A comparison of therapeutic hypothermia and strict therapeutic normothermia after cardiac arrest.

Abstract

In a recent high-quality randomised controlled trial (RCT), strict therapeutic normothermia (STN) following cardiac arrest with coma resulted in similar outcomes to therapeutic hypothermia (TH). We aimed to test the feasibility, reproducibility, and safety of the STN protocol outside of its RCT context. In two teaching hospital ICUs, we performed a before-and-after study comparing the previously International Liaison Committee on Resuscitation (ILCOR)-endorsed TH protocol to the recently studied STN protocol. The primary feasibility end point was the percentage of temperature recordings in the prescribed range in the first 24h of treatment. Secondary end points included pharmacological management and complications. We studied 69 similar patients in each group. We found no difference in feasibility as shown by the proportion of within range temperatures. However, the median doses of midazolam (37mg vs. 9mg, p=0.02), fentanyl (883μg vs. 310μg, p=0.01) and the use of muscle relaxants (84.1% vs. 59.4%, p=0.001) was greater with the TH protocol. Furthermore, shivering (52.2% vs. 18.8%, p<0.001), a composite of other pre-defined complications (66.7% vs. 47.8%, p<0.03) and fever in the first 96h (55.1% vs. 33.3%, p=0.01) were also more common with the TH protocol. The STN protocol was successfully reproduced outside of an RCT and appeared associated with fewer complications than the TH protocol. Our findings imply that the STN protocol may offer clinical advantages over the TH protocol.