Saturday, December 10, 2011

Why the Rumack-Matthew nomogram is a dangerous distraction

When a known or suspected overdose (OD) comes to the emergency department (ED), levels of several commonly ingested agents such as aspirin and tylenol are checked.  If the history or laboratory results suggest acetaminophen (APAP) overdose, the "standard of care" (if there is one - see:  Heard APAP OD review in NEJM, 2008) is to investigate the time of the ingestion and refer to the Rumack-Matthew nomogram (RMN) to determine whether toxicity is likely and to initiate a course of n-acetylcysteine (NAC) if the level is above the line on the nomogram.

Now there are several problems with the RMN that I will not fully delve into here. Suffice it to say that the RMN came from a review of a (striking!) 30 cases in a 1975 paper in Pediatrics.  The nomogram was extrapolated from this case review.  And while I'm no pediatrician, I suspect that pediatric poisionings are quite different from adult ones.  In any case, when the timing of ingestion is accurate the RMN has been a useful guide for ED physicians to determine who could be safely sent home and who would need to be admitted for treatment for APAP poisioning.

But Adults are different animals altogether.  Suppose that Candice Brown comes in with an intentional overdose after a psychologically traumatic break-up with her boyfriend.  She may have come to medical attention of her own volition or at the insistence of a suspicious relative.  Are we to trust her recollection or estimation of the time of ingestion?  What if she's prevaricating?  What if she was intoxicated and is mistaken?  Look at the nomogram.  A four hour error in the estimation of the time of overdose can be the difference between a decision to treat or not to treat. 

Then there are the polysubstance ODs who present with altered mental status (read "coma") from whom no reliable history is obtainable but who have a detectable APAP level and/or elevated LFTs (the latter often from shock liver).  In my neck of the woods, these are highly prevalent scenarios.  And here's the rub.  These patients are going to be admitted anyways.  So why the heck would we refer to the RMN to see if we should give them NAC?  You're going to give them oxygen, DVT prophylaxis, charcoal, perhaps antibiotics for "possible" aspiration, narcan, etc.  Why not  give NAC too in recognition of the imprecisions inherent in timing overdoses?  Why risk missing an opportunity to treat a serious malady with a benign therapy? Because of blind faith in a fragile nomogram, that's why.

I know, I know.  Oral NAC is miserable to take and induces vomiting.  Intravenous NAC is "expensive" ($400, about as much as a dose of linezolid or a transfusion of PRBCs) and sometimes causes anaphylaxis.  But then why, when we admit a gastrointestinal hemorrhage, don't we fret about $400 here and there as well as TRALI and all the other complications of transfusion amid the uncertainty of a safe hemoglobin level in the face of possible ongoing hemorrhage?  But if it's NAC, we're all up in a tizzy about anaphylaxis and cost.

Academic purists like to make the world more complicated than it is or needs to be, while simultaneously ignoring the imprecisions and uncertainties inherent in it.  And they like to inculcate in their students a sense of awe and reverence for tools such as the RMN, rather than to foster the use of common sense and engender healthy skepticism about dogma and tradition.  The RMN, for all its utility, should not be used as a substitute for good judgment and rational decision making.  If we're going to admit or observe somebody for an overdose, and they have a detectable tylenol level or elevated transaminases, we should give them the NAC until we discharge them with normal labs.  It's the uncommon sense thing to do.

No comments:

Post a Comment