Is “Epi” Killing Your Patient? The Evidence Behind Epinephrine in Cardiac Arrest

The theory behind giving epinephrine in a cardiac arrest is rooted in the idea that adrenergic stimulation of peripheral vasculature will increase diastolic pressure via vasoconstriction, thereby increasing coronary artery perfusion. It’s thought by some authors that this effect also increases cerebral perfusion, although this is controversial – others believe that the vasoconstrictive effects will actually cause microvascular ischemia.^[2]

The original data behind epinephrine’s efficacy is based on studies from the 1960s. Investigators asphyxiated dogs and then provided epinephrine, which seemed to improve survival.² The ACLS recommendation for epinephrine in cardiac arrest is at least partially based on this study and predates the FDA requirement that a drug must have demonstrated efficacy in humans before its use. Unfortunately, only one randomized, controlled trial (RCT) has been performed, which suggested that epinephrine may be of limited utility (to be discussed later in this post). No other RCTs have been performed because of now obvious ethical quandaries surrounding the withholding of long-standing, traditionally accepted treatment. Along those lines, the recommendation for epinephrine has been carried forward to current practice.^2,[3]

This is problematic, because despite early study and theory, there continues to be debate as to whether the use of epinephrine is favorable for in the long run – its side effects include increased myocardial work, metabolic demand, risk of worsening a tachyarrhythmia, and a reduction of microvascular brain perfusion. In the pulseless patient, it’s not difficult to imagine how these effects might worsen long-term outcome. In fact, data suggests that while epinephrine increases the likelihood that a patient achieves return of spontaneous circulation (ROSC) in the short term, it may actually worsen the neurological status and long-term survival prospects of the patient.^2,3

Fortunately, there has been some recent, critical appraisal of epinephrine for cardiac arrest. The idea that epinephrine just “kicks the can down the road” by increasing ROSC but may not improve (or may worsen) survival and/or neurological outcomes has been suggested by one RCT conducted to date. Jacobs et al enrolled 534 patients and found an increase in ROSC for those receiving epinephrine compared to placebo but no significant difference in survival to hospital discharge.^[4]

Meta-analyses have also been performed evaluating short- and long-term outcomes in cardiac arrest patients receiving epinephrine. The most impressive one spanned 14 studies and included 650,000 cases. The patients that received epinephrine had an odds ratio (OR) of 2.86 for ROSC, but no difference in neurological outcome or long-term survival.^[5]

A wealth of the data on epinephrine comes from a Japanese registry of pre-hospital cardiac arrests. Investigators used this data to statistically control for different variables and interventions and then extrapolated their effects on patient-centered outcomes. These are cohort studies, but they are extremely large (in the range of half a million patients or more). The result, depending on the particular manuscripts, was an increase in ROSC for patients that received epinephrine (OR 2.3) but a decrease in good neurological outcome (OR 0.31-0.79). Long-term survival was also decreased in patients that had received epinephrine (OR 0.46).^{[6],[7],[8],[9]} After propensity matching, this translates to a number needed to treat (NNT) of 13 for ROSC, but also a number needed to harm (NNH) of 52 for decreased 1-month survival and a NNH of 55 for worsened long-term neurological outcome.⁸ Of note, patients with a shockable rhythm were noted to have decreased ROSC, poorer neurological outcomes, and decreased long-term survival, suggesting that the risk of harm may even be higher depending on initial rhythm.^[10]

When you add this all up, it’s a pretty bleak picture. While epinephrine may help get the patient upstairs to an ICU, it may likely reduce the chance that the patient has any chance of meaningful survival.

Since it has been the standard of care for decades, many physicians will still push epinephrine despite the data. Let’s now take a look to see if there is an optimal timing and interval for administering epinephrine during resuscitation.

In order to discuss this topic, it’s important to first review the three phases of cardiac arrest. The first is the electrical phase – which lasts seconds to minutes – in which the patient is often in ventricular fibrillation. The effective treatment here is defibrillation. Next is the circulatory phase, which occurs within the first ten minutes of arrest and when coronary perfusion begins to decrease. This is when epinephrine is theorized to be most beneficial. The last and final phase is metabolic – this occurs as the oxygenation demands of the body begin to far outpace the oxygen delivery. It’s not a difficult mental leap to surmise how epinephrine, which increases metabolic demands, may be harmful in this phase.²

It’s theorized that epinephrine may be of more value earlier in the course of cardiac arrest. Data from large retrospective reviews are in agreement with this theory; those that receive epinephrine earlier in a code are more likely to have ROSC and improved neurological outcome compared to those that receive it later.^{9,[11],[12],[13],[14]} The aforementioned Japanese registry found that if epinephrine is given, there is a stepwise decrease in OR for good neurological outcome of 0.917 for every minute it is delayed.^[15] There are obvious confounding factors inherent to the trial’s retrospective design – such as the fact that time to chest compressions is likely to coincide with time to first epinephrine dose – so this data should be interpreted carefully. However, it does seem to make sense on a physiological basis.

With regards to interval, ACLS recommends giving epinephrine every 3 – 5 minutes in a cardiac arrest. However, recent studies have suggested that this may not be the optimal dosing interval. One prospective, observational study of 1,630 pediatric patients found a stepwise increase in both long-term survival and survival to discharge when epinephrine was given at intervals of 1 – 5 minutes, 5 – 8 minutes, or 8 – 10 minutes.^[16] A similar study on 20,909 adults had similar findings.^[17]

The interval is especially important in patients with a shockable rhythm. Investigators found that in patients who were defibrillated, those that received epinephrine within two minutes of being “zapped” had decreased ROSC, survival, and neurological outcome.^[18] This is in line with ACLS guidelines, which recommend epinephrine for shockable rhythms only after two shocks and two minutes of CPR have been performed.¹

Taken together, these data suggests that if epinephrine is to be given, it is more likely to be effective if given early for non-shockable rhythms, and at longer intervals than currently recommended by the guidelines. If defibrillation is indicated, hold off on epinephrine until after appropriate defibrillation and at least two minutes of CPR (one cycle) and then give it only if there continues to be evidence of no perfusion.

One of the justifications for using end-tidal CO₂ during resuscitation is the potential for quicker detection of ROSC during CPR – ROSC may correlate with a sudden increase in the parameter. This specific method of monitoring may allow for decreased epinephrine doses via earlier detection of ROSC^[19] – if you can detect ROSC during CPR, then epinephrine may be withheld. Of course, this requires ET intubation and a pause chest compressions, which is widely accepted to be harmful. Along the same lines, it’s reasonable to initiate arterial hemodynamic monitoring during CPR for use in administering (or witholding) epinephrine based upon diastolic blood pressure; however, there’s no data with regards to this practice.

It’s pretty discouraging that one of the primary medications that we have for cardiac arrest may be hurting our patients in the long run. The good news is that we still do have some other effective interventions under the right circumstances. Defibrillation is quite effective when indicated. Instead of vasoactive agents, perhaps we should be concentrating in high-quality chest compressions. And finally, in the event that you do get ROSC, targeted temperature management and coronary angiography should be considered in the appropriate patient.

If you’re not convinced by the data above, the PARAMEDIC-2 trial is an ongoing pre-hospital RCT of epinephrine in cardiac arrest in the UK.^[20]