The process by which medical procedures become established medical practice makes the drug-approval process look good. Before the FDA will approve a new drug, the pharmaceutical company must demonstrate that it does something better than the competition's, inspiring a lot of statistical ingenuity. ("We have a saying in research medicine," Roberts says. "If you torture the data long enough, it will confess to anything.") But a new medical device has to clear a lower regulatory bar – only that it's not dangerous – before it can be introduced to the public. The procedure itself is rarely systematically evaluated: A few influential doctors bring out something new, it catches on, and in a few years we have procedures like cardiac stenting to reduce heart attack risk or spinal-fusion surgery to relieve back pain that become institutional cash cows with little scientific evidence that they work as advertised.
Once a procedure becomes established, it's protected by a phalanx of moneyed interests. In How We Do Harm, Brawley recounts the tale of a fledgling federal agency, now called the Agency for Healthcare Research and Quality, which was created to review how well common medical treatments actually work. In 1995 it reported that the research on spinal-fusion surgery for back pain was unequivocal: It produced results that were no better or not much better than doing nothing. Outraged, the North American Spine Society convinced a group of Republican congressmen that the agency was wasting taxpayer money on shoddy research. After nearly losing its funding, the agency limped away from the inquisition with a 21 percent budget cut. "The self-serving surgeons were saying the hell with what the science says," Brawley wrote, "and everyone else was apathetic or worse."
Coronary Bypass Surgery and Stenting
Much of the profit and prestige of modern American medicine derives from two ways of dealing with coronary artery disease and the heart attack risk that goes with it. First came coronary bypass grafts in the 1960s and '70s – surgeons replace a segment of plaque-obstructed artery with a graft vessel, usually harvested from the thigh. Amazing stuff. As Hadler notes, the cardiac surgeons had cornered the market on "gold and glory" until the cardiologists, formerly relegated to mere diagnosis and pill-pushing, got into the act. No longer was it necessary to cut a patient open to get at the problem. The newly christened "interventional cardiologists" could thread a thin plastic tube, slipped in through the wrist or the groin, into the coronary vessels. In the 1980s and '90s, the usual procedure was a balloon angioplasty: inflating a small balloon attached to the tube to compress the plaque against the vessel walls. That's been largely replaced by stenting. After the balloon is inflated, a wire-mesh cylinder, a stent, is inserted to keep the vessel propped open. Stenting and similar procedures have flourished, accounting for more than a million procedures a year. If only they worked as well in the heart vessels as they do on paper. "The bleakest chapter in the entire history of Western medicine" is how Hadler, the Cassandra of American medicine, assesses both bypass surgery and the entire field of interventional cardiology.
David Letterman and Bill Clinton will happily tell you their surgeons saved their lives, but the research plainly argues that for most people, bypass surgery, for all its risks, doesn't produce much or any better results than less-drastic medical therapies.
When it comes to stenting, the best research suggests it's probably a good idea for the patient who is in the throes of a heart attack, and is arguably a good idea for the patient with "unstable" heart disease, at the highest risk for a heart attack. The problem is, about half of the 700,000 stenting procedures done in this country every year are done to patients in no immediate danger of a heart attack. Newman calls it "Whac-a-Mole," guessing which vessel might blow and putting a stent in it. Over the past few years, a pushback has been building, with a small but growing number of criminal cases brought against unscrupulous hospitals and doctors guilty of excessive or inappropriate stenting.
Prostate Cancer Treatment
Often lost in the shouting over PSA is a more fundamental question, on which hinges not only the worth of prostate cancer screening but of prostate cancer treatment itself. When screening and diagnosis works the way it is supposed to – when, say, the doctor has identified an early-stage cancer still localized in the prostate gland that is aggressive, fast-growing, and likely to kill you if not treated – will submitting to surgery or radiation meaningfully extend your life? The eminent urologist Dr. Paul Schellhammer puts it this way: "There is a kind of prostate cancer that can be cured but does not need to be; there is the kind of prostate cancer that needs to be cured and cannot be. We all hope there is a kind of prostate cancer that needs to be cured and can be cured." If that hope is not realized, and you can't effectively fight back when your life depends on it, then all the urology establishment's cheerleading about more-selective screening and new genetic tests to more accurately gauge a cancer's lethality doesn't amount to much.
The numbers do not encourage. In the past 20 years, there have been two huge clinical trials comparing the outcomes of men who had been screened with PSA (if they had prostate cancer, it was more likely to be treated early) with those who had not. The European study showed fewer deaths from prostate cancer in the PSA group but no difference in overall mortality. The American trial saw no difference in either measure, deaths from prostate cancer or total deaths. Another American trial zeroed in on the question at hand, comparing the outcomes of men who had been diagnosed with a PSA test and had their prostates surgically removed with those of PSA-diagnosed men who adopted the watch-and-wait approach. (Only 731 men were tracked in total because it proved so difficult to find men who were willing to watch and wait.) There was no statistically significant difference in the fates of the two groups. There was a suggestion that men with the more aggressive disease did better with surgery, but only just that.
Dr. Robert Mordkin, chief of urology at Virginia Hospital Center, says that when he has a patient with all the hallmarks of an aggressive, early-stage cancer, it would be reckless of him not to consider surgery despite possible side effects. (His clinical sense is that, after a skillfully done surgery, most men will regain their bladder control. Sex is a different story. The penis will be less responsive, though with the help of ED drugs, intercourse, he says, is often possible.) Even though this hypothetical patient has sky-high scores on the PSA and Gleason tests, Mordkin says, "This is a guy who can be cured."
But the epidemiologist asks: Then why haven't all these cures been captured in the rigorous studies that compare the outcomes of men who got the surgery with the men who didn't? "Until I see data from randomized controlled trials," Newman says, "why would I trust what a urologist says?" Almost unbelievably, millions of dollars and scores of studies haven't produced a knockout winner in this fight over the fate of your prostate. Dr. David Bostwick, the nation's preeminent prostate cancer pathologist, argues that many of the epidemiological studies that show screening and early-stage treatment of prostate cancer as having little or no benefit are hopelessly flawed. He believes that patient data accumulated by major cancer centers suggests that surgery and radiation are extending lives.
Here's the takeaway, such as it is: If you have been diagnosed with early-stage prostate cancer, make sure before you contemplate aggressive treatment that you fall in the group that is most likely to benefit from it – with terrible PSA and Gleason numbers. As to how much benefit, you better hope that the urological surgeons and pathologists are right. Brawley notes that the mortality rate for prostate cancer has dropped 20 percent over the past 20 years, but, he says, it's probably not attributable to increased screening and treatment, since the decline began before the advent of widespread PSA screening. "We hope that something we're doing is working," he says.
Disk Surgery for Lower-Back Pain
A disk is a hard-cartilage, fluid-filled shock absorber that sits between two vertebrae. Surgery to remove a damaged disk and fuse together the two vertebrae has been around for close to a century. But it wasn't until the 1970s that the idea that disk fusion might help relieve back pain caught on. If, when a doctor injected fluid into the interior of a damaged disk, it hurt, then maybe the disk was causing the pain. Then, when the MRI came along, you could see that disk damage in breathtaking detail.
But 40 years on we've learned that disk damage showing up on MRI is incredibly common. In one study in Spine, about a third of subjects under 40 without back pain symptoms showed damage. "If there's one thing we know in spine surgery, it's that abnormalities like degenerated disks do not correlate with back pain," says spine surgeon Dr. David Hanscom, author of Back in Control: A Spine Surgeon's Roadmap Out of Chronic Pain. "It's an urban legend that if everything else has been tried for back pain, then you try surgery. It doesn't work."
You can find small individual studies in which fusion surgery appears to be effective at relieving pain – the results are famously unpredictable. But if you look at the research as a whole, fusion tests out about as well as nonsurgical treatment, with positive results between 30 and 60 percent. And because they mostly track patients only for a matter of months, Hanscom says, what these studies only hint at is the long-term surgical fallout: worse pain, infection, the original vertebral fusion breaking down and requiring further fusions to stabilize the spine. Amazingly, the popularity of surgery for back pain continues to rise. According to one study, 200,000 spinal fusions were done in 2000; seven years later there were 500,000, the majority of them done at least in part to relieve chronic back pain. "Right now there are some very high-volume surgeons having major spine centers built for them," Hanscom says. "The hospitals can't get their hands out of the cookie jar."
What surgery can be good for, Hanscom and most any spine surgeon will tell you, is correcting a structural problem that shows up on imaging studies that matches the patient's symptoms. Garden-variety back pain doesn't rise to that level. True, a patient may indeed have a fractured disk that is pressing on a spinal nerve and causing numbness, pain, and muscle weakness, but those symptoms are experienced in the leg, not the back. "I tell those patients that I can do surgery to relieve the symptoms, but it's not going to relieve the back pain," Hanscom says. "And probably two-thirds of the time, that's their biggest complaint."
Hadler, who has contributed his own back-treatment exposé, Stabbed in the Back, regards disk surgery as the only true rival to coronary stents and grafts in the modern medical disaster department. As the North American Spine Society's congressional run around science seems to indicate, some bad ideas are too well defended to die.
The knee story, like the disk-surgery story, follows the pattern spelled out by Welch: "Look for more, see more, treat more." Damage visible on MRI, especially to the menisci, the two crescent-shaped, cartilage shock absorbers between the thigh and shin bones, has helped drive more than 500,000 arthroscopic knee surgeries a year in this country, making it the most common form of orthopedic surgery. With an arthroscope, a three-tubed scope slipped inside the knee, the orthopedist can remove torn pieces of menisci, smooth out aging connective tissues, and generally clean house more quickly and neatly, and with far less rehab time, than with the open knee surgeries that came before. But the research raises a question: Just because you can do something better, does it mean you should? In one 2002 study done by a VA hospital in Houston, the study group received the arthroscopic surgery and the control group got sham surgery – the telltale three-incision stigmata and nothing more. Both groups got about the same benefit in pain relief and mobility, results that were replicated in a Finnish study published nine months ago in the New England Journal of Medicine. In a multicenter study, also in the journal, last year, a group of patients who had a meniscal tear did no better after arthroscopic surgery than a similar group who received six weeks of physical therapy to strengthen the muscles around the knee. Consider that 465,000 "partial meniscectomies" are done in this country every year. "It's open season on the American knee," Hadler says.
A generation or two ago, the doctor was a god and his treatment decisions were commandments written in stone. Today we're coming to understand that even good doctors (and their patients) can be victimized by a health-insurance-driven system that rewards overtreatment. "Very few doctors are offered the opportunity to practice according to their conscience," Hadler says darkly. Still, it would be a mistake to assume any course of action your doctor wants to take is immediately suspect. "If your doctor is adamant about something," Welch explains, "you might want to listen. He might know what he is talking about." More important than whether you and your doctor agree on everything is the respectful give-and-take that should be the hallmark of the therapeutic relationship. As Hadler puts it, if medicine can change from a "telling what to do" profession to a "telling the patient the information" profession, we'll all be better off.