Why MVP wants to push fresher care-gap data into EMRs

Seven days. Some of the quality measures MVP Health Care tracks give a practice that long to act, and others give 30. Providers commonly complain, as David Raths of Healthcare Innovation noted, that payer data can lag by a few months.

Those windows already shape what the nonprofit insurer, which covers members in New York and Vermont, builds today. Hospital admit, discharge and transfer (ADT) alerts get their own tab on the dashboard MVP builds for the clinicians in its network. In an interview published Nov. 12, 2026, Healthcare Innovation spoke with Susan Kopp, MVP's director of provider analytics, innovation and delivery, and Collete Yao, vice president of enterprise AI and clinical data strategy, about building closer data-sharing relationships with provider organizations. The dashboards run on Arcadia's data platform, and the EMR step would use Arcadia's connectors. That raises a practical question for any health plan running provider analytics: which practices will let a payer's data into their EMR, and how quickly?

A dashboard organized around the patient

MVP started with individual reports in Arcadia, built with help from its internal risk and quality teams, and those grew into a custom provider dashboard. Alongside utilization and open quality gaps, it carries the ADT tab. The newest addition comes from one of MVP's pharmacy reporting systems and is meant to help practices act on medication adherence measures.

The utilization view goes further than a count. Kopp said the dashboard carries basic demographic data, emergency room use and hospitalizations, along with whether those events were avoidable. That flag suggests where earlier outreach might have mattered, which is a more useful question for a practice than how many of its patients ended up in a hospital bed.

MVP also built internal dashboards that closely match what providers see. That sounds minor, but it means the insurer's staff and the practice can look at nearly the same numbers when they discuss a patient. Kopp said the ADT tab puts context around each alert. It keeps that data in one place with the rest of the patient's story, which gives practices time to act on quality measures with seven-day and 30-day windows.

Arcadia has also been bringing in third-party data that Kopp called very important to MVP, and the platform now lets the insurer build its own custom dashboards and use words in its reporting that mean something to MVP users and to providers. Paired with near-identical internal and external views, that shared vocabulary could keep a conversation about a care gap from starting with a debate over what a label means.

Timing is the harder problem. For a quality measure on a seven-day clock, a payer report that runs a few months behind arrives after the deadline it was meant to help with. MVP combines data it generates internally with measures Arcadia calculates and aims to give providers the most current data it can. Kopp gave a concrete example: if a patient had an eye exam or an A1C test last week, that result counts toward the quality measure, and the practice does not contact the patient again about it.

The example shows why fresh data matters to a practice's workload. A stale gap list sends staff after work that is already done, and the patient gets a reminder they no longer need. For a payer judging its own provider analytics, a useful test is how quickly a completed service drops off the list.

Liaisons do the work the data can't

Kopp spent years on the provider side at regional health systems in the Greater Capital District. There, she told Healthcare Innovation, payer reports arrived first as PDFs and later as Excel files. Each patient was a line of data with no context, and practices were asked to act on it with little contact from the payer. The complaint is an old one. In 2019, Scott Weingarten, M.D., told the same publication that organizations get a lot of information, much of which is not actionable, and that working out what is actionable and important is the next step for data analytics.

Kopp, who has a clinical background, described a different arrangement at MVP. Teams of provider partner liaisons keep close relationships with practices and work with them to deliver data that has meaning, which turns the dashboard from a file a practice receives into a set of numbers both sides discuss.

The practices vary. Some are very small and some are very sophisticated. Kopp said the goal is always to get data into the provider's workflow, but that isn't always possible, so the liaison's job is to find the people in each practice who can act and agree with them on priorities. Providers work with many payers, and MVP wants its practices to know which items it considers most important.

Those priorities have money attached. Many providers are in MVP's value-based care program, especially in its government programs. Some earn quality incentives, and others take on upside and downside risk for their patient populations.

Where most health plans stand

MVP's approach looks more notable next to a vendor-commissioned survey that Arcadia, its platform vendor, released in August 2025, best read as a rough baseline. As Healthcare Innovation reported in an interview with Arcadia chief medical officer Luke Hansen, M.D., health plans rated themselves better at sending data out than at bringing outside data in.

Hansen has worked both sides of that divide. Before Arcadia he was Illinois market Medicare Advantage chief medical officer at UnitedHealth Group and a senior medical director for population health at Optum. He traced the gap less to technology than to how organizations are split: plans have business owners and clinical owners, and value-based care ties the two together far more tightly than the industry has historically expected.

He was also blunt about where plan analytics are strongest. Health plans are insurance companies at heart, he said, so their best tooling covers pricing risk and member acquisition. Managing clinical care is not their primary function, though they end up doing it for members whose risk they carry, partly because the provider side remains relatively reactive. MVP's dashboards aim squarely at the competency Hansen ranks weakest: turning plan data into clinical action at the practice.

That nine-point gap between sharing and integrating matters for any plan with AI ambitions. Hansen argued that AI tools can only be as good as the data underneath them, and that fragmented records keep a model from seeing the full patient history. He named document processing as probably the easiest AI use for plans and saw real opportunity in prior authorization advising and automation, with both depending on a good data asset.

Yao described MVP's AI investment the same way. A payer sees a member's contacts across many parts of the health system, she said, and MVP wants to use AI and clinical analytics to deliver those insights at the right time and in a usable form. The industry is still learning how to do that, in her telling, and the point is for the provider to have a meaningful conversation with the member at the right moment.

MVP's model may not work the same way everywhere. Hansen said regional plans tend to understand their communities and have long-standing ties with local providers, while national plans have more scale and resources. A liaison model built on relationships in the Capital District suits a regional plan. A national insurer covering thousands of practices would probably rely more on automation and less on personal contact.

Why the clinical record matters to a payer

Hansen compared the member profile a plan needs to an EHR, which he called a 360-degree longitudinal clinical record. More history makes the future easier to predict, he said, and a plan that wants to price risk, predict it and act proactively will struggle without that view.

The stakes have risen. Hansen pointed to Version 28 of Medicare Advantage risk adjustment, more regulatory scrutiny, and a retrenchment among plans and among providers less sure they want to participate in value-based arrangements. Together, he said, those forces push plans toward cost containment, where risk means more than HCC (Hierarchical Condition Category) coding and extends to social risk and everything that shapes how a person will use care.

Read against that, timely ADT alerts and avoidable-hospitalization flags look less like reporting extras and more like cost tools. It also suggests why EMR data is the prize in MVP's next step: it is the payer's most direct route into the kind of longitudinal clinical record Hansen describes.

EMR connectors and the trust question

The next step goes after a long-standing weakness. Yao said care-gap information from payers is often out of date because payers have not usually received EMR data. MVP plans to use Arcadia's EMR connectors to get clinical data faster, alongside the health information exchange data it already receives. In some cases, that would let MVP see a member's upcoming appointments, adjust what it sends based on that visit and deliver care-gap information directly into the practice's EMR.

Yao framed the aim as removing more friction so providers can focus on patient care, and MVP's executives also plan to build predictive analytics into their Arcadia work. The destination has been on the industry's list for years. In the same 2019 Healthcare Innovation roundup that quoted Weingarten, Debbie Zimmerman, M.D., then corporate chief medical officer of Lumeris, named embedding key data and analytics into clinician workflow to uncover gaps in care as one of the big opportunities. MVP's plan is a concrete version of that idea, with the gap arriving inside the practice's own system instead of in a separate report.

Who connects first depends on existing relationships. Yao said Arcadia already exchanges data with many provider groups. Some are in Medicare ACOs that use Arcadia, and some have EMR connections through Arcadia with other payers. She called those practices the easiest cases, where MVP mainly needs agreement to participate. Practices new to Arcadia, or with little experience sharing data with payers, will need more time. They need to be comfortable with both the information they send to MVP and the information MVP sends back into their EMR.

Yao also explained the platform decision. At other payers she had seen in-house builds, and MVP chose Arcadia instead as part of what she described as a provider-centric approach. The trust question shows why that choice matters. A shared platform comes with relationships that a custom-built system would have to create one practice at a time. This advantage depends on how much a payer's network overlaps with practices already using the same vendor through ACOs or other health plans.

If only the groups already connected to Arcadia join, the liaisons and the dashboard will still carry much of the work. The real test is whether practices not already connected to Arcadia sign on.