Chuck Bednar for redOrbit.com – @BednarChuck
Using a popular type of medication to treat heartburn or acid reflux disease could increase your risk of having a heart attack, researchers from the Stanford University School of Medicine have discovered in a new, large-scale, data-mining study.
Published in the June 10 edition of the journal PLOS One, lead author Dr. Nigam Shah, assistant professor of biomedical informatics, and his colleagues used a novel new technique for mining clinical data to investigate the link between proton pump inhibitors (PPIs) and adverse clinical outcomes in users of the blood thinner clopidogrel.
PPIs, which include omeprazole (Prilosec) had previously been associated with patients who take blood thinners like Plavix, but after reviewing the electronic health records of just under three million patients, Dr. Shah’s team found that PPI use was associated with a 20 percent increase in heart attack risk in adults, even if they did not use clopidogrel.
While Dr. Shah, who also serves as the assistant director of the Stanford Center for Biomedical Informatics Research, said that the link between PPIs and the increased chance of a subsequent heart attack is not necessarily causal. He claims the concerns should be taken seriously, as PPIs are very frequently prescribed by doctors and are available over the counter.
Benefits of medical data-mining techniques
Using the same data-mining techniques, the Stanford-led team also looked at another commonly prescribed class of heartburn drugs, known as H2 blockers, and found no association with elevated heart-attack risk. In an email, Dr. Shah told redOrbit that anyone who has been using PPIs “for a long duration” and “without a prescription” should “have a conversation with [their] doctor” and to “ask abound alternatives” to PPIs, such as the H2 blockers.
However, he added that the discovery of the link between PPIs and increased risk of heart attack, while obviously significant, was not the most important accomplishment of the research. Instead, he told redOrbit, “the key message for me is that data mining applied to electronic health records (EHRs) can provide insights into questions we haven’t yet tested in a clinical trial.”
Shah is a pioneer in the field of using data-mining techniques to find elusive but medically relevant trends. He had developed methods to search through large quantities of EHRs (not just the structured portions, but also free-form notes added by doctors as well) in search of unexpected health outcomes associated with medications or drug combinations.
“Most of the media buzz is going to latch on to the ‘heartburn drug might not be safe’ meme. However, that is an example of an insight obtained via data-mining. I’d encourage people to look beyond the immediate hype, and see the value of making such data mining routine,” he told redOrbit. “Imagine, if for whatever clinical situation you are in, your doctor could interrogate EMR data to figure out what happened to people like you, who got treated the way you are being treated. Wouldn’t it be awesome to know that? Wouldn’t be good to know what drug combinations (i.e. pairs or triplets of drugs) might be safe or unsafe?” Yes, yes it would.
—–
Follow redOrbit on Twitter, Facebook, Google+, Instagram and Pinterest.
Comments