Admission Bio-clinical Score to Predict One-year Outcomes in Coiled Subarachnoid Hemorrhage (SAH)

The investigators plan to screen all patients who were consecutively admitted from January 1, 2003 to our neurosurgical intensive care unit after a clinical diagnosis of SAH confirmed by cerebral angiography and treated with a coiling procedure. Patients for whom a decision was made by the treating physician to either forego any invasive treatment or to perform open surgical clipping were excluded. Clinical, biological and radiological characteristics will be recorded during the course of hospitalization. This observational study will be performed according to standard procedure.

Admission characteristics. At admission, the investigators plan to anonymously record age, sex, GCS, presence of motor deficit, presence of clinical seizure and WFNS score. The GCS refers to the value at admission before any treatment with sedative drugs. The cohort was divided into two groups, an admission coma group for patients with GCS<13 (WFNS 4 or 5), and a non-coma group for patients with GCS313 (WFNS 1 to 3).

Admission biological sampling of venous blood are routinely performed to measure S100beta and troponin per standard clinical practice in our ICU. Troponin I and S100beta levels will be considered high for values higher than 0.5 mg/L (corresponding to 5 times the maximum normal range) to be consistent with previous studies and to ensure high specificity.

Admission radiological characteristics are obtained from the admission CT-scan and the initial angiogram. The CT-scan will be reviewed by a neuro-radiologist and classified according to the modified Fisher score. Hydrocephalus and intraventricular hemorrhage at admission were also registered. Aneurysm site and number of aneurysms were obtained from the initial angiogram.

ICU Management. The ICU management will not be modified for this study. Briefly, a central venous line and an arterial catheter are routinely inserted at admission before coiling. An external ventricular drain (EVD) is inserted in patients with CT evidence of hydrocephalus, WFNS grade between III to V, or a transcranial Doppler pulsatility index greater than 1.2, suggestive of increased intracranial pressure. The EVD is routinely connected to an external pressure strain gauge to monitor ICP according to a recently published protocol.

After coiling, systolic arterial blood pressure is maintained between 130 and 150 mm Hg by a titration of IV norepinephrine. ICP elevation higher than 20 mm Hg is treated by cerebra-spinal fluid drainage, increased minute volume by mechanical ventilation when not already implemented, deepening of sedation and, rarely, moderate hypothermia (target core temperature between 35.5 to 36.5°C). An additional CT was performed whenever the clinical status deteriorated. Acetaminophen and insulin are used as needed to avoid hyperthermia and hyperglycemia, respectively. For ventilated patients, arterial carbon dioxide partial pressure (PaCO2) was maintained between 35 and 40 mm Hg and peripheral oxygen saturation (SPO2) above 97%, when deemed possible. Oral or enteral nutrition was started as soon as possible. All patients received seizure prophylaxis from admission.

Cerebral Vasospasm management strategy. The investigators used our standard strategy for cerebral vasospasm management. To summarize, all patients receive IV nimodipine at 2 mg/hr from admission until the end of the second week after admission, except during periods of uncontrolled hypoxia and ICP elevation. Patients are switched to oral nimodipine (60 mg orally every 4 hrs) at the time of ICU discharge to the ward, for a total treatment period of 21 days. The neurologic and hemodynamic status are assessed at least every 4 hours. TCD was performed at least once a day during the first 10 days. In unsedated patients, in the event of clinical deterioration, new symptoms (uncontrolled cephalalgia, confusion, seizure, motor deficit), mean TCD velocities above 120 cm/s or a greater than 50 cm/s change in mean TCD velocity (ref) or after an increase in S100(, cerebral angiogram was performed. As is standard practice, cerebral angiogram was performed after CT-scan ruled out other complications. In sedated patients, since clinical symptoms are not able to be assessed; only the TCD and S100( criteria were used to evaluate the need for cerebral angiogram. Each vasospasm episode is treated by intra-arterial nimodipine administration, followed by intra-arterial milrinone 2 mg if necessary (ref), and repeated each day if necessary. Mechanical angioplasty is used as second-line treatment when necessary. Additionally, hypertensive therapy is reinforced with a target SBP between 160 and 180 mm Hg. Occurrences of vasospasm and of ischemic vasospasm (neurological symptoms directly linked to artery spasm) were recorded.

Clinical Outcome. The primary outcome is systematically assessed using the Rankin outcome scale at ICU-discharge and then 6 months and 1 year later. The GCS, WFNS, and the Fisher scores will be compared to our new score for the prediction of 1-year mortality and 1-year full recovery (defined as Rankin 0 to 1).

Statistical Analysis. To identify factors that independently predicted 1-year mortality, multivariable analysis will be performed using backward stepwise logistic regression to enter variables that yielded p < 0.05 in the univariate analyses. The investigators will use an approximation of the odds ratio from the significant values of the stepwise logistic regression to construct a new score.

The investigators will construct receiver operating characteristic (ROC) curves of this score and compared the areas under the ROC curves (AUC) of the ABC, the WFNS, the GCS and the Fisher score to predict 1-year mortality and 1-year full recovery (defined as Rankin score 0-1). Statistical analyses will be performed using STATA version 11 (StataCorp, Texas, USA).