"But these studies typically require large numbers of patients, huge amounts of data, and thorough follow-ups, none of which comes easy or free. The upshot is fewer investigations into scenarios that are clinically important but unlikely to yield a profit for the firms funding them. Accordingly, researchers have been developing an option that uses real-world data from insurers to save patients from falling through the cracks."
""There's little incentive for pharmaceutical companies to test combinations of treatments, for example, because a manufacturer typically aims to market its own product rather than evaluate it alongside another," said Sebastian Schneeweiss, professor of medicine and epidemiology at Harvard Medical School. In addition, subsets of patients - such as pregnant women - are excluded from trials for ethical reasons, leaving gaps in knowledge."
""There's very tangible use cases where we would love to have trials, but we'll never have them, and we need to recognize this," said Schneeweiss, who also serves as chief of the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women's Hospital. "And at the same time, we're sitting on this huge pile of longitudinal and patient-level data, of which only a very small proportion is used today to turn into evidence.""
Randomized clinical trials remain the gold standard for establishing medication effects and form the basis for most U.S. approvals. Those trials require large patient numbers, extensive data, and long follow-ups, which limits studies of clinically important but commercially unattractive scenarios. Longitudinal, patient-level insurance data can help fill these gaps by enabling evaluation of treatment combinations and of populations excluded from trials, such as pregnant women. Leveraging insurer records could answer questions about cross-indications and generalizability much faster than new randomized controlled trials, yet most existing data remain unused.
Read at Harvard Gazette
Unable to calculate read time
Collection
[
|
...
]