.001 |
60.6 | 74.9 | 14.2 ± 16.7 <0.001 |
62.9 | 63.9 | 1.0 ± 16.1 0.726 |
| Physicians1 | 67.3 | 76.1 | 8.8 ± 14.3 <0.001 |
68.5 | 75.9 | 7.4 ± 12.9 <0.001 |
67.8 | 74.2 | 6.4 ± 9.9 <0.001 |
65.9 | 65.0 | −0.9 ± 14.1 0.610 |
| Non-physicians1 | 56.2 | 71.1 | 14.9 ± 19.9 <0.001 |
59.6 | 71.7 | 12.1 ± 16.9 <0.001 |
56.2 | 72.3 | 16.1 ± 16.1 <0.001 |
61.8 | 62.6 | 0.8 ± 14.8 0.719 |
| All participants | 62.3 | 73.7 | 11.4 ± 16.9 <0.001 |
63.1 | 73.0 | 9.8 ± 15.3 <0.001 |
61.4 | 72.4 | 11.0 ± 14.0 <0.001 |
62.8 | 63.6 | 0.8 ± 14.7 0.544 |
Physicians include primary care physicians, cardiology fellows-in-training, and physician residents, whereas non-physicians include advanced practice providers, nurses, and allied health professionals.
Primary Outcome
All intervention groups (question bank [Arm 1], lectures [Arm 2], and hybrid [Arm 3]) exhibited substantial improvements in the primary outcome of unadjusted score (P<0.01) and adjusted score (P<0.01) (Table 2 and Figure 2). The question bank, lectures, and hybrid intervention groups demonstrated significant average unadjusted score improvements of 11.4% (95% CI, 9.1 to 13.7; P<0.01), 9.8% (95% CI, 7.8 to 11.9; P<0.01), and 11.0% (95% CI, 9.2 to 12.9; P<0.01), respectively, whereas the control group exhibited a marginal and non-significant improvement of 0.8% (95% CI, −1.2 to 2.8; P=0.54). Subgroup analyses based on professional group revealed similar significant improvements (P<0.05), except for primary care physicians in the hybrid intervention (n=15; P=0.602) and advanced practice providers in the question bank (n=12; P=0.592) and lectures (n=17; P=0.073) interventions (Table 2 and Figure 2).
Figure 2. Subgroup analyses for mean difference in unadjusted scores.
The value to the left of each bar represents the number of participants.
1 Physicians include primary care physicians, cardiology fellows-in-training, and physician residents, whereas non-physicians include advanced practice providers, nurses, and allied health professionals.
Secondary Outcomes
All groups demonstrated significantly higher confidence levels compared to the pre-intervention test (P<0.01) (Table 3). Among intervention groups (question bank [Arm 1], lectures [Arm 2], and hybrid [Arm 3]), no significant differences in confidence levels were observed (P=1.00). Interpretation time did not significantly change within the intervention groups (P>0.05), contrary to the control group, which showed a significant reduction in time of 39 seconds (P<0.05).
Table 3.
Primary and secondary outcomes.
| Outcome | Question Bank (N=212) | Lectures (N=221) | Hybrid (N=221) | Control (N=209) | ||||
|---|---|---|---|---|---|---|---|---|
| Mean Difference | P Value | Mean Difference | P Value | Mean Difference | P Value | Mean Difference | P Value | |
| Primary outcomes | ||||||||
| Unadjusted score1 | 11.4 ± 16.9 | <0.001 | 9.8 ± 15.3 | <0.001 | 11.0 ± 14.0 | <0.001 | 0.8 ± 14.7 | 0.544 |
| Adjusted score2 | 12.3 ± 16.8 | <0.001 | 13.6 ± 15.7 | <0.001 | 15.4 ± 14.5 | <0.001 | −0.8 ± 15.3 | 0.570 |
| Secondary outcomes | ||||||||
| Confidence | 0.30 ± 0.58 | <0.001 | 0.24 ± 0.53 | <0.001 | 0.35 ± 0.62 | <0.001 | 0.20 ± 0.49 | <0.001 |
| Interpretation time (s) | 0 ± 79 | 1.000 | −6 ± 77 | 0.355 | 11 ± 72 | 0.055 | −39 ± 69 | <0.001 |
| Ventricular rate | 7.2 ± 20.3 | <0.001 | 8.3 ± 17.2 | <0.001 | 9.4 ± 16.1 | <0.001 | 0.8 ± 20.5 | 0.600 |
| QRS axis | 10.0 ± 24.1 | <0.001 | 7.5 ± 18.1 | <0.001 | 9.7 ± 17.0 | <0.001 | 4.3 ± 21.7 | 0.027 |
| Primary rhythm | 17.6 ± 21.9 | <0.001 | 15.0 ± 24.9 | <0.001 | 17.4 ± 23.4 | <0.001 | 3.7 ± 27.5 | 0.064 |
| Sinus rhythm | 18.5 ± 26.9 | <0.001 | 15.0 ± 30.1 | <0.001 | 18.6 ± 30.7 | <0.001 | 4.0 ± 34.5 | 0.101 |
| Atrial fibrillation | 23.0 ± 45.9 | <0.001 | 20.4 ± 41.6 | <0.001 | 21.9 ± 42.0 | <0.001 | 11.0 ± 41.4 | 0.002 |
| PAC/PVC | 13.4 ± 41.5 | <0.001 | 15.7 ± 36.6 | <0.001 | 15.5 ± 37.3 | <0.001 | 8.0 ± 40.3 | 0.023 |
| AV block | 15.6 ± 35.8 | <0.001 | 10.6 ± 33.3 | <0.001 | 12.5 ± 33.8 | <0.001 | 2.5 ± 36.5 | 0.405 |
| STEMI3 | 10.2 ± 32.4 | <0.001 | 9.1 ± 30.4 | <0.001 | 8.3 ± 30.9 | 0.001 | −3.4 ± 35.3 | 0.228 |
| Emergencies4 | 11.1 ± 28.6 | <0.001 | 8.7 ± 25.4 | <0.001 | 8.9 ± 26.8 | <0.001 | −2.3 ± 31.9 | 0.431 |
| BBB | 21.1 ± 44.2 | <0.001 | 19.3 ± 38.5 | <0.001 | 17.9 ± 42.9 | <0.001 | 4.8 ± 40.7 | 0.177 |
| LVH | 5.7 ± 33.7 | 0.040 | 6.9 ± 35.8 | 0.011 | 1.8 ± 28.9 | 0.408 | 5.1 ± 34.5 | 0.059 |
| Pericarditis | 14.6 ± 56.1 | 0.002 | 15.8 ± 59.3 | <0.001 | 20.4 ± 56.3 | <0.001 | −6.7 ± 60.4 | 0.172 |
Unadjusted score (%) reflects how often a participant correctly identified all ECG findings, assigning one point for a correct response and zero points for an incorrect response.
Adjusted score (%) represents how often a participant correctly identified ECG findings and incorporates a weighted-point based system depending on the clinical relevance of a finding as well as the potential to incur penalties for incorrect responses.
STEMI recognition, not localization, was evaluated.
Emergencies include STEMI2, ventricular tachycardia, ventricular fibrillation, and third-degree AV block.
Abbreviations: PAC, premature atrial complex; PVC, premature ventricular complex; AV, atrioventricular; STEMI, ST-elevation myocardial infarction; BBB, bundle branch block; LVH, left ventricular hypertrophy.
The intervention groups (question bank [Arm 1], lectures [Arm 2], and hybrid [Arm 3]) displayed a significant improvement in unadjusted scores for determining ventricular rate, QRS axis, and the primary rhythm as well as in accurately identifying sinus rhythm, atrial fibrillation, premature atrial and ventricular complexes, AV blocks, STEMIs, emergencies, and bundle branch blocks (P<0.05) (Table 3). No significant differences (P>0.05) were noted between the intervention groups in the post-intervention unadjusted scores for these outcomes, except for atrial fibrillation by the hybrid intervention group, which had a lower baseline unadjusted score. Intervention groups outperformed the control group in every evaluated outcome, except for LVH detection by the hybrid intervention group.
DISCUSSION
The EDUCATE Trial is the first international, prospective, randomized controlled trial to appraise the effectiveness of web-based, self-directed learning approaches in enhancing ECG interpretation proficiency among healthcare professionals. This study addresses the urgent need for practical, accessible, and evidence-based solutions to improve the ECG interpretation skills of various medical learners and professionals.
Main Findings
The EDUCATE Trial demonstrates the efficacy of web-based, self-directed learning approaches in enhancing ECG interpretation skills across a wide range of healthcare professionals. While no single method has proven most effective for teaching ECG interpretation (16–19), we observed that web-based interventions produced significant improvements in performance among all healthcare professional subgroups. We find it particularly encouraging that the post-intervention scores of primary care physicians, resident physicians, and allied health professionals surpassed the pre-intervention scores of cardiology fellows-in-training for each intervention group. We also noted significant improvements in primary rhythm analysis and the identification of critical findings, including STEMI and ECG emergencies. Such findings suggest that similar educational initiatives could potentially reduce instances of suboptimal care for high-acuity cardiovascular conditions. Moreover, these gains in ECG interpretation were achieved within a short period of time, and were coupled with an increase in self-confidence, which further highlights the effectiveness of web-based learning options in promoting ECG interpretation proficiency and autonomy.
Furthermore, this study supports and builds on previous research that highlighted significant gaps in ECG interpretation proficiency among healthcare professionals (6–14). Despite consistent improvements made by participants in the intervention groups, we observed persistent shortcomings in ECG interpretation performance. Several factors could have contributed to these lingering deficiencies. First, the educational content, developed well before the trial, was not specifically tailored to the pre- and post-intervention tests, allowing an imperfect alignment between the content and desired learning outcomes. Second, the voluntary nature of the trial resulted in varying engagement levels due to the absence of mandatory schedules. Third, the lack of actively involved instructors, who could provide regular updates and clarification, may have impeded learner engagement, accountability, and comprehension. Lastly, the random allocation of learners to education interventions that do not align with their individual training levels or learning preferences may have limited their engagement and the benefits derived from the resources. For example, participants with limited ECG interpretation experience may not fully benefit from an ECG interpretation practice question-based approach (i.e., question bank intervention) if they lack a solid foundational knowledge base.
Trial Implications
The trial findings offer valuable insights on how to achieve skill enhancement in the domain of ECG interpretation. Moreover, the study presents a viable solution to the pressing need of effective and accessible ECG interpretation training that caters to diverse healthcare learners and professionals. It also highlights the feasibility of integrating web-based learning resources into existing teaching frameworks, personal/work schedules, and care team processes.
Beyond the scope of ECG interpretation, this trial has broader implications for medical education and training. With digital technology advancements in educational environments, medical educators and learners are faced with new challenges (15). However, our work effectively demonstrates the feasibility and practicality of developing, implementing, and rigorously evaluating web-based educational approaches to enhance clinical skills and knowledge. Moreover, it provides a framework for addressing and evaluating other potential educational solutions aimed at addressing deficiencies in clinical competency and medical knowledge.
Trial Strengths and Limitations
The trial exhibits numerous primary strengths, including its randomized controlled design, international scope, large and diverse participant cohort, wide range of evaluated ECG findings, and incorporation of a user-friendly, universally accessible virtual testing environment and learning curriculum. These aspects enabled a comprehensive evaluation of the effectiveness of web-based, self-directed learning approaches in improving ECG interpretation proficiency without directly disrupting the existing educational and clinical responsibilities of healthcare professionals.
Several limitations of the trial must be acknowledged. First, was the observed attrition, with 343 (28%) participants not completing the study. This introduces possible performance differences between those who dropped out and those who completed the study, potentially affecting generalizability of the results. Also, the trial’s short duration restricts the assessment of long-term maintenance of ECG interpretation proficiency. The ECGs utilized in this study focused on commonly taught and encountered findings, which may limit representation of the full range of clinical ECG interpretation. Additionally, interpreting ECGs in a non-clinical setting without computerized ECG interpretation support may limit applicability of the findings to real-world scenarios. Furthermore, the study did not establish a link between improved ECG interpretation proficiency and patient care outcomes, though we suspect that improvements in ECG interpretation competency among healthcare providers would likely to benefit patients over time. Lastly, excluding cardiologists and emergency medicine physicians from the study population limits the generalizability of the findings to these healthcare professionals.
CONCLUSIONS
Web-based, self-directed educational interventions demonstrated significant improvements in ECG interpretation proficiency among a wide range of healthcare professionals. This success suggests that similar educational initiatives hold promise in addressing other gaps in clinical skill competency and knowledge.
Support:
The EDUCATE Trial is an investigator-initiated study supported by GE HealthCare (Milwaukee, WI) and NIH T32 HL007111. GE HealthCare provided ECGs for the trial, but did not participate in study implementation, data analysis, or reporting. The EKG Guy platform provided educational resources, supported implementation, and facilitated data collection and analysis.
Disclosures:
A.H.K. is the founder and CEO of The EKG Guy, and has received research funding from GE HealthCare (Milwaukee, WI).
Footnotes
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Kadish AH, Buxton AE, Kennedy HL, Knight BP, Mason JW, Schuger CD, et al. American College of Cardiology/American Heart Association/American College of Physicians-American Society of Internal Medicine Task Force; International Society for Holter and Noninvasive Electrocardiology. ACC/AHA clinical competence statement on electrocardiography and ambulatory electrocardiography: a report of the ACC/AHA/ACP-ASIM task force on clinical competence (ACC/AHA Committee to develop a clinical competence statement on electrocardiography and ambulatory electrocardiography) endorsed by the International Society for Holter and noninvasive electrocardiology. Circulation. 2001;104(25):3169–3178. [PubMed] [Google Scholar]
- 2.Salerno SM, Alguire PC, Waxman HS. Competency in interpretation of 12-lead electrocardiograms: a summary and appraisal of published evidence. Ann Intern Med. 2003;138(9):751–760. [DOI] [PubMed] [Google Scholar]
- 3.Balady GJ, Bufalino VJ, Gulati M, Kuvin JT, Mendes LA, Schuller JL. COCATS 4 Task Force 3: training in electrocardiography, ambulatory electrocardiography, and exercise testing. J Am Coll Cardiol. 2015;65(17):1763–1777. [DOI] [PubMed] [Google Scholar]
- 4.Kashou A, May A, DeSimone C, Noseworthy P. The essential skill of ECG interpretation: How do we define and improve competency? Postgrad Med J. 2020;96(1133):125–127. [DOI] [PubMed] [Google Scholar]
- 5.Kashou AH, Noseworthy PA, Beckman TJ, Anavekar NS, Cullen MS, Angstman KB, et al. Exploring Factors Influencing ECG Interpretation Proficiency of Medical Professionals. Curr Probl Cardiol. 2023;48(10):101865. [DOI] [PubMed] [Google Scholar]
- 6.Kashou AH, Noseworthy PA, Beckman TJ, Anavekar NS, Cullen MS, Angstman KB, et al. ECG Interpretation Proficiency of Healthcare Professionals. Curr Probl Cardiol. 2023;48(10):101924. [DOI] [PubMed] [Google Scholar]
- 7.Jheeta JS, Narayan O, Krasemann T. Republished: Accuracy in interpreting the paediatric ECG: a UK-wide study and the need for improvement. Postgrad Med J. 2015;91(1078):436–438. [DOI] [PubMed] [Google Scholar]
- 8.Novotny T, Bond RR, Andrzeja I, Koc L, Sisakova M, Finlay DD, et al. Data analysis of diagnostic accuracies in 12-lead electrocardiogram interpretation by junior medical fellows. J Electrocardiol. 2015;48(6):988–994. [DOI] [PubMed] [Google Scholar]
- 9.Kewcharoen J, Charoenpoonsiri N, Thangjui S, Panthong S, Hongkan W. A comparison between peer-assisted learning and self-study for electrocardiography interpretation in Thai medical students. J Adv Med Educ Prof. 2020;8(1):18–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fisch C The clinical electrocardiogram: a classic. Circulation 1980;62:1–4.7379272 [Google Scholar]
- 11.Wellens HJ. The electrocardiogram 80 years after Einthoven. J Am Coll Cardiol. 1986:7(3):484–491. [DOI] [PubMed] [Google Scholar]
- 12.Hurst JW. The interpretation of electrocardiograms: pretense or a well-developed skill? Cardiol Clin. 2006;24(3):305–307, vii. [DOI] [PubMed] [Google Scholar]
- 13.Salerno SM, Alguire PC, Waxman HS. Competency in interpretation of 12-lead electrocardiograms: a summary and appraisal of published evidence. Ann Intern Med. 2003;138(9):751–760. [DOI] [PubMed] [Google Scholar]
- 14.Chen Y, Kunst E, Nasrawi D, Massey D, Johnston ANB, Keller K, et al. Nurses’ competency in electrocardiogram interpretation in acute care settings: a systematic review. J Adv Nurs. 2022;78(5):1245–1266. [DOI] [PubMed] [Google Scholar]
- 15.Das T, Kaur G, Nematollahi S, Ambinder D, Shfer K, Sulistio M, et al. Medical education in the digital era: a new paradigm for acquiring knowledge and building communities. JACC Adv. 2022;1(2):100031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mahler SA, Wolcott CJ, Swoboda TK, Wang H, Arnold TC. Techniques for teaching electrocardiogram interpretation: self-directed learning is less effective than a workshop or lecture. Med Educ. 2001;45(4):347–353. [DOI] [PubMed] [Google Scholar]
- 17.Zhang H, Hsu LL. The effectiveness of an education program on nurses’ knowledge of electrocardiogram interpretation. Int Emerg Nurs. 2013;21(4):247–251. [DOI] [PubMed] [Google Scholar]
- 18.Montassier E, Hardouin JB, Segard J, Batard E, Potel G, Planchon B, et al. e-Learning versus lecture-based courses in ECG interpretation for undergraduate medical students: a randomized noninferiority study. Eur J Emerg Med. 2016;23(2):108–113. [DOI] [PubMed] [Google Scholar]
- 19.Fent G, Gosai J, Purva M. Teaching the interpretation of electrocardiograms: which method is best? J Electrocardiol. 2015;48:190–193. [DOI] [PubMed] [Google Scholar]
- 20.Rui Z, Lian-Rui X, Rong-Zheng Y, Jing Z, Xue-Hong W, Chuan Z. Friend or Foe? Flipped classroom for undergraduate electrocardiogram learning: a randomized controlled study. BMC Med Educ. 2017;17(1):53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Barthelemy FX, Segard J, Fradin P, Hourdin N, Batard E, Pottier P, et al. ECG interpretation in Emergency Department residents: an update and e-learning as a resource to improve skills. Eur J Emerg Med. 2017;24(2):149–156. [DOI] [PubMed] [Google Scholar]
- 22.Pourmand A, Tanski M, Davis S, Shokoohi H, Lucas R, Zaver F. Educational technology improves ECG interpretation of acute myocardial infarction among medical students and emergency medicine residents. West J Emerg Med. 2015;16(1):133–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kashou AH, Noseworthy PA, Beckman TJ, Anavekar NS, Cullen MS, Angstman KB, et al. Education curriculum assessment for teaching electrocardiography: Rationale and design for the prospective, international, randomized controlled EDUCATE Trial. J Electrocardiol. 2023;80:166–173. [DOI] [PubMed] [Google Scholar]
- 24.Boutron I, Altman DG, Moher D, Schulz KF, Ravaud P, Group CN. CONSORT Statement for Randomized Trials of Nonpharmacologic Treatments: A 2017 Update and a CONSORT Extension for Nonpharmacologic Trial Abstracts. Ann Intern Med 2017;167(1):40–47. [DOI] [PubMed] [Google Scholar]
- 25.Juszczak E, Altman DG, Hopewell S, Schulz K. Reporting of Multi-Arm Parallel-Group Randomized Trials: Extension of the CONSORT 2010 Statement. JAMA. 2019;321(16):1610–1620. [DOI] [PubMed] [Google Scholar]