Setting Up the Match
Internal alignment within drug companies that strengthens the case for investing in adherence programs is another challenge;
efforts that are disaggregated and fail to show a link between investment and outcome are victim to being seen as just another
public relations exercise that makes the programs targets for budget cutters. "Patient adherence initiatives often lack a
thorough understanding of the root causes of discontinuing treatment, or fail to engage patients sufficiently and to address
the issue holistically," says the Capgemini white paper. "Individual tactics are tried by different brands and then discontinued
as budgets and priorities shift, before their impact is known."
Source: Integrated Benefits Institute (EE = Employees)
Round 1: Building in Metrics
FICO's Health Care Practice division is now taking the same demographics tracking and analytics it has used since 1958 to
come up with credit scores and applying those skills towards calculating non-adherence risk.
— Daria Blackwell, Founder, Knowledge Clinic
"FICO has, for the past 10 years or so, been providing data-based marketing solutions in the healthcare space," explains Todd
Steffes, vice president of healthcare practice at FICO. Over the last few years, Steffes says, FICO's healthcare clients have
been more focused on consumer marketing, and on how to better understand patients—including their reasons for non-adherence.
"That's where this idea came from," says Steffes. "Could we apply our predictive modeling skill sets and the methodologies
around how we do the credit scores toward the pharma space and help answer those kinds of business questions?"
The process of creating a Medication Adherence Score (MAS) began by partnering with pharmacy benefit manager MedImpact, which
has access to claims data for patients across the country. Because the information from MedImpact is strictly demographic,
it's all de-identified and HIPAA-compliant. FICO then used that data to determine which mixes of demographics were adhering
to their medication in certain therapeutic areas. After a base study of 12 months, FICO can now use that data to predict the
likelihood of adherence in other patient groups with similar mixes of demographics.
"What you're doing is looking at those initial patients—how they actually adhere and actually perform—and then analyzing their
attributes and identifying which mix of those attributes is really predictive towards how they performed. That creates a scorecard
that allows you to do the same analysis—using just publicly available data—for future patients," explains Steffes.
The upshot? FICO can then take this predictive data to pharmaceutical marketers, managed care organizations, and eventually
physicians—all those who are at crucial points of contact with the patient—to help develop strategies to target the specific
patient populations that are least likely to adhere.