Targeted Temperature Management After Cardiac Arrest: How Low to Go?

Previous trials including Bernard SA, Gray TW, Buist MD, et al and a study by the Hypothermia after Cardiac Arrest Study Group have demonstrated a significant improvement in neurological function and survival when therapeutic hypothermia was employed following cardiac arrest. The optimal temperature remains unclear, and the mere avoidance of hyperthermia may explain the benefit of hypothermia when it is induced after arrest. In the November issue of the New England Journal of Medicine, the TTM (Target Temperature Management) Trial Investigators, led by Niklas Nielsen in Sweden, conducted a trial to investigate the benefits and harms of two temperature regimens for cardiac arrest patients.

The randomized, controlled, assessor-blinded multicenter trial was conducted in 36 intensive care units in Australia and Europe. Patients who sustained out-of-hospital cardiac arrest were randomized in a 1:1 ratio to induction of hypothermia with a target of either 33°C or 36°C. Hypothermia was maintained for 36 hours after randomization.  Ice packs, cold fluid, and intravascular (24% of subjects) or surface (76% of subjects)  temperature management devices were used. The primary outcome was survival, and the secondary outcome was a composite of poor neurological function or death, as determined by the Cerebral Performance Category (CPC) scoring systems and modified Rankin scale. A modified intention-to-treat analysis was performed, and Kaplan-Meier curves were compared. Logistic regression and Cox regression were used to adjust for age, sex, presence of a shockable rhythm, presence of cardiogenic shock, and time from cardiac arrest to return of spontaneous circulation.

Both treatment groups had similar prerandomization characteristics. No significant difference was detected in the primary outcome of death; 50% died in the 33°C group versus 48% in the 36°C group (P=0.51). In terms of secondary outcomes, patients in the 33°C group had CPC scores and modified Rankin scores that were not statistically different from those of the 36°C group. Adverse events occurred in 93% of patients in the 33°C group compared with 90% in the 36°C group (P=0.09).

The strengths of this trial include a robust randomized design, lack of contamination, and assessment of outcomes by blinded investigators. Nonshockable rhythms occurred in 20% of patients, making this trial more externally generalizable than the previous two hypothermia trials on which current advanced cardiac life support recommendations are based. The study was powered to find a 55% mortality reduction in the 33°C group; in previous trials, mortality was 49% to 55% (1,2), so in terms of the primary outcome of survival, this study was powered to simply replicate previous findings (at best). The secondary outcome of neurological disability—an outcome of perhaps more interest to intensivists—was assessed by scales that might not be sensitive enough to detect significant changes in health-related quality of life. Although a lower temperature failed to improve outcomes for cardiac arrest patients in this trial, mortality in both groups was similar to that in the previous major trials, supporting the benefit of some degree of hypothermia and avoidance of hyperthermia.

This Concise Critical Appraisal was authored by James Lantry, MD, who is a senior fellow in Critical Care Medicine at Division of Pulmonary and Critical Care.
Sam Galvagno, DO, PhD, is the editor of the Concise Critical Appraisals feature.

eNewsletter Issue:

2 thoughts on “Targeted Temperature Management After Cardiac Arrest: How Low to Go?

  1. Terry Forrette

    I also read this study but was unable to determine the rate of cooling to each target temperature. Additionally the appendices were not specific on which cooling technologies were used. Any help on this would be greatly appreciated.

    Thank you

  2. Sam Galvagno

    Dear readers,

    An error was made with this CCA. We stated that the study was powered to find a 55% mortality reduction. This is not quite correct. The trial was actually powered to find a relative risk reduction of 20% between the groups, assuming baseline mortalities of 55% and 44% in both groups. This would result in an 11% absolute risk reduction. This is quite consistent with the previous two trials that reported absolute risk reductions of 14-23%. Therefore, consistent with the overall rigorous design and execution of this study, in terms of the overall outcome, the study was powered sufficiently.

    For more details about this study, you can hear from Dr. Nielsen himself on “The Intensive Care Podcasts” (

    Happy Holidays!

    Sam G.

Comments are closed.