Therapeutic Hypothermia after Cardiac ARrest - Bernard et al.
Out of hospital cardiac arrest (OOHCA) is relatively common and associated with a poor prognosis. Much of the morbidity and mortality in initial survivors may be liked to anoxic neurological injury at the time of the arrest and persisting some hours after resuscitation. Studies in animal models and low quality human studies had suggested some benefit from therapeutic hypothermia in these cases. Bernard et al designed this RCT to try and provide more information on the potential benefits benefits.
Treatment of Comatose Survivors of Out-of-hospital Cardiac Arrest with Induced Hypothermia
Treatment of Comatose Survivors of Out-of-hospital Cardiac Arrest with Induced Hypothermia
What did they do?
Design: A single blinded randomised control trial looking at the described treatment intervention. Randomisation by date (odd versus even number days).
Control: ‘Normothermia’. Provided well protocolised care with no attempts to control temperature other than passive rewarming.
Intervention: Aggressively cooled to core temperature of 33 degrees Celsius. Ice packs used by paramedics prehospital and then in ED.
Population: 4 participating emergency departments in Melbourne, Australia. Patients received a well protocolised treatment plan, targeting specific physiological variables. All patients received lidocaine.
Inclusion criteria: Initial rhythm VF, successful return of spontaneous circulation, persisting coma after ROSC, brought to participating ED.
Exclusion criteria: Men under age of 18 years, women under age of 50 years (due to possibility of pregnancy), cardiogenic shock, alternative causes of coma, no ICU bed available.
Primary outcome: Neurological outcome good enough to allow discharge to home or rehabilitation facility was determined as a ‘good outcome’. Death or discharge to long term nursing care was determined as a ‘bad outcome’
Secondary outcomes: Looking at the haemodynamic, haematological and biochemical effects of the hypothermia.
Control: ‘Normothermia’. Provided well protocolised care with no attempts to control temperature other than passive rewarming.
Intervention: Aggressively cooled to core temperature of 33 degrees Celsius. Ice packs used by paramedics prehospital and then in ED.
Population: 4 participating emergency departments in Melbourne, Australia. Patients received a well protocolised treatment plan, targeting specific physiological variables. All patients received lidocaine.
Inclusion criteria: Initial rhythm VF, successful return of spontaneous circulation, persisting coma after ROSC, brought to participating ED.
Exclusion criteria: Men under age of 18 years, women under age of 50 years (due to possibility of pregnancy), cardiogenic shock, alternative causes of coma, no ICU bed available.
Primary outcome: Neurological outcome good enough to allow discharge to home or rehabilitation facility was determined as a ‘good outcome’. Death or discharge to long term nursing care was determined as a ‘bad outcome’
Secondary outcomes: Looking at the haemodynamic, haematological and biochemical effects of the hypothermia.
What did they find?
Numbers: 84 eligible patients over 33 month period. 7 patients excluded from analysis. Resulted in 43 patients in hypothermia group, 34 in normothermia group. 5 of these patients not adherent to protocol.
Results: ‘Good outcome’ better in hypothermia group (49% vs 26% of patients (p=0.046)).
Mortality rates lower in the hypothermia group but not statistically significant (51% vs 68% (p=0.145)).
Similar baseline characteristics though normothermia group more likely to receive bystander CPR (71% vs 49% (p=0.05)) and be male (79% vs 58% (p=0.05)).
Patient’s age and the time from collapse to ROSC both significantly affected the outcome, though not different between the 2 groups.
Systemic vascular resistance calculated as significantly higher in the hypothermia group during the first 12 hours. Cardiac index calculated as significantly lower at the time of admission to ICU (2.0 vs 2.6 l/min/m2 of BSA (p=0.01)). Blood glucose levels higher in the hypothermia group at the time of admission to ICU (median 16.2 vs 10.5 mmol/L (p=0.002)) and 6 hours post admission (16.0 vs 12.1 (p=0.02)).
Results: ‘Good outcome’ better in hypothermia group (49% vs 26% of patients (p=0.046)).
Mortality rates lower in the hypothermia group but not statistically significant (51% vs 68% (p=0.145)).
Similar baseline characteristics though normothermia group more likely to receive bystander CPR (71% vs 49% (p=0.05)) and be male (79% vs 58% (p=0.05)).
Patient’s age and the time from collapse to ROSC both significantly affected the outcome, though not different between the 2 groups.
Systemic vascular resistance calculated as significantly higher in the hypothermia group during the first 12 hours. Cardiac index calculated as significantly lower at the time of admission to ICU (2.0 vs 2.6 l/min/m2 of BSA (p=0.01)). Blood glucose levels higher in the hypothermia group at the time of admission to ICU (median 16.2 vs 10.5 mmol/L (p=0.002)) and 6 hours post admission (16.0 vs 12.1 (p=0.02)).
Is it any good?
Overall: Fairly strong
Strengths: Well protocolised care minimising confounding factors. Intervention started pre-hospital. Assessor of outcome blinded to intervention.
Weaknesses: Control arm mild pyrexia rather than normothermia. Some difference in the baseline characteristics of groups. Inclusion criteria resulted in a specific cohort of arrest patients so unclear how it can be extrapolated to other groups e.g. non-shockable rhythms. Only a relatively small number of patients, making it difficult to comment on the incidence of complications from hypothermia.
Strengths: Well protocolised care minimising confounding factors. Intervention started pre-hospital. Assessor of outcome blinded to intervention.
Weaknesses: Control arm mild pyrexia rather than normothermia. Some difference in the baseline characteristics of groups. Inclusion criteria resulted in a specific cohort of arrest patients so unclear how it can be extrapolated to other groups e.g. non-shockable rhythms. Only a relatively small number of patients, making it difficult to comment on the incidence of complications from hypothermia.
Final thoughts
A landmark paper with quite a significant result, strongly suggesting we should cool patients after OOHCA. Some important facets to this include the fact that this is a fairly narrow subset of OOHCA patients they are looking at, and it is not quite comparing therapeutic hypothermia with normothermia as it suggests, but actually slightly pyrexial patients.
Written: Tom Heaton
Reviewed: Not done
8th April 2016
Written: Tom Heaton
Reviewed: Not done
8th April 2016
References
- Bernard et al. Treatment of Comatose Survivors of Out-of-hospital Cardiac Arrest with Induced Hypothermia. NEJM. 2002. 346: 557-563
