ResearchPod Summary
This clinical practice guideline addresses the use of prophylactic antiseizure medications (ASM) in adult patients hospitalized with acute, spontaneous, nontraumatic intracerebral hemorrhage (ICH). Despite high rates of clinical and electrographic seizures in this population, clinical practice remains highly heterogeneous. The Neurocritical Care Society (NCS) conducted a systematic review and meta-analysis using the GRADE methodology to evaluate three primary questions: (1) whether to use ASM prophylaxis at all, (2) which medication to prefer (levetiracetam vs. phenytoin/fosphenytoin), and (3) the optimal duration of treatment.
The panel suggests against the routine use of prophylactic ASM in patients with acute nontraumatic ICH. Meta-analyses indicated that prophylactic ASM use provides no significant benefit in preventing early or late seizures or reducing mortality. Conversely, ASM use was associated with a higher risk of adverse events and a small but significant increase in poor functional outcomes at 90 days. When ASM is deemed necessary, the guidelines suggest levetiracetam (LEV) over phenytoin/fosphenytoin (PHT/fPHT) due to a more favorable pharmacokinetic profile, fewer drug-drug interactions, and a lower burden of adverse effects. Furthermore, the panel suggests limiting the duration of prophylaxis to 7 days or less, as longer courses do not appear to reduce late seizure risk and may increase the likelihood of adverse outcomes.
These guidelines provide a standardized, evidence-based framework for a common clinical dilemma in neurocritical care. By recommending against routine prophylaxis and favoring shorter, safer medication regimens, the NCS aims to reduce unnecessary drug exposure and potential harm. The guidelines also highlight the critical need for better risk stratification, suggesting that clinicians should consider continuous EEG monitoring to identify high-risk patients who might truly benefit from targeted, short-term treatment, rather than relying on blanket prophylactic strategies.
Alex: Welcome to another episode of ResearchPod. Today we're looking at a new guideline from the Neurocritical Care Society on seizure prophylaxis in nontraumatic intracerebral hemorrhage — and the central finding is genuinely counterintuitive.
Sam: It is. The paradox is this: up to nearly a third of these patients experience subclinical seizures detectable only on continuous EEG — yet routine prophylactic antiseizure medication has consistently failed to improve patient-centered outcomes. The thing you're trying to prevent is real, but the intervention isn't helping.
Alex: So the paper is essentially asking whether a widespread clinical practice is being sustained by habit rather than evidence?
Sam: That's a fair read. The guideline uses GRADE methodology to systematically evaluate the literature, and the conclusion is that routine prophylaxis provides no clear benefit on seizure prevention or mortality — while correlating with increased adverse events and worse functional outcomes. That's not a null result. That's a signal pointing in the wrong direction.
Alex: If the seizures are real and the medication isn't helping, what's the mechanism behind that failure? Why doesn't treating the EEG finding translate to better recovery?
Sam: The core problem is that you're treating a surrogate endpoint. The EEG seizure is the measurable signal, but suppressing it doesn't address the underlying hemorrhagic injury driving poor outcomes. And the treatment itself carries costs — most notably sedation, which directly impairs the neurological recovery you're trying to protect. Functional status, measured here by the modified Rankin scale, gets worse, not better. So you have a certain harm competing against an uncertain benefit, and the certain harm is winning.
Alex: That framing — certain harm versus uncertain benefit — is doing a lot of work. What does the randomized evidence actually look like? Is this a power problem, or is the null result robust?
Sam: The authors describe it as therapeutic inertia built on a weak evidentiary foundation. Retrospective data did suggest some reduction in early seizures, which is probably where the practice took hold. But when you look at the randomized evidence, it's neutral on efficacy and the harm signal — worse functional status — is more consistent across studies than any benefit signal. That asymmetry is the load-bearing finding here. It's not that the trials were underpowered to detect an effect; it's that the effect, where it appears, points toward harm.
AI-generated third-party summary by ResearchPod. Not official content or an endorsement by the paper authors or affiliated organizations.
Alex: That's a difficult position for a clinician at the bedside — a patient with a large hemorrhage, high seizure risk, and the instinct to do something. How do the authors translate this into actionable guidance?
Sam: They make a fairly direct recommendation: avoid routine prophylaxis. If a clinician judges that the clinical picture warrants treatment, the guideline favors levetiracetam over older agents — largely on a tolerability basis — and caps duration at seven days. The logic is to minimize exposure to a treatment whose benefit hasn't been established.
Alex: So the shift is from blanket prevention to a much more selective posture.
Sam: Exactly. And the mechanism for that selectivity is risk stratification. The guideline points to tools like the 2HELPS2B score, which uses continuous EEG features to identify patients who actually have elevated seizure burden — the subgroup where intervention might be justified — rather than applying a uniform policy to everyone with a hemorrhage. The idea is to reserve treatment for patients where the EEG evidence gives you a real signal, not just a theoretical risk.
Alex: That raises a practical question the guideline may not fully resolve — continuous EEG monitoring isn't universally available. Does the paper address what to do when that infrastructure isn't there?
Sam: That's a genuine gap. The guideline is written from a neurocritical care center perspective, where prolonged EEG monitoring is feasible. For institutions without that capacity, the risk-stratification framework becomes harder to operationalize. The authors acknowledge the evidence base is limited — much of it retrospective, with heterogeneous populations — and that's a real constraint on how confidently you can generalize the recommendations.
Alex: So what's the honest bottom line for someone reading this paper?
Sam: The bottom line is that the evidentiary case for routine prophylaxis was always weaker than practice patterns suggested, and this guideline makes that explicit. The randomized evidence doesn't support it, the harm signal is real, and the path forward is precision — using EEG-guided risk stratification to identify the patients who might actually benefit, rather than treating the entire population prophylactically. For most patients with nontraumatic intracerebral hemorrhage, the evidence now says: don't start the medication in the first place.
Alex: That's a meaningful recalibration of a common practice — and a useful reminder that a detectable finding on a monitor isn't the same as a finding that should drive treatment. Thanks for walking us through it, Sam, and thanks to everyone listening to ResearchPod.