Mortality and functional outcomes in patients with post-stroke epileptic seizures: a systematic review and meta-analysis
Shubham Misra1, Juan Vazquez2, Ece Eldem1, Lucas Scardua Silva3, Saba Mohidat4, L. Brian Hickman5, Erum Khan6, Melissa Funaro1, Clarissa Yasuda3, David Liebeskind5, Scott Kasner7, Nishant K. Mishra1
1Yale University, 2Albert Einstein College of Medicine, 3University of Campinas, 4The Royal Melbourne Hospital, 5University of California Los Angeles, 6B.J.Medical College,Ahmedabad, 7University of Pennsylvania
Objective:
To conduct a meta-analysis of published data to comprehensively investigate clinical and functional outcomes in post-stroke epilepsy (PSE) patients.
Background:
It is critical to highlight the burden of PSE on patient outcomes. Whereas PSE has been reported to worsen outcomes in stroke patients, published data are inconclusive.
Design/Methods:
We used Covidence to conduct a literature search of studies published until 1-March-2022. We used Joanna Briggs Institute tool for cohort studies to assess the risk of bias. Our outcome measures include mortality and poor outcomes (Rankin Scale score 3-6). We report Odds ratio (OR) and 95% Confidence Intervals (CI) using a random-effect meta-analysis. Publication bias was assessed using Egger’s test. We conducted the statistical analyses in R version 4.2.0. (PROSPERO ID: CRD42022308648)
Results:
We screened 3721 studies and included 71 articles (N= 29,759 patients with PSE; 17,42,083 patients without PSE). Articles reported 3938 early and 13,067 late seizures; 12,754 seizures were not classified as early/late. Risk of bias was high in two studies, moderate in 36 and low in 33. PSE was associated with greater odds of mortality (OR 2.09; CI 1.78-2.45; N= 52 studies) and poor outcomes (OR 2.39; CI 1.91-3.00; N= 19 studies). In subgroup analyses, post-ischemic stroke epilepsy was associated with mortality (OR 2.31; CI 1.88-2.83) and poor outcome (OR 2.60; CI 1.91-3.52). Post-hemorrhagic stroke epilepsy showed consistently significant OR for the poor outcome (OR 2.04; CI 1.33-3.13) but failed to reach statistical significance for mortality (OR 1.29; CI 0.96-1.74). No significant publication bias was observed for mortality (p=0.93) and poor outcome (p=0.17).
Conclusions:
Our meta-analysis suggests that PSE is associated with a doubled risk of death and severe disability. Prevention of PSE should be a high research priority. The role of stroke severity or lesion volume also requires further study.