Measuring Adherence

There are both direct and indirect measures of adherence. In some instances, providers might wish to measure adherence directly by measuring the concentration of drug levels in the blood. More commonly used, however, are indirect measures of adherence which include patient questionnaires, pill counts, refill rates, and clinical response.

Calculating Adherence

A variety of calculations have been introduced to calculate adherence based on pharmacy claims data, however, no single approach has been uniformly accepted.

Medication Possession Ratio (MPR)

The resulting MPR usually ranges from 0 to 1. A value of 1 corresponds to 100% adherence. It is possible for the results to be greater than 1 if patients get early refills or if they have only filled the medication once. Some researchers are concerned that this approach overestimates a patient’s adherence. Learn more about using MPR to evaluate your clinical services

Medication Persistence

Persistence is simply the denominator of the MPR equation and is used to calculate how long the patient has been taking the medication. It does not take into account any breaks in therapy. 


Proportion of Days Covered (PDC)

The resulting PDC ranges from 0 to 1. A value of 1 corresponds to 100% adherence. The Pharmacy Quality Alliance (PQA) recommends that PDC be used in measuring adherence. The organization determined that the PDC approach results in a more conservative estimate, especially in cases of frequently switched medications. Additionally, PDC is the adherence calculation used for CMS plan ratings.

These metrics define adherence as >0.8 or 80% of days covered. It is important to note that medications such as those for HIV/AIDS and birth control may require closer to 100% adherence for effectiveness.

Most adherence is measured via claims data and adherence can be wrongly represented using these calculations. Examples of this incorrect estimation include instances where the medication is automatically filled or the directions have changed and a new prescription has not been issued. Additionally, these methods do not take into account administration techniques or timing of dosing. The best way to assess adherence is to discuss medication-taking behaviors directly with the patient.