Don't be too eager to try the latest medical gadgets

Last fall, a manufacturer of implanted heart defibrillators recalled one of its electrodes after three years of marketing and after 268,000 of the devices had been implanted in patients. The electrode, made by Medtronic, was approved by the FDA in 2004 on the basis of laboratory tests, but no human studies.

For someone prone to dangerous heart rhythms, implanted defibrillators deliver a shock that restores a normal heartbeat, often saving a life. Unfortunately, when the electrode wires to the heart fail, they may not deliver the critical shock, or deliver shocks when they aren't needed, sometimes taking a life.

When concerns about the device arose, it was unclear despite years of use and thousands of patients if there was truly a performance problem. The company finally studied 25,000 patients and eventually recalled the defective electrode. Similar recalls have affected hundreds of thousands of defibrillators and pacemakers (without defibrillators) in recent years.

For patients with these devices, such recalls pose a terrible dilemma: Does the risk of removing a faulty pacemaker or defibrillator outweigh the risk of keeping it? When devices are recalled, the manufacturers typically offer a replacement device free. But they won't pay the doctor and hospital bills necessary to evaluate patients and perform the replacement.

Heart patients aren't alone in confronting this problem. Recalls of medical implants — from heart valves and defibrillators to artificial hips — are as common as drug recalls. And studies questioning the effectiveness of new devices appear regularly.

The medical-device industry, worth around $86 billion in the U.S. alone, has flown under the radar in comparison with drug companies. As it does for drugs, the Food and Drug Administration approves new medical devices.

But there's a double standard. Drug makers have to test their drugs in randomized trials to prove they work, even for "me too" drugs that mimic others on the market. In contrast, device makers generally need only argue that their products are "substantially equivalent" to something marketed before 1976, or one that's currently sold.

That's because the FDA first acquired authority over medical devices in 1976, and devices on the market before that were "grandfathered." Today, about 98 percent of new devices are approved based on "equivalence" to something already on the market. The bar is higher for the riskiest devices, but randomized trials aren't always required.

As for safety, former FDA commissioner Mark McClellan observed that "new devices are less likely than drugs to have their safety established clinically before they are marketed."

Excuse me? Given recent controversies with drugs like Vioxx, that pronouncement itself may be heart-stopping.

In 2008, it no longer makes sense to approve new medical devices because they resemble something marketed 30 years ago. It's time to require more rigorous proof of effectiveness, including randomized trials for implanted devices and head-to-head comparisons of competing devices.

The FDA estimates 300,000 deaths and injuries a year are associated with medical devices. In 2005, complications from medical devices, implants and grafts resulted in almost 616,000 hospitalizations, with hospital bills alone totaling more than $25 billion.

This year, Medtronic has also recalled some implantable infusion pumps, used to send pain medications or cancer chemotherapy right into the spinal canal. This recall affected thousands of patients.

Other studies this year have challenged the effectiveness of expensive gadgets used to monitor patient awareness during anesthesia, and have suggested that coronary stents, used to prop open clogged arteries, may not improve survival over drug therapy alone for patients with stable heart disease. Of course, there have been multiple recalls of stents themselves, involving tens of thousands of devices.

What do these problems mean for patients and policymakers?

First, patients shouldn't be too eager to try the latest gadgets. As a recent medical journal commentary noted, newly approved doesn't necessarily mean new and improved.

The FDA needs more funding to expand and strengthen the device-surveillance system. And both doctors and patients should be notified of device flaws directly.

Manufacturers should be required to pay the full costs of evaluating patients and, if necessary, replacing devices when recalls occur.

Let's subject medical implants to at least the same scrutiny as new drugs. After all, they're often meant to stay in the body for a lifetime — unlike drugs, which are just passing through.