Time to Cooling Is Associated with Resuscitation Outcomes

Abstract

Our purpose was to analyze evidence related to timing of cooling from studies of targeted temperature management (TTM) after return of spontaneous circulation (ROSC) after cardiac arrest and to recommend directions for future therapy optimization. We conducted a preliminary review of studies of both animals and patients treated with post-ROSC TTM and hypothesized that a more rapid cooling strategy in the absence of volume-adding cold infusions would provide improved outcomes in comparison with slower cooling. We defined rapid cooling as the achievement of 34°C within 3.5 hours of ROSC without the use of volume-adding cold infusions, with a ≥3.0°C/hour rate of cooling. Using the PubMed database and a previously published systematic review, we identified clinical studies published from 2002 through 2014 related to TTM. Analysis included studies with time from collapse to ROSC of 20-30 minutes, reporting of time from ROSC to target temperature and rate of patients in ventricular tachycardia or ventricular fibrillation, and hypothermia maintained for 20-24 hours. The use of cardiopulmonary bypass as a cooling method was an exclusion criterion for this analysis. We compared all rapid cooling studies with all slower cooling studies of ≥100 patients. Eleven studies were initially identified for analysis, comprising 4091 patients. Two additional studies totaling 609 patients were added based on availability of unpublished data, bringing the total to 13 studies of 4700 patients. Outcomes for patients, dichotomized into faster and slower cooling approaches, were determined using weighted linear regression using IBM SPSS Statistics software. Rapid cooling without volume-adding cold infusions yielded a higher rate of good neurological recovery than slower cooling methods. Attainment of a temperature below 34°C within 3.5 hours of ROSC and using a cooling rate of more than 3°C/hour appear to be beneficial.

Keywords: cardiac arrest; cooling strategies; human studies; postresuscitation cooling; temperature mechanisms.

FIG. 1.

Study selection flow diagram.

FIG. 2.

Rates of favorable recovery (CPC 1 or 2) (Safar, 1981) versus Cardiac Arrest Rhythm (Clinical Studies of PostResuscitation Cooling) as predicted by weighted linear regression analysis of clinical studies listed in Table 1 (weighted by number of patients in studies). Standard errors, indicated by the vertical lines, do not overlap for VT/VF rhythms, suggesting an advantage of the faster cooling approach for patients resuscitated from shockable rhythms. CPC, Cerebral Performance Categories; VF, ventricular fibrillation; VT, ventricular tachycardia.

FIG. 3.

Exploratory linear regression analyses of clinical studies showing chances of good outcomes (CPC 1 or 2) (Safar, 1981) of post-VT/VF patients treated with targeted temperature management versus (A) cooling speed (°C/hour) and (B) time delay from ROSC to target (hours). Regression line in plot (A) includes both rapid and slower cooling studies. Regression line in plot (B) includes only slower cooling studies; faster cooling studies in this plot are shown as boxes. ROSC, return of spontaneous circulation.