220 JACC Vol 7. No I January 1986220-30

Sudden Death and the Competitive Athlete: Perspectives on Preparticipation Screening Studies

STEPHEN E. EPSTEIN, MD, FACC, BARRY J. MARON, MD, FACC

Bethesda, Maryland

Sudden death in healthy athletes is uncommon but, when it occurs, the primary mechanism is cardiovascular. The major cause of sudden death in the young athlete is hypertrophic cardiomyopathy or related conditions characterized by left ventricular hypertrophy, aortic rupture due to cystic medial necrosis and congenital coronary artery abnormalities. In the middle-aged or older athlete, coronary artery disease is the most sig•nificant cause of sudden death.

Noninvasive screening procedures are currently available that can detect most subjects at risk of sudden death. However, although some potentially lethal dis•eases can be excluded by a relatively simple screening program, other diseases require expensive procedures, such as echocardiography and exercise electrocardio•graphic stress testing. This means that the sensitivity of detecting diseases leading to sudden death increases in

Sudden death can occasionally strike the competitive or recreational athlete. Such an individual is often young; by definition he or she is in apparent excellent health. The catastrophe, therefore, is totally unexpected and all the more tragic and alarming.

As described by the preceding review of the causes of sudden death in young competitive athletes (1), the vast majority of these athletes had underlying cardiovascular dis•ease that could have been responsible for their demise. The most common cause of sudden death in this group was hypertrophic cardiomyopathy. Other causes were idiopathic hypertrophy of the left ventricle (which may represent a variant of hypertrophic cardiomyopathy), aortic rupture due to cystic medial necrosis, congenital coronary artery ab•normalities and atherosclerotic coronary artery disease (Ta•ble I). Other reported series of middle-aged and older ath•letes who died suddenly (2,3) suggest that coronary artery

~rom the Cardiology Branch. NatIOnal InstItutes of Health, National Heart. Lung. and Blood Institute. Bethesda, Maryland.

Manuscnpt received January 22. 1985; revised manuscript received April 19. 1985. accepted July 2. 1985.

Address for reprints: Stephen E. Epstein, MD. National Im,btutes of Health. Building 10. Room 7B-15. Bethesda. Maryland 20205

© 1986 by the Amencan College of Cardiology

proportion to the financial resources that can be applied to the screening program. Thus, when a screening pro•gram designed to identify all cardiac diseases that have the potential to cause sudden death is planned by a com•munity, school or nonprofessional athletic team, the costs will almost undoubtedly be considered prohibitive. The practicality of applying a community- or school-initiated screening program can be questioned because of the very low incidence of sudden unexpected death in young healthy individuals. It is therefore likely that comprehensive screening programs will be confined to individuals or organizations with adequate financial resources. Less expensive, limited screening can be undertaken by in•dividuals or groups to identify some subjects at risk of sudden death during athletic competition.

(J Am Coil CardioI1986;7:220-30)

disease is far and away the most common cause of sudden unexpected death in this age group.

The fact that most young and, probably, older athletes who die suddenly during or shortly after intense physical exertion die from covert cardiac disease leads to two vital questions: I) are there noninvasive tests that can effectively detect the most common causes of sudden cardiac death in athletes, and 2) if such tests are available, are they eco•nomical enough to utilize as screening procedures? The following discussion deals with these issues. It does not attempt, however, to consider the screening of other clin•ically important cardiovascular disorders that do not seem to be common causes of sudden death during athletic com•petition (such as systemic hypertension), nor does it consider potentially lethal cerebrovascular disorders (such as berry aneurysms).

Hypertrophic Cardiomyopathy Physical examination. The sine qua non of hyper•

trophic cardiomyopathy is a left ventricle that is nondilated and hypertrophied (4-6). Figure I displays the characteristic auscultatory findings and carotid pulse contour of a patient

0735-1097/86/$3 50

JACC Vol 7, No I January 1986220-30

Table 1. Cause, of Sudden Death in Young Athlete, (:535 years old)

Most common causes Hypertrophic cardiomyopathy Idiopathic left ventricular hypertrophy* Aortic rupture (due to "cystic medial necrosis") Coronary artery disease Congenital coronary abnormalities

Less common causes Valvular aortic stenosis Prolonged QT syndrome Primary cardiac electrophysiologic abnormalitie, Cardiac sarCOidosis Myocarditis

'May be a van ant of hypertrophIC (ardlOmyopathy

who has hypertrophic cardiomyopathy with assocJated ob•struction to left ventricular outflow. The obstruction is dy•namic; it is absent in very early systole and can be altered by interventions such as the Val salva maneuver or changes in position such as standing or squatting (7), The dynamic nature of the obstruction is responsible for the characteristic rapid upstroke of the carotid pulse (which excludes fixed valvular or subvalvular aortic stenosis) and for the cardiac murmur that changes in intensity with such interventions. These findings are highly specific in that, when they are

Figure 1. Phonocardiogram and carotid pulse tracmg obtamed from a patient with obstructive hypertrophic cardiomyopathy Note the characteristic ",pike and dome" configuration of the carotid pulse contour. ECG = electrocardIOgram: 4LlCS = fourth left intercostal space.

5, AI 5,

.08--t t--Az

I:PSTEIN AND MARON ~CREENIN(; STl1DII-.~ FOR COMPHITIVE ATHLETI:-~S

221

present, the diagnosIs of hypertrophic cardiomyopathy is unequivocally established. However, although a substantial number of these patients have such classical findings under rest conditions, the majority do not because they have either no or minimal ba,al obstruction to left ventricular outflow (8,9). Thus, the sensitivity of the physical examination in this regard is low.

The chest roentgenogram may be useful in detecting pa•tient~ with hypertrophic cardiomyopathy because cardiac enlargement is often present (10). As with the physical examination, however, the majority of patients have either no cardiomegaly or very subtle abnormalities that may re•main undetected in large screening programs.

Echocardiography. In contrast, echocardiography is relatively sensitive and specific in detecting hypertrophic cardIOmyopathy. The finding of a hypertrophied nondilated left ventricle in the absence of other diseases that might produce this abnormality (for example, systemic hyperten•sIOn or valvular aortic stenosis) strongly suggests the di•agnosis of hypertrophic cardiomyopathy (4,11). The diag•no,is is virtually established when marked systolic anterior motion of the mitral valve with mitral-septal contact is pres•ent (II), a finding reflecting obstruction to left ventricular outflow.

There are certain limitations, however, to the use of echocardiography in diagnosing hypertrophic cardio•myopathy ill athletes. First, the absence of systolic anterior motion of the mitral valve does not exclude the diagnosis because the large majority of patients with hypertrophic cardiomyopathy do not have systolic anterior motion. Sec•ond, the anterior basal portion of the ventricular septum is

Ii. ..- 1I\i*~l""'.'''':''''''.-------~''t!-, ',j,'_~, ••• "''''';I.It _______ ... , '- ',:,..,' ',' .~

II: ! ! ! I; 'I! ! :! i I i! ' II , I

ECG" , , : 1

I

222 EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

usually hypertrophied to a greater extent than is the posterior left ventricular free wall-hence, the finding of asymmetric septal hypertrophy on M-mode echocardiographic study (Fig. 2). However, other cardiac or systemic conditions may re•sult in an asymmetric pattern of left ventricular hypertrophy. Furthermore, in patients with hypertrophic cardiomyopathy, maximal wall thickening is not always located in the anterior septum, but may sometimes be identified in the posterior portion of the septum or anterolateral left ventricular free wall, or be confined to the apical portions of the ventricle. Occasionally such left ventricular hypertrophy will be pres•ent even when the anterior septum is of normal thickness. In such instances, the M-mode study results may be normal and hypertrophic cardiomyopathy will be detected only with two-dimensional echocardiography (5,6). Third, some nor•mal athletes have septal and posterior wall thicknesses that exceed the upper limits of normal (although these thick•nesses rarely exceed 15 mm) (12-17). In addition, echo•cardiographic studies of a quality sufficient to definitively establish ventricular septal borders cannot always be obtained.

Finally, there is a gray area of increased septal thickness that, although raising the suspicion of hypertrophic cardio•myopathy, is not abnormal enough to establish the diagnosis unequivocally. For example, an increased ventricular septal thickness of 13 to 14 mm in an athlete (> 12 mm being abnormal in most adults) might be a manifestation of either a normal "athletic heart" or a heart with hypertrophic car•diomyopathy. When such a borderline echocardiogram is

Figure 2. M-mode echocardiogram demonstrating the classic findings of nonobstructive hypertrophic cardiomyopathy. The ven•tricular septum (YS) is markedly thickened with respect to the posterior left ventricular wall (PW), which is of normal thickness. MY = mitral valve.

,,,

.. t. " Ii

~'.,

II

~.

JACC Vol 7. No I January 1986220--30

obtained, additional information is necessary. In particular, two-dimensional echocardiographic study may detect other areas of the left ventricle with more marked hypertrophy that cannot be imaged with conventional M-mode echo•cardiography (5,6). Echocardiographic studies in relatives can also be of value in this circumstance. Because hyper•trophic cardiomyopathy is often transmitted as an autosomal dominant trait (8,9), the likelihood that slightly increased ventricular wall thicknesses in an athlete represents hyper•trophic cardiomyopathy increases if its morphologic expres•sion can be identified in the athlete's parents or siblings.

Electrocardiography. Electrocardiographic studies in patients with hypertrophic cardiomyopathy have revealed that approximately 90% of such individuals (I8) and more than 90% of those who die suddenly (19,20) have an ab•normal electrocardiogram. The patterns are quite diverse, and none is entirely specific for the disease. However, a normal electrocardiogram in an athlete would indicate a low probability that hypertrophic cardiomyopathy is present and a low risk for sudden death if it is.

Screening studies. Table 2 summarizes the relative power of screening studies in detecting hypertrophic cardio•myopathy. Combined M-mode and two-dimensional echo•cardiography provide the most sensitive and predictive in•formation; however, the cost of these tests is high and, as discussed later, probably precludes their use in community•based screening programs. The electrocardiogram yields a considerable number of nonspecific "abnormalities" that are not predictive of cardiac disease; however, it is suffi•ciently sensitive in detecting hypertrophic cardiomyopathy to make it an important screening study. particularly when cost factors must be considered. It is unlikely that many additional patients with hypertrophic cardiomyopathy would be detected by the concomitant use of chest roentgenograms and cardiac auscultation.

Table 2 also lists the possible screening benefits to be derived from the patient's history. The value of including a history to detect occult, potentially fatal, cardiac disease in "healthy" asymptomatic athletes is perhaps not as great as might be expected. A history of occasional exertional dyspnea, light-headedness or nonspecific chest discomfort may be of little value in athletes, who often routinely drive themselves to the point of physical exhaustion. The only historical information that should unequivocally alert the physician to serious cardiac disease is a history of syncope or a family history of hypertrophic cardiomyopathy or of premature sudden death ( <40 years of age) in a first degree or other close relative. A history of premature sudden death in family members is helpful in alerting the physician that hypertrophic cardiomyopathy must be considered. How•ever, although hypertrophic cardiomyopathy may assume a particularly malignant clinical course in certain families (21), such families are uncommon and such a history will be obtained in very few athletes with unsuspected hypertrophic

JACC Vol. 7. No.1 January 1986'220-30

EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

223

Table 2. Major Causes of Sudden Death in Athletes and Potential Efficacy of Screening Studies

Causes

Hypertrophic cardiomyopathy

Cystic medial necrosis

CAD/congenital coronary abnormahties

Aortic valve stenosIs

Screening Procedures

History* Chest X-ray film AuscultatIOn Echocardiogram 12 lead ECG

History Chest X-ray film Auscultation Echocardiogram 12 lead ECG

History Chest X-ray film Auscultation 12 lead ECG Exercise ECG RadlOnuchde studies

History Chest X-ray film Auscultation 12 lead ECG Echocardiogram

Predictive Sensillvity Value

Poor Poor Fair Poor Fair Excellent Excellent Excellentll Goodt Poor

Poor Poor Fair Excellent Poor Fair Excellent Excellent

Poor Poor Poor Fair**

Poor Poor Fair-good Poort Excellent Poort

Poor Poor Excellent Good§ Fair Poor Poor Poor

* Specifically , a history of syncope or sudden premature death in a close family member tBoth in making the diagnosis of hypertrophic cardiomyopathy and in identifying pallents at risk of sudden death. tOf increasing value in subjects older than 40 years of age when the prevalence of coronary artery disease increases; however, most abnormal responses will still be falsely positive. §The predlcllve value of auscultation depends on intensity of murmur; the high prevalence of softer innocent murmurs will necessitate a more sophisticated and expensive evaluation m many individuals. IIPredlclive value IS cntically dependent on the magnitude of ventricular septal thickness (septal thickness >20 mm would be highly indlcallve of hypertrophic cardiomyopathy; septal thickness of 13 to 14 mm, as an isolated clinical finding, would provide only suggestive evidence of disease). **Only if coronary calcifications are detected by fluoroscopy; poor If fluoroscopy is not carried out. CAD = coronary artery disease; ECG = electrocardiogram. The rating system is based on the followmg sequence: excellent> good > fair > poor.

cardiomyopathy. A history of premature sudden death in family members is not specific for hypertrophic cardio•myopathy but may suggest other diagnoses to be evaluated. For example, because sudden death occurs in families with the rare prolonged QT syndrome (22-24), the possibility that this syndrome is present should also be explored.

Aortic Rupture Due to Cystic Medial Necrosis

Chest X-ray film. Figure 3 demonstrates the character•istic chest roentgenogram of a patient with a dilated aortic root due to cystic medial necrosis. The ascending aorta is prominent, raising the suspicion of aneurysmal dilation. Once this abnormality is recognized, additional studies must be undertaken, including M-mode (Fig. 4) and two-dimen•sional echocardiography and, possibly, aortic root contrast angiography. However, the difficulty with relying on the chest X-ray study to detect occult cystic medial necrosis is that a large portion of the proximal aorta lies within the cardiac silhouette, and it is not uncommon for there to be

marked dilation of the root without extension of the dilation into the ascending aorta. Under these conditions, the chest X-ray film would not demonstrate aortic root dilation (Fig. SA and B).

Physical examination. Although auscultation may be of value in the diagnosis of aortic dilation, its sensitivity is probably low. The dilated aortic root may produce an early ejection sound and, once the root is sufficiently dilated, aortic regurgitation occurs. An unexplained aortic regur•gitation murmur, particularly when heard loudest along the right sternal border, should suggest aortic root dilation (the murmur of rheumatic aortic regurgitation, or that due to a bicuspid aortic value, is almost invariably loudest along the left sternal border). These findings, when present, are use•ful. However, in many patients cystic medial necrosis of the aorta cannot be detected on the basis of the physical examination alone.

Echocardiography. On the other hand, M-mode and, particularly, two-dimensional echocardiography are highly sensitive in detecting an abnormally dilated aortic root (Fig. 4). Although echocardiography is the best noninvasive pro-

224 EPSTEIN AND MARON SCRI-J:NING STUDIES FOR COMPETITIVE ATHLETES

Figure 3. Typical chest X-ray film of a patIent wIth cystIc medial necrosis of the aorta. showing a prominently dilated root and ascending aorta and enlarged left ventricle.

cedure for detecting aortic root dilation (Table 2), its cost efficiency is a prominent consideration.

Coronary Artery Disease Electrocardiographic stress test. The electrocardio•

graphic exercise test is potentially the most important screening test for detecting occult coronary artery disease in asymp•tomatic athletes. However, its low predictive value in asymptomatic subjects (who have a low pretest likelihood of disease) emphasizes the limits of reliability of less than perfect diagnostic tests (25-27), especially when applied to subjects with a low probability of coronary disease (such as athletes < 30 years of age). This concept, in fact. has been shown to be directly relevant to young athletes. Thus. in a study (28) in which electrocardiographic exercise stress testing was conducted in a group of 75 young asymptomatic athletes with no clinical evidence of coronary artery disease. 9% had 1 mm or greater ST segment depression provoked by exercise. Although the prevalence of silent coronary artery disease in asymptomatic individuals 35 to 55 years of age is higher. it is still low enough that a high percent of false positive tests can be anticipated (27).

Screening studies. The dilemma in the use of exercise testing in large scale screening programs. even for middle-

lACC Vol 7, No I January 1%6 220-.10

aged or older asymptomatic athletes. is exemplified by the findings of the Seattle Heart Watch Study (29,30). Among more than 2.000 men initially free of heart disease. most of whom were 35 to 55 years of age, the annual incidence of sudden cardiac death was only about 0.05%. Maximal exercise testing and risk factor analysis (positive family history, smoking, systemic hypertension and hypercholes•terolemia) identified I % of the population as a group with an 18 times greater risk than that of the remaining subjects either of developing angina or myocardial infarction or of dying (29). If it is assumed that a subgroup at an 18 times greater risk of sudden death can, in fact, be identified by such means. it would require 10.000 asymptomatic subjects, with an annual mortality rate of 0.05%, to undergo screening studies to identify the 100 men (I % ) who comprise this high risk group. Of these 100 men. only 1 will be destined to die. Ironically. even if this massive screening were carried out and a high risk group identified. the much larger "low risk" group, containing 9.900 subjects. would contain 4 (80%) of the 5 subjects destined to die, and nOlle of these 4 individuals would have been identified by screening stud•ies (Fig. 6).

Radionuclide techniques. Although the sensitivity and specificity for detecting coronary artery disease are higher using radionuclide techniques during stress testing, these procedures still yield a large percent of false positive results when employed in an asymptomatic population (27). They are also expensive if one is considering their use in routine screening programs. Thus. there are major problems inher-

Figure 4. M-mode echocardiogram demonstrating a dilated aortic root (Ao) in a patIent with cystic medial necrosis of the aorta. The aortic root measures 50 mm; upper limit of normal is about 30 mm. LA = left atrium; RVOT = right ventricular outflow tract.

lACC Vol 7. No I l&nuary 1986:220-30

EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

225

p .... ,,1 ••• '111'10.3""·011:0 ••• _

ent in existing screening techniques for detecting underlying coronary artery disease in asymptomatic athletes.

Congenital Coronary Artery Abnormalities Congenital abnormalities of the coronary arteries may be

commonly implicated as the cause of sudden death in young athletes (1). For example, a left coronary artery that arises anomalously from the right coronary cusp makes, at its origin, an acute angle with the aorta as it courses leftward to assume its usual left ventricular distribution. A similar situation exists when the right coronary artery arises from the left coronary cusp. It is hypothesized that with exercise the aorta dilates, thereby exacerbating the acute angle and possibly causing total or near total occlusion. Another pos•sible congential coronary artery abnormality that may lead to sudden death in the young athlete is absence or under•development of a major coronary artery, presumably leading to inadequate myocardial blood flow to a portion of myo•cardium (1,2).

Screening studies. The dilemma posed by congenital coronary artery anomalies in the athlete is that no infor•mation is currently available related to the efficacy of screen•ing studies in identifying these abnormalities. Thus, the frequency with which abnormalities occur on the 12 lead electrocardiogram or electrocardiographic stress test in sub•jects with congenital coronary anomalies is unknown. How•ever, such anomalies should at least be suspected in a young athlete who experiences exertional chest pain or syncope. In this setting, a 12 lead electrocardiogram, a two-dimen•sional echocardiogram (to determine whether the left main coronary artery is in the correct position) and an exercise stress test would constitute a proper noninvasive evaluation. The decision as to whether more extensive studies are in•dicated (such as radionuclide stress testing or even coronary

Figure 5. Chest X-ray film (A) and aortic root angiogram (B) from a patient with cys•tic medial necrosis of the aorta. A, The dilated aortic root lies entirely within the cardiac silhouette. The diagnosis would re•main undetected using only this technique. B, The aortic root is massively dilated and the sinuses of Valsalva are obliterated.

angiography) must be individualized and would depend on the likelihood that the symptoms have a cardiac basis.

Valvular Aortic Stenosis In the study on sudden death in young athletes previously

cited (I), none of the 29 deaths were due to valvular aortic

Figure 6. Practical implications of employing exercise screening plus risk factors to identify asymptomatic subjects older than 35 years of age at risk of sudden death. The overall annual mortality rate is 0.05%. A subgroup of 100 individuals can be identified with an 18 times greater risk of dying (annual mortality rate 0.77%) than the remainder of the population (annual mortality rate 0.043%).

EXERCISE PLUS

RISK FACTOR SCREEN

NEGATIVE

226 EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

stenosis. Because patients with aortic stenosis may die sud•denly, one reasonable explanation for the absence of this disease in a relatively large series of athletes dying suddenly is that aortic stenosis is an easy diagnosis to establish with a very simple screening procedure-cardiac auscultation. In a healthy individual with a normal cardiac output, the absence of a murmur or the presence of only a soft murmur excludes the presence of significant aortic stenosis. Hence, this disease is uniquely amenable to a screening program.

Although virtually all patients with aortic stenosis will be identified by such screening, there are many healthy young people with systolic murmurs whose innocence or significance can only be determined by an experienced phy•sician. After screening, therefore, many such young athletes would have to be referred to sUbspecialists for more exten•sive, and more costly, evaluations.

Other Abnormalities Several other cardiac abnormalities are probably rela•

tively uncommon causes of sudden death in athletes. Never•theless, it is important to consider these diseases from the standpoint of screening.

Mitral valve prolapse. The propensity of mitral pro•lapse to cause sudden death has been of major concern, mainly because of the high frequency of this abnormality in the general population. Although no studies are available documenting the prevalence of mitrql valve prolapse in ath•letes, screening studies (31,32) of apparently healthy young women have revealed that approxi'l1ately 5% have auscul•tatory or echocardiographic evidence of mitral valve pro•lapse. In addition, serious ventricl!lar arrhythmias, as well as sudden death, have rarely been associated with mitral valve prolapse (33-35). These findings raise questions as to whether this entity should be searched for in athletes and whether its presence in an otherwise healthy person indicates that competitive sports be avoided.

Although definitive data are unavailable, there is little question that only the very rare patient with mitral valve prolapse dies as a result of this abnormality (35). Because mitral valve prolapse appears so frequently in otherwise healthy persons, its finding in the absence of any serious symptomatic ventricular arrhythmias should probably be considered benign and not influence the physician's rec•ommendations concerning participation in athletic activities (36). A more conservative approach would be justified in the occasional athlete with mitral valve prolapse and sig•nificant symptoms or symptomatic ventricular arrhythmias. Given existing evidence, it appears unnecessary to routinely screen for mitral valve prolapse in asymptomatic athletes.

Wolff-Parkinson-White syndrome. Although patients with the Wolff-Parkinson-White pattern on the electrocar•diogram have been known to die suddenly, it appears that

JACC Vol 7. No I January I YXh 220-JO

the incidence of sudden death in such subjects is low (37). Moreover, several studies (37-39) suggest that those pa•tients with the Wolff-Parkinson-White syndrome who are most susceptible to ventricular fibrillation and sudden death are those exhibiting atrial fibrillation with a rapid ventricular response. Thus, if the electrocardiographic pattern of Wolff•Parkinson-White syndrome is detected in the absence of clinically apparent arrhythmias, no evidence exists to justify proscribing participation in strenuous athletic activities (36). ~t also does not appear necessary to screen for this abnor•mality in the asymptomatic athlete.

Primary cardiac electrical abnormalities. Many pre•sumably healthy individuals, particularly those who are older than 40 years of age, have frequent premature ventricular complexes. In the absence of both underlying cardiac dis•ease and symptomatic consequences of the arrythmia, the association of such complexes (or even ventricular tachy•cardia) with sudden death is extremely rare (36,40-44).

Thus, although an argument could be made for obtaining a standard 12 lead electrocardiogram, it is not necessary to employ ambulatory electrocardiographic monitoring to screen asymptomatic subjects who want to participate in a stren•uous athletic program. This procedure is not only expensive, but the results would be clinically relevant only (with rare exceptions) if serious structural heart disease was also present.

Implications of Cost-Effectiveness on Developing a Strategy to Detect Potentially

Lethal Cardiovascular Disease in the Athlete The feasibility of screeping competitive or recreational

athletes must be considered on two levels: 1) community•based screening in which schools or other organizations adopt routine screening procedures for members of their group who have decided to participate in athletic actiyites, and 2) individual-initiated screening in which the prospec•tive athlete, on the advice of his or her physician, determines which screening studies should be undertaken. The com•munity-based approach is, with the exception of financially well endowed professional organizations, almost invariably burdened by financial constraints. This also applies to the individual-initiated approach, so that the comprehensiveness of the screening program will ultimately be closely tailored to the subject's financial resources.

On the basis of informal surveys aided by the American Heart Association (Nancy V. Wilson) and the American College of Cardiology (David J. Feild), cost estimates for the screening studies discussed in this review were approx•imated. Two types of cost structure were developed: one for large volume community-based screening programs, and a second for screening studies undertaken for the individual athlete in a private physician's office. Although most echo•cardiographic evaluations now include both M-mode and two-dimensional studies, the data we developed fOT echo-

JACC Vol 7. No I January 1986.220-30

cardiographic testing were limited to M-mode studies be•cause the additional expense of two-dimensional studies would be prohibitive in most screening situations. The pre•cise costs vary considerably among different communities and areas of the country and will obviously fluctuate with changes in the economy. Thus, the estimates employed in the subsequent discussion are intended to describe only in a very general way the financial impact of screening pro•grams for athletes.

Athletes 35 Years of Age or Younger

Role of screening prognosis. It would not appear eco•nomically feasible for a high schooL college or athletic club to initiate a screening program that would identify most of the cardiovascular problems responsible for sudden death in young athletes (Table 3). Thus, each organization must determine for itself the type of program it is capable of financing. For example, although cardiac auscultation can detect all patients with aortic stenosis at low cost, such a screening procedure, even if it includes a cardiac history, will fail to detect a large number of patients with hyper•trophic cardiomyopathy, and virtually all patients with cys•tic medial necrosis, coronary artery disease or congenital coronary anomalies. If funds are available to permit a chest X-ray to be obtained, many individuals with cystic medial

EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

227

necrosis will be identified as will additional patients with hypertrophic cardiomyopathy. This screening program, however, will fail to detect most patients with hypertrophic cardiomyopathy, many with cystic medial necrosis and all with coronary artery abnormalities. If a 12 lead electrocar•diogram is added to the profile, most patients with hyper•trophic cardiomyopathy at risk of sudden death will be de•tected. However, the electrocardiogram will not detect any additional patients with cystic medial necrosis and only the occasional patient with coronary artery disease or a con•genital coronary artery abnormality. Echocardiography and exercise electrocardiographic stress testing are prohibitively expensive for most athletic organizations. However, be•cause most subjects with hypertrophic cardiomyopathy at risk of sudden death will probably be detected by an ab•normal rest 12 lead electrocardiogram (19,20), and because coronary artery disease is so uncommon in high school- or college-aged subjects, it is difficult to justify the costs of adding echocardiography and exercise electrocardiography to a large scale screening program.

It is unfortunate, given the large number of young people participating in athletic programs in any given high school or college, that all but the least expensive screening program would be beyond the financial resources of most institutions or individual families. Under these circumstances, families with adequate financial resources might choose to pay for additional screening studies to detect the diseases that might

Table 3. Possible Community-InitIated Screening Strategies for Identifying the Athlete at RIsk of Sudden Death

Screening Battery Cost*

History, auscultation 25

HI,tory. auscultation, 25 chest X-ray film 25

Total 50 HistOry, auscultation, 25

chest X-ray film. 25 12 lead ECG 25

Total 75 Hi,tory. auscultahon. 25

chest X-ray film. 25 12 lead ECG, 25 echocardlOgraphy (M-mode) 100

Total 175 History. aw,cultation, 25

che't X-ray film, 25 12 lead ECG, 25 echocardiography (M-mode) 100 exercise ECG 75

Total 250

Will Detect

All AS. 25'7< of all HCM. some CMN

All AS. 30'70 of all HCM. many CMN

All AS. most HCM at risk

of SD

All AS. mmt HCM. most CMN

All AS. most HCM. most CMN, ~20% of CAD/CAA

at w,k of SD

WIll Miss

75% of HCM. all CAD/CAA. Most CMN

70% of HCM, all CAD/CAA, many CMN

Many CMN. vIrtually all CAD/CAA

Virtually all CAD/CAA

~80% of CAD/CAA at risk of SD

*Costs In U. S. dollars are based on estimates of a large scale screening effort. rather than on the prevailing rates In a private office or clInic. AS = aortic stenosIS; CAA = coronary artery anomalies; CAD = coronary artery disease; CMN = cystic medial necrosIs; ECG = electrocardiogram; HCM = hypertrophic cardio•myopathy; SD = sudden death.

228 EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

predispose to sudden death. Table 4 outlines the costs and results of such studies. Although the decision to undergo such screening can be made only by the prospective athlete and his family, it must be emphasized that sudden death in young competitive athletes is extremely rare, and the like•lihood that any single "normal" young athlete will be at risk of sudden death is extraordinarily low.

Clinical implications. To place things in perspective, we have made some conservative estimations of the practical implications of etnploying a preparticipation screenihg ex•amination to identify young athletes at risk of sudden death. We assumed that in young asymptomatic individuals wish•ing to participate in an athletic program: I) the prevalence of congenital heart disease is about 0.5% (45), 2) approx•imately I % of such asymptomatic subjects have a cardiac abnormality that could potentially cause sudden death, and 3) approximately 10% of individuals with underlying car•diac disease that could cause sudden death will, indeed, die suddenly as a result of athletic competition. On the basis of these rough estimates and the assumption that screening procedutes could successfully identify all subjects with heart disease while producing no false positive results, it can be calculated that to identify a group of 1,000 athletes who have cardiovascular disease, of whom perhaps only I athlete will die, 200,000 competitive athletes will have to be screened (Fig. 7).

Athletes Older Than 35 Years of Age

Role of screening programs. Although similar consid•erations apply to middle-aged individuals, the threat of sud•den death is necessarily of greater concern because of the considerably higher prevalence of coronary artery disease in such subjects. Although a workable screening program including tests to identify underlying coronary artery disease would, therefore, be desirable (Table 4), not only would

0.5%

1%

~ATHLETES

.-:,,: .. \ WITH DISEASE

.:.10.; CAPABLE ......... OF CAUSING

SUDDEN 10% DEATH

JACC Vol 7. No I January 1986.220-30

Figure 7. Practical implications of employing screening studies to identify asymptomatic subjects younget than 30 years of age at risk elf sudden death (see text for details). CHD = coronary heart disease.

the costs be high, but similar methodologic problems would exist as suggested for young athletes. For example, the prevalence of underlying cardiac disease posing a threat of sudden death in a group of 35 to 55 yeat old asymptomatic men may be approximately 3% (25,46). This figure might double or triple if additional risk factors (such as systemic hypertension, hypercholesterolemia or history of smoking) are also present. Given this relatively low prevalence of disease, there would be major problems in identifying by large scale screening studies those asymptomatic middle•aged individuals who are at risk of sudden death (Fig. 6).

One last point should be addressed. This relates to the individual who comes to the physician seeking advice about

Table 4. Possible Individual-Initiated Screening Strategies for Identifying the Athlete at Risk of Sudden Death

Cost*

Athlete ::; 35 years old History and auscultation 25 Chest X-ray film 30 12 lead ECG 35 Echocardiography (M-mode) 200

Total 290 Athlete >35 years old

History and auscultation 25 Chest X-ray film 30 12 lead ECG 35 Echocardiography (M-mode) 200 Exercise ECG 200

Total 490

Will Detect

All AS, !nost HCM, most CMN

All AS, most HCM, most CMN,

-20% of CAD at fisk of SD

Will MISS

Virtually all CAD/CAA (but prevalence is low in young athletes)

-80% of CAD at fisk of SD

*Costs in U.S. dollars are based on rough estimates of what such studies might cost in a pflvate office or clinic. Abbreviations as III Table 3.

JACC Vol 7, No I January 1986220-30

whether to undergo stress testing, Although such a request may reflect only the patient's natural caution or awareness of developments in medicine, many persons are motivated to see their physician because they have experienced some disturbing or puzzling symptoms, This critical information may not always be transmitted to the physician; indeed, the individual may not be able to formulate verbally the source of anxiety concerning his or her physical well-being, There•fore, we believe that when an individual asks whether screening studies should be performed before participating in athletics, it is probably wise to proceed with such studies,

Conclusions The considerations presented in this review emphasize

the dilemmas posed in attempting to establish an econom•ically feasible and relatively successful large scale screening program to identify life-threatening cardiac problems in asymptomatic subjects who wish to participate in an athletic program, Noninvasive screening studies are available that can identify most such problems, but very few persons are at risk and comprehensive community-based screening pro•grams for large groups of athletes are economically im•practical. More modest programs, however, commensurate with the financial resources available, can be undertaken that will identify at least some of the athletes at risk, It is also possible for an individual with adequate financial re•sources to undergo a comprehensive (and unavoidably ex•pensive) noninvasive screening evaluation that can dem•onstrate whether he or she has a very low risk of sudden death (and thus can participate safely) or has an abnormality that poses some risk to life, It must be understood, however. that a "low risk" designation refers to probability, and can never entirely rule out the possibility of sudden death, and that a "some risk" designation is also simply a probability statement, often indicating that additional testing is neces•sary to establish more definitively the diagnosis as well as its functional implications.

We thank Imogene Surrey for a"l',tance In the typing of the manu.,cnpt

References Maron Bl, Ep,teIn SE, Roberts WC. Causes of ,udden death in com•petItIve athletes. 1 Am Coli CardlOl 1986;7'204-14.

2 Waller BF, Roberts WC Sudden death whIle running In condItioned runners aged 40 year, or over. Am 1 Cardlol 1980;45·12'l2-300.

3. Thompson PO, Stem MP, Wilham, p, Duncan K, Ha,kell WL, Wood PD. Death dunng jogging or running. A .,tudy of 18 ca,e, lAMA 1979,242: 1265-7

4. Maron Bl, Ep,teIn SE. HypertrophiC cardIOmyopathy' a dl,cu"lon of nomenclature Am 1 CardlOl 1979;43'1242-4

loPSHJN AND MARON SCREENING STUDlloS fOR COMPETITIVE ATHLETES

229

5. Maron Bl, Gottdlener IS, Bonow RO, Epstein SE. Hypertrophic car•diomyopathy with unu,ual location, of left ventncular hypertrophy undetectable by M-mode echocardlography- Identification by wlde•angle, two-dlmen'lOnal echocardlOgraphy. Circulation I 'l81 ;63:409-18

6. Maron Bl, Gottdlener IS, Epstein SE. Patterns and ,Ignlficance of the distnbutlOn of left ventncular hypertrophy In hypertrophiC cardiO•myopathy a Wide-angie, two-dlmen,lonal echocardlographlC ,tudy of 125 patients. Am 1 Cardlol 1981;48-418-28

7 Epstein SE, Henry WL, Clark CE, et al A,ymmetric ,eptal hyper•trophy. Ann Intern Med 1974;81 :650-80

8 Clark CE, Henry WL, Ep,teIn SE Familial prevalence and genetIc transmission of IdiopathIC hypertrophIC ,ubaortic stenm,l, N Engl 1 Med 1973,289:709-14.

'l. Maron Bl, Nlcho" PF, Pickle LW, Mulvihill JJ. Patterns of Inhen•tance In hypertrophiC cardiomyopathy: assessment by M-mode and two-dlmen,lonal echocardlography Am 1 Cardlol 1984,53:1087-94.

10. Frank S, Braunwald E Idiopathic hypertrohlc ,ubaortic stenosIs. Chn•lCal analYSIS of 126 patients with empha'l, on the natural history Circulation 1<)68;37 759-88

II Maron BJ, Epstein SE HypertrophiC cardiomyopathy: recent obser•vatlO", regarding the ,peclfiClty of three hallmarks of the disease' asymmetnc septal hypertrophy, septal dl,organizatlon and systolic anterior motion of the antenor mitral leaflet. Am 1 Cardiol 1980,45.141-54

12. Morganroth J, Maron Bl, Henry WL, Epstein SE Comparative left ventricular dimenSions In trained athletes. Ann Intern Med 1975;82:521-4.

13 Roeske WR, O'Rourke RA, Klein A, Leopold G, Karllner IS Non•Inva,ive evaluation of ventncular hypertrophy in profe"ional athletes Circulation 1976;53:286-91.

14. Allen HD, Goldberg Sl, Sahn 01, Schy N, Wokclk R. A quantitative echocardiographlc ,tudy of champion chIldhood sWimmers. Circula•tion 1'l77;55:142-5

15. Gilbert CA, Nutter DO, FeIner 1M, Perkin, lV, Heymsfield SB, Schlant RC Echocardiographlc study of cardiac dimen'lOns and func•tion In the endurance-trained athlete. Am 1 Cardiol 1977;40:528-33.

16 Parker BM, Londeree BR, Cupp GV, Dubiel IP. The nonInva,ive cardiac evaluatIOn of long-dl,tance runner,. Chest 1978;73.367-81.

17 Ikahelmo Ml, Palat,1 IJ, Takkunen JT NoninvaSive evaluation of the athletIC heart. spnnter, ver,us endurance runners Am 1 Cardiol 1979;44'24-30.

18 Savage DO, Seide, SF, Clark CE, et al. Electrocardiographic finding, In patient, With obstructive and nonobstructlve hypertrophiC cardiO•myopathy. Circulation 1978;58402-8.

19 Maron Bl, Roberts WC, Edwards JE, McAllister HA, Foley DO, Ep,tein SE Sudden death In patients With hypertrophiC cardiO•myopathy charactenzatlon of 26 patients Without functIOnal limita•tion Am 1 Cardlol 1<)78;41'803- IO

20 Maron BJ, Robert, WC, Epstein SE. Sudden death In hypertrophiC cardiomyopathy profile of 78 patient, Circulation 1982;65: 1388-94.

21 Maron Bl, Llp,on LC, Robert, WC, Savage OS, Epstein SE "Ma•lIgnant" hypertrophIC cardiomyopathy: Identification of a ,ubgroup oj famille, with unu.,ually frequent premature death. Am 1 Cardiol 1978;41.1133-40

22 lames TN QT prolongation and '>Odden death. Mod Concepts Car•dlovasc 01, 1<)75;38.35-8

23 Schwartz PF, Pentl M, Malliam A The long Q-T Syndrome Am Heart 1 1975;89'378-90.

24 Vmcent GM, Abtldskov JA, Burge" Ml. Q-T interval syndromes. Prog Cardiova,c DI, 1974;16.523-31.

25 Rifkin DR, Hood WB Jr BayeSian analYSIS of electrocardIOgraphic exerCI,e stre" te,tlng N Engl J Med 1977;297:681-6.

26 Ep,tem SE Value and limitations of the electrocardiographic response

230 EPSTEIN AND MARON SCREENING STUDIES FOR COMPETITIVE ATHLETES

to exercise in the assessment of patients with coronary artery disease. Am 1 Cardlol 1978;42:667-74.

27. Epstein SE. Implications of probability analysis on the strategy em•ployed for the noninvasive detection of coronary artery disease: role of single or combined use of exercise electrocardiographic testing, radionucllde Cineangiography, and myocardial perfusion imaging. Am 1 Cardiol 1980;46:491-9.

28. Spinto p, Maron B1, Bonow RO, Epstein SE. Prevalence and sig•nificance of the ST segment response to exercise In an athletic pop•ulation. Am 1 Cardiol 1983;51:1663-6.

29. Bruce RA, DeRouen TA, Hossack KF. Value of maximal exercise tests In risk assessment of primary coronary heart disease events in healthy men. Five years' expenence of the Seattle Heart Watch study. Am 1 Cardiol 1980;46:371-8.

30. Bruce RA, DeRouen TA, Blake B. Maximal exercise predictors of coronary heart disease events among asymptomatic men in Seattle Heart Watch (abstr). Circulation 1977;56(suppl I1I):I1I-15.

31. Markiewicz W, Stoner 1. London E, Hunt SA, Popp RL. Mitral valve prolapse in one hundred presumably healthy young females. Circu•lation 1976;53:464-73.

32. Procacci PM, Savran SV, Schreiter SL, Bryson L. Prevalence of clinical mitral-valve prolapse in 1169 young women. N Engl 1 Med 1976;294: 1086-8.

33. Schwartz MH, Teichholz LE, Donosco E. Mitral valve prolapse-a review of associated arrhythmias. Am 1 Med 1977;62:377-89.

34. 1eresaty RM. Mitral valve prolapse-click syndrome. Prog Cardiovasc Dis 1973; 15:623-52.

35. 1eresaty RM. Sudden death In the mitral valve prolapse-click syn•drome Am 1 Cardiol 1976;37:317-8.

36. Bethesda Conference 16: Cardiovascular abnormalities in the athlete:

JACC Vol 7. No I January 1986220-30

recommendations regarding eligibility for competition. 1 Am Coli Cardiol 1985;6:1185-232.

37. Gallagher 11. Pntchett ELC, Sealy WC, Kosell 1, Wallace AG. The preexcitation syndromes. Prog Cardiovasc DIS 1978;20:285-327

38. Farre 1, Ross DL, Wiener I, Bar FW, Vanagt E, Wellens H11 Elec•trophYSlOlogical studies in patients with Wolff-Parkinson-White syn•drome. Herz 1979;4:38-46.

39. Klein G1, Bashore TM. Sellers TO, Pritchett ELC, Smith WM, Gal•lagher 11 Ventricular fibrillation In the Wolff-Parkinson-White syn•drome. N Engl 1 Med 1979;301:1080-5.

40. Chiang BN, Perlman LV, Ostrander LD 1r, Epstein FH. Relationship of premature systoles to coronary heart disease and sudden death in the Tecumseh Epidemiologic Study. Ann Intern Med 1969;70:1159-66.

41. Hinkle LE, Carver ST. Stevens M. The frequency of asymptomatic disturbances of cardiac rhythm and conduction in middle-aged men. Am 1 CardlOl 1969;24:629-50.

42. Pedersen DH, Zipes DP, Foster PR, Troup P1. Ventricular tachycardia and ventricular fibrillation in a young population. Circulation 1979;60:988-97.

43. Schamroth L Ventricular extrasystoles, ventricular tachycardia, and ventricular fibrillation: clinical-electrocardiographic considerations. Prog CardlOvasc Dis 1980;23:13-32.

44. Moss A1. Clinical significance of ventricular arrhythmias in patients with and without coronary artery disease. Prog Cardiovasc Dis 1980;23:33-52.

45. Nadas AS. Pediatric Cardiology, 3rd ed. Philadelphia: WB Saunders, 1972:294.679.

46. Diamond GA, Forrester 1S. Analysis of probability as an aid in the clinical diagnosis of coronary-artery disease. N Engl 1 Med 1979;300: 1350-8.