Abstract P66: Clinical Predictors of Poor Outcomes in Patients Treated With Therapeutic Hypothermia
Objective: To determine the presence of clinical variables that may predict mortality or poor neurologic outcome during and after therapeutic hypothermia (TH) for cardiac arrest.
Background: Therapeutic hypothermia has been shown to reduce the degree of anoxic brain injury and improve mortality and neurologic recovery in clinical situations involving cardiopul-monary and cerebral resuscitation, but there are few studies looking at potential clinical markers of poor clinical outcome in this setting. Easily measured bedside markers for patient outcomes would be a useful adjunct in determining level of care of survivors of cardiac arrest
Methods: Retrospective medical records review of 41 consecutive survivors of cardiac arrest treated with TH from 2004 to the present.
Results: Patients in the study cohort were mostly Caucasian (96%) and elderly (mean age, 66 years). Most subjects had an out-of-hospital cardiac arrest that was witnessed, and most received bystander cardiopulmonary resuscitation (CPR). Fifty-seven percent (22 of 41 subjects) expired or experienced severe neurologic impairment. 46% of our patients had a non VT/VF arrest. Factors associated with a poor prognosis included:
a first rhythm at cardiac arrest other than ventricular tachycardia/fibrillation (p<0.01);
the presence of acute kidney injury during the first 72 hours in the ICU (p<0.001);
a Glasgow Coma Scale <8 determined 12 hours after warming (p<0.001) and cardiac arrhythmia requiring intervention in the first 24 hours (p .034).
The presence of any 2 of these factors markedly increased the risk of a poor outcome. Age, sex, arrest duration, and comorbidities did not predict outcome. Conclusion: Several simple, easily measured, and reproducible clinical markers may help predict mortality or poor neurologic outcome, during and after treatment, in patients managed with TH for cardiac arrest.