Analytics supports med adherence effort at Advocate Healthcare

Emerging program uses data from EHRs and other sources to identify patient trends and intervene with patients who aren’t following orders, says Tina Esposito.


As part of its analytics efforts, Advocate Healthcare has taken aim at medication adherence as a way to improve patient care and produce better outcomes.

The Illinois-based system, operating hospitals and physician group practices throughout the Chicago area, is focusing on an area that’s long been known to challenge care providers—ensuring that patients get prescriptions filled, and then take their medications at rates that make a difference.

What’s news is that Advocate is using data in its electronic health records systems, and from other sources, to be able to study adherence rates, find patients who are at risk for problems resulting from suboptimal adherence, and then give clinicians the tools to act upon that information.

Advocate Healthcare is using data from various electronic health records systems used at its facilities—Cerner, Allscripts, Meditech and eClinicalWorks—and blending in socio-economic data from Cerner, as it ramps up a program to increase patient medication adherence. It’s augmenting its research into medication adherence by assigning more talent to the problem, specifically a data analyst, statistician and data scientist.

The project includes development of a data warehouse and analytics platform with help from Amazon and Google, to leverage and analyze claims and clinical data, with the goal of identifying patients not adhering to their medication regimes.

The program is an extension of the 13-hospital delivery system’s adoption of value-based care, an initiative in full force now, said Tina Esposito, vice president of the Center for Health Information Services at Advocate, at Health Data Management’s Healthcare Analytics Symposium in Chicago. Advocate is seeing risk-based contracts increase, and its fee-for-service business is shrinking, from 82 percent in 2010 to 32 percent in 2015.

In addition to using EHR technology and analytics, Advocate also will develop a “deployment” strategy, an often-forgotten mechanism to not only give tools to physicians to assess medication adherence, but to help them understand how to meaningfully use the tools and communicate to patients the importance of adherence, according to Esposito. While more capabilities will be added to the EHR to support this, the intent is to not require more work by clinicians, she said.

The project focuses on describing and measuring patient medication adherence patterns to drive interventional strategies, said Darcy Davis, a data scientist at Advocate. Even if clinicians have information, new processes and workflows are necessary to help them understand what they can do with the information and determine the best actions to take. “Adherence controls both utilization and cost, and ultimately outcomes, and it is a major gap in clinical point-of-care,” she added.

There’s also a flip side to medication adherence that will be part of the program—identifying patients who are too adherent by taking their medication more often than medically necessary, and claims data can help make those identifications.

Guiding principles for the program include tools that aid physician judgment are more actionable than meeting certain thresholds, and patient needs must be central when discussing adherence patterns, as patients may be taking half of their medication because of financial costs. Claims-based adherence measures will show how often medication is available to a patient during a given time period.

Advocate also will graph patient adherence quarterly to assess the degree to which adherence may tail off. That’s important because patients are likely to appropriately take their medicine at least 70 percent of the time if prescribed one drug, but only 10 percent likely to take all their medicine if they have 10 drugs, Davis noted.

Leaders of the project shadowed physicians to see the tools and workflows they were using to assess medication adherence, and then brought them an early version of the new tool and asked for help in designing it, said Fran Wilk, clinical process designer at Advocate.

Characteristics physicians want in the tool include no additional login, building the tool within current workflows, minimal time to use the tool, the ability to flag patients of interest, near real-time data, integration with controlled substances and cost data formularies, among other factors.

More for you

Loading data for hdm_tax_topic #better-outcomes...