Testing in medicine is done for a reason. A medical test should be performed to establish or reject a diagnosis, quantify disease severity, monitor disease progression, or identify a treatment that could not have been determined before the test. The ultimate outcome of a useful test and the consequent intervention should be either that it (1) reduces morbidity (improves quality of life) or (2) reduces mortality (increases the quantity of life). How well a test accomplishes these outcomes must come with an acceptable risk/benefit ratio. Even if a test (or intervention) does not cause harm itself, the consequences of making an incorrect diagnosis resulting in a wrong course of therapy may cause harm. It is within this context that we argue that tilt table testing (TTT) for the workup of syncope should be abolished. We submit that the TTT fails to establish an explicit cause of syncope, is plagued by false positives, and never plays a role in guiding treatment. Given these limitations, we believe that TTT should stop being administered for these purposes.
Syncope is a condition with a high lifetime prevalence accounting for 1% to 3% of emergency room visits with a >30% rate of subsequent hospital admission. Much has been published regarding the cost of syncope to the health care system, and the subsequent high use rates of unnecessary testing, especially neurological imaging. The cornerstones in the initial evaluation of a patient presenting with syncope are the history, physical examination (including orthostatic vitals), and ECG. The diagnostic accuracy of using this clinical approach is 88%. Furthermore, neurally mediated syncope (NMS), also called vasovagal syncope, accounts for the majority of presentations. The patient's history is key to establishing the diagnosis of NMS. The presence of typical contextual triggers (pain, fear, emotional stress, phlebotomy/needle sticks, postexercise, heat, micturition, etc), a long history of syncopal events, the presence of prodromal symptoms (warmth or nausea), postsyncope fatigue, young age, and the absence of known heart disease are all suggestive of NMS.
Given the benign prognosis of NMS and the high diagnostic accuracy of the history, examination, and ECG in making a diagnosis, what does TTT add? The latest iteration of the US syncope guidelines recommend TTT in patients with suspected NMS who have an unclear diagnosis after initial evaluation. However, syncope during TTT is notoriously nonspecific. The European syncope guidelines note that whereas 92% of patients with NMS will have a positive TTT, 47% of patients with arrhythmic syncope also will have a positive TTT. The use of pharmacological provocation (eg, sublingual nitroglycerin or isoproterenol) increases the risk of precipitating syncope during TTT, even in individuals who have never had a syncopal event. Although such events are often called false positives, we contend that it simply induces a highly conserved reflex (ie, a fall in cerebral perfusion pressure from a decrease in cardiac output; heart rate or stroke volume, peripheral resistance, or both) regardless of whether or not the patient has ever had syncope as previously described by us.
It is important to emphasize that virtually everyone will have syncope given an adequate hemodynamic stress to a circulatory condition that exaggerates central hypovolemia (G-load, heat, exercise, etc). Syncope during orthostatic stress may be especially common in athletes with large compliant hearts (and therefore a large fall in stroke volume during orthostasis) and large vasodilatory capacity, thus providing further evidence that TTT-provoked syncope does not equate to having a disease. Therefore, the provocation of syncope during TTT does not help the physician in elucidating the mechanism for the patient's syncopal events, as the mechanism of syncope during TTT and during the patient's clinical events may be entirely different.
Indeed, given the primacy of symptoms for the diagnosis of NMS, it is difficult to imagine a scenario where the results of TTT will change a physician's pretest diagnosis. For example, if a patient has classic clinical features of NMS, a TTT that does not provoke syncope would not change that diagnosis. Conversely, for a patient with a concerning history of syncope, such as associated with palpitations or during exercise, or in a patient with a history of cardiovascular disease, the presence of NMS on TTT should never preclude a life-threatening pathophysiology; the case of former Boston Celtics basketball star Reggie Lewis serves as a stark reminder of this lesson. Although he had a concerning episode of syncope during a basketball game, a positive TTT led to the diagnosis of NMS. He subsequently died of cardiac arrest from ventricular fibrillation at the age of 27 years. Therefore, TTT at best adds little diagnostic value for determining the etiology of undiagnosed syncope, and at worst may lead to false reassurance in patients with malignant etiologies.
In what setting could TTT be helpful? Both US and European guidelines note that TTT plays a role in patients with orthostatic hypotension and postural orthostatic tachycardia syndrome.[2,4] However, a simple active stand test is more clinically relevant, reproduces the circumstances of the patient's complaint, and is just as sensitive but more specific than TTT for diagnosing these conditions. Moreover, an active stand test (5–10 min quiet standing) is simple and can be done cheaply in an office setting without specialized equipment. This test will also distinguish patients with early orthostatic hypotension versus delayed orthostatic hypotension (ie, blood pressure reduction >3 min after assuming the standing position). The active stand test should be done in all patients with suspected orthostatic hypotension or postural orthostatic tachycardia syndrome.
Some clinicians have used TTT in patients with known NMS to assess the effectiveness of treatment such as physical countermaneuvers, pharmacological measures, or pacing. However, the US syncope guidelines give this strategy a class III recommendation because the reproducibility, prognostic implications, mechanism, and timing of NMS can be quite variable.
Finally, the outcomes of TTT do not lead to therapeutic interventions that reduce morbidity, recurrent events, or mortality. For example, TTT is purported to be useful in distinguishing between the cardioinhibitory and vasodepressor predominant forms of NMS. However, we believe that the utility of this capability is limited because the mechanism by which syncope occurs in a laboratory setting can be quite different from what happens in the field. For example, if a patient has a cardioinhibitory response on the tilt table (especially after pharmacologic provocation), but not in the field, then pacing will not likely prevent future syncope. Additionally, in the current era of sophisticated ambulatory ECG monitoring devices and implantable loop recorders, we believe TTT is unnecessary and indeed can be misleading in making this diagnosis. Remote ECG monitoring allows for correlation of arrhythmic events with clinical events, thus ensuring higher specificity and clearly linking the patients' symptoms with their rhythm. These techniques (longer than 24 hours) should be the cornerstone of diagnostic testing in patients for whom the diagnosis cannot be established with confidence on initial evaluation, or for whom confirmatory testing is necessary.
In conclusion, the false positive rate in TTT is not trivial and can lead physicians and patients away from the true mechanism of syncope. We already have a problem with overuse of testing in the syncope patient. Let us stop contributing to this waste by abolishing the tilt table test for workup of syncope.
Circulation. 2020;141(5):335-337. © 2020 American Heart Association, Inc.