Wednesday, July 30, 2008

Determining Cause and Effect

I wouldn't want the FDA's job.

Today, the Chicago Tribune reported that of the 93 deaths from "tainted" heparin, that analphylaxis (a sudden allergic reaction with associated drop in blood pressure) was identified in 10 of the 93. Of these 10, 3 patients were conclusively found to have the "oversulfated chondroitin sulfate" contaminant in certain vials:
Of 10 reports of death from severe shock known as anaphylaxis or hypotension, three of those could be traced to lot numbers of Baxter products that tested positive for an animallike substance known as oversulfated chondroitin sulfate. Heparin lot numbers were not known for the other seven deaths from anaphylaxis or hypotension, so the FDA could not determine whether those 7 patients received heparin contaminated with the substance.
But what I found interesting were the number of deaths NOT found related to the contaminant. Of the remaining 83 deaths reported:
  • 45 didn't have enough data submitted to be able to determine a cause of death
  • 25 were causes unrelated to heparin, like kidney failure, pneumonia, and sepsis
  • 13 were related to bleeding, a known side effect of heparin

The FDA should be commended on their efforts to cull through a vast sea of medical information to attempt to determine the cause of death in the individuals affected and working so quickly to determine a cause of the contaminant. To me, finding three deaths in a sea of 90 deaths that were definitively determined to be related to the contaminant is like finding a needle in a haystack. Perhaps more importantly, by issuing their summative findings of the evalutation of the deaths, they shed light on the difficulty of monitoring and investigating drug reactions from community-based reporting.



Anonymous said...

I wouldn't want the FDA's job, either, but do you think they should do away with community-based reporting? If a patient reports a symptom to their physician and it is not a known side effect, what happens? My experience is that the physician will reply, "That is not supposed to happen", but then in a support group with others on the same medication, I will find out it happened to them as well.

DrWes said...


No, I certainly don't think removing community-based reporting should occur. My problem with such databases (like the FDA's MAUDE database) is there is never "closing of the loop." That is, there is no public notification of the results of investigations that occur based on those publically-submitted complaints so we can never determine if those complaints are legitimate.