Refill Metrics vs. Real Outcomes: When Medication Adherence Doesn't Equal Better Health

Medication adherence measures based on prescription refill history do not always equate to disease control or high-quality patient care, asserted Sarah J. Billups, PharmD, in a presentation at the American Society of Health-System Pharmacists (ASHP) 2025 Midyear Clinical Meeting and Exposition and in a subsequent interview with Pharmacy Times.

Adherence metrics are measured in "the percentage of days a patient has medication on hand," Billups said, describing the proportion of days covered (PDC) calculation used to measure adherence. "It has nothing to do with if you actually take the drug or not."¹

Billups summarized 2 studies, one from a Veterans Affairs database and one from a Medicare database, both of which showed lower medication possession ratios were linked to higher mortality, and higher possession was linked to a higher likelihood of low-density lipoprotein (LDL) control, an indicator of cardiovascular health, and lower overall health care costs. However, both studies have a flaw: that "high adherence [to medication] may itself be a marker for overall healthy behavior, a term known as the healthy adherer effect. In fact, high adherence to placebo has been associated with improved outcomes."²

"None of these studies controlled for social determinants of health," Billups said. "Things like income or education or other things going on ... that you just can't measure."

Looking at Refill Metrics vs Disease Control

Billups analyzed records of nearly 5000 Medicare Advantage patients eligible for at least 1 adherence metric for diabetes, cholesterol, or hypertension (some patient records were counted in more than 1 indication). Patients were classified as adherent if their PDC for medications in these indications was at least 80%. Disease control was assessed using commonly applied quality thresholds: blood pressure below 140/90 mmHg, hemoglobin A1c below 8%, and low-density lipoprotein (LDL) cholesterol below 100 mg/dL.

For hypertension (n = 2962), Billups said, “The people who passed the adherence measure…76% of them had a blood pressure that was less than 140 over 90.” Among those who failed the adherence measure, “71.5% had controlled blood pressure.” The results of being medically adherent, therefore, did not result in statistically significant improvement in hypertension control.

A similar pattern emerged in diabetes (n = 979). "The people who passed the adherent measure…70% of them had an A1c that was considered controlled," Billups said, "compared to 76% of the people who failed. "There's no correlation" between having enough medication on hand and the clinical improvement of the disease state.

“These are different things,” Billups said. “Adherence and disease state control are not the same.”

In the cholesterol arm (n = 3857), outcomes showed more alignment with the expectations but with important limitations. “Only about 2300 of them actually have an LDL measured,” Billups said. She called the measurement of LDL the best indicator of adherence for this patient population but noted that LDL is not always measured as part of a medical check-up, even for cholesterol patients. This, she said, limited the ability to check adherence. “If you want to know if a patient’s taking their statin, the best way to do that is to look at what their LDL is,” she said. “That tells you if they’re taking it or not.”

Limitations of Medical Adherence Metrics

Refill-based adherence measures are currently triple-weighted by the Centers for Medicare and Medicaid Services (CMS), and the result of this weighting has been that health systems push patients more strongly towards automatic refills. “It drives behavior and it drives resource use," Billups said. “If you wanted to perform perfectly on these measures ... what you need to do is get all of your patients who are on these chronic medicines and put them on automatic medication refills.” This, she said, looks good from a PDC and adherence metric, but "it has nothing to do with what [patients] are really doing."

Other tactics include granting extended-day prescription supplies and filling prescriptions through mail-order pharmacies, and especially outreach by telephone. “Pharmacists call [patients] and they remind them to pick up their prescription,” Billups said. “I’ve done it, and it’s extremely unsatisfying work.”

Pharmacies push this effort more strongly at the end of the calendar year so they can pass the year as a whole. "You’ve got people in December…you have to call her and get her to fill her medication by Tuesday, or she flunks the measure," Billups said. "That’s not a medical emergency. It just isn’t."

Moreover, "there’s a substantial number of people…who are failing the adherence measure [because they] have excellent disease state control, and we shouldn’t be bothering them," Billups said.

The bottom line is, the current adherence metric does not address clinical outcomes, and therefore patient health is secondary to getting prescriptions filled. The resources within the medical system are all aimed at a single goal: getting the prescriptions called in and, ultimately, picked up. "If those resources could be put somewhere else that would be more impactful in patient care, then it’s not great,” she said. "It's a missed opportunity."

Every time CMS changes their policy, though, it's an opportunity to improve clinical outcomes. Such a change will happen in 2026, when CMS will remove the triple-weighting from refill adherence metrics. Billups hopes pharmacies will take the chance to offer a variety of services to their patients, since a single goal such as automatic refills does not fill every patient need.

When asked what she would like to see changed, Billups said, "If it looks like there's an adherence concern, [pharmacists] can ask questions." As providers with access to patient refill data and who can offer the personal touch, pharmacists are ideally placed to ask patients about their barriers to care and to help remove them.

REFERENCES

1. Billups, SJ. "Applying a clinical lens to medicaton adherence quality measures - do they measure up?" Presented at: American Society of Health-System Pharmacists Midyear 2025 Clinical Meeting and Exposition; December 7-10, 2025; Las Vegas, Nevada.

2. Rodriguez F, Maron DJ, Knowles JW, Virani SS, Lin S, Heidenreich PA. Association of Statin Adherence With Mortality in Patients With Atherosclerotic Cardiovascular Disease. JAMA Cardiology. 2019;4(3):206. doi:https://doi.org/10.1001/jamacardio.2018.4936