Abstract
Objective To determine the effect of electrocardiogram (ECG) screening on prevention of sudden cardiac arrest and death (SCA/D) in young athletes and military members.
Data Sources MEDLINE, Embase, Cochrane CENTRAL, Web of Science, BIOSIS, Scopus, SPORT discus, PEDro, and clinicaltrials.gov were searched from inception to dates between 2/21/19 and 7/29/19.
Study Selection Randomized and non-randomized controlled trials, where pre-participation examination including ECG was the primary intervention used to screen athletes or military 40 years of age or younger. Accepted controls were no screening, usual care, or pre-participation examination without ECG. 3 published studies, and one conference abstract were identified for inclusion.
Data Extraction In all four studies, risk of bias was assessed with the Cochrane risk of bias tool, and found to be generally high. Two studies had data extracted for random effects meta-analysis, and the remaining study and conference abstract were included in narrative review.
Data Synthesis 4 studies (11,689,172 participants) were included, all at high risk of bias. Pooled data from two studies (n= 3,869,274; very low quality) showed a 42% relative decrease in sudden cardiac death, equating to an absolute risk reduction of .0016%. Uncertainty was high, with a potential 67% relative decrease to a 45% relative increase in those screened with ECG based on 95% confidence intervals (RR 0.58; 95%CI 0.23, 1.45). Heterogeneity was found to be high as measured with I2 statistic (71%).
Conclusion There is very low quality evidence ECG screening decreases risk of sudden cardiac death in young athletes and military members. Decisions need to consider evidence that ECG screening could also increase risk of sudden cardiac death based on the findings of meta-analysis.
PROSPERO Registration CRD42019125560
Key Points ECG screening of athletes has been shown to be more effective than history and physical examination alone to diagnose conditions which put the athlete at risk for sudden cardiac arrest or death (SCA/D). Few data are available to answer the question of the effectiveness of ECG screening in preventing SCA/D in young athletes.
We identified only four published accounts (3 full papers and one conference abstract) of non-randomized trials reporting on the effectiveness of ECG screening to prevent SCA/D in young athletes and military members. The quality of the published evidence is judged to be of very low quality to answer the question of whether ECG screening prevents episodes of SCA/D. No difference was identified between screened and non-screened athletes in data synthesis of two of the published articles eligible for meta-analysis (RR 0.58; 95%CI 0.23, 1.45).
- athlete
- military
- electrocardiogram screening
- sudden cardiac arrest
- sudden cardiac death
Introduction
Efforts to reduce the incidence of sudden cardiac arrest and death (SCA/D) in young athletes has led the European Society of Cardiology 1 to recommend electrocardiogram (ECG) screening as part of a pre-participation examination (PPE) of young competitive athletes prior to participation in 2005, and updates have confirmed their belief in screening with ECG2. Professional bodies around the world have followed this recommendation with statements of their own. With some in agreement including the International Olympic Committee 3; some agreeing but with limitations, such as the Australasian Society for Sports Physicians 4; while organizations in the United States have resisted calls for blanket screening 5,6. The evidence-base to support inclusion of ECG screening for reducing incidence of SCA and SCD in young athletes has not undergone systematic review. A previous systematic review assessed the effectiveness of ECG screening to detect potentially lethal cardiac disorders, but did not address the impact on SCA, SCD and the potential negative effects of ECG screening 7.
The provision of a systematic summary of existing data on the outcomes of ECG screening will provide both the public, health care practitioners, and policy makers with vital information about the health effects of ECG screening in these populations when compared with history and physical examination alone.
The aim of this study was to review all available evidence assessing the effect of the addition of ECG screening as part of PPE in young athletic, and military populations on incidence of SCA/D, and to synthesize available research to evaluate the effect of the addition of ECG on the occurrence of SCA/D.
Methods
This review is part of a project with two objectives: identifying the global incidence of SCA/D in athletes and military members8, and evaluating the effect of screening ECG on SCA/D in the same population. It was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)9 guidelines, and was registered at PROSPERO March 18th 2019 under CRD42019125560 10.
Data Sources and Searches
The search strategy was designed in conjunction with a medical librarian experienced in systematic reviews (MS), and the search strategy used combined the dual objectives into a single search. The search strategy is included in the supplementary appendix. We searched MEDLINE, Embase, Cochrane CENTRAL, Web of Science, BIOSIS, Scopus, SPORT discus, PEDro, between 2/22/19 and 3/1/19 and Clinicaltrials.gov on 7/29/19. Review articles and position statements were reviewed for eligible articles 11–13. There was no limitation on language or date of publication.
Study Selection
Studies eligible for inclusion were randomized and non-randomized controlled trials, where pre-participation examination including ECG was the primary intervention used to screen athletes or military 40 years of age or younger. Accepted controls were no screening, usual care, or pre-participation examination without ECG. Age 40 was selected due to the increased incidence of coronary artery disease as cause of SCA/D with increasing age, and the desire to focus on etiologies other than coronary artery disease 11,14.
The pre-specified primary outcome was the difference in SCA/D in athletes and military populations screened with ECG compared to control groups not screened with ECG. Secondary outcomes were planned but not carried out due to the lack of existing data in the literature. Details on these outcomes are listed in the supplementary appendix.
Data Extraction and Quality Assessment
Independent dual-investigator article screening, selection, risk of bias (ROB) assessments, and extraction was performed with Covidence (Covidence.org, Veritas Health Innovation, Melbourne, Australia 2018). The primary author (AL) screened all titles, and the second reviewer was from a team of three (CM, NP, VL). The Cochrane risk of bias tool 15, native to Covidence was used for ROB assessment. Disagreements were resolved by discussion with primary author (AL), and the second reviewer.
Data Synthesis and Analysis
Meta-analysis was performed with the statistical package native to Review Manager (Cochrane Collaboration, London, UK) using the random-effects Mantel-Haenszel method based on the clinical heterogeneity within the included studies16. Data are presented as relative risk with 95% confidence interval in those screened with ECG compared with those not screened with ECG where possible. Heterogeneity is reported with summary statistics I2 and Chi2, with pre-specified values of <30% considered low; 30-70% considered moderate; and >70% considered high. A p-value of 0.10 or lower for Chi2 statistic was indicated statistical heterogeneity. Sensitivity analysis and publication bias assessment (e.g. funnel plot asymmetry) were planned but not performed due the small number of included studies.
Ethical Approval
Ethical approval was not necessary as only publicly available data was included in this review.
Role of the Funding Source
No funding was received for this project.
Results
After removal of 10,780 duplicates, and addition of a further 11 titles after hand search, 20,059 titles and abstracts were screened from database searches, and additions from review of recent review articles. Full text screening was carried out on 322 articles, four of these met criteria for inclusion including three published articles, and one conference abstract. (Figure 1).
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow sheet
Included studies
The three studies and single conference abstract included were all non-randomized controlled studies. The three articles were on athletes17–19 and included a total of 6,431,380 subjects, and the included conference abstract was in military males only20 including 5,257,792 subjects. Two studies, and the conference abstract included a historical control group, comparing the rate of SCD in a control cohort prior to the implementation of ECG screening, to an intervention cohort after ECG screening began 17,19,20. The third included study compared two separate previously published cohorts of athletes, one that had not been screened with ECG as part of their PPE, and one that had been screened with ECG18. No studies were identified which reported an outcome of SCA.
Only two studies were included in meta-analysis, both on athletes17,18. In both studies included in meta-analysis it was unclear if the historical control group received a screening PPE, or no examination17,18. In those included in the meta-analysis, the intervention group in Corrado, et. al. (2009) received ECG screening as part of their PPEs; and Steinvil, et. al. (2011) included an intervention group that received PPE with ECG, and every four year exercise stress test in those aged 17-34, with annual stress test over this age.
In the two entries not included in the meta-analysis, the conference abstract20 did not provide extractable data for meta-analysis; and Maron, et. al. (2009) included a portion the Corrado, et. al. (2006) cohort in their analysis and was left out of meta-analysis to avoid double counting included subjects. The conference abstract including only military males20 compared conscripts prior to the initiation of ECG screening, to those after the initiation of ECG screening upon entry to the army. Maron, et. al. (2009) compared the rate of SCD in a cohort who had been screened with PPE including ECG, with the rate in a cohort that received a PPE without ECG. Full descriptions and characteristics of these four studies are presented in the supplementary appendix.
Risk of Bias Assessment
All included studies were found to have either unknown, or high risk of bias in the large majority of categories evaluated (Table 1). No included studies reported funding which was determined to increase their risk of bias. Further details on the risk of bias determination is included in the characteristics of included studies table in the supplementary appendix.
Risk of bias in included non-randomized trials in objective 2; based on Cochrane risk of bias tool
Effectiveness of ECG screening
Corrado, et. al. (2006) reports an analysis on SCD in athletes in the Veneto region of Italy before and after the initiation of a mandated ECG screening program. The authors report a primary outcome of an 89% decrease in incidence of SCD after the implementation of ECG screening program for athletes in 1982. This is done by comparing a two-year time period prior to the initiation of screening (1979-80), to the final two years of the screening period (2003-04). In our analysis, we compared the data from 23 years of athletes screened with ECG compared with the three years of athletes not screened with ECG showed a 63% decreased risk in the ECG screened group (RR 0.37; 95%CI 0.20 to 0.69).
Steinvil, et. al. (2011) compares SCD events in athletes reported in two Israeli newspapers covering 90% of the country, before and after the initiation of a cardiac screening program. The required intervention began in 1997, and the authors compared media reports in the newspapers for the 12 years before 1997, to the 12 years after the initiation of the law. The authors showed a non-significant 5% decrease in risk in those athletes undergoing cardiac screening with ECG and stress test, to those which did not receive this screening (RR: 0.95; 95%CI 0.43 to 2.13). The findings were highly uncertain, with a potential 57% decrease in risk of SCD to 113% increased risk noted in the 95% confidence interval.
A conference abstract by Abacherli, et. al. (2014) details a comparison of SCD in Swiss male military conscripts separated into age groups 16-19, 20-14, 25-29. The authors compare episodes of SCD after the initiation of ECG screening, compared to historical controls prior to ECG screening. A statistically significant reduction in the ECG screened 20-24 age group with a point estimate of 0.56 (CI: 95% 0.35 to 0.91) was reported. The same comparison in men aged 16-19 was found to be 0.89, and 25-29 was found to 1.04. These were described as non-significant, with only the point estimates, and no confidence intervals reported. The abstract is unclear as to whether the statistical method used was relative risk or odds ratio, making interpretation of the findings difficult. No extractable data were present. Contact with the author revealed that there was no full text article produced from this data, and the authors were unable to share the data at the time of contact.
Maron, et. al. (2009), in comparing a cohort of U.S. athletes in the state of Minnesota who have undergone PPE without ECG screening, with a proportion of ECG screened athletes from the Corrado, et. al. (2006) Italian cohort over a similar time period, report a 6% decrease in risk of SCD but estimates were also uncertain with a potential 59% decrease in risk to a 112% increase in risk in the 95% confidence interval (RR 0.94; 95%CI 0.41, 2.12).
Two studies 17,19 including athlete participants were included in meta-analysis (Figure 2). The results show a non-significant relative decrease risk of 42% for SCD in athletes screened by ECG but uncertainty was high, with a potential 77% relative decrease to a 45% relative increase in those screened with ECG (RR 0.58, 95%CI 0.23 to 1.45; I2 = 71%; Chi2 3.41, p=0.06). The heterogeneity present in the analysis was high by both I2, and Chi2 methods.
Meta-analysis of studies reporting data on outcomes of sudden cardiac death in athletes screened with ECG (experimental), compared to athletes not screened with ECG (control)
Corrado, et. al. (2006) reported data beyond the outcome of SCD, and this was done only in a portion of the intervention cohort, not allowing comparison with the control group. The authors report that 9% (3,914 of 42,386) of athletes were referred for further testing after their initial ECG screening and a further 2 % (879 of 42,386) were ultimately removed from sport. The authors did not describe the treatment rendered to athletes with positive screening ECG beyond the further diagnostic studies used and did not report on athletes returning to sport after treatment.
Quality of evidence
For the primary outcome of SCD we judged the evidence to be of very low certainty due mainly to high risk of bias, serious inconsistency, and serious imprecision.
Discussion
Our systematic, pre-registered, comprehensive, and up-to-date review found very low-quality evidence that ECG screening decreases risk of sudden cardiac death in young athletes and military members. However, caution is needed when considering this finding. Firstly, we were able to perform meta-analysis only with two studies. The absolute risk reduction from pooling these studies was 0.00157%. Using a single year assessment21, this would result in a number needed to screen to prevent one death in one year of 63,694. Secondly, only one of the four studies included reached statistical significance when evaluating the effect of ECG screening on SCD, and the remaining studies report confidence intervals which include both considerable decreased and increased risk with ECG screening. Thirdly, the findings of the included conference abstract are based on unpublished data20. And fourth, due to high risk of bias, high heterogeneity and poor precision of effect estimates, the overall certainty of evidence on the effectiveness of ECG screening was judged to be very low. Taken together, we have very low confidence that our findings would not change substantially with further high-quality research.
The existing evidence base to support the use of ECG screening to prevent SCD in athletes is largely confined to the data presented in the Corrado, et. al. (2006) article included in the review. There are significant methodologic concerns about this article, including the inherent bias, and likely confounding present, when comparing a small historical control group to a much larger intervention group some 20 years later. There have also been concerns raised about the transparency of the data reported, and further follow up data on the Italian screening program22. While there have been no recent controlled studies published, two recently published cohort studies may call into question the ability of screening athletes to prevent SCA/D. Both report on cohorts of mostly male, professional soccer players who underwent ECG screening as part of pre-participation examinations23,24. Both reported results with relatively high rates of SCA/D at 6.8, and 63 per 100,000 athlete years compared to accepted estimates of SCA/D in athletic populations. When considering the ability of ECG screening to prevent episodes of SCA/D, it is notable that published data on events of SCA/D suggest that approximately 60% of cardiac conditions which cause SCA/D in athletes may be identifiable with ECG screening25,26.
While there remains disagreement, and a general lack of empiric data to support the use of ECG to prevent SCA/D, multiple authors have advocated it’s use as an effective addition to the pre-participation examination due to the ability to better identify conditions putting athletes at risk of SCA/D27. A systematic review in 20157 compared the likelihood of history (hx), physical exam (PE), and ECG to identify potentially lethal cardiac disorders reporting the superiority of ECG in sensitivity (ECG=94%; hx 20%; PE 9%), positive likelihood ratio (ECG=14.8; hx=3.22; PE=2.93), and false positive rate (ECG=6%; hx=8%; PE=10%). A recent publication by the National Screening Committee in the United Kingdom28 has reviewed available data on ECG screening in athletes and recommended against its use based on the overall low incidence of SCA/D, confirmed by our systematic review8, as well as the lack of an effective screening test to identify those at risk of SCA/D. Results in a recent cohort study focusing on collegiate athletes in the U.S. comparing the results history and physical exam to additional ECG screening in the same cohort of patients29 show false positive rates of 33.3% for history alone, 2% with physical exam alone, and 3.4% with ECG alone. Sensitivity with ECG was reported as 100% compared to 15.4% for history, and 7.7% for physical exam.
As demonstrated in this review, there remains little trial data to compare information on how ECG screening affects athlete’s removal from sport, follow up treatment, and potential return to sport. There is cohort data on some of these outcomes, the most notable again being from the included Corrado, et. al. (2006) article which reports on a subgroup of the screened athletes included, with 9% found to be abnormal and receiving further testing, ultimately resulting in 2% excluded from sport. A 2014 scientific statement from the American Heart Association30 details a selection of published articles on ECG screening cohort studies, with no control groups and community ECG screening projects in athletes 12-25 years old. Articles reviewed report abnormal ECG rates from 2.5-25%, further testing of 2.5-24%, and disqualification from sport rates 0f 0.2-2% of those screening. More modern cohort studies referenced above 29,31 comparing history and physical exam to additional ECG screening in the same cohort of patients have shown lower levels of initially positive results in ECG screened athletes to those completing the American Heart Association32 history questionnaire, and physical exam alone. It is difficult to compare how these findings affected further testing and treatment, as all subjects received history, physical, and ECG screening interventions. Drezner, et. al. (2016) report the identification of 0.25% of athletes with a serious cardiac condition after full evaluation in collegiate athletes in the United States 29. Drezner, et. al. (2016) also report an average of 2.6 days out of sport to evaluate those with ECG abnormalities on screening. It is also notable, that within studies using ECG screening and reporting the cardiac abnormalities identified, that Wolff-Parkinson-White syndrome is often far and away the most frequent finding, often making up the majority of identifiable cardiac conditions considered serious 29,31,33. Concerns have been raised about identifying asymptomatic individuals with this condition on ECG and how large an impact on prevention of SCA/D34 this may provide.
Over the past decade, refinement of the ECG criteria for diagnosing these potentially lethal cardiac conditions has continued to improve the sensitivity, and decrease the false positive rates35. While these advancements in the diagnostic capability of ECG screening have occurred, there have been no controlled trials published on independent cohorts of patients comparing the ability of ECG to PPE with history and physical alone to prevent SCA/D. There is great need for a prospective study which tests the utility of screening ECG to prevent SCA/D in athletes.
Carrying out a prospective study on this topic would be a daunting task, and with the rarity of the condition may not be possible. To undertake such a project, one could consider randomizing clusters of high school and collegiate athletes to ECG screening with PPE compared to PPE. A model such as this may be even more feasible in the military, where large numbers of recruits enter into service every year. It would be possible to randomize subjects in this setting, which could lead to the data needed. Short of these two trials, it should be possible to compare the rates of SCA/D in a cohort study in the U.S. Many universities have transitioned to testing their athletes, and comparing the rates of SCA/D in these universities, to comparable universities who do not screen, would be technically feasible.
We believe the strength of this review lies in the breadth of the search for controlled trials of any kind which report on the ability of screening ECG in athletes or military members to prevent SCA/D. The primary limitation of our review is the low quality of evidence provided by the included studies leading to uncertainty for decision making. The limitations lie in both the paucity of, and the poor quality of the identified research reporting outcomes on SCA/D in our population.
Conclusion
There is very low-quality evidence ECG screening decreases risk of sudden cardiac death in young athletes and military members. Decisions regarding the use of screening ECG need to consider evidence here that ECG screening could also increase risk of sudden cardiac death. We have very low confidence that our findings would not change substantially with further high-quality research.
Data Availability
Reasonable request for data will be provided after author contact.
GRADE Summary of Findings Table
References
