Medical Reporting Beyond Hype and Hope

When I broke into the news business, the financial desk's primary source of breaking news was a Dow Jones wire clack-clack-clacking in the corner of the room. A bell rang whenever a major story broke. Sometimes two bells would go off, signaling a really big story. The day the stock market crashed in 1987, the newsroom sounded like St. Peters Square on Easter.

I imagine something comparable occurs these days when the advance copies of leading medical journals cross science editors' computer screens. Stories from the frontiers of medical research can make it onto page one—the most coveted real estate in daily journalism. News magazines have bolstered their sagging bottom lines with an endless stream of cover stories touting the latest breakthroughs in medicine.

But is this news all that it is cracked up to be? Have the reporters properly weighed the importance of the studies they're touting? Have they asked the tough questions of the researchers and their sponsors to figure out the significance of the results? Have they presented the data in a fashion that is meaningful to health-care consumers? And in an age when most clinical trials are sponsored by private companies, have they fully informed their readers of the researchers' conflicts of interest?

Too often, the answer to these questions is no. Take recent reports from the American Society of Clinical Oncology, which met in mid-May in Orlando. One leading paper reported on a Veterans Administration review of the experience of over 40,000 women in the south central U.S. “The women taking statins were half as likely to have breast cancer as women who were not taking the drugs,” the paper reported. Put that way, it sounds like a dramatic reduction. But elsewhere in the story, it was reported that 12 percent of the women were taking the cholesterol-lowering medications and that only 1.4 percent of the total group contracted breast cancer. Only by massaging the numbers could one figure out that physicians would need to put 700 women on statins to eliminate one cancer case (in medical parlance, this is called number needed to treat). It sounds a lot less impressive that way. But the number needed to treat would be a lot more meaningful to women, especially those on tight budgets wondering if it is worth $1,000 a year for a prescription.

Reporting of surrogate endpoints instead of primary endpoints is another way that readers get misled. Reports on cancer drug trials often fall into this trap. A “lifesaving” drug that shrinks tumors by 50 percent sounds a lot better than a chemotherapy agent that prolongs life by two months. The same can be said for bone density and fractures, blood pressure and strokes, and cholesterol levels and heart attacks. While there may be a minor yet statistically significant reduction in the primary endpoint, the trial sponsors prefer to promote the more dramatic-sounding secondary endpoint. Too many reporters prominently feature the less meaningful number, while leaving out or delaying until late in the story the real bottom line.

Sadly, the media have only lately come around to taking seriously the issue of conflicts of interest in medical science. Last July, the National Heart, Lung, and Blood Institute's National Cholesterol Education Project updated its guidelines for cholesterol management. The update, touted in the front page of every major U.S. paper, called for a dramatic reduction in the cholesterol levels now considered optimal for people who have never had heart disease but are considered moderately at risk. Prescribing physicians using these guidelines will likely put millions more Americans on these drugs in the next few years.

Yet three days after the report came out, reporters at Newsday broke the story that eight of nine physicians on the National Cholesterol Education Project panel had financial ties to statin manufacturers, which had the most to gain from the new guidelines. Writing in the Washington Post, former New England Journal of Medicine editor Jerome Kassirer asked, “Why should we swallow what these studies say?” The ensuing uproar contributed to a change in policy at the New York Times, which last fall circulated a memo to all reporters encouraging them to always report conflicts of interest of quoted sources in science stories, a policy that leading science and medical journals have had in place for many years.

In recent years the pharmaceutical and biotechnology industries have responded to complaints about the high cost of drugs by claiming they are needed to finance the medical miracles that are just around the corner. Meanwhile, the increase in life expectancy in the U.S. has slowed and still remains far below other advanced industrial countries. The number of new drugs coming out of industry labs, despite a slight uptick last year, is actually down from a decade ago. In a health-care environment that is increasingly cost-constrained, it shouldn't be too much to ask that medical reporters get beyond the hype and hope when reporting on the latest “breakthrough.”

This article first appeared in the Public Library of Science's PLoS Medicine in July 2005.

Posted by gooznews at July 26, 2005 11:27 AM