Order Code RL32826
CRS Report for Congress
Received through the CRS Web
The U.S. Approval Process for Medical Devices:
Legislative Issues and Comparison
with the Drug Model
March 23, 2005
Michele Schoonmaker
Specialist in Genetics
Domestic Social Policy Division
Congressional Research Service ˜ The Library of Congress

---

The U.S. Approval Process for Medical Devices:
Legislative Issues and Comparison with the Drug Model
Summary
In response to highly publicized cases concerning the safety of prescription
drugs and medical devices erupted, Congress convened several hearings in 2004 to
examine the effectiveness of the U.S. Food and Drug Administration’s (FDA) review
processes for new medical products. In three cases, information was available to
demonstrate that certain drugs might be unsafe and/or ineffective, yet they continued
to be marketed and prescribed. Legislation introduced in the 109th Congress (the
Fair Access to Clinical Trials Act, S. 470) seeks to improve drug safety by requiring
more transparency in disclosing pre- and postmarket clinical trials of FDA’s
regulated medical products, including drugs, devices, and biologicals (among other
trials). The legislation would also impose requirements for the disclosure of financial
conflicts of interest between investigators and manufacturers. The broad scope of
similar legislation proposed during the 108th Congress (S. 2933 and H.R. 5252) led
some to question the appropriateness of applying disclosure requirements to the
medical device industry, given differences in the way drugs and devices are approved
and in the nature of the industry and product development. Legislation was not
passed in the 108th, and debate on whether FDA’s pre- and postmarket review
processes are sufficient to protect the public from unsafe medical products is likely
to continue in the 109th Congress, particularly with regard to the balance between
timeliness and substance of premarket review, the adequacy of postmarket
surveillance mechanisms, and the disclosure of clinical data.
The medical device market is diverse: surgical and medical supplies constitute
the largest sector, followed by in vitro diagnostics (IVDs), cardiovascular devices,
orthopedic devices and diagnostic imaging. Medical devices can be legally marketed
in the United States in several ways. First, they are classified according to the risk
that is posed to the patient from their use or misuse. The classification determines
the type of premarket application, if any, that FDA will require. The higher the risk,
the more stringent the premarket review conducted for approval. If a product is
exempt from premarket review, the manufacturer need only register its facilities, list
its devices with FDA, and follow general controls requirements. If premarket review
is required, manufacturers can demonstrate substantial equivalence of their device
with a legally marketed device, or can demonstrate that it is safe and effective on its
own merits for the purpose intended by the company. Once approved, all
manufacturers are required to report serious adverse events associated with the use
of their devices to FDA. In addition, tracking is required for some medical devices.
In many ways, the device-approval process is more flexible than that for drugs, with
the majority of devices being cleared or approved without being evaluated in a true
clinical trial.
This report, which will be updated, describes FDA’s approval process for
medical devices, compares it with the approval process for drugs, and serves as a
primer for broader discussions of the impact of various legislative options aimed at
maximizing both patient safety and availability of beneficial medical products. For
more information about the U.S. drug approval process, see CRS Report RL32797:
Drug Safety and Effectiveness: Issues and Action Options After FDA Approval.

---

Contents
Overview of Legislative Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Speed of Approval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Postmarket Surveillance System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Off-Label Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Disclosure of Study Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Financial Conflict of Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
History of Device Legislation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Medical Device Approval Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Premarket Clearance or Approval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Device Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Marketing Applications for Medical Devices . . . . . . . . . . . . . . . . . . . 13
In vitro Diagnostic products (IVD) . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Post-Approval Requirements and Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Postmarket Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Compliance and Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Comparison of Device and Drug Models for Regulation . . . . . . . . . . . . . . . . . . 27
Industry and Product Related Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
FDA Review and Approval Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Premarket . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Postmarket . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
List of Tables
Table 1. Number of Marketing Applications for Medical Devices
Received, Approved and Denied by CDRH in 2003 . . . . . . . . . . . . . . . . . . 19
Table 2. Comparison of the Drug and Device Approval Processes . . . . . . . . . . . 31

---

The U.S. Approval Process for Medical
Devices: Legislative Issues and Comparison
with the Drug Model
Overview of Legislative Issues
Introduction
In response to highly publicized cases concerning the safety of prescription
drugs and medical devices erupted, Congress convened several hearings in 2004 to
examine the effectiveness of the U.S. Food and Drug Administration’s (FDA) review
processes for new medical products.1 At these hearings, some Members raised
questions as to FDA’s capacity or willingness to enforce safety requirements,
including FDA-ordered follow-up studies for products approved by the agency. One
overarching question for Congress is: Does FDA have the proper statutory tools and
regulatory authority it needs to fulfill its public health mission? If not, what
additional steps need to be taken? If it does, what are the barriers to enforcing the
current laws? Debate on whether FDA’s pre- and postmarket review processes are
sufficient to protect the public from unsafe medical products is likely to continue in
the 109th Congress, particularly with regard to the balance between timeliness and
substance of premarket review, the adequacy of postmarket surveillance and the
disclosure of clinical data. Two of the selected cases below illustrate the drug safety
questions that triggered congressional interest in FDA’s process for monitoring the
safety of medical products. Two other cases demonstrate how the questions and
issues surrounding FDA’s assessments of safety and efficacy may also extend to
medical devices.
Two hearings in 2004 investigated whether several drug manufacturers withheld
evidence from the public that their antidepressant drugs were unsafe, or, at best,
ineffective, in the treatment of pediatric patients.2 Beginning in June 2003, British
1 U.S. Congress, House of Representatives, Committee on Energy and Commerce,
Subcommittee on Oversight and Investigation, Publication and Disclosure Issues in
Anti-Depressant Pediatric Clinical Trials, Sept. 9, 2004 (Serial no. 108-121); and FDA’s
Role in Protecting the Public Health: Examining FDA’s Review of Safety and Efficacy
Concerns in Anti-Depressant Use by Children Sept. 23, 2004 (Serial no. 108-125); U.S.
Congress, Senate, Committee on Finance, FDA, Merck and Vioxx: Putting Patient Safety
First? Nov. 18, 2004, [http://finance.senate.gov/sitepages/hearing111804.htm].
2 Since 1987, nine drugs have been approved for treatment of depression in adults.
Questions about the safety and effectiveness of antidepressant drugs have been raised
publicly as early as 1991. Despite these questions and FDA’s silence on pediatric use, U.S.
(continued...)

---

CRS-2
drug regulators issued strong recommendations against prescribing certain
antidepressants for children.3 Following regulatory action in the United Kingdom,
questions arose concerning the FDA’s failure to take similar measures. On October
15, 2004, the FDA formally requested that manufacturers of the drugs include
expanded and prominent warnings in their labeling, and issued a Public Health
Advisory warning health care providers and patients of the added risk.
Another hearing received testimony concerning the anti-inflammatory drug,
Vioxx, which was primarily used to treat arthritis. On September 30, 2004, Merck
& Co., Inc. notified the FDA that it was withdrawing its product Vioxx4 from the
market in response to recent study results that indicated an increased risk of heart
attacks and sudden cardiac deaths in users. Some praised Merck’s quick and decisive
response; others (e.g., scientists, researchers, consumers, and physicians) claimed
that data indicating the adverse cardiac effects had been available for at least the last
three years. An estimated 93 million Vioxx prescriptions were reported to have been
written since the drug’s introduction in 1999.
Though not the subject of a congressional hearing in the 108th Congress, two
similar market withdrawals occurred in the medical device industry. In July 2004,
Boston Scientific Corporation withdrew two products from the market.5 The
Express2TM (Bare metal) Coronary Stent6 and the TaxusTM Express2TM (paclitaxel-
eluting) Coronary Stent were recalled because characteristics in the design resulted
in the failure of the balloon to deflate and impeded removal of the balloon after the
stent was placed in the coronary artery. Several years prior to the 2004 recall, device
failures involving the balloon were reported with a different product, the NIR ON
Ranger w/SOX cardiac stent. On September 17, 1998, Boston Scientific’s chairman
2 (...continued)
doctors wrote 10.8 million prescriptions for antidepressants for children and adolescents in
2002 (some children get more than one prescription per year). See “Helping Depressed
Children,” Nature, vol. 431, Sept. 9, 2004, p. 111.
3 Only Prozac was found to have a beneficial effect in children. See the Statement of the
Medicines and Healthcare Products Regulatory Agency at [http://www.mhra.gov.uk],
updated Feb. 2004.
4 Vioxx (rofecoxib) is a nonsteroidal anti-inflammatory drug (NSAID) that selectively
inhibits cyclooxygenase-2 (COX-2). NSAIDs are used primarily for pain relief. COX-2
inhibitors are thought to provide additional benefit over older NSAIDs (such as ibuprofen
and naproxen) by reducing the intensity and/or frequency of adverse gastrointestinal events.
Other COX-2 inhibitors are Pfizer’s Celebrex (celecoxib) and Bextra (valdecoxib).
5 See Medical Device Recalls, Class I Recall: Boston Scientific Express Coronary Stent, at
[http://www.fda.gov/cdrh/recalls/recall-071604.html], and Medical Device Recalls, Class
I Recall: Boston Scientific Taxus Express Coronary Stent, at [http://www.fda.gov/
cdrh/recalls/recall-070104.html].
6 A cardiac stent is usually a slotted stainless steel tube or coil meshwork that is inserted into
a blocked artery during an operation. Once the stent is inserted, the cardiologist uses saline
to inflate a balloon in the stent, causing it to expand and hold the artery open. Over time,
the artery may collapse or re-close (called restenosis). A drug-eluting stent is a normal
metal stent that has been coated with a drug known to interfere with the process of
restenosis.

---

CRS-3
and chief executive officer indicated the company’s awareness of potential problems
in a conference call to advisors and business partners.7 Despite the acknowledgment,
the company continued to ship thousands of stents while negotiating with FDA to
avoid a recall. On October 5, 1998, the company issued a recall of the defective
product. Also in 1998, the company disclosed an ongoing criminal investigation by
the Department of Justice surrounding the firm’s decision to delay the recall. Though
the company is still under investigation, a recent newspaper report stated that two
company officials would not face charges.8
Later in 2004, Access Cardiosystems, Inc. issued a worldwide recall of
approximately 10,000 automated external defibrillators (AEDs).9 AEDs are portable
devices used to restore normal heart rhythm to patients in cardiac arrest. The
problems leading to the recall involved faulty parts that failed to deliver shocks, or
in other situations, that resulted in the ON/OFF button becoming inoperative. The
company is no longer in business.
This report describes FDA’s approval process for medical devices, compares it
with the approval process for drugs, and serves as a primer for broader discussions
of the impact of various legislative options aimed at maximizing both patient safety
and availability of beneficial medical products. For background information about
the U.S. drug approval process, see CRS Report RL30989, The U.S. Drug Approval
Process: A Primer, by Blanchard Randall IV.
Speed of Approval
Thorough evaluation of a new medical product can take a long time. Adverse
events are often rare, and they may not become apparent until the product is used in
a large number of patients. Large studies are expensive for companies, and long-term
follow-up of patients can be difficult as they relocate or change health care providers.
In the past, FDA was often criticized for the length of time it took for a new product
to get approval. Health care crises, such as the war on cancer and the AIDS
epidemic, led to public outcries for faster access to new medicines. Industry claimed
FDA was killing or injuring patients by holding up beneficial new medicines. Others
were concerned that FDA action was hasty or lenient, bending to pressure from
industry.
Congress reacted in several ways. In 1992, the Prescription Drug User Fee Act
(PDUFA; P.L. 102-571), which authorized FDA to collect fees for reviewing new
drug applications, was enacted. The fees may be used to increase FDA’s technical
and human resources. PDUFA also established performance goals to reduce
approval times. In 1997, the Food and Drug Modernization Act (FDAMA;
7 Jeffrey Krasner, “Stents Sold Despite High Failure Rates: Boston Scientific Held Off
Recall Amid Alarming Data,” Boston Globe, Dec. 26, 2004, p. A1.
8 Ross Kerber, “Stent Firm Still Target of Federal Inquiry; Boston Scientific Says Two
Officials Won’t Face Charges,” Boston Globe, Nov. 13, 2004, p. C1.
9 See “Medical Device Recalls, Class 1 Recall: Access CardioSystems Automated External
Defibrillators,” at [http://www.fda.gov/cdrh/recalls/recall-110304.html].

---

CRS-4
P.L.105-115) gave FDA authority to establish procedures for expediting review of
new drugs and devices, with less stringent thresholds of evidence required to
demonstrate safety and effectiveness.10 In 2002, the Medical Device User Fee and
Modernization Act (MDUFMA; P.L. 107-250) accomplished for devices what
PDUFA did for drugs, authorizing FDA to collect fees for reviewing new device
applications and also establishing performance goals for reducing review time.
Postmarket Surveillance System
All of the efforts to decrease the time FDA took to review a new product by
easing premarket requirements indeed accelerated product approval time. However,
questions remained about the adequacy of FDA’s review process in protecting public
health, particularly when the modifications to the premarket approval process were
not balanced by increased diligence in postmarket surveillance. In a report to
Congress in March of 2004, FDA indicated that although companies had agreed to
conduct a total of 1,338 post-approval drug studies, 65% of these studies had not
begun, and just 33% were on schedule or had been completed. An internal FDA
study examining postmarketing requirements for 127 devices approved between
1998 and 2000 found that of 45 approvals that required manufacturers to conduct
postmarket review, 10 (or 22%) did not submit the required follow-up results.11
Recent events also raise questions about whether the current postmarket
surveillance mechanisms adequately identify unsafe drugs and devices. Are there
better mechanisms for identifying unsafe products? For example, FDA requires
manufacturers of certain medical devices to develop and maintain a tracking system
that will enable them to quickly locate patients using devices once they are
commercially available. Supporters of the regulation indicate that tracking
information can facilitate notifications and recalls in the event that serious risks
become evident. Opponents cite cost and difficulties in following and locating
patients over time as barriers to tracking a large number of devices. Some devices,
such as laboratory tests, may be almost impossible to track due to the large number
of users for high-volume tests. Tracking, as done for devices, is currently not
required of any drug; such tracking would likely be difficult to implement given the
high volume of drugs prescribed and the difficulty of determining patient
compliance. If products are identified as unsafe after marketing, what barriers exist
to FDA’s ability to correct problems (i.e., by restricting their use after approval or
removing them from the market)? Given differences in the way drugs and devices
are used, produced, and regulated, should remedies apply evenly across medical
products?
10 Known as the “least burdensome” approach, it mainly reduced the number of controlled
studies that constituted “valid scientific evidence” from two to one, and led to the
acceptance of surrogate endpoints (laboratory or physical findings associated with disease)
rather than clinical outcomes (such as improvement in survival, reduction in morbidity, etc.)
as a basis for establishing effectiveness.
11 R. Kerber, “22% of Medical Device Follow-Up Studies Left Undone,” Boston Globe, Dec.
24, 2004, p. D1.

---

CRS-5
Off-Label Use. Off-label use occurs when a medical product is used for a
condition or in a manner that was not approved by the FDA. It is a routine practice
in medicine that healthcare professionals have flexibility in prescribing interventions
that best meet the needs of their patients, regardless of the FDA-approved indications
described in the label. Off-label use is particularly prevalent in terminal care, such
as with cancer interventions. Manufacturers are required to report serious adverse
incidents resulting from off-label use, and the FDA can take action (such as ordering
labeling changes) to indicate the dangers of off-label use. However, the FDA argues
that it would be nearly impossible for either the manufacturer or the FDA to predict
and label against all potential off-label uses and issue warnings accordingly.
Following the congressional hearings, the FDA released a statement and plan
on November 5, 2004, regarding strengthening its post-approval safety program.12
Elements include sponsoring a study by the Institute of Medicine (IOM) to assess and
make recommendations about improving FDA’s drug safety system, especially
postmarketing monitoring. IOM is to also examine whether FDA should develop an
organizational unit to oversee drug postmarketing safety issues separate from the
office that conducts premarket assessments.13 These actions focus on drug safety
issues, and will not investigate aspects of the device approval process.
Disclosure of Study Results
All drugs and devices carry some risk; none are 100% safe and effective. Part
of FDA’s responsibility in both pre- and postmarket review is to ensure that the risk
is reasonable given the expected benefits for the intended patient population. Each
year, approximately 1-2 drugs and 6-8 devices are removed from the market for
safety concerns.14 What makes the cases in 2004 significant is, in part, the notion
that FDA and/or the manufacturer failed to disclose data concerning the potential
problems in a timely manner, exposing many patients to serious health risks.
Questions that Congress may wish to consider include: What is an appropriate
balance between the availability of a product and public safety? When do
manufacturers or the FDA have a duty to disclose any evidence of a potential
problem with a product? Were there signals identified during premarket review of
the products discussed above that could have prevented or minimized these cases had
postmarket surveillance and disclosure been in effect? Who should determine what
is an “acceptable risk” to a patient who may have no other alternative than to use a
potentially dangerous drug or device or suffer from their disease or condition?
Under the Adverse Event Reporting (AER) and Medical Device Reporting
(MDR) regulations for drugs, biologics, and devices, manufacturers are required to
12 Food and Drug Administration, FDA Acts to Strengthen the Safety Program for Marketed
Drugs, FDA Statement, Nov. 5, 2004, at [http://www.fda.gov/bbs/topics/news/2004/
NEW01131.html].
13 “FDA Drug Safety Review Will Weigh Need For A Separate Safety Agency,” The Green
Sheet, Nov. 15, 2004, p. 1.
14 U.S. General Accounting Office, Drugs Withdrawn from Market, Jan. 19, 2001 [GAO-01-
286R]; and Medical Device Recalls: An Overview and Analysis 1983-88, Aug. 1989
[GAO/PEMD-89-15BR].

---

CRS-6
report to FDA any serious adverse events associated with use or misuse of their
products. At any time they become aware that there is a potential concern with their
product, manufacturers can voluntarily notify health professionals and consumers,
and/or with FDA approval, change the labeling (written material supplied with the
product) to add precautionary or warning language that would improve safe and
effective use of their product. Alternatively, FDA can order that the label be changed
as soon as there is reasonable evidence that there is an association between the
product and a serious hazard.
The difficulty comes with the definition of “reasonable evidence”and the
boundaries of “association” with a serious adverse event. Because of the rarity of
most adverse events, it is often difficult to determine whether the product or some
other unknown factor is causing the event, particularly when the determination is
made outside of the controlled environment of a trial. Furthermore, no single trial
is likely to identify all possible adverse events. In the case of pediatric anti-
depressants, the adverse outcome — an increase in suicidal behaviors in young
people — was also associated with the condition that the drug was designed to treat:
depression. In recent years, the policy has been to err on the side of keeping a
product on the market, maintaining patient access for those who do experience
clinical benefits.
Financial Conflict of Interest. Several concerns have been raised as to the
potential conflicts of interest that may arise when clinical trial investigators,
Institutional Review Board (IRB) members (who approve protocols for the conduct
of clinical studies), and FDA advisory committee members hold financial interests
in the companies whose products are coming before them. With a declining rate of
growth in NIH funding for research to examine the safety and effectiveness of a new
drug, investigators are often given financial grants or other incentives (e.g., stock
options) for participating in the conduct of clinical trials of their products. While
FDA regulations require the disclosure of financial information for certain activities
(21 CFR Part 54, 21 CFR § 19.10), similar requirements applicable to study
investigators seeking to publish trial results or to IRB members are for the most part
lacking.
On February 28, 2005, Senator Christopher Dodd introduced S. 470, the Fair
Access to Clinical Trials (FACT) Act. This bill is similar to two bills introduced in
during the 108th Congress by Representative Edward Markey (H.R. 5252) and
Senator Dodd (S. 2933). The legislation seeks to address one aspect of improving
drug safety by requiring registration of all clinical trials of all FDA-regulated medical
products before the enrollment of human subjects, and the subsequent posting of their
results. The legislation would also impose requirements for the disclosure of
financial conflicts of interest for investigators.15 The broad scope of the proposed
15 Though FDA requires manufacturers to register their clinical trials with the government
[http://www.clinicaltrials.gov] less than half of trials are registered (see Sharon Vedantam,
“Drugmakers Prefer Silence on Test Data,” Washington Post, July 6, 2004, p. A1). A
centralized repository of clinical trial results does not exist. Medical journals tend to favor
publication of trials with positive results, making information from negative or inconclusive
trials difficult to find. FDA publishes summaries of safety and effectiveness for approved
(continued...)

---

CRS-7
legislation in the 108th Congress prompted the medical device industry to question
whether the remedy, drafted primarily in response to drug safety issues, applied to
some or all medical devices. Over time, differences in the nature of the drug and
device industries led to development of different models for FDA regulation of these
products. As the 109th Congress will likely continue to investigate FDA’s proper role
in protecting patient safety, it is important that policymakers be aware of these
differences, since they may wish to consider them in oversight or legislation.
Likewise, it is important to consider the appropriate balance between stringency and
leniency in premarket and postmarket review for both drugs and devices so that
oversight, legislation, or regulation can offer an optimum level of patient protection
without stifling product development or innovation.
History of Device Legislation
Though food has been regulated since early colonial times, and drugs since the
Drug Importation Act of 1848, medical devices did not come under federal scrutiny
until Congress passed the Federal Food, Drug and Cosmetic Act (FFDC) of 1938
(P.L. 75-717). At that time, few medical devices existed. In 1966, the Fair
Packaging and Labeling Act (P.L. 89-755) required all consumer products in
interstate commerce to be labeled accurately and truthfully, with FDA enforcing the
provisions on regulated medical products, including medical devices.
The Medical Device Amendments of 1976 (MDA; P.L. 94-295) was the first
major legislation passed to ensure safety and effectiveness of medical devices,
including diagnostic products, before they could be marketed. The amendments
required manufacturers to register with FDA and follow quality control procedures
in their manufacturing processes. Some products were required to undergo
premarket review by FDA, while others had to meet performance standards before
marketing. Devices already on the market in 1976 (“preamendment” or
“grandfathered” devices) did not have to undergo retrospective approval for
marketing. Instead, they were to be broadly classified by FDA into one of three
regulatory classes based on the risk they posed to the patient. Devices coming to
market after 1976 had to undergo pre-market review (unless they were exempt).
Devices could be “cleared” or “approved” by FDA either by demonstrating that they
were substantially equivalent to a preamendment device, or if the device (or its use)
were truly novel, by demonstrating that it was safe and effective on its own merit.
15 (...continued)
trials, but does not publish information about products that are not approved (or new uses
of approved products), or that are withdrawn by the manufacturer from FDA review. The
argument against publication of results for unapproved uses or withdrawn applications for
drugs is that the manufacturer may wish, after collecting more data, to resubmit the
application. In the device regulations, FDA can disclose non-trade secret information at any
time after making a decision as to whether to approve a product; however, they publish only
summaries of safety and effectiveness for approved products. Information regarding non-
approved or withdrawn applications may be requested under the Freedom of Information
Act (FOIA), but the process may be time-consuming. For more information, see CRS
Report RS21944, Clinical Trials Reporting and Publication, by Erin Williams and Susan
Thaul.

---

CRS-8
In 1990, the Safe Medical Devices Act (SMD Act; P.L. 101-629) established
postmarket requirements for medical devices. The SMD Act required facilities that
use medical devices to report to FDA any incident that suggested that a medical
device could have caused or contributed to the death, serious illness, or injury of a
patient. Manufacturers of certain permanently implanted devices were required to
establish methods for tracking the patients who received them and to conduct
postmarket surveillance to identify adverse events. The act authorized FDA to carry
out certain enforcement actions, such as device product recalls, for products that did
not comply with the law.
In 1997, the Food and Drug Administration Modernization Act (FDAMA; P.L.
105-115) mandated the most wide-ranging reforms in FDA practice since 1938. For
medical devices, provisions included measures to accelerate premarket review of
devices and to regulate company advertising of unapproved uses of approved devices.
In 2002, the Medical Device User Fee and Modernization Act (MDUFMA; P.L.
107-250) enacted three significant provisions for medical devices: (1) it established
user fees for premarket reviews of devices; (2) it allowed establishment inspections
to be conducted by accredited persons (third parties); and (3) it instituted new
regulatory requirements for reprocessed single-use devices.16 The additional funds
provided to FDA from user fees are intended to increase FDA’s technical and human
resources so that statutorily mandated deadlines may be met.
Medical Device Approval Process
Section 201(h) of the FFDC Act (21 USC § 321 (h)) defines a medical device
as
an instrument, apparatus, implement, machine, contrivance, implant, in vitro
reagent, or other similar or related article, including any component, part, or
accessory, which is —
(1) recognized in the official National Formulary, or the United States
Pharmacopeia, or any supplement to them,
(2) intended for use in the diagnosis of disease or other conditions, or
in the cure, mitigation, treatment, or prevention of disease, in man or
other animals, or
(3) intended to affect the structure or any function of the body of man
or other animals, and
which does not achieve its primary intended purposes through chemical action
within or on the body of man or other animals and which is not dependent upon
being metabolized for the achievement of its primary intended purposes.
Medical devices that are sold in the United States are regulated by two centers
within FDA: the Center for Devices and Radiological Health (CDRH) and the
Center for Biologics Evaluation and Research (CBER). CDRH is responsible for
regulating firms that manufacture, repackage, relabel, and/or import most medical
16 MDUFMA was amended in 2004 by the Medical Devices Technical Corrections Act
(MDTCA; P.L. 108-214) to clarify Congress’s intent and to improve and expand upon some
features of MDUFMA.

---

CRS-9
devices, including surgical instruments, implantable devices, diagnostic equipment,
clinical laboratory tests, and medical and non-medical radiation emitting electronic
products (such as lasers, x-ray systems, ultrasound equipment, microwave ovens, and
color televisions). CBER regulates medical devices involved in the collection,
processing, testing, manufacture and administration of licensed blood, blood
components, and cellular products, including HIV test kits used to screen donor
blood and cellular products, and to diagnose, treat, and monitor persons with HIV
and AIDS.
Premarket Clearance or Approval
In order to be marketed in the United States, a medical device must comply with
certain “controls” to ensure that they are used safely and effectively in the patients
for whom they were developed (i.e., the target population). The level of control, and
therefore the specific regulations that a manufacturer17 must follow, are determined
by the level of risk that the device poses to patients from its use or misuse.
Following the enactment of the MDA of 1976, FDA grouped devices that were
already on the market into 16 medical specialties referred to as panels.18 FDA then
established three risk classifications — Class I, II , and III — representing low-,
moderate- and high-risk categories. FDA classified in the Code of Federal
Regulations (CFR) approximately 1,700 different generic types of devices that were
already on the market. After they were classified in the CFR, these pre-1976 devices
could potentially serve as comparison devices in the premarket review of devices that
were brought to market after 1976.
Definitions. The following definitions are important in understanding the
various aspects of the medical device approval process:
Intended Use and Indications for Use. Often used interchangeably, the
intended use and indications for use provide the basis for risk classification, and
therefore the types of studies that are required to support approval or clearance of the
device, and the stringency of regulations with which the manufacturer will have to
comply. The intended use statement generally describes the device function or
physiological purpose (e.g., removes water from blood, cuts tissue, detects protein
in urine) and can include the indications for use (see 21 CFR § 807.92(a)(5)). The
indications for use include a general description of the disease or condition that the
device will diagnose, treat, prevent, cure, or mitigate (e.g., diabetes, stage III breast
cancer), a description of the patient population (e.g., patients with edema, juveniles,
17 The term “manufacturer” is used throughout this report for simplicity, but also includes
any person, organization, or sponsor that submits a marketing application to FDA for
marketing approval for a medical device.
18 The panels are found in 21 CFR Parts 862 through 892. They include clinical chemistry
and clinical toxicology devices, hematology and pathology devices; immunology and
microbiology devices; anesthesiology devices; cardiovascular devices; dental devices; ear,
nose and throat device; gastroenterology-urology devices; general and plastic surgery
devices; general hospital and personal use devices; neurological devices; obstetrical and
gynecological devices; ophthalmic devices; orthopedic devices; physical medicine devices;
and radiology devices.

---

CRS-10
women), any specific or special clinical circumstances (e.g., patients with blood
tumor marker levels greater than 4.0 ng/mL), clinical settings (e.g., clinical
laboratory, physician office, hospital), anatomical sites, or any other defining
information about how the device will be used (21 CFR § 814.20 (b)(3)(I)).
Predicate Device. The “predicate device” is any device that a manufacturer
references to obtain a clearance based on a determination of “substantial
equivalence” (see definition below). To be a predicate, the device must have either
been on the market before 1976, or it could have been cleared for marketing after
1976, but must have the same intended use as a device classified in the CFR. For
example, a device that measures glucose from blood to test for diabetes was sold in
1972. After the MDA of 1976, these devices were determined to be Class II and can
be found in the CFR:
21 CFR 862.1345 Glucose test system.
(a) Identification. A glucose test system is a device intended to measure glucose
quantitatively in blood and other body fluids. Glucose measurements are used in
the diagnosis and treatment of carbohydrate metabolism disorders including
diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and of
pancreatic islet cell carcinoma.
(b) Classification. Class II.
If a company wants to sell a new device to measure glucose in 2005, it is
required to seek FDA clearance. As long as the new device has the same intended
use as the older device (i.e., “... intended to measure glucose quantitatively in blood
and other body fluids. Glucose measurements are used in the diagnosis and treatment
of carbohydrate metabolism disorders including diabetes mellitus, neonatal
hypoglycemia, and idiopathic hypoglycemia, and of pancreatic islet cell carcinoma”),
the manufacturer of the new device can base its marketing application on data that
compares the new device with the old device, rather than conducting a new, large-
scale clinical study.
Substantial Equivalence. Substantial equivalence is the standard of
approval for most non-exempt (i.e., from premarket review) low- to moderate-risk
devices.19 Substantial equivalence is determined based on a comparison of the
performance characteristics of the new device with a predicate device. In making a
determination of substantial equivalence, the new device must either have the same
intended use and technological characteristics as the predicate device, or have
different technological characteristics that do not raise new questions of safety and
effectiveness.
The manufacturer decides what predicate device it would like to use in
comparison with its new device. However, FDA has the ultimate discretion in
determining whether the comparison is appropriate. Devices that are found to be
19 Many low-risk devices can be deemed “exempt” from premarket review. This means that
the manufacturer does not have to file a premarket notification (i.e., a 510(k)) or a premarket
application (PMA) with FDA demonstrating that the device is safe and effective.
Manufacturers of exempt devices are still required to comply with other regulations, known
as general controls.

---

CRS-11
substantially equivalent to another device receive FDA “clearance” rather than
“approval.”
Safe and Effective. Safety and effectiveness is a higher standard than
substantial equivalence. The evidence required to meet the standard may vary
according to the characteristics of the device, its conditions of use, the existence and
adequacy of warnings and other restrictions, and the extent of experience with its use
(21 CFR § 860.7(c(2)). FDA considers there to be reasonable assurance of safety
when it can be determined that the probable benefits to health that result from use of
the device as directed by the manufacturer outweigh any probable risks (21 CFR
§ 860.7(d)(1)). Investigations for safety can include animal studies, human studies,
and/or non-clinical in vitro studies (21 CFR § 860.7(d)(2)). FDA considers there to
be reasonable assurance of effectiveness when, based upon valid scientific evidence,
the use of the device in the target population according to the manufacturer’s
instructions will provide clinically significant results (21 CFR § 860.7(e)(1)). Valid
scientific evidence includes evidence from well-controlled clinical trials, other
carefully defined clinical investigations, well-documented case histories, and reports
of significant human experience. The evidence can be collected by the manufacturer
or a representative, and/or can be abstracted from the medical literature. Devices that
meet the standard of “safe and effective” receive FDA “approval.”
Device Classification. Medical devices are classified into a regulatory
category based on the risk that is posed to the patient with its use or misuse. A new
device can be classified based on comparison with a legally marketed device, by
regulation, or by a classification panel. If unsure what classification a product would
receive, a manufacturer may make a formal request for classification to FDA, to
which FDA must respond within 60 days.
Class I. Devices in Class I are those for which general controls (described
below) alone are sufficient to ensure safe and effective use of the product (21 CFR
§ 860.3(c)(1)). Many Class I devices are exempt from premarket review by FDA,
because FDA previously determined that they present low risk of illness or injury to
a patient (see 21 CFR Parts 862 to 892). Although a manufacturer of an exempt
device does not have to file a submission containing performance or other data with
FDA before marketing the device (as manufacturers of nonexempt devices do), it still
has to comply with the remaining general controls.
General controls are the minimum level of regulation that applies to all FDA
regulated medical devices.20 Manufacturers:
! are prohibited from selling an adulterated product,21
20 FDA, General Controls for Medical Devices, Apr. 16, 1998, at [http://www.fda.gov/
cdrh/devadvice/363.html].
21 A device is adulterated if it includes any filthy, putrid, or decomposed substance, or if it
is prepared, packed, or held under unsanitary conditions. The FDC Act further states that
a device is adulterated if its container contains any poisonous or deleterious substance, or
if its strength, purity or quality varies significantly from what the manufacturer claims. For
(continued...)

---

CRS-12
! are prohibited from misbranding a product,22
! must register their facility with FDA and list all of the medical
devices that they produce or process,
! must file the appropriate premarket submission with the agency at
least 90 days before introducing a non-exempt device onto the
market, and
! must report to FDA any incident that they are aware of that suggests
that their device may have caused or contributed to a death or
serious injury.
Class II. Devices in Class II are those for which general controls alone are not
sufficient to provide reasonable assurance of safety and effectiveness. Class II
includes devices that pose a moderate risk to patients, and may include new devices
for which information or “special controls” are available that will reduce or mitigate
the risk. Most Class II devices require premarket review; however, some are exempt
by regulation (21 CFR § 860.3(c)(2)).
Special controls can include any requirement that FDA deems necessary to
assure safe and effective use of a medical device: for example, performance
standards, postmarket surveillance requirements, patient registries, or the
development and dissemination of guidance documents. Special controls guidance
documents often contain information for the device manufacturer on FDA’s current
thinking of best practices to assure safety and effectiveness of a particular type of
device. These include, but are not limited to, guidance on the types of studies that
that the manufacturer could perform to support an application for marketing,
guidelines about labeling for use in a particular target population, or suggestions for
quality control procedures or manufacturing practice.
Class III. Class III medical devices include those for which general and/or
special controls are not sufficient to assure safe and effective use of the device. Class
III includes devices which are life-supporting or life-sustaining, and devices which
present a high or potentially unreasonable risk of illness or injury to a patient. New
devices which are not classified as Class I or II by another means, are automatically
designated as Class III unless the manufacturer files a request or petition for
reclassification under Section 513(f)(2) of the FFDC Act (See also 21 CFR
§ 860.3(c)(3)).
FDAMA gave FDA the authority to establish procedures for meeting with
manufacturers prior to preparing a submission.23 The procedures aim to speed the
21 (...continued)
higher class devices, a device can be considered adulterated if it fails to meet performance
requirements outlined in its approval, or if it is in violation of other Good Manufacturing
Practice requirements.
22 A device is misbranded when all or part of the labeling (i.e., the FDA-approved printed
material providing information about the device) is false, misleading, or missing.
23 For guidance on the procedures established, see Early Collaboration Meetings Under the
FDA Modernization Act; Final Guidance for Industry and CDRH Staff, Feb. 28, 2001, at
(continued...)

---

CRS-13
review process by giving FDA and a manufacturer the opportunity to address
questions and concerns about the device and/or the planned studies that will be used
to support the marketing application before the studies are initiated and the
application is submitted. For example, the “pre-IDE” process is an informal
“pre-submission” process. The “pre-IDE” process is so-called in name only;
submitting a pre-IDE does not mean that manufacturers are required to submit
subsequently an investigational device exemption (IDE) application.24 The pre-IDE
process is simply a means for FDA and industry to engage in dialogue about a new
device, before a study is initiated or a marketing application is submitted. The pre-
IDE process may involve sending analytical or clinical protocols to FDA for review
and comment before proceeding with studies, or meeting with FDA to discuss
protocols and/or possible regulatory pathways. This particular process is strictly
voluntary, and not binding on either FDA or industry. The benefits to manufacturers
include an opportunity to begin a dialogue with FDA, to promote greater
understanding of new technologies, to reduce the cost of research studies by focusing
on the important information needed for FDA approval (or clearance), eliminating
unnecessary or burdensome studies, and to speed the review process for the future
marketing application since FDA will already be familiar with the device.
Marketing Applications for Medical Devices. Prior to marketing a
medical device in the United States, a manufacturer must register their facility with
the FDA and list the devices that they commercialize. The classification of the
device (Class I, II or III based on the risk posed to the patient) determines whether or
not a premarket submission is required, and if so, what type of submission is
required. The following sections describe the types of premarket submissions that
FDA reviews for medical devices.
Premarket Notification (510(k)). A 510(k) submission is required for any
new, non-exempt low- or moderate-risk medical device that will be marketed in the
United States. The standard for clearance of a traditional 510(k) is substantial
equivalence with a predicate device. Though usually for Class I or II devices, an
older, preamendment Class III device may sometimes use a 510(k) submission. A
510(k) could also be used for currently marketed devices for which the manufacturer
seeks a new indication (e.g., a new population, such as pediatric use, or a new disease
or condition), or for which the manufacturer has changed the design or technical
characteristics such that the change may affect the performance characteristics of the
device.
There are several types of 510(k)s: traditional, abbreviated, special and de novo.
In a traditional 510(k), the manufacturer submits information about the performance
of the device under specific conditions of use. It also contains information about the
design of the device, characteristics of device components, representations of
packaging and labeling, a description and summary of the non-clinical and clinical
studies that were done to support the device performance characteristics, a
23 (...continued)
[http://www.fda.gov/cdrh/ode/guidance/310.pdf].
24 For more information on the IDE, please refer to the “Marketing Applications” section
that immediately follows this discussion.

---

CRS-14
description of means by which users can assess the quality of the device, and
information about any computer software or additional or special equipment needed.
Several administrative forms are also required.25 Most of the studies supporting a
510(k) submission are not true clinical studies. While FDA prefers to see data on
performance of the device in the actual intended population, substantial equivalence
in many cases, means only that the device performs in a similar fashion to the
predicate under a similar set of circumstances. As a result, many devices never have
to demonstrate safety and effectiveness through clinical studies.
FDA may take any of the following actions on a 510(k) after conducting its
review (21 CFR § 807.100(a)): find the device substantially equivalent to the
predicate and issue a clearance letter, find the device not substantially equivalent
(NSE) and issue an NSE letter prohibiting marketing, or request additional
information (with the final clearance decision pending review of that information).
A manufacturer generally has 30 days to provide any additional information, or the
FDA may issue a notice of withdrawal of the application (21 CFR § 807.87(l)). The
manufacturer may, at any time, withdraw its 510(k). FDA has 90 days to review a
traditional 510(k).26
Abbreviated and special 510(k)s were new approaches to premarket notification
that came from FDAMA of 1997, intended to streamline and expedite FDA’s review
for routine submissions meeting certain qualifications, thus leaving more reviewer
time for more complicated submissions. An abbreviated 510(k) uses guidance
documents developed by FDA to communicate regulatory and scientific expectations
to industry. Guidance documents have been prepared for many different kinds of
devices, and are available on FDA’s website. All guidance documents are developed
in accordance with Good Guidance Practices (GGP, 21 CFR § 10.115)), and many
with public participation or opportunities for public comment.27 In addition, FDA
can either develop performance or consensus standards or ‘recognize’ those
developed by outside parties (21 CFR Part 861). In an abbreviated 510(k), the
manufacturer describes what guidance document, special control, or performance
standard was used, and how it was used to assess performance of their device. Other
minimum required elements are the product description, representative labeling, and
a summary of the performance characteristics. FDA typically reviews an abbreviated
510(k) in 60 days.
The Quality System Regulation (QSR; 21 CFR § 820.30) is the regulation that
describes the good manufacturing practice (GMP) requirements for medical devices
(see the Manufacturing section of this report for more detail on QSR). A special
25 FDA, “Content of a 510(k),” June 30, 2003, at [http://www.fda.gov/cdrh/devadvice/
314312.html], and How to Prepare a Traditional 510(k), Jan. 4, 2005, at
[http://www.fda.gov/cdrh/devadvice/3143.html#link_4].
26 The FDA time clock (i.e., review cycle) begins when FDA receives the 510(k) and ends
with the date that FDA issues either a request for additional information or a decision. More
than one cycle may occur before FDA issues its final decision.
27 FDA continually accepts public comment on any draft or final guidance document.

---

CRS-15
510(k) utilizes the design control28 requirement of the QSR and may be used for a
modification to a device that has already been cleared. The modifications should not
affect the safety and effectiveness of the device. The special 510(k) allows the
manufacturer to declare conformance to design controls, without providing the data.
This type of submission references the original 510(k) number, and contains
information about the design control requirements. FDA aims to review most special
510(k)s in 30 days.
Under the FFDC Act, first-of-a-kind devices lacking a legally marketed
predicate would automatically be designated Class III. FDAMA amended Section
513(f) to allow FDA to establish a new, expedited mechanism for reclassifying these
devices based on risk, thus reducing the regulatory burden on manufacturers. The de
novo 510(k), though requiring more data than a traditional 510(k), often requires less
information than a premarket application (PMA).
In a de novo 510(k) process, the manufacturer submits a traditional 510(k) for
its device. However, because there is no predicate device or classification, the
agency will return a decision of not substantially equivalent. Within 30 days, the
manufacturer submits a petition requesting reclassification of its device into Class II
or I, as appropriate. Within 60 days, FDA will render a decision classifying the
device according to criteria in 513(a)(1) of FFDC Act. With approval, the FDA
issues a regulation that classifies the device. If the device is Class II, a special
controls guidance document is also developed that then allows subsequent
manufacturers to submit either traditional or abbreviated 510(k)s.29
In 2002, the MDUFMA authorized FDA to charge a fee for premarket reviews,
including non-exempt, non-waived 510(k)s. For FY2005, the standard fee is $3,502
and $2,802 for small businesses. FDA will not file an application unless the fee is
paid at the time of submission.
Premarket Application (PMA). A PMA is the most stringent type of device
marketing application required by FDA for new and/or high-risk devices. PMA
approval is based on a determination by FDA that the application contains sufficient
valid scientific evidence to assure that the device is safe and effective for its intended
use(s) (21 CFR Part 814). A PMA will contain the following information (amongst
other things): administrative requirements, summaries of non-clinical and clinical
data supporting the intended use and performance characteristics, detailed
information on the design of the device and a description of the device components,
instructions for use, representations of packaging and labeling, a description of means
28 Design controls are a series of pre-determined checks, verifications, and specifications
that are built into the manufacturing process to validate the quality of the product throughout
the process. These can include: defining the personnel responsible for implementing steps
in the development and manufacturing process, defining specifications and standards for
assessing the quality of the materials that go into making the product, designing
specifications for accepting and rejecting different batches or lots of final product,
requirements for maintaining appropriate records, etc.
29 Guidance for Industry and FDA Staff, New Section 513(f)(2) — Evaluation of Automatic
Class III Designation, Feb. 19, 1998, at [http://www.fda.gov/cdrh/modact/clasiii.pdf].

---

CRS-16
by which users can assess the quality of the device, information about any computer
software or additional or special equipment needed, literature about the disease and
the similar devices, information on the manufacturing process, and assurance of
compliance with QSR. In contrast to a 510(k), PMAs generally require some clinical
data, but can also use studies from the medical literature (a “paper PMA”). Approval
is based not only on the strength of the scientific data, but also on inspection of the
manufacturing facility to assure that the facility and the manufacturing process are
in compliance with the quality systems regulations (QSR: 21 CFR Part 820).
FDAMA made it easier for manufacturers to submit the required sections of a PMA
in a serial fashion as data are available (“modular PMA”).
When a PMA is first received, FDA has 45 days to make sure the application
is administratively complete. If so, the FDA formally files the application. If not, the
application is returned. FDA then has 75 days to complete the initial review and
determine whether an advisory panel meeting will be necessary. Advisory
committees, comprised of scientific, medical, and statistical experts, and industry and
consumer representatives, can be convened to make recommendations on any
scientific or policy matter before FDA. They allow for interested persons to present
information and views at an oral public hearing before the advisory committee (21
CFR Part 14). FDA typically accepts advisory committee recommendations for an
application (approvable, approvable with conditions or non-approvable); however,
there have been cases where the decision has not been consistent with the
recommendation (e.g., where the conditions for approval are so burdensome as to
practically present a non-approvable situation). CDRH will hold joint advisory
committee meetings with other centers where necessary. After FDA notifies the
applicant that the PMA has been approved or denied, a notice may be published on
the Internet (1) announcing the data on which the decision is based, and (2) providing
interested persons an opportunity to petition FDA within 30 days for reconsideration
of the decision.
Though FDA regulations allow 180 days to review the PMA and make a
determination (21 CFR § 814.40), in reality the review time could be much longer.
MDUFMA established performance goals to reduce the review time for PMAs.30
The FY2005 fees have been set at $239,237 for a standard PMA ($90,910 for small
business).
510(k), and PMA Supplements. Once a device has been cleared through
a 510(k) process or approved through the PMA process, the manufacturer can market
the device only for the intended use that the FDA cleared or approved. For example,
a device, such as a stent, approved to treat coronary artery disease may not be
marketed for treatment of blocked biliary ducts unless the manufacturer files
additional information with FDA to demonstrate that the device is safe and effective
for the new use. The information can be filed as a supplement to the original
application. Supplements are required not only for new uses of a cleared or approved
device, but also for design or manufacturing changes that may impact safety and
30 “FDA Announces New Goal for Reducing Total Device Review Times,” FDA News, Aug.
5, 2003, at [http://www.fda.gov/bbs/topics/NEWS/2003/NEW00929.html].

---

CRS-17
effectiveness (e.g., changing the type of metal or plastic on a device, using a different
antibody for diagnosis of a disease, etc.).
Fees for 510(k) supplements are the same as for the original application. If a
PMA supplement contains controversial or new information, it may be subject to
advisory committee review just like an original application. In that case, the user fees
are the same as the original. User fees to review other supplements are: $51,436 for
a traditional PMA supplement with a 180-day review time ($19,546 small business),
and $17,225 for a real-time supplement ($6,546 for small business).31
Investigational Device Exemption (IDE). An IDE allows an unapproved
device (most commonly an invasive or life-sustaining device) to be used in a clinical
study to collect data required to support a submission, most commonly a PMA, at
some later point in time.32 Investigational use can also include clinical evaluation of
certain modifications to or new intended uses of legally marketed devices (e.g.,
supplemental application). All clinical evaluations of investigational devices, unless
they are exempt, must have an IDE and be approved by an IRB, before the study is
initiated. The IDE permits a device to be shipped lawfully for investigation of the
device without requiring that the manufacturer comply with other requirements of the
FFDC Act, such as registration and listing. Manufacturers of devices with IDE’s are
also exempt from the quality systems regulations (QSR), except for the requirements
for design control.
While under investigation, manufacturers, sponsors, clinical investigators and
IRBs must comply with Good Clinical Practices, including all regulations that govern
the conduct of clinical studies:
! Investigational Device Exemptions (21 CFR Part 812) covering the
procedures for the conduct of clinical studies with medical devices
including application, responsibilities of sponsors and investigators,
labeling, records, and reports;
! Protection of Human Subjects (21 CFR Part 50) providing the
requirements and general elements of informed consent;
! Institutional Review Boards (21 CFR Part 56) covering the
procedures and responsibilities for institutional review boards
(IRBs) that approve clinical investigations protocols;
! Financial Disclosure by Clinical Investigators (21 CFR Part 54)
covering the disclosure of financial compensation to clinical
investigators which is part of FDA’s assessment of the reliability of
the clinical data; and
31 A “real time” supplement is “a supplement to an approved PMA or premarket report under
Section 515 of the FFDC Act that requests a minor change to the device, such as a minor
change to the design of the device, software, manufacturing, sterilization, or labeling, and
for which the applicant has requested and the agency has granted a meeting or similar forum
to jointly review and determine the status of the supplement.” The review time is generally
quicker (because the change is minor) compared to other types of supplements.
32 FDA, IDE Overview, July 8, 2003, at [http://www.fda.gov/cdrh/devadvice/ide/
index.shtml].

---

CRS-18
! Design Controls of the QSR (21 CFR Part 820 Subpart C) providing
the requirement for procedures to control the design of the device in
order to ensure that the specified design requirements are met.
Devices are exempt from IDE requirements when: testing is noninvasive, testing
does not require invasive sampling, testing does not introduce energy into a subject,
testing is not stand alone (i.e., is not used for diagnosis without confirmation by other
methods or medically established procedure) (21 CFR § 812.2(c)(3)).
Humanitarian Device Exemption (HDE). An HDE is an application that
is similar to a PMA, but exempt from the effectiveness requirements. An approved
HDE authorizes marketing of a humanitarian use device. A humanitarian use device
is intended to benefit patients in the treatment and diagnosis of diseases or conditions
that affect fewer than 4,000 individuals in the U.S. per year.
Before submitting an HDE application, the manufacturer submits a request for
a humanitarian use device designation to FDA’s Office of Orphan Products
Development (OOPD). The request includes: (1) a statement that they are requesting
a humanitarian use device designation for a rare disease or condition, (2) the name
and address of the manufacturer, (3) a description of the rare disease or condition for
which the device is to be used, (4) a description of the device, and (5) documentation,
with appended authoritative references, to demonstrate that the device is designed to
treat or diagnose a disease or condition that affects or is manifested in fewer than
4,000 people in the United States per year (see 21 CFR § 814.102(a)). In order for
a device to receive marketing approval under this regulation, there should not be
another legally marketed device available to treat or diagnose the disease or
condition. Once a device with the same intended use as the humanitarian use device
is approved or cleared, an HDE cannot be granted for the humanitarian use device.
The agency has 75 days from the date of receipt to review an HDE application.
This includes a 30-day filing period during which the agency determines whether the
HDE application is sufficiently complete to permit substantive review. FDA does
require that a manufacturer comply with the QSR that the agency deems most
relevant to the safety of the device. Alternatively, the manufacturer can request an
exemption. Supplements, and sometimes even a new HDE, are required for
additional indications.33
Table 1 shows the number of marketing applications that CDRH received,
approved and denied in 2003. Very few applications are actually denied approval.
Instead, an application may be deemed “not approvable” in its current form (i.e.,
typically because the data that the manufacturer provides are not sufficient to
demonstrate safety and effectiveness). Applications that are “not approvable” are
usually withdrawn by the manufacturer before a denial is rendered.
33 CDRH, FDA, Humanitarian Device Exemptions Regulation: Questions and Answers;
F i n a l G u i d a n c e f o r I n d u s t r y , J u l y 1 2 , 2 0 0 1 , a t
[http://www.fda.gov/cdrh/ode/guidance/1381.html].

---

CRS-19
Table 1. Number of Marketing Applications for Medical Devices
Received, Approved and Denied by CDRH in 2003
Number of applications
CDRH
Received
Original PMA
54
PMA Supplements
669
510(k)
4,247
Original IDE
242
Original HDE
10
Completeda
Original PMA
31
PMA Supplements
494
510(k)
4,132
Original IDE
246
Original HDE
2
Approvals + Approvable (number of decisions made)b
Original PMA
31 + 16 (57)
PMA Supplements
494 + 94 (635)
510(k): decision is substantially equivalent
3,522 (4,132)
Original IDE: “approvable” not a valid decision
146 (246)
Original HDE
2 + 0 (2)
Denials + Not Approvable
(Number of Decisions)
Original PMA
0 + 10 (57)
PMA Supplements
0 + 47 (635)
510(k): decision is not substantially equivalent
88
(4,132)
Original IDE: “not approvable” not a valid decision
78 (246)
Original HDE
0 + 0 (2)
Source: ODE and OIVD, Annual Report: Fiscal Year 2003, at [http://www.fda.gov/cdrh/annual/
fy2003/ode/2003.pdf].
a. The number completed is different than the number received because some applications were
received in 2002 and completed in 2003, others received in 2003 may not have been completed
in that year.
b. The number of decisions made can be greater than the number of applications completed because
in FDA accounting, an application could have more than one decision. For example, it could
have been “approvable” early in the year and later “approved.” An approval means that a
product can be marketed, approvable means that the manufacturer had to meet certain conditions
before the device could be marketed. Presumably, many applications that are determined to be
“not approvable” are subsequently withdrawn from FDA consideration by the manufacturer
prior to receiving the denial.
In vitro Diagnostic products (IVD). IVD products are “... those reagents,
instruments, and systems intended for use in diagnosis of disease or other conditions,
including a determination of the state of health, in order to cure, mitigate, treat, or
prevent disease or its sequelae. Such products are intended for use in the collection,
preparation, and examination of specimens taken from the human body (21 CFR

---

CRS-20
§ 809.3). IVDs are medical devices as defined in Section 210(h) of the FFDC Act,
and may also be biological products subject to Section 351 of the Public Health
Service Act. IVDs are different from other medical devices in that they do not act
directly on a patient to produce a result like an implantable, life-sustaining or other
device does. Instead, the risk to the patient is from the generation of inaccurate test
results (i.e., wrong answers) that lead to mismanagement of a patient’s condition.
IVDs (e.g., laboratory tests) may consist of general purpose reagents,34 analyte
specific reagents,35 general purpose or specific equipment, sometimes with computer
analysis software. Most stand-alone items of general purpose equipment, such as
automated clinical analyzers, are exempt Class I devices. However, if the equipment
performs a specific test, equipment plus the test becomes a test system. Test systems
are considered combination devices, and they are classified according to the risk level
of the highest of the two device classifications (i.e., an analyzer may be Class I
exempt, but if a manufacturer wishes to market it with an HIV test kit, the system
could be regulated as a Class III device and require a PMA). Most IVD products are
reviewed in CDRH’s Office of In Vitro Diagnostic Device Evaluation and Safety
(OIVD), and CBER’s Office of Blood Research and Review (OBRR). The
classification of existing IVDs can be found in 21 CFR Part 862, 21 CFR Part 864,
and 21 CFR Part 866.
Like other medical devices, IVDs are subject to premarket and postmarket
controls. IVDs are also subject to the Clinical Laboratory Improvement Amendments
(CLIA) of 1988.36 CLIA establishes quality standards for laboratory testing and an
accreditation program for clinical laboratories that perform testing using IVD
products. CLIA requirements vary according to the technical complexity in the
testing process and risk of harm in reporting erroneous results. The regulations
establish three categories of testing on the basis of the complexity of the testing
methodology: (a) waived tests, (b) tests of moderate complexity, and (c) tests of high
complexity. Manufacturers apply for CLIA categorization (determined by FDA)
during the premarket process. Under CLIA, laboratories performing only waived
tests are subject to minimal regulation. Laboratories performing moderate or high
complexity tests are subject to specific laboratory standards governing certification,
personnel, proficiency testing, patient test management, quality assurance, quality
control, and inspections.
34 A general purpose reagent is “a chemical reagent that has general laboratory application,
is used to collect, prepare, and examine specimens from the human body for diagnostic
purposes, and is not labeled or otherwise intended for a specific diagnostic application …
[General purpose reagents] do not include laboratory machinery, automated or powered
systems (21 CFR § 864.4010)).”
35 Analyte specific reagents (ASRs) are “antibodies, both polyclonal and monoclonal,
specific receptor proteins, ligands, nucleic acid sequences, and similar reagents which,
through specific binding or chemical reaction with substances in a specimen, are intended
for use in a diagnostic application for identification and quantification of an individual
chemical substance or ligand in biological specimens (21 CFR § 864. 4020(a)).”
36 FDA, Overview of IVD Regulation, Mar. 4, 2003, at [http://www.fda.gov/cdrh/
oivd/regulatory-overview.html#1].

---

CRS-21
CLIA categorization (defining regulatory requirements on the laboratory testing
process) do not always match FDA classification (defining regulatory requirements
on the tests kits or systems themselves). For example, a “waived” test under CLIA
is the simplest test to perform (usually by an untrained user), with the smallest
margin of error. As such, they receive little to no oversight under CLIA. However,
FDA may designate such a test as Class III, so that it undergoes rigorous review to
insure that it performs as advertised (i.e., with a small margin of error in the hands
of an untrained user). Most IVDs are exempt from IDE requirements.
Because the benefits and risks to the patient from use of IVDs are indirect (i.e.,
due to the use of the test result in patient management), FDA requires that the
companies demonstrate analytical test performance in patient samples that would test
along the continuum of positive and negative for the marker of interest. In addition,
FDA requires support for the clinical validity of the test (i.e., evidence that the
biological marker that the test is detecting actually is associated with the disease or
condition that the company wishes to market the test for in a predictable way). For
the many applications seeking clearance for an IVD, biological markers that the test
purports to measure may be relatively well characterized with respect to a disease and
patient population (such as the link between glucose measurement and diabetes). In
these cases, analytical studies using clinically derived samples (e.g., blood
specimens from healthy and diabetic individuals) suffice to show that the test is
actually detecting the marker. Sometimes, clinical samples can be supplemented by
carefully selected artificial samples, particularly if a disease, condition or marker is
rare. For example, if FDA were to review a genetic test to measure genetic markers
for drug metabolism, they may require the manufacturer to use actual patient samples
to demonstrate that they can detect common markers. FDA may, however, allow the
company to use artificial or ‘spiked’ samples to test for rare markers so that the
company would not have to test an overly burdensome number of clinical samples.
In this type of submission, the manufacturer could use medical literature to support
the clinical validity of the biological marker to the disease, and would not have to
conduct a clinical study to demonstrate that the test measures the marker and the
marker is associated with the disease.
For other IVDs, the link between analytical performance of the test in its ability
to detect a biological marker and the clinical validity of the marker is not well
defined. In these circumstances, new clinical information may be required. FDA
rarely requires prospective clinical studies for IVDs, but regularly requests clinical
samples with sufficient laboratory and/or clinical characterization to allow an
assessment of the clinical validity of a new device. For example, a company seeking
to market a test for a new tumor marker may use well-characterized, archived patient
samples collected as part of a completely separate study to demonstrate that their test
can classify patients in a predictable way. Clinical performance is usually expressed
in terms of clinical sensitivity and clinical specificity (when compared to a disease
or health state) or agreement (when compared to performance of a predicate device
or reference method). For most PMAs, manufacturers identify surrogate endpoints
(such as tumor shrinking or reduction in a tumor marker) and establish the device
performance in relation to those rather than to disease outcome (such as improved
survival).

---

CRS-22
Post-Approval Requirements and Issues
Once approved or cleared for marketing, manufacturers of medical devices must
comply with regulations on labeling and advertising, on production of their device,
and on postmarket surveillance of adverse events associated with the use of their
device.
Labeling. Like drugs and biological products, all FDA approved or cleared
medical devices are required to be labeled in a way that informs a user of how to use
the device in a safe and effective manner. Section 201(k) of the FFDC Act defines
a “label” as a: “display of written, printed, or graphic matter upon the immediate
container of any article.” Section 201(m) defines “labeling” as: “all labels and other
written, printed, or graphic matter upon any article or any of its containers or
wrappers, or accompanying such article” at any time while a device is held for sale
after shipment or delivery for shipment in interstate commerce. The term
“accompanying” is interpreted to mean more than physical association with the
product; it extends to posters, tags, pamphlets, circulars, booklets, brochures,
instruction books, direction sheets, fillers, webpages, etc. “Accompanying can also
include labeling that is connected with the device after shipment or delivery for
shipment in interstate commerce. According to an appellate court decision “Most,
if not all advertising, is labeling. The term ‘labeling’ is defined in the FFDC Act as
including all printed matter accompanying any article. Congress did not, and we
cannot, exclude from the definition printed matter which constitutes advertising.”
Labeling regulations pertaining to medical devices are found in the following
parts of Title 21 CFR:

! General Device Labeling (21 CFR Part 801)
! In Vitro Diagnostic Products (21 CFR Part 809)
! Investigational Device Exemptions (21 CFR Part 812)
! Good Manufacturing Practices (21 CFR Part 820)
! General Electronic Products (21 CFR Part 1010)
All devices must conform to the general labeling requirements. Certain
devices37 require specific labeling which may include not only package labeling, but
informational literature, patient release forms, performance testing, and/or specific
tolerances or prohibitions on certain ingredients.
Various sections of the QSR have an impact on labeling: Section 21 CFR
§ 820.80(b) requires the inspection and testing of incoming materials including
labeling; and 21 CFR § 820.70(f) requires buildings to be of suitable design and have
sufficient space for packaging and labeling operations; 21 CFR § 820.120 deals with
specific requirements for the control of labeling. This regulation applies to the
37 21 CFR §§ 801.405 to 801.437. Denture repair kits, impact resistant lenses in sunglasses
and eyeglasses, ozone emission levels, chlorofluorocarbon propellants, hearing aids,
menstrual tampons, chlorofluorocarbons or other ozone depleting substances, latex
condoms, and devices containing natural rubber.

---

CRS-23
application of labeling to ensure legibility under normal conditions of use over the
expected life of the device; and also applies to inspection, handling, storage, and
distribution of labeling. FDA considers a device to be adulterated if these
requirements are not met. These requirements do not apply to the adequacy of
labeling content, except to make sure the content meets labeling specifications
contained in the device master record. However, failure to comply with GMP
requirements, such as proofreading and change control, could result in labeling
content errors. In such cases, the device could be misbranded and/or adulterated.
Manufacturing. Like drugs and devices, medical device manufacturers must
produce their devices in accordance with Good Manufacturing Practice (GMP). The
GMP requirements for devices are described in the QSR, (Section 520 of the FFDC
Act; Part 820 of 21 CFR). The QSRs require that domestic or foreign manufacturers
have a quality system for the design, manufacture, packaging, labeling, storage,
installation, and servicing of non-exempt finished medical devices intended for
commercial distribution in the United States The regulation requires that various
specifications and controls be established for devices; that devices be designed and
manufactured under a quality system to meet these specifications; that finished
devices meet these specifications; that devices be correctly installed, checked and
serviced; that quality data be analyzed to identify and correct quality problems; and
that complaints be processed. FDA monitors device problem data and inspects the
operations and records of device developers and manufacturers to determine
compliance with the GMP requirements.38
Though FDA has identified in QSR the essential elements that a quality system
should have, manufacturers have a great deal of leeway to design quality systems that
best cover nuances of their devices and the means of producing them.
Postmarket Surveillance. Once their device is approved or cleared,
manufacturers must conduct postmarket surveillance studies to gather safety and
efficacy data for certain devices introduced into interstate commerce after January 1,
1991. This requirement applies to devices that:
! are permanent implants, the failure of which may cause serious
adverse health consequences or death;
! are intended for use in supporting or sustaining human life; or
! present a potential serious risk to human health.
FDA may require postmarket surveillance for other devices if deemed necessary to
protect the public health. The primary objective of postmarket surveillance is to
study the performance of the device after clearance or approval as it is used in the
population for which it is intended — and to discover cases of device failure and its
attendant impact on the patient.
Manufacturers may receive notification that their device is subject to postmarket
surveillance when FDA files (i.e., accepts) the submission, and again when a final
38 FDA, Good Manufacturing Practice/Quality Systems,at [http://www.fda.gov/cdrh/
comp/gmp.html].

---

CRS-24
decision is made. If notified, manufacturers must submit a plan for postmarket
surveillance to FDA for approval within 30 days of introducing their device into
interstate commerce.
MDUFMA authorized additional appropriations for postmarket surveillance —
$3 million for FY2003, $6 million for FY2004, and “such sums as may be necessary”
in subsequent years. For FDA to receive these resources in subsequent years, it must
submit to Congress by January 10, 2007, a study of:
! the effect of medical device user fees on its ability to conduct
postmarket surveillance;
! the extent to which device companies comply with postmarket
surveillance requirements;
! any improvements needed for adequate postmarket surveillance, and
the amount of funds needed to do so;
! recommendations as to whether, and in what amount, user fees
should be used for postmarket surveillance, if extended beyond
FY2007.39
Adverse Event Reporting. Section 519(a) of the FFDC Act as amended by
the SMDA of 1990 required FDA to establish a system for monitoring and tracking
serious adverse events that resulted from the use or misuse of medical devices. The
Medical Device Reporting (MDR) regulation is the mechanism that FDA and
manufacturers use to identify and monitor significant adverse events involving
medical devices, so that problems are detected and corrected in a timely manner.
User facilities (e.g., hospitals, nursing homes, clinical laboratories) are required to
report suspected medical device related deaths to both the FDA and the
manufacturers within 10 working days. User facilities may report medical device
related serious injuries only to the manufacturer within 10 days. Manufacturers must
file a summary of all medical device reports to FDA within 30 calendar days. User
facilities must file a summary report annually. Although the FFDC Act gives FDA
the authority to impose legal sanctions for not complying with MDR, FDA relies
largely on the goodwill and cooperation of all affected groups to accomplish the
objectives of the regulation. The searchable MDR database for devices is publically
accessible at [http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmdr/search.CFM].
Medical Device Tracking. Manufacturers must adopt a method of tracking
certain devices. These are devices:
! whose failure would be reasonably likely to have serious, adverse
health consequences; or
! which are intended to be implanted in the human body for more than
one year; or
! which are life-sustaining or life-supporting devices used outside of
a device user facility (21 CFR Part 821).
39 FDA, The Medical Device User Fee and Modernization Act of 2002 — “FAQs,”Aug. 1,
2003, at [http://www.fda.gov/cdrh/mdufma/faqs.html#postmarket].

---

CRS-25
The purpose of device tracking is to ensure that manufacturers of these devices can
locate them quickly once in commercial distribution if needed to facilitate
notifications and recalls in the case of serious risks to health presented by the
devices.40 A current list of the devices for which tracking is required can be found
at [http://www.fda.gov/cdrh/devadvice/353.html].
Compliance and Enforcement. Compliance requirements apply to both the
premarket approval process and postmarket surveillance. When a problem arises
with a product regulated by FDA, the agency can take a number of actions to protect
the public health. Initially, the agency tries to work with the manufacturer to correct
the problem on a voluntary basis. If that fails, legal remedies may be taken, such as:
asking the manufacturer to recall a product, having federal marshals seize products,
or detaining imports at the port of entry until problems are corrected. If warranted,
FDA can ask the courts to issue injunctions or prosecute individual company officers
that deliberately violate the law. When warranted, criminal penalties, including
prison sentences, may be sought.
Each center has an Office of Compliance (OC) which ensures compliance with
regulations while pre- or postmarket studies are being undertaken, with
manufacturing requirements, and with labeling requirements. The objectives of
CDRH’s OC’s Bioresearch Monitoring (BIMO) program are to ensure the quality and
integrity of data and information submitted in support of IDE, PMA, and 510(k)
submissions and to ensure that human subjects taking part in investigations are
protected from undue hazard or risk. This is achieved through audits of clinical data
contained in PMAs prior to approval, data audits of IDE and 510(k) submissions,
inspections of IRBs and nonclinical laboratories, and enforcement of the prohibitions
against promotion, marketing, or commercialization of investigational devices. Any
establishment where devices are manufactured, processed, packed, installed, used,
or implanted or where records of results from use of devices are kept, can be subject
to inspection.
The OC also reviews the quality system design and manufacturing information
in the PMA submission. It is to determine whether the manufacturer has described
the processes in sufficient detail and make a preliminary determination of whether
the manufacturer meets the QSR. If the manufacturer has provided an adequate
description of the design and manufacturing process, a preapproval inspection can
be initiated. Inspection is to include an assessment of the manufacturer’s capability
to design and manufacture the device as claimed in the PMA and confirm that the
quality system is in compliance with the QSR. Postapproval inspections can be
conducted within eight to twelve months of approval of the PMA submission. The
inspection is to primarily focus on any changes that may have been made in the
device design, manufacturing process, or quality systems.
40 OIVD, which reviews clinical laboratory test kits and equipment, has the capacity for both
premarket scientific review and pre — and postmarket compliance activity.

---

CRS-26
The compliance offices work closely with the Office of Regulatory Affairs
(ORA),41 which operates in the field to regulate almost 124,000 business
establishments that annually produce, warehouse, import and transport $1 trillion
worth of medical products. Consumer safety officers (CSOs) and inspectors typically
have conducted about 22,000 domestic and foreign inspections a year to ensure that
regulated products meet the agency’s standards. CSOs also monitor clinical trials.
Scientists in ORA’s 13 laboratories typically have analyzed more than 41,000
product samples each year to determine their adherence to the FDA’s standards.
Section 516 of the FFDC Act gives FDA the authority to ban devices that
present substantial deception or unreasonable and substantial risk of illness or injury.
Section 518 enables FDA to require manufacturers or other appropriate individuals
to notify all health professionals who prescribe or use the device and any other person
(including manufacturers, importers, distributors, retailers, and device users) of any
health risks resulting from the use of a violative device, so that these risks may be
reduced or eliminated. This section also gives consumers a procedure for economic
redress when they have been sold defective medical devices that present
unreasonable risks. Section 519 of the Act authorized the FDA to promulgate
regulations requiring manufacturers, importers, and distributors of devices to
maintain records and reports to assure that devices are not adulterated or misbranded.
Section 520(e) of the MDA, authorized FDA to restrict the sale, distribution, or use
of a device if there cannot otherwise be reasonable assurance of its safety and
effectiveness. A restricted device can only be sold on oral or written authorization
by a licensed practitioner or under conditions specified by regulation.
Warning Letter. A warning letter is a written communication from FDA
notifying a responsible individual, manufacturer, or facility that the agency considers
one or more products, practices, processes, or other activities to be in violation of the
laws that FDA enforces. The warning letter informs the recipient that failure to take
appropriate and prompt action to correct and prevent any future repeat of the
violations could result in an administrative or regulatory action. Although serious
noncompliance is often a catalyst for issuance of a warning letter, the warning letter
is informal and advisory. Warning letters are publically available on FDA’s website
at [http://www.fda.gov/foi/warning.htm].
Product Recall. A recall is a method of removing or correcting products that
FDA considers are in violation of the law.42 Medical device recalls are usually
conducted voluntarily by the manufacturer (21 CFR Part 7), after negotiation with the
FDA. Under 21 CFR Part 806, manufacturers (including refurbishers and
reconditioners) and importers are required to report to FDA any correction or
removal of a medical device that is undertaken to reduce a health risk posed by the
device. A recall may be a total market withdrawal or may be of a portion of product
41 See ORA at [http://www.fda.gov/ora/].
42 Recall does not include market withdrawal or a stock recovery. A market withdrawal is
a firm’s removal or correction of a distributed product for a minor violation that does not
violate the law and would not be subject to legal action by the FDA, e.g., normal stock
rotation practices, routine equipment adjustments and repairs, etc. Stock recovery involves
correction of a problem before product is shipped (i.e., is still in the manufacturer’s control).

---

CRS-27
(such as a single lot). In rare instances, where the manufacturer or importer fails to
voluntarily recall a device that is a risk to health, FDA may issue a recall order to the
manufacturer (21 CFR Part 810).
When a recall is initiated, FDA performs an evaluation of the health hazard
presented taking into account the following factors, among others:
! Whether any disease or injuries have already occurred from the use
of the product;
! Whether any existing conditions could contribute to a clinical
situation that could expose humans or animals to a health hazard;
! Assessment of hazard to various segments of the population, (e.g.,
children, surgical patients, pets, livestock, etc.), who would be
exposed to the product;
! Assessment of the degree of seriousness of the health hazard to
which the populations at risk would be exposed;
! Assessment of the likelihood of occurrence of the hazard;
! Assessment of the consequences (immediate or long-range) of
occurrence of the hazard.
Following the health hazard assessment, FDA assigns the recall a classification
according to the relative degree of health hazard. Class I recalls are the most serious,
reserved for situations where there is a reasonable probability that the use of, or
exposure to, a product will cause serious adverse health consequences or death.
Class II recalls are for situations where the use of, or exposure to, a product may
cause temporary or medically reversible adverse health consequences or where the
probability of serious adverse health consequences is remote. In a Class III recall
situation, the use of, or exposure to, a product is not likely to cause adverse health
consequences.
In addition to a warning letter or recall, FDA may also issue a public notification
or safety alert (e.g., “Dear Doctor” letter), to warn healthcare providers and
consumers of the risk of the device in question. The main page for recalls, market
withdrawals, and safety alerts for all FDA-regulated products is [http://www.fda.gov/
opacom/7alerts.html].
Comparison of Device and Drug Models
for Regulation
The regulatory system for drugs was created in order to reduce medical
“quackery” in the early 1900s. The first FFDC Act required drugmakers to prove to
FDA that their product worked for the use they were promoting, and required agency
approval for testing, production and marketing. At that time, few medical devices
existed. As a result of rapid advances in medical technology in the 1960s, the device
amendments were added more recently in the history of medical product regulation,

---

CRS-28
and were designed to facilitate rapid innovation that is characteristic of the industry.43
Because of differences in the evolution of statutory authority over drugs and devices
reflecting differences in the respective industries and products, different regulatory
review processes emerged. Functional review of drugs and devices is
organizationally separated within different centers of FDA (with coordination of
review for combination products, such as a dru- delivering device).
Industry and Product Related Factors
The medical device industry on a whole is much smaller than the
pharmaceutical industry, with estimated earnings of $80 billion in 2004 compared
to the drug industry’s estimated $222 billion (see Table 2).44 As a result, device
companies often do not have the economic or financial resources of multi-billion
dollar drug companies. Furthermore, the medical device market is highly
fragmented: surgical and medical supplies make up the largest sector, followed by
IVDs, cardiovascular devices, orthopedic devices, and diagnostic imaging.45
Although the largest companies dominate the market for devices in terms of sales,
it is often the small companies that make a significant contribution to early
innovation. They later will often partner with larger companies to bring products to
market, because small companies often lack access to capital, and often lack
resources to conduct clinical trials and navigate the regulatory and reimbursement
hurdles.
Devices are usually intended to treat fewer patients than drugs, and they are
often used for only a few years before being replaced with newer models or therapies.
Typically, there is initially a truly innovative device, which is then incrementally
modified several to many times over the course of its relatively short product life.
Given the dynamic process of improvement, regulatory requirements for incremental
modifications to approved devices are generally much less stringent than
modifications to approved drugs and/or their manufacturing process.
There are more incentives to encourage drug manufacturers to develop new
drugs built into the regulatory framework. The Orphan Drug Act (P.L. 97-414)
guarantees the developer of an orphan product seven years of market exclusivity
following the approval of the product by the FDA. During a time of market
43 R A. Merrill, “Regulation of Drugs and Devices: An Evolution,” Health Affairs, vol. 13,
no. 3 (1994): pp. 47-70.
44 U.S. Department of Commerce, International Trade Commission verbal estimates given
to Michele Schoonmaker, Jan. 2005. The $80 billion for devices includes both medical
devices and medical supplies. The $222 billion for drugs was based on the Department of
Commerce’s estimate of a $500 billion global market for drugs, of which they estimate that
the U.S. market is approximately 44% (same proportion as for devices). It was unclear if
the estimate for drugs included over-the counter drugs that are available without a
prescription. See also CRS Report RL3134, Health Expenditures in 2002. In this report,
it is estimated that $162.4 billion was spent in 2002 for prescription drugs in the United
States. Estimates for expenditures on over-the-counter drugs range from $5 to $19 billion.
45 The Legin Group, for AdvaMed, The Impact of Regulation and Market Dynamics on
Innovation: The State of the Industry (Washington, D.C., 2001).

---

CRS-29
exclusivity, FDA will not approve other competitive products for the same use.
Market exclusivity was also authorized by FDAMA wherein a drug manufacturer
could gain an additional six months exclusivity for performing studies in a pediatric
population at the agency’s request. Market exclusivity does not apply for devices.
Instead, under MDUFMA, device manufacturers can have application fees waived
for performing studies that the FDA requests (e.g., pediatric studies).
Market exclusivity does not extend a drug manufacturer’s patent (although FDA
will not approve a generic drug during a period of market exclusivity). Extensions
of patent beyond the 20 years presently granted are available for drugs and devices
with an NDA or PMA approved by FDA. These extensions are granted by the Patent
and Trademark Office (PTO), in consultation with FDA. The length of the extension
is variable, but it can be up to half of the time it takes to do a clinical trial. For drugs,
the clock starts when the Investigational New Drug (IND) application is approved
(i.e., before the clinical trial starts). For devices, the clock starts when the clinical
trial starts. Historically, the average patent extension has been greater for drugs
because of where the clock started, and because drug trials took longer to complete
than device trials.
FDA Review and Approval Processes
A greater number of drugs typically undergo more intense review compared to
devices, because there is no risk-based classification for them. Fewer devices, in
contrast, receive intense review. Most devices are cleared on substantial equivalence
based on comparative performance characteristics rather than actual clinical data.
Premarket. The premarket review phase for a drug is significantly longer
compared to that for a device. Both drugs and devices have a special regulatory
category for “investigational use” that allows for shipment of an unapproved product
for clinical research purposes. Clinical trials for drugs cannot begin without an (IND)
application filed with FDA. These are approved (or not) based on the preclinical data
provided. In contrast, IDEs are mainly required for a smaller group of devices,
usually those that are invasive, life-sustaining or high risk. Preclinical testing for
drugs is far more extensive than that for most devices (with the exception being new
implantable or life-sustaining devices, which may require animal studies). Drug
studies include computer modeling, animal testing, in vitro (i.e., test tube) studies,
and human studies, including toxicology studies, such as investigations into how the
drug is metabolized, where the drug and metabolites are distributed in the body, how
fast the drug and metabolites are cleared from the body. Preclinical data is almost
never required for Class I and II devices, or IVDs.
For both drugs and devices, valid scientific evidence is required to meet a
threshold for “safety” and “effectiveness.”46 FDAMA modified FFDC Act
requirements from two to one adequate and well-controlled study, and also permitted
use of surrogate endpoints for approval rather than clinical endpoints. The majority
46 FDA, Guidance for Industry: Providing Clinical Evidence of Effectiveness for Human
Drug and Biological Products, May, 1998, at [http://www.fda.gov/cder/guidance/
1397fnl.pdf].

---

CRS-30
of devices, especially diagnostic devices (all classes) and other devices that are
cleared under substantial equivalence, do not undergo “clinical trials” in the sense
that patients are not broken into two groups: one receiving the product (intervention)
and the other receiving an alternative. Studies for most devices are comparative
performance on the same samples or in a similar set of circumstances. Neither
similar drugs nor similar devices are routinely directly compared in order to assess
relative effectiveness of one product over another.
Drugs do not have a true “substantial equivalence” standard. Even generic
drugs, which are compounds similar to brand name pharmaceuticals in dosage,
safety, strength, route of administration, quality, performance and intended use,
undergo a similar review and approval process as the original. The application for
a generic drug is called an abbreviated new drug application (ANDA).47 The
standard for approval is “bioequivalence” which means that the manufacturer needs
to demonstrate that generic delivers the same amount of active ingredient to the
blood stream of healthy volunteers in the same amount of time as the original. All
drug manufacturers must undergo inspection (where device manufacturers can, but
often are not inspected), and must adhere to GMP and labeling regulations. The
target review time for an ANDA is 180 days although unlike a 510(k) for devices, the
review time for an ANDA is actually longer than the more stringent NDA.
User fees (i.e., fees paid to FDA by manufacturers applying for market approval)
for drugs are significantly higher than those for devices.
Postmarket. It has proved extremely difficult for FDA to remove a product
from the market once it has been cleared or approved. Questionable performance in
the premarket phase may generate restrictive labeling or indications and may trigger
a requirement for the manufacturer to conduct a postmarket study. Postmarket
studies are more common for drugs than devices. Alternatively, certain high-risk
medical devices can be subject to medical device tracking requirements. Devices are
more conducive to tracking requirements than are drugs because patient compliance
can often be verified (e.g., the patient has the device or does not, whereas it is
sometimes difficult to tell whether patients are compliant in taking their medication),
and because fewer devices need to be followed. FDA posts all recalls, market
withdrawals and safety alerts on its main website for all of its regulated products.48

47 FDA, Abbreviated New Drug Application Process for Generic Drugs, Dec. 30, 2003, at
[http://www.fda.gov/cder/regulatory/applications/anda.htm].
48 [http://www.fda.gov/opacom/7alerts.html].

---

CRS-31
Table 2. Comparison of the Drug and Device Approval Processes
Industry Variables
Drugs (CDER data)
Medical Devices (CDRH data)
U.S. Market size (2004)
$222 billion
$80 billion (includes medical supplies)
Average product lifetime
decades
years
Special incentives for development
Market exclusivity after approval
Patent extension for regulatory process
(determined by FDA), patent extension
(determined by Patent office)
for regulatory process (determined by
Humanitarian device exemption for rare
Patent office), pediatric exclusivity,
diseases, no pediatric exclusivity, but
orphan drug process for rare diseases,
waived fees for certain types of
waived fees for certain types of
submissions (such as pediatric studies,
submissions (agency defined).
agency defined).
FDA Variables
Premarket review
Required for all drugs
Some devices exempt
Type of submission and 2005 user fees
Not dependent on risk to patient
Dependent on risk-based classification;
— NDA (new drug application),
— PMA for high risk devices
— ANDA (abbreviated new drug
($239,237, standard; $90,910 for small
application) for generic drugs.
business).
User fees:
— 510(k) for low/moderate risk devices
— $672,000 for an application
($3,502 standard; $2,802 for small
requiring clinical data,
business).
— $336,000 for an application not
requiring clinical data or a supplement
requiring clinical data
— $262,200 (establishment
fees)
— $41,710 (product fees)

---

CRS-32
FDA Variables
Premarket review
Drugs
Medical devices
Total number of submissions, all types,
2,104 received, 2,019 decisions made
9,872 received; 9,570 completed; 4,740
in 2003
(a)
decisions made (b)
Total number of submissions, all types,
1,641 (2,019)
4,292 (4,740)
Approved/cleared in 2003
Total number of submissions, all types,
249 (2,019)
115 (4,740)
Approvable, in 2003
Total number of submissions, all types,
79 (2,019)
63 (4,740)
Not approvable in 2003
Total number of submissions, all types,
Withdrawn (drugs) Not approved
50 (2,019)
270 (4,740)
(devices) in 2003
Standard for approval
NDA, Efficacy supplements: safety and
PMA, de novo 510(k): safety and
effectiveness ANDA: bioequivalence
effectiveness other 510(k)s: substantial
equivalence.

---

CRS-33
FDA Variables
Premarket review
Drugs
Medical Devices
Types of studies
Safety and effectiveness: preclinical
Safety and effectiveness: preclinical and
and interventional clinical trials often
interventional clinical trials done for
done; data from at least one controlled
implantable or life-sustaining devices
clinical trial may suffice; approval can
may investigate clinical outcome), if a
be granted based on a surrogate
‘next generation’ studies endpoints are a
endpoint (21 CFR § 314.510), however,
surrogate marker of disease or health;
approval under this section is subject to
interventional trial rarely done for
postmarket studies to verify clinical
diagnostic devices; Substantial
benefit
equivalence: comparison of device
Bioequivalence: requires demonstration
performance to another legally marketed
that the generic form of an original
device using clinical or analytical
drug, used at the same dose intended
samples.
for the same population, delivers active
Summaries of approved/cleared
ingredient in the same amount at the
applications available on website
same rate.
Summaries of approved applications
available on website.
FDA Average Review time (days), 2003
NDA: 231 (priority, i.e., significant
PMA: 151
improvement), 357 (standard)
510(k): 76
ANDA: 510
Postmarket surveillance
MedWatch: searchable webpage for
MedWatch: searchable webpage for
medical product safety information
medical product safety information
Adverse Event Reporting (time indicates
Manufacturers report to FDA:
Manufacturers report to FDA:
days after becoming aware of an
— Serious and unexpected events
— Deaths, serious injuries,
incident)
within 15 calendar days
malfunctions within 30 calendar days
— Individual reports available on
— Serious events requiring remedial
website, but difficult to search
action: 5 working days
User facilities report to FDA:
— Death & serious injury: within 10
working days
Individual reports available on website
(Medical Device Reporting database)

---

CRS-34
Postmarket surveillance
Drugs
Medical Devices
Number of reports received, 2003
370,887:
64,369:
— 22,955 MedWatch reports directly
— 60,767 individual medical device
from individuals
adverse event reports from
— 144,310 manufacturer 15-day
manufacturers, user facilities, and
(expedited) reports
importers, and
— 58,998 serious manufacturer
— 3,602 voluntary reports from health
periodic reports
care professionals and the public
— 144,624 nonserious manufacturer
periodic reports
Tracking
Not required
Required for implantable, life-sustaining
or other high risk devices as defined
Number safety-based market
0
5
withdrawals in 2003
Sources: Center for Devices and Radiological Health (CDRH): Office of Device Evaluation/Office of In Vitro Diagnostic Device
Evaluation and Safety (ODE/OIVD), Annual Report: Fiscal Year 2003, at [http://www.fda.gov/cdrh/annual/fy2003/ode/]. And
CDRH, FY2003 CDRH Annual Report, at [http://www.fda.gov/cdrh/annual/fy2003/assuring.html#postmarket]. Center for Drug
Evaluation and Research (CDER), CDER New Drug and Biologic Approval Reports: 2003, (drug only), at [http://www.fda.gov/
cder/rdmt/default.htm], and CDER, “CDER Report to the Nation: 2003, at [http://www.fda.gov/cder/reports/rtn/2003/
rtn2003-3.HTM#AERS].
Notes: U.S. market size data were supplied by the Department of Commerce (see footnote 44). Other numerical data are for CDER
(drugs) and CDRH (devices) only; the percentage of devices regulated by the Center for Biologics Evaluation and Research (CBER)
is small compared to CDRH. At the time of writing this report, exact numbers of devices reviewed by CBER were unavailable for
2003. (a) The numbers reflect submissions for NDAs, resubmitted NDAs, manufacturing supplements, efficacy supplements (new
uses for approved drugs), and resubmitted efficacy supplements. Applications which are “withdrawn” or “not approvable” may
be resubmitted with new data at any time. (b) The number of decisions made includes data for: Premarket Applications (PMAs),
PMA Supplements, Humanitarian Device Exemptions (HDEs), HDE Supplements, Investigational Device Exemptions (IDEs), IDE
Amendments, and 510(k)s. Data do not include 4,424 decisions made for IDE Supplements, because the breakdown of data by
decision status was not available. An additional 648 applications were either withdrawn from FDA consideration by the
manufacturer or had another administrative action imposed that did not result in a decision. Data for approvable/not approvable
were not available for IDEs, IDE Amendments and 510(k) because that is not a relevant decision for those types of submissions.

---