January 05, 2005

Some new standard questions

Here's a good chance to practice our (hopefully) renewed skepticism in the face of the influence that big drug manufacturers have over clinical research.

The results of two new clinical studies on heart disease are grabbing headlines, and are likely to get a lot of public and professional attention. Both studies were sponsored by big pharmaceutical companies that make cholesterol-lowering statin drugs. According to the NYT, both studies suggest that C-reactive protein, or CRP, is an independent marker for important clinical endpoints--in the case of one study, heart attacks, and in the case of the other, heart disease progression. These studies might suggest that physicians should try to reduce their patients' CRP levels as well as their cholesterol levels. More aggressive use of statin drugs would be one way for physicians to do this.

The drug companies will almost certainly encourage patients, in the sonorous tones of the inevitable Pfizer ad, to "ask your doctor about CRP, and whether Lipitor is right for you." Bristol-Meyers-Squibb will trumpet the results of these studies in their Prevachol ads weighing down the next issue of American Family Physician. Both firms will certainly instruct their sales force to mention both studies in the detail lunches they provide in the back room of doctors' offices across the country.

Before the docs start reaching for their prescription pads, we can hope that they'll subject these studies to the usual close statistical and clinical scrutiny that responsible physicians should direct at any study. They'll note that neither study demonstrates that CRP is a cause of bad clinical outcomes. They'll realize that both studies looked at patients with known heart disease, and question the applicability of the results to healthy patients. Perhaps they'll also (God forbid) remember that exercise reduces CRP too.

Since we're wishing on stars, let's hope that they'll subject the studies to an economic and political scrutiny, too. How many studies may have been started or finished (but never published) which showed no independent correlation between CRP and any clinical endpoint? How much control did Pfizer and BMS have over the data in these two studies?

Seems to me, that's the physician's job. To exercise his or her best judgment on behalf of patients.

Posted by Carey at January 5, 2005 07:28 PM