How to Evaluate the Return on Investment of AI in MTM

Medication therapy management (MTM) puts pharmacists at the center of helping patients use medications safely and effectively. Through Medicare Part D and other programs, pharmacists review each patient’s regimen, address therapy problems, and reinforce adherence over time. CMS requires plans to provide an annual comprehensive medication review, which includes a pharmacist consultation and a written summary for the patient. Across the U.S., non-optimized medication use contributes to an estimated $528.4 billion each year in preventable complications, hospitalizations, and premature deaths.¹ The need for MTM is clear.

Investing in Better Care Transitions Leads to Better Outcomes

Most breakdowns in medication management happen at transition points: when a patient is discharged, when they pick up their first prescription, when they miss a refill, or when documentation is slow and incomplete. These are the critical pressure points were investment in strong MTM processes, supported by AI, can generate measurable returns across four key outcomes:

Fewer downstream medical events

One of the strongest signals of MTM’s value is its ability to prevent downstream medical events. In one Medicaid program, pairing AI with pharmacist phone calls made a big difference. When AI was used to guide telephonic medication management, total cost of care dropped by nearly 20%, and patients had fewer ER visits and hospital stays.²

Better performance on Star Measure

Adherence measures for diabetes, hypertension, and cholesterol medications are important for Star Ratings. Higher ratings strengthen a plan’s eligibility for quality bonuses and increase rebate percentages in Medicare Advantage. A large-scale study at Walgreens implemented an AI-powered adherence program that targeted patients filling medications for diabetes, hypertension, and statins. Among previously non-adherent patients, adherence improved by 23.9% for diabetic medications, 17.9% for blood pressure medications, and 25.3% for statins in the AI-supported group compared to those without AI targeting.³

Smarter Documentation, More Time for Care

With a streamlined CMR process and better documentation, pharmacists can complete more reviews or devote extra time to direct patient care. Studies of ambient AI scribes in outpatient care show meaningful efficiency gains, with less time spent on notes and lower documentation burden,⁴ findings that apply directly to pharmacist workflows. Applied to MTM, those gains can multiply pharmacist capacity and strengthen ROI.

Better Patient Experience Builds Better Adherence

When pharmacists have the time to truly listen, patients notice. By using AI to reduce documentation burden , pharmacists can reallocate that time to meaningful conversations: answering questions, addressing concerns, and tailoring advice to each person’s needs. The result is a care experience where patients feel understood and supported, which makes them more likely to stay adherent to their medications.

By investing in better care transitions MTM programs can reduce harm, improve adherence, strengthen Star Ratings, and deliver measurable ROI.

The CMS Enhanced MTM (EMTM) model, which ended in 2021 after a five-year test, offers a valuable benchmark. The program produced no statistically significant savings in Medicare Part A or B spending, and prospective and performance payments outweighed the small, non-significant reductions observed.⁵ It's clear that without precise targeting of patients and outcomes, even carefully designed MTM programs struggle to deliver ROI.

The Areas of MTM Where AI Can Create the Most Value

A single complete medication review (CMR) can be surprisingly time-intensive. Reports from practice settings estimate that a full CMR, including documentation, requires about 30 to 45 minutes.⁶ Meanwhile, ambient AI documentation tools in outpatient care have already shown measurable efficiency gains, reducing time spent in notes while improving provider experience and clinical notes.⁷ If these gains translate into MTM workflows, the implications are significant. Trimming just 10–15 minutes from each CMR could let a full-time pharmacist complete 4–6 more reviews per day without working longer hours, or redirect that time toward higher-value clinical care.

AI-powered data analytics are transforming MTM by moving beyond broad, rule-based screening toward more precise identification of patients at highest risk. Instead of relying solely on blunt criteria (such as age, number of medications, or chronic conditions). AI can synthesize diverse data streams, including claims histories, EHR encounters, and pharmacy fill patterns. This allows pharmacists to detect subtle signals of medication risk or early warning signs of non-adherence long before they surface in traditional reporting. Pharmacy teams and health plans can easily create a prioritized list that helps them focus their time on the patients most likely to benefit from MTM. Over time, this kind of precision targeting can reduce wasted effort on low-value encounters, improve adherence outcomes, and sharpen the impact of MTM programs on Star Ratings and total cost of care.

Of course, AI isn’t risk-free. The main risks are overreliance on automated outputs, bias in predictive models, and gaps in data quality. These can be managed with human-in-the-loop review, ongoing validation of AI performance, and structured documentation that can withstand audits.⁸

Proving ROI in AI-Driven Medication Therapy Management

When designing MTM interventions with AI, it’s essential to use evaluation methods that hold up under scrutiny. That means moving beyond anecdotal success stories and adopting rigorous approaches such as stepped-wedge rollouts across regions (to compare early vs. later adopters), difference-in-differences analyses against similar members not exposed to the AI, and cohort matching by risk score and baseline adherence. Outcomes should be measured in a hierarchy, starting with process measures like proportion of days covered (PDC), then advancing to ED visits, hospitalizations, and ultimately total cost of care.

Most organizations see workflow improvements, such as reduced documentation time, within the first few months. Clinical outcomes and Star measure improvements typically take 6–12 months to appear, since adherence and utilization trends need time to shift.⁴

Proving ROI in MTM means moving past anecdotes and showing, with rigorous methods, that AI can identify high-risk patients, improve adherence, and reduce costly utilization.

The CMS EMTM model shows why rigor matters. Success in MTM comes from reaching the right patients at the right moments, where the biggest impact can be made. Below are common pitfalls that can undermine those efforts and limit the returns on MTM investments:

Casting too wide a net

The CMS EMTM model showed that broad outreach doesn’t automatically save money. To move the needle, you need to zero in on members with high event risk and clear opportunities to fix medication problems.

Counting tasks instead of results

AI can speed up CMR documentation, freeing pharmacists’ time to reinvest into deeper patient engagement and better outcomes. That’s where the real ROI lies, in improved adherence and reduced hospital use. For perspective: in 2026, adherence carries triple the weight of CMR completion in Star Ratings (3 vs. 1).

No data backbone

If your patient notes are stuck in free-text notes, you can't learn from them or improve over time. Standardized fields and structured templates turn documentation into usable data for trends and predictions.

Underusing AI

Being overly cautious can leave valuable tools sitting idle. AI that helps structure notes, flag high-risk patients, or streamline outreach can be introduced incrementally with pharmacist oversight. The risk isn’t just misuse, it’s also missed use.

MTM is critical for keeping patients safe, yet it too often falls apart during transitions like discharges or missed refills. AI helps pharmacists step in at the right time, cut down on paperwork, and connect more deeply with patients, leading to better adherence, fewer ER visits, stronger Star Ratings, and a clear return on investment.