Discussion
The cumulative incidence of TLOC among young-adults (28.8%) and children (18.9%) from the SCD-SOS survey is high. Unexplained syncope was present in 14.8% of the individuals and it was independently associated with age<18 years-old, male gender, participation in competitive sports, major clinical high-risk features, syncope after exertion, fever context, lower number of syncopal episodes and a history of palpitations requiring medical care. These may be regarded as potential warning signs in individuals who are evaluated in a non-acute setting. Additionally, ECG abnormalities in this population with TLOC are rare and could not be independently associated with US.
Syncope is a common medical problem that affects individuals of all ages and is most frequently recognized as a benign event. (7) However, most of the previous studies of syncope were conducted in emergency departments or hospitals, and evidence about incidence and causes is lacking in the general population. In Italy, in a cohort of elderly patients admitted to an emergency department due to TLOC, (10) Rosso et al. found that 67% had reflex syncope, 10% had orthostatic hypotension and 15% had confirmed cardiac syncope. Accordingly, in a paper previously published by our group, (11) the majority of young-adults presenting with a TLOC episode, reported pre-syncopal symptoms and a context that was suggestive of a benign (non-cardiac) etiology. (7) In our SCD-SOS cohort, the majority of TLOC episodes were also benign. US was present in 14.8% and it was associated with major and minor high-risk clinical features, thus suggesting the presence of a serious underlying cause and a higher likelihood of a cardiac syncope. (7) Marijon et al. have shown that warning symptoms are frequent prior to sudden cardiac arrest, but the vast majority of them are not acted upon. (12) This evidence has triggered the acknowledgement of a new front in the fight against SCD - anticipation based on warning symptoms. (13) Therefore, we consider that early symptoms characterization and identification of the clinical features associated with US may contribute to improve outcomes related to SCD.
We found that male gender and less frequent syncopal events were independent predictors of unexplained syncope. Ungar et al. have reported that, in patients referred to the emergency department following a syncopal event, (14) male gender increased and the incidence of syncopal recurrences decreased mortality. On the other side, coherent with our results, in the Italian cohort, palpitations preceding syncope, the absence of autonomic prodromes and syncope during exertion were also considered independent predictors of cardiac syncope. (10) In our cohort, the presence of typical symptoms was used to classify other causes of TLOC, therefore we excluded the high-risk feature absence of warning symptoms from the multivariate analysis to avoid potential information bias.
Our findings showing that participation in competitive sports and the occurrence of syncope during exertion were independent predictors of US, may be difficult to interpret in the light of the current knowledge. Previous authors described that syncopal episodes in young athletes are in most cases non-cardiac and of reflex origin. (6) Additionally, in a Danish nationwide retrospective study in young-adults, Risgaard et al. found no differences in the incidence rate of SCD between competitive and non-competitive athletes aged 12-35 years.(16) On the other side, coherent with our results, there is some evidence showing that exertional syncope tends to be more frequently associated with structural heart disease and with an unfavorable outcome in adults (17) and mainly with channelopathies and other rhythm disorders in children. (18) Stronger evidence supports an association between SCD and sports practice, namely among athletes in the Veneto region, in whom 91% of the SCDs occurred during sports activity or immediately afterward. And in a recent British post-mortem study on the etiology of SD in sports, which included 69% of competitive-athletes, showing that the majority of athletes died during exertion (61%). (19)
Another independent predictor of US in our cohort was an history of fever related-syncope. Some of these individuals had ECG features potentially suggestive of Brugada patterns. Previous studies suggested a higher risk of fever-related symptoms and arrhythmic events in children and young-adults with Brugada Syndrome. (20) Finally, even though the ECG changes in a cohort of young-adults with a high proportion of athletes may be difficult to interpret, at least 0.8% of the participants had abnormalities in the ECG that may require further assessment to exclude the presence of intrinsic cardiac disease. (9) There was a negative association between US and Fridericia andBazett QTc, and a positive association with left ventricle hypertrophy in the baseline ECG in the univariate analysis, but after adjustment no ECG predictors of US were identified in the multivariate analysis. We hypothesize that for a cohort composed of predominantly healthy individuals, a much bigger sample would be necessary to further explore electrocardiographic predictors of US. More individuals would likely expand the statistical power of the sample and allow those analyses. However, this also appears to suggest that the effect size of frequent ECG traits as predictors of US is low.
We estimate that the costs for the Portuguese Health Services to screen 11 878 participants in the SCD-SOS survey would be 79 207,00\euro (2000,00\euro for the software and for placing online the questionnaires plus 6,50\euro per ECG (Diário da República 173/2018, Série I de 2018-09-07,https://data.dre.pt/eli/port/254/2018/09/07/p/dre/pt/html)). Given that the survey allowed the diagnosis of US in 466 individuals and the identification of relevant ECG abnormalities in 26 participants, the estimated cost per diagnosis for a diagnostic yield of 4.1% is 160,99\euro. In our opinion, this is an acceptable cost for identifying patients who are at risk of developing cardiac diseases and/or have the potential to suffer SCD, thus warranting further investigations and follow-up.