Early- versus late-initiation of therapeutic hypothermia after cardiac arrest: preliminary observations from the experience of 17 Italian intensive care units.

OBJECTIVES: Mild therapeutic hypothermia (TH) has been shown to improve neurologic outcome in patients experiencing cardiac arrest after return of spontaneous circulation (ROSC). The best timing to initiate TH is currently not known. The aim of this study by the ICE (Italian Cooling Experience) group was to investigate the relationship between the timing of initiation of therapeutic hypothermia (TH) and both patient survival and neurologic outcome.
METHODS: In this observational prospective clinical study we collected data on cardiac arrest patients admitted, after ROSC, to any of the 17 participating Italian intensive care units. Patients were managed according to routine clinical practice, including, in a group of patients, therapeutic hypothermia. Patients who underwent TH were classified, arbitrarily, into an early-initiation group (TH started <2 h since cardiac arrest) and a late-initiation group (TH started >2 h since cardiac arrest).
RESULTS: Intensive care unit (ICU) mortality was 47.4% for the early-initiation group and 23.8% for the late-initiation group (P=0.01). Six-month mortality was 60.8% for the early-initiation group and 40.5% for the late-initiation group (P=0.04). Cerebral performance category (CPC, a measure of neuro-cognitive outcome) at ICU discharge was 1 [1-2] for the early-initiation group and 1 [1-3] for the late-initiation group (P=0.57). At 6 months, CPC was 1 [1-1] for the early-initiation group and 1 [1-2] for the late-initiation group. DISCUSSION: Despite similar neurologic outcomes at every time point, mortality was significantly higher when therapeutic hypothermia was started within 2h of cardiac arrest than when it was started later. Due to the lack of possibility to control several putative confounding factors, such results should be considered as preliminary observations warranting further research.