As outlined in PDUFA VI, the FDA committed to hold public workshops and initiate and fund activities to address concerns and considerations in the use of Real World Data (RWD) for regulatory decision making. By the end of 2021, it is expected that the FDA will issue draft guidance on the use of Real World Evidence (RWE) in the assessment of safety and effectiveness in regulatory submissions.
I attended the workshop webinar on “Evaluating Real World Evidence (RWE) from Observational Studies in Regulatory Decision-Making” held on Feb 16 and 17, 2021, and hosted by Duke-Margolis Events. This workshop focused on in-depth studies conducted and supported by the FDA to assess the validity of using RWE compared to known randomized clinical trial (RCT) results. Scientists from Harvard Medical School, Brandeis University, Mayo Clinic, and others presented at this workshop, and panel discussions included representatives from the FDA. It was fascinating to see the amount of ongoing effort and the types of studies and in-depth statistical analyses that provide input to inform the next steps toward validating the use of RWE.
Topics presented were (1) principles for using rigorous observational study designs, (2) presentation of trial replication projects, (3) reactions to replication results, (4) key observations/themes from replication studies, and (5) opportunities, limitations, and next steps in the area of RWE.
Key learnings from the meeting included:
- Use of the term – Target Trial – This is a “trial” using a database (e.g., health records or claims databases) and should have a protocol with study design, inclusion and exclusion criteria, endpoints, and statistical analyses pre-defined. It was suggested to increase transparency by posting these trials to clinicaltrials.gov.
- Some studies showed a strong correlation with RWE and the RCT.
- Some outcomes are challenging to replicate with RWE because there may be a lack of details (e.g., information on deaths), and it is difficult to perfectly match the RWE populations, outcome measurements, time periods, and comparators to the RCT parameters.
- There is a continuous need to improve RWE data sources (e.g., standardization across databases and increase the type of information gathered).
- It is believed that as data and statistical methods improve, the reliability of causal inferences drawn from observational data will improve.
- Mitigating the introduction of bias is a real concern because you may know the RCT results as you design your RWE trial. You may want to look at the RCT results to inform your RWE trial’s design and then use another RCT to evaluate against your RWE trial.
- How data are gathered and processed in databases are not standardized; there are Center for Drug Evaluation and Research (CDER) initiatives ongoing evaluating how to collect RWD.
- There are probably two areas that we can trust:
- RWD for the control arm to augment RCT
- RWD for supplemental indications, label extensions including pediatrics, and for trials where you cannot enroll patients (e.g., ethical concerns), or patients are difficult to enroll (e.g., ultra-rare diseases)
In summary, the use of RWE to support regulatory submissions is an exciting and emerging area that is making progress. Both FDA divisions, CDER and the Center for Biologics Evaluation and Research (CBER), are supporting the investigation of RWD and are currently working on a portfolio of guidances for RWE. Contact us if you are interested in leveraging RWE to support your program. We provide guidance and advice on how to benefit from the use of RWE.
