Aims Long QT syndrome (LQTS), a potentially fatal disorder, has to be distinguished from non-fatal conditions. Our aim was to investigate whether history-taking can be used in identifying patients likely to have LQTS.
Methods and results We compared the characteristics of a group of LQTS patients with syncope patients presenting at the emergency department (ED) and vasovagal patients younger than 40 years of age. Thirty-two LQTS patients were included. We included 113 patients at the ED and 69 vasovagal patients. Family history of syncope, sudden cardiac death, or cardiovascular disease was found more often in LQTS patients. Palpitations were the only symptom reported more often in this group. Syncope while supine, during emotional stress and associated with exercise was also more common among LQTS. Standing as a trigger was found more often in ED and vasovagal patients.
Conclusion We conclude that a family history for syncope and sudden cardiac death, palpitations as a symptom, supine syncope, syncope associated with exercise, and emotional stress place patients at higher risk for LQTS. These findings should alert physicians to the potentially life-threatening illness of LQTS, and act accordingly by obtaining an electrocardiogram and paying specific attention to the QT interval.
Long QT syndrome
Syncope is a frequently occurring problem. At the emergency department (ED), 1–2% of all presentations are related to syncope.1,2 A general physician will see approximately two to nine patients per 1000 patient consultations per year with syncopal complaints.3 The majority of these patients will have reflex syncope, but serious causes have to be considered. The estimated life-time prevalence of reflex syncope among young adults is 39%.4
Patients with congenital long QT syndrome (LQTS) often present with syncope.5–7 LQTS is a rare disorder with an estimated prevalence of one gene carrier in 2000–5000 persons.7,8 Syncope was the initial symptom in 69% of young LQTS patients (median age 10 years, range 4 days to 19 years) and is present in 63% of established gene carriers.9,10 LQTS is a potentially fatal disorder causing 3000–4000 sudden deaths among children and young adults each year in the USA.7 The challenge for the general or ED physician is to identify the small number of LQTS patients within the large population of patients with reflex syncope.
History-taking plays a key role in the initial evaluation of syncope.11 Several triggers have been described for both reflex syncope and LQTS. Reflex syncope is associated with triggering events like prolonged standing, pain and emotion, and attacks are usually preceded by complaints of light-headedness, nausea, and blurring of vision. Within LQTS specific triggers have been linked to specific genes. LQT1 patients experience their syncopal episodes mostly during exercise, especially during swimming, whereas LQT3 patients are prone to events during sleep/rest.12,13 In LQT2 patients auditory stimuli for syncope seem to be more common.13 The diagnostic value of features from the history in LQTS are based on observational studies, expert opinion, and common sense. Reliable data about whether the frequency and type of triggers or the prodromal symptoms differ between two groups are not available.
Although an electrocardiogram (ECG) is an important test in the evaluation of syncope as well as in diagnosing LQTS, it is not always available or applied, especially in the general practice. We therefore sought to determine whether history-taking can be used as a tool in identifying patients who are more likely to have LQTS, thus prompting clinicians to perform an ECG and evaluate the QT interval.
Design and study population
We compared the characteristics of a group of LQTS patients (cases) with a consecutive heterogeneous group of patients with syncope presenting to the ED (ED-controls). The LQTS group (n = 41) consisted of all patients with LQTS who were symptomatic (syncopal complaints) and had been genotyped at the laboratory of the Academic Medical Centre (AMC) in Amsterdam up to July 2003. The genotypes identified were LQT1 in 12 patients, LQT2 in 28 patients, and LQT3 in one. All LQTS patients were classified according to the criteria for LQTS.14 We chose syncopal patients from the ED group as our main control group. These ED patients (n = 113) were consecutively recruited from the ED of the Academic Medical Centre in Amsterdam as part of the Dutch FAST-study (Fainting Assessment Trial), which is a cohort of all patients who presented to the AMC with syncope between 2000 and 2002.15 To test the robustness of the findings that were found in the comparison between LQTS and the heterogeneous group of ED patients, we added a second control group of young (below 40 years) vasovagal syncope patients (VVS group) derived from the same Dutch FAST cohort (n = 69). These patients were included from other departments than the ED and therefore did not overlap with the ED group. The age range of <40 years was chosen because it is one of the criteria for the diagnosis of LQTS.14
For the FAST study, we developed a standardized questionnaire to collect patient data on all aspects of their syncope. The questions were based on the literature dealing with syncope and personal experience. Questions were asked about the frequency of syncopal and pre-syncopal events, family history with regard to syncopal events, sudden death and cardiovascular disease, circumstances of syncopal events, symptoms accompanying syncopal events, and medication. Attending physicians filled in the questionnaire. A similar questionnaire was sent to LQTS patients.
Within the ED group, a diagnosis was made by the attending physician based on predefined criteria. The physicians could choose between a specific diagnosis, a highly likely (>80% certainty) diagnosis and no clear diagnosis. The most common diagnosis, vasovagal syncope, was defined as certain when there was a combination of prodromal symptoms and triggering circumstances, such as prolonged standing and emotion. This is in accordance with the definition used in the ESC Guidelines on Management of Syncope, which states that ‘The ‘classical vasovagal syncope’ is mediated by emotional or orthostatic stress and can be diagnosed by history-taking.11 A highly likely diagnosis was made when typical prodromal symptoms were present but an obvious trigger was missing.
The FAST protocol prescribed an ECG in all patients. To document a prolonged QT interval, we used the generally accepted diagnostic criteria of a corrected QT of >450 ms in males and >460 ms in female patients.7,14 This qualifies as a prolonged QT although it is not a necessary diagnostic of LQTS. No diagnosis of LQTS was made within the FAST cohort.
We used χ2 tests to determine whether the presence of symptoms and triggers differed between LQTS patients and control patients. Fisher's exact test was used when expected cell counts were smaller than 5. To express the strength of these relations, we calculated likelihood ratios (LR) with 95% confidence intervals by dividing the proportion of LQTS patients who reported a certain symptom/trigger by the same proportion in the control group. To compare the distribution of continuous variables between LQTS patients and controls, we used Student's t-test or a non-parametric alternative (Mann–Whitney) where appropriate. The positive predictive value (PPV; e.g. the probability that a patient with a specific trigger has indeed LQTS) would be a clinically useful measure to present. However, our case–control design prevents us from directly calculating predictive values from our data because of the excess of LQTS patients in our study. However, by using external data on the prevalence of LQTS from the literature we can calculate these probabilities by solving the following equation:
where prev denotes prevalence of LQTS in the population of interest; LR denotes likelihood ratio, the ratio of the frequency of a feature among LQTS patients to the frequency in non-LQTS patients.
This means that we need to combine our estimate of the LR with an estimate from the literature about the prevalence (proportion) of LQTS among the total population of syncope patients. The prevalence of LQTS in the general population varies between one in 2000–5000 persons.7,8 Taking into account the numbers from Vincent et al. who found that 63% of LQTS patients will have at least one syncopal episode and the findings from Ganzeboom et al.4 (39% prevalence of syncope in young adults), we estimated that one in every 1240–3100 syncopal patients will have LQTS.7,10 However, these calculations should be interpreted cautiously because sound data on the prevalence of LQTS is lacking. SPSS software (version 11.5.1) was used to analyse the data.
Questionnaires were sent to 41 LQTS patients. Five patients did not return the questionnaire leading to a response rate of 88% (36/41). Four patients only experienced pre-syncopal episodes and were subsequently excluded from the analysis. The data of the remaining 32 patients with LQTS were used in the analysis (Table 1).
Patient characteristics of long QT syndrome (LQTS) patients, emergency department (ED) patients, and patients with vasovagal syncope younger than 40 years
LQTS patients (n = 32)
ED patients (n = 113)
Vasovagal patients (n = 69)
P-values, LQTS vs. ED, LQTS vs. VVS
P = 0.001
P = 0.05
Age at time of questionnaire
P = 0.9
P < 0.002
Age at first syncope
P < 0.001
P = 0.49
Number of episodes
P = 0.002
P = 0.67
P25, P75: 25th and 75th percentile.
During the FAST study, we included 113 patients presenting with syncope at the ED (Table 1).
Of the 113 patients, 76 (67%) had vasovagal syncope as a certain or highly likely diagnosis. The other diagnoses were situational syncope (n = 2), orthostatic hypotension (n = 5), epilepsy (n = 6), hypoglycaemia (n = 1), and atrial fibrillation with hypotension (n = 1). In another patient a highly likely diagnosis of cardiac syncope was made, but additional history by an expert and tilt table test revealed a diagnosis of vasovagal syncope. In the remaining 21 patients no certain diagnosis could be established. These undiagnosed patients were followed for 2 years. Two patients were lost to follow-up, one died from a stroke, and no cardiac death occurred in the remaining 18 patients during this period.
Despite the FAST protocol which prescribes an ECG, this was available in only 61 (54%) of the 113 ED patients because the attending physician felt that the diagnosis of reflex syncope was certain. None of the patients with an ECG had a prolonged QT interval or revealed a cardiac cause for the syncope. The median age of the patients in whom an ECG was performed was 46 years, which was significantly higher than the median age (33 years) of the patients who did not receive an ECG (P < 0.001).
From 2000 until 2002, a total of 514 patients were included in the Dutch FAST-study. Hundred of them were diagnosed with vasovagal syncope and were younger than 40 years. Of these 31 patients were already included in our ED group, leaving 69 patients for comparison with the LQTS patients.
In 58 vasovagal patients (84%) an ECG was available. None of the patients with an ECG had a prolonged QT interval.
A personal and a family history of syncopal events was found more often in LQTS patients (72%) than in the ED patients (9%; P < 0.001), as was a family history of sudden cardiac death and cardiovascular disease (P < 0.001; Table 2).
Frequency of prodromal symptoms and family history in long QT syndrome (LQTS) patients, emergency department (ED) patients, and in patients with vasovagal syncope younger than 40 years
LQTS (n = 32)a
ED (n = 113)b
Vasovagal (n = 69)c
P-value, ED vs. LQTS
P-value, VVS vs. LQTS
LR ED (95% CI)
LR VVS (95% CI)
Blurring of vision
Wanting to lie down
Tingling around the mouth and/or fingers
Family history of syncopal episodes–no. (%)
Family history of sudden death–no. (%)
Family history of cardiovascular disease–no. (%)
FAST, Fainting Assessment Trial; LR, likelihood ratio; VVS, vasovagal syncope. Values in bold face are statistically significant.
aIn six patients data of at least one symptom are missing or patient does not know.
bIn nine patients data of at least one symptom are missing or patient does not know.
cIn four patients data of at least one symptom are missing or patient does not know.
Comparison with the vasovagal patients also revealed a significant difference in the family history of syncopal episodes, sudden death, and cardiovascular disease (P < 0.001, P < 0.001, and P = 0.04, respectively).
Palpitations (fast and/or irregular heart pounding) was the only symptom that was reported significantly more often in the LQTS group than in the ED group (44% vs. 21%, P = 0.01; Table 2). The LR was therefore 2.1 (95% CI 1.2–3.7), indicating that LQTS patients reported palpitations twice as often compared with ED patients. The frequency of other symptoms was not significantly different among the two groups.
In the vasovagal syncope group only nausea was seen significantly more than in the LQTS group (29% vs. 60%, P = 0.005, LR 0.56, 95% CI 0.38–0.81). All other symptoms were comparable between both the groups (Table 2).
Syncope while supine, especially during sleep or while lying in bed was significantly more common among LQTS patients (LR 12, 95% CI 5.6–24; Tables 3 and 4, Figure 1). Standing as a trigger was found significantly more in ED patients (P = 0.007), which is likely related to the high numbers of patients with vasovagal syncope in that group. Emotional stress was confirmed as a trigger significantly more often in LQTS than in ED patients (LR 2.5, 95% CI 1.7–3.7). Syncope during exercise was not a differentiating trigger, but syncope right after exercise was (P < 0.001). The frequencies of other triggers were similar in both the groups (Figure 2).
Frequency of triggers/circumstances in LQTS patients and in vasovagal patients
LQTS patients (n = 32)a
Vasovagal patients (n = 69)b
LR (95% CI)
PPV for having LQTS (range)
Associated with exercise
Bad night rest
Situational (micturition, defaecation, coughing)
Turning of the head
LR, likelihood ratio. Values in bold face are statistically significant.
aIn two patients data of at least one trigger are missing.
bIn two patients data of two triggers are missing.
Syncope upon standing was also more frequently reported by vasovagal patients (LQTS vs. VVS: LR 0.20, 95% CI 0.10–0.39) than in LQTS patients. Significantly more VVS patients reported having had syncope during venipuncture (LQTS vs. VVS: LR 0.85, 95% CI 0.75–0.97).
Usefulness in daily practice
The clinical usefulness of triggers in identifying high-risk patients for LQTS is given by the probability that a patient with a specific trigger has LQTS, i.e. the PPV.
The PPV of the symptom palpitations could lie between 0.07 and 0.15 depending on whether the prevalence of LQTS is closer to one in 1240 or one in 3100 syncope patients. A PPV of 0.07 indicates that, on an average, seven out of 100 patients presenting with palpitations before or during syncope will have LQTS. Syncope in the supine position has the highest PPV for LQTS ranging from 0.28 to 0.51 (Tables 3 and 4).
The presence of one or more of the three most important triggers (supine position, emotion, and exercise) is 2.3 times (95% CI 1.8–2.8; P < 0.001) more common among LQTS patients than among ED controls. The combination of syncope in supine position and a positive family history results in a very high LR of 41 (95% CI 10–163), and a PPV which varies between 0.57 and 0.78. Thus, on an average 57–78 out of 100 patients experiencing a syncopal episode in supine position with a positive family history for syncope or sudden death could have LQTS.
We examined the discriminative value of information derived from medical history-taking to identify patients at higher risk for LQTS within an unselected group of syncopal patients by focusing on family history, prodromal symptoms, and circumstances.
A family history of syncope, sudden death, and cardiovascular disease was more common in LQTS patients. Syncope in the supine position was more common in LQTS patients, as was syncope in association with exercise, emotional stress, and pain. Syncope upon standing was more often reported by ED and vasovagal patients. Most symptoms during syncope were comparable in LQTS and ED or vasovagal patients, except for palpitations that was reported more often in the LQTS group. The clear importance of these findings is that they can trigger physicians in general practice and at the ED to the potentially life-threatening illness of LQTS, and act accordingly by ordering an ECG, and carefully evaluating the QT interval in all such patients.
Although the age at first syncope was lower in this LQTS group than in the ED patients, LQTS also has to be considered in somewhat older patients. Goldenberg et al.16 showed recently that LQTS patients maintain a high risk for life-threatening events after the age of 40.
The first important feature is the family history in LQTS patients with regard to syncopal events, as well as sudden cardiac death and cardiovascular disease. Sudden cardiac death below the age of 40 is generally accepted as related to LQTS when the diagnosis is proven in families.17 Twenty-one LQTS patients (66%) reported sudden cardiac death in the family vs. 12 of the ED patients (10%). None of the patients in the ED group reported familial sudden death at a young age compared with 20 of the 21 LQTS patients. The high frequency of sudden deaths at young age in the LQTS group makes it highly likely that it is indeed related to the syndrome. This finding confirms one of the diagnostic criteria which has been established for diagnosing LQTS.7
The discriminative value of a positive family history has to be interpreted carefully. Families, in which a cardiac death occurs at a young age, will be more extensively evaluated for LQTS than families without a sudden cardiac death. This might lead to more case finding of LQTS in families with a positive family history than in families with a negative history, and therefore might overestimate its importance.
The overall description by LQTS patients of their events is remarkably similar to that of patients with vasovagal syncope. We can think of two possible explanations. First, episodes of torsades de pointes and less severe ventricular arrhythmias (as occurring in LQTS patients) can produce symptoms such as light-headedness, blurred vision, and loss of consciousness. These symptoms are similar to the well-known prodromi of vasovagal episodes, although the period of prodromal symptoms might be shorter in ventricular arrhythmias. Our data confirm the findings of Alboni et al.18 that palpitations were a predictor of cardiac syncope. Secondly, owing to the high prevalence of vasovagal syncope in the population, LQTS patients may also suffer from vasovagal episodes. Hermosillo et al.19 suggested a possible susceptibility of LQTS patients for neurally mediated syncope, but this study was performed on a very small group of patients without a control group. The high frequency of syncopal episodes during rest in LQTS patients, which are rare in vasovagal syncope, makes it unlikely that episodes in the LQTS group are solely reflex-mediated.
Four types of triggers were more common in LQTS patients. These were syncope during sleep or in the lying position, syncope associated with exercise, syncope in the setting of emotional stress, and syncope on turning of the head. Reliable data on episodes during sleep are difficult to obtain; some patients described episodes of dizziness or losing consciousness after waking up. Others blacked out after turning in bed. In 26 patients (81%) episodes were witnessed by others. Ventricular arrhythmias do occur during sleep, illustrated by the number of sudden deaths in LQTS families that occurred during sleep. Schwartz et al.12 found that 49% and 64% of the lethal events occurred during rest/sleep in LQT2 and LQT3 patients, respectively.
A comparable association of syncope during intense emotions in LQTS patients was observed in the study of Moss et al.6 Among LQTS patients they found that 47% had their episode during emotional circumstances; other circumstances included exercise in 41%, on awakening in 19%, during swimming in 15%, and during auditory stimuli in 8%.6 Schwartz et al.12 also reported that events associated with exercise, emotional stress, and at sleep/rest were important triggers. In our study swimming was a trigger in only 9% and auditory stimuli in 37% of the LQTS patients.
Usefulness in daily clinical practice
We have shown that a family history of sudden cardiac death is a very important historical feature that should alarm physicians. Furthermore, we have shown that some triggers are indeed useful in deciding which syncopal patients need further cardiac evaluation, in particular when syncope occurred in the supine position, during exercise or from any form of emotion (from pain to startle). On these patients, the physician should be alerted to the possibility of cardiac syncope, and importantly should consider LQTS.
Although an ECG was advised in the FAST study, it was performed in only 54% of the ED patients. This illustrates that physicians tend to omit performing an ECG in young patients with a suspected diagnosis of vasovagal syncope. Indeed the Guidelines on Management of Syncope suggests that further examination is unnecessary in most young patients without heart disease and if a definite diagnosis of reflex syncope can be made based on the history.11 As stated above, vasovagal episodes can present in a similar manner to arrhythmic syncope. Hence, an ECG with special interest in the QT interval, is of importance in all patients. Calculating the QTc (QT/√RR=QT interval corrected for the heart rate) though might be complicated even for a cardiologist, when in doubt the opinion of an expert with special interest in the field of LQTS should be asked.20
Strengths and limitations of the study
Family history, prodromal symptoms and circumstances, have been generally accepted as predictive for diagnosing syncopal patients, but have never been tested in two comparative groups as performed in this study.
The validity of our estimates of sensitivity and specificity depends on the selection of cases (LQTS patients) and controls (ED patients). Diagnostic case–control studies are susceptible to overestimation of diagnostic accuracy, particularly when healthy volunteers serve as the control group.21 We used a well-defined group of patients to serve as our control group: consecutive patients presenting with syncope in the ED. This group may be more complex than a typical reflex syncope group visiting the general practitioner. Our control group is therefore less likely to lead to over-optimistic measures of accuracy.
Because some of the LQTS patients filled in the questionnaire years after the first presentation, this might have caused a recall bias as far as the description of the episodes is concerned, although in LQTS, syncope is often a dramatic experience, and thus the impact is not likely to have been forgotten.19
A limitation of the study is that the patients in the ED were questioned by the attending physician at the moment of presentation and the LQTS patients filled in the questionnaire by themselves. In spite of the difficulties with self reports, the answers are comparable with what is known from the literature.22
A final limitation might be that in the LQTS patients, genotyping was used as a gold standard. The ED patients were diagnosed at first contact based on history, physical examination, and ECG alone. This may have caused a selection bias. In the FAST trial, those initial diagnoses were checked on accuracy through an average of 2-year follow-up. This follow-up was used as the gold standard for the control groups.15
We conclude that a positive family history for sudden cardiac death, palpitations prior to syncope, syncopal episodes in supine position, and syncopal episodes related to exercise and emotional stress are more common among patients with LQTS. These questions should be addressed in every syncopal patient and patients with these characteristics should receive further cardiac testing. Although an ECG is an inexpensive, relatively easy additional test that can be used for risk-stratifying, it is often not performed at the ED, especially in young patients. An ECG should be performed in all patients with syncopal episodes. If patients report high-risk triggers and have a positive family history for syncope and/or sudden death, this should prompt the physician to pay extra attention to the corrected QT interval or ask for an expert opinion.
This work was supported by a grant (NHS 99.181) from the Netherlands Heart Foundation, the Hague, the Netherlands.
. High diagnostic yield and accuracy of history, physical examination and ECG in patients with transient loss of consciousness in FAST: the Fainting Assessment Study. J Cardiovasc Electrophysiol 2008;19:48-54.
NancyColman, AnnemiekeBakker, MarkLinzer, Johannes B.Reitsma, WouterWieling, Arthur A.M.WildeEuropace(2009)11 (7):
937-943DOI: http://dx.doi.org/10.1093/europace/eup101First published online: 29 May 2009 (7 pages)