Study population

A retrospective review of all patients with aneurysmal SAH admitted to the Mount Sinai Medical Center between July 2000 and March 2009 was conducted. Patients with International Classification of Diseases, ninth revision, Clinical Modifications (ICD-9-CM) code 430 were identified in the Mount Sinai Neurosurgery Department physician database. A chart and head CT review was conducted to confirm the diagnosis of aneurysmal SAH in all subjects. Inclusion criteria for analysis were defined as: aneurysmal SAH, aneurysm repair via either clipping or coiling, presence of an ICH ≥30 ml or ICH with midline shift ≥5 mm, and ICH evacuation via craniotomy. Patients who underwent aneurysm coiling followed by ICH evacuation were compared with patients who underwent simultaneous aneurysm clipping and ICH evacuation. Exclusion criteria were: traumatic SAH, SAH secondary to vascular dissection, vasculopathy, arteriovenous malformation or other non-aneurysmal causes, and age <18 years. The study was approved by the Mount Sinai institutional review board.

Clinical management

All patients received nimodipine every 4 h for 21 days and phenytoin for 7 days for seizure prophylaxis. All patients received 0.9% normal saline at a rate of 1 ml/kg/h to maintain euvolemia. Aminocaproic acid was administered to prevent aneurysm rebleeding in patients who were within 72 h of aneurysm rupture and had no history of stroke, venous thromboembolism, myocardial infarction or hypercoagulable state. Aminocaproic acid was stopped within 1–3 h of angiography to minimize the risk of catheter induced spasm. Potential complications of aminocaproic acid include clot formation, focal ischemic deficits and vasospasm (either spontaneous or catheter induced). Early studies using tranexamic acid to prevent aneurysm re-rupture detected higher rates of infarct if continued through the vasospasm period.12 More recent randomized trials of tranexamic acid limited to the first few days after SAH detected reduced rebleeding rates without significant differences in delayed cerebral ischemia.13 ,14 An external ventricular drain was placed if the CT scan demonstrated acute hydrocephalus and/or the patient was non-command following or worse, or had a deteriorating neurological examination due to hydrocephalus. Mannitol (20%, 1 g/kg) or 23% hypertonic saline (30 ml) was administered if there was suspected clinical or radiographic signs of herniation and/or elevated intracranial pressure (ICP). Continuous hypertonic saline infusion (2% or 3%) was administered at a rate of 1 mg/kg/h and titrated to maintain a serum sodium of 150–155 mEq/dl or 20% mannitol 0.5 g/kg was administered every 6 h to maintain a serum osmolality of 300–320 mosmol/l if ICP was elevated. Serum sodium and/or osmolality were checked every 6 h to titrate osmotic therapy. Prior to aneurysm repair, systolic blood pressure was maintained below 160 mm Hg with nicardipine or labetolol continuous infusion.

Angiographic management

All patients underwent digital subtraction cerebral angiography under general anesthesia to characterize the aneurysm and assess for potential treatment with primary coil embolization with either Gugliemi detachable coils (Boston Scientific, Massachusetts, USA) or Orbit coils (Codman, Massachusetts, USA). The goal of treatment, whether endovascular or open, was complete aneurysm obliteration. The preferred strategy for treatment of aneurysmal SAH at our institution, including in the setting of ICH, was to coil embolize first, regardless of clinical or radiographic herniation, which was always initially treated with mannitol or hyptertonic saline and other conservative measures. Surgical hematoma evacuation with clipping was performed when aneurysm characteristics did not allow for endovascular protection (ie, wide neck, large aneurysm size) and when there were logistical concerns—namely, when the interventionalist was unavailable or there were anticipated delays in the angiography suite.

Surgical management

Patients were brought emergently from the angiography suite following coil embolization or diagnostic angiography to the operating room where all patients underwent an extended pterional craniotomy. A large curvilinear scalp flap was created followed by a fronto-temporal or fronto-temporal-parietal bone flap removal. The dura was widely incised in a cruciate fashion. Careful hematoma removal was performed where accessible. Surgical obliteration of the aneurysm, if not coiled, was performed with clipping. Partial lobectomy and/or craniectomy was performed based on intraoperative evaluation of brain edema. When significant brain swelling and edema were encountered, the bone flap was marsupialized in the abdominal wall. At the end of each case the dura was loosely reflected onto the brain surface, and dural substitutes or Gelfoam (Pfizer, New York, USA) was placed over it. An epidural drain was placed followed by galeal and scalp closure.

Postoperatively, all patients were managed in the neurointensive care unit with frequent neurologic examinations and/or ICP monitoring. All patients who had signs or symptoms of vasospasm were initially treated with hypertensive–hypervolemic therapy followed by angioplasty and/or intra-arterial vasodilator injection, if required. Patients who tolerated clamping of an external ventricular drain for >24 h underwent drain removal; otherwise, a ventriculo-pertioneal shunt was placed. After discharge, patients were followed in an outpatient setting by the primary neurosurgeon and interventionalist.

Clinical and radiographic measures

Baseline demographic data including age and gender, and admission radiographic data including modified Fisher score,15 ,16 ICH score, ICH volume (calculated using the ABC/2 method),17 ICH location, presence of intraventricular hemorrhage, midline shift (measured from the septum pellucidum at the level of the pineal gland) and aneurysm size were recorded. Admission neurological status was evaluated with the Hunt–Hess Scale18 and Glasgow Coma Scale.19 Clinical and radiographic evidence of uncal herniation was documented. Total anesthesia, angiography and operating room time were assessed using the anesthesia record. The time at presentation was the time the patient was first registered at Mount Sinai Hospital. Transfer times from outside hospitals were not included in these calculations. All angiograms and surgeries occurred at Mount Sinai Hospital. Treatment complications were identified through a chart audit.

Outcome measures

Discharge disposition was dichotomized as poor (expired, discharged to subacute care facility, chronic care facility, hospice or a skilled nursing facility) or good (discharged home, to an acute rehabilitation center, home with a healthcare aide or to an acute care facility). Three month GOS score was dichotomized as poor (expired or severely disabled: GOS 1–3) or good (GOS 4–5). The Mount Sinai Hospital data management services provided information on hospital and ICU length of stay (LOS) and hospitalization costs categorized as: total direct costs, overhead, ICU, surgical and imaging/angiography costs. All cost data were adjusted for inflation to $2009.

Statistical analysis

Demographic and radiographic characteristics, timing of procedures, clinical outcomes, LOS and cost were compared between patients who underwent aneurysm clipping and ICH evacuation and patients who underwent coiling with ICH evacuation using the Mann Whitney U non-parametric test for continuous non-normally distributed variables and the Fisher exact test for dichotomous variables. Logistic regression analysis was performed to determine predictors of dichotomized functional outcomes. All analyses were performed using SPSS V.17. Significance was set at p<0.05.