Pentobarbital coma for refractory intra-cranial hypertension after severe traumatic brain injury: mortality predictions and one-year outcomes in 55 patients.

OBJECTIVE: To identify predictors of mortality and long-term outcomes in survivors after pentobarbital coma (PBC) in patients failing current treatment standards for severe traumatic brain injuries (TBI). This is a retrospective cohort study of severe TBI patients receiving PBC at Level I Trauma Center and tertiary university hospital.
METHODS: Four thousand nine hundred thirty-four patients were admitted to the trauma intensive care unit with severe TBI (head Abbreviated Injury Scale >or= 3) between April 1998 and December 2004. Six hundred eleven received intracranial pressure (ICP) monitoring and 58 received PBC. Three patients underwent craniotomy for intracranial mass lesion and were excluded. The study group received standardized medical management for severe TBI including opiates, benzodiazepines, elevation of the head of bed, avoidance of hypotension and hypercapnia and hyperosmolar therapy (HOsmRx). In addition, 31 of 55 patients (56%) underwent placement of intraventricular catheters for cerebrospinal fluid drainage. If routine medical management and cerebrospinal fluid diversion failed to control ICP, then the patient was determined to have refractory intracranial hypertension (RICH) and PBC treatment was initiated. PBC was performed with pentobarbital infusion with continuous electroencephalogram monitoring to ensure adequate burst suppression. The measurements include serum sodium (Na) and osmolality (Osm) were assessed as indicators for initiation of PBC and to estimate the 50% mortality cut-points when controlling for ICP. Follow-up functional outcomes were assessed using the Glasgow Outcome Scale and stratified according to admission Glasgow Coma Scale score and Marshall computed tomography classification. Of the 55 PBC patients, 22 (40%) survived at discharge. 19 of 22 had long-term follow-up (1 year or more) available. Of these, 13 (68%) were normal or functionally independent (Glasgow Outcome Scale score 4 or 5). Serum Na and Osm were associated with death (p < 0.05) when controlling for ICP. The 50% mortality cut-points were Na of 160 mEq/L and Osm of 330 mOsm/kg H2O. Median minimum cerebral perfusion pressure after PBC was 42 mm Hg in survivors and 34 mm Hg in nonsurvivors (p = 0.013).
CONCLUSIONS: In patients with severe TBI and RICH, survival at discharge of 40% with good functional outcomes in 68% of survivors at 1 year or more can be achieved with PBC after failure of HOsmRx. Based on 50% mortality cut-points, analysis suggests the limits of HOsmRx to be Na of 160 mEq/L and Osm of 330 mOsm/Kg H2O. Maintenance of higher cerebral perfusion pressure after PBC is associated with survival. PBC treatment of RIH may be even more important when other treatments of RIH, such as decompressive craniectomy, are not available.