Big data unpacks blood pressure reduction role of drug combination

August 10, 2015 − by Suzanne Elvidge − in Big data, Big data in research, Drug development − No Comments

Doctors often prescribe triamterene along with the diuretic hydrochlorothiazide in high blood pressure because it protects the body’s levels of potassium. However, its role could be more important than that – in a recent big data study, triamterene boosted the blood pressure impact of hydrochlorothiazide alone. The results were published in the Journal of General Internal Medicine.

In a big data study carried out by researchers from the Regenstrief Institute and Indiana University, the team looked at the anonymised electronic medical records of 17,291 patients with hypertension from the Indiana Network for Patient Care. They compared the blood pressure data from patients who were taking hydrochlorothiazide, with and without triamterene, either alone or in combination with other antihypertensive medications.

“Physicians now know, beyond anecdotal evidence and studies that were too small to provide definitive findings, that there is an additional benefit from a drug we often prescribe, and perhaps should be prescribing more often,” says J Howard Pratt, of Indiana University School of Medicine.

Looking at the propensity score (which measures probability) of a patient receiving triamterene in each combination of medications, the researchers found that the mean blood pressure in the triamterene + hydrochlorothiazide group was 3.8 mmHg lower than in the hydrochlorothiazide only group (p < 0.0001). They also found that blood pressure was lower for patients taking triamterene with other medication combinations.

These findings, the first time they have been seen, suggest that adding in triamterene to other blood medications not only protects potassium levels but also improves the effect on lowering blood pressure.

“This study is a perfect example of how we can learn about the previously unknown therapeutic effects of drugs from big data. In this case, big electronic medical record data are being used to answer questions that may otherwise be unanswerable,” says Wanzhu Tu of Regenstrief Institute.

As Tu explains, companies are unlikely to carry out large scale and costly clinical trials on long-established generic drugs likely these, and small clinical trials don’t have enough power to reveal a drug’s effect.

“Observational studies based on big data, like ours, provide a viable alternative,” adds Tu.

Post a Comment

Your email address will not be published. Required fields are marked *