Acute Posterior Cranial Fossa Hemorrhage—Is Surgical Decompression Better than Expectant Medical Management?

BACKGROUND To compare the in-hospital mortality and institutional morbidity from medical therapy (MT), external ventricular drainage (EVD) and suboccipital decompressive craniectomy (SDC) following an acute hemorrhagic posterior cranial fossa stroke (PCFH) in patients admitted to the neurosciences critical care unit (NCCU). Retrospective observational single-center cohort study in a tertiary care center. All consecutive patients (n = 104) admitted with PCFH from January 1st 2005-December 31st 2011 were included in the study. METHODS All patients with a PCFH were identified and confirmed by reviewing computed tomography of the brain reported by a specialist neuroradiologist. Management decisions (MT, EVD, and SDC) were identified from operative notes and electronic patient records. RESULTS Following a PCFH, 47.8 % (n = 11) patients died after EVD placement without decompression, 45.7 % (n = 16) died following MT alone, and 17.4 % (n = 8) died following SDC. SDC was associated with lower mortality compared to MT with or without EVD (χ 2 test p = 0.006, p = 0.008). Age, ICNARC score, brain stem involvement, and hematoma volume did not differ significantly between the groups. There was a statistically significant increase in hydrocephalus and intraventricular bleeds in patients treated with EVD placement and SDC (χ 2 test p = 0.02). Median admission Glasgow Coma Scale scores for the MT only, MT with EVD, and SDC groups were 8, 6, and 7, respectively (ranges 3-15, 3-11 and 3-13) and did not differ significantly (Friedman test: p = 0.89). SDC resulted in a longer NCCU stay (mean of 17.4 days, standard deviation = 15.4, p < 0.001) and increased incidence of tracheostomy (50 vs. 17.2 %, p = 0.0004) compared to MT with or without EVD. CONCLUSIONS SDC following PCFH was associated with a reduction in mortality compared to expectant MT with or without EVD insertion. A high-quality multicenter randomized control trial is required to evaluate the superiority of SDC for PCFH.