Precision medicine research could help doctors make better prescription choices

A George Mason University study that examines the effectiveness of a commonly prescribed antidepressant could lead to comprehensive changes in how doctors prescribe medication.

Farrokh Alemi, a researcher in George Mason’s Department of Health Administration and Policy, and Aryan Mazloum, a first-year Health Services Research PhD student, began working on the study in October 2015.

They examined data from 1,933 patients with depression, which was collected by the National Institutes of Health (NIH) over a seven-year period using the Hamilton Depression Rating Scale.

In the study, Alemi used a new method he developed called stratified co-variant balancing. The method allows a patient’s concurrent diagnoses to be sorted out and held constant so only the effects of panic disorder can be seen.

Alemi and Mazloum found that Citalopram, sometimes marketed under the names Celexa or Cipramil, shouldn’t be prescribed for panic disorder-depressed patients because it isn’t effective.

Alemi said Citalopram was chosen for their study because it was the most prescribed antidepressant in the NIH data.

The Mason study takes some of the guess work out of prescribing antidepressants to patients, allowing clinicians to immediately eliminate a drug as an option, Mazloum said.

He explained that about 66 percent of people with depression do not benefit from the first antidepressant given to them by doctors. Sometimes patients have to try different drugs before they find one that helps, while still battling their symptoms, sometimes for months after their initial diagnosis.

“We have been able to personalize the method of analysis; you can make a decision about one person, whether they should take a medication or they should not, and that’s really the progress here,” Alemi said.

The goal is to use precision medicine to take the research they’ve done and go even further by applying it to people with other factors and concurrent conditions, such as bulimia or drug abuse, then use those factors to come up with treatment options specifically for that person, Mazloum said.

A comprehensive computerized decision support system would help doctors decide which medication to prescribe. Mazloum sees a future when statisticians would have one-on-one meetings with patients to understand what will work for them.

“Everybody here at Mason agrees that patients are the centerpieces,” Alemi said. “That’s why we are doing a project that will inevitably help the patient.”