The ILAE Nonepileptic Seizure Task Force has established criteria for “possible” and “probable” PNES relying only on event semiology and a nonepileptiform interictal EEG. The diagnosis of “clinically established” PNES requires visualization of a typical event (either in person or on video) by a clinician experienced in the diagnosis of epilepsy and separate non-video-EEG capture of a typical event without epileptiform abnormalities. The diagnosis of “documented” or definite PNES or NEPS requires video-EEG, interpreted by an experienced neurophysiologist, and allows a high degree of diagnostic certainty [37, 54]. The gold standard is capture of all typical episodes on video-EEG, without ictal EEG changes, and with normal awake EEG rhythms before, during, and after the event, in combination with a PNES-consistent clinical history and semiology [37]. During the event, the EEG may be obscured by artifact (Table 4.2).

There are important caveats in the use of EEG to diagnose PNES and NEPS. Scalp EEG has poor sensitivity (20–70%) for epileptiform activity in frontal lobe epilepsy [61] and in focal seizures without dyscognitive changes [62, 63]. For this reason, events that are clinically consistent with hypermotor frontal lobe seizures and simple partial seizures must be evaluated closely and considered seriously for epileptic etiologies even if video-EEG does not show epileptiform changes. As previously noted though, frontal lobe seizures are often very brief and therefore rarely a diagnostic consideration in the evaluation of potential SE. Simple partial SE can occur (including for example, epilepsia partialis continua, with ongoing involuntary focal motor activity without any alteration in consciousness) and is frequently without EEG abnormality, but any progression to impairment of consciousness should correlate with abnormalities on EEG. It is estimated that scalp EEG demonstrates diagnostic ictal or postictal features in >95% of focal dyscognitive seizures [64].

Conversely, it is important to avoid over-interpreting normal variants (e.g., wicket rhythm, rhythmic midtemporal theta of drowsiness, subclinical rhythmic electrographic discharges of adults (SREDA), positive occipital sharp transients of sleep, and nonspecific EEG abnormalities (e.g., nonfocal slowing) in patients being evaluated for SE versus NEPS. Nonspecific EEG abnormalities are common in both PNES and epilepsy [65], and the over-interpretation of nonspecific findings is one of the most common reasons for the misdiagnosis of patients as epileptic [66, 67]. Rather, specific epileptiform EEG changes (interictal spikes/polyspikes, sharp-and-slow waves; ictal rhythmicity and evolution; postictal focal or generalized slowing) should be identified to make a definite interpretation of epileptic seizures.

Although they are not necessary for the diagnosis of definite PNES or NEPS, there are other tests that may be useful in distinguishing between NEPS and SE, especially in the absence of EEG data. Elevated prolactin levels are highly specific (96%) for epileptic seizures, but sensitivity is best (60%) in generalized convulsive seizures, drops off in focal dyscognitive seizures (46%), and plummets in the setting of focal seizures without dyscognitive effects [38, 68]. The American Academy of Neurology (AAN) therefore classifies prolactin levels as a useful adjunct in differentiating generalized convulsive or focal dyscognitive epileptic seizures from PNES [68]. The AAN recommends that blood should be drawn 10–20 min after seizure onset and compared to a baseline non-ictal prolactin level, with a level at least twice normal suggestive of epileptic seizures. Importantly, prolactin has a short half-life, and its release may drop off during SE, resulting in a false negative if blood is drawn too long after seizure onset [69]. Use of dopamine agonists can also cause false negative results, while dopamine antagonists (and breast stimulation such as a suckling infant) can cause false positive results. Syncopal events may also raise prolactin levels [69].

Serum creatine kinase (CK) can also be used to distinguish between convulsive SE and NEPS, though its use has not been as extensively studied as that of prolactin [70]. CK levels usually elevate well above the normal range in patients with convulsive SE, while remaining within the normal range in patients with NEPS. Levels begin to elevate only 3 h after seizure onset, and remain elevated for approximately 36 h after cessation of seizures [70]. Serum cortisol, dexamethasone suppression tests, white blood counts, neuron-specific enolase, and brain-derived neurotrophic factor have not reliably distinguished between patients with NEPS and those with SE [37].

Structural neuroimaging has been only minimally useful in making the clinical distinction between epileptic and nonepileptic seizures. The majority of patients with epilepsy have normal magnetic resonance imaging (MRI) scans [71, 72], and a significant number of patients with PNES have incidental MRI abnormalities [72–74]. Functional neuroimaging studies have hinted at connectivity differences on a group level, but such techniques have not yet proven effective in distinguishing patients on an individual clinical basis [75].

Studies of personality measures such as the Minnesota Multiphasic Personality Inventory (MMPI) and MMPI-2 have suggested some differences in personality characteristics between patients with PNES and those with epileptic seizures but have failed to identify a single PNES personality profile and generally fail to show a combination of good sensitivity and specificity [76, 77]. Given that such assessments are time consuming and require the patient’s extensive cooperation, they may be more useful for guiding therapy and for research purposes than for making the initial diagnosis of NEPS versus SE.

In evaluating a case of potential SE or NEPS, other nonepileptic mimics of SE should be considered and excluded. These disorders can be conveniently subdivided between those that involve abnormal movements (and therefore mimic convulsive or focal motor SE) and those that involve only impaired responsiveness and therefore mimic nonconvulsive SE. Importantly, these disorders are more often encountered in critically ill patients in the ICU or in postoperative patients in the post-anesthesia care unit (PACU) than in newly presenting patients in the emergency department, whereas NEPS more often presents through the emergency department in patients who are not otherwise acutely ill. Nevertheless, “non-NEPS” mimics of SE can potentially arise anywhere and must be considered whenever the diagnoses of SE and NEPS are contemplated.

Once the definite diagnosis of NEPS has been made and epileptic seizures are ruled out, the first step of treatment is presentation of the diagnosis to the patient. Making and clearly delivering a positive diagnosis in a timely manner is important for the patient’s long-term prognosis and to minimize repeated and unnecessary use of healthcare resources [11, 90, 91]. This is one of the key tasks of the neurologist and must be performed in a supportive and nonjudgmental manner so as to facilitate adherence with the next stage of psychotherapeutic treatment. This may decrease the chances of repeated appearances in other emergency departments [92, 93].

Formal validated protocols for delivering the diagnosis of PNES are available [94, 95]. The key points are to acknowledge that the events are real and disabling, to give a positive diagnosis with a name for the condition, to explain that it is a common and recognized condition, to describe the psychological causes (stress, strong affect), to specify that it is not epilepsy and that anti-seizure drugs are not effective, to explain that psychological treatment is effective, to discuss referral to a behavioral health specialist, and to reassure the patient that improvement can be expected and complete resolution is possible [94]. We have also found it helpful to make comparisons with more commonly experienced responses to extreme stress, such as tachycardia, diaphoresis, and panic attacks.

If at all possible, behavioral health specialists should be involved in the early evaluation of the NEPS patient and in delivering the diagnosis to the patient [96]. This facilitates evaluation for common psychiatric comorbidities and suicidal or parasuicidal behavior (i.e., suicidal “gestures” unlikely to result in death, such as cutting) and helps present a single unified team to the patient, thus reducing the sense of being abandoned by neurologists and dismissed to psychiatrists. Patients with PNES have consistently poor adherence with psychotherapy, and one study showed that adherence is significantly improved when the referral happens within a single unified system rather than to an outside mental health specialist [97]. Introducing behavioral health specialists early allows them to use the NEPS episode as a teachable moment, motivating future psychotherapy. If mental health clinicians cannot be involved prior to delivery of the diagnosis, they should be involved as soon afterwards as possible. It is vitally important that there be clear and open communication between the referring neurologist and the behavioral health specialist (as well as the primary care physician) regarding the diagnosis and the basis on which it was made.

If epileptic seizures are entirely ruled out, another important early step is to stop all anti-seizure drugs (ASD), except those providing a documented psychopharmacologic benefit to the patient. A randomized controlled trial of immediate versus delayed withdrawal of ASDs in patients with newly diagnosed PNES showed improved outcome with fewer subsequent seizure-like events, less use of rescue medicines, and improved locus of control in the immediate withdrawal group (“locus of control” refers to individuals’ belief that they can control events affecting them) [98]. Stopping ASDs while the patient remains on video-EEG also allows the clinical team to evaluate for new interictal epileptiform abnormalities or epileptic seizures emerging in the absence of medications. If possible, we recommend allowing patients with newly diagnosed NEPS to remain hospitalized for at least 1 day after the diagnosis is made, to allow the withdrawal of medications, to allow patients to absorb the diagnosis, and to foster relations with the neurologists and mental health clinicians who will follow them as outpatients.

A minority of patients may experience resolution of PNES with no intervention beyond the delivery of the diagnosis, with 13–16% of patients becoming seizure free at a 6–12 month follow-up [99, 100], but the large majority of patients require further treatment. The best-studied treatment modality is cognitive behavioral therapy (CBT)-based psychotherapy, which has been shown to reduce seizure frequency and improve quality of life in two pilot randomized controlled trials [101, 102]. Other therapeutic modalities, including psychodynamic interpersonal therapy, have been shown to be effective in observational studies [100, 103, 104]. Psychopharmacologic agents such as SSRIs or neuroleptics may be indicated to treat comorbid psychiatric disorders but have not thus far been demonstrated to be effective in treating PNES directly [103]. There are no studies specifically regarding treatment of NEPS.