A recent article by Wilson and colleagues1 proposes the mandatory ECG screening of athletes as a means of reducing the risk of sudden cardiac death in this population. The idea of such screening originated with the studies of Dr Corrado and his colleagues in Italy.2 The concept has subsequently received support from several European groups, Dr Corrado being involved in some of these: the International Olympic Committee Medical Commission,3 the European Society of Cardiology4 and the Féderation Internationale de Football Association (FIFA).5

A proper starting point for a consideration of this issue is Bayes’ theorem. This theorem makes clear that mass screening cannot be effective and will inevitably lead to a high proportion of false positive results if one is using a fallible test procedure when attempting to detect a rare risk. Wilson and associates1 attempted to increase the frequency of their positive outcomes by assessing the frequency with which their ECGs detected certain “abnormalities” such as the Wolff–Parkinson–White (WPW) syndrome, thought to be associated with an increased risk of sudden death. One problem inherent in this approach is that most athletes have substantial left ventricular hypertrophy, and their ECGs (in essence the algebraic sum of 50 mV signals traversing both the left and right ventricles) are unlikely to conform to standards that in general have been developed on sedentary individuals. Perhaps for this reason, Wilson and associates1 diagnosed three times the anticipated number of WPW syndromes in their sample of athletes. Italian evaluations of ECG screening were based on a flawed sample of 785 ostensibly healthy athletes and 220 individuals who had been referred for cardiac evaluation; these studies also used an indirect criterion of success, namely a correspondence between reported ECG “abnormalities” and a combination of clinical judgment and echocardiographic criteria.6 However, the gold standard when assessing the success of athlete screening must be the ability of the physician to prevent sudden death. Unfortunately, we lack a good measure of risk severity. In Italy, the annual incidence of exercise-related deaths appears to be about 8.5 per 100 000 among athletes (who have undergone mandatory screening), as compared with 0.71 per 100 000 in the general population, most of whom have not been screened.7 In North America (where ECG screening of athletes is not required) the incidence of sudden death seems lower than that experienced by Italian competitors. But there can be little argument that the sudden death of an athlete is an extremely rare occurrence, both in Europe and in North America; in consequence, there is little likelihood that it will be detected by ECG screening. Italian investigators admitted that only 3% of their supposed “abnormal” records were true positive findings.7 Many false positive results could possibly be eliminated by further examinations such as echocardiography, but this would present an unwarranted cost to both the individual and the community. Many of those tested would face unnecessary restriction of their physical activity, iatrogenic disease and an impaired quality of life; in fact, 1.8% of Italian athletes were dissuaded from further sport participation,8 a figure grossly disproportionate to their risk of sudden death. Further, the community would need to provide substantial funds for additional testing.

Given that most North American authorities do not recommend the mass ECG screening of athletes,9 there seems an opportunity to compare the incidence of sudden exercise-related deaths in Italy with that in the western hemisphere. I am not aware of any statistics of this type suggesting that the risks are lower for Italian competitors. But even if ECG screening were to have some beneficial influence upon risk, cost-effectiveness would remain a major concern. Assuming an ECG cost of no more than $30 per patient, there would be an annual outlay of $3 million per 100 000 athletes. Moreover, in some 30% of those tested,9 10 costs would be augmented by a need for echocardiographic evaluation, at perhaps $300 per individual. This would bring the total annual budget to $13 million per 100 000 athletes. Critical issues in assessing the resulting benefits would be the annual risk – somewhere between 0.7 and 8.5 per 100 000 competitors – and the efficacy of ECG screening in reducing this risk. In order to reduce the cost per incident prevented to a level commonly accepted by health economists, it would be necessary to assume a high level of risk and an efficacy of at least 50%. If one could postulate that screening avoided four deaths per year in a population of 100 000 athletes (a saving of about 240 life-years), then, with a cost of $13 million per 100 000, the price would be $54 000 per life-year saved. Unfortunately, even this unlikely level of benefit would be more than offset by a decrease in the quality of life for those 1800 athletes for whom further sport participation was prohibited. If the quality of life in this group were reduced by an average of only 1%, then they would sustain a cumulative loss of 1080 quality-adjusted life years. Thus, the annual testing process would decrease the net prospects of the community by 740 quality-adjusted life years per 100 000 athletes, despite the large healthcare expenditure that had been incurred.

Plainly, there seems a need for more data and a critical analysis of findings before we can accept the wisdom of mass ECG screening for athletes.