Objectives: Mild therapeutic hypothermia after out-of-hospital cardiac arrest is usually achieved either by surface cooling or by core cooling via the patient's bloodstream. We compared modern core (Coolgard) and surface (Arctic Sun) cooling devices with a zero hypothesis of equal cooling, complications, and neurologic outcomes.
Design: Single-center observational study.
Setting: University hospital medical and cardiac intensive care units.
Patients: One hundred sixty-seven consecutive patients comatose after out-of-hospital cardiac arrest of all causes treated with mild therapeutic hypothermia in a 5-yr period.
Interventions: Nonrandomized allocation to core or surface cooling depending on availability and physician preference.
Measurements and main results: All out-of-hospital cardiac arrest patients' records were reviewed for relevant data regarding medical history, cardiac arrest event, prehospital care, in-hospital treatment, and complications. Survivor neurologic function was reassessed at follow-up after 6 to 12 months. Baseline patient and arrest episode characteristics were similar in the treatment groups. There was no significant difference in survival with good neurologic function, either to hospital discharge (surface, 34/90, 38%; core, 34/75, 45%; p=.345) or at follow-up (surface, 34/88, 39%; core, 34/75, 45%; p=.387). Time from cardiac arrest to achieving mild therapeutic hypothermia was equal with both devices (surface, 273 min, interquartile range 158-330; core, 270 min, interquartile range 190-360; p=.479). There were significantly more episodes of sustained hyperglycemia among the surface-cooled patients (surface, 64/92, 70%; core, 36/75, 48%; p=.005) and significantly more hypomagnesaemia among core-cooled patients (surface, 16/87, 18%; core, 27/74, 37%; p=.01).
Conclusions: In this study, surface and core cooling of out-of-hospital cardiac arrest patients following the same established postresuscitation treatment protocol resulted in similar survival to hospital discharge and comparable neurologic function at follow-up.