Therapeutic Hypothermia Following Cardiac Arrest

Hypothermia following cardiac arrest does not improve the chance of survival or favorable neurologic outcome

Benefits in NNT

No one was helped (chance of survival or favorable neurological function did not improve)
No one was helped

Harms in NNT

Not assessed
Not assessed
View As:


Long B, Gottlieb M. Therapeutic hypothermia following cardiac arrest. Academic Emergency Medicine. Published online July 31, 2023:acem.14785.

Study Population: 2,900 participants in 6 trials, most enrolling comatose survivors of out-of-hospital cardiac arrest

Efficacy Endpoints

Survival and favorable neurologic outcome

Harm Endpoints

Not assessed


Cardiac arrest and its aftermath represent common conditions managed in the emergency department. Timely cardiopulmonary resuscitation, rapid defibrillation of shockable rhythms, and post-arrest care are integral to improving outcomes.1 However survival remains poor, ranging between 6% and 20%.1, 2, 3 Several randomized trials in the past two decades have evaluated therapeutic hypothermia, also known as targeted temperature management (TTM), during the post-arrest phase to improve survival and neurologic outcomes.4, 5, 6, 7, 8 TTM has been refined over time with regard to patient selection, time of initiation, temperature goals, means and duration of cooling, and research design.9, 10, 11, 12, 13 Perhaps unsurprisingly therefore, trials during this period have yielded varying results. Herein, we summarize a systematic review of randomized trial data addressing TTM for survivors of cardiac arrest.11

The systematic review discussed here included 6 randomized trials (n = 2,900 participants) of comatose post-cardiac arrest patients. Trials randomized patients to receive either hypothermic TTM (32-34℃) versus normal-temperature TTM (36.5–38℃) in the hours after their cardiac arrest, or to receive hypothermic TTM versus no TTM.11 Included studies provided information concerning the timing of TTM initiation, target temperature, duration, method, and rewarming rate for cooled patients. The review did not include studies that used TTM for indications other than post-cardiac arrest, or with patients <18 years of age. The systematic review reported outcomes recommended by the International Liaison Committee on Resuscitation Advanced Life Support Task Force. These outcomes included short-term survival, mid-term favorable neurologic outcome, long-term survival, and long-term favorable neurologic outcome. The systematic review defined a favorable neurologic outcome as a modified Rankin Scale (mRS) score of 0-3 or a Cerebral Performance Category score of 1 to 2.11 Other outcomes included health-related quality of life, cognitive function, and anxiety and depression. Authors also reported subgroup analyses on shockable (ventricular fibrillation and pulseless ventricular tachycardia) or non-shockable (asystole and pulseless electrical activity) initial cardiac arrest rhythms. Finally, they conducted sensitivity analyses with exclusion of studies at high risk of bias and trials more than 10 years old. They also conducted sensitivity analysis comparing 36℃ versus 32-34℃.

Of note, the two earliest trials (Bernard 2002 and HACA 2002) enrolled almost exclusively patients with shockable rhythms4, 5 while a multicenter trial (HYPERION 2019)6 enrolled only patients with asystole or pulseless electrical activity. Other trials enrolled mixed populations with mostly shockable rhythms (72-100%).7, 9, 10 Only two trials included patients with in-hospital cardiac arrest;4, 6 the remaining trials studied patients after cardiac arrest outside of the hospital.5, 7, 9, 10

The findings of the meta-analysis indicates that TTM with a hypothermic target of 32-34℃ does not improve survival (5 trials; 2,776 participants) or favorable neurologic outcome at 90 to 180 days (5 trials; 2,753 participants). The results remained the same after authors performed sensitivity analyses. Evidence certainty was low for all outcomes.


There are several important considerations when interpreting these results. First, the certainty of evidence was low. This was driven by the risk of bias from lack of blinding, inconsistency of results and heterogeneity of the trials, and imprecision of effects leading to wide confidence intervals. Second, not all trials reported all outcomes sought in the review. Third, studies were performed decades apart as TTM changed, which may particularly have impacted the control groups. The HYPERION trial used a 36.5-37.5℃ target in the control group while Dankiewicz targeted 37.5℃, intervening only if the patient’s temp was greater than 37.8℃.6, 7 Other trials used no TTM in the control group or 37℃. Fourth, trials reported variable follow up durations. A fifth limitation was variations in time to achieving target temperature and durations of therapy. Sixth, authors were unable to conduct analyses separately for the participants whose arrests occurred in-hospital. Finally, this meta-analysis did not report adverse events which include arrhythmia, infection, and electrolyte abnormalities.12, 13

It is important to note the first two high profile trials of hypothermic TTM from 2002 found a benefit, but these trials were unblinded, single-center, proof-of-concept studies that enrolled less than 10% of screened patients, did not standardize approaches to withdrawal of care or control arm treatment, and were stopped early for logistical reasons.4, 5 Together they enrolled 213 subjects. By contrast, one TTM trial published in 2021 enrolled 1,821 participants and found no benefit with a target of 33℃ compared to normal temperature.7 The study was more rigorous, better controlled, and multi-center across 61 sites.

Also notable is a 2023 Cochrane systematic review of 12 RCTs and quasi-RCTs.12 The review found hypothermic TTM improved neurologic outcome (RR: 1.6, 95% CI 1.2 to 2.2) but also caused higher rates of adverse events.12 These results, however, incorporate several smaller, less controlled studies with significant bias and heterogeneity, and it is these studies that appear to drive the improvement. A second meta-analysis published in 2023 included 5 trials (n=3,614 patients) and found no improvement with TTM in survival or rate of favorable neurologic outcome, with a higher risk of arrhythmia.13

History suggests large replication studies often produce more reliable results, particularly when small early studies report benefits of a novel therapy.14 Based on current data it appears that induction of therapeutic hypothermia does not improve neurologic outcome or survival while multiple reviews suggest an increased risk of adverse events.12, 13 Therefore, we have selected a color recommendation of Red (benefits and harms equal or equivocal) for the use of therapeutic hypothermia in survivors of cardiac arrest. This recommendation is different from the previous NNT summary which had a green recommendation.15 The older summary was based on a 2009 Cochrane systematic review.16 Since then, several large RCTs have been published. The new data have more clearly challenged any claims of benefit due to hypothermia after cardiac arrest.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between and AEM.


Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD


August 17, 2023