Guidance Recap Podcast | Selecting, Developing or Modifying Fit-for-Purpose Clinical Outcomes Assessments
Thank you for joining us for another episode of the Guidance Recap Podcast. My name is Kylie Haskins, and I am today’s host. In this episode, I am excited to be talking with two people from FDA’s Center for Drug Evaluation and Research, Dr. David Reasner, the Director of the Division of Clinical Outcome Assessment and Dr. Laura Lee Johnson, the Director for the Division of Biometrics III. They will be sharing some thoughts with us on the newly published draft guidance titled, “Selecting, Developing, or Modifying Fit-for-Purpose Clinical Outcomes Assessments.” Welcome, Dr. Reasner and Dr. Johnson! Thank you for speaking with us today.
The first question is for Dr. Reasner. Before we talk about the guidance series and this guidance in particular, for listeners less familiar with this topic, what is a clinical outcome assessment?
A clinical outcome assessment (also called a C-O-A or “CO-Ah”) is a measure that describes or reflects how a patient feels, functions, or survives. Examples are patient reported abdominal pain, or how far you can walk in 2 minutes, or how clean your colon appears to be when a doctor performs a colonoscopy.
In the regulatory review process, a COA can be used to evaluate the clinical benefit and/or risk of a medical product being studied in a clinical trial. There are four types of COAs and they are categorized by who is reporting or performing the measurement: the patient is reporting their experience, the observer is reporting what they are seeing, the clinician is reporting their assessment, or the patient is performing a task.
It’s important to know what a COA is supposed to measure and the setting in which it will be used. What it’s supposed to measure is called the concept of interest and the setting in which it will be used is called the context of use.
Both the concept of interest and the context of use are essential to the success of the COA. For example, if you are developing a COA to evaluate a product intended to relieve abdominal pain in adults, a practical concept of interest could be patient-reported abdominal pain. However, the COA used for adults would not be appropriate if the context of use is in very young children. Very young children, because of their age, may not be able to reliably rate their own pain. However, it may be appropriate to use an Observer Reported Outcome measure where you ask a parent or other caregiver or observer like a teacher to rate the child’s signs or observable behaviors. In particular, those that indicate that the child may be in pain. For example, you might ask about observations of crying or holding the stomach. However, we caution against any proxy reporting by a parent or observer in which they answer questions as they think the child might answer. It is necessary for parents or observers to report only on signs and observable behaviors.
Before a COA is used in a clinical trial, it should be determined to be fit-for-purpose, which means the level of validation is sufficient to support its context of use. When you have a COA that is fit-for-purpose, you have enough evidence that the COA will provide meaningful information about the patient’s experience in the clinical trial in which the COA is used.
Now let’s focus on the guidance itself. Dr. Johnson, this is the third guidance published in the Patient-Focused Drug Development Guidance Series. Can you provide background about the series and describe the purpose of this guidance in the context of the series?
Sure, the patient-focused drug development guidance series consists of four methodological guidances that address, in a stepwise manner, how stakeholders can collect and submit patient experience data and other relevant information from patients and caregivers for medical product development and regulatory decision making.
Patient experience data are broader than just COAs. Patient Focused Drug Development Guidances 1 and 2 lay the foundation for describing best practices for methods of collecting input about what is meaningful to patients during medical product development.
Guidance 3 builds on the first 2 by providing recommendations for how to identify, modify, or develop a clinical outcome assessment for measuring a concept of interest.
From here, the final guidance—Guidance 4—will describe how the data collected from fit-for-purpose COAs in clinical trials are summarized and used to test a medical product and interpreted to support regulatory decision-making.
You can find more information about the FDA PFDD guidance Series by typing FDA and PFDD into any search engine and then going to the FDA PFDD website.
Dr. Reasner, how does all this inform the FDA review process?
Well, in general, the data collected using COAs is one part of many that FDA looks at when determining if a medical product should or should not be allowed to go on the U.S. market. Many times, the summary of these data leads us to what we call an indication, for example, what a drug is approved for, and the indication is further summarized in medical product labeling, package inserts, patient information sheets, printouts at your pharmacy, or more commonly today, online. This draft guidance is being jointly issued by three centers at FDA - CDER, CBER, and CDRH.
Dr. Johnson, what can patients take away from this guidance?
We hope that patients take away from this guidance that much of what is important to patients is important to regulators. Patient input strengthens the drug development process and helps to ensure that approved medical products provide treatment benefits that are meaningful to patients.
We think that another item that can be taken away from this guidance is the importance of developing these assessments in a rigorous manner that includes multiple different types of expertise. This can include working with experts in qualitative research, experts in quantitative research, biostatisticians, clinicians, and of course patients. It is really important to get this right because if it isn’t done right, the assessment may not measure what we think it is measuring, the measure may not be sensitive to important changes, or the measure may show that there is a change when there really isn’t. We hope that by providing clarity about FDA’s expectations we can expedite the process of incorporating COAs into clinical trials.
Dr. Reasner, can you highlight any new recommendations included in the draft guidance that you are looking for feedback on?
Well, draft guidance means we are looking for feedback on all the recommendations. It is worth noting that most of this is an encore. This draft guidance builds on the 2009 final guidance titled, “Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims” often times referred to as the PRO guidance, and once the PFDD series is complete, we expect that it will replace the 2009 guidance.
All of the basic principles from the 2009 guidance about what is important for modifying and developing new patient-reported outcome measures (which is one type of COA) are reflected in the current draft guidance, but broadened to be applicable to all COA types. The newest considerations in the guidance can be found in the section 4, “Developing the Evidence to Support the Conclusion that a COA Is Appropriate in a Particular Context of Use.” This section was added to ensure FDA's evaluation of measures are better harmonized with the current terminology and practices in other evidence-based-measurement fields like education and psychology. It was designed to provide a framework for validation that could be applied to all types of COAs, regardless of whether it's a multi-item patient-reported outcome measure, a performance-based task that people complete in the clinic, or any other COA type. You will also see appendices at the end to try to give some specific information for different types of COAs. The goal is to have a document that is not PRO-centric, and so, FDA is interested in feedback from the public to see if that goal has been achieved. Also, while we provide frameworks, we are hoping the draft guidance is clear that these are not simple checklists where you tick elements off. We are providing guidance on the information FDA will need to make its regulatory decisions.
Lastly, it’s important to stress that guidance 3 is about identifying, modifying, or developing COAs. One unintended consequence of the 2009 PRO guidance was that many sponsors assumed they had to create new PROs. The guidance is not just about developing new COAs. Rather, we’re trying to dissuade people from developing new COAs if there are good existing measures.
We look forward to feedback on the draft guidance, which we will use to help finalize guidance three. Positive and negative feedback are welcome, and we encourage submitting examples we can use, as well as exact wording recommendations with the line numbers and sentences indicated.
For our final question, Dr. Johnson, can you provide a couple of key items from the guidance that you especially want listeners to remember?
Yes, I’d be happy to. First, We’d like listeners to remember that before deciding on a COA for potential inclusion in a clinical trial, it is important to ensure that the foundational work is done to understand the patients’ experiences and to identify a concept of interest that is both important to patients and that is likely to be affected by the medical product under study in the trial.
Next, we’d like listeners to remember that sponsors should be clear, and make it clear to FDA, about what a COA is measuring and have evidence to support this. It should also be clear how the COA will be used in clinical research, and the population in which it will be used. There should be evidence to support that the COA is appropriate for that use, in that population.
A COA will be considered fit-for-purpose when the FDA review team determines there is sufficient evidence to support a clear rationale for the proposed interpretation and use of the COA in the clinical trial or a specific medical product development program. We realize COAs are used in registries and other data records that could be used for multiple purposes. The topics and considerations of what to consider in the rationale remain the same.
Lastly, we’d like listeners to remember that when selecting a COA, first consider whether an existing COA that measures the concept of interest in the intended context of use is available. While it still takes time and consultation with COA experts, using or modifying an existing COA may save time and other resources for patients, sponsors, and FDA—so long as there is strong support for the COA being fit-for-purpose in the relevant context of use. Regardless of the approach, it is important to describe the rationale for the measure’s use and to provide supporting evidence for this rationale.
Dr. Reasner and Dr. Johnson, thank you for speaking with us today about the draft guidance on clinical outcome assessments from the patient focused drug development series. We have learned so much from your informative discussion on the document. We would also like to thank the guidance working group for writing and publishing this draft guidance.
To the listeners, we hope you found this podcast useful. We encourage you to take a look at the snapshot and to read the guidance.