TTM 48

Targeted Temperature Management for 48 vs 24 Hours and Neurological Outcome After Out-of-Hospital Cardiac Arrest

Krikegaard. JAMA 2017: 25;318(4):341-350 doi: 10.1001/jama.2017.8978

Clinical Question

  • In adults with out of hospital cardiac arrest (OOHCA) due to a presumed cardiac cause, who are post ‘return of spontaneous circulation’ (ROSC), does cooling to 33°C for 48 hours, compared with 24 hours, improve improved neurological outcomes at 6 months?


  • Therapeutic cooling was established as an intervention for OOHCA patients and ROSC with the perceived benefit of reducing cerebral oxygen demand.
  • In the wake of the Targeted Temperature Management (TTM) trial, current guidelines, including the most recent European Resuscitation Council guidelines, now recommend TTM between 32-36°C, rather than the previously recommended 32-34°C(1).
  • While current guidelines also recommend a minimum duration of 24hrs for TTM, this recommendation in itself is low quality evidence and is primarily based on the protocol undertaken in what are considered the two largest and better quality RCTs in this area.
  • These trials were not however designed to look at duration of cooling specifically.
  • Although previous observational studies comparing 24 to 72 hours of hypothermia in OOHCA patients have failed to show any difference in mortality, randomized control trial (RCT) data looking at optimal periods of cooling as related to long term neurological outcomes is lacking and the optimal duration for cooling to any temperature, within the temperature bracket of 32-36°C remains poorly evidenced based.


  • Randomised controlled trial
  • Web based randomisation with variable block size
  • Stratified by age, hospital site, initial rhythm
  • Parallel groups
  • Pragmatic trial
  • Clinicians and patients/relatives non-blinded
  • Blinding of outcome assessors
  • Modified intention to treat analysis (included patients who initiated the intervention + did not withdraw consent)
  • Outcomes assessed at 6 months by telephone or person-to-person interview
  • Power calculation: 338 patients required to demonstrate an absolute difference of 15% (50% vs. 65%) in the primary outcome with a false negative rate of 20% and a false positive rate of 5%


  • 10 European Intensive Care Units (Denmark, Belgium, Finland, Estonia, Germany, Norway)
  • February 2013 – June 2016


  • Inclusion criteria:
    • Adults aged 18-80 years with OOHCA due to presumed cardiac cause with sustained ROSC and a Glasgow coma score (GCS) of ≤ 8.
  • Main exclusion criteria:
    • Unwitnessed asystole as initial presenting rhythm.
    • Estimated time from collapse to return of spontaneous circulation (ROSC) of >60min
    • Estimated time from OOHCA to cooling >240min
    • OOHCA due to presumed non-cardiac cause
    • Severe coagulopathy, not due to thrombolysis or anticoagulant therapy.
    • Suspected or confirmed intracerebral bleeding or acute stroke
    • Persistent cardiogenic shock
  • 907 patients assessed; 552 excluded
  • 176 and 179 patients randomised to 48hr and 24hr TTM groups respectively
  • Comparing baseline characteristics of 48hr vs. 24hr groups – no significant differences
    • Age: 61 vs. 60
    • Unwitnessed cardiac arrest: 9% vs. 8%
    • Shockable rhythm: 91% vs. 86%
    • Time to ROSC, median (IQR): 20 (15-30) vs. 21 (16-27)
    • Immediate coronary angiography: 83% vs. 82%
  • Comparing interventions in 48hr vs. 24hr group
    • Time from ROSC to start of TTM: 102 vs. 112min
    • Time from ROSC to achievement of target temp: 281 vs. 320min, p=0.01
    • Invasive cooling catheter: 65% vs. 59%
    • Temp at 60 hours: 37.0 vs. 37.2, p=0.02


  • TTM to 33°C (+/-1°C) for 48hrs.
    • Both surface or invasive cooling methods allowed.
    • Temperature measured according to local practices (bladder, oesophagus or intravascular)
    • Rewarming post period of TTM was to maximal rate of 0.5°C/hour


  • TTM to 33°C (+/-1°C) for 24hrs
    • Other specifics as for the intervention group

Management common to both groups

  • With the exception of brain death, refractory shock or multiple organ failure, treatment centres agreed to provide active treatment until 72hrs after normothermia.
  • For patients who remained unconscious despite cessation of sedation, the attending physician chose modality(s) of neurological assessment to include a combination of imaging (CT, MRI) as well as EEG and/or somatosensory evoked potentials (SEP)
  • Decisions to withdraw life support made according to local protocols


  • Primary outcome:
    • Favourable neurological outcome at 6 months based on Cerebral Performance Category scores – no significant difference
      • A score of 1 (able to work and lead a normal life) or 2 (moderate cerebral disability but able to carry out part time work) was considered favourable.
      • Comparing intervention vs. control group:
        • 69% vs. 64%, difference 4.9% (95% C.I. -5 to 14.8), RR 1.08, p=0.33
        • Lack of significant effect was sustained across subgroups
  • Secondary outcomes, comparing intervention vs. control group
    • 6-month mortality – No significant difference.
      • 27% vs. 34%, difference -6.5% (95% C.I. -16.1 to 3.1), RR 0.81 p=0.19
    • Time to death – No significant difference.
    • Other outcomes
      • Adverse events – Significantly higher in intervention group (97% vs. 91%, p=0.03)
        • Hypotension 62% vs. 49%, p=0.013
        • No significant difference in rates of pneumonia or bleeding between groups
      • Resource use
        • Ventilation time – significantly higher in the intervention group.
          • 120hrs V 87hrs (26% difference, 95% CI 16 to 36%, p = <0.001)
        • ICU length of stay was significantly higher in intervention group where as overall hospital stay was not significantly different between groups

Authors’ Conclusions

  • The authors conclude that this study did not show a difference for primary outcome, however they feel that their study may have had limited power to detect what may be a clinical important difference and suggest further research may be warranted.


  • Multicentre randomised trial across ten hospitals
  • Objective neurological assessment carried out by a blinded assessor
  • The primary outcome was highly relevant to patients (i.e based on functional outcome)
  • Protocol was designed with the intention of being pragmatic.


  • The authors point out that because sample size was calculated using a 15% absolute difference, the risk of type 2 error is increased. A trial to elucidate if the 5% difference in primary outcome detected could be significant would require approximately 10-fold more patients.
  • 60% of all OOHCA all patients were excluded from this trial and an upper age limit of 80 years was placed for inclusion. This limits the external validity.
  • The study protocol was written before the TTM trial was published. A 3rd group with temperature maintained at 36C for 48 hours would have been useful
  • In both the intervention and the control groups, following completion of the targeted temperature management, ~50% of patients had a temperature of >37C. Therefore this study could not tell us if any differences reported were from prolonged therapeutic hypothermia or the avoidance of pyrexia.

The Bottom Line

  • In patients with an out-of-hospital cardiac arrest of presumed cardiac origin, this study reported that maintaining temperature at 33 degrees for 24 vs 48 hours resulted in no significant difference in neurologic outcome.
  • If we consider adverse outcomes and implications for distributions of resources, these results may also point toward possible disadvantages to cooling for more than 24 hours.
  • My current practice is to use targeted temperature management to a target of 36C for 30 hours followed by avoidance of pyrexia for 24-48 hours. Based on this study I will not change this practice.
  • Hopefully the outcome of TTM2 will help shed some further light on best evidence based practice in this area

External Links


Summary author: Aoife Abbey @whistlingdixie4
Summary date: July 30th 2017
Peer-review editor:@davidslessor

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.