Serum Lactate-Based Stratification for Seizure Diagnosis in Resource-Limited Neurologic Emergency Settings
Osawa, S.-i.; Miyata, S.; Fujita, K.; Kawamura, T.; Suzuki, I.; Kimura, K.; Okushima, T.; Konn, A.; Endo, H.
Show abstract
BackgroundEpileptic seizures (SZ) and postictal neurological deficits are common stroke mimics that challenge acute clinical decision-making, especially in emergency settings with limited diagnostic resources. MRI or EEG is often unavailable at initial presentation. We aimed to determine whether routinely available serum lactate levels could distinguish SZ from stroke based on information accessible immediately on hospital arrival. MethodsWe retrospectively reviewed 22,430 consecutive emergency department visits and analyzed 637 patients transferred by ambulance with suspected neurologic emergencies identified via keyword screening at the regional emergency call center. All patients underwent venous blood gas analysis and non-contrast CT on arrival. We evaluated whether serum lactate and other routine laboratory variables (pH, actual base excess, glucose, WBC count, platelet count, and PT-INR) correlated with final neurological diagnoses. Classification and Regression Tree (CART) analysis was used to identify cutoffs predictive of SZ. ResultsAge, pH, lactate, and actual base excess were significantly associated with final diagnoses. Lactate levels were significantly higher in SZ and subarachnoid hemorrhage groups than in other groups. For distinguishing SZ from ischemic stroke (infarction and TIA), CART analysis yielded a serum lactate cutoff of 4.05 mmol/L. Among patients with lactate [≥]4.05 mmol/L and age <59.5 years, SZ probability was 100%. A four-quadrant model combining lactate and age stratified SZ probabilities from 9.1% to 100%. The area under the ROC curve for lactate in distinguishing SZ from stroke was 0.800. ConclusionsSerum lactate, when assessed upon emergency department arrival, may contribute significantly to differentiating seizures from acute ischemic stroke. A simple decision model using only lactate and age may aid in diagnostic stratification in neurologic emergencies--even in resource-limited settings where advanced imaging or EEG is not readily available. RegistrationNone.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Predictors of acute ischemic cerebral lesions in immune-mediated thrombotic thrombocytopenic purpura and hemolytic uremic syndrome 95%
- Endothelial receptor proteins in acute venous thrombosis and delayed thrombus resolution in cerebral sinus vein thrombosis 94%
- Serum anti-NMDA-receptor antibodies and cognitive function after ischemic stroke (PROSCIS-B) 93%
Similar papers in this journal
- Time is Brain: Detection of Nonconvulsive Seizures and Status Epilepticus During Acute Stroke Evaluation Using Point-of-Care Electroencephalography 96%
- Modified Rankin Scale Disability Status at Day 4 Poststroke is an Informative Predictor of Long-Term Day 90 Outcome 94%
- White matter lesions as a prognostic marker of recurrence in cryptogenic stroke with high-risk patent foramen ovale 94%
Similar papers in this journal
Similar papers in this journal
- Systemic Metabolic Alterations after Aneurysmal Subarachnoid Hemorrhage: A Plasma Metabolomics Approach 94%
- Isofurans and isoprostanes as markers of delayed brain injury and mitochondrial dysfunction following aneurysmal subarachnoid hemorrhage. A prospective observational study 93%
- Cerebral microbleeds in critically ill patients with respiratory failure or sepsis: a scoping review 93%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.