How Widespread Media Coverage, Increased Data Mining Opportunities, And Online Postings By Doctors Can Give Rise To So-Called False Positives
(Posted by Tom Lamb at DrugInjuryWatch.com)
The most recent headlines about Avandia are indicative of why improvements are needed as regards how the current system of reporting adverse drug reactions makes it difficult for the FDA to discern true "signals" of serious side effects from the so-called false positives.
A July 12, 2007 Associated Press (AP) article, "Avandia: Diabetes Drug Side Effect Reports Triple", provided us with some insight about how the media can affect the number of adverse drug reaction reports that are made to the FDA and why this phenomenon presents a problem:
In the month after a surprising analysis revealed possible heart risks from the blockbuster diabetes drug Avandia, reports of side effects to federal regulators tripled....
In the 35 days after May 21,[2007], when the New England Journal of Medicine published the analysis on the Internet, reports of heart attacks, deaths and hospitalizations leaped....
The drug's manufacturer, British-based GlaxoSmithKline PLC, insists that the drug is safe and effective.
"This is a very well-known phenomenon," where news reports lead to increased reporting, said company spokeswoman Mary Anne Rhyne. "It's good that there's awareness of the reporting system, but drawing conclusions on such data is inappropriate."...
"You really can't infer anything about incidence rates from that," because the spike in reports is likely due to the "publicity effect" of the study, said Dr. David Graham, an FDA drug safety expert.
Moving on, Ralph Edwards is director of the World Health Organization's drug monitoring center, which gets about 200,000 new reports of adverse drug events from 83 countries, including the U.S., each year. It is Edward's job to analyze this enormous amount of adverse event data in an attempt to identify emerging patterns of serious side effects; on average, he reportedly finds about 60 such drug safety signals a year. One way he accomplishes this is by using software programs that search methodically through this huge database looking for signs of possible drug dangers. Potentially, this type of data mining can detect rare but serious side effects which previously had not shown up in the standard clinical trials that are done before a drug is approved by the FDA.
Based on his work, Edwards wrote a paper that was published in Drug Safety, a relatively obscure medical journal from New Zealand, suggesting a possible link between statin drugs and and Lou Gehrig's disease -- known medically as amyotrophic lateral sclerosis or ALS -- a progressive neurodegenerative disease is almost always fatal. Edwards' paper received substantial attention when it was featured in an article, "A Risk in Cholesterol Drugs Is Detected, but Is It Real?", that appeared on the front page of the July 3, 2007 edition of The Wall Street Journal (WSJ).
This in-depth article by WSJ reporter Avery Johnson demonstrated how the data mining of large databases is becoming more prominent, and controversial, in detecting emerging drug safety issues. From that article, here is how the issue was explained by a representative of the drug company Pfizer, which makes the popular statin Lipitor:
Manfred Hauben, a Pfizer medical director, has long specialized in analyzing adverse events, and several years ago started using sophisticated software adopted by Pfizer.... Dr. Hauben says his enthusiasm for data mining has cooled.
He says with so many drugs and diseases, mere coincidence will create some worrisome-looking links. "Especially when you're looking for rare events, most positive results are going to be false positives," he says. Also, data mining presents "a lot of opportunities to retrofit an analysis to pre-existing expectations," he says. "Two different vendors' software can give two different results."
Lastly, in May 2007 the American Medical Association (AMA) announced it was beginning a collaboration with the start-up company Sermo that encourages AMA members to exchange ideas and information online with other doctors, including reports of possible drug side effects.
A May 30, 2007 AP article, "AMA To Use Web as Idea-Swapping Tool", reported that this type of online posting by doctors about drug reactions has caused some skepticism and concern:
Dr. William Maisel, a cardiologist at Beth Israel Deaconess who recently chaired a Food and Drug Administration advisory panel on heart stents, said he's skeptical the Web site will supplement government registries and industry systems to track unexpected drug reactions and device malfunctions.
"Incomplete, cryptic descriptions of drug adverse events or device malfunctions are not particularly useful in determining whether a true problem exists," Maisel said. "Without knowing the event rate — not just what went wrong, but out of how many patients exposed — it is difficult to separate a true problem from an expected event."
Alan Goldhammer, of the industry group Pharmaceutical Research and Manufacturers of America, worries that many doctors will essentially be spreading rumors about drug side effects and could "short-circuit" clinical trials and reviews by drug companies and the FDA.
For the reasons set forth above, as Congress works on legislation intended to improve the FDA's ability to monitor emerging drug safety issues, it should focus on how improvements can be made as regards the current system of reporting adverse drug reactions to the agency.