Original Research and Contributions

Special Report

Review Article

Case Report

Image Diagnosis

Nursing Research and Practice

Healthcare Communication


Narrative Medicine

Soul of the Healer

On the Cover

Improving Performance on Medicare Stars Medication Adherence Measures through Predictive Analytics


Heather Qian, MBA; Lynn Deguzman, PharmD, BDGP;
Scott Jacobs; Kevin Chiang, PharmD;
Chris Chang, PharmD, BCGP, PMP


Abstracts from the Kaiser Permanente 2018 National Quality Conference

From Colorado, Georgia, Hawaii, Northern California, Northwest, Southern California, Program Offices

Background: Low medication adherence is associated with worse clinical and quality outcomes. Accordingly, Centers for Medicare and Medicaid Services measures adherence to statin, blood pressure, and diabetes medications as part of its Medicare Part C and D Star Ratings. These 3 triple-weighted measures represent 11% of the Star Ratings portfolio. Annual changes and competitor improvements continue to raise the bar on performance required for 5 stars.
Methods: Population: Actionable patient lists were created by Pharmacy Analytical Services starting in 2015 and delivered through the centralized Permanente Online Interactive Network Tool (POINT) Medication Adherence Tool (MAT) for all Regions. These lists identified patients for intervention by regional pharmacy teams and included a date for each patient by which time refills must be made or the patient will be below the 80% adherence goal (per Medicare 5-Star program threshold) at year end. Interventions: Aside from live telephone calls, interventions included automated reminder calls, interactive texting reminders, batch kp.org messages, and outpatient pharmacy clinical services (OPCS) consultation modules. To prioritize outreach under limited resources, a predictive risk score was introduced in Q3 2017 to identify the highest-risk patients.
Results: Adherence scores have improved since the 2015 inception of the medication adherence programs, which leveraged analytic reporting through the POINT/MAT tool. Pharmacist and support staff outreach improved refill rates compared with a control group (39% vs 16% within 7 days, 46% vs 29% within 14 days). Telephone calls are among the most effective interventions for medication adherence. Across 7 Regions, the programs touched more than 160,000 patients in 2016 and 200,000 in 2017. Predictive risk modeling was implemented in Q3 2017. Because of data availability constraints since year end, the impact of the model is still being assessed, with results anticipated in Q1 2018. Outreach data and refill outcomes were collected. Work and partnerships are ongoing to enhance the predictive models in an iterative process.
Discussion: To improve quality outcomes while maintaining 5-Star ratings in medication adherence, pharmacy is building actionable analytics for adherence outreach and developing risk models for population management. This is made possible by internal partnerships within and between Regions, and with Program Office and data partners. Choosing to build analytics in-house leads to a continued cycle of investigative learning, monitoring outcomes, and development of workflows. Setting up analytic plans to track and to evaluate new tools (eg, predictive models) before tool implementation is important for value capture. Work is underway across the Regions to evaluate and enhance the new predictive models, leading to a more efficient and effective programwide effort.

Abstracts from the Kaiser Permanente 2018 National Quality Conference


Click here to join the eTOC list or text ETOC to 22828. You will receive an email notice with the Table of Contents of The Permanente Journal.


2 million page views of TPJ articles in PubMed from a broad international readership.


Indexed in MEDLINE, PubMed Central, EMBASE, EBSCO Academic Search Complete, and CrossRef.




ISSN 1552-5775 Copyright © 2021 thepermanentejournal.org

All Rights Reserved