Author + information
- Received July 9, 2012
- Revision received August 29, 2012
- Accepted September 4, 2012
- Published online December 4, 2012.
- Amir Halkin, MD,
- Arie Steinvil, MD,
- Raphael Rosso, MD,
- Arnon Adler, MD,
- Uri Rozovski, MD and
- Sami Viskin, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Sami Viskin, Department of Cardiology, Tel-Aviv Medical-Center, Weizman 6 Street, Tel-Aviv 64239, Israel
Objectives This study sought to estimate the costs of a national electrocardiographic (ECG) screening of athletes in the United States and the number of lives that would be saved by that program.
Background A single study from Italy suggests that mandatory ECG screening of athletes reduces their risk of sudden cardiac death. Based on that study, ECG screening of athletes is endorsed by the European Society of Cardiology, though not by the American Heart Association. The widespread application of ECG screening remains controversial because the absolute reduction of sudden cardiac death risk provided, and its economic ramifications, have not been studied in detail.
Methods A cost-projection model was based on the Italian study, replicating its data in terms of athlete characteristics and physician performance. The size of the screening-eligible population was estimated from data provided by the National Collegiate Athletic Association and the National Federation of State High School Associations. The costs of diagnostic tests were obtained from Medicare reimbursement rates.
Results A 20-year program of ECG screening of young competitive athletes in the United States would cost between $51 and $69 billion and could be expected to save 4,813 lives. Accordingly, the cost per life saved is likely to range between $10.6 and $14.4 million.
Conclusions Our cost-projection model suggests that replicating the Italian strategy of ECG screening in the United States would result in enormous costs per life saved.
Prevention of sudden cardiac death (SCD) among athletes is a universal goal, though the optimal strategy for its achievement is controversial (1–4). Specifically, mandatory electrocardiographic (ECG) screening of all competitive athletes is recommended by the European Society of Cardiology (ESC) (5), but not by the American Heart Association (6).
The ESC recommendations (5) are based primarily on a single Italian study (7) that suggests that ECG-based screening of athletes is beneficial. That study by Corrado et al. (7) reported a marked decline in the incidence of SCD among athletes following the implementation of an Italian law mandating ECG screening, amounting to a 79% relative risk reduction. However, the absolute risk reduction achieved through mandatory ECG screening in Italy, the cost of such a program, and its economic ramifications, were never studied in detail nor were they addressed in the ESC guidelines. Knowing the cost of the screening process, and the actual cost per life saved, allows for prioritization of protective strategies at the national level (8). Therefore, we created the present model to: 1) estimate the number of athletes that would need to undergo screening if mandatory screening was to be enforced in the United States; 2) compute the cost of this strategy; and 3) determine the number of lives that would conceivably be saved.
We used a cost-projection model to estimate the national annual expenditure related to mandatory ECG screening of competitive young athletes in the United States as well as the cost of saving a single athlete's life. The model reproduces the Italian experience, as reported by Corrado et al. (7), in terms of patient characteristics and physicians' performance. That study (7) demonstrated a statistically significant association between the onset of mandatory ECG screening of athletes in Italy and a reduction in their annual incidence of sudden death. Importantly, the survival benefit attributed to mandatory ECG screening became gradually evident over the course of 20 years of repeated annual screening (Fig. 1A) (7). Accordingly, the period of mandatory screening used in our model was 20 years (Fig. 1B).
The cost-projection model was structured on assumptions derived from the Corrado data (7) and were based on U.S. data concerning athletic activity and healthcare costs. First, we estimated the number of young athletes that would require ECG screening over 2 decades. Then, we calculated the type, number, and costs of tests that would be performed in this population, including secondary testing driven by abnormal ECG findings. Lastly, by extrapolating the Italian data (7), we estimated the number of lives saved by ECG screening.
Number of athletes to be screened over 2 decades
In keeping with the inclusion criteria of the Italian study (7), young athletes in the United States were defined as registered participants in organized high school, college, and professional sporting activities. Data pertaining to participation in competitive high school and intercollegiate sports, and the annual rates of new enrollment in these activities over the last decade, were obtained from the National Collegiate Athletic Association Membership Report (9) and the National Federation of State High School Associations Participation Data Report (10). Based on these data (Online Table 1), the mean annual increment in the number of young athletes over the last decade was 1.53% per year for high school athletes and 1.81% per year for college athletes. Assuming that the annual growth in the number of young athletes remains constant, we estimated that by 2013 (the arbitrary onset of the 2-decade screening period), the number of registered competitive athletes would be 8,568,369 (Online Table 1). We ignored the professional athletes population because <2% of college athletes ultimately join professional sporting leagues (11). For simplicity, the athlete population requiring screening at initiation of the program was rounded to 8,500,000. Thereafter, the number of new athletes (1.53% to 1.81% per year) would compensate for the 2% per year disqualification rate from sport activity that, based on data from Italy, is expected to result from ECG screening. Accordingly, in our model, 8.5 million athletes undergo annual ECG screening over the next 2 decades, 170,000 athletes (2%) are disqualified from sport participation every year, and the number of athletes requiring screening during subsequent years remains constant. Overall, 2 decades of screening results in 170 million screening processes and 3.4 million (2%) disqualifications.
Number and type of tests to be performed
Diagnostic-testing rates in the model reflect the findings of Corrado et al. (7). In accordance with Italian law, from 1982 to 2004, all Italian athletes underwent medical screening consisting of: 1) medical history (focusing on syncope, palpitations, exertion-related symptoms, and family history of SCD); 2) physical examination; and 3) resting ECG. These tests were repeated annually. Abnormal findings during basic screening led to further diagnostic tests. The number and type of “secondary” tests was retrieved from a subanalysis of 42,386 athletes who underwent screening in Padua, Italy, between 1982 (the year when mandatory screening started) and 2004 (the end of data collection by Corrado et al. ). Of these, 91% had normal ECG, whereas 9% had ECG abnormalities that drove further testing. The last group consisted of 7% of all athletes who had normal secondary tests and 2% of athletes eventually disqualified because of abnormal secondary tests (7). The type and number of secondary tests performed appear in Online Table 2. Overall, 9.2%, 3.1%, and 1.2% underwent echocardiography, exercise testing, and Holter monitoring, respectively. Additionally, 0.2% of athletes underwent 1 or more of the following: cardiac magnetic resonance imaging (MRI); cardiac catheterization; and/or electrophysiologic studies (EPS). Because the partition of these 3 tests was not specified, we averaged their costs (MRI, catheterization, and EPS) and assumed that 0.2% of screened athletes would undergo a single test of that cost. This assumption is conservative considering, for example, that patients with suspected right ventricular dysplasia often undergo both cardiac MRI and EPS.
Cost of the tests performed
Individual test costs in the United States were obtained from Medicare reimbursement rates (12) to which we added the Outpatient Prospective Payment System institutional reimbursement rates and the minimum unadjusted copayment, providing minimal and maximal total costs (Online Table 3).
As was already explained, cost projections assumed the following. 1) Annually, 8.5 million athletes would undergo ECG screening. 2) Every year, 91% would screen negative, be allowed to compete, and would be rescheduled for screening the following year. 3) Two percent would be disqualified after undergoing additional tests. According to Italian data, the partition of additional tests (for these 2% of patients) would be as follows: echocardiography for all; exercise testing for 82%; Holter recordings for 41%; and MRI, catheterization, and/or EPS in 5% (Online Table 2). 4) Seven percent would undergo secondary testing and would be allowed to compete, though they would be rescheduled for intensified follow-up consisting of: echocardiography for all; exercise testing for 19%; Holter monitors for 5%; and MRI, catheterization, and/or EPS in 1% (Online Table 2). The frequency of secondary testing in the last subgroup was not stated in the original report by Corrado et al. (7) and that information was retrieved from a separate report (by the same investigators) dealing with athletes who have abnormal ECG but normal echocardiogram on initial screening (13). In this population, additional tests were performed at least twice during follow-up of almost 10 years (13). Accordingly, we assumed that for the 7% with abnormal ECG who were not disqualified, physical examination and ECG would be repeated annually, whereas auxiliary testing would be repeated 4 times during the 20 years of screening. This too was a conservative assumption because the present guidelines (14) recommend thorough reevaluation every 2 years once a cardiomyopathy is suspected.
Number of lives saved credited to screening
The perception that ECG screening saves lives is derived from the study by Corrado et al. (7) comparing the incidence rates of SCD among athletes before and after the implementation of mandatory ECG screening. In that study, the pre-screening period was 3 years (1979 to 1981), whereas the post-screening period was much longer (1982 to 2004) (Fig. 1A). During the pre-screening period, the annual death rate actually increased from 3.6 per 100,000 athlete-years in 1979 and 1980 to 4.0 per 100,000 athlete-years immediately prior to mandatory screening initiation. During the post-screening period, the annual death rate decreased gradually, reaching a nadir of 0.43 per 100,000 athlete-years in 2001 to 2004 after 20 years of screening (Fig. 1A) (7).
As was already explained, without screening, the number of athletes (starting at 8.5 million) would increase by 2% per year. In this unscreened athlete population, the SCD rate would remain constant, at 4.0 per 100,000 athlete-years, which is the highest mortality value observed in the pre-screening period in Italy (red line in Fig. 1B). We then assumed that with screening and a resultant 2% per year disqualification rate, the number of athletes in consecutive years would remain constant (at 8.5 million) and the SCD rate would decrease over the next 20 years from 4 to 0.43 per 100,000 athlete-years. Furthermore, we assumed that the decrement in mortality rate over time would assume a linear function during the 20-year screening period (blue line in Fig. 1B). The difference between the annual mortality rate with and without screening could thus be calculated for each of the individual years of the 2-decade screening period (bidirectional arrows in Fig. 1B) and the total number of lives saved was the sum of the all these values. The cost per life saved was the ratio between total screening costs and the numbers of lives saved.
ECG screening of all young competitive athletes in the United States. How much will it cost?
A 20-year screening program including all registered high school and college athletes, beginning in 2013, would include 8.5 million athletes. During the course of 2 decades of screening, the number of athletes undergoing yearly screening would remain at 8.5 million per year because the number of new athletes would compensate for the 2% per year disqualification rate. Thus, 20 years of annual screening would culminate in 170 million screening processes. The number of tests (including annual ECG for all athletes and periodic secondary tests for a finite proportion of athletes) is shown in Table 1. Based on the number of tests (Table 1) and their cost (Table 2,Online Table 3), we estimate that a 2-decade ECG-screening program involving all registered competitive high school and college athletes in the United States would cost between $51 and $69 billion (approximately $2.5 to $3.4 billion per year).
ECG screening of all young competitive athletes in the United States. How many lives will be saved?
In our model, the population of screened and unscreened athletes initially consists of 8.5 million athletes in each group. In the unscreened group, the athlete population grows by 2% every year and the annual mortality rate remains constant at 4/per 100,000 athletes. Consequently, the number of fatalities gradually increases from 340 during the first year to 505 on the 20th year of follow-up (Table 3,Fig. 1B). In the screened group, the number of athletes screened yearly remains constant at 8.5 million (because the natural growth in athlete population offsets the 2% per year disqualification rate) and the mortality gradually decreases from 4 per 100,000 on day 0 to 0.43 per 100,000 athletes on the 20th year of screening. This represents a mean decrement in mortality of 0.1785 per 100,000 athlete-years. Accordingly, the number of fatalities decreases from 340 during the first year to 37 during the 20th year of screening (Table 3, Fig. 1B). The number of lives saved accredited to screening increases from 22 per year after 1 year of screening to 469 per year on the 20th year of screening. The total number of lives saved over 20 years of screening is 4,813 lives (Table 3). Based on the “minimum” and “maximum” prices for the total 20-year screening (Table 2) and the total number of lives saved (Table 3), the cost per life saved ranges from $10.6 to $14.4 million.
We estimate that a 20-year program of ECG screening of young competitive athletes in the United States would cost between $51 and $69 billion and would save 4,813 lives. Accordingly, the costs-per-lives saved would be >$10 million. This cost is significantly higher than previously reported (15). That study was also modeled on the Corrado data (7), but it was based on the assumption that a 1-time screening process would result in a mortality reduction similar to that observed in Italy with 20 years of annual screening, thereby reducing the true cost of screening and artificially enhancing cost-effectiveness (15). Importantly, the ECG manifestations of the main causes of exercise-related SCD (hypertrophic cardiomyopathy and right ventricular dysplasia) always develop gradually and may therefore become evident only during repeated screening.
Cost-effectiveness analyses should ideally be based on randomized controlled trials, or at the very least, on observational studies for which consensus exist. Our cost calculations were based on a single retrospective study from Italy (7), which has triggered considerable debate (3,16,17). In particular, the mortality rate reported in Italy for the pre-screening period (4 per 100,000 athletes) (7) has been criticized as excessively high in comparison with other studies (3,16–18). In fact, the low mortality rate credited to 20 years of continuous screening in Italy (0.4 per 100,000 athletes) does not differ substantially from that reported in other countries were mandatory ECG screening is not routinely performed (16,18). Conceding that the “true” mortality rate without ECG screening is not as high as was reported in Italy by Corrado et al. (7) would diminish the absolute mortality reduction credited to ECG screening and would skyrocket cost-per-life-saved estimations. Certainly, many will argue that before cost-effectiveness analyses are performed, the “effectiveness” of ECG screening for saving lives should be more clearly demonstrated. Nonetheless, the Italian study (7) used in our calculations is the cause-de-exist of the ESC paper endorsing mandatory ECG screening of athletes (5), and our study is intended for those who believe the findings reported in the Italian study are correct. In other words, those who claim (based on Italian data) that ECG screening saves lives should take a good look at our analysis to realize how much that strategy costs.
Our expenditure estimation (>$10 million per life saved), high as it may be, probably underestimates what mass ECG screening of athletes would cost in the United States. First, expensive tests are likely to be used more liberally in the United States. For example, in a prospective study at the University of Virginia (19), 15% of all athletes were referred for echocardiographic examination and 2% underwent cardiac MRI studies (vs. 9% and 0.2%, respectively, in Italy ). Mass screening in nonuniversity facilities is likely to result in even higher rates of additional testing. Second, African American athletes have a higher prevalence of ECG abnormalities requiring additional tests (20), and these athletes represent a high proportion of athletes in the United States, but not in Italy.
The price of ECG screening goes beyond monetary expenses. A 2/thousand suicide rate, reported for disqualified athletes in Italy (21), is higher by several orders of magnitude than the SCD rate among unscreened athletes (7). Although these suicide events may not necessarily be due to disqualification, they must serve as a reminder of the profound emotional and financial implications expulsion might have, particularly when resulting from screening that was enforced rather than solicited. The long-term risk for asymptomatic disease carriers, identified solely through mass-screening processes, may not necessarily replicate the natural history of patients reported in clinical studies; thus, physicians advocating screening should acknowledge the potential for overtreatment.
First, data presented reflect “price” rather than “cost” (22). For example, the $700 price tag for an echocardiogram is higher than the actual cost of the test. However, Medicare reimbursement rates are the best approximation of the anticipated expenditure that would result from widespread ECG screening. Furthermore, Medicare prices have been used for cost-effectiveness calculations in numerous studies (23). Second, costs could be significantly reduced by reducing the frequency of screening or by using a more restrictive definition of “abnormal ECG,” which, in turn, would reduce the number of expensive auxiliary tests, as recently endorsed by the ESC (24). However, it is difficult to predict the impact that such changes would have on mortality. Third, we calculated “cost-per-lives-saved” rather than “cost-per-life-year-saved.” The cardiomyopathies identified by screening are progressive in nature. Also, it is unreasonable to expect that all disqualified athletes will stop exercising completely (3); yet, the degree of sportive activity that is completely risk-free remains undefined. Finally, the majority of disqualified athletes would not qualify for defibrillator implantation if truly asymptomatic. For all these reasons, athletes disqualified from sport participation should not be expected to have a normal life span.
The staggering costs of widespread ECG screening of athletes (>$5 billion over the next 2 decades, and >$10 million per life saved) have profound implications for society. To put this expenditure into perspective, the cost of alternative strategies to reduce SCD should be considered. For example, the Medical Emergency Response Plan for Schools is a public initiative to prepare high schools for life-threatening emergencies occurring on-campus. An American Heart Association expert panel estimated that a preventive program consisting of widespread cardiopulmonary resuscitation training, effective communication systems within campuses, and external automatic defibrillators operated by lay rescuers would result in a cost-per-life-saved of $1.5 to $3.3 million (25). Given the limited resources available to healthcare in the United States and the limited life-saving potential of ECG screening, it is clear that mandating such a program, rather than applying it selectively, is likely to hinder the penetration of other preventive measures for cardiac arrest victims.
For supplemental tables, please see the online version of this article.
Preventing Sudden Death of Athletes With Electrocardiographic Screening What Is the Absolute Benefit and How Much Will it Cost?
The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Halkin and Steinvil contributed equally to this paper.
- Abbreviations and Acronyms
- electrocardiographic/ electrocardiogram(s)
- electrophysiologic studies
- European Society of Cardiology
- magnetic resonance imaging
- sudden cardiac death
- Received July 9, 2012.
- Revision received August 29, 2012.
- Accepted September 4, 2012.
- American College of Cardiology Foundation
- Chaitman B.R.
- Myerburg R.J.,
- Vetter V.L.
- Steinvil A.,
- Chundadze T.,
- Zeltser D.,
- et al.
- Corrado D.,
- Pelliccia A.,
- Bjornstad H.H.,
- et al.
- Maron B.J.,
- Thompson P.D.,
- Ackerman M.J.,
- et al.
- Rembold C.M.
- National Federation of State High School Associations
- National College Athletic Association Research
- Centers for Medicare and Medicaid Services
- Gersh B.J.,
- Maron B.J.,
- Bonow R.O.,
- et al.
- Viskin S.,
- Halkin A.,
- Steinvil A.,
- Zeltser D.
- Schmied C.,
- Zerguini Y.,
- Junge A.,
- et al.
- Ezekowitz J.A.
- Corrado D.,
- Pelliccia A.,
- Heidbuchel H.,
- et al.,
- Section of Sports Cardiology, European Association of Cardiovascular Prevention and Rehabilitation
- Hazinski M.F.,
- Markenson D.,
- Neish S.,
- et al.