Institutional review of predictors of in-hospital and 30-day discharge mortality in neurosurgical patients.

Abstract Objective Risk of adverse events from neurosurgical diagnoses is high. It is not well described whether there are any demographic, admission, or discharge factors that are associated with inpatient or post-discharge mortality outcomes in neurosurgical patients. The aim of this study is to identify the differences in predictors of mortality during inpatient stay and within 30 days of discharge. Methods This was a single-institution, retrospective cohort analysis of mortality. Our patient cohort of 11,477 was defined as all adult patients who were discharged (dead or alive) from an inpatient stay between January 1, 2014, and December 31, 2018, and were either admitted to a neurosurgical service or underwent a neurosurgical procedure during that admission. Results Out of 11,477 patients, 224 (1.95%) and 290 (2.53%) died inpatient and within 30 days of discharge, respectively. In multivariate analysis, the independent predictors of inpatient mortality were older age, female gender, diagnostic group, high present on admission severity of illness (POA-SOI) and present on admission risk of mortality (POA-ROM), intensive care unit (ICU) care, and palliative care consult (all p  Conclusion Older age and palliative care consult were significant predictors of both inpatient and within 30 days of discharge mortality. Admission SOI (>3) and ROM (>3) and ICU care were significant predictors for inpatient mortality while discharge disposition (home health, skilled nursing facility) was important for 30-day mortality.