Hagihara: Prehospital epinephrine use and survival among patients without-of-hospital cardiac arrest

Hagihara. JAMA 2012;(307)11:1165-1168

Clinical Question

• In patients with out-of-hospital (OOH) cardiac arrest is the use of adrenaline, compared to not using adrenaline, associated with an improved mortality?

Design

• Prospective
• Non-randomised
• Cohort study with concurrent control group

Setting

• Pre-hospital services, Japan
• 01.01.2005-31.12.2008

Population

• Inclusion: adult patients who had an OOH cardiac arrest prior to arrival of emergency services
• Exclusion: >60 minutes elapsed from call to scene arrival or >480 minutes from call to hospital arrival; missing data on adrenaline administration
• 417 188 patients

Intervention

• Patients treated with epinephrine
  • From April 2006 certified paramedics were permitted to use adrenaline which led to the development of the intervention and the control group

Control

• Patients not treated with adrenaline

Outcome

• Comparing patients who received adrenaline with patients who did not receive adrenaline:
  •  the Odds Ratios (95% C.I.) were the following for:
    • Return of spontaneous circulation (ROSC) before hospital arrival
      • unadjusted OR 3.75 (3.59-3.91)
      • adjusted OR 2.36 (2.22-2.5)
    • Survival at 1 month
      • unadjusted OR 1.15 (1.07-1.23)
      • adjusted OR 0.46 (0.42-0.51)
    • Good neurological outcome (CPC 1 or 2)
      • unadjusted OR 0.61 (0.53-0.7)
      • adjusted OR 0.31(0.26-0.36)
  • Similar results were found in a propensity matched subgroup (n=26,802)

Authors’ Conclusions

• Use of prehospital adrenaline was significantly associated with increased ROSC but decreased chance of survival and good functional outcome

Strengths

• Huge numbers of patients
• Similar results found when analysed with different models

Weaknesses

• Non-randomised
• Baseline characteristics generally favour adrenaline group with greater rates of bystander CPR (45% vs. 36%), use of public AED (0.8% vs. 0.4%), VF/VT (13.7% vs. 7.2%)
• Data on in-hospital CPR not included in analysis
• Number of adrenaline doses given not recorded
• Included pre-hospital cardiac arrests only
• One month survival influenced by in-hospital care which varies between centres