ECG interpretation is the skill of reading the electrical tracing produced by the heart and translating its waves, intervals, and segments into a clinical picture. A standard 12-lead ECG records the heart’s electrical activity from twelve different angles simultaneously, and each angle reveals something different about the heart’s structure, rhythm, and blood supply. The process sounds mechanical, but in practice it is full of subtlety: patterns that look alarming can be harmless, normal variants can mimic disease, and small technical errors during recording can send clinicians down the wrong diagnostic path.
What the Waves and Intervals Actually Represent
Every heartbeat generates a predictable sequence of electrical events, and each one leaves a signature on the ECG tracing. The P wave represents the electrical activation of the upper chambers. The QRS complex, the tallest and sharpest deflection, marks activation of the lower chambers. The T wave reflects the electrical recovery of those same lower chambers. The flat segments between these waves matter just as much: the PR interval tells you how long the signal takes to travel from the upper to the lower chambers, the QRS duration tells you whether the signal is spreading through normal pathways, and the QT interval captures the full cycle of activation and recovery.
These intervals have accepted normal ranges, but even high-quality ECG machines can disagree about where a wave starts and ends. A comparison of seven commercial ECG interpretation programs found that while most agreed closely on average measurements, the extent of large, clinically meaningful errors varied by a factor of two between programs. QT interval measurements showed the widest spread, with the gap between the smallest and largest deviation from the population mean reaching about 12 ms.1PubMed Central. Comparison of PR, QRS, and QT interval measurements by seven ECG interpretation programs That may sound small, but in situations where a few milliseconds determine whether a QT interval is dangerously long, these machine-to-machine differences matter.
Reading the Signs of a Heart Attack
The most urgent application of ECG interpretation is recognizing an acute heart attack, specifically the kind where a coronary artery is completely blocked. The ECG in this setting typically follows a recognizable sequence: it begins with tall, peaked T waves, progresses to ST-segment elevation (the segment between the QRS and T wave lifts above its normal baseline), and can eventually produce deep Q waves that reflect dead heart tissue.2PubMed. The evolution of electrocardiographic changes in ST-segment elevation myocardial infarction T-wave inversion and ST depression can appear before, during, or after the event, adding complexity to the picture.
Crucially, the specific leads showing these changes point to which coronary artery is blocked. Changes in leads I, aVL, and the chest leads V1 through V4 are strongly associated with a blockage in the left anterior descending artery, the vessel supplying the front wall of the heart. Changes in leads II, III, and aVF point toward the right coronary artery or the circumflex artery, which supply the bottom and back of the heart.3PubMed. Electrocardiographic localization of coronary artery narrowings: studies during myocardial ischemia and infarction in patients with one-vessel disease This localization is not just academic: it tells the team heading to the catheterization lab where to look first and what kind of complications to expect.
Speed matters enormously here. Getting an ECG done quickly after a patient walks through the emergency department door directly influences how fast treatment begins. Studies of triage protocols designed to shorten the time between arrival and ECG recording have consistently shown that cutting even a few minutes off that interval translates into faster balloon times, the moment the blocked artery is physically reopened. One intervention reduced the median time from arrival to artery reopening from 85 minutes to 75 minutes simply by streamlining the path to the ECG machine.4PubMed Central. Improving Door-to-balloon Time by Decreasing Door-to-ECG time for Walk-in STEMI Patients Another showed a drop from 81 to 70 minutes after restructuring how chest pain patients were routed.5Scientific Reports. A modified cardiac triage strategy reduces door to ECG time in patients with ST elevation myocardial infarction
Wide QRS Tachycardias and Why Getting Them Right Matters
When the heart beats fast and the QRS complex is wider than 120 milliseconds, the clinician faces one of the most consequential decisions in ECG interpretation: is this a rhythm originating from the ventricles, or is it a rhythm from above the ventricles that just looks wide because the electrical pathway is abnormal? The distinction is critical because giving a drug meant for the second type to a patient actually experiencing the first can cause hemodynamic collapse.6PubMed Central. Differential Diagnosis of Wide QRS Tachycardias
Several diagnostic algorithms have been developed over the decades to help clinicians work through this problem step by step. When tested head to head, these algorithms perform similarly on the whole, with sensitivities around 90%, but their specificity is lower and a significant fraction of cases still get misclassified. One comparison found that the percentage of incorrect classifications averaged roughly 25% across two well-known approaches.7EP Europace. Differential diagnosis of wide QRS tachycardias: comparison of two electrocardiographic algorithms In clinical practice, when uncertainty persists, the safest default is to treat a wide QRS tachycardia as if it is ventricular in origin until proven otherwise.8PubMed Central. Wide Complex Tachycardia Differentiation: A Reappraisal of the State-of-the-Art
Conduction Blocks and Accessory Pathways
The heart’s electrical wiring system can develop blocks at various points. A right bundle branch block, one of the most common conduction abnormalities, occurs when the normal pathway to the right ventricle is interrupted. This forces the right ventricle to be activated by a slower, roundabout route, producing a characteristically widened QRS complex with distinctive changes in wave direction.9PubMed Central. Right Bundle Branch Block: Current Considerations A right bundle branch block can appear in perfectly healthy hearts, but it can also signal underlying heart or lung disease, so context matters.
On the opposite end of the spectrum, some people are born with extra electrical connections between the upper and lower chambers. Wolff-Parkinson-White syndrome is the best-known example. The hallmark ECG finding is a “delta wave,” a slurred upstroke at the beginning of the QRS complex caused by early activation of the ventricle through the accessory pathway.10PubMed. Wolff-Parkinson-White (WPW) syndrome: the detection of delta wave in an electrocardiogram (ECG) People with this pattern can experience rapid heart rhythms when the extra pathway forms part of a short circuit, and in rare cases these episodes can become dangerous. Identifying the location of the accessory pathway from the ECG helps guide decisions about whether and where to perform an ablation procedure.11European Journal of Clinical Investigation. Localization of accessory pathways in Wolff‐Parkinson‐white syndrome using ECG‐based multi‐task deep learning
What Electrolytes and Medications Do to the ECG
The ECG is surprisingly sensitive to the body’s internal chemistry. Potassium levels, in particular, leave clear fingerprints on the tracing. When potassium rises too high, the T waves become tall and peaked, the QRS complex widens, and the P wave shrinks. In severe cases, the QRS can widen so dramatically that it merges with the T wave, a pattern that signals imminent cardiac arrest.12PubMed Central. ECG frequency changes in potassium disorders: a narrative review Low potassium produces its own set of changes, including flattened T waves and the appearance of U waves after the T wave. Emergency physicians routinely glance at the ECG for these patterns before lab results come back, because the turnaround time for a blood draw can be much longer than the time it takes to acquire a 10-second ECG.
Medications represent another major source of ECG changes. Over a hundred drugs on the market can prolong the QT interval, including widely prescribed antibiotics, antidepressants, and heart rhythm medications.13PubMed Central. Drug-induced QT interval prolongation and torsades de pointes: Role of the pharmacist in risk assessment, prevention and management A prolonged QT interval matters because it creates the conditions for a distinctive and dangerous arrhythmia called torsades de pointes, a rapid polymorphic rhythm that is almost always triggered by drugs rather than arising spontaneously.14PubMed Central. Drug-Induced QT Prolongation And Torsades de Pointes The risk is especially high when the corrected QT interval exceeds 500 ms, when multiple QT-prolonging drugs are taken together, or when electrolyte levels are low. Genetic susceptibility also plays a role: some people carry the same gene variants found in congenital long QT syndrome, making them more vulnerable to drug-induced prolongation even at standard doses.15PubMed. Predicting the Unpredictable: Drug-Induced QT Prolongation and Torsades de Pointes
When Normal Hearts Produce Abnormal-Looking ECGs
One of the trickiest areas of ECG interpretation is distinguishing genuinely pathological findings from normal variants that look alarming. Athletes are a classic example. Sustained training enlarges the heart, and these structural changes show up on the ECG as voltage increases, repolarization changes, and sometimes patterns that overlap with hypertrophic cardiomyopathy, a genetic heart muscle disease that is a leading cause of sudden cardiac death in young athletes.16PubMed Central. Distinguishing hypertrophic cardiomyopathy from athlete’s heart: a clinical problem of increasing magnitude and significance One study found that measuring the combined amplitude of Q and S waves in lead III could help separate the two: athletes averaged much lower combined amplitudes than patients with hypertrophic cardiomyopathy, and adding a specific cutoff to existing screening criteria improved detection sensitivity from about 64% to 70% with only a small trade-off in specificity.17Heart. Large Q and S waves in lead III on the electrocardiogram distinguish patients with hypertrophic cardiomyopathy from athletes
Early repolarization is another common source of confusion. This benign variant causes slight ST-segment elevation, which can look similar to acute pericarditis or even a heart attack. A prospective study of over 100 patients found that the ECG alone correctly identified early repolarization about 97% of the time, but its reliability for confirming acute pericarditis was much lower, at only 20%.18PubMed. Repolarization variant vs acute pericarditis. A prospective electrocardiographic and echocardiographic evaluation In other words, when the ECG said “this is just early repolarization,” it was almost always right; but when it suggested pericarditis, it was wrong four out of five times. Several features help tilt the diagnosis one way or the other. ST-segment depression in lead V1, PR-segment deviations in both limb and chest leads, and a horizontal ST vector pointing to the left all favor pericarditis, while a vertical ST vector and a flat ST segment in V6 favor benign early repolarization.19PubMed. Differential characteristics of the electrocardiogram in early repolarization and acute pericarditis Despite these clues, the overlap remains significant enough that some patients with early repolarization undergo unnecessary workups when clinicians interpret the pattern as pathological.20PubMed Central. Early Repolarization vs. Acute Pericarditis Morphology: A Case Report of Electrocardiographic Mimicry
Detecting Structural Heart Disease on Paper
Clinicians have long used voltage criteria on the ECG to estimate whether the left ventricle is enlarged, a condition called left ventricular hypertrophy. The problem is that these criteria have consistently low sensitivity. The ECG can detect increased voltage when the heart muscle is thick, but it misses a large fraction of cases that imaging would catch.21PubMed Central. The role of ECG in the diagnosis of left ventricular hypertrophy Despite this limitation, the ECG retains real prognostic value. A large study of over 9,000 patients with high blood pressure found that those who met two different ECG criteria for left ventricular hypertrophy simultaneously had more than three times the risk of stroke, heart attack, or cardiovascular death compared with those who met neither criterion.22PubMed Central. Combining ECG Criteria for Left ventricular Hypertrophy Improves Risk Prediction in Patients With Hypertension The ECG may not be a great tool for measuring how thick the wall is, but the electrical abnormalities it picks up carry independent prognostic weight.
Technical Mistakes That Change the Diagnosis
Before you can interpret an ECG, you need a correctly recorded one, and this is where things go wrong more often than most people realize. Misplacing the electrode stickers on the chest or swapping limb lead cables produces tracings that can mimic disease. One study found that applying the wrong electrode configuration creates a 17% to 24% chance that the resulting interpretation will differ from what the correct placement would have shown.23PubMed. The effects of electrode misplacement on clinicians’ interpretation of the standard 12-lead electrocardiogram Swapping the right-arm and left-arm leads, for example, can flip the axis and create the appearance of pathological Q waves in lead I. Placing chest electrodes one rib space too high can flatten or invert T waves in ways that mimic ischemia.
A systematic review of machine learning approaches to detecting electrode misplacement found that algorithms could identify most types of swaps with high accuracy, but left-arm/left-leg interchange remained difficult even for software to catch.24PubMed. Machine learning techniques for detecting electrode misplacement and interchanges when recording ECGs: A systematic review and meta-analysis Other common artifacts include muscle tremor, which creates a jagged baseline that can obscure P waves, and electrical interference from nearby equipment. Recognizing these artifacts is a core part of ECG interpretation, because failing to identify them means either missing a real finding hidden underneath or, worse, treating a recording error as if it were a disease.25PubMed Central. Technical mistakes during the acquisition of the electrocardiogram
What Artificial Intelligence Adds
Automated ECG interpretation has existed for decades, but recent AI models built on deep neural networks have pushed accuracy well beyond older rule-based systems. One comprehensive AI program demonstrated an area under the curve of 0.98 or higher for 62 out of 66 diagnostic categories, with sensitivity at or above 95% across all categories tested.26Cardiovascular Digital Health Journal. A comprehensive artificial intelligence–enabled electrocardiogram interpretation program In a head-to-head emergency department comparison, a deep learning algorithm achieved overall accurate interpretation for about 72% of ECGs, compared with roughly 60% for a conventional algorithm, with the biggest gains coming in specificity and positive predictive value rather than raw sensitivity.27PubMed. A deep neural network learning algorithm outperforms a conventional algorithm for emergency department electrocardiogram interpretation
Perhaps more interesting than pattern matching is what AI can extract from the ECG that human eyes cannot. A landmark study trained a neural network on nearly 45,000 paired ECGs and echocardiograms and found that the algorithm could detect reduced heart pumping function from the ECG alone with an area under the curve of 0.93, sensitivity of about 86%, and specificity of about 86%. Even more striking, patients whose AI screen was falsely positive — meaning the algorithm flagged them but their pump function was currently normal — went on to develop reduced function at four times the rate of those with a negative screen.28Nature Medicine. Screening for cardiac contractile dysfunction using an artificial intelligence–enabled electrocardiogram A population-level validation confirmed these results, with the AI maintaining an area under the curve of 0.97 for detecting poor pump function even in a general screening population where only about 2% of people had the condition.29PubMed Central. Artificial Intelligence-Augmented Electrocardiogram Detection of Left Ventricular Systolic Dysfunction in the General Population
A meta-analysis pooling thirteen studies of AI-based heart failure detection from ECG data reported pooled sensitivity of 93% and specificity of 95%.30PubMed Central. Efficacy of AI Models in Detecting Heart Failure Using ECG Data: A Systematic Review and Meta-Analysis These tools are starting to shift the ECG from a diagnostic test — “what is wrong right now?” — toward a screening tool: “what is likely to go wrong soon?”
How Smartwatch ECGs Compare
Consumer devices like the Apple Watch can now record a single-lead ECG from the wrist. This is fundamentally different from a clinical 12-lead recording. A smartwatch captures roughly the equivalent of lead I, which limits what it can detect. An emergency department study found that the Apple Watch ECG performed similarly to a standard ECG when distinguishing normal from abnormal rhythms, but there was a significant difference in its ability to identify specific abnormal findings within that single lead.31PubMed. Diagnostic accuracy of apple watch ECG outputs in identifying dysrhythmias: A comparison with 12-Lead ECG in emergency department In practical terms, a smartwatch ECG is useful for capturing an irregular rhythm while it is happening (especially atrial fibrillation, which is what most smartwatch algorithms are designed to detect), but it cannot localize a heart attack, assess for hypertrophy, or identify most conduction abnormalities. Think of it as a single snapshot from one camera versus a full panoramic view from twelve.
Pediatric ECGs Play by Different Rules
Children’s ECGs are not just smaller versions of adult tracings. The normal ranges for heart rate, intervals, and wave amplitudes change substantially with age, and patterns that would be alarming in an adult are completely expected in a child. T-wave inversion in leads V1 through V3, for example, is normal in children between about five and nine years old, and inversion isolated to V1 persists as a normal finding through the teenage years.32PubMed. Normal electrocardiographic patterns in school-aged children: Data from a sports cardiology screening program In an adult, T-wave inversion in V1 through V3 would raise immediate concern for ischemia or cardiomyopathy.
Age-specific reference standards have been developed using Z-scores across over a hundred ECG variables, covering everything from wave amplitudes in each lead to PR and QT intervals, all stratified by age.33PubMed. Electrocardiogram Standards for Children and Young Adults Using Z-Scores Applying adult criteria to a pediatric ECG is a reliable way to generate false alarms. This is one reason pediatric ECGs are often read by specialists or with dedicated pediatric software, and it is a common pitfall when adult-trained clinicians encounter a child’s tracing for the first time. The electrical dominance of the right ventricle in newborns, the gradual leftward shift of the axis through childhood, and the faster resting heart rates all produce patterns that adult interpretation frameworks would flag as abnormal. Knowing which findings are age-appropriate is as much a part of ECG interpretation as recognizing disease patterns.