Abstract T P282: Withdrawal of Life Sustaining Therapy and In Hospital Death after Intracranial Hemorrhage

Abstract

Introduction: Despite relatively high in hospital mortality, studies on withdrawal of life sustaining therapy (WOLST) and physician ability to predict in hospital death after intracranial hemorrhage remain sparse. Methods: Prospectively collected data from 2008-2011 was reviewed in patients with intraparencymal (IPH), subdural (SDH), and subarachnoid (SAH) hemorrhage. Baseline demographic data, medical comorbidities, and admission clinical status were assessed to determine predictors of WOLST. Statistically significant variables were included in a forward stepwise multivariable logistic regression model to identify independent predictors of WOLST. The model was then tested as a predictor of in hospital death amongst patients who received maximal medical and surgical therapy. Results: 379 patients were enrolled: 127 (33.5%) SDH, 133 (35.1%) SAH, and 119 (31.4%) IPH. 31 (8.2%) patients underwent WOLST and died, 36 (9.5%) continued maximal therapy and died in hospital, and 312 (82.3%) continued maximal therapy and survived to discharge. Patients who underwent WOLST had more medical comorbidities, worse baseline functional status, worse clinical admission status, and were more likely to have IPH. There were no ethnic or level of education predictors of WOLST. Additionally, there were no differences in hospital complications, cost, or length of stay amongst patients who underwent WOLST versus those who died despite maximal therapy. On multivariable analysis, admission GCS, herniation, and history of cardiovascular disease predicted WOLST (Area under the curve [AUC]=0.83, all p<0.01). This model predicted in hospital death amongst patients who received maximal therapy with a sensitivity of 76.5% and specificity of 83.5% (AUC=0.89, all p<0.001). Conclusions: Physicians identified intracranial hemorrhage patients who would die irrespective of maximal medical and surgical therapy with a high sensitivity and specificity. Physicians may be able to transition to WOLST more quickly to attenuate hospital complications, cost, and length of stay.