Marlborough explained that the traditional approach of copying and pasting protocols from a Word document in one trial to the next without questioning what should be included and why, can consistently result in trial procedures that are not associated with a study’s primary/secondary endpoints, or safety or regulatory objectives.
Data from the Tufts Center for the Study of Drug Development, in fact, showed that in a typical trial “approximately 25% of the procedures are not associated with a key primary or secondary objective, safety, or regulatory requirement,” Marlborough writes in a new book, to be released Tuesday, called Re-Engineering Clinical Trials.
As far as what those other procedures included, she gave the example of exploratory endpoints or endpoints to support publications afterwards, or to evaluate a patient’s quality of life. “You would probably justify keeping that data in, but I would look to collect and manage it differently, and honestly there’s a chunk of data that’s basically in there for no reason,” she said.
When you look at what industry has done traditionally -- when data and technology are integrated into a clinical trial -- everything starts after the protocol is designed, Marlborough added.
Data Upfront
But the use of internal and external data, such as insurance data or electronic medical/health record data, to help determine protocol feasibility upfront can help determine whether there are sufficient number of patients that meet the trial’s eligibility criteria and where the highest concentrations of patients are.
“What many sponsors overlook is the huge amount of data they have internally available to them,” Marlborough writes. “Every organization that conducts clinical trials has data that can provide insight into the likelihood of success, including (but not limited to)”:
• Enrollment rates for trials with similar inclusion/exclusion criteria;
• Dropout rates in studies utilizing specific clinical procedures;
• Most common reasons for amending a protocol in a specific therapeutic area; and
• How common a particular clinical procedure is in a specific therapeutic area
And though the cost of building a database to study how design data and metrics can aid in protocol development may be expensive and time-consuming upfront, the benefits over the long run are worth the investment, she said.
“You can get the benefit [from such a database] pretty quickly,” Marlborough told us.