Targeted Temperature Management
The trials by the HACA group and by Bernard and colleagues had provided fairly convincing evidence for the benefits of targeted temperature management (A.K.A. therapeutic hypothermia) after cardiac arrest. However, there were certain concerns about these trials (VF rhythm only, pyrexia in control group) that left some questions unanswered. The TTM group recognised these limitations and conducted this trial to try and provide answers.
Targeted temperature management after cardiac arrest.
Targeted temperature management after cardiac arrest.
What did they do?
Design: Multicentre, single blinded, randomised control trial. Patient assessment and analysis of results blinded to allocation
Intervention: Targeted temperature of 33 or 36oC. Rapid cooling with ice, cold fluids, or temperature management systems at the discretion of the treating site, with sedation. Intervention period of 36 hours – cooling for 28 hours then 0.5oC per hour rise to 37oC. Mandated sedation stopped at 36 hours. Target of temperature below 37.5oC until 72 hours. Neurological assessment of patients who remained unconscious at 72 hours by blinded clinician with recommendation for withdrawal of care if appropriate (final decisions remained with the clinical team)
Inclusion Criteria: Out of hospital cardiac arrest (any rhythm), aged 18 or over, depressed conscious level (GCS <8), presumed cardiac cause, ROSC of > 20 mins after resuscitation.
Exclusion Criteria: Unwitnessed arrest with initial asytolic rhythm, delay of > 240 minutes from ROSC until screening, temperature below 30oC, known or suspected intracranial haemorrhage or stroke.
Population: Patients from 36 ICUs in Europe and Australia.
Primary Outcome: All-cause mortality through end of trial
Secondary Outcome: Composite of poor neurological function or death at 180 days. Defined by a Cerebral Performance Category score of 3 to 5 and a score of 4 to 6 on the modified Rankin scale. Mortality and neurological function also examined individually.
Intervention: Targeted temperature of 33 or 36oC. Rapid cooling with ice, cold fluids, or temperature management systems at the discretion of the treating site, with sedation. Intervention period of 36 hours – cooling for 28 hours then 0.5oC per hour rise to 37oC. Mandated sedation stopped at 36 hours. Target of temperature below 37.5oC until 72 hours. Neurological assessment of patients who remained unconscious at 72 hours by blinded clinician with recommendation for withdrawal of care if appropriate (final decisions remained with the clinical team)
Inclusion Criteria: Out of hospital cardiac arrest (any rhythm), aged 18 or over, depressed conscious level (GCS <8), presumed cardiac cause, ROSC of > 20 mins after resuscitation.
Exclusion Criteria: Unwitnessed arrest with initial asytolic rhythm, delay of > 240 minutes from ROSC until screening, temperature below 30oC, known or suspected intracranial haemorrhage or stroke.
Population: Patients from 36 ICUs in Europe and Australia.
Primary Outcome: All-cause mortality through end of trial
Secondary Outcome: Composite of poor neurological function or death at 180 days. Defined by a Cerebral Performance Category score of 3 to 5 and a score of 4 to 6 on the modified Rankin scale. Mortality and neurological function also examined individually.
What did they find?
Numbers: 950 patients enrolled (predetermined target). Final numbers: 473 patients in 33oC group (with 3 not receiving the intervention), 466 patients in 36oC group (with 4 not receiving the intervention).
Results: There was no significant difference in the primary outcome between the groups. Mortality was 235/473 patients (50%) in the 33oC group and 225/466 patients (48%) in the 36oC group. The p value for this was 0.51.
This provided a hazard ratio of 1.06 (95% CI 0.89 – 1.28) for cooling to 33oC.
There was no difference between the groups in terms of their neurological status using either the CPC or modified Rankin criteria.
The Hazard ratio for death or poor neurological outcome in the 33oC group was 1.04 (95% CI 0.89 – 1.17), with a p value of 0.67).
16 patients in the 33oC group were rewarmed early at the discretion of the treating physician (arrhythmia the most common reason).
There was a non-statistically significant increase in serious adverse events in the 33oC group. One or more serious events occurred in 439/472 patients (93%) in this group compared to 417/464 patients (90%) in the 36oC group. This produced a hazard ratio of 1.03 (95% CI 1.00 – 1.08, p = 0.09).
Results: There was no significant difference in the primary outcome between the groups. Mortality was 235/473 patients (50%) in the 33oC group and 225/466 patients (48%) in the 36oC group. The p value for this was 0.51.
This provided a hazard ratio of 1.06 (95% CI 0.89 – 1.28) for cooling to 33oC.
There was no difference between the groups in terms of their neurological status using either the CPC or modified Rankin criteria.
The Hazard ratio for death or poor neurological outcome in the 33oC group was 1.04 (95% CI 0.89 – 1.17), with a p value of 0.67).
16 patients in the 33oC group were rewarmed early at the discretion of the treating physician (arrhythmia the most common reason).
There was a non-statistically significant increase in serious adverse events in the 33oC group. One or more serious events occurred in 439/472 patients (93%) in this group compared to 417/464 patients (90%) in the 36oC group. This produced a hazard ratio of 1.03 (95% CI 1.00 – 1.08, p = 0.09).
Is it any good?
Overall: Very Good
Strengths: Blinding of the clinical assessors and continuation of blinding through the whole analysis phase.
A robust pragmatic design.
A relatively large number of patients.
Weaknesses: Targeted quite a significant difference in outcome with their power calculation, hence the exact meaning of a negative result is slightly debatable; not finding a difference is different from there not being a difference.
Still some patient selection via the exclusion criteria e.g. unwitnessed arrests
Strengths: Blinding of the clinical assessors and continuation of blinding through the whole analysis phase.
A robust pragmatic design.
A relatively large number of patients.
Weaknesses: Targeted quite a significant difference in outcome with their power calculation, hence the exact meaning of a negative result is slightly debatable; not finding a difference is different from there not being a difference.
Still some patient selection via the exclusion criteria e.g. unwitnessed arrests
Final Thoughts
A really important paper in the hunt for answers about the best management approach to patients after cardiac arrest. It was a very pragmatic study that strongly suggests that there is no significant benefit to cooling to 33oC, and instead the target should be 36oC. However, it certainly isn't incontrovertible proof of this and there are still unanswered questions about the benefits of cooling.
Written: Tom Heaton
Reviewed: Not done
25th April 2016
Written: Tom Heaton
Reviewed: Not done
25th April 2016
