Commenters on FDA PPI Guidance Document Are Clear: We Need More Specific Expectations from FDA
In my last blog, I focused on the urgent need for FDA to share their thinking about what they actually want from patient preference studies conducted by industry and how they expect to use patient preference information (PPI) in regulatory decision-making. Here, I want to look more at the specific feedback FDA received from those who commented on FDA’s request for input into proposed changes to the PPI Guidance. The feedback was substantive and thoughtful. The hope is that FDA takes the feedback seriously and makes meaningful updates to the PPI Guidance that will ultimately be of benefit to patients.
Eleven individuals or organizations commented on the FDA docket. You can find all of the comments from the original, and most recent, opening of the docket here. Be sure to sort the comments by date to find the most recent submissions. There were several themes among the comments: a need to use existing best practice recommendations; the need to increase the expectation and accountability of those conducting PPI studies to engage patients in the study design process; more transparency around how FDA will use PPI data in its decision-making; a desire to see the guidance expanded to therapeutics beyond medical devices; consistency in attribute definition and utilization; and greater clarity around FDA’s expectations for study design, sample size calculations, the use of qualitative vs quantitative methods, and how to handle preference heterogeneity.
Many commenters noted the need to include the role of caregivers, family, and other stakeholders in the assessment of preference. Currently, the guidance is largely silent on the value of these types of data to FDA. Further, the majority of commenters made reference to ISPOR and IMI-PREFER’s best practice recommendations for patient preference research in regulatory decision-making. There is a sincere desire to see these guidelines incorporated into the FDA PPI Guidance document. Similarly, several commenters noted the work done by MDIC on the original Patient-Centered Benefit-Risk (PCBR) Framework and the more recent Patient Preference in Clinical Trials framework project. There are many resources available to FDA to further enhance the practices recommended in the PPI Guidance, including an ICH proposal to harmonize the use of patient perspectives by global regulatory authorities.
While a number of commenters provided actual language for use in the proposed updates to the guidance, the most common feedback was a request for greater clarity around FDA’s expectations. Very few medical devices have been cleared or approved with PPI data. One commenter noted that the “best” studies conducted so far have been partially funded by, and in collaboration with, FDA. For example, the original Obesity Study, the Heart Failure study, and the enormous Parkinson’s study conducted by the Fox Foundation. All of these studies had very large sample sizes that are unattainable for most industry sponsored PPI endeavors. And, while their results were informative and clinically meaningful, they do not provide realistic examples for industry to emulate. Given these examples, there is a need for FDA to provide more specificity on its expectations for what it will consider to be sufficien PPI data in a regulatory submission.
Several commenters mentioned the need for guidance on how attributes in PPI studies should be defined. A novel suggestion was to generate an “attribute library”; another was to harmonize attribute definitions. These ideas have fascinating potential to increase comparability across studies and consistency in PPI data collected for regulatory purposes. Development of such a harmonized attribute repository could be done in partnership between FDA and MDIC along with other organizations. An attribute that could benefit from this harmonization might be “invasiveness”, which is a useful attribute for implants, minimally invasive surgery, and other devices or procedures that require surgical openings or small ports in the body. The repository could be dynamic and grow as new attributes were developed or existing attributes evolved with better understanding from patient insights.
Most importantly, commenters want greater clarity from FDA on what it actually wants from industry. There is a need to expand the guidance to go beyond medical devices and include PPI study guidance for drugs, biologics, and combination products. Similarly, there is a need for much greater clarity on FDA expectations for sample sizes, study methodology, design, and how to cope with, and report on, preference heterogeneity. It is also important for us all to know: does FDA ever really want qualitative data? There are many qualitative methods out there that can generate exceptional insights into patient preferences. In addition, qualitative interviews should be used during the design of a preference study to ensure it is designed with patient input and the attributes are understandable. However, the PPI guidance does not provide sufficient clarity on whether FDA will consider qualitative PPI data to ever be sufficient for regulatory decision-making.
Finally, there is a desire for greater transparency in how FDA will consider and use PPI data in regulatory decision-making. Currently, there is no statutory requirement that FDA consider PPI data in its review of a device submission. One commenter noted that results from PPI studies should be provided back to the patient community to increase transparency and ensure patients are informed about the results of the research in which they participate. Of equal importance is assurance for the life sciences industry that if a company invests in the conduct of a rigorous PPI study that it will prove meaningful and useful to FDA in the regulatory review process.