How often can real-world evidence replace clinical trial data?
Start adding items to your reading lists:
Create your account
Save this item to:
This item has been saved to your reading list.
Senior Manager, Health Research Institute, PwC USOctober 25, 2019
Life sciences companies seeking to swap clinical trials for analyses of data extracted from insurance claims and electronic health records (EHR) systems – so-called real-world evidence (RWE) – may find those opportunities are few and far between, according to a new study published in JAMA Network Open, JAMA’s open-access medical journal.
Studying 220 clinical trials published in seven high-impact general medical journals in 2017, researchers concluded that just 15 percent of them could have been entirely conducted using claims and EHR data.
The team, which included researchers from large health systems, government providers and a pharmaceutical company, asked whether each trial’s goals – assessing the interventions, indications, inclusion and exclusion criteria and endpoints – could have been gleaned from insurance claims, EHR data or both.
RWE, abundant and being collected at great cost in time and money by the US healthcare industry, is being carefully examined as a possible source of valuable data for trials and, critically, regulatory reviews. Earlier this year, the FDA published draft guidance exploring how RWE could support approvals, even if the data are unlikely to form the whole basis for approval for products.
The study authors noted one key caveat that could boost the use of RWE studies for some drugs: that many data did not exist at the time of trial publication in EHRs and claims databases because the products had not been approved by the FDA yet. That could indicate that the data may exist in greater quantities after approval and after patients start filling their prescriptions.
HRI impact analysis
While troves of data exist, they may not capture all of the necessary information that companies would need for regulatory submissions, such as a request for a new indication or a labeling change.
The researchers note that these constraints may limit the use of RWE to studies that complement, rather than replace, clinical trials. Such studies already have been used in support of some drug approvals. Companies that want to use this data may wish to work with healthcare systems and providers to create incentives for the capture of that data.
The JAMA Network Open study does not mention other data issues that could prevent the use of data for clinical studies. For example, data collected by busy clinicians in a hospital may not be accurate, or they may not be as accurate as data collected by specially trained clinicians in a clinical study environment.
Claims databases may contain errors or coding that maximizes accuracy for billing but not for medical decision-making. RWE also could point to correlations and not causations in outcomes, which could be hard to tease apart.