FDA’s process for allowing medical devices to be sold
fails to ensure safety and effectiveness
Medical devices range from crutches and hearing aids to knee replacements and heart valves. They all have to be reviewed by the Food and Drug Administration (FDA) before being sold in the U.S. There are two paths to approval: pre-market approval (PMA) for high risk devices and the “510(k) pre market notification for approval” for lower risk devices (named after section 510(k) of the Food, Drug, and Cosmetic Act). High risk devices are those that are life-sustaining or permanently implanted in patients. More than 90 percent of devices are cleared through the 510(k) notification path, a process that does not require clinical trials but merely a showing that the new device is similar to an existing one.
A study of medical device recalls in the last five years by the National Research Center for Women and Families found that over 70 percent of recalls for “serious health problems or death” were cleared through the 510(k) process.[1]
Consumers Union’s Safe Patient Project has called on Congress to make all implantable or life-sustaining medical devices subject to a more rigorous review by the FDA that increases the safety requirements for these devices. We also called for a better system for tracking devices after approval so patients and doctors can be notified when safety problems arise.[2]
History of FDA’s medical device approval
When the Food, Drug and Cosmetic Act of 1938 was passed it did not require pre-market approval of medical devices. They were not as complicated or widely used as they are today. The FDA only had the authority to stop continued sale of a device already being sold. .During the 1940s and 1950s, the agency removed from the market scores of fraudulent or “quack” devices. [3]
The original medical device law was passed after reports that the Dalkon Shield contraceptive, an intrauterine contraceptive device (IUD) with design flaws, was seriously harming a disproportionate number of women in the early 1970’s. It was later discovered that A.H. Robbins Company, the manufacturer of the device, knew it was unsafe. Three hundred thousand American women testified about being injured by the device and filed claims against the manufacturer, which sold it to an estimated 2.5 million women over four-years. The injuries included ectopic pregnancies, pelvic infections, and infertility due to the design of the device.[4]
Only one study was performed on the device before it was allowed on the market and only for the purpose of determining if it prevented pregnancy.[5] The company’s claims that the Dalkon Shield would reduce infections and the number of involuntarily discharges of the device were later proven to be untrue. Labeling and marketing of the device did not provide full information about the safety issues known by the company. There was no research or requirement for the company to show the device was safe.
In June 1974 Robbins voluntarily removed the Dalkon Shield from the market at the urging of the FDA since the agency had no authority to stop the sale of the devices. They could only sue the company to get the device off the market.
In the early 1970’s a government study group was formed (later known as the Cooper Commission) to determine if medical devices were harming people. The Cooper Commission found that more than 700 deaths and 10,000 injuries were associated with medical devices, including more than 512 deaths and injuries associated with heart valves, 89 deaths and 186 injuries linked to heart pacemakers and ten deaths and 8,000 injuries attributed to intrauterine devices (IUD’s). The commission’s recommendations were part of the device amendments added to the Federal Food, Drug and Cosmetic Act in 1976.[6]
The amendments provided FDA the authority to review and approve medical devices before they could be sold and marketed in the US. The FDA could ban a device, require manufacturers to notify users of risks and repair or replace products or give refunds. And finally, the FDA could enforce regulations without going to court before proceeding with an action.
In 1994 the FDA proposed a user fee paid by medical device companies in order to defray costs associated with the growing number and complexity of device approval submissions. The Medical Device User Fee Act (MDUFA) was modeled on the 1992 Prescription Drug User Fee Act (PDUFA). However, it was opposed by the medical device industry and did not go into effect until 1998. The original bill required that the law expire every five years and it must be renewed to stay in effect. In 2012 the law is expected be renewed for the third time.
Medical Device classifications
The FDA classifies medical devices from low to high risk (Classes I, II and III). Their classification determines if they are approved through the longer more rigorous pre-market approval (PMA) or the 510(k) fast track approval. Class I and II devices may be approved through 510(k).
Class I devices are low risk medical devices such as tongue depressors, reading glasses, and bandages. Devices in this category are not individually regulated, but they are subject to “general controls”—that is, regulations designed to ensure the safe manufacturing and correct labeling of all medical devices.
Class II devices are defined as those that are intermediate risk or present a greater risk of harm than class I devices. The FDA has the authority to require additional information such as clinical trials and post market performance prior to clearing these devices. Class II devices are not considered life-sustaining and include diagnostic devices such as CT scans and MRI’s, as well as implantable devices such as hip and knee joints and surgical mesh that .
Class III devices are designated as high-risk devices and are life-sustaining or necessary to prevent impairment of health. They include implantable devices such as pacemakers, stents, heart valves and some joint implants. All products in this class are intended to go through the premarket approval pathway, requiring review for safety and effectiveness before they can be approved.
The device classification system is far from perfect – it is based on the risks of harming a patient and the amount of information required to adequately approve the device. Even though a class III high risk device approved prior to 1976 did not go through the same rigorous approval process required after the 1976 law passed, they remain on the market. More troubling, current law allows for many class II implantable devices to be approved based on their similarity to a device approved prior to the 1976 law.
One example is silicone breast implants, which began to get attention as an increasing number of women were being harmed by silicone rupturing or leaking into the body. This brought attention to the FDA’s classification and approval of implantable devices. As class II devices, their approval was based on similarities to pre-1976 approved breast implants and were not evaluated for safety and effectiveness. In 1988, the U.S. Food and Drug Administration (FDA) investigated breast-implant failures and the subsequent complications, and re-classified breast implant devices as Class III medical devices, and required manufacturers prove the implant was safe and effective.
Two pathways for approval of devices
Pre-market approval (PMA)
High risk (Class III) devices must have pre-market approval before being sold in the U.S.
Device makers must submit evidence that the device is safe and effective for novel or high-risk Class III devices.
Class III devices are subject to more stringent controls and requirements. These include implanted and life-supporting or life-sustaining devices, although not all implantable devices are Class III. For these devices, comprehensive evaluation, including data from clinical studies, is required to ensure safety and effectiveness. This involves bench and animal tests, clinical trials, the submission of a Pre-Market Approval Application (PMA), and in many cases review by an outside advisory panel. Class III devices comprise less than 1% of marketing applications received by the agency. Examples of devices in this category include heart valves, defibrillators, and computerized microscopes that automatically read Pap smears.[7]
Pre-market notification; the 510(k) process
Section 510(k) of the Food, Drug and Cosmetics Act requires manufacturers to notify the FDA of their intention to market a new product based on a previously approved device like it. The FDA determines if the device is “substantially equivalent” to that device currently marketed device, known as a predicate.
This approval pathway was intended for low risk Class I devices but includes some Class II devices that are surgically implanted such as surgical mesh and some hip and knee replacements if they are based on a predicate device. More than 80 percent of Class II devices are now approved through the 510(k) process. [8]
Problems with the 510(k) approval process
As stated in an Institute of Medicine 2011 report, “Stimulated by reports of problems with several 510(k)-cleared devices, the public, legislators, the Government Accountability Office, the Department of Health and Human Services Office of the Inspector General, and the courts, including the Supreme Court, have all questioned the logic and value of the 510(k) clearance process being used by a federal agency charged with responsibility for protecting and promoting the public’s health.”
- The FDA has the authority to request clinical data to establish substantial equivalence if the new device has different technology or design but must ask for it in the least burdensome way for the manufacturer.
- The 510(k) review does not require evidence that the device is safe and effective if it can be compared to a product already on the market before 1976, even though these pre-1976 devices were approved without safety testing.
- The GAO found that more than 700 devices approved through this process are recalled each year, affecting millions of people. One example of a high profile recall involving a device approved through the 510(k) process is Johnson & Johnson’s artificial hip in August 2010. The metal on metal construction of the artificial hip caused release fragments of metal into the body due to the friction between the metal socket and ball. 93,000 patients worldwide had that implant.[9] Unlike drugs, where people can stop taking them if they experience adverse reactions, an implanted medical device has to be surgically removed; if the device fails, significant medical intervention may be needed.
- The predicate device upon which a new device is cleared may never have been through a clinical trial. Devices may vary quite considerably from the initial predicate, as each newer model with modifications from the initial predicate can then become the predicate for subsequent models. Once a predicate device has been removed from the market, it is not necessarily removed as a predicate, and subsequent devices approved based on that recalled predicate are not automatically reviewed or removed from the market. The FDA’s 501(k) process was also used to clear both Johnson & Johnson’s DePuy hip implants
Solutions to improving system
- Reclassify all permanently implantable devices and all life-sustaining devices to Class III, requiring more stringent review, including safety and effectiveness testing.
- Exclude pre-1976 devices from serving as predicates, so devices being submitted for approval have to show substantial equivalence to a more current device.
- Eliminate the “least burdensome” requirement when the FDA asks for additional information in its review of 510(k) devices.
- Raise the standard for PMA approval of devices from “reasonable assurance” of safety to “substantial evidence” of safety to bring it in line with the standard for drugs.
- If a device has been recalled from the market it should no longer be used by the FDA as a predicate device to approve similar devices and devices approved based on that recalled device should be automatically reviewed for safety.
January 19, 2012
[1] Zuckerman; Brown; Nissen,; Medical Device Recalls and the FDA Approval Process; Arch Intern Med. Published online February 14, 2011. doi:10.1001/archinternmed.2011.30
[2] Lisa McGiffert; DeAnn Friedholm; Consumers Union; http://www.consumersunion.org/pub/core_health_care/018231.html
[3] Peter Barton Hutt, “A History of Government Regulation of Adulteration and Misbranding of Medical Devices,” Food Drug Cosmetic Law Journal 44 (1989): 105; and Peter Barton Hutt and Richard A. Merrill, Food and Drug Law: Cases and Materials, 2d ed. (Westbury, N.Y.: Foundation Press, 1991), p. 735.
[4] http://www.nytimes.com/1987/12/06/magazine/the-sad-legacy-of-the-dalkon-shield.html
[5] Sobol, Richard B. (1991). Bending The Law: The Story Of The Dalkon Shield Bankruptcy. Chicago: The University of Chicago Press.
[6] David L. Stepp; The History of FDA Regulation of Biotechnology in the Twentieth Century
[7] FDA
[8] IOM study 7/9/11, p. 20, prepublication copy.