Arctic Sun 5000 for therapeutic hypothermia after cardiac arrest

Pittl et al. (2013)

Study size, design and location

80 patients.

Germany, single centre.

Prospective, randomised (1:1) study on patients who survived in-hospital and out-of-hospital cardiac arrest.

Intervention and comparator

Intervention: Arctic Sun (n=40).

Comparator: Coolgard endovascular system (Zoll) (n=40).

All patients were cooled at 33°C core body temperature for 24 hours followed by active rewarming.

Patients were followed-up for 72 hours.

Key outcomes

Neuro-specific enolase levels at 72 hours did not differ significantly between groups. Overall survival was not significantly different between groups.

Survival with good neurological outcome (CPC 1–2) during hospitalisation was identical in both groups (35.9%). There was no significant difference in cooling rate between the 2 devices.

Stability at the target temperature was significantly better in the Coolgard group. Significantly more Coolgard patients had bleeding complications compared with Arctic Sun patients.

Strengths and limitations

The study used 1:1 randomised methods but was not double blinded or adequately powered to detect differences in clinical outcomes. Some time recordings may not be reliable because they were based on witness or paramedic testimony reporting the patients' time of collapse from cardiac arrest.

Heard et al. (2010)

Study size, design and location

64 patients.

US, 6 university-affiliated hospitals.

Multicentre randomised (1:1) trial of haemodynamically stable comatose survivors of out-of-hospital cardiac arrest.

Intervention and comparator

Intervention: Arctic Sun (n=34).

Comparator: standard post-resuscitative care inducing hypothermia using cooling blankets and ice (n=30).

Key outcomes

The proportion of subjects cooled below the 34°C target at 4 hours was not significantly different between groups.

The median time to target temperature was significantly faster in the Arctic Sun group.

At 3 months follow-up, overall survival was not significantly different between the 2 groups. Survival with good neurological outcomes (CPC 1–2) was also not significantly different between groups. Rates of adverse events were not significantly different between groups.

The rate of overshooting (cooling to below 32°C) during induction was significantly higher in the standard cooling group.

Strengths and limitations

The study used 1:1 randomisation methods used at each centre but was not double blinded. Most centres were inexperienced in therapeutic hypothermia, and some centres had a gap of several months between recruitments.

There was no standardised protocol for pharmaceutical paralytics, which varied between centres. The study was also supported by company grant.

Tomte et al. (2011)

Study size, design and location

167 patients.

Norway, single centre.

Comparative 2-armed observational study enrolling consecutive comatose patients after out-of-hospital cardiac arrest treated with mild therapeutic hypothermia.

Intervention and comparator

Intervention: Arctic Sun (n=92).

Comparator: Coolgard (n=75).

Key outcomes

There were no significant differences in survival to final hospital discharge with good neurologic function (CPC 1–2) between groups.

Lengths of intensive care unit stay and respirator dependency, rates of mild therapeutic hypothermia discontinuation, and post-cooling fever were not significantly different between the 2 groups. There was no significant difference in time to achieve target temperature (34°C) between the 2 groups.

Infection rates and antibiotic treatment, arrhythmias, level of haemodynamic support, transfusion rates, presence of seizures or shivering, use of paralytic agents, and need for dialysis were also comparable between the 2 groups.

Strengths and limitations

Consecutive patient enrolment may have minimised patient selection bias.

All out-of-hospital cardiac arrest patients in the study period were included, creating a large patient cohort. The study was not randomised or blinded, and the hospital initially used Coolgard before Arctic Sun was introduced, which may have introduced bias in terms of clinical developments in hospital standard care.

The study was underpowered to assess neurologic-intact survival as an outcome between the 2 groups.

Shinada et al. (2014)

Study size, design and location

51 patients.

Japan, single centre.

Consecutive enrolment of patients who did not regain consciousness after return of spontaneous circulation following cardiogenic cardiac arrest.

Intervention and comparator

Intervention: Arctic Sun 2000 (n=40).

Comparator: conventional standard cooling blankets (n=11).

Key outcomes

Time to achieve the target temperature was significantly shorter with Arctic Sun than with standard cooling blankets.

There was no significant difference between groups in number of patients achieving CPC 1–2 at 30 days' hospitalisation.

There were no differences in the incidence of complications during hypothermia, including pneumonia and bleeding, between the 2 groups.

Minimal body temperature achieved during therapeutic hypothermia was significantly lower in the standard cooling group.

Strengths and limitations

Consecutive patient enrolment may minimise patient selection bias and all patients meeting inclusion criteria in the study period were included. However, the study was non-randomised, retrospective and unblinded. There were also more patients in the Arctic Sun arm than the much smaller comparator arm.