USC Schaeffer Center on How Quantitative Patient Preferences Are Helping Clinical and Regulatory Decision-Making
Barry Liden, Director of Public Policy for the USC Schaeffer Center for Health Policy & Economics, tells us how quantitative patient perspective information can aid in regulatory decision-making, illuminate new treatments, and shape clinical trial design.
What are some methods for obtaining PPI?
The most kind of robust method, which tends to be perceived by regulators as the gold standard, is the discrete choice experiment, where you ask patients to choose between two to three “products,” or combinations of attributes. When thinking about attributes, I like to use a transportation example: when thinking about transportation options (train, car, plane, etc.), we usually consider at least three attributes of time, cost and ease/comfort. But in a discrete choice experiment, the bundles of attributes would be mixed into different combinations, and the survey respondent would pick the best combination in each choice. That’s repeated several times until they’ve weighed in on multiple attribute combinations. This approach is preferred by regulators because it minimizes bias and more accurately reflects the decision-making context that patients might normally face.
One method that is less cumbersome/cognitively burdensome is best-worst scaling: out of multiple attributes, you are choosing one as “best” and one as “worst”, through multiple iterations, resulting in a forced ranking of the attributes. FDA has said that they would be accepting a best-worst scaling approach for certain applications.
In the context of clinical trials, this is useful when you’re discussing with the FDA the endpoints you intend to measure. If you can show that some of the attributes – or endpoints – are not as important to patients, you can argue for smaller or less data collection, which is ultimately less burdensome on patients.
"In the context of clinical trials, this is useful when you’re discussing with the FDA the endpoints you intend to measure."
Can you go deeper into FDA’s appetite for PPI?
FDA is very open to doing this and they're encouraging sponsors to use this kind of data to help them. The FDA’s desire to incorporate patient perspectives into their risk-benefit decision-making process started during the HIV/AIDS crisis, where they met considerable criticism from the HIV/AIDS patient community that they were being too risk-averse. FDA has experimented with different techniques for gathering patient input, but those have mostly yielded qualitative responses. FDA faced a difficult challenge: how can they incorporate patient input in the regulatory context when they have to be evaluating clinical trial data?
The Center for Devices and Radiological Health (CDRH) went on the journey to figure out how that input can be quantified, and since then, CDER and CBER have adopted those approaches. The EMA has also adopted similar guidelines for drug makers. It is a growing experience though. When the industry works with reviewers who have never seen or dealt with patient preference data, it requires some education and coaching on behalf of some of the leadership within the agency to help reviewers better understand it. When regulators do understand, they're grateful for the insights and the unique applicability of patient preference data in regulatory decision-making.
What types of trials or devices are more preference-sensitive?
Where there are clear differences between alternative treatments, and/or when a therapy includes some significant risk, patient preferences are most valuable. As an example, if you’re working on a compound with potential neurological impacts, that would automatically raise scrutiny among regulators.
However, in a patient population where they’ve been suffering significantly, their perspective on and tolerance for risk in that area might be extremely high. That’s a scenario where patient preference data can help inform the regulator’s perspective. This is also the case in oncology and cardiology, among others, where patients might have extreme discomfort and diminished quality of life, and may therefore be more open to accepting the risks of tradeoffs.
"In a patient population where they’ve been suffering significantly, their perspective on and tolerance for risk in that area might be extremely high. That’s a scenario where patient preference data can help inform the regulator’s perspective."
What other insights can be generated through the course of surveying patients in this way that would be beneficial to pharma?
One is finding out the “Why,” of a patient’s preference. A good example is when working at my previous organization, we had a medical device with an increased risk of stroke as compared to the existing technology. There was a lot of perceived hesitancy on behalf of regulators to allow the product to get on the market with that increased risk.
We did studies that discovered patients in this population were willing to tolerate a much higher level of stroke risk than anybody would have guessed. We found out from qualitative interviews that these patients had been living on a day-to-day basis with an increased risk of stroke.
They had been told by their doctors that if they didn't take their daily medication, they could have a stroke. We discovered through additional prodding that patients had developed a tolerance for this risk over time because of their repeated exposure to it. The risk is cumulative over time, so as they get older, their risk of stroke gets higher and higher. In the early days of our technology’s development, the risk of stroke was roughly 5-8% higher than the alternative approach. Doctors thought it was a significant issue, but the patients thought that if it was less than 10%, it wasn't even on their radar screen.
"When regulators do understand, they're grateful for the insights and the unique applicability of patient preference data in regulatory decision-making."
Do you have any examples of where you are already seeing the value of PPI?
You can discover new treatments. One company I’ve worked with was working on a drug that was designed to treat a relatively rare condition. They were doing some qualitative patient preference research and asking patients to sign on to a website once a week and answer questions. One patient mentioned, “I'm really sorry. I haven't logged on lately. But I haven't slept for the last three days,” and when asked why, they said it was because of the pain.
The next round of questions, the company asked the patients, “Have you been having a hard time sleeping?” And every one of them has had major problems with sleeping. Research revealed that if someone has gone three days or more without sleeping, the endocrine system starts to react as if you have type 2 diabetes. You end up developing other comorbidities as a result of a failure to sleep.
It completely changed the company’s approach to the indications for the drug. It actually allowed them to expand the use of that drug to a broader population. If they could ease that pain, they could reduce sleep, they could reduce the side effects of Type 2 diabetes.
What are you finding is the biggest challenge in collecting patient preference information?
The biggest challenge to anything new is overcoming the momentum of the incumbent approach. Try to work with experts to find ways to lower the cost of investment and to provide those decision makers with real time experience and involvement in the process.
To conclude, what is the value of patient preference information?
There are the benefits we already discussed: value for that study, the possibility of discovering new solutions. But there is also the positive impact on employees involved in the drug discovery process, to get closer to the ultimate beneficiary of our research. Seeing how therapies have helped patients gives greater inspiration to continue to do the work.
Connecting patient advisory boards and patient listening sessions with real data that shows how you can move the needle on product approvals and market access shows why we’re doing this and the potential financial benefits for further investment in R&D and greater discoveries.
