Flu vaccinations have declined 3.4% over the past six years for adults over age 65. One plan wanted to reverse this trend and improve flu vaccination rates for Medicare members. They engaged MedOrion to run a flu program for 225,000 members over a six-month period from October 2025 to March 2026 across a continuously shifting population. This program was designed with a matched control group and saw an overall 1.8% improvement in vaccination rate for the “test” population. MedOrion segmented the population using its proprietary behavioral analysis to target individuals based on motivators to act.

Adherence Rate by Decision Factor

Behavioral segmentation drove the individualized approach, with vaccination rates reflecting how well each segment responded to MedOrion messaging over time. For example, the “At Risk” segment had a 41% adherence rate, despite representing just 13% of the population.

Adherence by Decision Factor

The Results

MedOrion’s test group achieved a 29.1% vaccination rate versus the control group’s 27.4% rate. When looking at performance by risk group, for the Top PCT group, the test group achieved a 45.2% rate vs the control’s 42.3% and for the Bottom PCT, the test group achieved a 13% rate versus the control’s 11.7% rate. 

Our Signal-Based Approach

MedOrion improves your Star ratings through individualized member outreach at scale. Our Electronic Behavioral Record (EBR) combines real-time clinical, behavioral and situational signals with AI-assisted prioritization to create a precise profile of each member. Rather than static, pre-defined journeys, we help you understand why your members are not seeking care and then address their individual barriers. The result is an improved member experience and better health outcomes.

Medication adherence is one of the most stubborn problems in Medicare Advantage, with non-adherence driving poor outcomes, higher costs, and lower Star ratings across the PDC measures for diabetes, hypertension, and high cholesterol. What’s harder to solve is the why behind each individual member’s gap. That’s precisely where most outreach falls short. It is possible to see medication adherence improvement across multiple conditions. It happens when you stop treating non-adherence as a single problem and start addressing it as an individual, multi-causal challenge. 

The Scale of the Challenge

One national plan with 3.5 million eligible members worked with MedOrion to improve adherence in diabetes, hypertension, and high cholesterol. All conditions that have an outsized impact on both member health and plan ratings. The program produced meaningful year-over-year improvements in adherence,with particularly strong conversion rates among members who had previously been non-adherent.

For diabetes, 7.79% of non-adherent members became adherent during the program. The overall population adherence rate improved by 4.82% year-over-year.

For high cholesterol, 6.31% of non-adherent became adherent, with an overall population improvement of 2.78%.

For hypertension, 6.21% became adherent.

Equally important: slip rates (were adherent and became non-adherent) held low and consistent across all three conditions, ranging from 2.44% to 2.56%. Improving adherence is one challenge. Maintaining it is another. The program addressed both.

Why Signal-Based Outreach Works Differently

Traditional adherence programs tend to treat non-adherence as a uniform gap to be closed with uniform outreach. A member misses a refill; they receive a reminder. The problem is that a missed refill can mean a dozen different things, from a cost barrier, to a side effect concern, to a transportation challenge, to a belief that the medication isn’t necessary, or simply a busy week. Generic reminders don’t move members whose barrier is something the reminder never addresses.

MedOrion’s approach operates on three layers of signal that work together continuously.

Real-time clinical signals track what has and hasn’t happened within the Stars measure framework. A prescription that wasn’t filled, a refill gap that’s opening, a PDC rate trending toward non-adherence before the measurement window closes.

Situational signals surface why adherence may be breaking down for a specific member. Like whether cost sensitivity, low care engagement, or recent clinical events are creating barriers that a standard outreach cadence would miss entirely.

Behavioral signals determine how to reach each member effectively. Which channel, what message framing, and what timing will actually move this person rather than a persona model that approximates them.

The outcome of layering these signals is individualized member journeys: the right message, through the right channel, at the right moment; all chosen automatically for each member and continuously refined as outreach outcomes feed back into the model.

What This Means for Stars

For plans trying to hold or improve their ratings, adherence performance isn’t a secondary concern, it’s often the margin between a three-star and four-star rating and the revenue that difference represents. Moving 6–8% of non-adherent members to adherent, at scale, across multiple conditions simultaneously, is the kind of performance that shows up in ratings. And, helps you stay there.

Every year, Medicare Advantage plans pour resources into CAHPS preparation. They analyze prior year scores, identify at-risk populations, stand up targeted campaigns, and coach call center teams. And every year, many of these same plans are surprised when results don’t reflect the effort.

By the time a survey lands in a member’s mailbox, the experience it measures is already over. Last-minute outreach doesn’t reverse perceptions built across a year of touchpoints. The verdict has already been rendered.

What CAHPS Is Actually Measuring

To understand why most CAHPS interventions fall short, it helps to look at what the survey is actually measuring beneath the surface of each question.

Care Coordination asks whether a member’s doctor had their medical records, whether medications were reviewed, and whether the primary care provider was informed about specialist visits. These aren’t soft perception questions about communication quality; they’re questions about whether an annual wellness visit happened and whether the care team functioned as a coordinated unit when it did. Plans that increase annual wellness visit completion rates almost invariably see Care Coordination scores follow. The survey question is a proxy for the clinical event.

Getting Care Quickly asks whether members received care when they needed it. When a member defaults to the emergency room for a non-emergency, they wait hours, feel frustrated, and report a negative experience. When directed to urgent care, they’re often seen within the hour. The care is the same. The experience, and the CAHPS response, is very different. What looks like a satisfaction problem is actually a care navigation problem.

Getting Needed Care assumes that members are actively engaged with a primary care provider. A member with no established PCP cannot report positive experiences about referrals, specialist access, or care continuity because they aren’t seeking care in the first place. Their absence from the care system reads as dissatisfaction on the survey.

To improve CAHPS, you have to change the actual sequence of clinical events, care decisions, and plan touchpoints that a member experiences across a full year.

The Hidden Culprit: Friction

Member experience research consistently points to a concept that health plans underweigh: reducing the effort required to resolve a problem or complete a task is a more powerful driver of satisfaction than any single positive experience.

Member friction accumulates across dozens of annual touchpoints like the welcome call that never comes, the annual benefit notification that’s too overwhelming to understand, the care gap reminder that misses the actual reason the member has the gap. Each friction point is a small withdrawal from the trust account that CAHPS ultimately measures.

Reducing that friction requires identifying where it occurs, understanding why it occurs for each individual member, and intervening before it calcifies into a negative perception.

Five Things Plans Can Do Right Now

1. Make Wellness a Priority The annual wellness visit is the single clinical event that drives performance across care coordination, medication reviews, specialist referrals, and preventive screenings simultaneously. Plans have seen AWV completion drive measurable lifts across three or more CAHPS and HOS domains at once.

2. Stop Running Parallel Campaigns When Stars, quality, and care delivery teams each run their own outreach with their own data, members receive uncoordinated touchpoints from a plan that appears not to know their situation. That perception becomes a CAHPS response. Coordinating outreach logic across functions by ensuring the right department leads at the right moment is foundational.

3. Redirect Emergency Room Utilization Target members with patterns of avoidable ER use and educate them on appropriate care settings before the next urgent need arises. One regional PPO plan moved 1,900 members to more appropriate care settings, generating an estimated $1.25 million in cost savings while improving the member experience.

4. Treat Benefit Confusion as a Care Gap When proactive call center outreach was deployed to members most likely to complain about their experience, one finding stood out: 50% of those members didn’t fully understand their own benefit design. Individualized benefit education delivered through email, bi-directional texting, and responsive call-backs addresses a friction point that no care gap reminder would ever reach.

5. Address Mental Health Barriers Studies show that 20–30% of Medicare Advantage members experience depression or anxiety, with half remaining undiagnosed or untreated. A six-month pilot covering 340,000 members demonstrated that proactively removing barriers to mental health conversations produced a 27.6% new depression diagnosis rate among at-risk members, a 5% increase in AWVs, a 5% increase in cancer screenings, and a 3% increase in diabetic eye exams. None of those were the direct target of the program. They were the natural outcome of a member who felt supported.

The Signal Framework That Makes It Possible

The plans building durable CAHPS performance have moved from campaign-based engagement to a continuous, signal-driven model that operates in three layers.

Clinical signals identify what has or hasn’t happened: a missed wellness visit, a specialist referral that didn’t result in a completed appointment, a prescription that wasn’t filled.

Situational signals reveal why: is a mobility limitation making transportation the real barrier? Is a recent diagnosis creating emotional overwhelm that’s preventing action on a care recommendation?

Behavioral signals determine how to intervene: what tone, what channel, what message framing will meet this member where they are rather than where a persona model assumes them to be.

When a national health plan deployed this approach across a 150,000-member HMO pilot focused on CAHPS recovery, digital-first outreach combined with proactive call center engagement uncovered that half of likely complainers didn’t understand their own benefit design, while the other half faced scheduling, transportation, and access barriers. Separate root causes requiring entirely different interventions. The outcome: a 5% improvement in mock survey rates, a 15% reduction in overall complaints, a two-star improvement per measure, and an overall four-star rating by the end of MY2023. A rating the plan had not expected to achieve.

The Stakes Are Rising

As CMS moves toward a fully-digital Stars ecosystem and NCQA sunsets hybrid reporting, the gray areas that allowed retrospective management of experience measures are disappearing. Rising cut points mean that the difference between a three-star and four-star rating, and the revenue that separates them. is increasingly decided by experience measures that cannot be addressed in the final weeks of a measurement year.

The plans best positioned for this environment are those that have redesigned the member journey itself: reducing friction at the touchpoints that matter, coordinating care delivery across departments, and building a signal infrastructure that understands each member as their needs evolve.

When the journey is designed well, the survey takes care of itself.

Every September, Medicare Advantage plans send a compliance document members rarely read. One national plan decided to do something different by turning the Annual Notice of Change (ANOC) into a CAHPS advantage.

As CMS refines the Medicare Advantage Stars program, the Health Outcomes Survey (HOS) has emerged as a critical driver of plan performance. Unlike process measures, HOS directly captures whether health plans are improving member health over time, making it a unique predictor of plan quality. The mental health components have become particularly significant. Studies show 20-30% of Medicare Advantage members experience depression or anxiety, with half remaining undiagnosed or untreated.

Members with untreated mental health conditions score lower across all HOS domains. And, research from Health Affairs confirms that addressing mental health gaps improves not only HOS scores but also drives measurable gains across HEDIS, CAHPS, and medication adherence measures. For Stars leadership, HOS improvement demands longitudinal engagement that changes behaviors and outcomes. There’s no way to predict which members will be surveyed, so plans must create positive experiences for their entire population.

Read on to find out how one national health plan moved from reactive HOS management to proactive, measure-level execution, and achieved:

• 5% increase in annual wellness visits; 81% of which occurred after the engagement.
• 27.6% received a new depression diagnosis, supporting early identification and care.
• 3.2% higher positive responses to B11A question in the PE VAT survey.

Stopping a Ripple Effect

One national plan knew that members with depression scored significantly lower across all HOS measures, but traditional care management wasn’t moving the needle. These members weren’t just scoring lower on mental health questions, it was creating ripple effects across physical health status measures, preventive care, and quality of life assessments. The plan launched a six-month pilot that began in July 2025 for 340,000 members identified as at risk of scoring low on the IMMH HOS measure. The strategy: proactively engage all members with personalized support, removing barriers and normalizing mental health conversations.

To move HOS scores, the strategy was anchored in the following principles:

• Members with or without a diagnosis of depression face different barriers to action.
• Members without a depression diagnosis would have a different engagement strategy than those that had a depression diagnosis.
• Beyond clinical indicators, the program needed to overcome emotional blockers like stigma, low self-efficacy, and lack of resource knowledge.
• Engagement must feel supportive, not intrusive and focus on calls-to-action about improving mental health.

The Decision Logic

MedOrion’s approach is to tackle the measure itself, with member engagement as a natural output of that strategy. “Predictive analytics identifies risk, but that’s only the starting point,” says Shai Levi, Chief Product Officer at MedOrion. “To improve HOS, we had to identify what specific barriers each member faces. And then, for this program it was about encouraging proactive conversations with their primary care provider. Helping members recognize early warning signs that were worth discussing. And, providing reassurance, resources and clear next steps. “

First, the program identified each member’s clinical profile. Some have diagnosed depression needing treatment support. Others have multiple chronic conditions increasing depression risk. The next layer is understanding what the circumstances are within that clinical profile. What’s happening in each member’s life right now? Newly diagnosed with a chronic condition? Recent loss of a spouse? Managing multiple conditions while caregiving? Stopped taking medications or missed appointments? These factors drive timely, relevant support.

After identifying both the clinical and situational indicators of any measure, the program needs to tackle the behavioral profiles of a member. Does stigma prevent seeking help? Do they lack knowledge about covered benefits? Are they struggling with transportation or digital literacy barriers? These inputs, overlayed with embedded Stars measure logic, create a unique outreach strategy for each member based on a series of approved plan messages and channels.

Every Member, Every Time

The program began with digital outreach to all members, using behavioral signals to determine optimal message, timing, and channel. For members with significant unmet mental health needs, proactive telephone outreach uncovered barriers:

• Low Self-Efficacy members (16%): Overwhelmed managing multiple conditions, didn’t know where to start
• Resourced Members (22%): Had means but lacked awareness that mental health benefits were fully covered
• Chronic Illness and Isolation (34%): Wanted support but faced barriers like rural location or transportation
• Engaged with Limited Resources (28%): Actively managing health but struggling with literacy or digital access

MedOrion created tailored pathways for each segment. For low self-efficacy members, messages emphasized small steps and normalized conversations. For resourced members, communications highlighted ease of access. For isolated members, outreach focused on telehealth and transportation assistance.

From Diagnosis to Treatment

This targeted program successfully normalized mental health conversations and drove sustained behavior change. For the 27.6% percent of the population without a previous depression diagnosis, receiving one during this engagement is a critical first step towards long-term improved outcomes.

“Getting members to open up about mental health is the foundation,” explains Dave Burianek, Chief Strategy Officer at MedOrion. “Diagnosis leads to treatment, treatment leads to better outcomes, and better outcomes lead to improved HOS responses. But none of that happens without first removing behavioral barriers.”

Beyond that direct mental health impact, the membership showed meaningful improvements in health behaviors correlating with positive HOS outcomes:

• For the initial “low efficacy” segment, along with an 8% increase in annual wellness visits, there was a 5% increase in breast cancer and colorectal cancer screenings and a 3% increase in diabetic eye exams.
• For those experiencing chronic illness, there was an improvement in getting specialists appointments. And for those that had limited resources, they also saw a 3.5% increase in annual wellness visits.

These cross-measure improvements reinforce that addressing mental health creates a halo effect across the entire member experience, which is exactly what HOS measures.

The future of CAHPS success requires appropriate clinical management at the member level, which translates into better experiences. Care Coordination questions ask if the doctor has medical records, if the member has discussed medications, and if they knew about specialist care. Those are all required components of an annual wellness visit. Annual wellness visits create the interventions that CAHPS Care Coordination measures. Getting Care Quickly asks “When you needed care right away, did you get it as soon as needed?” When members are using the emergency room for non-emergencies, they wait hours. When they use urgent care instead, they are often seen more quickly. Different care setting. Different perception. Different CAHPS response. Getting Needed Care questions assume members are actively seeking care. If they’re not engaged with a primary care provider, they can’t report positive experiences about access. In order to do that, members need help in nudging them to receive appropriate care in the first place.

Approaching CAHPS starts with understanding what member behavior drives positive responses. In order to create positive responses, and experiences, you first need to address member barriers before they turn into negative experiences. Find the barrier. Improve the outcome.

A Regional PPO Takes on Annual Wellness Visits

Knowing that annual wellness visits are a key positive indicator for Care Coordination, that urgent care use could impact Getting Care Quickly and that nudging members towards seeking care would help address Getting Needed Care, a regional PPO plan ran a 2025 program that set out to:

• Educate about annual wellness visits to improve medication reviews, specialist coordination and medical record reviews.
• Refocus ER utilization by re-directing members to more appropriate care settings.
• Improve the relationship between members and their primary care providers.

The Signal Strategy

Since annual wellness visits are not a measure in and of themselves, but drive the performance of multiple other measures, MedOrion’s approach was to focus on finding the barriers to care.

“Annual wellness visits are a strong indicator of the overall health of that individual,” says Dave Burianek, Chief Strategy Officer at MedOrion. “Those that are low utilizers of overall care tend to present with some really complex health issues when they finally do seek care. Increasing annual wellness visits is about getting ahead of future problems down the road. People think, ‘I’m feeling fine. I don’t have time to go in.’ We really need to start breaking down that assumption.”

First, the program targeted members with no annual wellness visits and “low-utilizers,” those defined as having multiple open gaps, minimal prior engagement and sporadic healthcare use. Then, behavioral analysis looked at inputs based on income, language, and access as well as comfort with digital communications. Together they create an individualized profile that targeted the care barriers for each member.

The Results

• An estimated $1.25 million in cost savings.*
• 2992 members completed an annual wellness visit, and there was a 2% increase in primary care visits for the hardest to reach population.
• 1,900 members were redirected away from the emergency room to more appropriate care settings

As member journeys become more clinically complex, getting members to do what you need them to do isn’t just about identifying gaps and moving members to take action. Member engagement success today requires coordination across quality, experience and operations teams that is strongly anchored in organizational priorities and clinical outcomes. The need for high-level coordination also requires a different level of partnership. We understand the challenges of today’s Medicare Advantage market, and work with plans to solve Stars issues at the measure level, freeing you to focus on more strategic priorities, with the confidence that we’ll improve your measure scores and overall Star rating, one member at a time.

*Beckman AL, Becerra AZ, Marcus A, et al. Medicare Annual Wellness Visit Association With Healthcare Quality and Costs. Am J Manag Care. 2019;25(3):e76-e82.

Study finding: 5.7% reduction in total healthcare costs over 11 months following AWV.

Nearly a decade after CMS launched the Meaningful Measures Initiative, quality measurement has fundamentally shifted from retrospective, process-based reporting to outcomes that reflect real patient experience and clinical impact. Meaningful Measures 2.0 and NCQA’s acceleration toward Electronic Clinical Data Systems (ECDS) signal an irreversible move toward fully digital, patient-centered measurement built on electronic clinical data, patient-reported outcomes, and real-time performance feedback.

As CMS targets a fully digital Stars ecosystem by 2030 and NCQA sunsets hybrid reporting by 2029, plans are already seeing material performance divergence exposed by ECDS transitions. At the same time, rising Stars cut points and the outsized financial influence of triple-weighted measures have narrowed the margin for error, making accuracy, speed, and data accountability mission-critical for plans reliant on bonus payments to fund benefits and member experience investments.

This evolution has also changed the rules of member engagement. Static, campaign-based outreach models cannot keep pace with members whose clinical needs, access barriers, and experience drivers shift continuously across the year. As experience and patient-reported outcomes such as CAHPS take on greater weight, projected by some analysts to approach 40% of Stars by 2029, plans must move toward real-time, signal-driven, coordinated engagement strategies.

By combining real-time clinical signals with behavioral and situational insights, and applying AI-enabled analytics across quality, pharmacy, and care coordination, plans can identify the member’s needs sooner, determine the next best action, and intervene without increasing member abrasion. In fact, the plans best positioned for the next phase of Stars will treat member engagement as a dynamic, precision-driven capability that at its best can anticipate gaps by:

• Leveraging real-time signals allowing for quicker interventions.
• Utilizing AI and advanced analytics to create deeper insights and faster decisions.
• Individualizing outreach and coordinating care to move the member to action.

Rebuilding Member Engagement

Member engagement hasn’t evolved much from the early days of preventive screening and gaps reminders. While today’s approach now includes “digital channels” like text and email in addition to mail-based and call center outreach, it still relies heavily on predictive analytics and persona-based outreach that cannot adapt to rapidly changing health status or identify the real reasons gaps remain open, often catching risk too late. And many organizations still run multiple campaigns targeted at individual members, increasing abrasion. Successful member engagement programs require a different level of systems logic built across departments that operates on real-time clinical, situational and behavioral signals that adjust with and understand the member’s changing needs.

Complex Clinical Signals

Clinical signals based on data and member interaction monitoring are foundational to all predictive analytics models, identifying when something has or hasn’t happened, like usage of a concurrent medication or missing a mammogram. However, without an overlay of Stars measure logic, they cannot address the “blind spots” inherently built into the Stars measures themselves.

Each Stars measure requires a different strategy, and this is especially true when harmonizing claims data and pharmacy activity. Traditional “personalized engagement” strategies will look for a gap and outreach to the member on that gap but lack inherent understanding of the measure itself. For example, a member who is currently on anti-anxiety medication and has a previous history of falls and hospitalization will be at risk for Poly-ACH. Proactive outreach to that member about the risk of taking opioids with an anti-anxiety can work to prevent the member from entering the denominator in the first place if they find themselves hospitalized again. For those that do enter the denominator, immediate triage needs to take place to prevent them from entering the numerator.

AI-assisted outreach becomes a powerful tool in this new era of complex clinical conditions allowing plans to coordinate measure outreach not only based on the complexity of a single measure, but also for multiple measure gaps. By coordinating these efforts across departments, and creating a more wholistic view of the member, plans not only maximize increasingly scarce resources, but create a member experience that naturally reduces abrasion and is more impactful.

Assessing the Situation

Clinical signals enhanced with situational signals provide instructions on not only what to ask the member to do, but how to do it. Overlaying the initial clinical signal with previous claims data, member complaint logs, care coordination call insights and previous survey responses creates a better understanding of the underlying situation for each member, improving the efficiency of outreach.

Situational signals can enhance existing persona work by going beyond basic segmentation to infer context. For example, signal systems can take members with a medication adherence gap and:

• identify a subset of people that are not engaging with the “90-day mail order pharmacy is convenient” messaging
• have claims activity pointing to a disability
• infer that for those members it may be an actual mobility and transportation issue, not a convenience issue
• tailor messaging towards transportation and access.

These situational signals can also inform the right mechanism for outreach. In the previous, concurrent medication example, AI-assisted outreach can identify that the member has entered the numerator and needs an immediate phone call rather than an informative letter.

Understanding the Behavior

Behavioral signals round out the member engagement puzzle as they help uncover underlying motivations for why the member is struggling to close the gap. They also help inform the right tone and best mechanism for outreach.

Is the member overwhelmed by a recent diagnosis and struggling to cope with filling prescriptions? Are they struggling financially and unable to afford medication or appointments? Do they actually feel fine and don’t see the need to go to the doctor or fill a prescription? These nuances help cater messaging in a way that shows the member they are understood and meets them in the moment.

Particularly in the member experience Stars measures, “personalization” needs to move beyond language of preference and preferred outreach channel to a deep understanding of each member as they change in real-time with health status.

The Power of Care Coordination

As member journeys become more clinically complex, getting members to do what you need them to do isn’t just about identifying gaps and moving the members to act. It’s about providing coordinated resources that support the action.

A fully operating signal system identifies, for example, the absence of a colorectal screening, understands that the member is feeling overwhelmed by the thought of a colonoscopy and can also recommend and deliver an at-home FIT test to efficiently close the gap. It can also focus on coordinating with primary care providers to help bridge the gap between them and specialists, as members notoriously forget or omit things between visits.

It Starts and Ends with Signals

As the health care industry shifts away from retrospective, process-based reporting to fully digital, outcomes-driven measurement centered on patient experience, clinical quality, and value-based care, plans leveraging real-time claims data paired with deep member insights will outperform their peers.

Static segmentation, retrospective outreach and siloed measure campaigns can’t address the growing clinical complexity of members and their rapidly changing expectations and needs. The future of member engagement is signal-driven, coordinated and patient-centric. By embracing AI-assisted analytics and harmonizing quality, pharmacy and care coordination efforts, health plans can prioritize the most impactful gaps, reduce member abrasion and improve outcomes at scale.

MedOrion was recently featured on BrightSpots in Healthcare’s Feb. 12 episode, Beyond Segmentation: How Medicare Advantage Engagement is Being Rebuilt. The conversation focused on how health plans are moving beyond static segmentation and retrospective reporting towards signal driven member engagement that support better decisions across quality, experience, and operations. This episode is designed for leaders who recognize that the engagement approaches that worked even a few years ago are no longer sufficient. As Medicare Advantage continues to evolve, Stars measures are shifting away from administrative mechanics and toward clinical outcomes, pharmacy performance, and CAHPS. That shift is forcing a deeper rethink of how plans identify priorities, engage members, and translate insight into action.

As CMS accelerates the shift towards experience-driven Stars measures, improving CAHPS has become one of the largest controllable drivers of Medicare Advantage revenue, retention and competitive advantage. Some industry analysts suggest that experience measures will comprise nearly 40% of Stars performance by 2029.

CAHPS captures where member experience falls short in real life, bringing meaningful impacts to both revenue and member retention. Last minute interventions don’t solve the perceptions members have built over time. Retrospective chart chasing is no longer an option as measures become electronic. And, there is no way to tell which members will be surveyed.

Improvement in these experience measures demands cross-functional alignment across service lines and consistent communication. The same message for all members will no longer get you there, and member personas often lack the unique situation of the “person.” AI-generated content is not yet mature, and the associated regulatory concerns make customized approaches challenging for marketing and legal departments.

For Medicare Advantage plans facing CAHPS-driven Stars risk, MedOrion provides a measurable, scalable path from reactive management to measure level execution. This case study explores how one plan achieved CAHPS measure improvement.

Sinking Stars

2023 was an inflection point for one national health plan. Having lost their 5-star rating, they knew that CAHPS, especially Getting Needed Care and Care Coordination had driven the decline, but it was hard to know how to reverse it. “PCP Informed about Specialist Care was one of their consistently lowest performing items,” recalls Dave Burianek, Chief Strategy Officer at MedOrion. “The health plans own internal Stars Predictive Model predicted at best they would do 3.5 stars for MY2023.”

Beginning with a 150,000 member pilot for their HMO population, MedOrion came with a strategy that seems simple, but can be incredibly complex to execute, they would need to address all measures for all members.

CAHPS Measure Improvement

CAHPS measures are difficult to address because there is no way to pro-actively identify who receives a survey. To create a meaningful lift, they would need to impact a broad population, not just those at risk for low scores. To move the needle on Stars, improving CAHPS meant targeting all members to address complaints before they happen, with a strategy anchored in three truths:

• Each measure needs to be treated uniquely.
• Each member needs to be treated uniquely.
• Create the least member abrasion possible.

The Decision Logic

MedOrion’s approach is to tackle the measure itself, with member engagement as a natural output of that strategy.

“Predictive analytics is baked into all member engagement programs, including ours. But, the solution needs to go beyond just predicting who might do what. For CAHPS, we go beyond just trying to identify who is most likely to score low, to figure why they will score low, which we refer to as a ‘situational indicator.’ And, then the potential barrier to closing that gap, which is something we refer to as a ‘behavioral indicator.’ The interventions are then tailored accordingly, and we’re continually monitoring those members and can quickly identify when something has changed and we need to intervene again,” says Shai Levi, Chief Product Officer at MedOrion.

The Clinical Indicator

First, the program needs to identify what clinical profile members will have within the measure. For Getting Needed Care, for example, some members will have a chronic condition and need to see a specialist. Some members will be healthy, but need preventative screening. The way those members are approached will be different.

The Situational Indicator

The next layer is understanding what the circumstances are within that clinical profile. Does the member need a specialist because they are newly diagnosed? Are they presenting with multiple conditions? Do they need a preventive screening but have a new diagnosis? Are they recently out of active treatment and don’t know they need to go back to regular screening?

The Behavioral Indicator

After identifying both the clinical and situational indicators of any measure, the program needs to tackle the behavioral profiles of a member. Is the member overwhelmed with a recent diagnosis? Are they feeling fine and don’t thing they need screening? Are they struggling to get transportation to appointments?

These inputs, overlayed with embedded Stars measure logic, create a unique outreach strategy for each member based on a series of approved plan messages and channels.

Every Measure, Every Member, Every Time

Since low satisfaction in areas like benefits communication and access to care negatively impact CAHPS, success means first tackling complaints before they happen. The start of the program was simple, begin outreach to all members digitally using signal equations to determine the best message and method. With the decision logic in place, members would receive personally relevant messaging. And an increased focus would be placed on those most likely to complain. For that population, the engagement would also include a phone call.

Focus on Complaints

The call center reached out to individuals most likely to complain to ask about their current experience with the plan using an anticipated barrier model. 50% of those contacted complained about not fully understanding their benefit design. The remaining half struggled with a combination of scheduling appointments, getting the correct tests and treatments or scheduling transportation. A smaller subset had unrelated complaints that are were not addressed on the CAHPS survey.

Armed with that information, MedOrion created a decision model targeting members on unique aspects of Getting Needed Care, with the test group performing better than control on “access to services,” “logistics, “quality – plan services” and “quality service provider.”

Early Indicators

The plan saw a 5% improvement in mock survey rates, with a 15% reduction in overall complaints. By September 2024, the plan saw a 2% improvement in GQC for the test group and a 5% improvement in GNC. The control groups, in contrast, saw less than 1% improvement in both measures. Specialist visits from May – August exceeded the control group every month.

By the end of MY2023, the plan had moved two-stars per measure (2024 CAHPS season), and achieved an overall 4-star rating for MY2023 beating their predicted expectations, which they maintained in MY2024.

Going All In

In 2025, the plan decided it was time to apply the success of the pilot to their entire membership. Leveraging MedOrion’s unique measure framework, real-time data refreshes, early indicators and AI-assisted prioritization, they began individualized outreach tied to claims, diagnoses and behavioral cues.

Early Results

To determine year-over-year lift, the original test group was followed continuously through 2025, as the program was rolled out to everyone. That subset saw additional increases in care coordination, improving an additional .4%, double the prior year. The previous control group now on the program improved by .7%.

Both groups also improved in Getting Needed Care with the previous control group improving .6% and the original test group improving more than double with a 1.4% increase.

How We Can Help

As member journeys become more clinically complex, getting members to do what you need them to do isn’t just about identifying gaps and moving members to take action. Member engagement success today requires coordination across quality, experience and operations teams that is strongly anchored in organizational priorities and clinical outcomes. The need for high-level coordination also requires a different level of partnership.

We understand the challenges of today’s Medicare Advantage market and can help improve your measure scores and overall Star rating, one member at a time.

In 2025, Medicare Advantage plans encountered two new measures in the Part D Star Ratings Program related to medication use: Concurrent Use of Opioids and Benzodiazepines (COB) and Polypharmacy: Use of Multiple Anticholinergic Medications in Older Adults (POLY-ACH). These medication safety measures provide an important signal to health plans, pharmacy teams and provider groups about what it will take to maintain high performance in the years to come. The change could place millions in quality bonus payments at risk for plans that rely heavily on traditional adherence and retrospective review models.

These measures are designed to address the growing population of seniors at risk for health complications due to concurrent or multiple drug use. The COB measure takes on the increased risk of visiting the emergency department, being admitted to the hospital for a drug-related emergency, and dying of drug overdose.¹ And, POLY-ACH the increased risk of dementia, cognitive disorders and other induced brain changes that come with use of anticholinergic medications². As inverse measures, each patient that hits the numerator criteria counts as a failure, and once a patient reaches that threshold, they remain in the numerator for the measure period unless they qualify for an exclusion, like hospice.

A member’s health status can change rapidly and for health plans tackling these measures at scale, real-time proactive monitoring, timely member outreach, and provider and pharmacy engagement are critical keys to success.  

Understanding these measures through the eyes of the member

For both measures, members enter a denominator first, meaning they have not yet failed the measure, but are at an increased risk for doing so. The denominator for these measures is built on two circumstances:

1. Two or more prescription fills for a qualifying medication (opioids, anticholinergic, benzodiazepines, etc.) on different service dates, and
2. Meeting cumulative days of supply thresholds or age requirements

The member fails the measure, and are counted in the numerator, if they meet the following criteria:

1. COB: 30 or more cumulative days of overlapping opioid and benzodiazepine use.
2. Poly-ACH: 30 or more cumulative days of overlapping use of two or more unique anticholinergic medications, with at least two fills of each.

These measures are particularly challenging as members are often long-term users of certain medications without fully understanding the cumulative effects of concurrent use. Or, an acute event, such as hospitalization, can cause them to fall out of compliance with the measures very quickly. Anticholinergic medications, many of which are available over the counter as antihistamines, sleep aids, motion sickness medications and cold medications are difficult to track through medical records or claims data as they do not generate formal prescriptions, or claims, for health plans to mine. A member who was recently hospitalized for a fall and prescribed opioids to manage pain may not properly report their regular use of depression or anxiety medication. Due to the abrupt nature of these changes, members may move into non-compliance simply with a single prescription fill.

Why traditional medication adherence programs don’t work

Traditional medication adherence programs offer a straightforward approach to medication management. There is no debate about the efficacy of the medication prescribed or the risks involved. The issue centers around member’s non-compliance with the doctor’s guidance. Traditional programs address this challenge by nurturing the member along a very linear path encouraging them to take their medications as prescribed.

These new measures bring an additional layer of complexity as they tackle drug interactions that could potentially have an impact on the member, that may also be prescribed by multiple providers. In a peer-reviewed study, as many as 51% of older patients were taking medications that were not recorded with their primary care physician.³  With providers sometimes unaware of the concurrent prescriptions, it removes them as a first line of defense, leaving the pharmacist or member on the front lines to challenge the doctor’s prescription. The pharmacist is put in a position to negotiate with the member to return to their doctor for guidance. And the member may feel that they are challenging the doctor’s guidance.

These new measures require knowing not only whether the member is taking the medication, but also if they are at risk for taking concurrent medications. Advanced logic must be built into any claims monitoring to accommodate these variables, while also considering the distinct role of the provider in these measures.

This environment poses critical challenges for compliance as the member must be informed, the care coordinated, and the systems in place to monitor the patient. Because health plans are often the only part of the larger system equipped to fill in all the knowledge gaps, they become the default integrator in a fragmented care system.

Engaging with Members: Decision Intelligence Designed for Poly-ACH and COB

Members that understand the effects of taking concurrent medications and are approached in a manner that is most meaningful to them are more likely to comply with the measure requirements. Talking through relevant substitutions, preparing them for and offering support to address withdrawal symptoms and reinforcing the need for ongoing communication with their provider can empower members to stay active in their medication management.

Key components of an effective member engagement strategy include:

1. Real-Time Decision Intelligence: Continuous evaluation of new claims that dynamically assign members to the correct clinical outreach stage accounting for those at risk for entering the denominator and those who are in the denominator but have not yet reached 30 days prevents numerator entry. It is also critical to create a strategy for “non-avoidable” members who will need step-down therapy and thus automatically enter the numerator.
2. Clinical Stage Calibration: Appropriate member segmentation in to “at risk,” “awareness,” and “highest risk” to deliver the right clinical and communication intervention at each stage reducing member abrasion and ensuring accurate messaging.
3. Medication Class Segmentation: Messages tailored to the exact medication class as effective outreach keeps in mind the differences between muscle relaxants and pain medication overlap vs antihistamine and antidepressant overlaps providing clarity and priority in communication.
4. Care Coordination Support: Messaging prompts that support overall care coordination and reduce risk, including encouragement to use one pharmacy, dispose of discontinued medication and to bring a full medication list to all doctor’s visits to support providers and reduce provider abrasion.
5. Event-Based Outreach: One-time urgent member outreach within the first 7 days of overlap to provide a three-week intervention window.
6. Real-Time Reporting: Continuous insights into which members have entered which clinical stage, the progression rates, prevented transitions into the numerator and overall overlapping day reductions inform adjustments and decisions.

Metrics That Matter

There are several key performance indicators that plans should track to best address these new measures:

Risk Identification:

1. The rate of new concurrent use within 30 days of post-discharge or a recent medical procedure or surgery.
2. Social Determinant of Health (SDoH) factors that could positively correlate to a high likelihood of reaching 30 days of overlap including access to mental health providers and opioid prescribing rate).

 Prescribing Behavior:

3. The percentage of concurrent use initiated by the same provider (move to provider section).

 Operational Exposure:

4. Members with mail order pharmacy as they are more likely to fill 90-day refills.
5. Strategies for members in avoidable vs. non-avoidable medication classes, stratifying the top medication classes responsible for the highest cumulative overlapping days.

One national plan running engagement programs related to these measures saw an immediate uplift in call center volume from members seeking guidance within weeks of starting their first member engagement campaign.

As CMS moves away from operational measures to focus on better care coordination, it will become increasingly difficult for plans to achieve higher Stars scores. The shift to triple-weighted COB and Poly-ACH measures illustrates that medication safety is no longer just about adherence, but navigating a complex environment of concurrent prescribing, fragmented care and rapidly shifting member risk. Plans that have strong alignment around resilient quality programs, operational efficiency, clinical excellence and best-in-class member experiences will be able to drive measures performance and do so in a way that keeps members happy, healthy and loyal for years to come.

MedOrion helps plan operationalize individualized engagement at scale bridging clinical measure logic and behavioral intelligence with AI-assisted member outreach to improve Stars ratings and improve member outcomes. To find out how MedOrion’s industry-leading technology platform can help your plan, request a demo today.

Fill out the form below to have a copy of this emailed to your inbox.