Abstracts

Rationale: In catastrophic neurologic injury, withdrawal of care (WOC) is often considered. While WOC is based on the perception of poor prognosis, the question of whether it creates a self-fulfilling prophecy has been raised in conditions such as traumatic brain injury and stroke. Similar to these, status epilepticus (SE) has also been associated with poor outcome. However, the effect of WOC on mortality in SE is unclear. The goal of this study was to describe the final cause of death in adults with SE, and to determine the impact and associated clinical characteristics of WOC on mortality rates. Methods: A single-center retrospective study at an urban academic medical center was conducted between 1/1/2005-10/31/2012. Subjects aged?-18 years that were managed for and diagnosed with SE were included. Data was collected on age, gender, Glasgow Coma Scale (GCS), Acute Physiology and Chronic Health Evaluation II (APACHE II), history of epilepsy, etiology of SE, refractory SE (RSE), in-hospital mortality and cause of death. Pearson's ? and t-tests were used as appropriate. Binary logistic regression analysis adjusted for covariates, and p < 0.05 was considered significant. Results: Of 151 subjects, 68 male, mean age 59 years, there were 61(40.4%) in-hospital deaths, 45(74%) due to WOC. The remaining causes were cardiogenic (N=11,18%) and respiratory (N=2,3.2%), with sepsis, brain death and seizures individually comprising 1.6% (N=1 each). Excluding WOC subjects, in-hospital mortality fell to 15%(N=16/106). In the sub-group without cardiac arrest (CA), in-hospital mortality was 25%(N=26/104), of which 14%(N=15) was due to WOC; mortality decreased to 12%(11/89) excluding WOC subjects. For the total cohort, compared to subjects with non-WOC death, there was no significant difference in age, gender, history of epilepsy, GCS, APACHE II, and SE etiologies which were acute, infectious, autoimmune, or due to stroke or mass lesion. WOC was more common in subjects with metabolic seizure etiologies (N=6,?=8.97,p=0.003), CA (N=30,?=6.05,p=0.014), and RSE (N=45,?=5.82,p=0.016). Metabolic etiologies (OR 6.50,p=0.005) and CA (OR 4.40,p=0.018) remained significantly associated with WOC after univariate but not multivariate adjustment. Conclusions: Withdrawal of care is a major determinant of mortality in SE, regardless of anoxic brain injury, and is associated with metabolic dysfunction and cardiac arrest. The impact of WOC on mortality rates in SE should be taken into consideration in clinical decision making. Funding: N/A