Approval processes different for devices and drugs.

Approval Processes Different for Devices and Drugs

On March 29, 1984, the U.S. Food and Drug Administration approved the first two magnetic resonance imaging scanners for marketing in this country. Ten months later, the National Blue Cross and Blue Shield Association recommended against coverage. "It is premature to accept MRI as a standard, clinically effective, diagnostic technique even for general applications to the central nervous system."

In June 1985, BC/BS updated its assessment and recommended that MRI be considered" generally accepted medical practice for diagnostic conditions of the posterior fossa (cerebellum and brainstem) and high (C1 through C4) cervical cord and demyelinating diseases of the white matter of the brain."

The Health Care Financing Administration did not announce its positive coverage decision for MRI until November 1985, 20 months after FDA approval. The case of MRI illustrates an increasingly prevalent phenomenon. Unlike for drugs, third-party layers do not consider FDA approval of a device an automatic signal for a positive coverage decision. Indeed, HCFA, in its January 30, 1989, Federal Register proposed rule, stated that "FDA approval for the marketing of a medical decvice will not necessarily lead to a favorable coverage recommendation...."

The Blue Cross and Blue Shield Association also has coverage criteria that extend beyond FDA approval. In addition to FDA approval, BC/BS suggests that a device must provide a benefit that is attainable outside an investigational setting. This additional scrutiny of FDA-approved devices establishes a hurdle that a technology must clear before it can become widely available to patients.

Third-party payers have rationalized this extra layer of analysis by pointing to the less stringent FDA approval process for devices compared to drugs. The Food, Drug, and Cosmetic Act, passed in 1938 and amended in 1962, charged the FDA with ensuring the safety and effectiveness of prescription drugs marketed in this country. In 1976, the Medical Device Amendments charged the FDA with the same responsibility for medical devices. Medical devices were very broadly defined and could include anything from a tongue depressor to an MRI unit. Three classes of devices were established, reflecting the degree of control needed to ensure their safety and effectiveness by the Amendments.

Class I devices are those not intended for use in support of human life and offer little risk to human health. Tongue depressors fall into this category. Class I devices are subject only to general controls that address issues of adulteration, misbranding, and registration.

Class II devices are defined in basically the same way as Class I devices. However, mere compliance with general controls is not considered sufficient to ensure safety and effectiveness. Theoretically, Class II devices must comply with performance standards developed for each device. However, the FDA has not yet developed formal performance standards to distinguish devices in these classes, and Class I and Class II devices are currently subject to the same general controls.

A Class III device is defined as one that may be used in supporting or sustaining human life, may be of substantial importance in preventing impairment of human health, or may present a potential unreasonable risk of injury. Two different approval processes--the premarket approval (PMA) and the premarket notification of 501(K) process-are available to Class III devices. To ensure the safety and effectiveness for a premarket approval, the FDA requires the manufacturer to provide reasonable assurance that the device is safe and effective for its intended use.

These safety and effectiveness requirements differ somewhat from those for drugs. Approval of a drug is based on evidence from controlled clinical trials, whereas the evidence necessary to establish the safety and effectiveness of a device is broadly defined as "valid scientific evidence" derived from well-controlled investigations, partially controlled studies, studies and objective trials without matched controls, and reports of significant experience with a marketed device. The underlying philosophy of approval of a Class III medical device is that an increase in the severity of illness or the risk associated with a device requires an increase in the evidence supporting its safety and effectiveness. This evidence, along with supporting materials, must be submitted to the FDA in the form of a premarket approval application.

In 1988, only 46 of the more than 5,000 devices approved for marketing by the FDA went through the premarket approval process. The rest were cleared through the 510(K) process. This process exempts a device from the more rigorous PMA process if it is judged to be substantially equivalent to a device marketed in the United States prior to May 28, 1976. Substantial equivalence means that the device performs the same function as its predecessor and that it possesses similar technical and performance characteristics. Under the 510(K) process, examination is confined primarily to differences between the earlier and new devices that might affect the safety and effectiveness of the device.

Clearance for marketing through the 510(K) provision is understandably preferred by manufacturers because it eliminates the need for development of primary data and extensive evidence of the device's safety and effectiveness. The FDA has interpreted the 510(K) provision liberally because of the agency's staff limitations and the need to get devices to market in a timely fashion. In addition, the FDA has informally created the hybrid 510(K) or mini-PMA process. This is essentially the same process as a 510(K) except the manufacturer submits some data on the safety and effectiveness of a device. In this way, more devices can be approved through the much quicker 510(K) process and can avoid the lengthy PMA approval process. For example, in 1988, the average time for PMA review of a medical device was 262 days, compared with 78 days for review of a 510(K) application.

In summary, in comparison to the approval process for drugs, the PMA process does not require well-controlled trials and the 510(K) process does not formally require any evidence regarding the intended safety and effectiveness of a device. This discrepancy in the approval process between drugs and devices has prompted third-party payers to frequently withhold coverage approval of a device until further evidence of safety and effectiveness is presented.

William T. McGivney, PhD, is Director of the Division of Health Technology, American Medical Association, Chicago, III.