Background: Although the level of evidence of improvement is significant in cardiac arrest patients resuscitated from a shockable rhythm (ventricular fibrillation or pulseless ventricular tachycardia [VF/Vt]), the use of therapeutic mild hypothermia (TMH) is more controversial in nonshockable patients (pulseless electric activity or asystole [PEA/asystole]). We therefore assessed the prognostic value of hypothermia for neurological outcome at hospital discharge according to first-recorded cardiac rhythm in a large cohort.
Methods and results: Between January 2000 and December 2009, data from 1145 consecutive out-of-hospital cardiac arrest patients in whom a successful resuscitation had been achieved were prospectively collected. The association of TMH with a good neurological outcome at hospital discharge (cerebral performance categories level 1 or 2) was quantified by logistic regression analysis. TMH was induced in 457/708 patients (65%) in VF/Vt and in 261/437 patients (60%) in PEA/asystole. Overall, 342/1145 patients (30%) reached a favorable outcome (cerebral performance categories level 1 or 2) at hospital discharge, respectively 274/708 (39%) in VF/Vt and 68/437 (16%) in PEA/asystole (P<0.001). After adjustment, in VF/Vt patients, TMH was associated with increased odds of good neurological outcome (adjusted odds ratio, 1.90; 95% confidence interval, 1.18 to 3.06) whereas in PEA/asystole patients, TMH was not significantly associated with good neurological outcome (adjusted odds ratio, 0.71; 95% confidence interval, 0.37 to 1.36).
Conclusions: In this large cohort of cardiac arrest patients, hypothermia was independently associated with an improved outcome at hospital discharge in patients presenting with VF/Vt. By contrast, TMH was not associated with good outcome in nonshockable patients. Further investigations are needed to clarify this lack of efficiency in PEA/asystole.