Factors Impacting Hydrocephalus Incidence in Intracerebral Hemorrhage: A Retrospective Analysis

Abstract

To test which intracerebral hemorrhage (ICH) characteristics impact incidence of hydrocephalus and characterize subsequent impact on outcomes. A search of the electronic medical record of Sinai Grace Hospital between January 2009 and April 2018 using International Classification of Diseases, Ninth Revision and Tenth Revision codes for ICH identified 847 patients. After excluding patients with hemorrhagic conversion of stroke, subarachnoid hemorrhage, and traumatic hemorrhage, 560 patients remained for analysis. Generalized linear modeling was used to assess variance in modified Rankin Scale (mRS) score and length of stay. Incidence of hydrocephalus on arrival varied with ICH volume (P < 0.001), intraventricular hemorrhage (IVH) status (P < 0.001), bleed location (P < 0.001), and external ventricular drain (EVD) status (P < 0.001). An EVD was inserted in 47% of patients presenting with IVH (n= 102/217), while 4% of patients without IVH received an EVD (n= 14/343) (P < 0.001). Hemorrhage locations had different rates of EVD placement: thalamic 43%, basal ganglia 22%, cerebellar 28%, brainstem 21%, lobar 7% (P < 0.001). Shunt dependency did not vary between bleed locations (P= 0.072). Variance in mRS score was explained by IVH, bleed location, hydrocephalus on arrival, and ICH volumes. In particular, cerebellar hemorrhage location was associated with better outcomes (mean discharge mRS score of 3.3 vs. 3.9, P < 0.001). Bleed characteristics affect incidence of hydrocephalus on admission, rates of long-term shunt dependency, and outcomes. Hemorrhage location did not predict shunt dependency; however, it did predict outcomes. Specifically, cerebellar ICH was associated with a better discharge mRS score.