Patient-Reported Outcomes to Support Regulatory and Payer Discussions

How can biotech companies use patient-reported outcomes to bring better drugs forward for patients? 

Generally, patient-reported outcomes (PROs) provide complementary information to measured or clinician-assessed functions or symptoms. First, PROs need to be tailored to the specific disorder and, where possible, specifically developed to assess complaints, and symptoms for single diseases will be preferred. For example, PROs may cover a variety of disease-specific functional outcomes and consequences of diseases on activities of daily living, like motor symptoms and cognitive symptoms. For many diseases, there are also disease-specific quality of life outcomes, which can sometimes overlap with functional outcomes, like for example the Sarcopenia Quality of Life (SARQoL) instrument. We also use generic quality of life measures, such as EuroQol and SF-36. These instruments are also suitable for QALY (quality-adjusted life years) calculations for health technology assessment (HTA)/HEOR applications to support pricing and reimbursement discussions, by quantifying the benefits in a way that allows for economic analysis. In some areas, specific PROs exist to assess side effects of drugs (classes). 

In addition, caregiver-reported outcomes for patients who cannot fill their own PROs – like children or patients with dementia – can assess caregiver observations, supplementing the clinician-rated or objectively measured functions or effects. Sometimes, PROs allow us to collect information to quantify important confounders, for example in our current obesity study, we’re collecting information through diaries on diet and exercise, because these are important factors that influence the efficacy of the outcomes we are measuring. Body weight and muscle strength, some of our key efficacy endpoints, are obviously confounded by changes in diet and exercise and these are not assessed as outcome parameters, but as confounders.  

It’s also useful to collect information from patients about how frequently they use healthcare resources to support your pricing and reimbursement negotiations, which would be especially relevant if an intervention leads to reduction in healthcare use. 

 

How do you integrate patient-reported outcomes into your clinical trials?

Mostly, I begin by identifying PROs from literature and then talking with experts and obtaining the scientific advice of regulators. Then oftentimes, suppliers need to be selected that can collect the data in a validated way. If the PRO doesn’t have the needed validation, you need to work with the developers of the PRO to improve on the data. Where PRO instrument validation was not performed to FDA standards, you need specialist methodologists in qualitative research who can help in designing anchoring exercises and exit interviews in the trials in a way that addresses validation issues and can help in assessing the clinical relevance of changes in PRO scores. This is our approach in the sarcopenia development program.

"The choice of a particular instrument is influenced by the presence of translations and completeness of validation of an instrument, especially per recent FDA requirements – a well documented development history including demonstration of content validity is needed."

How should CMOs think about which PROs to include in their trials?

It starts with identifying the purpose of including PROs. Are they supportive or pivotal? Why would PROs be needed – usually due to limitations of more “objective” methods? The choice of a particular instrument is influenced by the presence of translations and completeness of validation of an instrument, especially per recent FDA requirements – a well documented development history including demonstration of content validity is needed. The CMO needs to be critical by thoroughly evaluating content validity – as the FDA will too! The hierarchy in statistical analysis needs to be defined and, if primary, the Minimally Clinically Important Difference (MCID) and variability need be assessed to judge suitability. The sensitivity to drug and non-drug treatment effects needs to be assessed, for example, based on historical trial results.The phase of development is important; when you start in phase 2, novel instruments might be piloted and validated to be ready for phase 3.

Thereafter, the CMOs, in coordination with their clinical operations colleagues and advisory boards, need to ensure that the totality of all the measures proposed for a study will be doable for patients and not overly burden them. Also, before protocols are even designed, CMOs should be having discussions with patient representatives and patient organizations to get their thoughts as well.  

For many indications, the FDA and EMA still have a longstanding preference for more objective, measurable outcomes. For example, you can ask a patient if they think they can walk 400 meters, but the proof of the pudding is in the eating, so you need to do a performance test (400 meter walk test) and not a PRO. Furthermore, as a CMO you need to think about the main objectives of the study and who you want to convince with the study. Obviously, the three main “clients” for what we do are regulators, payers and healthcare professionals. Payers and healthcare professionals also generally favor objective measures over PROs to quantify benefits (outside of very specific fields like pain) or need generic QoL measures to be used in HTA applications. 

"CMOs, in coordination with their clinical operations colleagues and advisory boards, need to ensure that the totality of all the measures proposed for a study will be doable for patients and not overly burden them."

What are some of the mistakes people make when using PROs?

Not considering enough which instrument to use and whether alternative non-PRO type measures are available for the concept of interest, and not checking in-depth the regulatory acceptability of candidate PROs. Especially nowadays, the FDA has a very rigorous assessment of outcomes to be used for regulatory decision making, and many historically developed instruments do not satisfy these. Usually, the instruments are designed in a way to minimize or at least standardize and harmonize problems, like recall period where someone may be filling out a diary right before a visit instead of daily. If you have everyone fill out the same instrument with the same timing, you reduce that issue. Also, many PROs are now digital, which provides a validated timestamp for when they are completed, resolving the issue of retrospective completion. But again, digital completion or an initially paper-designed instrument may need a body of validation to be performed. Some studies require information to be collected multiple times a day, and controls and monitors can be built into devices to alert patients and improve compliance. 

Also, when using PROs, everyone needs to be trained: the investigator, sites and patients. You have to get their consent in a way where they understand what they are getting into. 

"When using PROs, everyone needs to be trained: the investigator, sites and patients. You have to get their consent in a way where they understand what they are getting into."

What are some mistakes people make on the clinical operations side of integrating PROs?  

I’ve encountered situations where staff involved in the execution of trials at the sponsor, CRO or sites underestimate or don’t familiarize themselves with the validation requirements underlying PROs. PROs are scientifically designed instruments that undergo rigorous assessments and validation, including translations. I’ve seen people take a PRO and send it through Google Translate before submitting it to an ethics committee to run it in countries without a formal translation. That’s not how this works. 

It can also be difficult to identify the originators of PROs and contact them to be able to use them in our studies. For example, there are two competing entities arguing that they have the right to the SF-36. Sometimes, people don’t realize that even simple concepts like patient global impression of change – where you ask patients how they rate their change versus baseline – is a PRO, and therefore needs to be included in anchoring exercises for other PROs. Especially for digital outcomes, there are strict requirements for verifying the authenticity of materials like e-signatures that need to be built into systems.

In conclusion, using PROs is unavoidable in most drug development programs, and CMOs are intricately involved in the selection, validation and clinical anchoring and interpretation of PRO data, which in many fields are critically important to document efficacy of novel drugs to the satisfaction of regulators, payers and clinicians.

In This Article