Health IT, Patient Engagement

Study highlights pitfalls with medication adherence in low-income populations

While it was hard to engage patients in large numbers, those who did participate saw marked gains in their medication adherence, though clinical measures did not really improve.

Brian Dixon - Indiana University

Brian Dixon

MedCity News has partnered with BioCrossroads to provide coverage focused on Indiana’s next generation of growth and innovation in life sciences.

Think medication adherence is tough among ordinary patients with Type 2 diabetes? Try doing it with a low-income population.

Academicians at Indiana University discovered that in a study published last month in the open-access Journal of Medical Internet Research. They discussed their findings Tuesday at HIMSS16 in Las Vegas.

While it was hard to engage patients in large numbers, those who did participate saw marked gains in their medication adherence. Still, clinical measures, including hemoglobin A1c, blood pressure, lipid control, body-mass index and visits to the emergency department, didn’t budge, reported David Marrero, director of IU’s Diabetes Translational Research Center, and Brian Dixon, a research scientist at the university’s Richard M. Fairbanks School of Public Health and the Regenstrief Institute.

Marrero and Dixon conducted a nine-month study at three primary care clinics at Eskenazi Health, the large safety-net health system in Indianapolis.

The researchers collected patient records and lab results from the Regenstrief electronic health record, medication data from Surescripts and patient-reported barriers to medication adherence from a custom-built portal. They also built an integrated dashboard with information to spark physician-patient discussions about medication adherence for patients with Type 2 diabetes.

After nine months, adherence for meds improved by 6-20 percent, depending on the drug class, but patient outcomes didn’t budge from the baseline. This may be due to the small size of the sample pool, or to physician disinterest, Marrero and Dixon surmised.

Regenstrief enrolled 96 patients in the study after identifying more than 2,300 candidates in the EHR. “This being a pilot, 100 patients is sufficient,” Dixon said. But it could be a challenge getting more people to participate if and when the institute decides to expand the program to its general population of diabetic patients.

Physicians at the Eskenazi clinics didn’t use the portal much at all, though Regenstrief had to build a custom one for this study because the health system’s regular portal vendor didn’t cooperate with the study, Dixon said. He wondered if nurse practitioners, pharmacists or social workers might be better case managers and thus more likely to log into the portal.

Dixon said he would like to expand and diversify the study base by including additional health systems in future research.

As for barriers, the results were somewhat surprising. The audience guessed cost and fear of side effects would be the biggest roadblocks to medication adherence. They were wrong.

Based on patient responses, the No. 1 roadblock barrier to medication adherence was aversion to medications in general, followed by patients simply forgetting to take their meds. Cost came in third, while access to a pharmacy was fourth.

In Indianapolis, there are about “five good pharmacies downtown,” Marrero said, explaining the access issue. But low-income populations often cannot get to them, “We have a really crappy bus system,” he quipped.