Yep, The HEAT Trial is hot, hot, hot.
From February 2013 to July 2014, 700 med/surg ICU patients 16 years of age or older in 23 Australian & New Zealand ICUs receiving antibiotics for known or suspected infection were enrolled 1:1 to receive 1 gram of IV acetaminophen or placebo every 6 hours. The study groups had similar characteristics at baseline. Exclusion criteria only being acute brain disorder or liver dysfunction. Rescue cooling measures were allowed for temperatures >39.5 C (103.1F).
There are a few interesting take-aways from this paper-
1) The number of ICU free days to day 28 did not differ significantly between the two groups (23 vs 22 days, acetaminophen vs placebo respectively, p = 0.07). Nor did ICU length of stay or mortality at both 28 day & 90 days. However, acetaminophen was associated with shorter ICU stay for survivors, and longer ICU stays for non-survivors – consistent with physical cooling measures for mechanically vented septic shock patients as well as with prior ICU studies.
2) Of the patients discharged from the ICU, there was no difference between the two groups in terms of CRP, CK, or creatinine levels (though the placebo group was more likely to have a fever on the day of discharge from the ICU- 23% vs 12%).
3) There were no significant differences between the two groups for the use of nsaids or rescue cooling. Liver dysfunction led to discontinuation of the acetaminophen 8.1% of the time, and discontinuation of placebo 9.9% of the time (not statistically significant). There was one death associated with marked hyperthermia – a 52 year old patient with community acquired pneumonia with a core temperature of 42.7C (108.9F). This patient was given 9 doses of study medication, of which 5 were acetaminophen. It is unclear exactly how this mix up occurred, but regardless, you can not blame the death on lack of acetaminophen.
With an NNT somewhere between 350 and infinity, urgently lowering a temperature with acetaminophen does not matter in the ICU. Considering that there were no significant differences in the number of rescue cooling attempts or nsaid use between the two groups, we are simply attempting to buff the chart with a meaningless intervention, with an increased cost (no matter how small), without appreciable benefit except to make ourselves feel better. Kudos to the authors for challenging dogma. Now if we can repeat this study design on the wards…