We should send more patients with syncope home.
The data are clear and compelling. Syncope is incredibly common, representing up to five percent of all ED encounters. Overwhelmingly, these patients will do fine. The incidence of serious outcomes is well below one percent in nearly every cohort. Even in higher-risk populations, the likelihood of an adverse event barely rises to clinical significance. Nonetheless, syncope remains a condition marked by widespread practice variability and high rates of admission, about 50 percent in most investigations.
The profound gap between the admission rate and the probability of bad outcome leaves much room for improvement. It seems like it would be easy to dictate a switch to increased discharge rates for these patients, but that approach ignores the deeply ingrained practice patterns and experiences of EPs throughout the country. Suddenly shifting from predictable admission to routine discharge of a population once perceived to be at risk for adverse events requires an intermediary, a medical and medicolegal buttress as we refine the clinical practice of emergency medicine.
To understand the way forward, it's critical to examine where we are. Our high rates of admission for syncope are driven by fear of ventricular dysrhythmia and sudden cardiac death. Legendary EM educators have done much to drive increased recognition of electrocardiographic harbingers of such an outcome, and most EPs can recognize sodium channelopathies and accessory pathways without a moment's hesitation.
Clinical decision algorithms have appeared and then faded, always falling short of the perfection we demand and doing little to temper our fears of concealed cardiomyopathy or occult arrhythmia. Would the patient we discharged be the one to drop dead hours later from his run of V-tach?
A Diagnostic Waypoint
Brain (or B-type) natriuretic peptide is secreted by the ventricles in response to myocyte stretch. It has permeated EDs and inpatient wards around the globe for its lately-lambasted place in evaluating the dyspneic patient, but continues to find early purchase in a number of other roles as a prognostic marker in pulmonary embolism, a therapeutic guidepost in pulmonary hypertension and a diagnostic waypoint in the increased rate of discharge of patients with syncope. A normal BNP effectively rules out the presence of hidden cardiomyopathy or cloaked structural catastrophe, and it reliably rises in the setting of life-threatening cardiac dysrhythmia.
A small prospective study in 2009 by Pfister, et al., demonstrated a cardiac etiology of syncope in nearly one-third of included patients. Those adjudicated to a cardiac cause had consistently higher NT-pro-BNP levels than their noncardiac comparators. (Int J Cardiol 2009;133:51.)
A larger study of patients with already deemed medium- or high-risk syncope (patients over 60 or with history or physical findings that would give nearly anyone pause before hitting “discharge”) found that a BNP <300 pg/ml reliably precluded serious outcomes in all but a small handful of patients: a 74-year-old who had been admitted with a stroke and pneumonia; a 66-year-old who died nearly three months later (on another hospital admission) and on whom no autopsy was performed; and a 91-year-old and a 76-year-old whose “adverse event” consisted of asymptomatic heart block noted on outpatient Holter monitoring. (Emerg Med J 2007;24:270.)
The ROSE (Risk Stratification of Syncope in the Emergency Department) study, a prospective, observational study of 529 adults presenting to the ED with syncope, found that a BNP <300 pg/ml had a 99.8% NPV for all-cause death at one month following presentation. (J Am Coll Cardiol 2010;23;55:713.) Patients in whom the BNP did not predict a serious event had GI bleeds, stroke, and subarachnoid hemorrhage—dangerous entities, to be sure—but irrelevant confounders when the diagnostic question we ask of the test is limited to cardiogenic syncope only.
BNP isn't the hero we need, but it's the one we deserve. The test is neither a catchall nor a safety blanket, and should only be incorporated in the narrow context of evaluating for the possibility of occult arrhythmogenic syncope in the patient who otherwise seems safe for discharge. The low risk of adverse outcome following ED presentation for syncope suggests that a nihilistic and minimalist approach is safe—however unpalatable—but incorporation of BNP into the evaluation provides a clinically reassuring and medicolegally justifiable waypoint as we strive to reexamine the practice of syncope over-admission.
Share this article on Twitter and Facebook.
Access the links in EMN by reading this on our website or in our free iPad app, both available at www.EM-News.com.
Comments? Write to us at firstname.lastname@example.org.