Abstract
BACKGROUND: Subarachnoid hemorrhage (SAH) represents a severe stroke subtype. Our study aims to develop gender-specific prognostic prediction models derived from distinct prognostic factors observed among different-gender patients. METHODS: Inclusion comprised SAH-diagnosed patients from January 2014 to March 2016 in our institution. Collected data encompassed patients' demographics, admission severity, treatments, imaging findings, and complications. Three-month post-discharge prognoses were obtained via follow-ups. Analyses assessed gender-based differences in patient information. Key factors underwent subgroup analysis, followed by univariate and multivariate analyses to identify gender-specific prognostic factors and establish/validate gender-specific prognostic models. RESULTS: A total of 929 patients, with a median age of 57 (16) years, were analyzed; 372 (40%) were male, and 557 (60%) were female. Differences in age, smoking history, hypertension, aneurysm presence, and treatment interventions existed between genders (p < 0.01), yet no disparity in prognosis was noted. Subgroup analysis explored hypertension history, aneurysm presence, and treatment impact, revealing gender-specific variations in these factors' influence on the disease. Screening identified independent prognostic factors: age, SEBES score, admission GCS score, and complications for males; and age, admission GCS score, intraventricular hemorrhage, treatment interventions, symptomatic vasospasm, hydrocephalus, delayed cerebral ischemia, and seizures for females. Evaluation and validation of gender-specific models yielded an AUC of 0.916 (95% CI: 0.878-0.954) for males and 0.914 (95% CI: 0.885-0.944) for females in the ROC curve. Gender-specific prognostic models didn't significantly differ from the overall population-based model (model 3) but exhibited robust discriminative ability and clinical utility. CONCLUSION: Variations in baseline and treatment-related factors among genders contribute partly to gender-based prognosis differences. Independent prognostic factors vary by gender. Gender-specific prognostic models exhibit favorable prognostic performance.