Clinical Electrocardiography: A Simplified Approach, 7th Edition (2006)


Chapter 23. Limitations and Uses of the ECG

Throughout this book, the clinical uses of the ECG have been stressed. This review chapter underscores some important limitations of the ECG, reemphasizes its utility, and reviews some common pitfallsin its interpretation.


Like most clinical tests, the ECG yields both false-positive and false-negative results. A false-positive result is an apparently abnormal ECG in a normal subject. For example, prominent precordial voltage may occur in the absence of left ventricular hypertrophy (see Chapter 6 ). Furthermore, Q waves may occur as a normal variant and therefore do not always indicate heart disease (see Chapters 8and 9 ). Falsenegative results, on the other hand, occur when the ECG fails to show evidence of some cardiac abnormality. For example, some patients with acute myocardial infarction (MI) may not show diagnostic ST-T changes, and patients with severe coronary artery disease may not show diagnostic ST depressions during stress testing ( Chapter 9 ).

The diagnostic accuracy of any test is determined by the percentages of false-positive and false-negative results. The sensitivity of a test is a measure of the percentage of patients with a particular abnormality that can be identified by an abnormal test result. For example, a test with 100% sensitivity has no false-negative results. The more false-negative results, the less sensitive is the test. Thespecificity of a test is a measure of the percentage of false-positive results. The more false-positive test results, the less specific is the test.

As just noted, both the sensitivity and specificity of the ECG in diagnosing a variety of conditions, including MI, are limited. Clinicians need to be aware of these diagnostic limitations. The following are some important problems that cannot be excluded simply because the ECG is normal or shows only nondiagnostic abnormalities:



Prior MI



Acute MI[*]



Severe coronary artery disease



Significant left ventricular hypertrophy (LVH)



Significant right ventricular hypertrophy (RVH)



Intermittent arrhythmias such as paroxysmal atrial fibrillation (AF), paroxysmal supraventricular tachycardia (PSVT), ventricular tachycardia (VT), and bradycardias



Acute pulmonary embolism



Pericarditis, acute or chronic

*  The pattern of acute MI may also be masked in patients with left bundle branch block, Wolff-Parkinson-White preexcitation patterns, or electronic ventricular pacemakers.



Although the ECG has definite limitations, it often helps in the diagnosis of specific cardiac conditions and sometimes aids in the evaluation and management of general medical problems such as life-threatening electrolyte disorders ( Box 23-1 ). Some particular areas in which the ECG may be helpful are as follows:



Myocardial infarction Most patients with acute MI show diagnostic ECG changes (i.e., new Q waves and/or ST elevations, hyperacute T waves, ST depressions, or T wave inversions). In the weeks and months after an acute MI, however, these changes may become less apparent and, in some cases, may disappear. 
ST segment elevation in a right chest lead (e.g., V4R) in a patient with acute inferior infarction suggests associated right ventricular ischemia or infarction (see Chapter 8 ). 
Persistent ST elevations several weeks after an MI should suggest a ventricular aneurysm.



Pulmonary embolism 
A new SIQIII or right bundle branch block pattern, particularly in association with sinus tachycardia, should suggest the possibility of acute cor pulmonale resulting from, for example, pulmonary embolism (see Chapter 11 ).



Pericardial tamponade 
Low QRS voltage in a patient with elevated central venous pressure (distended neck veins) and sinus tachycardia suggests possible pericardial tamponade. Sinus tachycardia with electrical alternans is virtually diagnostic of pericardial effusion with tamponade (see Chapter 11 ).



Aortic valve disease 
LVH is seen in most patients with critical aortic stenosis or severe aortic regurgitation.



Mitral valve disease 
ECG signs of left atrial enlargement (abnormality) and RVH strongly suggest mitral stenosis ( Fig. 23-1 ). 
Frequent ventricular premature beats (VPBs) and nonspecific ST-T changes, particularly in a younger patient, should prompt a search for mitral valve prolapse.



Atrial septal defect 
Most patients with a moderate to large atrial septal defect have a right bundle branch block pattern.



Severe hyperkalemia, a life-threatening electrolyte abnormality, virtually always produces ECG changes, beginning with T wave peaking, loss of P waves, QRS widening, and finally asystole (see Chapter 10 ).



Renal failure 
The triad of LVH (caused by hypertension), peaked T waves (caused by hyperkalemia), and prolonged QT interval (caused by hypocalcemia) should suggest chronic renal failure.



Thyroid disease 
The combination of low voltage and sinus bradycardia should suggest possible hypothyroidism. (“Low and slow—think hypo.”) 
Unexplained AF (or sinus tachycardia at rest) should prompt a search for hyperthyroidism.



Chronic lung disease 
The combination of low voltage and slow precordial R wave progression is commonly seen with chronic obstructive lung disease.



The ECG-CHF (congestive heart failure) triad of relatively low limb lead voltage, prominent precordial voltage, and slow R wave progression suggests an underlying dilated cardiomyopathy (see Chapter 11 ).

BOX 23-1 

ECG as a Clue to Acute Life-Threatening Conditions without Primary Heart or Lung Disease

Cerebrovascular accident (especially intracranial bleed)

Drug toxicity


Tricyclic antidepressant overdose

Electrolyte disorders









Endocrine disorders






FIGURE 23-1  This ECG from a 45-year-old woman with severe mitral stenosis shows multiple abnormalities. The rhythm is sinus tachycardia. Right axis deviation and a tall R wave in lead V1 indicate right ventricular hypertrophy. The prominent biphasic P wave in lead V1 indicates left atrial abnormality (enlargement). The tall P wave in lead II may indicate concomitant right atrial enlargement (biatrial abnormality). Nonspecific ST-T changes and incomplete right bundle branch block are also present. The combination of right ventricular hypertrophy and left atrial abnormality (enlargement) is highly suggestive of mitral stenosis.



The ECG may also provide important and immediately available clues in the evaluation of such major medical problems as syncopecomashock, and weakness.


Fainting (transient loss of consciousness) can result from primary cardiac factors and various noncardiac causes. The primary cardiac causes can be divided into mechanical obstructions (aortic stenosis, primary pulmonary hypertension, or atrial myxoma) and electrical problems (bradyarrhythmias or tachyarrhythmias). The noncardiac causes of syncope include neurogenic mechanisms (e.g., vasovagal attacks), orthostatic (postural) hypotension, and brain dysfunction from vascular insufficiency or metabolic derangements (e.g., from alcohol or hypoglycemia).

Patients with syncope resulting from aortic stenosis generally show LVH on their resting ECG. Primary pulmonary hypertension is most common in young and middle-aged adult women. The ECG generally shows RVH. The presence of frequent VPBs may be a clue to intermittent sustained VT. Evidence of previous Q wave MI with syncope should suggest the possibility of sustained monomorphic VT. Syncope with QT(U) prolongation should suggest torsades de pointes, a potentially lethal ventricular arrhythmia (see Chapter 16 ). A severe bradycardia (usually from heart block or sick sinus syndrome) in a patient with syncope constitutes the Stokes-Adams syndrome (see Chapter 17 ). In some cases, serious arrhythmias can be detected only when long-term monitoring is performed. Syncope in a patient with ECG evidence of bifascicular block (e.g., RBBB with left anterior fascicular block) should prompt a search for intermittent second- or third-degree heart block or other arrhythmias. Syncope in patients taking dofetilide, quinidine, sotalol, and related drugs may be associated with torsades de pointes or other arrhythmias.

Selected patients with unexplained syncope may benefit from invasive electrophysiologic testing. During these studies, the placement of intracardiac electrodes permits more direct and controlled assessment of sinoatrial (SA) node function, atrioventricular (AV) conduction, and the susceptibility to sustained ventricular or supraventricular tachycardias.


An ECG should be obtained in all comatose patients. If coma is from MI with subsequent cardiac arrest (anoxic encephalopathy), diagnostic ECG changes related to the infarct are usually seen. Subarachnoid hemorrhage or certain other types of central nervous system pathology may cause very deep T wave inversions (see Fig. 9-11 ), simulating the changes of MI. When coma is associated with hypercalcemia, the QT interval is often short. Myxedema coma generally presents with ECG evidence of sinus bradycardia and low voltage. Widening of the QRS complex in a comatose patient should also always raise the possibility of drug overdose (tricyclic antidepressant or phenothiazine) or hyperkalemia. The triad of a wide QRS, a prolonged QT interval, and sinus tachycardia is particularly suggestive of tricyclic antidepressant overdose (see Fig. 10-5 A).


An ECG should be obtained promptly in patients with severe hypotension because MI is a major cause of shock (cardiogenic shock). In other cases, hypotension may be caused or worsened by a bradyarrhythmia or tachyarrhythmia. Finally, some patients with shock from noncardiac causes (e.g., hypovolemia or diabetic ketoacidosis) may have myocardial ischemia and sometimes MI as aconsequence of their initial problem.


An ECG may be helpful in evaluating patients with unexplained weakness. Elderly or diabetic patients, in particular, may have relatively “silent” MIs with minimal or atypical symptoms, such as the onset of fatigue or general weakness. Distinctive ECG changes may also occur with certain pharmacologic and metabolic factors (e.g., hypokalemia or hypocalcemia) that cause weakness ( Chapter 10 ).



You can minimize errors in interpreting ECGs by taking care to analyze all the points listed in the first section of Chapter 22 . Many mistakes result from the failure to be systematic. Other mistakes result from confusing ECG patterns that are “look-alikes.” Important reminders are provided in Box 23-2 . Some common pitfalls in ECG interpretation are discussed in the following sections.

BOX 23-2 

Some Important Reminders

Check standardization.

Exclude limb lead reversal. (For example, a negative P wave with a negative QRS complex in lead I suggests a left/right arm electrode switch.)

Look for hidden P waves, which may indicate atrioventricular (AV) block, blocked atrial premature beats, or atrial tachycardia with block.

With a regular narrow-complex tachycardia at about 150 beats/min at rest, consider atrial flutter with 2:1 block versus paroxysmal supraventricular tachycardia or (less likely) sinus tachycardia.

With group beating (clusters of QRS complexes), consider Mobitz type I (Wenckebach) block or blocked atrial premature beats.

With wide QRS complexes and with short PR intervals, consider the Wolff-Parkinson-White preexcitation pattern.

With wide QRS complexes without P waves or with AV block, think of hyperkalemia.

Unless recognized and corrected, inadvertent reversal of limb lead electrodes can cause diagnostic confusion. For example, reversal of the left and right arm electrodes can cause an apparent rightward QRS axis shift as well as an abnormal P wave axis that simulates an ectopic atrial or junctional rhythm ( Fig. 23-2 ). As a general rule, when lead I shows a negative P wave and a negative QRS, reversal of the left and right arm electrodes should be suspected.

FIGURE 23-2  Whenever the QRS axis is unusual, limb lead reversal may be the problem. Most commonly, the left and right arm electrodes become switched so that lead I shows a negative P wave and a negative QRS complex.

Voltage can appear abnormal if standardization is not checked. Many ECGs are mistakenly thought to show “high” or “low” voltage when the voltage is actually normal but the standardization marker is set at half standardization or two-times normal sensitivity.

Atrial flutter with 2:1 block is one of the most commonly missed diagnoses. The rhythm is often misdiagnosed as sinus tachycardia (mistaking part of a flutter wave for a true P wave) or PSVT. When you see a narrow-complex tachycardia with a ventricular rate of about 150 beats/min, you should always consider atrial flutter (see Fig. 15-3 ).

Coarse atrial fibrillation and atrial flutter are sometimes confused. When the fibrillatory (f) waves are prominent (coarse), the rhythm is commonly mistaken for atrial flutter. With AF, however, the ventricular rate is erratic, and the atrial waves are not exactly consistent from one segment to the next. With pure atrial flutter, the atrial waves are identical from one moment to the next, even when the ventricular response is variable ( Fig. 23-3 ).

FIGURE 23-3  Atrial flutter with variable block (A) and coarse atrial fibrillation (B) are often confused. Notice that with atrial fibrillation the ventricular rate is completely erratic and the atrial waves are not identical from segment to segment, as they are with atrial flutter.

The Wolff-Parkinson-White (WPW) pattern is frequently mistaken for bundle branch block, hypertrophy, or infarction because the preexcitation results in a wide QRS complex and may cause increased QRS voltage, T wave inversions, and pseudoinfarction Q waves (see Fig. 12-3 ).

Isorhythmic AV dissociation and complete heart block can be confused. With isorhythmic AV dissociation, the SA and AV node pacemakers become “desynchronized,” and the QRS rate is the same as or slightly faster than the P wave rate (see Chapter 17 ). With complete heart block, the atria and ventricles also beat independently, but the ventricular rate is much slower than the atrial (sinus) rate. Isorhythmic AV dissociation is usually a minor arrhythmia, although it may reflect conduction disease or drug toxicity (e.g., digitalis, diltiazem, verapamil, and beta blockers). Complete heart block is always a major arrhythmia and generally requires pacemaker therapy.

Normal-variant and pathologic Q waves require special attention. Remember that Q waves may be a normal variant as part of QS waves in leads aVR, aVL, aVF, III, V1, and occasionally V2 (see Chapter 8 ). Small q waves (as part of qR waves) may occur in leads I, II, III, aVL, and aVFas well as in the left chest leads (V4 to V6). These “septal” Q waves are less than 0.04 second in duration. On the other hand, small pathologic Q waves may be overlooked because they are not always very deep. In some cases, it may not be possible to state definitively whether or not a Q wave is pathologic.

Mobitz type I (Wenckebach) AV block is a commonly missed diagnosis. “Group beating” is an important clue to the diagnosis of this finding (see Chapter 17 ). The QRS complexes become grouped in clusters because of the intermittent failure of AV nodal conduction.

Hidden P waves may lead to mistakes in the diagnosis of a number of arrhythmias, including blocked APBs, atrial tachycardia with block, and second- or third-degree (complete) AV block. Therefore you must search the ST segment and T wave for buried P waves (see Fig. 18-3 ).

Multifocal atrial tachycardia (MAT) and AF are often confused because the ventricular response in both is usually rapid and irregular. With MAT, you need to look for multiple different P waves. With AF, you must be careful not to mistake the sometimes “coarse” f waves for actual P waves.

LBBB may be mistaken for infarction because it is associated with poor R wave progression and, often, ST segment elevation in the right chest leads.

U waves are sometimes overlooked. Small U waves are a physiologic finding, but large U waves (which may only be apparent in the chest leads) are sometimes an important marker of hypokalemia or drug toxicity (e.g., sotalol). Large U waves may be associated with increased risk of torsades de pointes (see Fig. 16-17 ).

Severe hyperkalemia must be considered immediately in any patient with an unexplained wide QRS complex, particularly if P waves are not apparent. Delay in making this diagnosis can be fatal because severe hyperkalemia may lead to asystole and cardiac arrest while the clinician is waiting for the laboratory report (see Figs. 10-5 and 10-6 ).



The normal ECG patterns seen in children differ considerably from those in adults. The topic of pediatric electrocardiography falls outside the scope of this book (see Bibliography ), but a few critical points of difference between pediatric and adult ECGs are mentioned briefly. The normal ECG of a neonate resembles the pattern seen in RVH, with tall R waves in the right chest leads and right axis deviation. The QRS complex, however, is very narrow at about 0.04 sec ( Fig. 23-4 ).

FIGURE 23-4  The ECG of this healthy neonate shows a pattern resembling that of right ventricular hypertrophy, with tall right precordial R waves and right axis deviation. This pattern reflects the physiologic predominance of the right ventricle during fetal development. Notice also the sinus tachycardia and relatively narrow QRS complexes.

During the first decade of life, the T waves in the right to middle chest leads are normally inverted. This pattern, which sometimes persists into adolescence and adulthood, is called the juvenile T wave variant. Children and young adults may also have high voltage in the chest leads as a normal variant (see Chapter 6 ).



Like most clinical tests, the ECG yields both false-positive and false-negative diagnoses. For example, not all Q waves indicate myocardial infarction (MI) and not all patients with actual MI show diagnostic ECG changes. A normal or nondiagnostic ECG also does not exclude left ventricular hypertrophy, right ventricular hypertrophy, intermittent life-threatening bradyarrhythmias or tachyarrhythmias, pulmonary embolism, or pericarditis.

Despite limitations in sensitivity and specificity, the ECG can provide valuable information in a wide range of clinical situations, including the evaluation of general major medical problems such as syncope, coma, weakness, and shock.





Which arrhythmias can lead to syncope (fainting)? (Questions 2 and 3):



A normal exercise tolerance test excludes significant coronary disease.



The more false-positive results associated with a particular test, the less sensitive the test is.