Tuesday, January 04, 2005

Small tests for big drugs?

From article "Merck Mess Shows FDA's Flaws" in TheStreet.com (a whole series is being published this week, see sidebar on page 1 of the article):

Prior to FDA approval, Psaty [an epidemiologist and drug-safety specialist who testified at the recent Senate hearings on drug approval] explained, Vioxx had been tested -- at common doses of 12.5 and 25 milligrams -- on just 371 patients for at least one year. The only large study to explore the drug's cardiac risks had yet to be completed.

As a medical informatics specialist used to designing EMR systems meant to capture clinical information on thousands or tens of thousands of patients (or more) routinely, the Clinical Trials and FDA approval processes seem like fossils from the 1950's.

Clearly, if medicine is to become more data-driven, industry and government can do far better than this.

It might help to clean out some of the deadwood that allow drugs for millions to be approved after tests in hundreds.

-- SS


KCFleming said...

Yeah, lots of stuff to rip on here. However, the "Pharma is evil" mantra has the potential to bite us back in the end.

I note the same article cited the dangers of Serevent for asthma. Great. So we'll have fewer asthma drugs and fewer pain medicines too. Goodbye Vioxx, Celebrex and Alleve. Hello, um, Darvocet, Oxycodone and Neurontin. What great alternatives. And risk-free, too!

Look, as we've discussed before, the cardiac risks from Vioxx are interesting, but it is not some thalidomide-like horrific effect. The public reaction seems to me to be quite excessive given the acknowledged risk.

I'll say it again: Be careful what you wish for. This drug safety stampede is looking more irrational every day. If we're not careful, we'll sacrifice many useful medicines on the altar of perfect safety. I'll not agree we've gained that much, and the schadenfreude at Pharma's losses will be short-lived indeed.

InformaticsMD said...

Still...industry and FDA can do better than finding the testing of this level of patients for major new drugs satisfactory.

As we start tinkering with genomics, where our knowledge is even more limited than in the traditional medicinal chemistry-based discovery approaches of the last century, more widepread testing may become even more imperative.

-- SS