Do Outreach Calls Improve Medication Adherence? What Trials Show

Updated October 2026. Written by Matteo Grassi, clinical psychologist and cofounder of HANA Health.

Medication adherence is taking a drug as prescribed, usually measured by the share of days covered by fills. Outreach calls can improve it, but the effect is usually small. In a 2017 randomized trial of 735,218 Medicare Part D patients, pharmacist reminder calls raised strong adherence by about 1 percentage point over a year.

Calls change behavior when they remove a specific barrier, and that's a clinical idea before it's a technology idea.

Key facts

• Largest trial: 735,218 Medicare Part D patients who were 3 days late on a refill were randomized to pharmacist calls or no calls. Refills within 14 days rose by 6.09 percentage points, and average days covered rose by 0.856 percentage points over 365 days (Taitel et al., Patient Preference and Adherence, 2017).
• Single scripted call: In 2,378 adults with diabetes, one call from an educator or pharmacist did not significantly improve adherence, persistence, A1C, blood pressure or LDL (O'Connor et al., Diabetes Care, 2014).
• Automated calls for new statins: Automated calls plus a letter lifted statin pickup from 26.0% to 42.3% among 5,216 patients who hadn't filled a new prescription (Derose et al., JAMA Internal Medicine, 2013).
• Extra nurse calls after three pharmacy calls: Pickup was 25% with the extra calls and 24% without (Fischer et al., JMCP, 2015).
• Fading effects: One telephone counseling session improved statin adherence at 6 months but not at 12 or 60 months (Qvist et al., 2020).
• Big picture: A Cochrane review of 182 trials found inconsistent results, and even the best interventions didn't produce large gains (Nieuwlaat et al., Cochrane, 2014).
• Stakes: Each of the three Part D adherence measures carries a weight of 3 in the 2027 Star Ratings (CMS, 2027 Star Ratings Measures and Weights).

What did the largest trial of pharmacist calls find?

It found a real but small effect. In the 2017 Taitel study, patients who were 3 days late refilling a maintenance drug were randomized to pharmacist calls or usual care. The intervention group had 367,631 patients and the control group 367,587.

The early result was the strong one. Within 14 days of the expected refill date, 6.09 percentage points more intervention patients had refilled. That's the moment a reminder fits best: the patient meant to refill and forgot.

The long run was smaller. Over 365 days, average proportion of days covered rose by 0.856 percentage points. The share of patients at 80% or higher rose by 0.97 percentage points. Persistence over 270 days rose by about 2 days.

Two limits matter. The analysis couldn't tell nonadherence from switching pharmacies, changing plans or death, and the authors worked at the pharmacy chain that ran the program. Treat it as the best large scale evidence we have, not as a promise.

Why did a single scripted call fail in diabetes?

Probably because a short script doesn't fix the reasons people stop. The O'Connor trial randomized 2,378 adults who had just been prescribed a new class of medication for high A1C, blood pressure or LDL. The call lasted under 5 minutes at the median, with up to three attempts.

On intention to treat analysis, the call didn't improve first fills, persistence or the clinical numbers. Results held across age, sex, race, ethnicity and study site. The authors concluded that wide use of this strategy didn't look warranted.

The Qvist trial in Denmark points the same way. A single counseling call 3 months after vascular screening improved statin adherence at 6 months, with a risk difference of 9.8%. By 12 and 60 months the effect was gone.

Here's the pattern. One touch can nudge. It rarely changes a habit.

Do automated calls work, and for whom?

They can work for one specific problem: a new prescription that was never picked up. In the Derose trial at Kaiser Permanente Southern California, 2,606 members got automated calls, followed a week later by a letter if they still hadn't filled. 2,610 members got nothing.

Statins were dispensed to 42.3% of the intervention group and 26.0% of controls, an absolute difference of 16.3 percentage points. The gap in dispensing persisted up to a year. That's a much larger effect than the refill reminder trial, because the target was different. Starting a drug is a one time event with an obvious action, and a reminder can trigger it.

Now the counterweight. In the Fischer trial, patients who still hadn't picked up a new prescription after three pharmacy calls were randomized to extra nurse calls. Pickup was 25% with the calls and 24% without. Those who ignore three calls need something else, and the authors suggested going back to the prescriber.

Here's how the evidence sorts by call type:

• Late refill reminder call from a pharmacist: small gain in days covered, larger gain in refills within 2 weeks (2017).
• Automated call plus letter for an unfilled new statin: large gain in first fills (2013).
• Extra nurse call after three failed pharmacy calls: no change in pickup (2015).
• One scripted call about a new diabetes medication: no significant change in adherence or clinical control (2014).
• One counseling call after screening: statin gain at 6 months, none by 12 months (2020).

I'm leaving out one claim that circulates in this area, that automated IVR calls show no effect. The Derose trial contradicts it for new statin pickup, so the honest summary is mixed, not null.

What does the Cochrane review say about adherence interventions overall?

It says the field hasn't found a reliable fix. The 2014 Cochrane review covered 182 randomized trials. The studies differed so much that the authors couldn't pool them in a meta analysis.

Among the 17 trials at the lowest risk of bias, the interventions were usually complex. They combined tailored, ongoing support from pharmacists or other allied health professionals, often with education, counseling or daily treatment support. Only five of the 17 improved both adherence and clinical outcomes, and no shared feature explained why.

The authors wrote that current methods are mostly complex and not very effective. That review is older than some of the trials above, so newer work may shift the picture. But nothing here supports the idea that a phone call, by itself, closes the adherence gap.

Why do three Star adherence measures make small gains matter?

Because the measures are scored at a cliff edge and weighted heavily. CMS assigns a weight of 3 to each of the three Part D adherence measures (diabetes medications, RAS antagonists for hypertension, and statins) in the 2027 Star Ratings. Most process measures carry a weight of 1. Weights change by year, so check the current technical notes before you plan around them.

A patient counts as adherent when proportion of days covered reaches 80% or higher, as the Taitel paper describes. That threshold is why the 0.97 percentage point gain in that trial matters more than it sounds.

Here's an illustration, with arithmetic. It uses the trial's effect, not a HANA result.

• A practice has 2,000 attributed patients on a measured drug class.
• The trial raised the share at 80% or higher by 0.97 percentage points.
• 2,000 × 0.0097 = about 19.4 patients.
• So the expected gain is about 19 more adherent patients out of 2,000.

Nineteen patients can be the difference between two star bands for a plan, or a bonus tier for a group. They're also under 1% of the panel. Both things are true. For how plans turn Star results into practice payments, see how Medicare Advantage plans pay practices for Star ratings, and for the measures themselves, see the Star Ratings adherence measures.

What this means for your practice

Plan on small, targeted gains, not a broad fix. A few moves follow from the evidence.

• Aim calls at a moment. Late refills and unfilled new prescriptions responded best. Generic check ins did not.
• Ask what the barrier is. Cost, side effects, confusion and forgetting need different answers. A script that only reminds fixes only forgetting.
• Plan a second step. When calls fail, route the patient to the prescriber, a 90 day supply or a pharmacist, as the Fischer authors suggested.
• Measure against the trials. If a program claims much more than the effects above, ask for the control group.
• Connect it to other measures. Adherence work sits beside blood pressure control, covered in closing blood pressure control gaps between visits, and the wider ACO quality measures for practices.

HANA builds patient outreach for practices. We aren't putting a result number on this page, and the trials above are the fair benchmark for any program, ours included. You can read more on the HANA Remote page.

Frequently asked questions

How much can a refill reminder call raise adherence?

In the largest randomized trial, about 0.9 percentage points of average days covered over a year, and about 1 percentage point more patients at 80% or higher. Refills within 14 days rose by about 6 percentage points. Results from other populations and call designs may differ.

Are automated calls as good as live calls?

The evidence doesn't answer that directly, because the trials tested different problems. Automated calls plus a letter worked well for unfilled new statin prescriptions. A live pharmacist call helped late refills modestly. No trial above compares them head to head.

Why do adherence gains fade after a call?

Most trials measured a short window, and the one that followed patients longer, the Danish counseling trial, saw the statin gain disappear by 12 months. Habits and barriers return when contact stops. That's a reason to think about follow up cadence, not one off calls.

Do outreach calls help patients who never fill a new prescription?

Sometimes. Automated outreach worked for first fills of a new statin in one large trial. For patients who ignored three pharmacy calls, an extra nurse call didn't help, so a different route may be needed.

Is a high Star adherence score proof that calls worked?

No. Scores reflect many forces, including prescriptions, benefits, cost and patient mix. Only a randomized or well matched comparison can attribute a change to calls.

Sources

1. Taitel M, Mu Y, Gooptu A, Lou Y. Impact of late to refill reminder calls on medication adherence in the Medicare Part D population: evaluation of a randomized controlled study. Patient Preference and Adherence, 2017. https://pmc.ncbi.nlm.nih.gov/articles/PMC5338963/
2. O'Connor PJ, Schmittdiel JA, et al. Randomized trial of telephone outreach to improve medication adherence and metabolic control in adults with diabetes. Diabetes Care, 2014. https://pmc.ncbi.nlm.nih.gov/articles/PMC4751474/
3. Derose SF, et al. Automated outreach to increase primary adherence to cholesterol lowering medications. JAMA Internal Medicine, 2013. https://pubmed.ncbi.nlm.nih.gov/23403978/
4. Fischer MA, et al. A randomized telephone intervention trial to reduce primary medication nonadherence. Journal of Managed Care and Specialty Pharmacy, 2015. https://doi.org/10.18553/jmcp.2015.21.2.124
5. Qvist I, et al. Randomised trial of telephone counselling to improve participants' adherence to prescribed drugs in a vascular screening trial. Basic and Clinical Pharmacology and Toxicology, 2020. https://doi.org/10.1111/bcpt.13459
6. Nieuwlaat R, et al. Interventions for enhancing medication adherence. Cochrane Database of Systematic Reviews, 2014. https://doi.org/10.1002/14651858.cd000011.pub4
7. Centers for Medicare and Medicaid Services. 2027 Star Ratings Measures and Weights, 2026. https://www.cms.gov/files/document/2027-star-ratings-measures.pdf

This article is educational and isn't billing, coding, clinical or legal advice. Check current CMS guidance and your MAC before you bill.

Want to talk through outreach for your practice? Book a discovery call.