Lilja: Cognitive Function in Survivors of Out-of-Hospital Cardiac Arrest After Target Temperature Management at 33ºC Versus 36ºC

Lilja et al. Circulation.2015, Published online February 13

Clinical Question

• In patients with out-of-hospital (OOH) cardiac arrest does a target temperature management (TTM) of 33°C vs. 36°C improve cognitive function at 6 months?

Design

• Randomised controlled trial
• Treating clinicians non-blinded
• Face-to-face follow up at ~180 days
  • examiners blinded for TTM allocation but not whether or not patient was a control
• With study sample size, >80% power to detect a 10% difference between the groups based on the Rivermead Behavioural Memory Test profile score

Setting

• 5 European countries
• Out of 36 sites who participated in the TTM Trial, twenty sites completed this follow-up study
• Follow-up occurred between June 2011 – September 2013

Population

• Follow-up study of TTM Trial:
  • Inclusion: adult patients who were unconscious (GCS < 8) on admission to hospital after OOH cardiac arrest of presumed cardiac cause.
  • Exclusion: > 6 hours from return of spontaneous circulation to enrolment screening, unwitnessed arrest with systole, hypothermia < 30°C, suspected or known intracranial haemorrhage.
• In addition a control group was included:
  • cohort of patients who had a ST elevation myocardial infarction treated with percutaneous coronary intervention
  • patients matched for age, gender, country and time-point of hospitalisation
  • 119 control patients
• 652 patients treated with therapeutic hypothermia included
  • comparing 33°C vs. 36°C group no significant difference in co-morbidities, bystander CPR, shockable rhythm
• 342 patients treated with TTM alive at 180 days of whom 287 completed follow up
  • comparing 33°C vs. 36°C  vs. control group
    • age: 63 vs. 60 vs. 64, p=0.04
    • previous ischaemic heart disease: 28% vs. 19% vs. 13%, p=0.01
    • bystander CPR: 74% vs. 84% vs. N/A, p=0.03
    • 1st monitored rhythm shockable: 93% vs. 96% vs. N/A, p=.032

Intervention

• 28 hours of temperature management with a target of
  • a) 33°C
  • b) 36°C
• followed by gradual warming to 37°C at 0.5°C per hour and avoidance of pyrexia for 72 hours from randomisation

Control

• Recruited from cohort of ST-elevation myocardial infarction patients treated with percutaneous coronary intervention
  • matched for age, gender, country and time-point of hospitalisation

Outcome

• Primary outcome: Rivermead Behavioural Memory Test (RBMT)
  • no significant difference between 33°C vs. 36°C, p=0.56
• Secondary outcomes:
  • Comparing 33°C vs. 36°C,
    • no significant difference in
      • Executive dysfunction as tested by Frontal Assessment Battery (FAB), p=0.43
      • Symbol Digit Modalities Test (SDMT) assessing mental speed and agitation, p=0.31
  • Including only patients who completed follow-up and the matched STEMI controls
    • STEMI controls had significantly better SDMT scores than 33°C and 36°C treatment groups (unadjusted p=0.0001, covariate-adjusted p=0.002)
    • No significant difference in:
      • RBMT (unadjusted p=0.46, covariate-adjusted p=0.06)
      • FAB, (unadjusted p=0.22, covariate-adjusted p=0.008)
      • Hospital Anxiety and Depression rating Scale (HADS),
        • depression, p=0.99
        • anxiety, p=0.89

Authors’ Conclusions

• Survivors of patients with OOH cardiac arrest who were treated with targeted temperature management of 33°C vs. 36°C had no difference in cognitive outcome

Strengths

• Randomised controlled trial
• Blinding of assessor of primary outcome to TTM allocation
• Low number of patients lost to follow up
• Investigated multiple assessment measures of cognitive outcome
• Use of CONSORT diagram

Weaknesses

• Due to the nature of the study baseline differences in cognitive function cannot be assessed. With relatively small numbers of patients included there is a greater probability of random differences in baseline cognitive function between the treatment groups
• RBMT test may be insensitive for very mild impairment. The authors stated that the updated RBMT-3 version was not available when this study commenced
• FAB has not been validated in an OOH cardiac arrest population
• Proportionately fewer TTM patients with poor neurological recovery (cerebral performance category 3 or 4) were followed-up than those with good recovery. This will bias toward less cognitive impairment in the OHCA group compared to the STEMI-control group, but should still allow accurate comparison between 33ºC and 36ºC. It will under-estimate cognitive impairment from cardiac arrest. (see table for data)