Hypothermic to ischemic ratio and mortality in post-cardiac arrest patients

We studied the associations between ischemia and hypothermia duration, that is, the hypothermic to ischemic ratio (H/I ratio), with mortality in patients included in a trial on two durations of targeted temperature management (TTM) at 33°C.

The TTH48 (NCT01689077) trial compared 24 and 48 hours of TTM in patients after cardiac arrest. We calculated the hypothermia time from return of spontaneous circulation (ROSC) until the patient reached 37°C after TTM and the ischemic time from CA to ROSC. We compared continuous variables with the Mann-Whitney U test. Using COX regression, we studied the independent association of the logarithmically transformed H/I ratio and time to death as well as interaction between time to ROSC, hypothermia duration, and intervention group. We visualized the predictive ability of variables with receiver operating characteristic curve analysis.

Of the 338 patients, 237 (70%) survived for 6 months. The H/I ratio was 155 (IQR 111-238) in survivors and 114 (IQR 80-169) in non-survivors (P < .001). In a Cox regression model including factors associated with outcome in univariate analysis, the logarithmically transformed H/I ratio was a significant predictor of outcome (hazard ratio 0.52 (0.37-0.72, P = .001)). After removing an outlier, we found no interaction between time to ROSC and intervention group (P = .55) or hypothermia duration in quartiles (P = .07) with mortality. There was no significant difference in the area under the curve (AUC) between time to ROSC and H/I ratio (ΔAUC 0.03 95% CI -0.006-0.07, P = .10).

We did not find any consistent evidence of a modification of the effect of TTM based on ischemia duration.