In a landmark study of a new cardiovascular device unveiled last month, patients received anesthetics, had a large-bore catheter inserted through a cut into one of their major arteries, and had dye injected into their bloodstream. Their surgeons worked on them for about an hour, with unnecessary pokes and prods, while a monitor displayed the false progress using radioactive fluoroscopy.
The patients weren't being treated. They had agreed to undergo the angiogram procedure without knowing if they'd get the real treatment. They were part of a Food and Drug Administration-approved study of a new medical device from Medtronic to treat serious high blood pressure that is resistant to conventional medicines. Some patients were randomly assigned to this sham surgery. They were placebos.
Was their sacrifice worth it? That is a question many patients may want to consider as the FDA insists on a new study methodology with uncertain benefits. The methodology's high costs mean that some new products will reach patients many years later than they need to—or not at all.
Everyone is familiar with the long-standing practice of randomly assigning patients to placebo pills in clinical trials that test new drugs. One group of patients will get the experimental medicine, and another group will get a dummy sugar pill. The FDA began encouraging some medical-device makers to use dummy surgical procedures a few years ago, and a review of Clinicaltrials.gov shows that dozens of such studies are under way.
The goal is to isolate the observed effect of a new treatment from other factors that could affect the results. For instance, the blood-pressure device works by ablating, or destroying small nerves in arteries that feed the kidneys with blood. The activity of these nerves contributes to hypertension. So the FDA would want to know things like whether patients getting the real procedure and device achieved lower blood pressure from the psychological influence of the surgery, rather than from the new device.
Yet other scientific approaches also allow us to get rigorous evidence of safety and benefits, while enabling patients to get real rather than fake therapies. In the case of the high-blood-pressure device, 10 years ago the FDA would have requested a reasonably sized "non inferiority" trial. That is a study in which patients getting the new procedure are randomly compared with patients getting a different standard treatment for the same condition—to see if the new device is about as effective as conventional treatment.
New device technologies can offer other advantages, such as faster recovery times, lower risk to patients, or easier use by doctors and patients. Non-inferiority studies are often better for evaluating these secondary benefits.
But for several years the FDA has clamped down on these approaches in favor of sham surgeries, which it sees as more statistically scrupulous and free from bias. The high-blood-pressure device, for example, is already available in Europe, where regulators approved it based on traditional studies. The FDA disregarded those results in favor of the new and larger trial using a sham.
Preliminary results from the sham study suggested that the device might not deliver the hoped-for benefits. While some people think the problem wasn't with the device but more with the way the procedure was designed in that trial, the negative results are already emboldening proponents of sham studies.
Sham procedures could be defensible in certain narrow cases—for example, when a minimally invasive (very low risk) device is used to affect symptoms that are subjective and hard to measure. One example is devices aimed at treating pain. Another is when surgery is already under way, and the sham might be an experimental step that the surgeon forgoes in some patients.
Yet research that introduces harm or risk with no opportunity for benefit would seem to conflict with the principles governing research on humans. Some of these are reflected in the Declaration of Helsinki, an international treaty concerning the conduct of medical research. Other experiments using sham surgeries are obligating patients to undergo unnecessary anesthetics, radiation, abdominal incisions, endoscopy and injections into the rectum, to mention a few examples. The needless cutting means pain as well as the risk of anesthesia and infection.
The FDA tries to address ethics issues by letting patients who get sham treatments eventually join the real treatment group. But this often requires a second surgery. The sham trials can also be costly because they involve unnecessary operations. They are hard to recruit for when patients know they may get a fake surgery and are reluctant to consent to being cut unnecessarily.
All of this raises development costs—and it encourages firms to skip the U.S. market and commercialize new products overseas. This can suppress innovation. When a sham trial doesn't produce positive results, the company may have exhausted its resources and have no capital left to refine a good idea into a beneficial product.
Instead of clinging to inflexible testing requirements, the FDA should allow trials that are feasible, reflect clinical practice and are morally defensible. There are methods for evaluating science that don't require such contrived experiments on people. The agency doesn't need to rely on research models that raise the opportunity costs so high that some valuable treatments or devices may never become available to patients.
Dr. Gottlieb is a physician and resident fellow at the American Enterprise Institute. He also consults with firms that invest in medical device companies.