Secondary Logo

Share this article on:


Pinto, Daniel; Liotta, Eric; Maas, Matthew; Prabhakaran, Shyam; Holl, Jane; Naidech, Andrew

doi: 10.1097/01.ccm.0000528795.53927.7e
Research Snapshot Theater: Neuroscience

1Northwestern University, Chicago, IL, 2Northwestern, Chicago, IL, 3N/A, Chicago, IL, 4Northwestern Feinberg School of Medicine, Chicago, IL

Learning Objectives: Prophylactic seizure medications (usually levetiracetam) are administered to approximately 40% of patients with intracerebral hemorrhage (ICH). Several patient attributes (characteristics) are associated with an elevated risk of seizures (patient age < 65 years, hematoma volume > 10 mL, lobar hematoma location), but it is not known which have priority for the clinical decision to administer prophylactic seizure medication, or if other attributes have priority.

Methods: We performed an adaptive decision choice experiment from clinicians drawn from the NIH Stroke Trials Network. Selected attributes were known to be associated with an elevated risk of seizures in the CAVE Score (lobar hematoma, age < 65 years and hematoma location), a clinical manifestation of seizure (depressed consciousness), and patient race. Respondents were asked to choose between multiple pairs of attributes requiring a trade off as to which would be more appropriate to receive prophylactic levetiracetam (e.g., a patient with lobar hematoma and age > 65 years versus a patient of age < 65 years and hematoma volume > 10 mL) that adapted based on prior responses, based on the Potentially All Pairwise RanKings of all possible Alternative (PAPRIKA) algorithm. The IRB approved the study, and forbade collection of identifiers.

Results: Twenty-seven StrokeNet physicians completed the survey. Priorities were normally distributed. In descending order, lobar hematoma had the highest priority (30 ± 6%) followed by depressed Glasgow Coma Scale (24 ± 7%), hematoma volume > 10 mL (19 ± 4%), white ethnicity (14 ± 4%) and age > 65 years (13 ± 5%). Attribute weights were not correlated with each other by Spearman’s rank (P> 0.2 for all). (A second survey wave is planned for August, 2017.)

Conclusions: Among established risk factors for seizures after ICH, lobar hematoma location was over twice as important for the decision to administer prophylactic levetiracetam as other CAVE Score criteria, correlating with a recent publication on objective practice (Disparities in the Use of Seizure Medications After ICH, Stroke 2017). Future research on the use of seizure medications should account for lobar hematoma location and depressed consciousness because these attributes most influence decision-making. Adaptive decision choice experimentation is a useful method to determine which attributes have priority for clinical decision-making. Future research will identify clusters of physicians with different priorities.

Copyright © by 2018 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.