In the hours after return of spontaneous circulation, the temperature of a comatose post-arrest patient tends to move in one direction: upwards. It happens without an infection, often within the first day, and it happens in a brain that has just survived a global ischaemic insult and is least able to tolerate additional metabolic load. Fever avoidance in this patient is not the absence of an intervention — it is an intervention in its own right, and it lasts longer than the cooling phase most protocols are built around.
This article looks at what post-arrest fever is, what the evidence says about it, and why treating it is a temperature-control task rather than a pharmacological one.
What the term describes
“Ischaemic fever” is a practical shorthand for a temperature rise that follows an ischaemic insult rather than an infection. In the literature the same phenomenon appears under several names — post-arrest hyperthermia, post-hypothermia fever, rebound hyperthermia — depending on whether it occurs spontaneously after resuscitation or after the rewarming phase of an active protocol.
The mechanism is inflammatory rather than infectious. Whole-body ischaemia followed by reperfusion produces a systemic response resembling sepsis, with circulating inflammatory mediators, vasodilation and pyrexia, in the absence of a pathogen. A raised temperature in the first days after cardiac arrest should therefore be read as part of the post-arrest syndrome until proven otherwise, not automatically as a sign of infection. Infection does occur in these patients and still has to be excluded — but waiting for a microbiological answer before controlling the temperature means leaving the brain exposed in the meantime.
Why fever is more dangerous here than in almost any other patient
The secondary injury that follows reperfusion — inflammation, excitotoxicity, oxidative stress, cerebral oedema — is temperature-sensitive. Higher temperature increases cerebral metabolic demand at exactly the point when perfusion and oxygen delivery are least reliable, and accelerates the biochemical processes driving the damage. The same physiological argument that once justified deep hypothermia works, in reverse, against allowing fever.
Two features of this population make the problem harder to manage than fever elsewhere in the hospital. First, global ischaemia can impair central thermoregulation itself, so the patient’s own set-point control is part of what has failed. Second, sedation, analgesia and neuromuscular blockade blunt the usual outward signs, so a rising temperature may be visible only on the monitor — which is an argument for continuous measurement rather than intermittent spot checks.
What the evidence and the guidelines actually require
The association between post-arrest fever and worse neurological outcome was described early: Zeiner and colleagues reported it in 2001, and it has been reproduced since in the specific setting of fever occurring after the rewarming phase, where Bro-Jeppesen and colleagues found post-hypothermia fever associated with increased mortality after out-of-hospital cardiac arrest.
The large randomised trials are frequently misread on this point. The TTM trial compared 33 °C with 36 °C and found no significant mortality difference; TTM2 compared hypothermia at 33 °C with normothermia and found no significant difference in six-month mortality. In both cases the comparator arm was actively temperature-managed with early treatment of fever — not standard care without temperature control. The conclusion supported by these trials is that deep hypothermia is not mandatory; the conclusion they do not support is that the temperature can be left alone.
The ERC–ESICM guidelines on temperature control after cardiac arrest state this directly: in patients who remain comatose after resuscitation, fever above 37.7 °C should be avoided for at least 72 hours. That figure is worth reading carefully. It sets a threshold well below what would be called a fever on a general ward, and it sets a duration three times longer than the maintenance phase of a classic hypothermia protocol.
Why antipyretics alone are not a strategy
Paracetamol and non-steroidal anti-inflammatory drugs work by lowering a prostaglandin-mediated set-point. That is the right mechanism for an infectious fever in an otherwise intact patient. It is a poor match for a temperature rise driven by systemic inflammation in a patient whose central thermoregulation may itself be damaged, and the response in this population is inconsistent.
There is also a control problem, independent of pharmacology. An antipyretic is an open-loop intervention against a moving target: it is given at intervals, its effect cannot be titrated in real time, and nothing in the loop reacts when the temperature starts climbing again between doses. Passive measures share the same weakness in the opposite direction — an unregulated cooling blanket or ice packs can push a patient below normothermia, and unintended hypothermia brings its own complications.
Maintaining a defined upper limit for three days is a closed-loop task: measure continuously, compare against a set target, and adjust output automatically in both directions.
What 72 hours of active fever avoidance involves in practice
Four things have to be in place, and none of them is pharmacological.
Continuous core temperature measurement, with a defined reference site and, in many units, a second site as a cross-check if a probe becomes displaced. A device operating in normothermia mode with automatic feedback, so that the target is held rather than chased — the same target temperature management equipment used for induction and controlled rewarming, simply set to a different objective. An anticipated shivering strategy, because counter-regulation appears at the upper end of normothermia as well, not only during cooling. And a record of the temperature curve, which matters both for clinical review and for demonstrating protocol compliance afterwards.
Whether that control is delivered by a surface system such as a non-invasive patient cooling device or by an intravascular one is a unit-level decision based on protocol, vascular access and staff training. The requirement that does not vary is feedback control over the full 72 hours.
The phase nobody plans for
Fever avoidance has a consequence that shows up in procurement rather than in the guidelines. A protocol built around 24 hours of maintenance occupies a device for roughly a day per patient. A protocol that also holds normothermia and blocks fever for 72 hours after ROSC occupies it for three.
Units that read the post-TTM2 evidence as “we cool less now” and sized their equipment accordingly are the ones most likely to run short of it. The active phase became less aggressive; the total period during which the patient needs a controlled temperature became longer. That is the operational question worth asking before the next post-arrest admission, not after it.
What this means in practice
A unit reviewing its post-arrest pathway should be able to answer four questions. What is our upper temperature limit, and is it set at the guideline threshold rather than at a general-ward definition of fever? How is temperature measured, and is the measurement continuous? How long do we maintain control after ROSC, and does the protocol say 72 hours? And what is the first response when the temperature starts to rise — a dose of an antipyretic, or a device already in the loop?
Fever after cardiac arrest is predictable, measurable and preventable. Of everything an intensive care team does for a post-arrest patient, this is among the least dramatic and the most consistently rewarded.
Sources and further reading
- Zeiner A, et al. Hyperthermia after cardiac arrest is associated with an unfavorable neurologic outcome. Archives of Internal Medicine (now JAMA Internal Medicine), 2001.
- Bro-Jeppesen J, et al. Post-hypothermia fever is associated with increased mortality after out-of-hospital cardiac arrest. Resuscitation, 2013.
- Nielsen N, et al. Targeted temperature management at 33 °C versus 36 °C after cardiac arrest. New England Journal of Medicine, 2013.
- Dankiewicz J, et al. Hypothermia versus normothermia after out-of-hospital cardiac arrest (TTM2). New England Journal of Medicine, 2021.
- Nolan JP, et al. ERC–ESICM guidelines on temperature control after cardiac arrest in adults. Resuscitation, 2022.
- American Heart Association — guidelines for cardiopulmonary resuscitation and emergency cardiovascular care, adult post-cardiac arrest care.
- International Liaison Committee on Resuscitation (ILCOR) — consensus on science with treatment recommendations.