MANAGED CARE April 2010. ©MediMedia USA
Twenty-two percent of patients who start taking warfarin, a commonly prescribed blood thinner, end up in the hospital within six months, according to health plan data compiled by Medco Health Solutions. That’s because it is exceptionally difficult to administer the proper dose, owing to great genetic variation among patients.
A study conducted by the PBM asked whether determining a patient’s genotype can make a difference in hospitalization rates, “which is an outcomes question important to payers,” says Robert S. Epstein, MD, chief medical officer at Medco Health Solutions.
It turns out that there are two genes that can help predict the proper dose, even beyond clinical factors like age and weight. These genes determine how well the enzymes CYP2C9 and VKORC1 function. These enzymes contribute significantly to how sensitive a patient is to warfarin, and, as a result, the dose that should be used.
“The study is interesting because it wasn’t designed to determine if genotyping keeps the patient’s INR [ International Normalized Ratio] in the normal range,” says Epstein. The INR is used to monitor the effectiveness of blood thinning and measures how thick the blood is.
This comparative effectiveness study examined the effect of incorporating genetic information into the management of warfarin treatment on hospitalization during the first six months of therapy.
Results were evaluated in two ways: One included all hospitalizations from the time of the initial warfarin treatment, even if genotyping had not yet occurred, and was defined as the intention-to-treat (ITT) analysis; the other only included hospitalizations that occurred after genotyping and was called the per protocol (PP) analysis.
“In the ITT analysis, every single patient who agrees to participate gets counted regardless of when he got his genotype test done,” says Epstein. “It’s the most conservative way of looking at your data. Say a patient agrees to have his genotype test done next week, but the next day, he suffers a bleed. His adverse event is still counted. Per protocol is where you only follow patients according to the protocol. In this case, we followed people forward in time after they were genotyped. Then we compared that to the control group.”
The results showed that for the ITT analysis, patients who had been genotyped had a 31 percent lower rate of hospitalization for any cause, and a 28 percent lower rate of hospitalization because of bleeding or blood clots, than for patients who had not been given a genetic test. The difference was even more pronounced for the PP analysis, with genotyped patients having a 33 percent lower rate of hospitalization for any cause and a 44 percent lower rate of hospitalization for bleeding than the control group.