The Chart Before the Laboratory: Diagnostic Stewardship in Pediatric Breakthrough Seizures

Abstract

Children with established epilepsy with breakthrough seizures are frequently seen in emergency departments (EDs) and in inpatient services. There is often a reflex from the frontline providers toward a broad workup, but the evidence suggests it is frequently excessive.In one cohort of pediatric breakthrough seizure visits, laboratory testing was performed in nearly 87% of encounters. While abnormal results were common, they changed acute management in roughly 3% of presentations.1 A separate pediatric seizure cohort similarly concluded that the history preceding the event should guide which laboratory studies are obtained, and not the characteristics of the seizure itself.2 For neuroimaging, the American Academy of Neurology’s evidence-based review found that emergency computed tomography (CT) changes acute management in only about 3% to 8% of children presenting with seizures. They also found that an abnormal neurologic examination, a predisposing history, or a focal onset predicted a meaningful scan.3Diagnostic stewardship may provide a more useful framework for determining which tests to order.4 The same principle anchors broader value-based frameworks for acute pediatric seizure care.5 For this edition of Bending the Value Curve, we present an illustrative composite case in which the highest-value diagnostic step is a structured look at the chart.This case is constructed for educational purposes and does not describe an identifiable patient.A 12-year-old boy with focal epilepsy presents to the ED after the parents report 3 breakthrough seizures over the preceding day. The events involved right arm stiffening with impaired awareness, followed in 2 instances by secondary generalization with full-body tonic-clonic activity lasting under 2 minutes. The child is maintained on maximum doses of lamotrigine, levetiracetam, and clobazam. Family endorses adherence to the current antiseizure medication regimen. The last dose adjustment was a weight-adjustment of the levetiracetam 3 months ago. The child is not febrile in the ED, and the family denies any recent illness. The parents say, “There has been a change in the seizures—there’s more of them, and the last one looked bigger than usual.”In the ED, the child is awake, alert, and at his neurologic baseline by the time of evaluation. The team’s initial plan includes a comprehensive metabolic panel, complete blood count, ammonia, magnesium, calcium, lactate, urine toxicology, a stat electroencephalogram (EEG), a noncontrast head CT, an empirical levetiracetam load, and admission for observation.Before they order this, the bedside team opens the chart and reads the most recent pediatric neurology note from 3 months ago. The note describes the child’s typical seizures as focal seizures of the right arm stiffening with impaired awareness, with occasional secondary generalization, occurring 2 to 5 times per week. A video-EEG from the prior calendar year captured several of these events, confirming the events as seizures. The neurologist’s last note documents an explicit shared decision with the family: tolerating 2 to 5 seizures per week was preferred by the family rather than adding a fourth antiseizure medication (ASM), because earlier polytherapy trials had been associated with sedation and worsening school performance.The initial plan would have committed the child to costly testing, whose overall yield in this setting is low.1,2 Testing would have included a CT scan with ionizing radiation and a suspected 3% to 8% chance of altering management in a child without new focal findings,3 and an EEG, ASM load, and admission of a child who was already back to baseline. None of it would have answered a clinical question. Escalation in this setting may also have gone against the family’s expressed goals of care for ASM management in their last clinic visit.Most clinicians know the chart matters, and the escalation of tests may be more structural and psychological than negligent. Order sets that bundle a “seizure panel” as a single click, the time pressure of a full department, and overall clinician discomfort with seizures that pushes toward testing as a hedge against uncertainty may be the biggest drivers of over-testing.6,7 A structured approach can help frontline clinicians manage this uncertainty.Prior long-term EEGs document which clinical events have an EEG correlate (ie, which events are seizures). Prior neurology notes often describe the child’s typical semiology and frequency. For a child with epilepsy controlled on only one ASM, the goal seizure frequency is often zero. For a child who is taking 3 or more ASMs, the goal may have shifted to a balance between seizure burden and antiseizure medication-related side effects. In the latter case, breakthrough seizures congruent with prior typical seizures, in a child who has returned to baseline, at a frequency at or below the documented baseline, do not always represent a need to escalate treatment. With this in mind, a family always has the right to change their minds on what level of seizure burden is acceptable, and should be asked pointedly in a child with a high seizure burden on several ASMs if they want to make adjustments to the ASM regimen before calling a neurologist under the presumption that they do. In contrast, the presentation features that should always prompt more testing or treatment escalation arise if a child with breakthrough seizures has not returned to baseline, has a new focal deficit, injury from a seizure, a frequency clearly exceeding the documented baseline, or has a semiology incongruent with their prior seizure etiology. A semiology’s congruence with the prior seizure etiology helps the provider qualify a family’s comment that a seizure looks different. If a child with epilepsy secondary to a prior left-sided middle cerebral artery stroke were to have a seizure with semiology described as right arm stiffening with loss of awareness, which could generalize and become generalized tonic-clonic, a description of a similar seizure semiology except “longer” or with “more intense drooling than normal” should not generally be regarded as a new/unexpected seizure semiology. In contrast, a description of a focal seizure (eg, right arm stiffening or jerking followed by whole body shaking or loss of consciousness) in a child with a reported generalized epilepsy syndrome (eg, juvenile myoclonic epilepsy) should be treated as a meaningfully new or unexpected semiology. This could trigger additional diagnostics such as neuroimaging (to look for evidence of a new or evolving intracranial process) or EEG (to provide evidence toward a focal or generalized epilepsy), with the clinical context determining which of these workups would best be performed on an outpatient basis.Treatment should never be delayed for a level draw in status epilepticus, airway compromise, or hemodynamic instability. In these situations, the American Epilepsy Society algorithm takes precedence.8 For the stable child, a preloading ASM level—drawn before any additional medication is given—can answer a defined question. In some pediatric cohorts, persistent nonadherence has been documented in more than half of children with newly diagnosed epilepsy within the first 6 months of therapy,9 and the breakthrough seizure visit is often the moment that pattern becomes visible. A subtherapeutic level in a child whose dose has not changed may be the entire explanation for the visit. Even a level that returns after discharge has value in future encounters to inform the outpatient neurologist’s next decision.10 When laboratory work is indicated beyond obtaining ASM levels, it should be matched to the medication and to a clinical question. Examples of targeted labs include ammonia and liver enzymes for a child on valproic acid with new lethargy or asterixis11 and serum sodium level for a child on oxcarbazepine with unexplained confusion or vomiting.12 In contrast, the yield of a generic “seizure panel” in this population is low.1,2Urgent EEG earns its place when the child does not return to neurologic baseline within the expected postseizure window. Ask a family how long their child is usually confused and/or sleepy after a seizure and/or after receiving rescue ASM. Confusion or sleepiness beyond this time raises the concern for ongoing electrographic seizures (also called nonconvulsive status epilepticus). This is a good example of when an EEG can answer a pointed question (namely, is this child presently experiencing nonconvulsive status epilepticus).13 If the child is at neurological baseline and the event is congruent with typical seizure semiology, a recent outpatient EEG often answers the relevant question—whether anything has materially changed—better than a new acute recording. An EEG obtained during an acute encounter is not a neutral snapshot of habitual seizure burden. Instead, the EEG can underestimate seizure burden after benzodiazepine exposure and can overestimate seizure burden during an unusually active illness window. This is often a meaningful statement to anticipatorily give the family, as they may wonder why neurology does not want an EEG during every hospital encounter for a breakthrough seizure. Imaging follows the same logic. In established epilepsy, an emergent CT obtained during a seizure visit shows an acute abnormality in only about 8% of scans, and in patients without a history of brain tumor or seizure-related head trauma, that figure falls below 3%.14 Pediatric data specific to breakthrough seizures remain limited, and these particular figures derive largely from adults; they nonetheless describe the closest matched population to the child in this case. However, new focal deficits, sustained altered mental status, trauma, a new/unexpected semiology, or concern for an acute intracranial process move imaging from optional to indicated.The team did not order the comprehensive panel, the EEG, the head CT, or the (prelevel) levetiracetam load. They drew levetiracetam, lamotrigine, and clobazam levels (the child was clinically stable). The child was observed in the ED for a few hours and remained at neurologic baseline. The child’s levetiracetam level (which was a rapid level at this institution) returned during the encounter as below the detectable threshold. The child was then loaded with levetiracetam, placed back on his previously charted levetiracetam dosage, and counseled on the importance of medication adherence. The family left the ED after significantly fewer tests were ordered, with a clear explanation for why there were breakthrough seizures, what could be done to address the breakthrough seizures that were within their expressed goals of care, and a clear rationale for why more labs, imaging, EEG, admission, and a new ASM were not pursued.Undirected testing after a breakthrough seizure is frequently low yield: laboratory panels and emergency CT change management in a small percentage of children. Instead, let presentation features—return to baseline, focal deficit, new/unexpected semiology, injury, frequency above baseline—rather than the seizure itself, decide what to order.1,2,3For a child with established epilepsy who is back to baseline, the most valuable first diagnostic act is often a structured look at prior neurology notes, prior EEG, documented baseline frequency, and the family’s stated goals of care.ASM levels are most useful when ordered prospectively and matched to a defined question; even a level that returns after discharge can change subsequent care by documenting adherence.Urgent EEG and acute imaging earn their place by answering questions that change management. A recent representative prior EEG often answers the baseline question better than a new acute recording, provided the child has returned to baseline, and there is thus no concern for nonconvulsive status epilepticus. New neuroimaging is most useful when history and/or examination suggest a new or evolving intracranial process.Is the child back to neurologic baseline?Is this event outside the child’s documented typical/expected seizure semiology?Is the current frequency above the documented baseline?Is there a new focal deficit, injury, or sustained alteration in mental status?Have prior video-EEG or neurology notes already characterized this event?Are there established therapeutic goals of seizure duration and/or frequency—agreed with the family and outpatient team—that limit further ASM escalation? Does the family want to readdress this?Would an ASM level, a targeted medication-specific laboratory test, an EEG, or imaging change the next clinical decision?Is treatment urgent enough that chart review or level draws should not delay care (status epilepticus, airway compromise, hemodynamic instability)?