Measuring ROI Has Always Been a Struggle for Health Plans and Vendors. Now the Stakes Are Higher Than Ever. How Do We Credibly Show Value?

One of the most critical variables to consider when comparing potential solutions for Medicare Advantage plans is how various vendors measure their performance and demonstrate value. Measurement is really hard. It requires dedicated data science and actuarial expertise, in addition to supporting  data infrastructure and processes. Even if these things are available,  it can be extremely difficult to get to a true apples-to-apples comparison and to credibly evaluate the value that vendor point solutions  bring to the table.  

Complexity is further heightened as value-based care is increasingly dominating the Medicare landscape, reshaping the relationship between vendor, provider and payer., Value-based arrangements encourage providers to deliver more effective services, reducing medical cost for both patients and the healthcare system as a whole by aligning financial rewards with quality of care While some argue that value based care arrangements have not delivered on their promises, other’s tout their success. 

With CMS’s Medicare Advantage and Part D Final Rule for CY2025, supplemental benefits have entered the value-based arena. The Final Rule implements stringent standards for supplemental benefits for the chronically ill (SSBCI), which now must meet a legal requirement for improving the health and well-being of its enrollees. As a result, the stakes for successfully measuring and reporting value have been dramatically raised, both for health plans and their vendor partners.

Moving forward, supplemental benefits must work and have credible proof of value.

As part of reasonably expecting that each chosen benefit improves or maintains the health or overall function for SSBCI enrollees, MA plans are required to maintain a bibliography of research studies to back up that belief. Consequently, it’s likely that vendors will need to increase their research efforts to credibly measure the degree to which their products improve or maintain the health or overall function, and plans will need to have in place proper measurement strategies to ensure that these products continue to provide value. Further, if vendors want to remain competitive in this increasingly value-based and outcomes-focused environment, they will need to adopt robust outcomes measurement strategies, ideally combined with a performance guarantee (PG) that puts their fees at risk.

•  With this new reality as a backdrop, how can MA plans cut through the noise and understand which vendors have value creation methodologies that truly work, are linked to vital business metrics, and meet the standards of the Final Rule?

 

• Meanwhile, what steps should vendors take with their approach to ensure all impact is measured and documented to the standard of scientific research?

Not All Measurement is Credible Measurement

To evaluate any service, we must start by asking: Is this program making a difference in members’ lives? We look for tangible signs – are hospitalizations decreasing, are chronic conditions better managed? Next, we dig deeper: Is this improvement directly linked to the care provided, or could it be due to other factors?  This impact must be quantified. How much is the service reducing costs, improving health, or enhancing member satisfaction? The degree to which this impact is observed along a specific, measurable healthcare outcome is critical to proving value.

The core of outcome reporting involves establishing an incremental average effect estimate, or the average impact per member enrolled in the program. For example, Belle’s findings demonstrate a total cost of care reduction of $800 – 1500 per engaged member per year, depending on the overall risk status of the population.  Vendors may describe a reduction in adverse events, change in clinical biomarker value, or reduction of symptoms associated with disease. While all of these outcomes fulfill the CMS requirement, plans are more likely to welcome vendors with endpoints that follow specific criteria aligned with the plan’s performance metrics and regulatory reporting framework.

A compelling  impact metric must be objective. Improvements in symptom reduction or member wellness that may be self-reported are subjective measures that alone do not demonstrate value creation. Health plans need to see meaningful, measurable outcomes that directly link to financial metrics such as medical cost savings, increased risk-adjusted revenue, or improved member retention. It’s crucial that improvements are tied to specific interventions or services provided by the vendor, rather than general trends or placebo effects. Moreover, outcomes should be assessed using standardized, validated measurement tools rather than proprietary or subjective scales that lack broader industry recognition or comparability. For example, a service may purport an improvement in cardiovascular health through increased exercise fitness on a scale designed by the vendor. These statements may be true, but the measurement is credible when the increase in perceived fitness has a causal link to reduced adverse cardiovascular outcomes, which directly map to medical cost savings.

Improving Outcomes With Respect to What Standard?

Any measurement of change must be compared to a standard, ideally utilizing a control group of plan members eligible for but not receiving the supplemental benefit. While randomized control trials (RCTs) are the gold standard, they’re often impractical for supplemental benefits. Ethically, members cannot be excluded for covered healthcare services they are eligible for, especially if it will benefit them. Additionally, given low enrollment rates for many supplemental benefits, cutting the population sometimes even in half will decrease revenue and may not meet the sample size requirement for many such studies.

Instead, we can implement quasi-experimental designs by controlling for confounding factors like demographics and health history in a baseline period. Various statistical and machine-learning methods can ensure observed differences during the treatment period are attributable only to the program itself. In cases without a traditional control group, robust within-subject designs like mixed effects models can be used. Crucially, vendors must demonstrate their metrics are based on rigorous comparisons to established baselines, not subjective assessments or simplistic before-after comparisons. This ensures that reported outcomes reflect genuine program impacts rather than unreliable estimates.

Who Is Being Measured and for How Long?

To ensure consistent impact, the MA plan must understand what population is analyzed to create the incremental effect estimate. Ideally, we would want to measure each member enrolled in the program.  However, any data-driven strategy will require a minimum amount of information needed to determine impact. For many causal inference methods, knowledge of the member prior to enrollment in the supplemental benefit program is critical to creating an accurate baseline. Not all members will have enough prior data to meet this criteria, and therefore would be excluded from measurement. The best methodologies will make targeting equivalent to measurement, and exclude those without data beforehand. Additionally, the vendor and plan must also agree to a reasonable time limit needed to measure, as members who either leave the plan or the supplemental benefit early may have met the initial criteria for measurement but did not receive the full treatment.

A Final Word, for Now

Belle’s measurement plan considers all of the above components, and our rigorous yet conservative measurement approach results in a 2:1 ROI for standard MA plans and a 4:1 ROI for “high risk” C-SNP plans. This allows for a performance guarantee that puts 100% of our fees at risk. We believe this approach is an effective, and necessary, model that creates a win-win scenario for MA plans and their vendors in a time when the stakes truly have never been higher.

In our upcoming webinar and blog series, we’ll dive deeply into how to credibly measure and report on point solutions. In the meantime, you can contact us for a 1:1 conversation on measuring ROI.