Originally published April 1996
The primary mission of the Office of Device Evaluation (ODE) of FDA's Center for Devices and Radiological Health (CDRH) is to ensure timely delivery of safe and effective medical products to patients. One aspect of this process involves ensuring that each device does what its manufacturer claims it will do and that manufacturers support such claims with valid scientific evidence. Approximately 10% of all submissions ODE receives for marketing authorization include clinical data as part of this scientific evidence.
Since the initiation of the medical device program in the United States in 1976, valid scientific evidence of safety and effectiveness from clinical trials has been required to support premarket approval (PMA) applications. However, the Safe Medical Devices Act of 1990 (SMDA) mandated that clinical trials also be conducted to support a claim of substantial equivalence for some Class II premarket notification (510(k)) submissions.1
To understand the need for supplying clinical data, it is necessary to consider the definitions of safety and effectiveness in FDA's statute and implementing regulations.2Safety is defined as "evidence that the risks presented to an individual from the use of a product are not unreasonable as supported by information from preclinical and clinical evaluation of the product." Effectiveness is defined in terms of "benefit derived to a significant portion of the affected population through use of the product." In essence, FDA defines safety and effectiveness in terms of a benefit-to-risk assessment of the use of a product based on valid scientific evidence submitted to support a marketing application. This includes the results of clinical trials conducted with the new product.
Products can reach the marketplace through the 510(k) process if a legally marketed predicate Class I or Class II product exists or if the predicate is a preamendment Class III device. FDA determines product classification by evaluating available evidence that establishes whether a given product type is safe and effective. In turn, a device's classification determines the kinds of controls (general for Class I versus special for Class II) that are needed to provide reasonable assurance of its safety and effectiveness. For a given 510(k), demonstrating substantial equivalence is intended to determine whether the new and predicate devices are sufficiently comparable that an independent classification procedure--establishing new safety and effectiveness and controls--is therefore not necessary for the new product.
As the medical device program has matured, the use of the 510(k) route to market has expanded to encompass significant technological developments. This results in differences between a given device and its legally marketed predicate that require additional clinical data to demonstrate equivalence. Through clinical trials, technological differences that could affect safety and effectiveness, and that might otherwise render a device not substantially equivalent, can be evaluated and demonstrated to have no significant effect. Clinical trials can also be used as a special control to allow a device type to be down-classified when it requires only specific clinical data to demonstrate equivalence.
CLINICAL TRIALS SCIENCE
Clinical trials science is an established discipline that provides consistent guidelines for all product types. Scientific principles direct how a study should be designed and evaluated, the basic requirements to include in a clinical study, and the appropriate and valid mechanisms for evaluating the trial data.
Clinical trials should contain well-defined hypotheses that address the intended use of the test product in the population for which it has been developed. Trials should also identify appropriate controls (historical or concurrent) and use a sample size based on the expected effect of the test device compared to that of the control product or procedure. Clinical trials should include a plan for data analysis that takes into account the amount and types of data to be collected. There are instances, however, where all of these elements may not be applicable. It is the responsibility of the trial sponsor to justify why a specific element is inappropriate or unethical.
As has been described in this series, no single valid clinical trial design is appropriate for all circumstances. The scientific evidence needed from a clinical trial should be the driver of the trial's design and evaluation. The claim a sponsor wants to make for a given product and the patient population in which it will be used define the design and scope of a clinical trial. The disease to be diagnosed or treated, the risks associated with the disease and with the product, and other products available to treat this condition also affect design and scope. In addition, the type of data a trial generates will affect the data assessments that can and should be made at the trial's conclusion. Sharply focused questions for a given trial strengthen trial design and analysis.
Because clinical trials are experiments that involve humans, they must be conducted ethically and appropriately to protect the subjects. To be ethical, a trial must develop useful data. It is unethical to conduct human research for its own sake, without a goal for the result of the study. Goals can be to advance knowledge about the natural history of a disease or to assess the diagnosis of or intervention in the given disease state. When conducting research on a new product, a trial's goal is most often to develop valid scientific evidence that demonstrates safe and effective use to support entry of the product into the marketplace. This is true for all health-care products: drugs, biologicals, and medical devices.
ROLE OF ODE
Within ODE the understanding and use of clinical trials in support of marketing applications have evolved. Early on, observational trials conducted on small numbers of patients were the accepted norm for clinical trials. Currently, trials are more frequently concurrently controlled and designed for statistical evaluation. As the industry has evolved and its products have become more sophisticated, the complexity and importance of medical device clinical trials have also evolved.
This increasing complexity has also led, unfortunately, to delays in the initiation of clinical trials. A recent informal evaluation of the causes of such delays revealed several common problems related to getting trials under way. Common problems include
- Lack of a clear hypothesis.
- Lack of a mature, well-thought-out trial design that includes how patients will be identified and assessed.
- Insufficient sample size to support sponsor claims.
- Weak plans for statistically evaluating trial results.
Trials that cannot produce the data needed to bring the product to market waste time and money and, most importantly, raise questions about ethical patient enrollment.
FDA's medical device program assists sponsors in identifying the questions to address before ODE will grant a new product marketing access. These questions may form the basis on which a manufacturer designs a clinical trial and determines the direction of data development. To obtain data to support safety and effectiveness, sponsors should design clinical trials that incorporate these regulatory questions as well as other issues to be examined. However, FDA does not dictate the details of clinical trial design because the development of each product is unique.
IDE PROGRAM CHANGES
During the past year, the administration of FDA's investigational device exemption (IDE) program has changed significantly. Changes include encouraging preIDE submission contact between ODE staff and sponsors and increased use of pilot trials to gather human data with new devices.3
Encouraging earlier interaction between a device sponsor and the reviewing division enables ODE to provide better guidance to the submitter and more background on new products to the review staff. An important issue to discuss during pre-IDE meetings is the timing of an IDE submission. The product design should be sufficiently mature so that only minimal changes take place prior to the submission of the marketing application. Any changes should not significantly affect patient outcomes. If substantive changes are made during a clinical trial that result in significantly different operating characteristics or expected patient outcomes, additional clinical data may be required before ODE will grant market entry. Clinical trials need to address a device's promotional claims and also FDA's need for sufficient safety and efficacy data. Early contact not only serves to identify these needs and the most efficient means of clinical testing but also reduces the time to initiation of a clinical trial.
Increased use of feasibility trials is being implemented in order to eliminate delays in the initiation of clinical investigations caused by poor clinical trial design. To address IDEs for which no significant safety issues exist that should preclude patient exposure to the device, ODE has allowed sponsors to begin trials in a limited number of patients while developing the definitive trial design. Early feasibility trials can answer device design questions about user interfaces or instructions for use. Sponsors can also use feasibility trials to assess data collection forms or identify problems not anticipated during bench assessments or in animal studies.
These and other changes to the program have increased the IDE approval rate from less than 30% in fiscal year (FY) 94 to more than 60% in FY95 (see Figure 1).4 Clinical trials of new devices now start earlier and because of improved communication are better designed. It will take several years, however, to determine whether these changes result in shorter development times for new products, which is the overall goal behind ODE's increased attention to the IDE process.
One focus for course development in the CDRH staff college is training in the design and assessment of clinical trials. In order to evaluate a clinical trial, reviewers must understand the theory of clinical trial design, the logistical elements of conducting a trial, and its proper assessment. To this end, reviewers can attend a basic course in clinical trials, a course in basic statistical methods, and a refresher course on the review process.
The basic clinical trials course is team-taught by staff from throughout the agency. Lectures address clinical trial theory; regulatory requirements of IDEs, 510(k)s, and PMAs; practical issues of trial design and conduct; statistical considerations in trial design and assessment; and the appropriate role of the reviewer in these activities. The course acquaints new reviewers with the FDA resources available to assist in evaluation of trial design and assessment.
Among the most challenging topics in the basic course are those that address appropriate trial controls, the amount of data required to support a safety and effectiveness claim, the variety of valid trial designs, and the trial design flexibility available to sponsors. The course discusses actual IDE trials and uses practical exercises to help reviewers learn how to evaluate design approaches. This course was offered twice in FY95 with a focus on enrolling experienced reviewers. In FY96 ODE plans to offer this module several times and to encourage newer reviewers to attend.
The course in basic statistics is also team-taught by senior staff within the agency. Members of the staff college, senior ODE reviewers, and ODE program operations staff will teach a review process course, which includes modules on IDE, 510(k), and PMA review; documentation of the review process; and writing the summary of safety and effectiveness data.
CLINICAL TRIALS GUIDANCE
ODE's outreach to the industry regarding clinical trials requirements for medical devices is also a focus of current and future CDRH initiatives.
Guidances. Guidance documents are one way ODE consistently communicates regulatory requirements. Product-specific guidances (see box) address clinical trial needs for given product types and outline preclinical requirements as well as the format of marketing submissions. CDRH and ODE plan to make the process of guidance development more accessible as well as to develop guidance documents in more areas of device evaluation.
In September 1993, CDRH sponsored a workshop on clinical trials. One aspect of the center's work at that time involved developing a general clinical trials guidance document for industry use. Following the workshop, the center received numerous comments regarding that document. In FY96, the center plans to issue a revised draft of the guidance that focuses on the statistical aspects of trial design. In addition, CDRH will issue companion documents addressing the clinical considerations of general device trials and a special document on trial design for in vitro diagnostics. CDRH is also currently planning a second clinical trials workshop.
Industry Access. During the past several years, ODE has used public panel meetings and open workshops to involve a wide group of individuals with varying expertise in the review process. These meetings have served as a forum for discussion of the specific FDA requirements for safety and effectiveness for a given type of medical device and the extent of data needed to develop appropriate instructions for the health-care community. Such discussions have included appropriate technical information development and labeling for the reuse of hemodialysis filters, the reclassification of immunohistochemical stains, and the testing and labeling of devices containing natural latex.
Clinical Community. ODE has also increased outreach efforts to practice societies and associations in order to involve the clinical community in the development and evaluation of new technologies. Using FDA's Office of Health Affairs and its own resources, ODE has increased the participation of health-care practitioners in both the pre- and postmarket assessment of devices. As a result, the number of individuals cleared to act as special government employees and to participate in panel meetings as consultants and voting members has also increased. These individuals can be called upon to respond directly to ODE review questions that involve products before they reach the final PMA review stage. ODE has greatly benefited from this valuable clinical expertise.
CDRH and particularly ODE have focused much effort on device clinical trials design and assessment. The program is intended to facilitate better, more appropriate assessment of new technologies so that safe and effective products reach consumers within reasonable time frames.
Specifically, ODE has developed new guidances for industry, increased the training of CDRH review staff, and stepped up its involvement and communication with the health-care community, the ultimate users of medical device technologies.
1. Food, Drug, and Cosmetic Act 515(c) and 515(d).
2. Code of Federal Regulations, CFR 21 860.7.
3. "Goals and Initiatives for the IDE Program," Bluebook memo D95-1, Rockville, MD, FDA, Center for Devices and Radiological Health, Office of Device Evaluation, July 12, 1995.
4. Annual Report, Rockville, MD, FDA, CDRH, ODE, 1995.
Susan Alpert is director of the Office of Device Evaluation at FDA's Center for Devices and Radiological Health (Rockville, MD).
ODE GUIDANCES FOR CLINICAL TRIALS
The Office of Device Evaluation (ODE) has more than 200 guidance documents available that provide information relevant to clinical trials. The following list represents a sampling that have been published in the past two years.
Office of Device Evaluation
"Goals and Initiatives for the IDE Program" (July 1995)
"Availability of Investigational Devices" (May 1995)
"IDE Refuse to Accept Procedure" (May 1994)
Division of Cardiovascular, Respiratory, and Neurological Devices
"Coronary and Cerebrovascular Guidewire" (January 1995)
"Cranial Electrotherapy Stimulators" (August 1994)
"Interventional Cardiology Devices" (May 1994)
"Coronary and Cerebrovascular Guidewire" (March 1994)
"Replacement Heart Valves" (December 1993)
Division of General and Restorative Devices
"Spinal Fixation Device Systems" (July 1995)
"Saline-Filled Silicone Breast Implants" (December 1993)
Division of Ophthalmic Devices
"PRK Laser IDEs/PMAs" (July 1995)
Division of Reproductive, Abdominal, Ear, Nose and Throat, and Radiological Devices
"Benign Prostatic Hyperplasia" (November 1994)
"Vasovasotomy" (November 1993)
All ODE guidance documents are available from the Division of Small Manufacturers Assistance (DSMA). To obtain copies, contact DSMA via electronic docket: 800/252-1366 or 301/594-2741; Facts on Demand (telefax): 800/899-0381 or 301/827-0111; phone 800/638-2041 or 301/443-6597; or by mail at 1350 Piccard Drive, Rockville, MD 20850-4307.