“Everything that shakes or faints need not be epilepsy,” the French Child Neurologist Jean Aicardi once said. Unfortunately, an incorrect diagnosis of epilepsy not only exposes a child to the side effects of antiepileptic drugs but also eliminates an opportunity to treat his or her true condition.
The problem is more common than you might think. Research from the U.K. has shown that 25 to 30 percent of children seen for epilepsy turn out not to have it. Children seen in the Boston Children’s Hospital Epilepsy Center for a “CIBAS” consultation (“could it be a seizure?”) also frequently have a different diagnosis.
EEGs are often to blame for a mistaken diagnosis of epilepsy. The electrical signals they pick up are relatively weak and have a low signal-to-noise ratio. The signals are easily confounded by movement, tensed muscle and other artifacts. Because of these limitations, we often repeat EEG testing or do prolonged monitoring to capture an actual event. Interpretation is another pitfall. Even in experienced hands, interobserver variability is high and false-positives are common.
The diagnostic yield of EEG really depends on the pre-test probability of epilepsy — that is, your prior degree of c
ertainty about whether your patient has it. Getting a better history can increase your certainty. For example:
#1: Parent: “The babysitter saw a staring spell that lasted 10 seconds, then he was awake and talking but unable to move.”
#2: Babysitter (directly on phone): “He stared off, moved his head to the right, was unresponsive, then became lethargic and floppy and mumbled for a few minutes.”
The second conversation supports a diagnosis of epilepsy more strongly than the first, and this makes the same EEG results more informative.
Before you think “epilepsy”
Before ordering an EEG, try to make a diagnosis other than epilepsy. A variety of other conditions, some of them benign, can produce seizure-like events. With the use of cell-phone videos, parents can capture spells for your review. Avoid diagnostic trials of antiepileptic drugs, especially if the events are infrequent and non-threatening.
Sometimes, it is better not to place a definite label on a type of event. For example, staring spells or shaking that are infrequent and do not appear threatening can be managed as “paroxysmal events of uncertain etiology” until a specific diagnosis is made. Frequent or apparently life-threatening events call for referral to a specialist and sometimes admission for video-EEG monitoring.
Events that look like seizures
- Sleep myoclonus, or involuntary muscle jerks or twitches. This benign condition in infants occurs only during sleep, never involves the face and doesn’t waken the child. The twitches typically occur in the first hour of sleep and can go on for 30 minutes or longer. They can migrate between limbs and can vary in intensity. Waking the child during the event is an easy diagnostic test: If the movements stop, it’s not a seizure. This video taken by a parent shows benign sleep myoclonus:
- Parasomnias, including night terrors and somnambulation, can look like nocturnal frontal-lobe seizures. Children with night terrors may wake up in agitation, sit up in bed, scream, mumble, moan and sleepwalk, perspiring with a rapid heartbeat. Typically they do not arouse in full, do not appear to recognize their parents and don’t remember anything in the morning. Many parasomnias go away on their own and are often harmless, requiring only reassurance.
- Head banging. Parents may observe infants and young children rhythmically banging their head or legs or rocking their body during sleep or while falling asleep. This is more commonly seen with autism, developmental delay and abuse or neglect. It can last minutes to hours but may appear violent, but usually doesn’t cause injury.
- Post-feeding episodes. Extreme sleepiness after meals or during diaper changes are sometimes thought to represent seizures. Some infants may stare and become pale, flushed or cyanotic. An old term for this was “vagotonia,” signaling the excessive vagal response. Children with Sandifer syndrome may show discomfort or pain, stiffen and assume odd postures (twisting of the neck, arching of the back) after feeding to relieve discomfort from reflux. Empiric treatment or further GI investigation for GERD is recommended. This parent video shows what opisthotonus looks like (backward arching of the head, neck, and spine), though it’s unclear in this case whether reflux is the cause:
- Convulsive syncope. Fainting can be followed by convulsions and other motor phenomena (eye movements, head turning, flexion of upper limbs, atonia and non-rhythmic and non-sustained myoclonic or clonic movements). Some children even bite their tongue (typically the tip) and lose bladder control. Note that the confusion after these events (seeming dazed or briefly disoriented) is different from post-ictal symptoms (agitation, inability to arouse, oxygen debt).
Syncope has many causes. Explore whether the child had pre-syncopal symptoms (dizziness, graying/blacking out of vision, nausea, clamminess or cold sweat, tinnitus or hearing sounds as if remote, perceived leg weakness) and potential syncope triggers (heat, pain, etc.). If syncope is triggered by being startled or during exercise, think of cardiac arrhythmias such as long QT syndrome.
- Breath-holding spells: Similar to syncope followed by convulsions, sustained breath-holding can trigger an anoxic seizure in some infants and young children, typically after prolonged, intense crying. If parents observe this, assess the child for iron deficiency anemia. This parent video shows a reflex anoxic seizure triggered by unexpected pain:
Still other phenomena that can look like seizures
- Masturbation/self stimulation: Parents may observe a tightening of the legs and rhythmic movements, sometimes accompanied by decreased responsiveness, grunting, staring and facial flushing in children as young as 4 months, especially when the child is bored or stressed. The key isthat the “episodes” can be interrupted. They are a normal part of sexual development, but if very persistent, think of stress or abuse. Home videos can be very helpful in making the diagnosis.
- Shuddering attacks: Occasionally infants and children up to age 5 have episodes of rapid shivering. These typically last several seconds and involve the head, shoulders and sometimes trunk, as if someone poured cold water down the child’s back. Attacks can happen up to 100 times a day but are almost always benign — again, you can ask the parents to videotape the episodes. EEG is rarely needed:
- Stereotypies are seen at all ages, especially in neurologically impaired children. Parents may note staring, rhythmic movements (flapping, head-shaking, rocking), eye movements, tongue thrusting, tonic posturing and more. Stereotypies increase when a child is excited or upset, and may be comorbid with tic disorder, OCD and ADHD. They can be considered similar to tics, but should raisesuspicion if they correlate with cognitive decline or significant new school problems.
- Paroxysmal movement disorders: There is a long list to consider, including tics, episodic ataxia, paroxysmal torticollis, paroxysmal dyskinesias and dystonias, hyperekplexia, and conditions involving eye movements: paroxysmal tonic upgaze (generally benign) and oculo-gyric crisis, a painful dystonic eye movement that can be treated with anticholinergics or antihistamines.
- Pseudo-seizures and non-epileptic seizures occur in middle childhood and adolescence. Typically, these do not occur by conscious choice. More rarely, children may stage seizures as a way of acting out or seeking attention, or the events may be triggered by abuse or mental health issues. If so, children who receive appropriate and early mental health interventions have a good prognosis. Sometimes, however, pseudo-seizures can coexist with a true seizure disordes. Features that help distinguish between epileptic seizures and psychogenic non-epileptic seizures are found in the table below.
The bottom line
When it comes to seizures, the devil is in the details. When ordering an EEG, be prepared to accept false-positive ab
normal results. Put all test findings in the appropriate clinical context, and don’t be too quick to diagnose a child with epilepsy if you don’t have complete diagnostic certainty. If you do suspect epilepsy, try first to make the diagnosis clinically, and never try antiepileptic drugs as a means of making the diagnosis. A detailed history is crucial.
The Handbook of Pediatric Neurology, edited by Dr. Peters, has a chapter on Epilepsy and Paroxysmal Events with a detailed section on non-epileptic events.
To refer a patient or request a consult, please call 617-355-7970.