© 2004 by European Society of Cardiology
ESC GUIDELINES
Guidelines on Management (diagnosis and treatment) of syncope – update 2004*
The Task Force on Syncope, European Society of Cardiology,**
Manuscript submitted 25 August 2004. Accepted after revision 25 August 2004.
Key Words: Syncope, Guidelines, European Society of Cardiology
Preamble
Guidelines and Expert Consensus documents aim to present all the relevant evidence on a particular issue in order to help physicians to weigh the benefits and risks of a particular diagnostic or therapeutic procedure. They should be helpful in everyday clinical decision-making.
A great number of Guidelines and Expert Consensus Documents have been issued in recent years by the European Society of Cardiology (ESC) and by different organisations and other related societies. This profusion can put at stake the authority and validity of guidelines, which can only be guaranteed if they have been developed by an unquestionable decision-making process. This is one of the reasons why the ESC and others have issued recommendations for formulating and issuing Guidelines and Expert Consensus Documents.
In spite of the fact that standards for issuing good quality Guidelines and Expert Consensus Documents are well defined, recent surveys of Guidelines and Expert Consensus Documents published in peer-reviewed journals between 1985 and 1998 have shown that methodological standards were not complied with in the vast majority of cases. It is therefore of great importance that guidelines and recommendations are presented in formats that are easily interpreted. Subsequently, their implementation programmes must also be well conducted. Attempts have been made to determine whether guidelines improve the quality of clinical practice and the utilisation of health resources.
The ESC Committee for Practice Guidelines (CPG) supervises and coordinates the preparation of new Guidelines and Expert Consensus Documents produced by Task Forces, expert groups or consensus panels. The chosen experts in these writing panels are asked to provide disclosure statements of all relationships they may have which might be perceived as real or potential conflicts of interest. These disclosure forms are kept on file at the European Heart House, headquarters of the ESC. The Committee is also responsible for the endorsement of these Guidelines and Expert Consensus Documents or statements.
The Task Force has classified and ranked the usefulness or efficacy of the recommended procedure and/or treatments and the Level of Evidence as indicated in the tables below:
|
|
Scope of the document
The purpose of this document is to provide specific recommendations on the diagnostic evaluation and management of syncope. The document is divided into four parts: (1) classification, epidemiology and prognosis; (2) diagnosis; (3) treatment; and (4) special issues in evaluating patients with syncope. Each part reviews background information and summarizes the relevant literature. The details of pathophysiology and mechanisms of various aetiologies were considered to lie outside the scope of this document. Although the document encompasses many of the important aspects of syncope, the panel recommendations focused on the following main questions:
- What are the diagnostic criteria for the causes of syncope?
- What is the preferred approach to the diagnostic work-up in various subgroups of patients with syncope?
- How should patients with syncope be risk stratified?
- When should patients with syncope be hospitalized?
- Which treatments are likely to be effective in preventing syncopal recurrences?
Method
The methodology for writing this document consisted of literature reviews and consensus development by the panel assembled by the European Society of Cardiology. The European Society of Cardiology guidelines for the management (diagnosis and treatment) of syncope were published in August 2001 [1]
. Since then, numerous clinical trials and observational studies have been published or presented, some of which alter the recommendations made in the original document.
Therefore, the Task Force on Syncope of the European Society of Cardiology met in August 2002 and developed a comprehensive outline of the issues that needed to be revised in the document. Subgroups of the panel were formed and each was assigned the task of reviewing the literature on specific topics and of developing a draft summarizing the issue. Each subgroup was to perform literature searches on MEDLINE and to supplement the search by documents from their personal collections. The panel reconvened in September 2003, reviewed the draft documents, made revisions whenever appropriate and developed the consensus recommendations. The panel discussed each recommendation and arrived at consensus by obtaining a majority vote. When there was divergence of opinion, this was noted. Since the goal of the project was to provide specific recommendations for diagnosis and management, guidelines are provided even when the data from the literature was not definitive. It must be pointed out that most of the recommendations are based on consensus expert opinion. All the members of the panel reviewed final drafts and approved the final document.
With respect to the initial document, the following sections (and recommendations) were widely revised in the 2004 update:
- Classification of transient loss of consciousness
- Epidemiological and prognostic considerations
- Initial evaluation and diagnostic flow
- Prolonged electrocardiographic monitoring
- Electrophysiological study
- ATP test
- Ventricular signal averaged electrocardiogram, T wave alternans
- Exercise testing
- Neurological and psychiatric evaluation
- Treatment of neurally-mediated (reflex) syncope
- Syncope in the older adult
- Syncope in paediatric patients
- Driving and syncope
- Glossary of terms.
Furthermore, since the strategies for the assessment of syncope vary widely among physicians and among hospitals in Europe, we recognised the need to coordinate the evaluation of syncope. We sought to define ESC standards for the management of syncope and we proposed a model of organization for the evaluation of the syncope patient. A new section was thus added to the document on this topic.
A major issue in the use of diagnostic tests is that syncope is a transient symptom and not a disease. Typically patients are asymptomatic at the time of evaluation and the opportunity to capture a spontaneous event during diagnostic testing is rare. As a result, the diagnostic evaluation has focused on physiological states that could cause loss of consciousness. This type of reasoning leads, of necessity, to uncertainty in establishing a cause. In other words, the causal relationship between a diagnostic abnormality and syncope in a given patient is often presumptive. Uncertainty is further compounded by the fact that there is a great deal of variation in how physicians take a history and perform a physical examination, the types of tests requested and how they are interpreted. These issues make the diagnostic evaluation of syncope inordinately difficult. Consequently, there is a need for specific criteria for diagnosis from history and physical examination, and clear-cut guidelines for choosing tests, test abnormalities and how to use the results in establishing a cause of syncope. This document has tried to provide specific criteria by using the literature as well as a consensus of the panel.
A further concern about tests used for evaluation of the aetiology of syncope is that measurements of test sensitivity are not possible because of a lack of reference or gold standard for most of the tests employed for this condition. Since syncope is an episodic symptom, a reference standard could be an abnormality observed during a spontaneous event. This is possible, for instance, if syncope occurred concurrently with an arrhythmia detected by an implantable loop recorder. However, most of the time decisions have to be made based on the patient's history or abnormal findings during asymptomatic periods. To overcome the lack of a gold standard, the diagnostic yield of many tests in syncope has been assessed indirectly by evaluating the reduction of syncopal recurrences after administration of the specific therapy suggested by the results of the tests which were diagnostic.
A major problem with the literature on syncope is that it has been defined variably in the past. Definitions did not always include a restriction that transient loss of consciousness should be due to cerebral hypoperfusion, with the result that many other causes of loss of consciousness could also be interpreted as ‘syncope’, including concussion, vertebrobasilar TIAs and epilepsy, or even stroke, hardly ever even associated with unconsciousness. In many cases it was not clear whether ‘syncope’ was interpreted in this very wide sense, here ‘transient loss of consciousness’ (TLOC), or was used in the restricted sense as in these guidelines. Obviously, mixing epilepsy, concussion or even stroke with syncope proper seriously detracts from the value of epidemiological and prognostic estimates.
The literature on syncope testing is largely composed of case series, cohort studies, or retrospective analyses of already existing data. The impact of testing on guiding therapy and reducing syncopal recurrences is difficult to discern from these methods of research without randomization and blinding. Because of these issues, the panel performed full reviews of the literature for diagnostic tests but did not use predefined criteria for selection of articles to be reviewed. Additionally, the panel did not feel that an evidence-based summary of the literature was possible.
In assessing treatment of syncope, this document reviews the few randomised-controlled trials that have been reported. For various diseases and disorders with known treatments (e.g., orthostatic hypotension, sick sinus syndrome) those therapies are reviewed and recommendations are modified for patients with syncope. Most studies of treatment have used a non-randomized design and many even lack a control group. The interpretation of these studies is very difficult but their results were used in summary recommendations of treatment.
The strength of recommendations has been ranked as follows:
- Class I, when there is evidence for and/or general agreement that the procedure or treatment is useful. Class I recommendations are generally those reported in the sections labelled as ‘Recommendations’ and in the tables.
- Class II, when usefulness of the procedure or treatment is less well established or divergence of opinion exists among the members of the Task Force.
- Class III, when the procedure or treatment is not useful and in some cases may be harmful.
The strength of evidence supporting a particular procedure/treatment option has been ranked as follows:
- Level of Evidence A = Data derived from multiple randomized clinical trials or meta-analyses
- Level of Evidence B = Data derived from a single randomized trial or multiple non-randomized studies
- Level of Evidence C = Consensus Opinion of experts.
When not expressed otherwise, evidence is of type C
Part 1. Classification, epidemiology and prognosis
Definition
Syncope (derived from the Greek words, ‘syn’ meaning ‘with’ and the verb ‘kopto’ meaning ‘I cut’ or more appropriately in this case ‘I interrupt’) is a symptom, defined as a transient, self-limited loss of consciousness, usually leading to falling. The onset of syncope is relatively rapid, and the subsequent recovery is spontaneous, complete, and usually prompt [2–5]. The underlying mechanism is a transient global cerebral hypoperfusion.
In some forms of syncope there may be a premonitory period in which various symptoms (e.g., light-headedness, nausea, sweating, weakness, and visual disturbances) offer warning of an impending syncopal event. Often, however, loss of consciousness occurs without warning. Recovery from syncope is usually accompanied by almost immediate restoration of appropriate behaviour and orientation. Retrograde amnesia, although believed to be uncommon, may be more frequent than previously thought, particularly in older individuals. Sometimes the post-recovery period may be marked by fatigue.
An accurate estimate of the duration of syncope episodes is rarely obtained. However, typical syncopal episodes are brief. Complete loss of consciousness in vasovagal syncope is usually no longer than 20 s in duration. In one videometric study of 56 episodes of shortlasting severe cerebral hypoxia in adolescents induced by instantaneous deep fall in systemic pressure using the ‘mess trick’ syncope occurred in all without any premonitory symptoms and myoclonic jerks were present in 90 percent; the syncope duration averaged 12 s (range 5–22) [4]. However, rarely syncope duration may be longer, even lasting for several minutes. In such cases, the differential diagnosis between syncope and other causes of loss of consciousness can be difficult [5].
Pre-syncope or near-syncope refers to a condition in which patients feel as though syncope is imminent. Symptoms associated with pre-syncope may be relatively non-specific (e.g., ‘dizziness’), and tend to overlap with those associated with the premonitory phase of true syncope described earlier.
Brief overview of pathophysiology of syncope
Specific factors resulting in syncope vary from patient-to-patient, but several general principles are worthy of note.
In healthy younger individuals with cerebral blood flow in the range of 50–60 ml/100 g tissue/min – that represents about 12–15 percent of resting cardiac output – minimum oxygen requirements necessary to sustain consciousness (approximately 3.0–3.5 ml O2/100 g tissue/min) are easily achieved [6]. However, in older individuals, or those with underlying disease conditions, the safety margin for oxygen delivery may be more tenuous [7,8].
Cerebral perfusion pressure is largely dependent on systemic arterial pressure. Thus, any factor that decreases either cardiac output or total peripheral vascular resistance diminishes systemic arterial pressure and cerebral perfusion pressure [9]. In regard to cardiac output, the most important physiological determinant is venous filling (pre-load). Therefore, excessive pooling of blood in dependent parts of the body or diminished blood volume may predispose to syncope. Cardiac output may also be impaired due to bradyarrhythmias, tachyarrhythmias, or valvular disease. In terms of peripheral vascular resistance, widespread and excessive vasodilatation may play a critical role in decreasing arterial pressure (a main cause of fainting in the reflex syncopal syndromes). Vasodilatation also occurs during thermal stress. Impaired capacity to increase vascular tone during standing is the cause of orthostatic hypotension and syncope in patients taking vasoactive drugs and in patients with autonomic neuropathies [10]. Cerebral hypoperfusion may also result from an abnormally high cerebral vascular resistance. Low carbon dioxide tension is probably the main cause, but sometimes the cause remains unknown.
A sudden cessation of cerebral blood flow for 6–8 s has been shown to be sufficient to cause complete loss of consciousness [2]. Experience from tilt testing showed that a decrease in systolic blood pressure to 60 mmHg is associated with syncope [11]. Further, it has been estimated that as little as a 20 percent drop in cerebral oxygen delivery is sufficient to cause loss of consciousness [2]. In this regard, the integrity of a number of control mechanisms is crucial for maintaining adequate cerebral oxygen delivery, including: (a) cerebrovascular ‘autoregulatory’ capability, which permits cerebral blood flow to be maintained over a relatively wide range of perfusion pressures; (b) local metabolic and chemical control which permits cerebral vasodilatation to occur in the presence of either diminished pO2 or elevated pC02; (c) arterial baroreceptor-induced adjustments of heart rate, cardiac contractility, and systemic vascular resistance, which modify systemic circulatory dynamics in order to protect cerebral flow; (d) and vascular volume regulation, in which renal and hormonal influences help to maintain central circulating volume.
Whatever the mechanism, a transient global cerebral hypoperfusion to critical values induces a syncopal episode. Risk of failure is greatest in the elderly or critically ill patients [5,11]. Ageing alone has been associated with diminution of cerebral blood flow [7]. Additionally, certain common disease states may diminish cerebral blood flow protection. For example, hypertension has been associated with a shift of the autoregulatory range to higher pressures, while diabetes alters the chemoreceptor responsiveness of the cerebrovascular bed [8].
Classification
Syncope must be differentiated from other ‘non-syncopal’ conditions associated with real or apparent transient loss of consciousness (Fig. 1). Table 1.1 and Table 1.2 provide a pathophysiological classification of the principal known causes of transient loss of consciousness. The subdivision of syncope is based on pathophysiology as follows:
|
|
|
- ‘Neurally-mediated (reflex) syncope’ refers to a reflex response that, when triggered, gives rise to vasodilatation and/or bradycardia; however the contribution of each of these two factors to systemic hypotension and cerebral hypoperfusion may differ considerably. The triggering events might vary considerably in individual patients. The ‘classical vasovagal syncope’ is mediated by emotional or orthostatic stress and can be diagnosed by history taking. ‘Carotid sinus syncope’ is defined as syncope which, by history, seems to occur in close relationship to accidental mechanical manipulation of the carotid sinuses, and which can be reproduced by carotid sinus massage. ‘Situational syncope’ refers to those forms of neurally-mediated syncope associated with specific scenarios (e.g., micturition, coughing, defaecating, etc.). Often, however, neurally-mediated reflex syncopes have ‘non-classical’ presentations. These forms are diagnosed by minor clinical criteria, exclusion of other causes for syncope (absence of structural heart disease) and positive response to tilt testing or carotid sinus massage. Examples of non-classical vasovagal syncope include episodes without clear triggering events or premonitory symptoms.
- ‘Orthostatic hypotension’ refers to syncope in which the upright position (most often the movement from sitting or lying to an upright position) causes arterial hypotension. This occurs when the autonomic nervous system is incapacitated and fails to respond to the challenges imposed by upright posture. A second major cause is ‘volume depletion’ in which the autonomic nervous system is itself not deranged, but is unable to maintain blood pressure due to decreased circulating volume. Note that vasovagal syncope can also be provoked by standing (e.g., soldiers fainting on parade), but these events are grouped under ‘neurally-mediated (reflex) syncope’.
- ‘Cardiac arrhythmias’ can cause a decrease in cardiac output, which usually occurs irrespectively of circulatory demands.
- ‘Structural heart disease’ can cause syncope when circulatory demands outweigh the impaired ability of the heart to increase its output.
- ‘Steal’ syndromes can cause syncope when a blood vessel has to supply both part of the brain and an arm.
Several disorders may resemble syncope in two different ways. In some, consciousness is truly lost, but the mechanism is something other than cerebral hypoperfusion. Examples are epilepsy, several metabolic disorders (including hypoxia and, hypoglycaemia) and intoxications. In several other disorders, consciousness is only apparently lost; this is the case in ‘psychogenic pseudo-syncope’, cataplexy and drop attacks. In psychogenic pseudo-syncope patients may pretend to be unconscious when they are not. This condition can be seen in the context of factitious disorders, malingering and conversion. Finally, some patients may voluntarily trigger true syncope in themselves to attract attention, as a game, or to obtain some other advantage. Table 1.2 lists the most common conditions misdiagnosed as the cause of syncope. A differentiation such as this is important because the clinician is usually confronted with patients with sudden loss of consciousness (real or apparent) which may be due to causes not associated with decreased cerebral blood flow such as seizure and/or conversion reaction.
Note that the term ‘syncope’ should only be used when cerebral hypoperfusion was at least likely to have caused the loss of consciousness. If this is not the case, and no other cause is at least probable, then the term ‘transient loss of consciousness’ should be used to avoid limiting the scope of diagnostic thinking (see the ‘glossary of terms’).
A major limitation of this classification is the fact that more than one pathophysiological factor may contribute to the symptoms. For instance, in the setting of valvular aortic stenosis or left ventricular outflow tract obstruction, syncope is not solely the result of restricted cardiac output, but may be in part due to inappropriate neurally mediated reflex vasodilation and/or primary cardiac arrhythmias [12]. Similarly, a neural reflex component (preventing or delaying vasoconstrictor compensation) appears to play an important role when syncope occurs in association with certain brady- and tachyarrhythmias [13–15].
Epidemiological considerations
Numerous studies have examined epidemiological aspects of syncope and delineated the multiple potential causes of syncope. However, some reports have focused on relatively select populations such as the military or tertiary care medical centres or solitary medical practices. For example, a survey of 3000 United States Air Force personnel (average age 29 years) revealed that 27% had experienced a syncopal spell during their lifetime [16]. Application of these findings to medical practice is limited not only by the nature of the environment in which patients were enroled, but also the variable manner in which symptoms were evaluated.
In terms of studies examining a broad population sample, the Framingham Study (in which biennial examinations were carried out from 1971 to 1998 in 7814 free-living individuals) 822 (10.5%) reported at least one syncopal event during an average of 17 years [17]. The incidence of a first report of syncope was 6.2 per 1000 person-years. Assuming a constant incidence rate over time, the authors calculated a 10-year cumulative incidence of syncope of 6%; this means a 42% prevalence of syncope during the life of a person living 70 years. However, the incidence was not constant, but increased more rapidly starting at the age of 70 years. Indeed, it was 11 per 1000 person-years for both men and women at the age of 70–79 and 17 per 1000 person-years for men and 19 per 1000 person-years for women at the age 
80.
Among the elderly confined to long-term care institutions, the annual incidence may be as high as 6% with a recurrence rate of 30% [18]. Several reports indicate that syncope is common presenting problem in the health care settings accounting for 3% to 5% of emergency room visits and 1% to 3% of hospital admissions [19–21]. Other studies in specific populations provide insight into the relative frequency with which syncope may occur in certain settings. Several of these reports may be summarized as follows:
- 15% of children before the age of 18 [22]
- 25% of a military population aged 17–26 [23]
- 20% of air force personnel aged 17–46 [24]
- 39% of young medical students (median age of 21 years) with a prevalence in females twice as high than in males [25]
- 16% during a 10-year period in men aged 40–59 [26]
- 19% during a 10-year period in women aged 40–49 [26]
- 23% during a 10-year period in elderly people (age>70) [18].
However, the majority of these individuals probably do not seek medical evaluation.
In summary, even if some variability in prevalence and incidence of syncope is reported, the majority of studies suggest that syncope is a common problem in the community, long-term care institutions, and in health care delivery settings.
Prognostic evaluation
Mortality
In the Framingham study [17], the participants with syncope from any cause, compared with those who did not have syncope, had 1.31 increased risk of death from any cause, 1.27 for non-fatal myocardial infarction or death from coronary heart disease, and 1.06 for fatal or non-fatal stroke. Patients with cardiac syncope had the highest risk of death from any cause (hazard ratio of 2.1) and cardiovascular events (hazard ratio of 2.66). Studies in the 1980s showed that one-year mortality of patients with cardiac syncope was consistently higher (ranging between 18 and 33%) than patients with non-cardiac cause (0–12%) or unexplained syncope (6%) [19,20,27–29]. One-year incidence of sudden death was 24% in patients with a cardiac cause compared with 3–4% in the other two groups [28,29]. When adjustments were made for differences in baseline rates of heart and other diseases, cardiac syncope was still an independent predictor of mortality and sudden death [28,29]. However, a more recent study [30] directly compared the outcomes of patients with syncope with matched control subjects without syncope. Although patients with cardiac syncope had higher mortality rates compared with those of non-cardiac or unknown causes, patients with cardiac causes did not have a higher mortality when compared with their matched controls that had similar degrees of heart disease. This study showed that presence of structural heart disease was the most important predictor of mortality. In a selected population of patients with advanced heart failure and a mean ejection fraction of 20%, the patients with syncope had a higher risk of sudden death (45% at 1 year) than those without (12% at 1 year); admittedly, the risk of sudden death was similarly high in patients with either supposed cardiac syncope or syncope from other causes [31].
Structural heart disease is a major risk factor for sudden death and overall mortality in patients with syncope. The association of syncope with aortic stenosis has long been recognised as having an average survival without valve replacement of 2 years [32]. Similarly, in hypertrophic cardiomyopathy, the combination of young age, syncope at diagnosis, severe dyspnoea and a family history of sudden death best predicted sudden death [33]. In arrhythmogenic right ventricular dysplasia, patients with syncope or symptomatic ventricular tachycardia have a similarly poor prognosis [34]. Patients with ventricular tachyarrhythmias have higher rates of mortality and sudden death but the excess mortality rates depend on underlying heart disease; patients with severe ventricular dysfunction have the worst prognosis [35]. Some of the cardiac causes of syncope do not appear to be associated with increased mortality. These include most types of supraventricular tachycardias and sick sinus syndrome.
A number of subgroups of patients who have an excellent prognosis can be identified:
- Young healthy individuals without heart disease and normal ECG. The 1-year mortality and sudden death rates in young patients (less than 45 years of age) without heart disease and normal ECG is low [36]. Although comparisons have not been made with age and sex matched controls, there is no evidence that these patients have an increased mortality risk. Many of these patients have neurally mediated syncope or unexplained syncope.
- Neurally mediated syncope. A large number of cohort studies in which the diagnosis has been established using tilt testing show that the mortality at follow-up of patients with neurally mediated syncope is near 0% [37]. Most of these patients had normal hearts. None of these studies report patients who died suddenly. In the Framingham study [17], there was no increased risk of cardiovascular morbidity or mortality associated with vasovagal (including orthostatic and medication-related) syncope during an average 17-year follow-up.
- Orthostatic hypotension. The mortality rates of patients with orthostatic hypotension depend on the causes of this disorder. Some causes (e.g., volume depletion, drug induced) are transient problems that respond to treatment and do not have long-term consequences. Other diseases causing primary and secondary autonomic failure have long-term consequences and may potentially increase mortality depending on the severity of the underlying disease. In elderly patients with orthostatic hypotension, the prognosis is largely determined by co-morbid illnesses.
Syncope of unknown cause is a heterogeneous group at intermediate risk. The definition of unexplained syncope largely depends on the diagnostic accuracy employed. An approximately 5% first year mortality in patients with unexplained syncope has been a relatively consistent observation in the literature [19,20,28,29,38]. As referred to earlier [17], the participants with syncope of unknown cause, compared with those who did not have syncope, had 1.31 increased risk of death. The group of patients with unknown syncope, in general, is heterogeneous probably including patients with a benign cause of syncope as well as those with an undiagnosed cardiac cause. The consequence is that this group shows an intermediate risk between the cardiac and neurally-mediated groups. The remaining challenge is how to identify those who are at high risk of death although the presence or the absence of structural heart disease seems to be an important determinant of risk. Although the mortality is largely due to underlying co-morbidity, such patients continue to be at risk of physical injury, and may encounter employment and life-style restrictions.
Recurrences
Approximately 35% of patients have recurrences of syncope at 3 years of follow-up; 82% of recurrences occur within the first 2 years [29,39]. Predictors of recurrence of syncope include having had recurrent syncope at the time of presentation (four or more episodes in one study [39]) or a psychiatric diagnosis [39–41]. In one study [42], more than five lifetime episodes gave a 50% chance of recurrence in the following year. In another study [40], age 45 years was also associated with higher rates of syncopal recurrence after controlling for other risk factors. After positive tilt table testing the patients with more than six syncopal spells had a risk of recurrence of >50% over 2 years [43].
Recurrences are not associated with increased mortality or sudden death rates, but patients with recurrent syncope have a poor functional status similar to patients with other chronic diseases.
Risk stratification
One study has developed and validated a clinical prediction rule for risk stratification of patients with syncope [36]. This study used a composite outcome of having cardiac arrhythmias as a cause of syncope or death (or cardiac death) within 1 year of follow-up. Four variables were identified and included age 45 years, history of congestive heart failure, history of ventricular arrhythmias, and abnormal ECG (other than non-specific ST changes). Arrhythmias or death within 1 year occurred in 4–7% of patients without any of the risk factors and progressively increased to 58–80% in patients with three or more factors [34]. The critical importance of identifying cardiac causes of syncope is that many of the arrhythmias and other cardiac diseases are now treatable with drugs and/or devices.
Physical injury
Syncope may result in injury to the patient or to others such as may occur when a patient is driving. Major morbidity such as fractures and motor vehicle accidents were reported in 6% of patients and minor injury such as laceration and bruises in 29%. There are no data on the risk of injury to others. Recurrent syncope is associated with fractures and soft-tissue injury in 12% of patients [39].
Quality of life
A study that evaluated the impact of recurrent syncope on quality of life in 62 patients used the Sickness Impact Profile and found functional impairment similar to chronic illnesses such rheumatoid arthritis, low back pain, and psychiatric disorders [44]. Another study on 136 patients with unexplained syncope found impairment on all five dimensions measured by the EQ-5D instrument, namely Mobility, Usual activities, Self-care, Pain/discomfort, Anxiety/depression. Furthermore there was a significant negative relationship between frequency of spells and overall perception of health [42].
Economic implications
Patients with syncope are often admitted to hospital and undergo expensive and repeated investigations, many of which do not provide a definite diagnosis. In a study, performed in 1982, patients often underwent multiple diagnostic tests despite which a cause of syncope was established in only 13 of 121 patients [45]. With the advent of newer diagnostic tests (e.g., tilt testing, wider use of electrophysiological testing, loop monitoring) it is likely that patients are undergoing a greater number of tests at considerably higher cost. In a recent study, based on administrative data from Medicare, there were estimated to be 193,164 syncope hospital discharges in 1993 in the USA [46]. The cost per discharge was calculated to $4132 and increased to $5281 for those patients who were readmitted for recurrent syncope. This figure underestimates the true total cost associated with syncope because many patients with syncope are not admitted to hospital for either investigation or therapy. In the UK [47] the overall cost per patient was £611, with 74% attributed to the costs of hospital stay alone. Cost per diagnosis of patients admitted to hospital was £1080.
Part 2. Diagnosis
The diagnostic strategy based on the initial evaluation
Fig. 2 shows a flow diagram of an approach to the evaluation of transient loss of consciousness (TLOC).
|
Initial evaluation
The starting point for the evaluation of syncope is a careful history and physical examination including orthostatic blood pressure measurements. In most young patients without heart disease a definite diagnosis of neurally mediated syncope can be made without any further examination. Apart from this case, a 12-lead ECG should usually be part of the general evaluation of patients. This basic assessment will be defined as Initial evaluation.
Three key questions should be addressed during the initial evaluation:
- Is loss of consciousness attributable to syncope or not?
- Is heart disease present or absent?
- Are there important clinical features in the history that suggest the diagnosis?
Differentiating true syncope from other ‘non-syncopal’ conditions associated with real or apparent transient loss of consciousness (TLOC) is generally the first diagnostic challenge and influences the subsequent diagnostic strategy (see classification in Part 1 and Table 2.1). The symptoms surrounding the loss of consciousness accurately discriminate between seizures and syncope [48]. Apart from the prognostic importance of the presence of heart disease (see Part 1, prognostic stratification), its absence excludes a cardiac cause of syncope with few exceptions. In a recent study [49], heart disease was an independent predictor of cardiac cause of syncope, with a sensitivity of 95% and a specificity of 45%; by contrast, the absence of heart disease allowed exclusion of a cardiac cause of syncope in 97% of the patients. Finally, accurate history taking alone may be diagnostic of the cause of syncope or may suggest the strategy of evaluation (see Part 2, initial evaluation). It must be pointed out that syncope may be one of the accompanying symptoms which occur at the presentation of certain diseases, such as aortic dissection, pulmonary embolism, acute myocardial infarction, outflow tract obstruction, etc. In these cases, priority must be given to specific and immediate treatment of the underlying condition. Those issues are not addressed by this report.
|
The initial evaluation may lead to certain or suspected diagnosis or no diagnosis (here termed as unexplained syncope).
Certain diagnosis
Initial evaluation may lead to a certain diagnosis based on symptoms, signs or ECG findings. The recommended diagnostic criteria are listed in the section under Initial Evaluation. Under such circumstances, no further evaluation of the disease or disorder may be needed and treatment, if any, can be planned.
Suspected diagnosis
More commonly, the initial evaluation leads to a suspected diagnosis, when one or more of the features listed in Tables 2.2 and Table 2.3 are present. The recommended diagnostic work-up is listed in the section under Initial evaluation. A suspected diagnosis needs to be confirmed by directed testing If a diagnosis is confirmed by specific testing, treatment may be initiated. On the other hand, if the diagnosis is not confirmed, then patients are considered to have unexplained syncope and are evaluated as follows.
|
|
Unexplained syncope
The initial evaluation may lead to no diagnosis (here termed as unexplained syncope). The strategy of evaluation varies according to the severity and frequency of the episodes. In patients with unexplained syncope the likely diagnosis is neurally-mediated. The tests for neurally mediated syncope consist of tilt testing and carotid massage. The majority of patients with single or rare episodes in this category probably have neurally mediated syncope and tests for confirmation are usually not necessary. If it is not clear that it was syncope, the term transient loss of consciousness (TLOC) is preferable and reappraisal is warranted.
Re-appraisal
Once the evaluation, as outlined, is completed and no cause of syncope is determined, re-appraisal of the work-up is needed since subtle findings or new historical information may change the entire differential diagnosis. Re-appraisal may consist of obtaining details of history and re-examining patients as well as review of the entire work-up. If unexplored clues to possible cardiac or neurological disease are apparent, further cardiac and neurological assessment is recommended. In these circumstances, consultation with appropriate speciality services may be needed. An additional consideration is psychiatric illness. Psychiatric assessment is recommended in patients with frequent recurrent syncope who have multiple other somatic complaints and initial evaluation raises concerns for stress, anxiety and possible other psychiatric disorders.
Diagnostic yield and prevalence of causes of syncope
Data from seven population based studies [19,20,27,29,41,50,51] showed that the history and physical examination identified a potential cause of syncope in 726 (45%) of 1607 patients whose primary disorder can be diagnosed. However, the diagnostic criteria for vasovagal syncope, which represent the most frequent cause of loss of consciousness, have been varied among studies. While some studies have used precipitating events for diagnosing vasovagal syncope, others have used only the presence of prodromal symptoms which may lack specificity.
The diagnostic yield of electrocardiography and rhythm recordings obtained in the emergency department is low, ranging between 1% and 11% (mean 7%) [19,27,29,51]. The most common diagnoses included ventricular tachycardia, bradyarrhythmias and, less commonly, acute myocardial infarction.
Similarly routine blood tests (blood count and tests for electrolyte and glucose level) rarely yield diagnostically useful information. They usually confirm a clinical suspicion of hypoglycaemia, when loss of consciousness is associated with confusion, salivation, tremors, hunger, hyperadrenergic state and serum glucose value is <40 mg/dl. Syncope due to acute severe anaemia and bleeding may be diagnosed by clinical features and confirmed by a complete blood count.
The cause of syncope remains unknown despite a complete work-up in a substantial proportion of patients. For example, in five studies [19,27,29,50,51], performed in the 1980s, the cause of syncope could not be determined in 34% of cases (range 13–41%) and in four recent studies [49,52–54] the cause of syncope could not be determined in 20% of cases (range 17.5–26%).
The prevalence of the causes of syncope has been evaluated in six population-based studies of unselected patients [19,27,29,41,50,51] for a total of 1499 patients, performed in the 1980s. The most common cause was neurally-mediated and orthostatic hypotension which accounted for 381 cases (37%). The second most common cause was cardiac which accounted for 246 cases (17%) with a primary arrhythmic mechanism being responsible in 195 (13%). Neurological and psychiatric causes were found in 150 cases (10%). In four recent studies [49,52–54], for a total of 1640 patients, neurally-mediated and orthostatic hypotension accounted for 917 cases (56%). The second most common cause was cardiac which accounted for 233 cases (14%) with a primary arrhythmic mechanism being responsible in 182 (11%). Neurological and psychiatric causes were found in 155 cases (9%). Thus, the percentage of neurally-mediated syncope increased and that of unexplained syncope decreased compared with the older studies. In more recent studies there was a more extensive use of carotid sinus massage and tilt testing [49]. This suggests that when specific tests are used, reflex syncope or autonomic failure are even more frequent and that carotid sinus massage and tilt testing are useful to discover those particular forms when the history alone is not diagnostic.
Initial evaluation
The following section provides specific recommendations about how to use the history, physical examination and ECG for making certain or presumptive diagnoses of syncope.
History and physical examination
The history alone may be diagnostic of the cause of syncope or may suggest the strategy of evaluation. The clinical features of the presentation are most important, especially the factors that might predispose to syncope and its sequelae. Some attempts have been made to validate the diagnostic value of the history in prospective and case-control studies [5,27,41,48,49,55,56].
The important parts of the history are listed in Table 2.1. They are the key features in the diagnostic work-up of patients with syncope. When taking a history, all the items listed in Table 2.1 should be carefully sought.
Apart from being diagnostic, the history may guide the subsequent evaluation strategy. For example, a cardiac cause is more likely when syncope is preceded by palpitations or occurs in the supine position or during exercise. Conversely, a neurally-mediated mechanism is likely when predisposing factors, precipitating events and accompanying symptoms are present and the patient has recurrent syncopal episodes over several years.
Physical findings that are useful in diagnosing syncope include cardiovascular and neurological signs and orthostatic hypotension. For example, the presence of a murmur or severe dyspnoea is indicative of structural heart disease and of a cardiac cause of syncope. Table 2.2 lists how to use the history and physical findings in suggesting various aetiologies.
Baseline electrocardiogram
An initial ECG is most commonly normal in patients with syncope. When abnormal, the ECG may disclose an arrhythmia associated with a high likelihood of syncope, or an abnormality which may predispose to arrhythmia development and syncope. Moreover, any abnormality of the baseline ECG is an independent predictor of cardiac syncope or increased mortality, suggesting the need for pursuing evaluation for cardiac causes in these patients. Equally important, a normal ECG is associated with a low risk of cardiac syncope as the cause, with a few possible exceptions, for example in case of syncope due to a paroxysmal atrial tachyarrhythmia.
Arrhythmias that are considered diagnostic of the cause of syncope are listed below. More commonly, the baseline ECG leads to a suspected cardiac arrhythmia, which needs to be confirmed by directed testing (Table 2.3).
|
|
Echocardiogram
Echocardiography is frequently used as a screening test to detect cardiac disease in patients with syncope. Although numerous published case reports have suggested an important role of echocardiography in disclosing the cause and/or mechanism of syncope, larger studies have shown that the diagnostic yield from echocardiography is low in the absence of clinical, physical or electrocardiographic findings suggestive of a cardiac abnormality [59
–61]. In patients with syncope or pre-syncope and normal physical examination, the most frequent (from 4.6% to 18.5% of cases) finding is mitral valve prolapse [59]. This may be coincidental as both conditions are common. Other cardiac abnormalities include valvular diseases (most frequently aortic stenosis), cardiomyopathies, regional wall motion abnormalities suggestive of myocardial infarction, infiltrative heart diseases such as amyloidosis, cardiac tumours, aneurysms, atrial thromboembolism and other abnormalities [62–66]. Even if echocardiography alone is only seldom diagnostic, this test provides information about the type and severity of underlying heart disease which may be useful for risk stratification. If moderate to severe structural heart disease is found, evaluation is directed toward a cardiac cause of syncope. On the other hand, in the presence of minor structural abnormalities detected by echocardiography, the probability of a cardiac cause of syncope may not be high, and the evaluation may proceed as in patients without structural heart disease. Examples of heart disease in which cardiac syncope is likely include:
- cardiomyopathy with episodes of overt heart failure
- systolic dysfunction (ejection fraction <40%)
- ischaemic cardiomyopathy following an acute myocardial infarction
- right ventricular dysplasia
- hypertrophic cardiomyopathy
- congenital heart diseases
- cardiac tumours
- outflow tract obstruction
- pulmonary embolism
- aortic dissection.
|
Carotid sinus massage
It has long been observed that pressure at the site where common carotid artery bifurcates produces a reflex slowing in heart rate and fall in blood pressure. In some patients with syncope, especially those >40 years, an abnormal response to carotid massage can be observed. A ventricular pause lasting 3 s or more and a fall in systolic blood pressure of 50 mm/Hg or more is considered abnormal and defines the carotid sinus hypersensitivity [67,
68].
The carotid sinus reflex arc is composed of an afferent limb arising from the mechanoreceptors of the carotid artery and terminating in midbrain centres, mainly the vagus nucleus and the vasomotor centre. The efferent limb is via the vagus nerve and the parasympathetic ganglia to the sinus and atrioventricular nodes and via the sympathetic nervous system to the heart and the blood vessels. Whether the site of dysfunction resulting in a hypersensitive response to the massage is central at the level of brainstem nuclei or peripheral at the level of carotid baroreceptors is still a matter of debate [68–70].
Methodology and response to carotid sinus massage
Carotid sinus massage is a tool used to disclose carotid sinus syndrome in patients with syncope.
Protocol
In most studies the carotid sinus massage is performed in the supine position; in others, it is performed in both supine and upright positions (usually on a tilt table). Continuous electrocardiographic monitoring must be used. Continuous blood pressure monitoring, for which a non-invasive measurement device is best suited, should also be used as the vasodepressor response is rapid and cannot be adequately detected with devices which do not measure continuous blood pressure. After baseline measurements, the right carotid artery is firmly massaged for 5–10 s at the anterior margin of the sternocleomastoid muscle at the level of the cricoid cartilage. After one or two minutes a second massage is performed on the opposite side if the massage on one side failed to yield a ‘positive’ result. If an asystolic response is evoked, to assess the contribution of the vasodepressor component (which may otherwise be hidden) massage is usually repeated after intravenous administration of Atropine (1 mg or 0.02 mg/kg body weight). Atropine administration is preferred to temporary dual chamber pacing as it is simple, non-invasive, and easily reproducible [71]. The response to carotid sinus massage is generally classified as cardioinhibitory (i.e., asystole), vasodepressive (fall in systolic blood pressure) or mixed. The mixed response is diagnosed by the association of an asystole of 3 s and a decline in systolic blood pressure of 50 mmHg on rhythm resumption from the baseline value.
There are two widely used methods of carotid sinus massage. In the first method, massage is performed only in the supine position and pressure is applied for no more than 5 s. A positive response is defined as a ventricular pause 3 s and/or a fall in systolic blood pressure 50 mmHg. Pooled data from four studies performed in elderly patients with syncope show a positive response rate of 35% (235 of 663 patients) [72–75].However, abnormal responses are also frequently observed in subjects without syncope. For example, an abnormal response was observed in 17–20% of patients affected by various types of cardiovascular diseases [76], and in 38% of patients with severe narrowing of the carotid arteries [77]. Moreover, the diagnosis may be missed in about one third of cases if only supine massage is performed [78,79].
In the second method, reproduction of spontaneous symptoms is required during carotid massage (80). Eliciting symptoms requires a longer period of massage (10 s) and massage performed in both supine and upright positions [81,82]. A positive response was observed in 49% of 100 patients with syncope of uncertain origin [83] and in 60% of elderly patients with syncope and sinus bradycardia [84], but only in 4% of 101 control patients without syncope pooled from three studies [82–84]. In an intrapatient comparison study [82], the ‘method of symptoms’ appears to carry a higher positivity rate (49% vs 41%) in patients with syncope and a lower positivity rate (5% vs 15%) in patients without syncope than the first method. In a large population of 1719 consecutive patients with syncope unexplained after the initial evaluation (mean age 66±17 years), carotid sinus hypersensitivity was found in 56% and syncope was reproduced in 26% of cases [85]. Among the positive tests, the response was cardioinhibitory in 46%, mixed in 40% and vasodepressor in 14%. The positivity rate increased with age, ranging from 4% in patients <40 years to 41% in patients >80 years. The test was positive only in the upright position in 49% of patients.
Whatever method is used, increasing importance has been given to the execution of the massage also in the upright position, usually using a tilt table [78,79,85,86]. Other than a higher positivity rate compared with supine massage only, the importance of performing upright massage is due to the better possibility of evaluating the magnitude of the vasodepressor component and of reproducing symptoms. Underestimated in the past, actually a vasodepressor component of the reflex is present in most patients with an asystolic response [86]. A correct determination of the vasodepressor component of the reflex is of practical importance for the choice of therapy. Indeed, pacemaker therapy has been shown to be less effective in mixed forms with an important vasodepressor component rather than in dominant cardioinhibitory forms [71,72].
In conclusion, carotid sinus syndrome is a frequent cause of syncope, especially in the elderly. The syndrome is misdiagnosed in half of the cases if the massage is not performed in the upright position.
Reproducibility
A concordance between abnormal and normal responses during a second carotid sinus massage was reported in 93% of cases [76]. In another study [81], a pause >3 s was repeatedly reproduced in all patients who were referred for implantation of pacemaker because of severe carotid sinus syndrome.
Complications
The main complications of carotid sinus massage are neurological [87]. In three studies, neurological complications were reported respectively in: 7 among 1600 patients (5000 massages) with an incidence of 0.45% [87], 11 in 4000 patients (16,000 massages) with an incidence of 0.28% [88], and in 3 among 1719 patients with an incidence of 0.17% [85].
Even if these complications are rare, carotid massage should be avoided in patients with previous transient ischaemic attacks or strokes within the past 3 months (except if carotid Doppler studies excluded significant stenosis) or in patients with carotid bruits [87]. Rarely carotid massage may elicit self-limited atrial fibrillation of little clinical significance [67,72]. Since asystole induced by the massage is self-terminating shortly after the end of the massage, no resuscitative measures are usually needed.
Personnel
As it carries potential hazards, the test should be performed by physicians who are aware that complications, especially neurological, may occur.
Relationship between carotid sinus massage and spontaneous syncope
The relationship between carotid sinus hypersensitivity and spontaneous, otherwise unexplained, syncope has been demonstrated by pre-post comparative studies, two controlled trials, and a prospective observational study (Level B). Pre-post comparisons were made by analysing the recurrence rates of syncope in patients treated by pacing in several non-randomized studies [89,90]. These studies show fewer recurrences at follow-up. One non-randomized comparative study of patients receiving a pacemaker and untreated patients showed syncope recurrence rates to be lower in paced than non-paced patients [91]. Brignole et al. [81] undertook a randomized study in 60 patients; 32 patients were assigned to the pacemaker arm and 28 to the ‘no treatment’ group. After a mean follow up of 36±10 months, syncope recurred in 9% of pacemaker group versus 57% in the untreated patients (p<0.0002). Finally, in patients implanted with a pacemaker designed to detect asystolic episodes, long pauses (6 s) were detected in 53% of the patients during 2 years of follow-up, suggesting that a positive response to carotid massage predicts the occurrence of spontaneous asystolic episodes [92].
|
Tilt testing
Background
On moving from supine to erect posture there is a large gravitational shift of blood away from the chest to the distensible venous capacitance system below the diaphragm. This shift is estimated to total one half to one litre of thoracic blood and the bulk of the total change occurs in the first 10 s. In addition, with prolonged standing, the high capillary transmural pressure in dependent parts of the body causes a filtration of protein-free fluid into the interstitial spaces. It is estimated that this results in about a 15–20% (700 ml) decrease in plasma volume in 10 min in healthy humans [10]
. As a consequence of this gravitationally induced blood pooling and the superimposed decline in plasma volume, the return of venous blood to the heart is reduced resulting in a rapid diminution of cardiac filling pressure and thereby in a decrease in stroke volume. Despite decreased cardiac output, a fall in mean arterial pressure is prevented by a compensatory vasoconstriction of the resistance and the capacitance vessels in the splanchnic, musculo-cutaneous, and renal vascular beds. Vasoconstriction of systemic blood vessels is the key factor in the maintenance of arterial blood pressure in the upright posture. Pronounced heart rate increases are insufficient to maintain cardiac output [10]. The rapid short-term adjustments to orthostatic stress are mediated exclusively by the neural pathways of the autonomic nervous system. During prolonged orthostatic stress, additional adjustments are mediated by the humoral limb of the neuroendocrine system [10]. The main sensory receptors involved in orthostatic neural reflex adjustments are the arterial mechanoreceptors (baroreceptors) located in the aortic arch and carotid sinuses. Mechanoreceptors located in the heart and the lungs (cardiopulmonary receptors) are thought to play a minor role. Reflex activation of central sympathetic outflow to the systemic blood vessels can be reinforced by local reflex mechanisms like the venoarteriolar reflex. The skeletal muscle pump and the respiratory pump play an important adjunctive role in the maintenance of arterial pressure in the upright posture by promoting venous return. The static increase in skeletal muscle tone induced by the upright posture opposes pooling of blood in limb veins even in the absence of movement of the subject [8]. Failure of the above discussed compensatory adjustments to orthostatic stress is thought to play a predominant role in a large number of patients with syncope. This forms the basis for the use of tilt testing in the evaluation of patients with syncope. There is a large body of literature on the mechanisms involved in vasovagal syncope induced by tilt testing. Yet many unanswered questions remain regarding the multiple potential causes and the underlying pathophysiology. The panel did not consider an extensive review of pathophysiology as one of the goals of the consensus process. Excellent reviews are available [93–96].
Tilt test protocols
In 1986 Kenny et al. [97] observed an abnormal response to tilt test in 10 of 15 patients with syncope of unknown origin. This response consisted of hypotension and/or bradycardia. They also performed the test in 10 healthy controls without previous syncope, and an abnormal response was provoked in only one. In this study, the authors used an inclination of 60° during 60 min of tilt duration. Since then, tilt testing has been used extensively by many authors proposing different protocols for diagnostic, investigational and therapeutic purposes. Tilt testing protocols have varied with respect to many factors including the angle of tilting, time duration and the use of different provocative drugs.
In 1991, Fitzpatrick et al. [98] showed that the use of a bicycle saddle with the legs hanging free for tilt testing gave a low specificity when compared with footboard support. They also showed that tilting at an angle of less than 60° resulted in a low rate of positive responses. Analysing the time to positive responses, they reported a mean time of 24±10 min and proposed 45 min of passive tilting as an adequate duration for the test since this incorporated the mean duration to syncope plus two standard deviations. This method is widely known as the Westminster protocol. They reported a rate of positive responses in patients with syncope of unknown origin of 75% and a specificity of 93%.
In 1989, Almquist et al. [99] and Waxman et al. [100] used intravenous isoprenaline tilt testing. In the study of Almquist [99], after 10 min of passive tilt test without drugs, patients were returned to the supine position and isoprenaline infusion at initial doses of 1 µg/min was administered. When patients achieved a stable increase in heart rate they were tilted again. This manoeuvre was repeated at increasing doses up to 5 µg/min. With this protocol 9 of 11 patients with syncope of unknown origin and negative electrophysiological study showed hypotension and/or bradycardia, whereas such responses were found in 2 of 18 control subjects. In 1992, Kapoor et al. [101] using an isoprenaline tilt test at 80°, in which the drug was administered in progressive doses from 1 to 5 µg/min, without returning the patient to the supine position before each dose increase, reported a low specificity (between 45% and 65%). In 1995, Morillo et al. [102] and Natale et al. [103] proposed a ‘shortened’ low-dose isoprenaline tilt testing, in which, after 15–20 min of baseline tilt at 60–70°, incremental doses of isoprenaline designed to increase average heart rate by about 20–25% over baseline (usually 
3 µg/min) were administered without returning the patient to the supine position in one study, or returning to the supine position in the other. With this protocol, the rate of positive responses was of 61% with a specificity of 92–93%.
In 1994, Raviele et al. [104] proposed the use of intravenous nitroglycerin infusion. With their protocol, 21 of 40 (53%) patients with syncope of unknown origin had positive responses with a specificity of 92%. Ten of 40 patients (25%), had progressive hypotension without bradycardia. This response was classified as an exaggerated response consisting of an excessive hypotensive effect of the drug. More recently, Raviele et al. [105] have used sublingual nitroglycerin instead of an intravenous infusion. After 45 min of baseline tilting, 0.3 mg of sublingual nitroglycerin was administered. With this protocol, the overall rate of positive responses in patients with syncope of unknown origin was 51% (25% with baseline tilt test and 26% after nitroglycerin administration) with a specificity of 94%. An exaggerated response was observed in 14% of patients and 15% of controls. The main advantage of sublingual nitroglycerin is that venous cannulation is not needed for the protocol. Oraii et al. [106], Raviele et al. [107] and Graham et al. [108] have compared the isoprenaline test with the nitroglycerin test, with similar rates of positive responses and specificity, but with a lower rate of side effects with nitroglycerin. The optimal duration of the unmedicated phase before the administration of sublingual nitroglycerin has not been fully established. Bartoletti et al. [109] compared the effect of an unmedicated phase of 45 min versus 5 min on the overall positive rate of the nitroglycerin test. The test with the short passive phase was associated with a significant reduction in the rate of positive responses, and they concluded, that at least some baseline unmedicated tilt testing is needed. Recently, many authors have used a shortened protocol using 400 µg nitroglycerin spray sublingually after a 20 min baseline phase. Pooled data from three studies [110–112] using this protocol, in a total of 304 patients, showed a positive response rate of 69% which was similar to the positive rate of 62% observed in another 163 patients from three studies [109,112–113] using a passive phase duration of 45 min and 400 µg nitroglycerin spray administration. With this protocol, specificity remained high, being 94% in 97 controls [110–112]. Thus a 20 min passive phase before nitroglycerin administration appears to be an alternative to the more prolonged 45 min passive phase. This method is known as the Italian protocol. Sensitivity and specificity values are also similar in the older patients but the older patients have more exaggerated responses than younger [114]. The difficulty of a clear cut distinction between exaggerated and positive responses partly gives an explanation of the lower specificity values reported by some recent reports [108,115].
Clomipramine, a central serotoninergic agent, is another provocative agent. The drug is given intravenously administered during the first 5 min of tilting at a dose of 5 mg (1 mg/min). Following this, the patients remain in the upright position for a further 15 min or until syncope occurs [116,117]. Thus the test is time-saving. Its sensitivity ranged between 64% and 83%, the specificity was 93%. Further experience is needed before the test can become widely accepted.
Other drugs used as provocative agents during tilt testing include isosorbide dinitrate [118,119], edrophonium [120,121], and adenosine; the latter is discussed in another section.
Irrespective of the exact protocol, some general measures may be suggested when tilt testing is performed. Many of the following rules were published in 1996 as an expert consensus document [122]. The room where the test is performed should be quiet and with dim lights. The patients should fast for at least 2 h before the test. The patients should be in a supine position 20–45 min before tilting. This time interval was proposed to decrease the likelihood of a vasovagal reaction in response to venous cannulation [123,124]. With the protocols that do not use venous cannulation, time in supine position before tilting can be reduced to 5 min. Continuous beat-to-beat finger arterial blood pressure should be monitored non-invasively. Invasive measurements of arterial blood pressure can affect the specificity of the test, especially in the elderly [123] and in children [124]. Although intermittent measurement of