Abstract
Background and Aims: Medication adherence studies usually rely on a narrow aspect of a patient’s medication history. Studies using administrative claims data such as insurance pharmacy claims cannot compute adherence for patients who don’t fill their medication orders or who use spousal benefits. Studies using medication orders from the electronic health record (EHR) assume that all patients fill their medication orders. Linking EHR medication orders to pharmacy claims can create a more robust picture of a patient’s medication history. The aim of this study was to create an algorithm for linking EHR orders to pharmacy claims to create a new metric of medication adherence – the eMPR (electronic Medication Possession Ratio).
Methods: Patients included in this study had a Geisinger primary care clinic provider while enrolled with the Geisinger Health Plan. Overlapping information across the EHR medication order database and the pharmacy claims database were used to identify likely matches between a prescription and subsequent fill. The overlapping data includes medical record number, order and fill dates, days supply, and National Drug Code (including medication name and dose). To evaluate the usefulness of the algorithm, the percent of claims that matched to an order was calculated.
Results: The linking algorithm is currently being tested as part of ongoing research to be completed by the time of the HMORN Annual Meeting (results pending). Several challenges identified include:
-
Developing a method of linking EHR orders to pharmacy claims;
-
Creating a link for medication orders and multiple refills, as each generates unique pharmacy claims; and
-
Linking generic forms of the prescribed medication to name-brand prescriptions written.
Initial results indicate a high degree of matching which fluctuates based on assumptions provided to the algorithm.
Conclusions: Medication orders and pharmacy claims can be used to create a link between a patient’s order and their pharmacy claim or “fill.” The addition of medication orders to the HMORN Virtual Data Warehouse may be valuable for improving the quality of medication adherence research within the HMORN, and lay the foundation for an eMPR which can be widely used in many settings.
- Received May 27, 2010.
- Accepted May 27, 2010.
