An MD&DI June 1997 Column
Pilot studies can help manufacturers not only make better devices, but also save time, money, and resources. Considerable time and financial resources are spent on clinical research programs intended to demonstrate the safety and effectiveness of medical devices. Careful planning of such programs is essential to meet the manufacturers' goals for their products. The goals of any given clinical research program can vary widely. A sponsor's objective may be to obtain FDA approval for a new Class III product using data from a single pivotal trial that incorporates all the desired indications and label claims critical to success in the marketplace. Or the goal may be to conduct a focused postmarket surveillance and device retrieval program to monitor safety and collect data for use in the design of future products. Whatever the goal, the design of an efficient and successful medical device clinical trials program will almost always involve the use of pilot studies.
MEDICAL DEVICE CLINICAL TRIALS
Medical device clinical trials fall into three categories, distinguished primarily by differences in trial size and research objectives. They are: pilot studies, pivotal trials of safety and effectiveness, and postmarketing studies. Pilot studies, sometimes called feasibility studies, are conducted when additional information is needed on a product or on trial management and logistics before a larger-scale clinical study is begun. Pilot studies are usually single-center trials involving a limited number of subjects and are designed to accomplish any number of objectives within a clinical development program.
Pivotal trials of safety and effectiveness are larger, controlled studies designed most often as hypothesis-testing studies to support the submission of a premarket approval (PMA) application for a new device, a conformity assessment for CE mark registration in Europe, or a 510(k) premarket notification when clinical data are necessary to establish substantial equivalence.
Postmarketing studies, like pilot studies, can vary in size and design, depending on the objective of the research. They can be narrow in scope when designed to gather additional information on a specific safety or performance issue. Other goals of postmarket research include performing comparative studies with alternative or competitive treatments or devices directed at providing support for pharmacoeconomic claims, comparative effectiveness claims, or failure analysis investigations.
MEDICAL DEVICE CLINICAL TRIALS: THE GOALS OF PILOT STUDIES
Pilot studies are typically carried out when there is little or no experience using the device with humans, or if specific information is needed before finishing an investigational plan to demonstrate a device's safety and effectiveness in a pivotal trial. Pilot studies, however, are appropriate at any time in the life cycle of a product when research, marketing, or design objectives cannot be met without data from additional clinical experience.
Unlike pivotal trials, pilot studies are not usually analytical trials designed to test a hypothesis (e.g., that the product performance exceeds that of a competitor). Instead, they usually focus on gathering data for use in optimizing the design of subsequent clinical research or device modifications. Some common objectives of pilot studies for medical devices are summarized below. Although these objectives are discussed separately, multiple objectives can, and should, be combined in a single trial. This can save time, money, and resources.
Preliminary Safety Evaluations. Small pilot studies are used to conduct preliminary evaluations of the safety of new products for which clinical experience is minimal or nonexistent. These are generally open-label studies (i.e., they have no control group) conducted at a single site with fewer than 25 subjects. The design of such trials incorporates measures to minimize risk to the subjects, and their initiation depends upon adequate nonclinical performance and safety testing. Preliminary safety evaluation trials usually focus on the use of the device for a single indication. However, they may include subjects with a range of demographic, risk-factor, and disease-status characteristics, the purpose being to help identify unanticipated adverse events and those characteristics of patients that are most likely to raise safety issues.
Preliminary Performance Evaluations. Preliminary performance evaluations are often paired with preliminary safety evaluations as two primary objectives of a single trial. For the purposes of clinical research studies, performance evaluations of a medical device can be assessed from three perspectives: those of the patient, the user, and the device.
Evaluating patient outcomes provides performance data that help sponsors select the best clinical outcome measure to use in a subsequent pivotal trial to demonstrate effectiveness. This also provides data on the clinical benefit of the device and an estimation of variance on this and other variables, data that are critical for calculating sample sizes in a pivotal trial.
From the perspective of a user, usually a health-care professional, the preliminary performance evaluation will include gathering data on such design features as ease of use, reliability, and control, and on the suitability of the instructions for use.
Finally, the preliminary performance of the device itself indicates how well the product meets its performance specifications in clinical use.
Validation of Outcome Measures. When outcomes are being selected for use as end points in a clinical study, whether for a pivotal trial or for postmarketing research, those outcomes must have clinical significance.1 In some cases, such as the consideration of a reduction in pain as an effectiveness end point for trials of joint replacements, the clinical significance of an outcome measure is well established. But for some devices, sponsors will develop label claims that reflect new or special features of the product. If these an-ticipated claims are critical to establishing the effectiveness of the device, the clinical significance of outcomes and claims must be established, a process that can sometimes be achieved in pilot medical device clinical trials.
For example, consider an injectable product used to treat urinary incontinence. The definition of effectiveness for this product will not hinge on the statistical difference between the subject's condition before and after treatment, or even on a comparison of the subject treated with the new product versus the currently available alternative treatment. Rather, the determination of effectiveness hinges on the clinical significance of the outcome achieved. Is a 25% reduction in the occurrence of urinary incontinence clinically significant? A pilot study can address and justify the subsequent articulation of effectiveness criteria for use in a pivotal medical device clinical trial. Establishing the clinical significance of an outcome measure is a key issue in the design of a pivotal trial, not only because it leads to the definition of effectiveness and the trial success criteria, but also because it will influence the sample size, study rationale, and choice of a control group.
Validation of Methods for Measuring Outcomes. Distinct from selecting and validating the clinical significance of an outcome is the process of developing a method to reliably measure it. For many clinical outcomes, this does not pose a problem because validated measuring instruments are available for obvious end points, such as a reduction in cholesterol levels. Much more difficult is the reliable assessment of subjective end points, such as cosmetic outcomes, pain, and quality of life. Some end points have not yet been the subject of a controlled study in a clinical setting, making it necessary for sponsors to develop new or improved methods to assess outcome. Such methodologies can be validated in pilot studies.
MEDICAL DEVICE CLINICAL TRIALS: REGULATORY GUIDANCE
There is little formal guidance from FDA on the need for pilot studies, or on methods for their use in medical device clinical trials. However, FDA's Center for Devices and Radiological Health provides some information on the need for such preliminary research in a clinical trial guidance document and a related blue book memorandum.2,3 Although some question the advisability of leading medical device clinical trials into a multiple-phase research program comparable to that for drugs, it is clear that FDA is not advising manufacturers to do so. Certainly the enhanced rigor required by FDA in the design and conduct of pivotal clinical research can require the judicious use of pilot studies, but such studies are not a criterion for approval. Medical devices requiring premarket approval are still subject to only one well-controlled trial to demonstrate safety and effectiveness, whereas for drugs at least two such trials are required.4
MEDICAL DEVICE CLINICAL TRIALS: CONCLUSION
Medical device clinical trials are ultimately designed to gather data to meet FDA product approval requirements and specific marketing objectives. It should also be apparent that valid and credible clinical research can only help in the design of new products to improve health and patient care. The use of pilot studies to clarify the most appropriate indications for use, to identify and validate primary end points, to help choose a control group, to provide data for sample-size calculations, and to define success criteria is becoming an increasingly valuable and requisite step in clinical development programs for Class III products.
1. "Clinical Utility and Premarket Approval, PMA Memorandum #P91-1," Rockville, MD, FDA, Center for Devices and Radiological Health (CDRH), Office of Device Evaluation, 1991.
2. "Clinical Trial Guidance for Nondiagnostic Medical Devices," Rockville, MD, FDA, CDRH, Division of Biostatistics, 1995.
3. "Guidance on the Review of Investigational Device Exemption Applications for Feasibility Studies," Memorandum #D89-1, Rockville, MD, FDA, CDRH, May 1989.
4. "Statement Regarding the Demonstration of Effectiveness of Human Drug Products and Devices," Federal Register, 60 FR:39180 39181.
Karen Becker Witkin is a principal with the Weinberg Group, Inc. (Washington, DC, and Brussels), and manages the firm's worldwide health-care practice. She has written a section of and edited the forthcoming book titled Clinical Studies of Medical Devices: Principles and Case Studies.
Copyright ©1997 Medical Device & Diagnostic Industry