Monday, February 25, 2013

JAMA & Procalcitonin

Someday, I'll publish another article summary that doesn't involve a conflict-of-interest skewering.  I'm really not as angry as Rob Orman says I am.  This article, at least, is directly relevant to the Emergency Department.

There's been significant research into biomarkers for infectious/inflammatory processes, with the goal of identifying a sufficiently sensitive assay to use as a "rule-out" for serious infection.  The goal is to use such an assay to prevent the overuse of antibiotics without increasing morbidity/mortality.  This is a good thing.

Procalcitonin is the latest darling of pediatrics and intensive care units.  However, to call the literature "inconclusive" is a bit of an understatement – which is why I was surprised to see an article in JAMA squarely endorsing procalcitonin-guided antibiotic-initiation strategies.  After all, I've previously covered negative trials in this blog (pubmedpubmed).  However, these authors seem to have intentionally narrowed their trial selection to exclude these trials – and publish no methods regarding their systematic selection of articles.

The disclosures for all three authors includes "BRAHMS/Thermofisher".  Who is this, you might ask?  Google points me to: – where BRAHMS/Thermofisher will sell you one of seven procalcitonin assays.  JAMA, third-ranked medicine journal in Impact Factor, reduced to advertising masquerading as peer-reviewed science.

"Clinical Outcomes Associated With Procalcitonin Algorithms to Guide Antibiotic Therapy in Respiratory Tract Infections"


  1. Misleading science is probably worse than bad science.

    1. Probably? No, it's definitely worse.

  2. Glad to see someone shine a light on this.

  3. Good point Ryan, as usual!
    After I read carefully those papers, seems that Jensen was also paid by that Brahms company for the negative paper too! is it the same?

  4. It ought to be the same company. This is not a topic I'm innately familiar with, other than to be aware that it's hardly settled science - which makes the JAMA article at least befuddling, at the minimum.

  5. Yeah, I think that anyway the best way to judge a biomarker value is to try it on patients and then make a local retrospective evaluation. Opinions in papers, good or bad are just that, opinions! I know it's not used as much in USA as in Europe, any of you tried it?


Comments on posts > 10 days old will be moderated; blame spammers.