Patient preference information: Researchers offer insights on medical device clinical trial design

Regulatory NewsRegulatory News | 06 September 2022 |  By 

The US Food and Drug Administration (FDA) has increasingly called for the inclusion of patient preference information (PPI) to inform its regulatory decision-making for medical devices. A recently published report offers insights on the opportunities to leverage PPI, such as how to maximize stakeholder engagement, identifying novel endpoints and statistical considerations.
The article, published in Therapeutic Innovation & Regulatory Science on 27 August, looks at recent advancements in the use of PPI for medical device development, including lessons learned by industry, regulators and other stakeholders from the Medical Device Innovation Consortium’s (MDIC) patient preference information in the design of clinical trials (PPI-CT) framework.
MDIC is a public-private partnership between the FDA, industry, patient advocates, academia and other stakeholders aimed at advancing the regulatory science behind medical device and diagnostics development. The group has been influential over the past decade in working with the FDA to develop new regulations in using patient preference data for regulatory uses.
In 2015, MDIC began several projects aimed at studying how PPI could be used to develop new products, which resulted in the PPI-CT framework. Based on the framework, the report’s authors have highlighted some steps that PPI proponents, and especially device makers, should consider when incorporating PPI into clinical trials.
First, they recommend gauging regulatory interest in using PPI. The authors note that officials at the Center for Devices and Radiological Health (CDRH) have shown a strong interest in using PPI and have issued guidances in recent years on how to use such data in regulatory applications. The authors also note that its critical sponsors engage the agency early on when developing research plans.
“A key ingredient for success in these efforts is early buy-in from critical stakeholders about the importance of ensuring that a clinical trial design reflects the tradeoffs patients are willing to accept as a function of the magnitude of gains and decrements in health,” they said. “Involving these stakeholders helps to ensure that patient preference studies are designed and positioned to collect PPI that best addresses the pertinent research questions.”
The report also offers recommendations on how to identify novel endpoints for PPI in clinical studies.
“Endpoints are most useful as attributes in a patient preference study when they can be translated into attributes that are understandable and meaningful to patients [34],” the authors wrote. “Once attributes that matter to patients (including potentially different sub-groups of patients) have been identified through patient preference studies or other approaches, researchers can work with their clinical trial teams and regulators to appropriately incorporate them as endpoints in the clinical trial.”
They note that doing a patient preference study early in the product development process is a key opportunity to help researchers develop novel endpoints that may be used to collect PPI. In this early stage, researchers should use a “bottom-up” approach whereby they get patient input on what’s meaningful to them in terms of successful treatment of their condition, rather than just looking for traditional endpoints.
The authors go on to state that clinical attributes such as survival and pain, which are typically considered high priority for patients, may be used as primary or secondary clinical trial endpoints. However, they’re not always the endpoints that patients prioritize. As an example, sometimes factors such as independence are a higher priority for patients.
With that in mind, the study authors said stakeholders should take into consideration aligning patient preference study attributes and traditional clinical trial endpoints.
“Identification of specific elements of a composite endpoint to define attributes for the patient preference study could be discussed with regulators, with the goal of aligning on an approach to evaluate and weight measurable factors (often symptoms) that matter most to patients,” they said. “It is important to note that patient preference studies typically do not define the entire sphere of relevant endpoints; rather, they may suggest additional outcomes or help to prioritize outcomes already identified as relevant.”
The authors also looked at how researchers can ensure the PPI being evaluated is applicable to the study population. A key part of this is to ensure participants are recruited in a timely manner and can provide accurate reporting. The authors acknowledge that it may be hard to identify the right participants and get them to provide accurate reports due to their condition for multiple reasons. For such patients, the authors said it may be useful to use a “confirmed diagnosis” method, which includes engaging patients who have been referred to by a physician or whose electronic medical records verify the diagnosis for the condition being studied.
“Because obtaining confirmed diagnoses can be time consuming and costly, patient preference studies may also collect data from individuals who ‘self-report’ or self-identify that they have a certain condition,” the study authors added. “In these cases, researchers may want to look for secondary data (e.g., the channel through which a patient was contacted, such as a patient organization network) and supporting information from participants (e.g., information about their symptoms or treatments that may be unique to the relevant condition) that can be used to increase confidence about the diagnosis.”
One of the most important areas that MDIC has been working on in the past few years is developing new statistical tools to evaluate studies that have small patient populations. With medical devices in general, it may not always be feasible to conduct large clinical trials; this is especially true if those devices are being developed for a small patient population.
In such small studies, the authors note that the Bayesian decision analysis (BDA framework developed by researchers from the Massachusetts Institute of Technology (MIT) could be used to set the statistical significance threshold for a clinical trial in a “systematic, quantitative, patient-centered, and transparent” manner.
“The methodology attempts to balance the consequences of approving an ineffective and possibly harmful treatment (false approval) against the consequences of rejecting an effective treatment (false rejection) such that the overall expected utility of a clinical trial is maximized,” they said. “In addition to incorporating PPI related to risk-tolerance among patients, the BDA framework can also analyze tradeoffs relating to time preferences among patients (e.g., how long would patients be willing to wait for a novel device). This framework might be particularly valuable in diseases for which clinical trial recruitment can be challenging, such as rare disease or conditions with a high mortality rate.”
Leveraging Patient Preference Information in Medical Device Clinical Trial Design


© 2023 Regulatory Affairs Professionals Society.

Discover more of what matters to you