Diagnostic Value of Neurological Studies in Diagnosing Syncope. A Systematic Review

Abstract Background Syncope is common and approaches to establishing etiology remain a matter of clinical and financial importance. Patients often undergo comprehensive neurologic investigations despite a lack of compelling indications. The aim was to determine the prevalence of use and diagnostic yield of electroencephalography (EEG), head computed tomography (CT), head magnetic resonance imaging (MRI), and carotid Doppler ultrasound (CUS). Methods We conducted a systematic search in EMBASE, PUBMED and Cochrane from 1970-2015 for studies reporting the use of EEG, CT, MRI, and CUS in diagnosing the cause of syncope. The inclusion criteria were 1) observational and randomized trials, 2) frequency of use of investigations, and 3) diagnostic yield. Diagnostic studies of the more general Transient Loss of Consciousness were excluded. Results Of 149 screened studies, 15 studies having 6944 patients met the criteria. No studies met all 6 prespecified quality descriptors. The mean prevalence of test use were EEG 17.0%, CT, 57.3%, MRI 10.5%, and CUS 17.8%. The articles reported the likelihoods of a test providing diagnostic information for syncope etiology were EEG 1.35%, CT 1.18%, MRI 3.74%, and CUS 2.4%. Only 2 new and informative results were noted in 6334 tests. Conclusions Neurologic investigations for assessment of patients deemed to have syncope are used widely and widely ineffective. Neurologic investigations should be obtained only with a very high degree of clinical suspicion.