Incidence of sudden death

Specific data on unexpected sudden death in children and teenagers are not easy to find. Corrado et al reported a rate of 1.0/105 in Italians up to the age of 35 years.1 The risk increased with age in this population so the proportion younger than 20 years was probably less than half. A report from an International Olympic Committee working group found that 40% of such deaths occurred under the age of 18 years.2 Maron reported a sudden death risk of 0.46/105 person years in high school athletes.3 Van Camp reported a rate of 0.4/105 in high school and college athletes, being 0.75/105 in males and 0.13/105 in females.^w2 A study of cardiac arrest in schools estimated an annual incidence of 0.18/105 person-years among students.^w3

It is reasonable to conclude that the overall average risk of sudden cardiac death in apparently normal children and teenagers is around 0.4/105 age specific person years or 1:250 000. This represents about 2% of all deaths in this age group: 50% of deaths at the same age are unnatural (that is, due to accident, trauma, homicide, suicide etc).4 A rate of sudden death of 0.4/105 equates to two or three deaths during secondary school years for every 100 000 children, or three or four deaths during teenage.

Cardiac causes of sudden death

The data available come mainly from populations studied up to the age of 35 years. In a registry of young athletes in the USA hypertrophic cardiomyopathy accounted for 34–44% of sudden deaths.5 Other leading causes included congenital coronary anomalies (17%), myocarditis (6%), and arrhythmogenic right ventricular cardiomyopathy (4%). Only 2% were ascribed to ion channelopathies and only 3% of victims had “normal hearts”. In a study of a similar population in Italy the leading causes of sudden death were premature coronary artery disease (22%), arrhythmogenic right ventricular cardiomyopathy (14%), myocarditis (12%), mitral valve prolapse (10%), hypertrophic cardiomyopathy (9%), and congenital coronary anomalies (8%).1 In this series only 0.3% of deaths were attributed to long QT syndrome and 7% were unexplained.

The number of sudden deaths with no explanation seems low in both these reports, given the fact that most primary arrhythmias leave no trace after death. Other experience suggests the true proportion may be as high as 30%. An Australian study of sudden cardiac death at age 5–35 years found 25% due to coronary artery disease, 12% to myocarditis, 6% to hypertrophic cardiomyopathy, 2% to coronary anomalies, and 29% were unexplained, presumed to be due to arrhythmia.6

In children and teenagers we might expect no deaths from premature coronary artery disease and very few from arrhythmogenic right ventricular cardiomyopathy. We would expect the main structural causes to be hypertrophic cardiomyopathy and anomalous coronary arteries and a higher proportion to be due to primary cardiac arrhythmias (leaving an apparently normal heart after death) (box 1).

Sudden death from cardiomyopathy

Hypertrophic cardiomyopathy

Hypertrophic cardiomyopathy is said to be common in adults with a prevalence of 1:500.^{w4 w5} The disease prevalence at a younger age is probably much lower. A Japanese study using ECG screening found 1:15 000.^w6 There are a few reports that enable us to estimate the risk of sudden death from previous undetected hypertrophic cardiomyopathy. A Finnish study of all deaths from undiagnosed cardiomyopathy found a rate of 0.09 per 105 person years at age 0–20 years.^w7 A report from northern England showed a rate of 0.075 per 105 person years age 1–20 years.4 An Anglo-Swedish report found a rate of 0.11 per 105 at age 8–16 years.^w8 In Veneto, Italy in 1979–1996, the risk of sudden death from hypertrophic cardiomyopathy at age 12–35 years was 0.05 per 105 person years.^w9 Taken together, these reports suggest that the risk of sudden death from hypertrophic cardiomyopathy in childhood and teenage is no higher than 0.1 per 105 or one in a million age specific person years.

Arrhythmogenic right ventricular cardiomyopathy

Arrhythmogenic right ventricular cardiomyopathy is mostly a disease of young adults. The prevalence is said to be 1:5000 but is probably much lower in children. Arrhythmogenic right ventricular cardiomyopathy caused 14% of sudden deaths under 35 years of age in Italy,^w10 10% in France,^w11 but only 4% in the USA.5 In a French report of sudden deaths from arrhythmogenic right ventricular cardiomyopathy the average age was 34 with about 15% of sudden deaths under the age of 21 years.^w11 This equates to an age specific risk of around 0.08 per 105 per year. In the same study only 3.5% of deaths occurred during physical exercise.

Sudden death from other cardiovascular abnormalities

Congenital coronary artery anomalies

Anomalous origin of the right or left coronary artery from the contralateral sinus of Valsalva usually only comes to light after sudden death.7 The prevalence of these anomalies, and therefore the risk, is difficult to ascertain. Coronary angiographic studies suggest that around 1:1000 patients have anomalous origin of the right coronary artery from the left sinus and around 1:2000 have origin of the left coronary artery from the right sinus, but such reports may exhibit ascertainment bias.^w12 A small study of children referred for echocardiography found 4 in 2400, or around 1.7:1000.8 The cumulative lifetime risk is difficult to calculate. Corrado et al reported 21 sudden deaths from congenital coronary anomalies in 34 million person years at age 12–35 years, or 0.05 per 105.9 If the prevalence is indeed around 1:1000, the individual risk is fairly low.

Marfan syndrome

Sudden death in Marfan syndrome is mainly due to aortic dissection.^w13 In reports of causes of sudden death from the USA,5 Italy1 and Australia6 aortic dissection caused 2%, 5% and 5% of sudden deaths at age <35 years, respectively. No data for deaths under 20 years are given, but they are probably much lower. It is not clear what proportion of these deaths from aortic dissection were due to Marfan syndrome. Prospective reports of series of patients with Marfan syndrome showed very few sudden deaths, but the numbers are small.^w13

Sudden death from primary arrhythmia

Most published reports of causes of sudden death contain surprisingly few “normal” hearts. It is now well recognised that unexplained sudden death is mainly due to primary cardiac arrhythmia. In the UK the recent National Service Framework chapter on arrhythmias and sudden cardiac death recommends that the coroner’s pathologist should preserve tissue for DNA analysis.10

Congenital long QT syndrome

Long QT syndrome is mostly a disease of children and young adults and is now recognised as the leading arrhythmic cause of sudden death in the young. In common with other primary arrhythmias it leaves no trace after death, so retrospective diagnosis is possible only if the disease is identified in other family members through screening or as a result of genetic analysis of autopsy tissue. Gene mutations causing long QT syndrome are thought to be present in around 1 in 5000 in the population.^w14 The number of cases in which long QT syndrome presents with sudden death is very difficult to assess because of the difficulty of retrospective diagnosis. In a study of unexplained sudden deaths, long QT mutations were identified in 20%.11 Another report on swimming related sudden deaths found 70% had long QT mutations.^w14

Brugada syndrome

Brugada syndrome is a rare disease, mostly causing sudden death at rest from ventricular fibrillation in young adults. The population prevalence of the disease is unknown. A Brugada-like ECG may be present in up to 100 per 105 adults.^w16 Brugada syndrome is rarely diagnosed in children, although the original series did include three children.^w17 A recent report of 30 cases of Brugada syndrome in children was compiled from 13 European centres, a measure of the rarity of the disease.^w18 We have no estimate of the disease frequency in children, but two Japanese studies of children with an average age of 9 years found a “Brugada-like” ECG in 5–10 per 105.^{w19 w20} Sudden death from Brugada syndrome in childhood is probably very rare.

Catecholaminergic polymorphic ventricular tachycardia

Catecholaminergic polymorphic ventricular tachycardia is a rare disease occurring mostly in children. The diagnosis in life is usually made after presentation with syncope and demonstration of characteristic bidirectional ventricular tachycardia on exercise or ambulatory ECG recording. The resting ECG is normal. The high mortality risk in early life suggests that most cases in a population will be new mutations. This is confirmed by observations that 30% of cases are familial^w21 and 70% of cases are due to new mutations.12 The number of cases that cause sudden death is not known. In a study of swimming related sudden death about 20% of cases had a ryanodine receptor RyR2 mutation on postmortem genetic analysis.^w15 Patients diagnosed in life mostly present with syncope before the age of 10 years,12 so one might suspect presentation with sudden death would be similarly early in life.

Wolff–Parkinson–White syndrome

Postmortem diagnosis of Wolff–Parkinson–White syndrome in the absence of a diagnosis in life is almost impossible. A population based study of subjects up to the age of 77 years, which just pre-dated the widespread introduction of radiofrequency ablation, found two deaths (both in adults) in 1338 patient years.^w22 Sudden death in asymptomatic Wolff–Parkinson–White syndrome is probably rare, suggesting it is an unlikely explanation for unexplained sudden death in children and teenagers.13

Sudden cardiac death from other causes

Commotio cordis

Commotio cordis describes sudden death from ventricular fibrillation triggered by blunt trauma to the chest.^w23 As such it mainly represents extreme misfortune as there is no primary cardiac abnormality. It is a sporadic event and not amenable to prediction from screening. Over one third of cases occur outside competitive sports activities.^w23

Myocarditis

Myocarditis accounts for 6–12% of sudden deaths in young athletes up to the age of 35 years. A North of England study found six sudden deaths in around eight million age specific person years at age 1–20 years, a risk of 0.074 per 105.4 In a Finnish study of myocarditis at age 1–24 years the incidence of sudden death was 0.16 per 105 age specific person years, but only about one third of cases met the Dallas criteria.^w24 Myocarditis is a sporadic event and is not amenable to screening.

Current recommendations for screening

Recommendations for screening for potentially fatal cardiac disease in the young have been published in the USA and Europe, concentrating mainly on pre-participation screening of young athletes. An American Heart Association Scientific Statement advises screening of children in middle and high school (grades 7–12, ages 12–18 years) and recommends an eight point health questionnaire and a four point clinical examination.5 The report recognises that implementation of screening is incomplete and does not recommend an ECG—a point that has produced debate.14 15 ^w25 A report from the European Society of Cardiology proposes screening from the age of 12 or 14 years to include a health questionnaire, clinical examination and a 12 lead ECG, and recommends that screening be repeated every 2 years.16 The International Olympic Committee endorsed these guidelines in the Lausanne Recommendations.2 The guidelines are followed in much of the Italian population but not elsewhere in Europe.^w26 In the UK there is no systematic pre-participation screening although some sports teams and clubs do undertake this.

Feasibility of screening in children and teenagers

Wilson and Junger set out the criteria for a screening programme on behalf of the World Health Organization in 1968. These were extended and modified in the UK by the National Screening Committee in 2003 to establish 22 criteria for appraising the viability, effectiveness, and appropriateness of a screening programme.17 The eight criteria most relevant to screening for latent heart disease in children and teenagers are shown in box 2.

The US and European published guidelines for screening are influenced by the reports of causes of sudden cardiac death between the age of 12 and 35 years where premature coronary artery disease, arrhythmogenic right ventricular cardiomyopathy, and hypertrophic cardiomyopathy predominate. In the pre-adult population ion channelopathies and congenital coronary anomalies are likely to be more important causes of death. Before deciding whether screening should be recommended it is important to consider the diagnoses being sought and how well they comply with the criteria for screening (tables 1 and 2).

Although the guidelines address screening of “young competitive athletes”, these are not clearly defined. In the UK all children take part in regular vigorous sports activities at school and are expected to perform to their maximum ability. To that extent all of them are young competitive athletes. There is no convincing evidence of an increased risk with an increased level of performance at this age so the logical step, if screening is recommended, is that it should be applied to the whole population. The US and European guidelines recommend screening from 12 or 14 years of age. We know that many of the target diseases cause sudden death before that age.

In an Anglo-Swedish report on children with hypertrophic cardiomyopathy diagnosed in life, the highest sudden death rate was at age 10–15 years.^w8 In a US report the age group with the highest mortality was 5–15 years.^w27 It is not certain whether the same would apply to asymptomatic undiagnosed cases, but these findings provide no support for starting screening at 12 or 14 years.

Congenital long QT syndrome is a disease of the young. International registry data show that >75% of patients with LQT1 (the most common type of long QT syndrome) have their first cardiac event (syncope, cardiac arrest or sudden death) before 15 years of age.^w28 It seems probable the same will apply to those suffering sudden death without diagnosis. Again there is no support here for screening from the age of 14 (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

An ECG from a 1-year-old girl with a patent ductus arteriosus. She was discharged from follow-up after ductus ligation and the abnormality of the ECG was not appreciated at the time. She died suddenly in the school playground at the age of 7 years. In retrospect the QT prolongation is obvious.

As mentioned above, most patients with catecholaminergic polymorphic ventricular tachycardia present before the age of 10 years. The age spectrum of sudden death from anomalous origin of the coronary arteries is not known but deaths are reported in children and there is no reason to suppose that the risk only appears above the age of 14 years.

The logic of repeating an ECG every 2 years in all apparently normal children is unclear. There is no evidence of benefit from this strategy and the yield from repeat examinations would be low, as diagnoses such as long QT syndrome and Wolff–Parkinson–White syndrome would presumably be made on the first examination.

If screening were to be implemented, would it be effective?

For latent heart disease to be detected by history, questionnaire and physical examination, it needs to have a high incidence of prodromal symptoms, or a strong likelihood of a positive family history, or abnormal physical signs. The evidence is that this type of screening performs poorly with many false positive and false negatives. In a study of more than 2700 children, Wilson et al concluded “family history and personal symptom questionnaires alone are inadequate to identify people with disease associated with sudden cardiac death”.18 If screening included an ECG, we would expect to detect diseases such as long QT syndrome and Wolff–Parkinson–White syndrome but to miss important diagnoses such as anomalous origin of a coronary artery and catecholaminergic polymorphic ventricular tachycardia. The performance of the ECG in early detection of hypertrophic cardiomyopathy in children is unknown. As mentioned above, the disease frequency in adults is reported to be 1: 500.^{w4 w5} An Italian screening programme identified the disease in 1:1500 of those screened under the age of 35 years, so it is possible that a significant number of cases were missed.^w9 Two screening programmes of children using ECG, with a combined total of >8300 subjects, found no cases of hypertrophic cardiomyopathy—either because the disease frequency is low or because the test is ineffective.18 ^w29 Italian experience also shows that arrhythmogenic right ventricular cardiomyopathy is often missed by ECG screening.1 Thus universal ECG screening in children would be expected to find long QT syndrome and Wolff–Parkinson–White syndrome, but to miss anomalous coronary arteries and perhaps many cases of cardiomyopathy.

In a population of 100 000 school children about 300 will have latent heart disease potentially linked to sudden death (fig 2). Of these, perhaps 30 might be “at risk” and the sudden death rate will be around 0.4 per year. If screening was performed and included an ECG, around 2.5–5% would be abnormal and would require further evaluation.^w30 The 2500–5000 abnormal ECGs would include around 150 true positives, but there would be 150 false negatives with a normal ECG. All those with an abnormal ECG would require expert evaluation by a paediatric cardiologist. Mass screening with history, examination and ECG might be possible for around €30 (£27, US$44) per screen. The European recommendations are for screening in alternate years so there would perhaps be three such evaluations per child during secondary school years. The basic screening for each cohort of 100 000 would cost €9 000 000 (£8 million, US$13 million), with additional follow on costs for the expert evaluation of 2500–5000 children.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Potential for screening for latent heart disease in schoolchildren.

Prospects for prevention of sudden death

The crucial question is if the diagnosis is made earlier, will the outcome be improved? There is no doubt that the prognosis for long QT syndrome is dramatically improved by β-blocker treatment. Catecholaminergic polymorphic ventricular tachycardia also responds well to treatment with β-blocker treatment, although many patients require implantable cardioverter-defibrillator (ICD) implantation in later childhood or teenage. Wolff–Parkinson–White syndrome is detectable on ECG, but the individual risk in the absence of symptoms is very low.13 “Risk stratification” is also unhelpful so early diagnosis may have little benefit.^w31 Diagnosis of an anomalous coronary artery by this sort of screening is very unlikely. Although it is a potentially lethal abnormality the individual risk is probably low (as discussed above) and pre-symptomatic diagnosis raises the dilemma about whether the risk of surgery is greater or less than the risk of sudden death. Myocarditis and commotio cordis, being sporadic events, are not amenable to screening, even though they are significant causes of sudden death.

There is more difficulty in assessing the possible benefit of presymptomatic diagnosis of hypertrophic cardiomyopathy. It is undoubtedly an important cause of sudden death but the population risk is very low, of the order of one per million age specific person years. The only intervention shown to improve survival is ICD implantation, which is only recommended in the presence of two or more risk factors.^w32 Patients detected by screening are at significantly lower risk than those presenting with symptoms.^{w5 w33} It is unclear whether restricting activity after diagnosis reduces risk. The restrictions imposed by published guidelines are severe and come as a great disappointment to most newly diagnosed children and their families.^{w34 w35} Restricting exercise and banning sporting activities certainly stops death during sports, but there is no good evidence that it reduces risk overall. Maron et al reported that only 16% of sudden deaths in hypertrophic cardiomyopathy occur during moderate-severe exertion.^w27

Although screening for latent heart disease is often promoted using hypertrophic cardiomyopathy as an example, the benefit is unproven. A recent report from the UK National Screening Committee concluded that “Population-based screening for hypertrophic cardiomyopathy should not be instituted outside the context of carefully designed research projects” and that “Pre-participation screening of athletes for hypertrophic cardiomyopathy and other causes of sudden cardiac death should not be instituted at present”.19 A systematic review published in 2003 concluded that “pre-participation screening for athletes does not satisfy the basic requirements for medical screening as described by the United States Preventive Services Task Force”.20 It is not clear whether publications since then have significantly altered the situation.

Conclusions

There is undeniably a potential benefit from early diagnosis of some primary arrhythmias such as long QT syndrome and catecholaminergic polymorphic ventricular tachycardia, although the latter is unlikely to be detected on screening. The benefit of presymptomatic diagnosis of other diseases such as hypertrophic cardiomyopathy, Wolff–Parkinson–White syndrome, or coronary artery anomalies is unproven. Current recommendations for screening, if implemented, would probably detect fewer than half of children potentially at risk, partly because they would start too late in life and also because causes of sudden cardiac death in children differ from those in adults. More specific information on the prevalence and spectrum of latent cardiac disease in children is needed before screening can be recommended.