6.2 Temperature Control (01.K)

Core temperature is a controllable secondary injury. This section covers temperature control after return of spontaneous circulation (ROSC), fever after stroke and brain hemorrhage, why routine hypothermia is not the default after ICH, SAH, or TBI, cooling devices, and shivering. Independent OpenExamPrep teaching here covers Temperature control as listed in the ABPN Content Specifications. It is not an AHA course manual and not an ABIM product.

Why the wording changed from "cool to 33" to temperature control

For a decade, many ICUs treated targeted temperature management (TTM) as a synonym for 33 °C. Two large trials broke that habit. TTM (Nielsen and colleagues, NEJM 2013) compared 33 °C with 36 °C after out-of-hospital arrest and found no difference in survival or neurologic function. TTM2 (Dankiewicz and colleagues, NEJM 2021) compared hypothermia at 33 °C with normothermia targeting 37.5 °C (active fever prevention) in about 1,900 comatose adults after out-of-hospital cardiac arrest. Six-month mortality was 50% versus 48%. Functional outcomes did not differ. The hypothermia arm used more sedation and had more arrhythmias. Those two trials are why a 2026 examination item that still says "every ROSC patient must be cooled to 33 °C" is testing an outdated slogan, not current resuscitation science.

HYPERION (Lascarrou and colleagues, NEJM 2019) found more favorable 90-day outcomes with 33 °C than with 37 °C in a smaller trial limited to nonshockable rhythms. It does not override TTM2's larger mixed-rhythm result, and it does not recreate a mandate to cool everyone to 33 °C. Mention it so you can explain why some units still individualize a lower target; do not use it to discard fever prevention as the evidence-based floor.

AHA 2023 ACLS focused update: the language you must know

The 2023 American Heart Association Focused Update on Adult Advanced Cardiovascular Life Support (Perman and colleagues, Circulation 2023) replaced loose "TTM to 33" teaching with temperature control. The update's own training summary states that temperature control includes hypothermic temperature control, normothermic temperature control, and temperature control with fever prevention. You choose one temperature between 32 °C and 37.5 °C and hold that constant temperature.

Memorize the operational rules, not a brand-name device:

• Who: All adults who do not follow commands after ROSC, irrespective of arrest location (in-hospital or out-of-hospital) and irrespective of presenting rhythm. This is a Class 1 recommendation for a deliberate temperature-control strategy.
• What number: Select and maintain a constant temperature between 32 °C and 37.5 °C (Class 1). The update states there is insufficient evidence to name one mandatory therapeutic temperature inside that band.
• How long: It is reasonable to maintain temperature control for at least 24 hours after the target is reached (Class 2a).
• Protocols: Hospitals should develop protocols for post-arrest temperature control (Class 1). A verbal "we usually cool them" is not a protocol.
• Fever: It is reasonable to prevent fever in patients still unresponsive to verbal commands after the initial temperature-control phase (Class 2b). Fever after ROSC associates with worse outcome; the update still treats prevention or treatment of fever as reasonable even though a dedicated "fever prevention improves survival" trial is not what TTM2 primarily tested.
• Spontaneous hypothermia: Patients who arrive already cold and do not follow commands should not routinely be rewarmed faster than 0.5 °C per hour (Class 2b).
• Prehospital: Routine rapid cold intravenous fluid for prehospital cooling after ROSC is not recommended.

Continuous core temperature (bladder, esophageal, or vascular) is how you know you are actually holding a constant number. Axillary stickers lie. If you pick 36.5 °C, the nursing order is not "keep them under 38 if you can." It is a controlled band around the chosen setpoint.

What this is not

It is not mandatory cooling to 33 °C. It is not permission to ignore a temperature of 39 °C because "TTM2 said normothermia is fine." Normothermia in TTM2 meant active prevention of fever, including a cooling device in a large fraction of the 37.5 °C arm. It is not a reason to freeze a following, conversant patient. If the patient follows commands, the post-arrest temperature-control indication in the focused update does not apply in the same way; still treat fever as you would in any ICU.

Stroke fever, QASC, INTREPID, and why SHINE sits next to the thermometer

Fever after ischemic stroke, ICH, and SAH is common (chemical meningitis after SAH, central fever, aspiration, line infection, venous thromboembolism, drug fever). Observational series repeatedly associate fever with hematoma expansion, delayed cerebral ischemia, longer ICU stay, and worse functional outcome. Association is not the same as "induced hypothermia saves the hemisphere," which is why the next paragraphs separate treating fever from inducing 33 °C.

The Quality in Acute Stroke Care (QASC) cluster trial treated fever, sugar, and swallow as one nursing bundle and reduced death and dependency. That is the bedside triad: treat temperature ≥37.5–38 °C with a protocol (usually starting with acetaminophen and a workup for infection), keep glucose in a moderate band, and do not feed an unscreened mouth. SHINE belongs in that sentence even though it is a glucose trial. SHINE showed that driving glucose to 80–130 mg/dL with intravenous insulin did not improve 90-day function and caused severe hypoglycemia. The lesson for temperature items is the same shape: secondary-injury numbers that look prettier are not automatically outcome-positive. Fever still gets treated because it raises cerebral metabolic demand and is consistently associated with worse brain-injury physiology. That is not a license to induce hypothermia after every infarct, and it is not a license to run an insulin drip to 100 mg/dL while you cool the patient to 33 °C "to be thorough."

INTREPID (Greer and colleagues, JAMA 2024) randomized critically ill adults with acute ischemic stroke, ICH, or SAH to automated surface fever prevention targeting 37.0 °C for up to 14 days versus standardized treatment once temperature reached 38 °C. Fever burden fell (primary outcome). The principal secondary analysis of 3-month modified Rankin Scale did not show a functional benefit. So: preventing and treating fever is feasible and reduces hours spent hot; prophylactic automated normothermia has not yet been shown to improve recovery compared with treating fever when it occurs. On the examination, "fever is harmless after ICH" is false; "every vascular brain injury must be kept at 33 °C" is also false; "INTREPID proved outcome benefit for device-based fever prevention" overclaims the trial.

Ischemic-stroke hypothermia trials (including EuroHYP-1 and earlier ICTuS work) have not made routine cooling standard. Shivering, pneumonia, and failure to reach target plagued those studies. Treat fever; do not promise families that a cooling catheter will reverse the infarct.

Induced hypothermia versus fever control after ICH, SAH, and TBI

Fever is harmful in these diseases as a physiologic stress: higher cerebral metabolic rate, higher ICP in some patients, more vasospasm associations after SAH, and worse observational outcomes. Routine induced hypothermia is not standard for ICH, SAH, or TBI as a disease-modifying therapy.

After TBI, two trials should be in your pocket. Eurotherm3235 (Andrews and colleagues, NEJM 2015) used hypothermia (32–35 °C) as a stage-2 treatment for intracranial hypertension. The trial stopped early. The hypothermia group had worse Extended Glasgow Outcome Scale results at 6 months (adjusted common odds ratio 1.53 toward worse outcome) and higher mortality. Favorable recovery was about 26% with hypothermia versus 37% with standard care in the reported split. POLAR (Cooper and colleagues, JAMA 2018) tested prophylactic hypothermia after severe TBI and did not improve favorable outcomes. Brain Trauma Foundation 4th Edition material does not recommend prophylactic hypothermia to improve survival or neurologic outcome. Hypothermia can still appear as a rescue ICP tool in a stepwise algorithm after sedation, osmotherapy, and CSF drainage — that is a last-rung physiologic maneuver, not a day-1 protocol for every GCS 7.

After ICH and SAH, induced hypothermia has been studied in small or special-population protocols (sometimes for ICP or for refractory fever). It is not guideline-standard disease-modifying care. You treat fever, search for infection and VTE, and you do not cool a stable Hunt-Hess 2 patient to 33 °C because a review article mentioned neuroprotection in rats.

Rewarming, when you have cooled, is slow (often 0.25–0.5 °C per hour) to avoid rebound ICP, vasodilation, and shivering. Electrolytes shift during cooling (potassium falls) and rebound during rewarming.

Surface versus endovascular devices

You need a deliberate method, not a bag of ice on the groin and hope.

Surface systems include hydrogel wraps (for example Arctic Sun-style gel pads), circulating-water blankets, and, for induction only, ice packs and fans. Advantages: no central venous cooling catheter, faster to start, usable when vascular access is ugly. Disadvantages: skin injury, less tight control in obese patients, shivering from cold skin, and slower induction than a well-running intravascular catheter. Watch pressure points and pad-adhesive burns.

Endovascular heat-exchange catheters (for example Thermogard-style devices) sit in the inferior vena cava through a femoral sheath and circulate closed-loop fluid. Advantages: rapid induction, tight control, less shivering from skin cold in some patients. Disadvantages: central venous complications — insertion injury, bloodstream infection, catheter-related thrombosis, and the need to keep a large femoral line in a DVT-prone neuro patient. They are not first-line for a patient who only needs acetaminophen for a 38.2 °C infectious workup.

Cold intravenous crystalloid can drop temperature quickly in the emergency department. The AHA 2023 update rejects routine prehospital cold-fluid loading. In-hospital cold-fluid boluses still appear in some induction protocols; they are a volume load and a potassium shift, not a standalone 24-hour strategy.

Whatever device you pick, the AHA rule is the same: one constant temperature, at least 24 hours in the post-ROSC comatose adult, then a planned rewarm or a planned fever-prevention phase. Write the target on the order, not "TTM on."

Shivering protocol

Shivering doubles or triples oxygen consumption, raises ICP, and fights your cooling device. Score it (the Bedside Shivering Assessment Scale, BSAS, runs 0–3). Treat in tiers. A widely used neuro-ICU sequence, popularized in temperature-management programs (Badjatia and colleagues and subsequent unit protocols), is:

Paralysis without sedation is awareness. Paralysis as step 1 is how you miss seizures and create ICU-acquired weakness.

Do not stack a paralytic because the Arctic Sun alarm is ringing before you have given acetaminophen and turned on counterwarmth. Do not use shivering as an excuse to skip a fever workup: blood cultures, chest imaging, urine, lines, and sinus or wound checks still matter after SAH and TBI.

Worked bedside scenarios

A 62-year-old remains unresponsive to commands 2 hours after ROSC from a shockable out-of-hospital arrest. Temperature is 36.4 °C. The AHA 2023-consistent plan is a protocolized constant target between 32 and 37.5 °C for at least 24 hours — for example 36.5 °C with a device and fever prevention — not an automatic 33 °C induction and not "watch and treat if they hit 39."

A 40-year-old with severe TBI has ICP 28 mmHg after sedation, osmotherapy, and an open EVD. Someone proposes 33 °C as the next ICP step. You may discuss hypothermia as a rescue in a local algorithm, but you must know Eurotherm3235 found worse outcomes when hypothermia was used as staged ICP therapy. It is not a benign default.

A Hunt-Hess 3 SAH patient hits 38.8 °C on day 6. Treat the fever, culture and image, and do not start an endovascular cooling catheter to 33 °C as "vasospasm prophylaxis."

Exam traps

Mandatory cooling of every ROSC patient to 33 °C. Ignoring fever because TTM2 used a 37.5 °C arm (that arm prevented fever). Rewarming a spontaneously hypothermic comatose arrest patient with a forced-air warmer at 1 °C every 15 minutes. Routine hypothermia after ICH or TBI "for neuroprotection." Paralytics as the first shivering drug. Prehospital cold-fluid boluses as required care. Claiming INTREPID proved better 90-day recovery with device-based fever prevention.