Abstract
BACKGROUND: Post-traumatic epilepsy (PTE) is a recognized complication of traumatic brain injury (TBI), yet its risk following mild and moderate TBI remains underappreciated. Although mild TBI represents the majority of cases in clinical practice, a subset of patients develop unprovoked seizures months or even years post-injury. This review aims to synthesize current evidence on the incidence and predictors of PTE in mild and moderate TBI and to propose a clinically actionable decision-support tool for early risk stratification. METHODS: We performed a narrative review of peer-reviewed studies published between 1985 and 2024 that reported on the incidence, risk factors and predictive models of PTE in patients with mild (Glasgow Coma Scale [GCS] 13-15) and moderate (GCS 9-12 or imaging-positive) TBI. Data from 24 studies were extracted, focusing on neuroimaging findings, early post-traumatic seizures, EEG abnormalities and clinical risk factors. These variables were integrated into a rule-based algorithm, which was implemented using Streamlit to enable real-time clinical decision-making. The decision-support tool incorporated five domains: injury severity, early post-traumatic seizures, neuroimaging findings (including contusion location and hematoma type), clinical and demographic variables (age, sex, psychiatric comorbidities, prior TBI, neurosurgical intervention) and EEG abnormalities. RESULTS: PTE incidence following mild TBI ranged from <1% to 10%, with increased risk observed in patients presenting with intracranial hemorrhage or early seizures. From moderate TBI, incidence rates were consistently higher (6-12%). Key predictors included early seizures, frontal or temporal contusions, subdural hematoma, multiple contusions and midline shift. Additional risk-enhancing factors included prolonged loss of consciousness, male sex, psychiatric comorbidities and abnormal EEG patterns. Based on these features, we developed a decision-support tool that stratifies patients into low-, moderate- and high-risk categories for developing PTE. CONCLUSIONS: Even in non-severe cases, patients with mild and moderate TBI who exhibit high-risk features remain vulnerable to long-term epileptogenesis. Our proposed tool provides a pragmatic, evidence-based framework for early identification and follow-up planning. Prospective validation studies are needed to confirm its predictive accuracy and optimize its clinical utility.