1093. Factors Associated with Increased In-Hospital Mortality Following Aneurysmal Subarachnoid Hemorrhage

Authors: Jonathan Christian Dallas; Campbell Liles, BS; Stephen Gannon, BS; Chevis Shannon, MBA, MPH, DrPH; Rohan Chitale, MD; Matthew Fusco, MD (Nashville, TN)

Introduction: Aneurysmal subarachnoid hemorrhage (SAH) is a devastating yet unpredictable disease associated with exceedingly high morbidity and mortality. However, outcomes are also highly variable. Prior studies searching for predictors of negative outcomes have been limited by sample size or population characteristics. The purpose of this study is to identify predictive factors associated with increased in-hospital mortality following aneurysmal SAH in a diverse and representative patient population. Methods: Patients hospitalized for aneurysmal SAH treatment were identified from the 2012-2015 National Inpatient Sample (NIS). Inclusion criteria selected for patients with (1) subarachnoid hemorrhage (ICD9 diagnosis: 430) and (2) surgical aneurysm repair (clipping or endovascular; ICD9 procedure: 3951, 3952, 3972, or 3979). Exclusion criteria removed patients with a zero-day length of stay, head trauma, or AVM/AVF. Independent variables included sociodemographics, hospital characteristics, treatment type, neurovascular comorbidities index (NCI), and the NIS-SAH severity score (NIS-SSS, a previously validated synthetic Hunt and Hess scale equivalent). Missing data was imputed via the “multivariate imputation by chained equations” method. The primary outcome was in-hospital mortality, a dichotomous variable. Analysis of each predictive variable was performed via both univariate and multivariate logistic regression, and associated odds ratios (OR) were obtained. Results: 5355 unweighted patients met the inclusion criteria. The overall in-hospital mortality rate was 12.1%. Following multivariate logistic regression, variables predictive of increased in-hospital mortality included increased age ( OR=1.02/year, P<0.001 ) and NIS-SSS ( OR=1.30/point, P<0.001 ), whereas factors predictive of decreased mortality included African-American race ( OR=0.67, P=0.010 ), private insurance ( OR=0.73, P=0.021 ), urban teaching hospital status ( OR=0.32, P=0.035 ), transfer from an acute care hospital ( OR=0.080, P=0.024 ), and increased NCI ( OR=0.94/point, P<0.001 ). Conclusion: Multivariate logistic analysis identified a number of significant predictors of in-hospital mortality following aneurysmal SAH. These predictors can be used to both better determine patient prognosis and improve outcomes by addressing potentially modifiable variables, such as patient transfer status.