If I were a literary critic, I'd be a deconstructionist. Inspired by Don Quixote, I like to take shots at established ideas, all the while espousing the purity of my motives. So when advanced cardiac life support (ACLS) experts trumpeted amiodarone as a revolutionary drug, I went out of my way to find things wrong with it.
It was actually pretty easy.
Limitations & patient outcome
The first major paper on amiodarone in prehospital cardiac arrest (the ARREST trial) noted an improved survival rate to hospital admission of roughly one-third in patients given prehospital amiodarone instead of a placebo (an inactive treatment) in treating ventricular fibrillation. The study noted that survival to ED discharge improved from 34% in patients given no active drug to 43% in patients given amiodarone. What this really means is that roughly one additional patient in 10 actually survives the ED stay; there were no significant differences in long-term survival. Amiodarone might make you better for a few hours, but nobody left the hospital tap dancing.
There are other problems with the work. The demographics of the study don't likely reflect the entire population of prehospital cardiac arrests. For example, between 59% and 68% of arrest victims received bystander CPR. Even the improved survival within the ED occurred in only a small fraction of the total patients presenting with cardiac arrest during the study period. Of the nearly 4,000 patients suffering from prehospital cardiac arrest during the study, only 504 (13%) qualified for the study with continued ventricular fibrillation or pulseless ventricular tachycardia despite three countershocks, endotracheal intubation and initiation of an IV line with a first dose of epinephrine. From the practical side, amiodarone costs a lot and is difficult both to mix and to administer.
And then there were the incessant, mind-numbing ads proclaiming that ACLS recommended the administration of amiodarone over lidocaine. In fact, this wasn't true, as even the American Heart Association (AHA) noted. Amiodarone was given a class IIa recommendation because it had been formally studied. Lidocaine was considered "indeterminate" simply because it has not been evaluated in a similar fashion (Unlike Kiefer Sutherland in "Flatliners," not many people volunteer to be put into v-fib). Nobody had ever actually compared the two drugs in a head-to-head contest. They were both in the same box within the v-fib algorithm with an "OR" right between the two.
Some felt that amiodarone was the centerpiece of a cumbersome and profiteering new approach by the AHA to the process of ACLS guideline development and education. One of my paramedics said it well: "The best recommendations money can buy." We already had lidocaine on board, which was cheap and could do double duty as both an antiarrhythmic and a preintubation agent. So no one shed any tears when we opted to keep faith with lidocaine and forgo amiodarone for now.
I'm still not an amiodarone fan, but some recent articles and discussions have made me take a second look. Results of the ALIVE trial, which directly compared amiodarone with lidocaine in the management of prehospital cardiac arrest, indicate that survival rates to hospital admission reach 22.8% in patients given amiodarone and 12% in those given lidocaine. Roughly 25% of prehospital cardiac arrest victims qualified for administration of one of the two drugs, and a little work with the math indicates the rate of survival to ED discharge rate goes from roughly one in 10 to two in 10. Unfortunately, the question of long-term outcome remains.
I've always tried to look at the big picture. Maybe this is a result of a career in emergency medicine where we look at big, life-threatening things like airways before worrying about the strange, alien pulsations in the left fourth toe. It's just a matter of perspective. In medicine, the proper perspective is always the patient and always the long term. Does the patient get better from what you do? In what way do they improve and for how long? But I'm starting to think that a detailed focus on the short-term segments of care might yield more overall long-term benefit than focusing on overall patient outcomes.
Interestingly, I held this same view 10 years ago. I recently reviewed some of my first papers on prehospital care. I had compared two different air medical crew configurations and tried to see if one was better by comparing the patient's Revised Trauma Scores before and after transport. One of the criticisms of the work was that it didn't look at the big picture; what's the point of any difference if it's not linked to eventual patient outcome? Aren't we missing the forest for the trees? It made sense, so I shifted my views; and the new paradigm was reinforced throughout my future training in both clinical research and public health.
What we control
The issue of control is the key. If I want to see how well amiodarone works in my EMS system, I must look at what elements of the patient's care I can control. In reality, not very many. I can't dictate the time and place of the cardiac event or determine the health status of the patient; I can't define driving times or traffic conditions; I can't dictate ED or other post-resuscitative care. The only thing I can control is that brief prehospital encounter where the patient receives the drug. That is not outcome. Outcome is the sum of all these parts, of which only one is significantly affected by my actions.
I've seen this in my own work with levalbuterol, a new nebulized agent for respiratory distress. While using this drug, paramedics were able to improve the patient's respiratory status more than by using albuterol. But hospital admission rates stayed the same, no matter which agent was given in the field. Why? Because the emergency physicians and the hospitals were running off a different set of rules and conditions and were using treatments other than levalbuterol for ED care. I was trying to prove that levalbuterol would reduce admissions when I had no control over the admission process. It simply couldn't be done.
What I can control is the prehospital medical care provided to the patient. So it makes sense for me to occasionally stick to the trees and ignore the forest. My efforts should focus on developing a protocol to maximize the potential outcomes of this phase of care. For me to evaluate my efforts in total isolation of overall outcome is wrong, but to denigrate prehospital contributions to care in light of variables beyond our control is similarly flawed reasoning. So if amiodarone can enhance prehospital survival, the next step is for others to investigate additional parameters that influence overall survival, including EMS contact systems, citizen CPR education and post-resuscitation care.
Optimizing every step of the process should result in overall changes in survival. Darwinian evolutionary theory contends that multiple small changes result in large adaptations and enhanced survival of biologic species. I think he's got it right.
At least this intellectual adventure has forced me to rethink some cherished concepts, and that's what science is all about.