Gaps, Tensions, and Conflicts in the FDA Approval Process: Implications for Clinical Practice

Richard A. Deyo, MD, MPH


J Am Board Fam Med. 2004;17(2) 

In This Article

Things the FDA Does Not Do

There are some potentially valuable functions the FDA does not perform. For example, it does not approve old drugs and devices. Some medical products in wide use were marketed before FDA approval was required, and their use is "grandfathered" in.

The FDA makes no judgment about the value for money of a new drug or device. Dr. Larry Kessler, Director of the Office of Science and Technology in FDA's Center for Devices and Radiologic Health, says if a manufacturer wanted to market "a gold-plated biliary stent that costs a million dollars a pop—works great—FDA has to approve it. It's a lousy buy because the $127 version works almost as well. But FDA has to approve it. Medicare may decide it's not cost-effective and refuse to pay for it, but FDA cannot address cost-effectiveness" (L. Kessler, personal communication). In truth, even Medicare cannot make reimbursement decisions based on cost-effectiveness, although private health plans and state Medicaid programs can.

The FDA does not determine whether one blood pressure drug is better than another for reducing the risk of blood pressure complications (like strokes and congestive failure). It does not require that drugs prove this effect, nor does it require head-to-head comparisons of competing drugs or devices. Dr. Kessler says, "The reason is that we could be seen as favoring product A over product B. And FDA always, always, always shies away from that" (L. Kessler, personal communication).

Some consequences of this policy were illustrated by results of recent clinical trials. In the ALLHAT trial, diuretic therapy was found to be more effective at preventing cardiovascular complications of hypertension than were calcium channel blockers or angiotensin-converting enzyme (ACE) inhibitors.[7] Because no adequate comparisons were previously available, and because the newer drugs were heavily marketed, diuretics had come to be used in only a minority of patients, whereas calcium channel drugs and ACE inhibitors (at 10 to 20 times the cost) had steadily gained market share.[10] Similarly, a recent study demonstrated no advantage of ticlopidine over aspirin for preventing recurrent strokes among African Americans, despite ticlopidine's substantially higher price and its risk of serious adverse events.[11]

The FDA does not approve every use to which a medical product might be put. A drug can be marketed after approval for treating one condition, and doctors can legally use it for others. Gabapentin (Neurontin) for example, is approved as an adjunct for treating seizures and for management of postherpetic neuralgia, but many physicians use it to treat long-term pain and psychiatric problems. Unfortunately, there may be little if any scientific evidence to support off-label uses. "When you routinely recommend some off-label use for your patients—for which there aren't data to prove this is the right thing—when does that really become experimentation without informed consent?" asks Dr. Kessler (L. Kessler, personal communication).

The FDA does not approve television or magazine ads for new drugs before they are aired or printed, although companies are required to submit ads to the agency at the time they are first disseminated. Because of their wide exposure, direct-to-consumer broadcast ads are all reviewed, although some print ads are not.[12] FDA can only request that a company pull ads that are judged misleading after they are already in use. In late 2001, the Department of Health and Human Services instructed FDA not to issue regulatory letters until they are reviewed by the agency's Office of the Chief Counsel. This has delayed letters from 2 to 12 weeks: long enough in some cases for misleading ads to complete their planned broadcast life cycle. FDA cannot verify that it receives all new ads from drug companies and has issued 6 regulatory letters since 1997 citing companies for failing to submit ads when they were first disseminated.[12]

FDA has few resources for routine monitoring of ads and often issues warning letters only after competitors complain. For the interval 2000 to 2002, The FDA web site lists 222 letters to drug makers for warnings or violations of advertising rules. The agency has repeatedly contacted several companies, including Pfizer, Schering-Plough, Merck, and Glaxo Wellcome (now GlaxoSmith-Kline) for violations. These include improper claims and minimizing drug risks.[12] The agency cannot levy fines, and several drug companies have aired new misleading ads even after being cited for violations, according to a 2002 report by the Congressional General Accounting Office.[12]

The FDA recalls drugs or devices if new evidence emerges suggesting they are unsafe. However, it generally does not recall drugs or devices because of accumulating evidence they do not work. The marketplace is judged sufficient to accomplish this goal, and it sometimes does. But if this process is slow, tests and procedures may continue in use for years after they have been found to be ineffective or inferior to alternative products.

The FDA does not regulate new surgical procedures in any way. It does regulate devices, such as surgical implants. Examples would be metal hip replacements or cardiac pacemakers. It also regulates new surgical instruments, such as the fiberoptic scopes that are increasingly used for minimally invasive surgery. But if a surgeon develops a new approach or technique that does not involve a new device, it falls outside the jurisdiction of the FDA.