Preventing sudden cardiac death in the young: Is electrocardiogram screening the most effective means?

The sudden and unexpected death of a young person is an infrequent event, although it has become an important public health issue. Factors contributing to the increasing profile are the long-standing debates over pre-participation screening for young athletes, the apparent success of this screening in some jurisdictions, as well as concerns about the possible risk of stimulant medication for attention-deficit hyperactivity disorder. In fact, this latter consideration resulted in a recent document (1) from the American Heart Association that devoted a large portion to discussion of the issues related to the usefulness of electrocardiogram (ECG) screening.

DEFINITIONS

Screening refers to a process whereby individuals are invited to undergo tests to separate a target population into those with higher or lower probabilities of disease (2). Early detection of the target diseases should be possible by reasonable clinical methods before the usual time of symptom development. The goal of a screening program is improvement in the clinical outcome (such as survival and quality of life).

Sudden cardiac death (SCD) is defined as natural death due to cardiac causes within 1 h of the onset of symptoms (3). If the event is unwitnessed, the definition is often extended to 24 h. What is being referred to in the present article is a sudden, unexpected, nonviolent, natural death of an individual without a previously known underlying cardiac diagnosis, in whom a thorough autopsy reveals an underlying cardiac condition as the likely cause of death, or no abnormalities. The latter is required because of the increasing realization that many SCDs may be due to arrhythmias originating at the cellular or ionic channel level.

A framework to understand the issues can be constructed by examining the epidemiology of and conditions associated with unexpected SCD in the young, and the ability of the ECG to identify these conditions. ECG screening in newborns and sudden infant death syndrome prevention are separate issues and will not be discussed.

EPIDEMIOLOGY

Although there is a wide range of estimates of the incidence of SCD in the young population – ranging from 0.8 per 100,000 to 6.2 per 100,000 – the majority of deaths in the young are both not sudden and noncardiac (1,4,5). Population-based studies (6–8) show that SCDs usually account for less than 20% of the total, with only approximately 10% of these being due to previously undiagnosed cardiac disease. This could be as high as 40% if those with negative autopsies are included (presumed arrhythmia). The prevalence of a negative autopsy leading to no diagnosis is also quite variable and is less common when investigations on extended family members are performed after a negative autopsy on the index case (9). Cardiac diagnoses are made at autopsy less frequently in the one- to 14-year-old age group compared with the 15- to 19-year-old group; however, there are more unexplained (ie, negative autopsy) cases in the younger group (10).

Of equal importance is that despite the emphasis on sports-related deaths from many groups (11,12), the majority (90% to 95%) of SCDs in young people are not related to participation in competitive sports (13). Although the rate of SCD in nonathletes is low, the number of deaths within the population is high (14). SCD is considerably more frequent in males than females in both sporting- and nonsporting-related deaths (13,14).

CONDITIONS

A list of the most common diagnoses found at autopsy is provided in Table 1. The prevalence of the diagnoses differs among regions. For example, hypertrophic cardiomyopathy (HCM) causes approximately 30% of SCDs in young athletes in the United States (15), whereas arrhythmogenic right ventricular cardiomyopathy (ARVC) is the most frequent cause (22%) in the Veneto region of Italy (16). The distribution of cardiac diagnoses is different between the athletic and nonathletic populations. In one study (13) from the United States, HCM was responsible for only 3% of SCDs in nonathletes. In Italy, HCM was found in only 2% of athlete deaths, but 7.3% of nonathlete deaths (14). Coronary artery anomalies may cause between 10% and 20% of SCDs (5). Other diagnoses are less common and are responsible for less than 5% of all SCDs (5). Myocarditis is implicated in the majority of autopsy series, and is usually clinically unsuspected. Previously unsuspected genetic arrhythmias such as congenital long QT syndrome (LQTS), catecholaminergic polymorphic ventricular tachycardia (CPVT) and Brugada syndrome are causes that may account for as many as 40% of patients with a normal heart at autopsy (17–19). Although important, commotio cordis (nonpene-trating blunt trauma to the anterior thorax causing arrhythmia) is an acute event and will not be discussed further in the context of ECG screening (20).

TABLE 1.

Most common causes of sudden cardiac death in the young

Cardiomyopathies

Hypertrophic cardiomyopathy

Arrhythmogenic right ventricular cardiomyopathy

Dilated cardiomyopathy

Nonspecific left ventricular hypertrophy

Myocarditis

Congenital coronary artery abnormalities

Anomalous origin of the coronary artery from the incorrect sinus

Coronary artery disease

Tunnelled coronary artery

Arrhythmia syndromes

Wolff-Parkinson-White syndrome

Congenital long QT syndrome

Catecholaminergic polymorphic ventricular tachycardia

Brugada syndrome

Structural abnormalities

Previously unrecognized congenital heart disease

Aortic rupture

ECG IDENTIFICATION

Central to the question of the potential usefulness of the ECG as a screening tool for the prevention of SCD is whether the ECG can identify the conditions of interest. The ECG is abnormal in 95% of individuals with clinically diagnosed HCM (21); however, it is abnormal in as few as 38% of those who are identified with HCM by genetic testing (22). It is presumed, but not known, that an abnormal ECG is a prerequisite for identifying those with HCM whose first presentation is SCD, although the ECG is abnormal in the majority of individuals with HCM who subsequently experience SCD (21). The ECG abnormalities in children with clinically diagnosed HCM are non-specific, and further testing would be required for diagnosis (23). ARVC is classically characterized by ECG changes such as inverted T waves to the midprecordial leads, widened or fragmented QRS complexes, and abnormal complexes on the terminal portion of the S wave in the right chest leads (epsilon waves). However, these abnormalities are usually not present in their full form in the young and in those with early disease. Even in Italy, individuals who passed screening and subsequently died from ARVC were shown, in retrospect, to have nonspecifically abnormal ECGs that did not lead to an accurate diagnosis, despite further testing in some (14). Congenital coronary artery abnormalities are usually not detected by ECG, even if the individuals come to clinical assessment before SCD (24).

The ECG was once considered to be the diagnostic tool for congenital LQTS; however, it is now known that 10% to 30% of gene-positive individuals may have a corrected QT interval within the normal range. Follow-up studies implicate very prolonged corrected QT intervals (longer than 500 ms) as a risk factor for symptoms and/or SCD in males, but not females, and it is hoped that these abnormalities would be detected on a screening ECG. However, even among cardiologists, interpreting ECGs correctly in LQTS has been clearly shown to be problematic (25). CPVT is diagnosed clinically by the presence of ventricular arrhythmias under conditions of sympathetic stimulation (exercise and excitement). The resting ECG is usually normal and it would be fortuitous, but not expected, for a screening ECG to uncover CPVT. Brugada syndrome is an uncommon channel abnormality characterized by an ECG appearance of right bundle branch block and an ST segment elevation that may be intermittent and only evident during situations such as fever or concomitant drug therapy.

PREVIOUS ECG ABNORMALITIES IN INDIVIDUALS WITH SUBSEQUENT SCD

A limited amount of data that examines this question both in the sporting population and the general population are available. The Italian data showed that 25% of SCD victims had abnormal ECGs at preparticipation screening, but were allowed to compete after further evaluation (14). Most of the abnormalities undetected by further examination were cardiomyopathies and coronary artery anomalies. A population-based autopsy study (26) from Sweden that used previous ECGs obtained for medical or military service purposes showed ECG abnormalities in 82% of the individuals, in whom ECGs were available (41% of those with SCD). The ECGs were abnormal in 88% and 82%, respectively, of those subsequently shown to have ARVC or HCM at autopsy. Abnormal ECGs were also found in one-third of the 28% of individuals with a normal (negative) autopsy. A previous study (27) showed that an initially normal ECG may become abnormal before SCD in patients subsequently diagnosed by autopsy as having ARVC or HCM.

Retrospective studies (24,28) have also identified that many victims of SCD have had previous symptoms such as chest pain, palpitations or syncope, which potentially could have led to an early, correct diagnosis.

INCLUSION OF AN ECG IN SCREENING PROTOCOLS

Population-based ECG screening occurs in some international jurisdictions. In Italy, the screening is restricted to the preparticipation examination of athletes (12 years of age to 35 years of age) (14), whereas in Japan, it is mandated for all school-aged children (29). Both systems showed that the detection of significant disease is an uncommon event. Recent follow-up data (14) from Italy suggested that the SCD rate decreased significantly from 3.6 per 100,000 person-years to 0.43 per 100,000 person-years since the implementation of the program. However, it is not clear from the published report whether this decrease in the rate of SCDs applies equally to all age groups (for example, those younger than 18 years of age). It is worth noting that the rate of SCD finally determined in the Italian study is not dissimilar from the baseline rates seen in the United States. Although the rate of SCD in the nonathletic, nonscreened population has remained stable, the number of deaths has increased substantially. The current rate of SCDs in nonathletes is now more than double that of athletes (0.9 per 100,000 person-years versus 0.43 per 100,000 person-years) and the number of deaths is more than 10 times greater (14).

Studies (30) of smaller populations have confirmed that there is a low prevalence of important cardiovascular diseases that are identified, and the majority of studies have not been designed to test whether the SCD rate is reduced by screening with an ECG. In older cohorts, the resting ECG may not be as useful as an exercise ECG for identifying individuals who should be disqualified from participation in sports (31). Discussions of the cost-effectiveness of ECG screening are dependent on the jurisdiction considering the program. The potential cost is not considered prohibitive in Europe (11), while it is considered prohibitive in the United States (32). No appropriate analysis has been published that examines the cost implications in Canada.

INCLUSION OF ECHOCARDIOGRAM IN ROUTINE SCREENING

No large, population-based studies have examined the question of whether the echocardiogram provides better sensitivity and specificity in the general population to diagnose conditions that may cause SCD. In studies with a few thousand participants, HCM and Marfan syndrome are rarely, if ever, identified (33). In highly trained athletes, it was concluded that the ECG is useful to select those who require echocardiography to exclude HCM, but is not cost-effective on its own (34).

INTERNATIONAL RECOMMENDATIONS FOR PREPARTICIPATION SCREENING

The American Heart Association currently does not support the addition of the ECG to the preparticipation examination and cites both scientific and practical reasons (32). The final conclusion was ‘complete and targeted personal and family history and physical examination’; however, the addition of the echocardiogram and/or ECG was ‘optional’ (32). The European Society of Cardiology has crafted a consensus statement informed by the Italian experience that recommends a common European protocol based on the ECG (11). The initial screening includes family and personal history, physical examination and the ECG.

CONCLUSION AND CONSIDERATIONS

Although there are some important take-home messages (Table 2), there are insufficient data to either fully endorse or reject the widespread implementation of a separate screening program in Canada, with the express purpose of decreasing SCD in the young. It does seem clear from the available data that any considerations to decrease the numbers of young victims of SCD cannot be restricted to a narrow subset of the population (for example, athletes). The most appropriate time and age for testing, as well as the best protocol, remains to be determined. The most efficient and effective venue is also unclear. Rather than creating specific screening clinics, it may be a consideration to first use a ‘case-finding’ approach that uses the expertise of the primary care physicians who ideally have access to longitudinal clinical follow-up data, a thorough family history and a relationship with the individual, which would facilitate information gathering and sharing, and the opportunity for diagnosis (2,35).

TABLE 2.

Take-home messages for electrocardiogram screening for prevention of sudden cardiac death (SCD) in the young

Many of the conditions associated with SCD will be detected There will be a significant false-positive rate There will be an important false-negative rate Some, but not all, SCDs may be prevented