Scandinavian Clinical practice guidelines for therapeutic hypothermia and post-resuscitation care after cardiac arrest


Prof. Maaret Castrèn
Department of clinical science and education
Karolinska Institutet


Background and aim: Sudden cardiac arrest survivors suffer from ischaemic brain injury that may lead to poor neurological outcome and death. The reperfusion injury that occurs is associated with damaging biochemical reactions, which are suppressed by mild therapeutic hypothermia (MTH). In several studies MTH has been proven to be safe, with few complications and improved survival, and is recommended by the International Liaison of Committee on Resuscitation. The aim of this paper is to recommend clinical practice guidelines for MTH treatment after cardiac arrest from the Scandinavian Society of Anaesthesiology and Intensive Care Medicine (SSAI).

Methods: Relevant studies were identified after two consensus meetings of the SSAI Task Force on Therapeutic Hypothermia (SSAITFTH) and via literature search of the Cochrane Central Register of Controlled Trials and Medline. Evidence was assessed and consensus opinion was used when high-grade evidence (Grade of Recommendation, GOR) was unavailable. A management strategy was developed as a consensus from the evidence and the protocols in the participating countries.

Results and conclusion: Although proven beneficial only for patients with initial ventricular fibrillation (GOR A), the SSAITFTH also recommend MTH after restored spontaneous circulation, if active treatment is chosen, in patients with initial pulseless electrical activity and asystole (GOR D). Normal ethical considerations, premorbid status, total anoxia time and general condition should decide whether active treatment is required or not. MTH should be part of a standardized treatment protocol, and initiated as early as possible after indication and treatment have been decided (GOR E). There is insufficient evidence to make definitive recommendations among techniques to induce MTH, and we do not know the optimal target temperature, duration of cooling and rewarming time. New studies are needed to address the question as to how MTH affects, for example, prognostic factors.

Sudden cardiac arrest survivors suffer from an ischaemic brain injury that may lead to poor neurological outcome and death. The injury during cardiopulmonary resuscitation (CPR) and successfully restored spontaneous circulation (ROSC) is described as a global ischaemia–reperfusion injury. This initiates a cascade of deleterious inflammatory reactions in the body that may continue for several days. Treatment directed at minimizing the inflammatory response and cell death in the reperfusion period may improve outcome following cardiac arrest. Until recently, post-resuscitation treatment was regarded as the ‘weak link in the chain of survival’.1 However, the introduction of mild therapeutic hypothermia (MTH), defined as a reduction of body temperature to 32–34 °C, following cardiac arrest, has emphasized the importance of appropriate post-resuscitation treatment. The main protective effect of MTH is a reduction of the global cerebral injury, such as a reduction of the following: body and cerebral metabolism,2,3 apoptosis,4–6 influx of Ca2+ into the cell,7 intra- and extracellular acidosis,8,9 accumulation of the exitotoxic neurotransmitter glutamate,10–12 release of glycine,13 inflammation,14,15 nitric oxide production7 and free radical production.11,16 These are all factors associated with poor outcome. It will also reduce the disruption in the blood–brain barrier as well as vascular permeability and thereby decreased cerebral oedema formation.17–19

MTH has been recommended by the International Liaison Committee on Resuscitation (ILCOR) since 2003.20 The aim of this paper is to review the recent publications on MTH after cardiac arrest, and recommend Clinical Practice Guidelines for MTH treatment from the Scandinavian Society of Anaesthesiology and Intensive Care Medicine (SSAI). The proposed management strategy represents a consensus from evidence and protocols in the participating five Nordic countries.


Relevant studies were identified after two consensus meetings of the SSAI Task Force on Therapeutic Hypothermia (SSAITFTH) and via a literature search from the Cochrane Central Register of Controlled Trials and Medline. The following search words were used: therapeutic hypothermia, induced hypothermia, post-resuscitation, cardiac arrest, ventricular fibrillation, cardiopulmonary resuscitation, CPR, outcome, hypoxia–ischaemia and brain. Evidence was assessed and consensus opinion was used when high-grade evidence was unavailable. A management strategy was developed as a consensus from the evidence and the protocols in the participating countries.

The grading system used is presented in Table 1,21 with grading of recommendation (GOR) from A to E, and grading of evidence (GOE) from I to V. Only clinical studies are graded. In this SSAI Guideline Recommendation, we have attempted to address several important aspects regarding post-resuscitation care and treatment with MTH such as how, when, for how long, to whom, methods, timing of other interventions including monitoring, prognostication as well as complications and side effects.

Table 1. 
Grading system used.
1. Grading of recommendations
 A. Supported by at least two level I investigations
 B. Supported by one level I investigations
 C. Supported by level II investigations only
 D. Supported by at least one level III investigation
 E. Supported by level IV or V evidence
2. Grading of evidence
 I. Large, randomized trials with clear-cut results; low risk of false-positive (α) error or false-negative (β) error
 II. Small, randomized trials with uncertain results; moderate-to-high risk of false-positive (α) and/or false-negative (β) error
 III. Non-randomized, contemporaneous controls
 IV. Non-randomized, historical controls, and expert opinion
 V. Case series, uncontrolled studies and expert opinion


The support of MTH after cardiac arrest is based on three randomized-controlled studies22–24 considered to be GOE I–II. There have also been several previous studies with historical controls25–27 considered as GOE IV, and experimental studies.28–31 Recently, several observational studies, with historical controls32–35 (GOE IV), uncontrolled studies36,37 (GOE V) and a European registry39 (GOE V), all show that MTH is feasible, safe, has few side effects and seems to contribute toward improved survival.

Which patients to cool?

ILCOR recommended in 2003 that all comatose cardiac-arrested patients with initial ventricular fibrillation (VF) should be cooled for 12–24 h.20 Further, they stated: ‘For any other rhythm, or cardiac arrest in hospital, such cooling may also be beneficial’.20 Recently, observational studies32,34 (GOE IV) and two registry reports39,40 (GOE V) have reported the feasibility of treating patients with non-VF cardiac arrest. Although the prognosis for this group of cardiac-arrested patients is worse,39 recent survival data on non-VF patients are promising38 (GOE V). In addition, despite being a different aetiology, the promising results of 72 h of MTH use on newborns after asphyxial cardiac arrest in two randomized studies41,42 (GOE I) favour the use of MTH in the presence of global cerebral ischaemia.

Recommendation: Although proven beneficial only for comatose patients with initial VF (GOR A), we recommend MTH after ROSC, if active treatment is decided, in comatose patients with initial pulseless electrical activity and asystole (GOR D). Normal ethical considerations, premorbid status, total anoxia time and general condition should decide whether active treatment is required or not.

Methods for MTH

Several different methods can be used to achieve MTH. It is not the aim of this paper to present all the available different cooling methods, but rather to recommend the interested reader to a recent review by Holzer.43 In Table 2, however, we provide a brief overview of the advantages and disadvantages of different cooling methods, ranging from simple external methods to advanced invasive techniques. A combination of different methods may be necessary, at least during induction of MTH. No specific methods can be recommended, because there are only a few studies comparing feasibility and efficacy44,45 (GOE V), and no studies have evaluated implications on survival between different cooling methods. Some differences in cooling rate and stability have been described44,45 (GOE III, IV). Recently, a pilot study identified significant differences in the rating of key nursing aspects of different cooling methods46 (GOE V).

Table 2. 
Methods for induction and maintenance of mild therapeutical hypothermia.
Simple external cooling
Ice-bags, cold wet blankets, (ice-cold water, alcohol, etc.)Simple, non-invasive
easy initial cooling
prehospital use
Long time to reach target temperature
local wounds
body fat isolates
Advanced external cooling
Cooling blankets, pads, dress or similar devices (mainly water-filled or other cooling elements)
more advanced systems such as cooling tents (with cold air), cooling beds/mattresses (with cold water)
Relatively stable
excellent for maintenance
easy to apply and use
can be easily combined with ice-cold fluids
very fast cooling with some devices
prehospital use (some of the devices)
Varying effect, depending on contact area, circulating substance and system used
some devices not for prehospital use
costs (system dependent)
cooling tent and beds/mattresses requires space
Infusion of cold fluids
0.9% saline peripheral intravenously, 40 ml/kg, 1–3 lFast, easy and cheap induction of cooling
prehospital use
(requires cooling capacity in the ambulance)
Not sufficient for maintenance (can be used in combination with other methods)
large volume contraindicated if major left heart pump failure
Endovascular cooling
A saline-filled catheter for thermo-regulationStable
excellent for maintenance
less shivering
Invasive procedure
skilled personnel for catheter placement (anaesthesiologists)
not for prehospital use

MTH treatment can be divided into three parts: rapid induction, stable and controlled maintenance and controlled rewarming. Induction can easily be induced with ice-cold i.v. fluids (30–40 ml/kg)35,47–50 (GOE II–V) or removal of clothes and with the use of ice packs, placed in the groins, armpits and around the neck and head33,51 (GOE IV and II). Although traditional external cooling has proven its feasibility33,51,52 (GOE IV, V, II), its use may lead to overcooling53 (GOE V). Ice-cold i.v. fluids alone are not sufficient for keeping the patients in a stable hypothermic state over time54 (GOE V). Shivering can be avoided by primarily deepening the sedation and, if necessary, single dosages/infusion of muscle relaxants. Rewarming can be performed with external or invasive cooling devices or with other more conventional heating systems.

To avoid overcooling, appropriate temperature monitoring is absolutely necessary, but studies focusing on the optimal temperature monitoring sites during cooling after cardiac arrest are lacking. Tympanic, nasopharyngeal and rectal probes are useable, but the bladder, oesophagus or blood (pulmonary artery or Picco catheter) are most frequently used for temperature control during MTH maintenance.55

Recommendation: We recommend that a rapid, efficient and safe cooling strategy be used (GOR D). There are no specific methods for MTH that can be recommended. Each institution should use a method (or a combination of methods) that suits their infrastructure, logistics, financial resources and treatment plan.

When, and for how long?

The real benefit of rapid cooling, the optimal target temperature, the duration of the cooling and the rewarming phase are currently unknown. Animal studies have demonstrated that hypothermia initiated during or immediately after ROSC is associated with better organ preservation and increased survival.56,57 Recently, in 49 patients treated with MTH, early achievement of the target temperature was an independent factor for good outcome58 (GOE V). However, larger trials are warranted to assess the advantage of early vs. late cooling in humans. It is worth noting that although it took 8 h to reach the target temperature in the HACA trial, there was still a beneficial effect on cerebral outcome and death compared with normothermic patients21 (GOE I).

The optimal duration of therapeutic hypothermia is undetermined. All recent clinical studies have maintained the target temperature for 24 h32–35 (GOE III–V), as in the HACA study22 (GOE I). A European hypothermia registry study39 (GOE V) reported that the majority of the patients were MTH treated for 24 h. However, 12-h protocols have also been used.23,33 Newborns with asphyctial cardiac arrest have been successfully treated for 72 h41,42 (GOE I), and it may be that even adults with a severe reperfusion injury due to hypoxic-induced cardiac arrest should be treated with MTH of a longer duration. Further studies are warranted to address this important question.

The rate of rewarming is not known, but the traditional recommendation is 0.3–0.5 °C20 (GOE I–V). Rebound hyperthermia should be avoided.59,60

Recommendation: We recommend that MTH be initiated as early as possible after the decision has been made (GOR E), and be maintained for 24 h (GOR A–E).

Timing of other interventions

Comatose patients in the post-resuscitation period are critically ill and require extensive intensive care treatment depending on the cause of the arrest, the severity of the post-resuscitation disease and myocardial dysfunction61,62 or the presence of general complications. Each hospital should have a well-defined standardized plan, including MTH, for intervention and treatment according to the local conditions, infrastructure and logistics (GOE IV). Such a plan enables doctors and nurses to focus on when and how to treat, and monitor, the patients under different circumstances.33Table 4 summarizes these interventions, and the most important ones are as follows:

Table 4. 
A standardized treatment plan during the postresuscitation period.
ReperfusionEarly reperfusionPCI, thrombolytic agents
TemperatureAs soon as possible
32–34°C for 24 h
(12–72 h?)
Cold fluids, blankets, ice packs, external and internal devices
Blood pressureMAP≥65–70 mmHgVolume, vasopressors, inotropic agents, IABP. Drug treatment according to local procedures (dopamine, norepinephrine, epinephrine, dobutamine and levosimedane)
Pulse40–100/minVolume, sedation, glycerylnitrate, and β-blockers
CVP8–12 mmHg (individual differences)Volume, vasodilatation
Arterial blood gasesSO2: 95–98Respirator adjustment (FiO2, PEEP, avoid hyperoxia
paO2: 10–15 kPa 
pCO2: 5.0–6 kPaAvoid hyperventilation
pH>7.1, BE>−10Natriumbicarbonate, Trometamol
HaemoglobinNormal valuesTransfusion if necessary
ElectrolytesNormal valuesSubstitution or specific treatment if required (obs hyperkalaemia during rewarming)
Diuresis>0.5–1.0 ml/kg/hVolume, furosemide, vasopressors
Blood glucose5–8 mmol/lActrapid (observed hypoglycaemia)
FluidsPositive fluid balanceCristalloids, colloids
NutritionEarly nutritionEnteral 10 ml/h and G5% 1–1.5 l/24 h
SedationMAAS 0, no pain, comfort, no shiveringFentanyl and propofol or midazolam (or other according to local protocol) (drug doses: titrate to wanted effect)
Muscle relaxationAvoid shiveringIf needed, rocuronium or cisatracurium
SeizuresEarly diagnosis, treatment or preventionIncrease sedation, or specific anticonvulsive medication.
Rewarming0.3–0.5°C/h until 37°CDepending on equipment used for MTH
PrognosticationNot earlier than 72 h post arrestEEG, SEP, clinical signs (with caution), NSE, Protein s-100

1. Diagnosing the cause of the cardiac arrest. Early coronary angiography with subsequent PCI should be recommended for primary resuscitated cardiac-arrested patients suspected to have myocardial infarction or severe coronary heart disease34–36 (GOE IV, V). MTH can be initiated and maintained during coronary angiography and PCI34,63 (GOE IV) Pre- or in-hospital thrombolysis is recommended for patients with ST elevation if there are no facilities for immediate PCI. Other causes of arrest, both of cardiac and non-cardiac aetiology, should also be diagnosed and receive specific treatment, if possible. A CT scan of the brain must be considered early if a cerebral haemorrhage or a similar condition is suspected.

2. Early stabilization and normalization of haemodynamics. The optimal blood pressure after ROSC is not known, but both the damaged brain and the heart must be adequately perfused without placing undue strain on the heart. A positive fluid balance, use of vasopressors, inotropic drugs and/or the intra-aortic ballon pump are required depending on the clinical situation and cause of the arrest34,62 (GOE IV–V). Cardiogenic shock should not be considered as a contraindication for MTH treatment34,36,62,64 (GOE IV–V).

3. Mechanical ventilation. For controlled ventilation, mechanical ventilation should be initiated as early as possible, aiming for normo-oxygenation and -ventilation. Both hyperoxia65 and hypoxia should be avoided. Hyperventilation may reduce cerebral perfusion,66 and hypoventilation may have serious negative effects on cardiopulmonary interactions, causing a further increase in tissue acidosis (GOE II).

4. Treat hyperglycaemia. Non-intervention studies have shown increased mortality among cardiac arrest patients with high levels of blood glucose60,67,68 (GOE IV–V). Aggressive insulin treatment in patients with critical illness has reduced long-time mortality69 (GOE I). Thus, hyperglycaemia should be avoided and treated with insulin infusions (GOR B). We do not know the optimal target, but a strict protocol does not seem necessary and may lead to increased numbers of dangerous hypoglycaemia.70 A target level of 5–8 mmol/l seems appropriate (GOE II, IV). Hypoglycaemia may also occur during rewarming because insulin sensitivity and secretion will increase with temperature.

5. Monitoring. These patients, therefore, require routine intensive care monitoring, which can be divided into two categories (Table 3): recommended (general intensive care monitoring) and optional (both more advanced haemodynamic monitoring and cerebral monitoring). For diagnostics, control of the myocardial dysfunction and optimizing therapy, at least echocardiography, should be performed daily for the first days after a cardiac arrest34 (GOE IV).

Table 3. 
Monitoring for the postresuscitation period.
 Arterial catheter
 O2 saturation
 Continuous ECG
 Central venous pressure
 Temperature (bladder, oesophagus)
 Arterial blood gases (pH, BE, pCO2, pO2)
 Blood glucose, electrolytes and general blood sampling
 X-ray thorax
 Echocardiography (daily for the first days)
 sVO2 (from the central venous line)
 PA catheter/PICCO or other cardiac output monitoring
 EEG (on indication/continuously): early seizure detection and treatment
 SSEP: prognostication (after day 3)
 NSE/protein s-100: prognostication

6. Awakening and extubation. When the patient reaches normothermia, the sedation may be discontinued. The haemodynamic, respiratory and neurological status must be monitored carefully before the patient is extubated. Because of reduced elimination caused by MTH, the different drugs used for sedation will have prolonged effects, which may prolong awakening and time for extubation.71,72

Recommendation: Standardized treatment with a focus on early treatment of the cause and normalizing haemodynamics and metabolism is required (GOR C-E). Although difficult to prove through randomized studies, we recommend that all patients be monitored routinely with general intensive care monitoring following a local standardized treatment protocol (GOR E). The additional use of more advanced haemodynamic monitoring and/or cerebral monitoring must be decided based on local use and protocols. Frequent blood gas analyses for control of oxygenation, ventilation, blood sugar and electrolytes are necessary, especially during cooling and rewarming (Table 4).


Time from arrest until completed rewarming under normal circumstances is at least 36 h on average. Further, clinical prognostication on the first 2–3 days after a cardiac arrest is difficult73–75 (GOE IV). Experienced physicians reviewed clinical data in resuscitated patients 24 h after cardiac arrest and were able to predict the clinical outcome correctly in only 52% of the patients75 (GOE IV). However, signs such as persisting coma after discontinuation of sedatives (prolonged effects during therapeutic hypothermia), no signs of breathing, absence of papillary light reflexes or corneal reflexes and no motor response to pain on days 2–3 post arrest are clinical signs of a bad outcome73,74 (GOE IV). In addition, somatonsensory-evoked potentials on day 3 are the best predictors of a bad outcome76,77 (GOE II). These studies, however, are from the era before MTH treatment, and we do not know how MTH will affect these prognostication aspects. As metabolism and drug elimination is reduced, it is reasonable to assume that prognostication may also be delayed. Recent reports show good survival even in patients with initially bad prognostic signs78,79 (GOE V).

Another important aspect is EEG, either continuously or when indicated, to detect, prevent and treat early occurrence of seizures. In a Swedish study, continuous EEG was used as a predictor of outcome80 (GOE V). Moreover, biochemical tests such as s-100 and NSE have also been used successfully81 (GOE II), and in a recent study NSE corelated with time to reach the target temperature58 (GOE V). The prognostic value of CT and MRI is not known today, but both are recommended if a cerebral cause or complication is suspected. More studies are warranted on prognostication in patients treated with MTH.

Recommendation: Although not much data are available, we recommend that the outcome prognostication following MTH should not be initiated before 72 h post arrest to avoid early treatment withdrawal in patients who may still recover (GOR C). We recommend the use of EEG, SEP and biochemical markers if available (GOR D).

Adverse effects and complications of MTH

Although MTH exerts various effects on several organ systems, clinical studies show that MTH does not increase the risk or the number of complications compared with similar patients not treated with MTH22,34,35,39,82 (GOE I–V). Induction of hypothermia may affect haemodynamic stability. However, echocardiographic and invasive studies during infusion of cold fluids to induce hypothermia several hours after ROSC show that haemodynamic stability is preserved49,50 (GOE V), and a trend towards improved cardiac output and increased arterial pressure after infusion of ice-cold saline has been reported47 (GOE V). Comatose cardiac arrest victims often suffer from a post-resuscitation sepsis-like syndrome with reduced systemic vascular resistance50,82 (GOE V), and cooling may therefore be beneficial32,34 (GOE IV). A reduced heart rate may also be beneficial, because β-blocker use in the early post-resuscitation period has been associated with an improved outcome68 (GOE V). MTH may have an effect on the platelet count and clotting time, increasing the risk of haemorrhagic complications, and may impair immune function and increase the risk of infections. However, this has not been confirmed in clinical studies22,23,34,35,39 (GOE I–V). MTH may induce hyperglycaemia by decreasing insulin sensitivity and secretion, but this is usually easy to control with insulin treatment. Prolonged drug effects due to decreased clearance should be kept in mind71,72 (GOE IV).

Pneumonia, due to aspiration and/or mechanical ventilation, may be the most important complication during the post-resuscitation period, occurring in approximately 50% of patients.21,33 The rate is similar in patients treated and not treated with MTH22,33,34 (GOE 1, IV).

Renal failure has not been reported to be present more often in patients treated with MTH than in ICU patients in general35 (GOE IV). MTH may cause tubular dysfunction and increase diuresis, and may lead to hypophosphataemia, hyponatraemia (may increase cerebral oedema due to hypoosmolality), hypomagnesaemia, hypocalcaemia or hypokalaemia71 (GOE IV). MTH-related deaths have not been reported.

Conclusion: MTH after cardiac arrest appears to be safe, well tolerated and not associated with more complications than in patients not treated with MTH (GOR A).

Recommendation: Careful control of blood glucose, electrolyte levels and fluid balance is advised during induction, maintenance and the rewarming phase of MTH (GOR E).

Summary of the recommendations

Although proven beneficial only for patients with initial VF (GOR A), the SSAITFTH also recommends MTH after ROSC, if active treatment is decided, in patients with initial pulseless electrical activity and asystole (GOR D). Normal ethical considerations, premorbid status, total anoxia time and general condition should decide whether active treatment is required or not. MTH should be part of a standardized treatment protocol, and initiated as early as possible after indication and treatment has been decided (GOR E). There is insufficient evidence to make definitive recommendations among the techniques to induce MTH, and we do not know the optimal target temperature, duration of cooling and rewarming time. New studies are needed to address the question as to how MTH affects, for example, prognostic factors.