Some Remarks about Medical Devices

Medical devices came under official FDA control in 1976 with the Medical Device Amendments. The definition of medical devices encompasses everything from tongue depressors to artificial hearts, all of which must go through some FDA approval procedure. Although the regulation of devices has not been as well studied as the regulation of drugs, it is clear that FDA control of such devices has led to similar costs, especially device lag and device loss (Campbell 1999; Higgs 1995c).

An especially absurd example of device delay occurred to the Sensor Pad. The Sensor Pad is so simple it hardly justifies the term device: it is two sheets of sealed plastic that sandwich a silicon lubricant. With the Sensor Pad, a woman can more easily detect unusual breast lumps in a self-examination. Although the product is simple, it is quite useful and can save lives through early detection of breast cancer. The Sensor Pad was invented in 1986 by Earl Wright of Inventive Products and was submitted to the FDA for approval. The FDA, however, could find no other substantially equivalent product on the market and thus automatically classified the Sensor Pad as a high-risk, Class III device. Before being allowed to sell the Sensor Pad, Inventive Products had to submit a premarket approval application to the FDA.

Inventive Products submitted hundreds of pages documenting the safety and effectiveness of the Sensor Pad (both of which were evident to anyone who used the product for sixty seconds), yet years went by, and still the FDA wanted more documents. As it turned out, the FDA decided that in order for the device to be approved, Inventive Products needed to show not simply that the Sensor Pad helped women to detect breast lumps but that it reduced the mortality of breast cancer. Proof of this kind would require extensive clinical trials involving thousands of women tracked over many years—all this in order to get permission to sell a device substantially less complicated and less dangerous than a toaster oven.

Canada, western Europe, Japan, and other countries quickly approved the Sensor Pad. In Canada, a country that is hardly known as the Wild West of medical devices, the Sensor Pad was approved in thirty days with approximately half a dozen pages of documentation. Frustrated at the delay, Inventive Products defied the FDA and in 1988 began selling the Sensor Pad to hospitals.

Hospitals bought several hundred thousand Sensor Pads and gave them to women, some of whom would later testify in Congress that the product had saved their lives. Doctors too began to use the device and reported positive results. In 1989, however, the FDA raided Inventive Products and a number of hospitals (!) and “on behalf of American women” confiscated the Sensor Pads. Several lawsuits and several years later the product was still not available in the United States.

In April 1994, the situation began to change when the Wall Street Journal ran a story on the Sensor Pad that was later discussed in Congress by Representative John Duncan of Tennessee (the Congressional Record contains Duncan’s remarks and the Wall Street Journal story). Later that same year, ABC’s 20/20 did an investigative report on the FDA highlighting the Sensor Pad fiasco as well as other examples of device delay (see, for example, the Baby Ventilator Incident). The next year, Congress held hearings at which a number of women and doctors testified in favor of the Sensor Pad. Finally, after nearly ten years of delay and several million dollars in legal and other costs, the Sensor Pad was approved in December 1995. But the mandarins at the FDA had the last word: they decreed that the Sensor Pad was to be used only with a doctor’s prescription!

The lesson of the Sensor Pad is not simply one of FDA intransigence. The larger lesson reveals itself when we consider that if Inventive Products had not defied the FDA, there would have been no women ready and able to testify in Congress that the Sensor Pad saved their lives. The media paid attention because it could place a human face on the Sensor Pad story. If Inventive Products had not defied the FDA, knowledge of the Sensor Pad would not have leaked out from behind the FDA’s wall of silence. (It’s noteworthy that, since the Safe Medical Devices Act of 1990, the FDA can fine companies up to $1 million for an alleged violation of the Food, Drug, and Cosmetic Act of 1938 related to medical devices. Inventive Products probably could not afford to defy the FDA today.) Tragically, the usual situation is for information to remain locked behind FDA doors. We rarely get to try products that the FDA fails to approve, and we never get to try products that are never brought into existence because FDA rules and regulations have made research and development uneconomic. As a result, the public remains ignorant of the true costs of FDA power. (The section FDA Incentives discusses this problem at greater length.) It is only in rare cases, when the FDA withdraws approval from a product (recall the lotronex example) or bans an already existing product, that the true costs become clearer.

The Question of Software

According to the FDA, software is a medical device and is regulated accordingly. Because medical software is often new, it can, like the Sensor Pad, be classified as a high-risk Class III device even if it is not involved in high-risk procedures. As with the regulation of other devices and drugs, FDA regulation of software can (and does) often result in less safety. Volokh (1997), for example, noted that

In X-ray therapy for breast cancer, the affected tissue can be irradiated from three different directions, and to avoid overdosing or underdosing the patient, the doctor has to be sure that there are no gaps or overlaps of the X-ray beams. Once the doctor knows which treatment machine is to be used, and what the patient’s measurements are, he can figure out where the patient should be positioned and how the beams should be directed. But this calculation, when performed manually on a calculator, is tedious, error-prone, and time-consuming, taking up to half an hour.

A radiologist developed some software to simplify the calculations. Initially, the software even received FDA clearance. Dr. Martin Weinhous, chief medical physicist in the radiation oncology department at the Cleveland Clinic in Ohio, noted, however, that as the software evolved, “the FDA began to require more of its creator. Eventually, the FDA required that [the developer] test his software tool against every possible PC and Macintosh configuration, even though the inherent risk is small. Faced with unrealistic GMP requirements, he ceased manufacturing of the device. So we in radiation oncology have no choice but to continue to use a less sophisticated and more error-prone method” (quoted in Volokh 1997).

The most serious problem with software regulation is the potential for the FDA to use software regulation as a means of regulating medical practice. Congress has always insisted that the FDA cannot regulate doctors. As a result, once a drug or device has been approved for some use it can be used off-label for any use. As is discussed above and at greater length in Tabarrok (2000), the off-label use of drugs has been a tremendous boon to patients. The FDA has always been uncomfortable with off-label use, however, and has tried to suppress such use whenever possible. With the growing use of software as a component in devices and drug delivery systems, the FDA believes it has found a way to limit off-label uses. Essentially, it is demanding that the software in any device be written so that the device can be operated only for FDA-approved uses. Excimer lasers, for example, have many different uses in eye surgery, but when the FDA approved excimer lasers, they physically and legally tied use of the laser to software that was limited to FDA-approved procedures (Volokh 1997).

The FDA claims, though, that it has not regulated medical practice in the laser case, and this claim has a certain logic. Approved devices may be used for unapproved uses, but unapproved devices cannot be used at all. In tying the excimer laser to restrictive software, the FDA simply approved a narrowly defined product; technically it did not regulate medical practice. The argument illustrates that freedom of choice has survived in the medical arena only because it has hidden itself in the interstices of policy that the FDA once found too costly to fill.

John Calfee explained how this practice could greatly expand:

Imagine that the oncologist using a combination drug therapy uses a computer-controlled device to administer the drugs. Suppose the computer software determines all dosages, and does so according to settings provided by the physician. Now suppose that when the FDA approves the drug-administration device for marketing, it also approves the software in every detail. If the FDA forbids physicians or others from reprogramming the device, it could effectively tell doctors how to administer the drug and could even exercise considerable control over which kinds of patients receive the drug and even which illnesses are treated. (quoted in Volokh 1997)

It should be noted that, although the FDA might, within the framework of existing legislation, be much more oriented toward letting manufacturers, doctors, and patients engage in consensual transactions and might recommend liberalizing amendments, it cannot rightly ignore the restrictions that legislation places on Americans. The fundamental flaw of federal drug policy lies, therefore, not within the FDA itself but within the enabling legislation.

Third-Party Certification and Review

The European Union (EU) maintains strict quality-control standards for medical devices, including extensive reporting requirements, but does so with very little government involvement. Low risk products can be marketed under a self-certification system. Higher-risk products can be marketed after being certified by “notified bodies” that test products and certify that EU standards have been met. Notified bodies are themselves certified by governments. Approval of a product by a notified body is essentially equivalent to government approval.

In 1996, under pressure from Congress, which was considering much stronger legislation, the FDA introduced a third-party certification system that was later extended by the 1997 Modernization Act. Manufacturers have not extensively used the third-party certification program, however, because the FDA has limited its use to relatively simple Class I products. The program and the larger European experience on which it is based does, however, suggest a model for FDA reform based on the creation of drug-certifying bodies. We take up reform options in the next section.