Functional neurological disorder in the emergency department

We provide a narrative review of functional neurological disorder (FND, or conversion disorder) for the emergency department (ED). Diagnosis of FND has shifted from a “rule‐out” disorder to one now based on the recognition of positive clinical signs, allowing the ED physician to make a suspected or likely diagnosis of FND. PubMed, Google Scholar, academic books, and a hand search through review article references were used to conduct a literature review. We review clinical features and diagnostic pitfalls for the most common functional neurologic presentations to the ED, including functional limb weakness, functional (nonepileptic) seizures, and functional movement disorders. We provide practical advice for discussing FND as a possible diagnosis and suggestions for initial steps in workup and management plans.


INTRODUC TI ON
This review aims to make the recognition of FND more accessible to emergency physicians, such that they can consider it as a likely or suspected diagnosis. We discuss in detail positive clinical signs observed in the most common FNDs presenting to the ED. Common diagnostic pitfalls are addressed as well as an approach to diagnostic testing. We then discuss how to have a conversation with patients about a possible FND diagnosis and first steps in management.

A brief word on terminology
Terminology regarding functional disorders has evolved over time.
Some terms, including "psychogenic," "psychosomatic," and "conversion" disorder, along with "somatization," presume an exclusively psychological cause, which is often not evident. "Nonorganic" suggests a dualism of brain and mind and "medically unexplained" suggests a problem where we have no idea about etiology, diagnosis, or treatment. Terms like "hysteria" or "pseudoseizures" are pejorative or suggest a problem that is faked. The research community have supported the use of the term "functional neurological disorder" as one that is etiologically agnostic. FND seizures will be referred to in this paper as functional seizures but are alternatively referred to in the literature as dissociative, psychogenic, or nonepileptic seizures or attacks.
Factitious disorder is the deliberate feigning of symptoms without external motivators, while malingering is deliberate feigning for the purposes of secondary gain such as financial benefit. These are distinguished from FND by their intentionality-FND symptoms are unintentional and involuntary (see "Dealing with doubt" section).

EPIDEMI OLOGY
The overall prevalence of FND in the ED has been reported as 0.4% to 4%, although studies likely underestimate rates due to inconsistency in diagnostic coding and underrecognition. 5,6 Patients with FND account for 9% of all acute neurological admissions. 7 Functional seizures represent around 10% of all seizures in the ED, 8 and of patients presenting with refractory status epilepticus resulting in intensive care unit care, 25% have FND seizures and not epilepsy. 9 Up to one-third of patients with functional seizures will develop functional status epilepticus, 10 often with accompanying ED visits. Of patients presenting with acute onset motor or sensory symptoms, up to 25% of cases have been found to be stroke mimics, with about one in 10 of those representing patients with functional neurological symptoms. [11][12][13] Patients with functional disorders, including FND, have a higher utilization of ED care correlating with higher health care costs, even after they have received a diagnosis. 3, 14 Moeller et al., 15 when examining diagnostic accuracy of neurological disorders in the ED, found that functional disorders were the leading cause of misdiagnosis of neurological presentations.
Costs of ED treatment for FND in 2017 among around 40,000 adults and children from a population of around 130 million U.S. citizens was $163 million, compared to $135 million for refractory epilepsy. 5

PATHOPHYS IOLOGY
Previous etiological ideas for FND were exclusively psychological.
New ideas about the pathophysiology of FND retain the importance of psychological models, but introduce a neurobiological perspective that places FND at the interface of the brain and mind. [16][17][18][19] Research using functional imaging suggests that these disorders are associated with dysfunction of brain networks involved in attention and perception, sense of agency, and prior sensorimotor expectations ( Figure 1). A number of functional neuroimaging and neurophysiologic studies have demonstrated differences in activations between patients with FND, healthy controls, and participants asked to feign symptoms. Symptom generation and maintenance is likely due to a combination of predisposing, precipitating, and perpetuating factors. These arise from the patient's biology, cognition, environmental factors, previous experiences, and in some cases acute triggers, which are more often a pathophysiological experience such as injury or migraine than a psychological one. [20][21][22][23] Dysregulation of attention is a major component of FND. Most people are likely familiar with the effect of focused attention on the self altering the outcome of an intended action-for example, being more likely to mix up one's words during a public speaking engagement. Our nervous system is designed to balance "bottom-up" sensory information traveling from the body to the brain with "top-down" predictions about what that sensory information will be. Dysregulation of this system in patients with FND is supported by electrophysiologic studies. 24,25 There appears to be an abnormally high amount of involuntary attention directed toward symptom-related prior beliefs and expectations, serving to reinforce and perpetuate symptoms. 26 This may explain why FND symptoms tend to improve with distraction, which physiotherapists capitalize on to treat FND motor symptoms. 27,28

MAK ING THE D IAG NOS IS OF P OSS IB LE OR LI K E LY FN D
The basis for FND diagnosis is the demonstration of clinical features of internal inconsistency (reversibility) and/or to a lesser extent incongruency with known patterns of structural neurological disease. 29 This is done primarily by looking for positive clinical signs of these disorders. 29 No clinical sign in isolation should be taken as confirmation of a functional disorder. Importantly, the need for a stressor preceding onset of physical symptoms has been removed from the DSM-5. In the absence of an established therapeutic F I G U R E 1 Decreased functional connectivity between the right temporoparietal junction and bilateral sensorimotor regions in patients with functional movement disorder*. Contraction of one side of the platysma, creating the effect of a facial droop.

++
Hip abductor sign 37 Return of strength to hip abduction in the weak leg with contralateral hip abduction against resistance ++ Give-way/collapsing weakness 35,41,42 Strength is initially normal and then collapses with resistance.

++
Dragging monoplegic leg 20,35 Plegic leg is dragged behind body often with hip internal or external rotation and without hip circumduction.

++
Drift without pronation 35,43 Isolated downward arm-drift without associated pronation. + Global pattern of weakness 35,44 Equal weakness of both flexor and extensor muscles, both proximally and distally.

+
Motor inconsistencies 45 Inability to produce one movement, while using the same muscles to produce a different movement. For example, a patient may have difficulty dorsiflexing while supine, but be able to stand on heels without difficulty. + +++ = highly reliable; ++ = reliable; + = suggestive. a Reliability determined based on available clinical data 34 and author consensus.

TA B L E 1 Selected clinical signs in functional weakness
relationship, as would be typical in the ED, we suggest avoiding routinely questioning patients about past trauma. While it is a risk factor for FND, occurring in 10% to 30%, diagnosis should not be based on its presence or absence, and harm can be done by bringing this up with patients if they are not prepared to talk about it.
In gathering the history, care should be given, as always, to taking the patient's symptoms seriously. Practically, this can include making statements indicating that these symptoms are familiar, that this is a real problem, and that you believe them. 30 It is important to ask about the amount of disability the symptoms are causing for the patient on a day-to-day basis. 31

Functional limb weakness
Functional limb weakness is one of the most common presentations of FND to the ED 2 and can present similarly to a variety of structural disorders including stroke and demyelinating lesions. About half of patients with functional limb weakness will present with acute onset of symptoms. 32 One or any combination of limbs can be affected, although unilateral symptoms are the most common. 20,33 Often when there is only one limb that feels weak, subtle weakness will also be found in the other ipsilateral "normal" feeling limb on examination. 34 Patients may subjectively note that their limb feels heavy, like it is "not there," or "not a part" of them. 34 If the upper limb is affected, patients may report frequently dropping things. If the lower limb is affected, patients may drag their leg behind them, 20,35 or find their knee giving way leading to falls. 34 20 Similarly, drift without pronation as a sign of functional arm weakness has a reported high specificity of 93% to 95%. 38 However, most providers would agree that this can be seen in clinical practice in a variety of non-FND patients. Caution in interpretation should be taken when only one positive sign is present, when they are only mildly positive, or when there is significant pain. Patients with neglect or apraxia may also have falsely positive signs. 34 We present the reliability of these signs in Table 1 as a composite of the data available and author consensus based on clinical experience.

How do I know it's not a stroke?
Stroke and transient ischemic attacks, as well as other stroke mimics, will necessarily be on the differential for acute onset neurological symptoms, and typical stroke protocol should be followed in the initial workup of these patients. Data from a systematic review and a meta-analysis show that FND represents between 7% and 15% of stroke mimics, making it only slightly less common than stroke mimics related to migraine or seizure. 12,13 If the diagnosis remains uncertain, patients can usually be treated safely with tPA: the rate of symptomatic intracerebral hemorrhage in stroke mimics is 0% to 0.5%, with systemic hemorrhage and angioedema being similarly rare. [46][47][48][49][50] Other potential harms of giving tPA to a patient with FND include increased cost, with one study showing a median excess cost for stroke mimics given tPA to be over $5,000 USD per admission 51 as well as a potential for adverse psychological impact. On balance, it is likely best to err on the side of overtreating, rather than undertreating, with tPA in cases of uncertainty when patients otherwise meet criteria for thrombolysis.

Functional sensory loss
Sensory symptoms in FND range from pain or a "pins and needles" sensation, to heaviness or numbness. 52 conjunction with sensory changes and can help put the sensory symptoms in a broader clinical context. 54 Sensory testing on examination is necessarily subjective and prone to bias, on the part of both the patient and the examiner. 55 The clinical signs for functional sensory loss have not been found to be reliable in terms of differentiating from structural sensory loss. 53 For example, reliability for splitting of vibration sense across the sternum or forehead varies widely across studies, ranging from 50% to 95% for sensitivity and 14% to 88% for specificity. 38

Functional seizures
Functional seizures are perhaps the most well studied of all functional disorders, and several attempts have been made to determine the reliability of various distinguishing features from epilepsy. Patients often report warning symptoms of autonomic arousal prior to the event. [56][57][58][59][60] They may also report dissociation-a feeling that the world or their body is disconnected from them 56,61,62 (a note of caution: symptoms of autonomic arousal and dissociation can also precede focal onset seizures as well as syncopal episodes).
A detailed history from the patient and any witnesses to the event should be taken, going over any warning symptoms, ictal features, and postictal state. Examining any video the patient or their family members have of similar events can help greatly with diagnosis. 63 Table 2 lists selective features that have been shown to be useful in differentiating between functional and epileptic seizures. The sum of the clinical signs and history, rather than one clinical sign provided, should be taken as a whole to determine whether the episode is likely a functional seizure. 64 We strongly discourage maneuvers that may harm an individual, such as dropping the patient's arm on to their face. These tests are diagnostically unhelpful because they will often be negative While the majority of functional seizures are convulsive, thrashing, or tremulous events, about 30% of patients will have events that resemble syncope, in which they fall down, are still, and are unresponsive. 38 For these types of events, a phenotype of sudden collapse to the ground, with eyes closed, and documentation of 2 or more minutes of loss of consciousness is highly specific for a functional disorder etiology. 38,66,67 Research on biomarkers to differentiate functional from epileptic seizures has thus far not proven helpful. Serum lactate and prolactin levels may be raised in epileptic seizures compared to functional seizures, but levels are highly dependent on timing in relation to the seizure and can be elevated in functional seizures. [68][69][70] For example, one study asking participants to feign a seizure demonstrated an increase in lactate levels from baseline. 71 Similarly, elevation of creatine kinase or white blood cell count, while possibly more common after an epileptic seizure in comparison to a functional seizure, are nonspecific and should not be relied upon for diagnosis. 68

Functional movement disorders
Functional movement disorders are the second most common cause of acute movement disorders presenting to the ED. 78 The primary characteristics of functional movement disorders are that they diminish or resolve with distraction and/or entrain (change frequency to match that of other motor tasks). [79][80][81] Movements may be sudden in onset and have spontaneous remissions. The affected body part may change over time. Do not assume that just because the movement appears to be "bizarre" that it relates to a functional disorder.
Many movement disorders can appear strange, such as task-specific dystonia or stiff person syndrome, emphasizing the need for a neurologist to usually be involved in making a diagnosis.
In the case of functional tremor, it may be present at rest, with sustained postures, or on action: look for variability in frequency, rhythm, and axis or direction (but not amplitude, because this can vary in a number of tremor etiologies). 82 Improvement with distraction may be seen while taking a history or may require the examiner to ask the patient to perform other motor tasks with a nonaffected body part. 79 Entrainment can be demonstrated by asking the patient to copy a rhythmic movement with an unaffected limb, such as finger tapping. 83 In functional tremor, tremor will either improve or change to match the frequency of the voluntary movement or the patient will have trouble copying the movement.

FND and suspected cauda equina syndrome
Over 50% of patients presenting with cauda equina syndrome (CES) will have normal imaging ("scan-negative CES"). 84,85 Recent studies have pointed to a high frequency of associated FND symptoms and signs, especially lower-limb weakness FND signs, in these patients. 86 Patients with scan-positive CES are more likely to have diminished or absent ankle jerks than scan-negative patients (78% vs. 12%).
Abnormal anal sphincter tone on digital rectal examination and high postvoid residual volume (200 or 500 cc) have not been shown to be clinically useful differentiators. 87 Ultimately, given the potential morbidity of CES, no historical features or clinical signs remove the need for urgent neuroimaging. If imaging fails to identify a structural etiology, however, then discussing FND as a possible contributor to symptoms may be appropriate.

Diagnostic pitfalls
The diagnosis of possible or likely FND should usually be made on the basis of positive clinical features, usually from the physical examination (including seizure semiology), not from the clinical history. 1. Presence of psychiatric comorbidity: A diagnosis of FND should not be based on the patient having a psychiatric disorder such as anxiety, depression, or a personality disorder. 2. Failure to consider structural disease comorbidity: One of the commonest risk factors for FND is the presence of minor or major disease comorbidity such as multiple sclerosis, stroke, or epilepsy. Therefore, even in a patient with clear FND, always consider whether they may have an additional medical or neurological condition. 3. Putting too much weight on the presence or absence of "stress": A diagnosis of FND should not be based on the presence of an obvious life event or stressor nor should it be discarded due to lack of recent stress. Similarly, just because the patient attributes their symptoms to stress does not mean this is the case. 4. La belle indifférence: i.e., the patient seemingly not caring about their symptoms is not a reliable marker for FND and occurs just as commonly in structural disorders. 88 5. The patient is not a young female: FND should not be excluded based on demographics. Patients can be male or female, young or elderly, and from diverse socioeconomic backgrounds. 6. The patient seems too "normal": patients with FND may be nice, normal people too! Comorbid personality disorder may also be present in patients with FND at rates increased from those in the general population. 94 Despite the higher rate of psychiatric disorders in the FND population, not all FND patients have a psychiatric diagnosis (indeed, up to one-third may not). As such, psychiatric comorbidity is best seen as a risk factor, rather than a causative factor, for FND. In patients who do present with clear psychiatric symptoms, ensuring that these are optimally managed is often necessary for patients to engage meaningfully in therapy for FND symptoms.

DE ALING WITH DOUBT: IS MY FND PATIENT FAKING HIS OR HER SYMP TOMS?
In the ED setting, perhaps more than any other, the issue arises as to do. This is not the same as observing variability that the patient is aware of. Frank deception remains rare, and the error of considering that someone is feigning when they are not is one that every doctor should strive to avoid.

MANAG EMENT
Recognition of FND is one of the first challenges, especially in "acute stroke" or "status epilepticus" presentations. Generally, we recom- The diagnosis of possible or likely FND can be delivered in the same manner as diagnosing any other condition ( Table 4). The clinician should explain to the patient the name of the diagnosis and how the diagnosis was made and provide some basics regarding pathophysiology. In explaining how the diagnosis was made, it is often useful to demonstrate to the patient any positive physical signs on their examination, such as a Hoover's sign. 98 In the case of functional seizures, review semiological features that are strongly suggestive of FND rather than focusing on why it is not epilepsy.
Any specific concerns the patients may have had about alternative diagnoses should be addressed. In explaining pathophysiology, it can be effective to use analogies, such as comparing the brain to a computer and explaining that FND is "software problem" of the brain (Table 5).

Referral
Assessment by a neurologist is usually necessary to confirm the diagnosis, arrange therapy, and identify any concurrent neurological disorders. Once the diagnosis of FND is confirmed by a neurologist, typical avenues for treatment include physiotherapy or psychological therapy. 27,99 There is increasing evidence of effectiveness of these approaches, which should ideally be delivered in a multidisciplinary team. 100

CON CLUS ION
Functional neurological disorder is a disabling and distressing condition that commonly presents to the ED and can take many forms. As in the ED by disclosing the potential diagnosis to patients in a clear manner, providing a brief explanation for why this diagnosis is suspected, and referring on to neurology for further treatment.

CO N FLI C T O F I NTE R E S T
The authors have no potential conflicts to disclose. TA B L E 4 Key elements to include and to avoid in discussing a possible diagnosis of FND

Do include Avoid
• The name of the diagnosis • How the diagnosis was made (including sharing positive diagnostic signs) • A brief explanation of pathophysiology • Tell the patient their symptoms are real and not imagined • Emphasize that these symptoms are common • Emphasize that symptoms are potentially reversible and therefore could improve • Offer further resources to learn more • Only an explanation of what they do not have • Attributing symptoms to psychological problems or stress • Saying or inferring that this is "imagined," "all in their head," or voluntary in some way • Misattribution of symptoms • Using negative investigations as evidence of the diagnosis Abbreviation: FND, functional neurological disorder.

TA B L E 5
Examples of ways to explain the diagnosis of possible FND "You likely have functional neurological disorder, or FND, which is causing your weakness. I can see from your examination that your nervous system is not damaged; however, it's struggling in getting its messages through. Can you see how the more you try, the worse your leg weakness gets, but when you are focused on your other leg it works much better? [demonstrate Hoover sign] What this tells me is that your brain is having difficulty sending messages to your leg, but that improves when you are distracted. It's like the opposite of phantom limb pain. Your brain thinks the leg isn't there even though it is. It shows us that there is no damage to your nervous system and the problem is potentially reversible." "Seizures/attacks in FND are caused by a 'trance-like' state in the brain called dissociation. The brain shuts itself down temporarily, often in response to a 'red-alert' state and this becomes a reflex or habit, which is why it keeps happening."