Adrenaline has been an integral component of advanced life support from the birth of modern cardiopulmonary resuscitation (CPR) in the early 1960s. In guidelines written originally in 1961, Peter Safar recommended the use of very large doses of adrenaline: 10 mg intravenously or 0.5 mg intracardiac (Safar, 1964). Data from animal studies show that when injected during cardiac arrest, adrenaline increases aortic relaxation (diastolic) pressure, thereby increasing coronary and cerebral blood flow (Michael et al, 1984; Brown et al, 1987). The inclusion of adrenaline in resuscitation guidelines has always been based mainly on animal data, not from evidence derived from humans.
Recent prospective randomised trials (Olasveengen et al, 2009; Jacobs et al, 2011) and observational studies (Hagihara et al, 2012) have challenged the value of using adrenaline in cardiac arrest and this view is supported by the findings of recent systematic reviews of adrenaline in out-of-hospital cardiac arrest (Patanwala et al, 2013; Lin et al, 2014). These studies indicate that rates of return of spontaneous circulation (ROSC) are increased with adrenaline but long-term survival (survival to discharge neurologically ‘good’) is either no better or worse. In other words, more hearts are restarted with adrenaline but at the cost of more brain injury.
In a study undertaken in Oslo, Norway, 851 patients with out-of-hospital nontraumatic cardiac arrest (all rhythms) were randomised to intravenous (IV) cannulation with injection of drugs (including adrenaline) versus no IV cannula or drugs until after ROSC had been achieved (Olasveengen et al., 2009). The patients in the IV group had a higher rate of ROSC but this was seen only in the patients with initial non-shockable rhythms (asystole and PEA); in those patients with an initial rhythm of VF/VT the ROSC rate was no different. There was no significant difference between the IV and no IV groups in survival to hospital discharge, survival with favourable neurological outcome and the survival at one year. This study included the use of all drugs versus no drugs—it was not simply an adrenaline versus no adrenaline trial; furthermore, the analysis was undertaken according to the group to which patients were allocated (‘intention to treat’ analysis, which is the appropriate way of analysing the results of a randomised controlled trial such as this), not on the basis of whether or not they received drugs. In a follow up analysis of this Norwegian study, outcomes were examined according to whether the patient had actually received adrenaline (Olasveengen et al, 2012). Treatment with adrenaline was associated with a greater chance of being admitted to hospital but a reduced chance of survival to hospital discharge (reduced by half in those given adrenaline). These effects persisted after adjustment for factors that might bias the results, i.e. confounders such as rhythm, response interval, witnessed arrest, gender, age and tracheal intubation.
To date, the only double-blind randomised placebo-controlled trial of adrenaline in out-of-hospital cardiac arrest was undertaken in Western Australia (Jacobs et al, 2011). The investigators had planned originally to enrol 5 000 patients; unfortunately, several ambulance services were unable to participate and ultimately only 534 patients were included in the final analysis. The rate of ROSC was three times higher in those receiving adrenaline but survival to hospital discharge was no different between the groups.
The Ontario Prehospital Advanced Life Support (OPALS) Study used a before-and-after design to evaluate the impact of adding tracheal intubation and drug administration to an optimised rapid defibrillation programme (Stiell et al, 2004). The rate of admission to hospital increased significantly in the ALS phase but the rate of survival to hospital discharge was unchanged. During the ALS phase, 95.8% of patients received adrenaline. Given that other drugs and tracheal intubation were also included in the ALS phase, it is difficult to make firm conclusions about the impact of adrenaline but the results are notably similar to those documented in later prospective controlled studies.
Out-of-hospital cardiac arrest registries include data on large numbers of cardiac arrests and this can be analysed in an attempt to determine the impact of therapeutic interventions. Such observational studies enable large quantities of data to be collected but they rely on statistical risk-adjustment to remove inherent biases. The largest observational study to date on the use of adrenaline in cardiac arrest involved 417 188 OHCAs in Japan (Hagihara et al, 2012). The authors concluded that use of adrenaline was associated with a risk-adjusted ROSC rate 2.5 times higher but a one-month survival rate approximately half of that achieved in those not given adrenaline. In contrast, a more recent analysis of the same database (but confined to a more recent period) concluded that long-term survival was similar among the adrenaline and no-adrenaline groups (Nakahara et al, 2013). In another observational study from Japan, OHCA patients receiving adrenaline had a significantly lower rate of neurologically intact one-month survival than those not receiving adrenaline (Hayashi et al, 2012).
The findings from randomised trials and observational studies indicate that giving adrenaline in OHCA increases the rate of ROSC but that longer-term outcomes (survival to hospital discharge and neurologically favourable survival) are either worse or, at best, neutral. The observational studies showing worse long-term outcome after adrenaline may be misleading if there are confounders that have not be fully adjusted for in the statistical analyses. For example, most of these studies do not adjust for the duration of the resuscitation attempt, yet patients who respond rapidly to initial treatment (e.g. defibrillation) will have the best outcomes. In general, only those with longer resuscitation attempts might be expected to receive adrenaline.
It is entirely possible that adrenaline is genuinely harmful when given in cardiac arrest. Evidence from an animal study indicates that although adrenaline improves myocardial blood flow during ventricular fibrillation (VF) cardiac arrest, it does not improve myocardial oxygen balance; in other words, an increase in the amount of oxygen consumed by the heart exceeds any increase in the amount delivered (Ditchey and Lindenfeld, 1988). Of those patients who reach hospital alive after OHCA, but who subsequently die before hospital discharge, the majority die from neurological injury (Lemiale et al, 2013). Although adrenaline may improve blood pressure and global blood flow to the heart and brain during CPR, animal evidence suggests that adrenaline may reduce blood flowing in the capillaries (the microcirculation) of the brain (Ristagno et al, 2009); thus oxygen delivery to cells may be reduced by adrenaline and neurological injury therefore made worse.
Adrenaline has adverse effects on the myocardium: in an analysis of the Norwegian IV versus no IV trial adrenaline increased the frequency of transitions from PEA to ROSC and extended the time window for ROSC to develop; however, this was at a cost of greater cardiovascular instability after ROSC, with a higher rate of re-arresting (Nordseth et al, 2012). Adrenaline is associated with the development of lactic acidosis and high lactate concentrations and slow lactate clearance after ROSC are associated with poor outcome.
The accumulating evidence highlights the urgent need for further appropriately powered high-quality randomised controlled trials (Perkins et al, 2014). This need will be met by the PARAMEDIC-2: The Adrenaline Trial, which will involve five ambulance service NHS Trusts (London Ambulance Service, North East Ambulance Service, South Central Ambulance Service, Welsh Ambulance Services, and West Midlands Ambulance Service) and will enrol 8 000 OHCA patients. Approval for this trial has been granted by South Central Research Ethics Committee. These patients will receive randomly either adrenaline or placebo (water for injection); paramedics giving the study drug and those undertaking follow-up assessment will be unaware of the group to which the patient has been allocated (i.e. they will be ‘blinded’). Study drug (adrenaline or placebo) will be supplied in identical packs of 10 syringes. If patients survive to be admitted to a hospital ward, the research paramedics employed for the trial will make contact with the patients (or their personal or professional representatives) and seek informed consent to participate in the follow-up of the study, which involves completing quality of life questionnaires. The outcomes that will be measured include survival to hospital admission, discharge, 30 days, 3 months, 6 months, and 12 months; and neurological status and quality of life.
Members of the public who are concerned about being enrolled in the trial will have the opportunity to opt out. The mechanism for this is being finalised but is likely to involving them being given a wrist band that will be easily be identifiable by attending paramedics. This opt out mechanism has already been used successfully by the Resuscitation Outcomes Consortium in North America.
The Adrenaline Trial should answer once and for all the question of whether giving doses of adrenaline 1 mg during CPR improves long-term outcome from OHCA. As such, it is the most important cardiac arrest research study ever to be undertaken in the United Kingdom. We anticipate that patient enrolment into the study will start in October 2014.