Patients presenting with transient alteration of neurologic function offer neurologists the opportunity to do what they do best: Solve a medical mystery and make a diagnosis. In general, the history will provide the essential clues, with ancillary testing providing only a modicum of additional assistance.

A careful history in the setting of a possible first seizure will frequently disclose prior spells that provide important diagnostic clues. Often, however, patients will not have recognized the relevance of these previous symptoms, and the neurologist must ask about them specifically. In fact, many typical epileptic phenomena such as déjà vu and morning myoclonus (“I’m just clumsy when I first get up”) may be considered normal by the person experiencing them or, alternatively, not neurologic in origin (e.g., “panic attacks”). A history that clearly indicates alteration of awareness is very helpful, but many patients—perhaps most—do not recognize brief changes in the degree of awareness.

Relatively few patients with epilepsy are able to recall all of their seizures, and a substantial number cannot remember any of them. Blum et al.⁶ prospectively determined seizure awareness in 31 patients admitted to an epilepsy monitoring unit. Twenty-three had epileptic seizures, but only 26% of these were aware of all of their seizures (regardless of seizure type), and 30% were not aware of any. Thus, obtaining additional information from observers is crucial. For recurrent spells, video-electroencephalogram (EEG) recording in an epilepsy monitoring unit is the best way—and sometimes the only way—to make a definitive diagnosis. Health professionals who have spent substantial time in one of these units know that findings can be surprising, and can include unexpected psychogenic spells, subclinical seizures, or substantial impairment of awareness during “auras,” to name a few.

Any highly stereotyped, recurrent, brief alteration of neurologic function without an obvious alternative explanation can be of epileptic origin. Table 1 lists some of the disorders that can mimic epilepsy in adults along with the clinical features that are most useful diagnostically. Table 2 provides similar information for children and adolescents. The most important condition to consider and recognize promptly is syncope, especially cardiac syncope due to a life-threatening arrhythmia that may be fatal in a subsequent spell.³¹ Table 3 compares the clinical characteristics of syncope and seizures to assist in distinguishing the two.

An incorrect diagnosis of epilepsy creates a significant problem with serious consequences.²³ About one fourth of patients with a diagnosis of epilepsy have been misdiagnosed^29,30; the rate of misdiagnosis is higher in children.³⁷ One should always remain open to the possibility that the initial diagnosis was incorrect, especially when antiepileptic drug treatment is ineffective, and take full advantage of future spells to help confirm or modify the diagnosis. The majority of patients misdiagnosed as having epilepsy are eventually found to have either psychogenic nonepileptic seizures (PNES) or syncope.^{11,25,29,30,40} Misinterpretation of EEG findings or overreliance on the EEG frequently contributes to misdiagnosis.^11,30 The presence or absence of interictal epileptiform discharges on EEG is not definitive and may be misleading.^5,19 Interictal epileptiform discharges (IEDs) occur in about 0.5% of healthy adults and in 1.9% to 3.5% of normal children.²⁶ Multifocal IEDs and IEDs occurring over the frontal and anterior temporal regions are highly correlated with clinical seizures,^10,18 but only about 40% of children with central-midtemporal spikes and 50% with occipital spikes had seizures in one study.¹⁸ When EEGs are interpreted by physicians without special training, a number of benign or normal patterns are commonly misinterpreted as epileptiform. These include benign epileptiform transients of sleep (also termed small, sharp spikes), wicket spikes, hyperventilation-induced high-voltage paroxysmal slow waves, various artifacts (such as overfiltered muscle potentials), repetitive vertex waves, especially in children, and the 6/sec “phantom” spike-wave phenomenon.³⁸ One study reported that 54% of patients referred to an epilepsy center with a previous diagnosis of epilepsy and wicket rhythms on EEG did not have epilepsy on further evaluation.¹⁹ Benbadis et al. reported 15 patients who were eventually diagnosed with PNES but had previously carried a diagnosis of epilepsy based on EEGs misinterpreted as epileptiform.⁵ Of the 15 records reviewed, the patterns that were incorrectly considered epileptiform were wicket spikes (n = 1), hypnagogic hypersynchrony (n = 1), and hyperventilation-induced slowing (n = 1). In the other 12 records, simple fluctuations of sharply contoured baseline background activity and fragmented alpha activity, not identifiable specific variants, were misinterpreted as epileptiform. Repeating the EEG or having it reinterpreted at a tertiary epilepsy center by a certified electroencephalographer can be helpful in problematic cases. It is also important to remember that EEGs can never rule out epilepsy. Even with repeated EEGs or prolonged monitoring, a significant number of patients with epilepsy (10% to 19%) will have no interictal epileptiform discharges²⁶; even ictal recordings may not have an identifiable scalp correlate in many frontal lobe seizures, as well as in simple partial seizures from any location.

Zaidi et al.⁴⁰ investigated the utility of noninvasive cardiovascular tests (including head-up tilt test and carotid sinus massage during continuous electrocardiography, EEG, and blood pressure monitoring) in 74 patients with apparent treatment-resistant epilepsy. An alternative diagnosis was found in 31 (41.9%) patients, including 13 (36.1%) of 36 patients taking an antiepileptic drug (AED). The most common alternative diagnosis was vasovagal syncope (25.7% of patients), followed by carotid sinus hypersensitivity (9.5%). At follow-up at about
10 months, all patients in whom an alternative diagnosis was made had subjective improvement, and 61.3% were symptom free. Of the 13 patients who were taking AEDs and for whom an alternative diagnosis was identified, 11 had successfully stopped their medications. These findings underscore the importance of cardiovascular testing early in the evaluation of a patient with “blackouts.” This is especially true in patients with cardiac risk factors, in whom it is essential to exclude cardiac arrhythmias.

The most challenging and most common alternative diagnosis in epilepsy monitoring units is PNES. Approximately 20% to 40% of patients admitted to such units have PNES,^3,4,13 including a substantial portion of those referred for epilepsy surgery. If seizures do not respond to trials of two AEDs, the patient should be referred for video-EEG monitoring, both for definitive diagnosis and, if epilepsy is confirmed, initial surgical evaluation. The chance of complete seizure control after failing two AEDs is <10%.²²

The following sections discuss differential diagnosis based on clinical presentation. Because PNES can manifest with any of these presentations, they are addressed separately at the end and more extensively in Chapters 207 and 282.