A Danish study in Resuscitation sheds more light on the link between fever and mortality in post-ROSC patients. It is interesting as it fits nicely with the recently published, and highly controversial, study in NEJM that suggested hypothermia isn’t working but temperature control might.
Emergency medicine social media is still discussing the NEJM study that suggests hypothermia isn’t working. One possible explanation is that it is rather temperature control and avoiding hyperexia that is benefiting the patients. Perhaps that is where therapeutic hypothermia really works.
Fever (defined as a core temp of >38°C ) has an incidence in the post-cardiac arrest period of 20-80%, depending on what studies you rely on. Earlier studies have observed how this is a negative prognostic marker. Unfortunately those studies were small and from before therapeutic hypothermia was implemented.
This study looks at the association between fever and outcome in the period after therapeutic hypothermia.
The danish study is a prospective, observational, cohort study from the famous Copenhagen University Hospital (Rigshospitalet). It enrolled adult comatose out-of-hospital cardiac arrest patients with sustained ROSC since 20 minutes. It excluded patient’s with cardiogenic shock and where therapeutic hypothermia < 34 °C wasn’t achieved.
Their post- cardiac arrest protocol seems fairly standard. Therapeutic hypothermia was 24 hours with a goal temperature of 33°C. All patient’s had Cefuroxime 1,5 g three times daily. Treatment of post-hypothermia fever was not part of their protocol.
Patients were then sub-grouped into patients who developed fever after the 24h period of therapeutic hypothermia and those who didn’t. Primary endpoints were 30 day mortality and neurologic outcome at discharge.
From a 6 year period, after exclusions, the group enrolled 270 patients that had undergone the full 24 hour protocol. Of these 136 developed fever after therapeutic hypothermia. 134 patients remained normothermic.
Overall mortality in the whole group was 29%.
Mortality at 30 days in the group who had post-hypothermia fever was 35%.
Mortality at 30 days in the non-fever group was 22%
75% of patients in the non-fever group were discharged with a ‘good’ neurological outcome. Compared to 61% of patients in the fever-group.
More evidence, and better evidence, associating fever with poor outcome. This time the association is seen even after a 24h period of therapeutic cooling.
The study doesn’t, however, say anything about the the causal relationship between fever and outcome.
For example, fever after a cardiac arrest could be caused by ischaemic lesions in the thalamus’ thermoregulatory system. In those cases, fever could merely be a marker of serious ischaemic brain injury and treating fever might not help.