Seizure incidence of angiogram-negative subarachnoid hemorrhage: An updated meta-analysis

BACKGROUND: To provide updated information on seizure events and patterns in patients with angiogram-negative subarachnoid hemorrhage based on the initial hemorrhage patterns: perimesencephalic subarachnoid hemorrhage (PMH) vs non-PMH. METHODS: A review of online database literature from January 1990 to November 2017 was systematically performed. In case of heterogeneity <50%, a fixed effect model was used. Publication bias was determined using Begg funnel plot and the trim-and-fill method. RESULTS: A total of 9 studies with 645 patients were included for final analysis after excluding one study without any seizure within either cohort. PMH patients had lower seizure rates (odds ratio, 0.393; 95% CI, 0.158-0.978) compared with non-PMH patients. The funnel plot showed a relatively asymmetric pattern, suggesting possible publication bias. After correction of the forest plot, the adjusted odds ratio was 0.362 (95% CI, 0.148-0.886), indicating significant relationships between PMH and lower incidence of seizure. CONCLUSION: PMH is associated with lower seizure risk than non-PMH. However, possible publication bias could be a concern to the interpretation. Additional meta-analyses based on individual patient data from prospective large-scale studies are necessary.