Getting study teams to consider using readily available data to help design and enroll clinical trials can be more difficult than it should be.
To encourage using real-world data, managers should start as early as possible in the design of a trial and use case studies to help staff understand what options are available. Creating a pilot for early phase or feasibility trials or using case studies where the data have worked are good techniques to help, said Sudha Raman, an assistant professor in the department of population health sciences at Duke University.
Raman, speaking at a Clinical Trials Transformation Initiative (CTTI) webinar, suggests asking three questions when considering using commonly available data like electronic health records and insurance claims in a clinical trial design:
“Once you have identified what kind of commonly available data to use, you then need to set up a discussion about how to use it concerning the scientific objectives, the endpoints, and the operational considerations,” said Raman.
Using commonly available data early on in a trial might actually replace current recruiting methods, said Jack Sheehan, Janssen Scientific Affairs director of real-world value and evidence in neuroscience, who also spoke at the webinar.
“Janssen recently needed patients on a specific background medication and were unsure how common that would be worldwide,” said Sheehan. The clinical team “looked at databases in the U.S. and Europe to get a better handle on that.”
But using medical records and insurance claims to support recruitment is not always perfect, said Sheehan. Claims and electronic health records data are often incomplete and not filed in a timely manner. “The level of recency of data is very different in HIV than it is for flu,” said Sheehan.
“Sometimes flu inclusion criteria specify having the virus within the last month” and claims data won’t be up to date.