Syncope is a common problem in childhood. About 15% children under the age of 18 years have had at least one syncopal episode.Reference Driscoll, Jacobsen, Porter and Wollan 1 – Reference Wieling, Ganzeboom and Saul 3 Syncope in children and adolescents can be caused by a variety of underlying diseases. According to our study and other research, the most common cause of syncope is vasovagal syncope.Reference Wong and So 4 – Reference Zhang, Du, Wang, Du, Wang and Tang 6 Other causes include orthostatic hypotension, cardiac and postural orthostatic tachycardia syndrome, and epileptic seizure or psychogenic pseudosyncope. Cardiac syncope represents 2–6% of paediatric cases.Reference Driscoll, Jacobsen, Porter and Wollan 1 , Reference Zhang, Du, Wang, Du, Wang and Tang 6 , Reference Wren 7 The causes of cardiac syncope are mainly arrhythmia and structural heart diseases. The prognosis of cardiac syncope and vasovagal syncope is different, but the two kinds of syncope in clinical practice are similar. Therefore, it is important to distinguish cardiac syncope and vasovagal syncope. History taking plays a key role in the initial evaluation of syncope. 8 In adult patients, a point score based on clinical history predicting arrhythmias for patients with unexplained syncope in the emergency department has developed. For example, Sarasin et alReference Sarasin, Hanusa, Perneger, Louis-Simonet, Rajeswaran and Kapoor 9 found that abnormal electrocardiogram, a history of congestive heart failure, and age greater than 65 years were predictors of arrhythmic syncope. Rosso et al also developed a diagnostic point score to discriminate cardiac syncope from syncope of other causes. They found that abnormal electrocardiogram and/or heart disease, palpitations before syncope during effort or in supine position, absence of autonomic prodromes, and absence of predisposing and/or precipitating factors were the predictors of cardiac syncope.Reference Rosso, Ungar and Maggi 10 However, this score is of limited value when evaluated in the children and adolescents with syncope because, as our previous study indicated, the causes of syncope in children and adolescents differed from those of adult patients.Reference Qingyou, Junbao, Jianjun and Wanzhen 11 The better understanding of clinical characteristics of cardiac syncope in children and adolescents is an important issue to improve the diagnosis of syncope in children. The present study was, therefore, undertaken to identify clinical features of patients with cardiac syncope and explore the useful history findings predictive of the cause of syncope.
Methods
Study subjects
The study was prospective and included consecutive patients less than or equal to 18 years of age with suspected syncope admitted to the Pediatric Syncope Unit of five hospitals in Beijing, Shanghai, Changsha, and Wuhan in China from October, 2006 to June, 2010. This investigation was sponsored by the Pediatric Syncope and Related Disorders Clinical Study Group.Reference Zhang, Du, Wang, Du, Wang and Tang 6 All patients were affected by transient loss of consciousness, which, on initial evaluation, was attributed to a syncopal condition or a syncopal condition not to be excluded. Patients with symptoms that were clearly compatible with seizure disorders, vertigo, coma, or shock were not included. All patients underwent a standardised initial evaluation consisting of history taking, physical examination, orthostatic blood pressures, and standard electrocardiogram. According to the initial evaluation, patients were classified into three groups: those with a cause of syncope strongly suspected according to disease-specific diagnostic criteria; those with a specific entity causing syncope suspected by suggestive signs or symptoms that required confirmation by selected diagnostic procedures – such as echocardiography for suggested hypertrophic cardiomyopathy; and those with the cause of syncope undetermined. When the results of the initial evaluation were inconclusive, further clinical and laboratory investigations in patients for whom a cause of syncope was strongly suspected were guided by a diagnostic hypothesis, which was based on the abnormalities found in the initial clinical assessment.
Diagnostic criteria
Diagnostic criteria of the causes of syncope were developed before recruitment on the basis of the best evidence-based recommendations, and assignment of a cause was based on strict adherence to these criteria. In particular, mechanical cardiac syncope was diagnosed in the presence of outflow obstruction of heart disease, such as hypertrophic cardiomyopathy or idiopathic pulmonary hypertension.Reference Zhang, Du, Wang, Du, Wang and Tang 6 , 8 , Reference Rosso, Ungar and Maggi 10 Arrhythmic syncope was diagnosed on the basis of a standard electrocardiogram, electrocardiogram monitoring, 24-hour Holter recordings, or during the electrophysiological study. A diagnosis of arrhythmic syncope was considered with electrocardiogram findings – with 12-lead electrocardiogram and Holter recorder – of sinus pause greater than or equal to 3 seconds, sinus bradycardia less than or equal to 45 beats per minute, Mobits 2 or advanced 2:1 block, complete atrioventricular block, non-sustained ventricular tachycardia greater than 5 beats per minute, congenital long-QT syndrome with prolonged QT interval corrected for heart rate by Bazett's formula greater than 480 milliseconds, and episodic ventricular tachycardia, particularly torsade de pointes. Diagnostic results of electrophysiologic studies included prolonged corrected sinus-node recovery time indicative of sinus-node disease, prolonged H–V interval, and superventricular tachycardia greater than or equal to 180 beats per minute.Reference Zhang, Du, Wang, Du, Wang and Tang 6 , Reference Rosso, Ungar and Maggi 10 , Reference Alboni, Brignole and Menozzi 12
A diagnosis of vasovagal syncope was madeReference Zhang, Du, Wang, Du, Wang and Tang 6 , Reference Alboni, Brignole and Menozzi 12 , Reference Calkins, Shyr, Frumin, Schork and Morady 13 if a trigger – such as prolonged standing, mental stress, and pain – was associated with prodromal signs and symptoms – such as nausea, light-headedness, and diaphoresis called typical vasovagal – or if a less typical history was confirmed by a positive table tilt test and/or information from follow-up, and no other cause of syncope was found after additional testing and follow-up – called tilt-induced syncope.
Owing to the fact that typical vasovagal syncope patients were diagnosed only on the basis of history, we used typical vasovagal syncope patients as control group to find clinical characteristics of cardiac syncope in children and adolescents.
Postural orthostatic tachycardia syndrome is defined as the development of orthostatic intolerance symptoms accompanied by a heart rate increase of at least 30 beats per minute, or a rate that exceeds 120 beats per minute, which occurs within the first 10 minutes of standing or head-up tilt test, in the absence of other chronic debilitating disorders, prolonged bed rest, or medications that impair autonomic reflexes.Reference Zhang, Du, Wang, Du, Wang and Tang 6
Orthostatic hypotension was defined as a persistent decrease of more than 20 millimetres of mercury of systolic blood pressure or decrease of more than 10 millimetres of mercury of diastolic blood pressure within 3 minutes after the start of head-up tilt test or standing.Reference Zhang, Du, Wang, Du, Wang and Tang 6
Situational syncope was diagnosed if syncope occurred during or immediately after micturition, defecation, coughing, or swallowing.Reference Zhang, Du, Wang, Du, Wang and Tang 6
Syncope questionnaire
All patients completed a 118-item questionnaireReference Alboni, Brignole and Menozzi 12 – Reference Sheldon, Rose, Connolly, Ritchie, Koshman and Frenneaux 14 developed with formative input from Calkins et al.Reference Calkins, Shyr, Frumin, Schork and Morady 13 It included age, gender, duration of disease, attack frequency, predisposing factors, prodromal symptoms of syncope, syncope duration, body position at the onset of syncope, any associated symptoms, the duration of the loss of consciousness, a previous history of syncope related to heart disease, and family history of sudden death or syncope. A standard 12-lead electrocardiogram was obtained on all patients. The variables investigated are reported in Table 2.
This study was approved by the ethics committee of all participating hospitals, and for each child informed parental consent and the child's agreement were obtained.
Statistical analysis
Continuous variables were summarised as mean scores and standard deviations from the mean. For the purpose of comparing the distribution of continuous variables, a t-test or a non-parametric alternative, Mann–Whitney, was used where appropriate. The χ2 method or Fisher test was used to compare categorical variables, as appropriate. Logistic regression analysis was used to find out the independent predictors for cardiac syncope. SPSS software (version 13.0) was used to analyse the data (p less than 0.05 was defined as the level of significance).
Results
The diagnosis of cardiac syncope in children
During the recruitment period, data from 275 patients were collected. At the end of the diagnostic evaluation, vasovagal syncope was established in 132 (48.0%), including 55 cases with typical vasovagal syncope and 77 with tilt-induced vasovagal syncope. Postural orthostatic tachycardia syndrome was diagnosed in 67 (24.4%), situational syncope in 8 (2.9%), orthostatic hypotension in 7 (2.5%), cardiac syncope in 31 (11.2%) including 22 patients caused by arrhythmias (8.0%) and 9 caused by mechanical cardiac syncope (3.2%), and the cause of syncope remained unexplained in 15 (Table 1). In 15 (5.5%) patients, non-syncopal cause of loss of consciousness was confirmed by investigations. Electroencephalography, brain computed tomographic scan, and magnetic resonance imaging were applied in the case of suspected epilepsy or migraine; laboratory blood tests were performed in the case of suspected hypoglycaemia and severe anaemia, and psychiatric consultation in the case of psychogenic pseudosyncope.
Causes of syncope (n = 275).
Table 2 shows the diagnosis, history of electrocardiogram performance, and diagnostic tests needed to confirm the diagnosis of 31 cases of cardiac syncope. Of the 31 cases of cardiac syncope, case 1 and case 24 – congenital long QT syndrome – had a family history of sudden death; 5 patients had a previous history of heart disease, of which case 8 had a history of myocarditis, case 9 had a history of surgery of total anomalous pulmonary venous return, cases 11 and 28 had a history of ventricular septal defect repair, and case 22 had a history of surgical correction for transposition of the great arteries.
The diagnosis and diagnostic tests needed to confirm the diagnosis of 31 cases of cardiac syncope.
AF = atrial fibrillation; Angiog = pulmonary arterial angiography; AT = atrial tachycardia; CAVB = complete atrioventricular block; CVH = combined ventricular hypertrophy; ECG = electrocardiogram; Echo = echocardiography; EPS = electrophysiological study; F = female; HCM = hypertrophic cardiomyopathy; Holt = Holter electrocardiogram monitoring; IPAH = idiopathic pulmonary arterial hypertension; LQTS = long QT syndrome; M = male; PSVT = paroxysmal supraventricular tachycardia; PVT = paroxysmal ventricular tachycardia; RVH = right ventricular hypertrophy; SSS = sick sinus syndrome; VT = ventricular tachycardia
Univariate analysis of clinical features of cardiac syncope compared with typical vasovagal syncope in children and adolescents (Table 3)
Using univariate analysis, children with cardiac syncope were younger than those with vasovagal syncope in terms of age at the first syncope. The mean age of children with cardiac syncope was 8.5 years old, whereas it was 11.6 years old in children with vasovagal syncope.
Predictors of cardiac syncope on univariate analysis in children and adolescents with syncope.
ECG = electrocardiogram
Most cardiac syncope was triggered by exercise, whereas most vasovagal syncope was triggered by prolonged standing, fear or pain emotion, or warm-crowded place. Syncopal spells in supine position were common in cardiac syncope, but the spells of the majority of children with vasovagal syncope occurred while standing (96.4%). Only 2 out of 55 children with vasovagal syncope had syncopal episodes while in the supine position. The number of children with cardiac syncope who had prodromal symptoms was significantly less than that in children with vasovagal syncope. Only 51.6% of children had prodromal symptoms with cardiac syncope, whereas 87.3% of children with vasovagal syncope had prodromal symptoms. The frequency of dizziness and chest discomfort in children with cardiac syncope was significantly lower than that of children with vasovagal syncope. Other prodromal symptoms such as palpitations, headaches, and gastrointestinal symptoms occurred equally in children with cardiac syncope and vasovagal syncope.
In witnessed syncopal spells, the duration was less in cardiac syncope than in vasovagal syncope patients (1.8 versus 4.8 minutes, p less than 0.001).
More associated symptoms occurred in children with cardiac syncope than in those with vasovagal syncope. In particular, syncope with incontinence was seen more often in children with cardiac syncope than in children with vasovagal syncope, but there was no significant difference in the frequency of physical injury between cardiac syncope and vasovagal syncope.
Family history of sudden death or syncope in both groups of children was less, but only three children – two with cardiac syncope and one with vasovagal syncope – were found to have positive family histories. In 31 cases of cardiac syncope, 5 had previous history of heart diseases, but only 2 of 55 cases of vasovagal syncope had previous history of heart disease (p less than 0.05). Children with cardiac syncope had a history of abnormal electrocardiogram more frequently than those with vasovagal syncope (93.5% versus 9.1%, p less than 0.001).
Logistic analysis of clinical features in children with cardiac syncope and vasovagal syncope
Multivariate logistic regression analysis showed that history of electrocardiogram abnormalities and exercise-triggered syncope were independent predictors of cardiac syncope. Their odds ratio values and 95% confidence intervals were 180.670 and 19.9842–1633.574 (p less than 0.001), and 18.145, 1.840–178.892 (p less than 0.05), respectively. The sensitivity of history of abnormal electrocardiogram for predicting cardiac syncope was 93.5% and specificity 90.9%. The sensitivity of exercise-triggered syncope spell for predicting cardiac syncope was 61.3% and specificity 85.5%.
Discussion
Syncope is common in children, although the majority of paediatric syncope is benign. Vasovagal syncope is the most common cause of syncope in children. However, under some rare conditions, syncope may be a symptom of a potentially fatal heart disease.Reference Wieling, Ganzeboom and Saul 3 , Reference Zhang, Junbao, Jiong, Yonghong, Wanzhen and Xinhua 5 – Reference Wren 7 , Reference Soteroades, Evans and Larson 15 Therefore, in clinical practice, children presenting with syncope present a diagnostic dilemma. For the purpose of excluding heart disease as a cause in children, expenses on medical resource utilisation, associated with syncope management, are enormous.Reference Wren 7 , Reference Sarasin, Hanusa, Perneger, Louis-Simonet, Rajeswaran and Kapoor 9 , Reference Soteroades, Evans and Larson 15 , Reference Massin, Malekzadeh-Milani and Benatar 16 Many potential diagnostic approaches to this symptom complex remain difficult, expensive, and often unrewarding. History taking plays a key role in the initial evaluations of syncope. 8 Colman et al compared the characteristics of patients with prolonged QT syndrome with vasovagal syncope who were younger than 40 years of age, presenting to the emergency department. They found that a family history for syncope and sudden cardiac death, palpitation as a symptom, supine syncope, syncope associated with exercise, and emotion stress placed patients at higher risk for long QT syndrome. These findings should alert physicians to the potentially life-threatening illness of long QT syndrome. However, their study did not include other causes of cardiac syncope.Reference Colman, Bakker, Linzer, Reitsma, Wieling and Wilde 17 In previous studies, clinical variables have been found, which are predictive of cardiac syncope, and developed a diagnostic point score to discriminate cardiac syncope from syncope of other causes. However, their study results were applicable only to adult patients.Reference Sarasin, Hanusa, Perneger, Louis-Simonet, Rajeswaran and Kapoor 9 , Reference Rosso, Ungar and Maggi 10 The causes of syncope in children and adolescents differ from those of adult patients.Reference Qingyou, Junbao, Jianjun and Wanzhen 11 If we find the characteristics that may be shown as cardiac syncope based on a history of patients, the further diagnosis and treatment of syncope will be greatly improved. Therefore, the study was designed to compare characteristics of the history of the children who were diagnosed with cardiac syncope with that of the children with vasovagal syncope to find out the features that were important to the diagnosis of cardiac syncope.
Through the univariate analysis of the study, we found that the children with cardiac syncope were younger than those with vasovagal syncope, because vasovagal syncope mostly occurred in adolescents.Reference Qingyou, Junbao, Jianjun and Wanzhen 11 Most of the syncopal episodes in children with cardiac syncope were triggered by exercise. This result was consistent with previous studies in adults, and was consistent with the features of cardiac syncope noted in most of the syncope guidelines we reviewed. 8 – Reference Rosso, Ungar and Maggi 10 , Reference Alboni, Brignole and Menozzi 12 , Reference Qingyou, Junbao and Chaoshu 18 , Reference Strickberger, Benson and Biaggioni 19 Syncopal spells in children with vasovagal syncope were prevalent in the upright position, whereas cardiac syncope attacks could occur in the supine position, which, again, was consistent with the features of cardiac syncope that was noted in reviewed syncope guidelines. 8 , Reference Strickberger, Benson and Biaggioni 19 Prodromal symptoms of syncope were relatively rare in cardiac syncope, when compared with vasovagal syncope, in children. The symptoms of autonomic nervous activation such as diaphoresis, nausea, vomiting, blurred vision, and pallor occurred similarly in both groups. There was activation of the autonomic nervous system in both types of syncopal attacks, which was similar to the results of Alboni's study.Reference Alboni, Brignole and Menozzi 12 Results of most studies and syncope guidelines indicated that the presence of palpitation could be used as one of the key clinical features in cardiac syncope. 8 , Reference Rosso, Ungar and Maggi 10 , Reference Strickberger, Benson and Biaggioni 19 Colman et alReference Colman, Bakker, Linzer, Reitsma, Wieling and Wilde 17 also fund that palpitation, as a symptom, increased the risk of long QT syndromes. However, this study revealed that children with both types of syncope had a similar incidence of palpitation. The reason for such inconsistency was, mainly, due to the fact that children and caregivers did not clearly understand the meaning of palpitation. Therefore, this is why palpitations are not reliable prodromal symptoms in childhood cardiac syncope.
Syncope, accompanied with incontinence, was more common in children with cardiac syncope, when compared with children with vasovagal syncope. This, however, was inconsistent with Alboni's research.Reference Alboni, Brignole and Menozzi 12 Recently, Costanino et alReference Costantino, Perega and Dipaola 20 found an increased likelihood of adverse events within 10 days, in patients with syncope and accompanied by physical injury. Our study did not find that children with cardiac syncope had higher risk of injury, when compared with children with vasovagal syncope. Family history of sudden death was considered to be a feature of cardiac syncope, but this study did not find the similar results, perhaps because of fewer cases of cardiac syncope. It was known that the leading causes of cardiac syncope in children were mostly congenital and hereditary diseases, such as congenital long QT syndromes and hypertrophic cardiomyopathy.Reference Wren 7 , 8 , Reference Strickberger, Benson and Biaggioni 19 Therefore, prospectively, children with a family history of sudden death should be carefully and thoroughly evaluated. The children who had a previous history of heart disease suffered from cardiac syncope more often than those with vasovagal syncope, which was consistent with a large number of research studies. Most cardiac syncope patients had a history of abnormalities on a standard 12-lead electrocardiogram, which can be easily obtained, and can establish the diagnosis of some cardiac anomalies, such as long QT syndromes and third-degree atrioventricular block. In addition, mechanical heart diseases can be manifested on electrocardiogram, such as hypertrophic cardiomyopathy and idiopathic pulmonary arterial hypertension. In our study, we found that the prevalence of history of abnormal electrocardiogram findings in children with cardiac syncope was higher than that with vasovagal syncope, and it predicted cardiac syncope in children and adolescents with high sensitivity and specificity. Most researchers and the syncope guidelines of European Society of Cardiology and American Heart Association believe that electrocardiogram examination is necessary and is the most basic examination for patients with syncope.Reference Zhang, Du, Wang, Du, Wang and Tang 6 – 8 , Reference Massin, Malekzadeh-Milani and Benatar 16 , Reference Strickberger, Benson and Biaggioni 19 However, it should be noticed that the electrocardiogram findings of 2 cases in 31 cases of children with cardiac syncope were normal. Therefore, for patients with syncope with normal electrocardiogram findings, further evaluation is warranted, if there is other evidence of cardiac syncope, such as syncope during effort.
Through multivariable factors analysis, we found that the independent factors predicting for the cardiac syncope were history of abnormal electrocardiogram findings and exercise-triggered syncope, with the odds ratio value of 180.670 and 18.145, respectively. The sensitivity and specificity of prediction of a history of abnormal electrocardiogram were all above 90%. The sensitivity of exercise-triggered syncope for predicting cardiac syncope was 61.3%, and the specificity reached 85.5%. Therefore, for children with these two characteristics of syncope, the possibility of cardiac syncope should be strongly considered. Further examination for heart disease should be carried out, including Holter electrocardiogram monitoring, echocardiography, and cardiac electrophysiological examination, in order to determine the cause of syncope in children.
Our study still has limitations. First, in syncope no diagnostic “gold standard” has been determined and diagnosis often remains presumptive. Although the diagnostic criteria adhered strictly to the recommendations of guidelines, another competitive diagnosis cannot be excluded in some cases. Second, we did not develop a risk score based on clinical and history of electrocardiogram factors predicting cardiac syncope in children and adolescents. This would decrease the value of this study; however, this is the objective of a continuing study.
In conclusion, through our study, compared with vasovagal syncope in children, the patients with cardiac syncope have cardinal clinical features, especially history of abnormal electrocardiogram and exercise-triggered syncope. For children with syncope, a detailed medical history and routine electrocardiogram examination are necessary for identification of these clinical features, and further examination of heart disease for children with the above characteristics may enhance the efficiency and the accuracy of the diagnosis and save the patient and the family from intrusive procedures, lost time and money, and better utilise medical resources. The results of our study also can help general paediatricians to decide whether to hospitalise their patients with syncope.
Acknowledgement
The work was supported by National Twelfth Five-Year Plan for Science and Technology Support (2012 BA103B03), Beijing Science and Technology Project (D10100050010059) and Capital Medical Scientific Development Project (2007–2003).
