A pharmacy can make hundreds of refill calls and still know very little about whether its adherence program helps the patients it was built to serve. Activity is not an outcome. A useful program measures contact, barriers, actions, and follow-through without turning a refill date into a claim about how a patient takes a medication.
Answer first: measure medication-adherence services as a patient-support process, not a call-volume campaign. Use dispensing data to identify possible gaps, confirm the patient’s situation through a respectful conversation, document the barrier and action, and review a small set of process and outcome signals over time. CMS Part D adherence measures are valuable context, but they do not replace clinical judgment, patient preference, or a pharmacy’s own privacy and documentation obligations.
This article is general educational information, not individualized medical, legal, billing, or payer advice. A refill history cannot establish whether a patient took a medication as prescribed. Use pharmacist judgment, current plan requirements, applicable law, and the patient’s care plan.
Contents
- What a measure can—and cannot—say
- Design the service before the dashboard
- Choose meaningful signals
- Document the intervention path
- Review results without misreading them
- Protect privacy and patient choice
- Owner checklist
- FAQ
Key takeaways
- A refill gap is a prompt for outreach or review, not proof of nonadherence.
- Separate program activity, reach, barrier resolution, and dispensing-based measures.
- Document what the patient says, what the pharmacy did, and the next owner—without overstating the result.
- Use payer measures as context, then verify the definitions and data windows that apply to the program.
- Do not make adherence outreach coercive; patient choice and clinical appropriateness remain central.
What a measure can—and cannot—say
Medication adherence is clinically important, but the measure chosen determines what the team can responsibly conclude. A pharmacy dispensing record can show that a prescription was filled, reversed, transferred, returned, or not refilled within an expected window. It cannot by itself show ingestion, persistence after a clinical change, an adverse effect, affordability, a hospital stay, a prescriber instruction, or a patient’s decision to stop therapy. Those possibilities are why an outreach program needs a conversation and a documented escalation path.
CMS uses adherence measures in its Part D Star Ratings materials, including measures for certain medication classes. Those measures have technical definitions, eligibility criteria, exclusions, calculation periods, and plan-level uses. An independent pharmacy should not assume that an internal refill report duplicates a CMS measure or that a payer’s performance report applies the same logic. Read the current technical materials and the applicable contract or program documentation before promising a result or tying staff compensation to a metric.
For operations, the more useful question is: what did the patient need, what barrier was identified, what action was completed, and what still requires follow-up? A patient who declines a refill after a prescriber-directed change is not a failed outreach. A patient who cannot afford a medication, cannot obtain an authorization, has trouble opening a container, or has an unresolved adverse-effect concern may need a different pathway than a routine reminder.
Design the service before the dashboard
Define the purpose first. A pharmacy may seek to reduce missed refill opportunities, support transitions of care, improve medication synchronization, identify affordability barriers, or meet a documented service obligation. Each aim needs a different eligible population, staffing model, script, documentation field, and escalation rule. One generic queue tends to hide those differences and rewards the fastest contact instead of the most useful intervention.
Build a plain-language operating definition for the team. Specify which patients enter the queue, which events trigger review, who may contact the patient, what information can be discussed, what counts as a completed outreach, and how a pharmacist, prescriber, payer, caregiver, or other resource is involved when needed. Clarify the difference between a reminder, a refill request, a clinical question, a prior-authorization issue, and a safety concern.
| Signal | What it can support | What it cannot prove | Useful next step |
|---|---|---|---|
| Refill gap or late pickup | A reason to review or contact the patient | That the patient stopped taking therapy | Confirm status and ask about barriers |
| Reached patient | Contact was made | That a problem was solved | Document concern, action, and follow-up |
| Patient reports barrier | A patient-identified need | Clinical causation or payer responsibility | Route to the appropriate owner |
| Refill dispensed | A dispensing event occurred | Medication ingestion or treatment success | Continue appropriate support and monitoring |
| Resolved referral | The assigned task was closed | Long-term persistence | Review outcome at the next defined point |
Do not enroll people simply because they are easy to reach. Use risk-informed criteria that fit the pharmacy’s capacity and documented purpose. A small program may start with patients who have a recurring refill gap, a recent unresolved access issue, a requested synchronization review, or a clinician referral. The goal is to make the queue explainable and manageable, not to create an exhaustive list that the team cannot safely work.
Choose meaningful signals
Use a balanced scorecard. Activity signals can include eligible patients identified, outreach attempts, successful contacts, pharmacist escalations, prescriber messages, and referrals completed. Process-quality signals can include time from flag to first review, percentage of cases with a documented outcome, unresolved cases past the pharmacy’s target, and repeat barriers by category. Dispensing-based signals can be reviewed separately using the applicable calculation rules.
Measure patient experience where the pharmacy can do so respectfully. A brief question—whether the contact helped the patient understand the next step or solve a practical problem—can reveal more than a large call count. Do not require a patient to disclose more than is necessary to receive help, and do not represent an operational survey as a clinical outcome study.
Keep denominators visible. A rise in completed contacts may reflect more staffing, a narrower eligible list, a seasonal campaign, or a better phone-number file. A drop in refill gaps may reflect a formulary change, a prescriber change, or fewer eligible patients. Trend review should include the population, time window, data source, and known changes; otherwise the dashboard invites confident but unsupported conclusions.
Document the intervention path
Document enough to make the next step safe. A useful note captures the trigger, contact status, patient-stated barrier when volunteered, action taken, pharmacist involvement, referral or message destination, and follow-up owner. Use objective language. Write that the patient reported a cost concern or said a prescriber had changed the regimen; do not write that the patient is noncompliant, unmotivated, or refusing care unless the record supports a precise and appropriate statement.
Build distinct routes for common barriers. A coverage denial may require a payer or prescriber route. A stock issue may require an inventory or transfer decision. A medication question, adverse-effect concern, or possible therapy change requires pharmacist assessment and may need prescriber communication. A transportation, language, device, or caregiver issue may require a different support service. The route is often more valuable than a polished reminder script.
Close the loop explicitly. Assign a date and owner for unresolved cases, then review the queue before it becomes a record of abandoned work. If an action cannot be completed, record the limitation rather than silently marking the case complete. This protects the patient and produces a more honest measure of what the program can and cannot accomplish.
Review results without misreading them
Review the program monthly or on another deliberate cadence that fits volume. Start with a small sample of cases, not only totals. Can the team reconstruct why the patient entered the queue, what happened during outreach, whether the right person handled the question, and whether the follow-up occurred? Sample review identifies problems in scripts, routing, documentation, and staffing that a summary percentage will miss.
Then examine patterns. Repeated affordability concerns may justify a better benefits-verification or assistance workflow. Repeated unreachable contacts may point to contact-data quality or an outreach method that patients do not prefer. Repeated unresolved clinical questions may show that pharmacist coverage is insufficient at the time the queue is worked. Treat these as hypotheses to test, not proof of a single cause.
Do not use a dispensing-based metric to punish a patient or override a clinical plan. A medication may be paused, replaced, intentionally discontinued, obtained elsewhere, or not appropriate to refill. The pharmacy’s responsibility is to make a safe, respectful pathway visible—not to make every line trend upward.
Use a decision log for measurement changes
When the pharmacy changes a queue rule, report format, contact channel, or follow-up interval, note the date and reason. Without that record, an owner may mistake a design change for a change in patient behavior. For example, narrowing the queue to patients with a documented prior barrier may reduce the number of outreach attempts while increasing the proportion of cases that need pharmacist involvement. Neither movement is automatically good or bad; it must be interpreted against the service purpose.
Set a limited number of review questions before looking at the dashboard: Did patients receive a timely and appropriate response? Did the team document the barrier and next step? Were clinical questions routed to a pharmacist? Did open cases close through a defined action, referral, or patient decision? What prevented completion? This approach limits the temptation to add every available data field and helps the team improve a process that people can actually operate.
Where a pharmacy reports performance to a payer, partner, or employer, preserve the source data, calculation version, and reporting period used for that report. Reconciliation is especially important when a payer record differs from the dispensing system. Investigate the discrepancy through the appropriate contract or data process; do not silently alter patient documentation to make two systems agree.
Finally, give the team permission to flag a measure that no longer supports safe care. A concise report with a known limitation is more useful than a polished dashboard that encourages unsupported conclusions. This protects patients, staff, and the credibility of the service when decisions are reviewed later.
Protect privacy and patient choice
Adherence outreach uses sensitive health information. The pharmacy should apply its privacy policies to calls, voicemail, text, portal, caregiver communication, and documentation. Confirm the right person before discussing details, use approved communication methods, and escalate questions about authorization or disclosure. A successful call does not justify sharing more information than the patient has authorized or the situation requires.
Patient choice also matters operationally. Offer a clear way to decline reminders or choose a preferred contact method where the pharmacy can support it. A patient may decline synchronization, prefer to speak with a pharmacist, or choose to contact the prescriber directly. Record that choice and avoid repeatedly generating the same outreach without a clinical or operational reason.
Owner’s measurement checklist
- State the service purpose, eligible population, and limits in writing.
- Verify payer or CMS measure definitions before using their labels or calculations.
- Separate outreach activity, barrier resolution, dispensing events, and patient-reported experience.
- Create documented routes for affordability, access, clinical, and operational barriers.
- Use objective notes and assign an owner for every open follow-up.
- Review a sample of cases as well as the aggregate dashboard.
- Protect privacy across every communication channel.
- Revise the workflow when recurring exceptions show that it is not serving patients well.
Frequently asked questions
Does an on-time refill prove adherence?
No. It shows a dispensing event, not how or whether a medication was taken. Use it as one signal within a patient-centered review.
Should every patient with a refill gap receive the same script?
No. The script should route the conversation safely, but the follow-up should reflect the patient’s stated barrier, current therapy, preferences, and clinical needs.
Can a pharmacy use a CMS adherence metric as its internal dashboard?
Only after verifying the current definition, eligible population, data source, and calculation rules. An internal report may not match a CMS or payer measure.
Conclusion
A strong adherence program measures useful work without claiming more than the data can show. It identifies a possible gap, listens for the actual barrier, assigns a safe next step, and learns from unresolved cases. For a complementary approach to selecting patients by risk rather than volume, see Dispense Times’ adherence-program workflow article.
