Thursday, June 20, 2013

Logic Based Medicine: The Case of Activated Charcoal

Like all good things, Evidence Based Medicine (EBM), when taken to far, runs the risk of making us overwrought and becoming cliche.  I think we are reaching this point.  Given Ioannidis' meta-research findings that most published research findings are false (does he consider the irony that that may apply to his findings too?) the corrupting and corrosive influence of industry on research programs and guideline construction, the biases of academic researchers intent on grants, prestige and promotions (as well as honoraria to supplement paltry academic salaries - I was there once, and I did it too), and the zealousness of "experts" who wish to interpret the evidence in the form of an edict for all to follow (euphemistically called "guidelines"), and several other disturbing trends, it becomes apparent that in the end we must rely upon our own judgment and logic to discern the proper path to follow.  And so it is with Activated Charcoal (AC) administration, an agent used in overdoses and toxic ingestions that has a remarkable capacity to adsorb ingested substances and theoretically limit their toxicity.  (It is not barbecue charcoal, the photo is tongue-in-cheek.)

Because there are no clinical trials to offer guidance, only studies that use volunteers to demonstrate reduction in gastrointestinal absorption of substances when AC is administered at different times after toxic ingestion, recommendations for the use of AC vary.  In Utah, the poison control center advises dogmatically that this agent is useful only when given in the first hour after ingestion, because of these volunteer studies.  There are several problems with this approach.

First, it presumes that we can properly time the ingestion.  Patients that take overdoses are notoriously poor historians, and often no information about the timing (or dose or even substance) of toxic ingestion is available.  When timing information is available, it is often based only on guesswork or is otherwise suspect.

Second, it presumes that there is external validity to the volunteer studies.  The volunteers in those studies almost assuredly had normal gastrointestinal motility (GIM), that is, they digest and absorb the overdose quickly, within one hour.  For several reasons, overdose patients most often do not have normal GIM:
  • Coingested medications and chronic polypharmacy influence and also interact to alter GIM
  • Many agents taken chronically or in overdose such as narcotics and any agent with anticholinergic properties slow GIM
  • Underlying disease states such as diabetes mellitus affect GIM
  • Gut hypoperfusion resulting from shock and other hemodynamic perturbations associated with the overdose alter GIM
  • The quantity of the overdose taken in "real life" markedly exceeds the quantity utilized in the volunteer studies (which were necessarily limited by IRB safety concerns)
So it would seem that sometimes people get a bit too excited about their knowledge of some purported "fact" - too excited to recognize the shortcomings and limited applicability of that "fact."  "There's no evidence for XYZ," we often hear said.  Not to mention that absence of evidence is not evidence of absence, this statement attempts to abdicate us of our responsibility to judge whether the evidence we do have is worthy of guiding our actions, and of our responsibility to act in some manner regardless of the quality and quantity of evidence that we have.

The only reason to not give AC to all but the most remote overdoses is because we worry about side effects of the medication.  The most worrisome of these would be aspiration of the AC which can cause severe pulmonary complications.  But the risk of aspiration can be assessed like any other risk in medicine.  Rather than have a mindless universal rule that says "time the ingestion and if it was less than an hour ago, give AC", it would seem to me that we should say "estimate the timing of the overdose and its seriousness, and estimate the risk of aspiration.  If the risks of AC administration are exceeded by its benefits, give AC."  This would make the decision to give AC akin to many other complicated and nuanced decisions we make every day in medicine.  So, if a patient has a potentially lethal overdose of amitryptyline (Elavil) [a drug that slows gastric motility] that was taken somewhere between 1-6 hours ago, and the risk of aspiration is judged to be acceptable, AC should be given.  Alternatively, if the patient took an overdose of alprazolam (Xanax) between 6-12 hours ago, AC would not be given.  Alprazolam does not alter GIM, it impairs airway protective reflexes, and patients usually recover from it without sequelae if hypoxemia is avoided.  The risks of charcoal appear to exceed the benefits in that case.

Based on my own personal experience, I have a strong clinical intuition that patients who receive AC recover more rapidly and with fewer sequelae.  Do I have evidence for this?  No, and I don't need to.  I'm using the time-honored, but largely abandoned tradition of logic.


  1. Unfortunately, the world is full of doctors reporting excellent results from many treatments that turn out to be no better than placebo (if they are ever put to the test). If they are not put to the test, it is assumed that they work, and this drives much of the medicalisation, overtreatment and overdiagnosis in medicine today

  2. I recently admitted a young man after he took 48 Benadryl. When he presented with normal mental status, Poison Control (PC) told he ER not to give AC, he was outside the window, just observe him. Six hours later when he was [predictably] floridly anticholinergic, they called PC back, and PC said something like "Oh, yeah, because of the anticholinergic effects of Benadryl, there can be delayed absorption and a late toxidrome. Give him Ativan and physostigmine." So let me get this straight, we recognize that Benadryl slows gut motility, but we don't give AC early in the course, before there are symptoms. We wait until the predictable late absorption happens, then we give other treatments and antidotes after the kid is tied to the bed and freaking out, and he stays 48 hours in my ICU while he recovers. [By the way, what evidence is there for physostigmine that is paramount to the evidence for AC?] To quote the late James Trafficant, "BEAM ME UP!"