Patient engagement has become one of healthcare's most cited performance metrics, but it's also one of the least defined. Health plans structure contracts around it. Providers use it to evaluate care management platforms. State Medicaid agencies incorporate it into performance-based payment arrangements. Yet the term means something different to nearly every organization using it.

This lack of consensus leaves the door open for confusion, miscommunication, and inconsistencies. When a health plan compares two care management vendors both reporting 80% engagement, those figures might reflect entirely different realities. One vendor may count a single monthly app login as engagement, while another may require sustained interaction across multiple channels, demonstrated health literacy improvement, and measurable progress on health goals. Without standardized definitions, the comparison is meaningless, which poses a significant problem when organizations structure payment arrangements based partly on engagement performance.

"How the word 'engagement' is used creates so much confusion in the digital health space," said Sara Greenbaum, Waymark's Head of Operations. She described a spectrum where at one end, organizations claim they "engaged a patient" the moment they have a simple conversation, regardless of outcome or whether the patient actually receives care. At the other end of that spectrum sits a different approach: building relationships where care teams and patients work together to solve at least one barrier to getting and staying healthier.

Research from Waymark co-founder Sanjay Basu, published in Mayo Clinic Digital Health, provides comprehensive analysis of how digital health companies define engagement. The systematic review examined 30 digital health startups, analyzing 64 engagement definitions from publicly available sources. The findings confirm what many care delivery professionals already suspect: the industry operates without shared standards.

The Five-Level Maturity Model for Defining Engagement

The research identified three types of engagement definitions: explicit definitions that clearly articulate measurement approaches (18.8%), implicit definitions that describe metrics without defining the concept (45.3%), and non-definitions that use engagement without substantive clarification (35.9%). The research proposes a five-level maturity model that maps this spectrum:

• Level 1 – Using engagement as a marketing term without measurement
• Level 2 – Tracking basic behavioral metrics
• Level 3 – Codifying an explicit definition with one to two dimensions
• Level 4 – Standardizing a comprehensive definition across multiple dimensions
• Level 5 – Using an optimized definition incorporating multiple dimensions with clear outcome linkage

The study found that 82.8% of digital health companies operate at Level 1 or Level 2. Only 1.6% have reached Level 5, the level at which Waymark performs.

What does Level 5 measurement actually look like in practice, and how does it produce engagement rates that correlate with clinical outcomes? Waymark's approach demonstrates the concrete differences that separate Level 5 organizations from those operating at lower maturity levels.

How Level 2 Measurements Fall Short

Most organizations cluster at Level 2, tracking basic behavioral signals: app logins, session duration, click-through rates, appointment attendance. While these metrics offer objective measurement, they capture only surface-level interactions. Under this definition, a patient with diabetes who logs in daily but demonstrates limited understanding of their condition, experiences frustration with the platform, and feels disconnected from their care team registers the same engagement score as a patient with identical login frequency who shows improved health literacy, reports satisfaction with care, and actively partners with a care team member.

For health plans and providers serving patients receiving Medicaid benefits, this limitation becomes particularly problematic. Members managing chronic conditions while navigating housing instability or transportation barriers may show low platform utilization—and that's not because they're disengaged from care, but because they face structural access constraints. A missed appointment may reflect transportation unavailability rather than lack of health management commitment. Behavioral metrics alone cannot differentiate between members who are disengaged and those who are engaged but facing barriers.

Organizations reaching Level 3 or 4 typically establish explicit engagement definitions before building measurement systems. This approach reflects what Sara identifies as an essential mindset shift. "You need a clear definition of the outcomes you're trying to drive for a patient before you can do anything else," she explains. "What is the ultimate goal and outcome you want to drive as an organization for patients, and how do you think about the right milestones and points in time to meet those goals?"

How Waymark Defines and Measures Engagement at Level 5

According to Sara, Waymark measures engagement across integrated dimensions that directly predict clinical and utilization outcomes:

• Behavioral engagement: Waymark tracks member interactions across channels, and also contextualizes them in our reporting. A missed appointment triggers different care team responses depending on whether it reflects transportation barriers, scheduling conflicts, or clinical concerns. The platform captures not just that an interaction occurred, but also what that interaction accomplished.
• Social determinants: Waymark defines and measures barriers to care: housing instability, transportation access, food insecurity, pharmacy access. Care team members help members resolve these barriers, and those interventions become measurable engagement components rather than unmeasured "soft" activities.
• Patient’s care goals: While many care goals outlined by patients to Waymark’s care team members are related to social determinants of health, some are not. Regardless, meeting patients where they are includes ensuring that their priorities are folded into their care pathways.

Why Multi-Dimensional Engagement Data Matters for Improved Patient Targeting

The research found that companies with explicit definitions were significantly more likely to incorporate dimensions beyond behavior. Academic frameworks typically emphasize cognitive engagement (attention, comprehension, information processing), affective engagement (emotional connection, satisfaction, motivation), and social engagement (interaction with care teams, peer support, community participation). While these dimensions may sound theoretical, they translate directly into predictive indicators that matter for health plan performance.

Predictive analytics deployment is an example of this. Machine learning algorithms identifying high-risk populations depend entirely on input data quality. When engagement data consists primarily of basic behavioral signals (clicks, logins, session duration, etc.) predictive models will remain similarly limited. Effective risk stratification requires data that captures not just whether a patient interacted with a touchpoint, but whether the interaction improved their understanding of their condition, strengthened their relationship with their care team, or built sustainable health management behaviors.

For health plans investing in data science to identify members at risk for acute care events, engagement measurement sophistication directly impacts prediction quality. Models incorporating only behavioral data miss important indicators of patient activation, which influences whether a high-risk member successfully manages their condition or experiences an avoidable emergency department visit.

Waymark's reinforcement learning models use comprehensive engagement data to predict which members need which interventions. This multidimensional input enables the models to differentiate between members at high risk for acute care events who need intensive care management, members facing primarily social determinant barriers who need resource navigation, members with low health literacy who need education-focused interventions, and members demonstrating strong activation who can succeed with lighter-touch support. This differentiation directly impacts both outcomes and resource efficiency.

What Level 5 Organizations Can Demonstrate

Level 5 organizations define their outcomes first, then build engagement measurement to track progress toward those outcomes. For patients receiving Medicaid benefits, this might include clinical outcomes like diabetes control or blood pressure management, utilization outcomes like reduced avoidable emergency department visits, cost management, and patient activation improvements. It also requires measuring social determinants of health and defining the barriers that prevent patients from accessing care and managing chronic conditions.

This comprehensive approach enables care teams to differentiate between types of member needs. When a care team member helps a patient resolve pharmacy access issues, secure transportation to appointments, address food insecurity, connect to primary care, or navigate housing stability challenges, these interventions become measurable components of engagement rather than unmeasured "soft" activities. The engagement data then feeds back into predictive models, improving the organization's ability to identify which interventions drive outcomes for which member populations.

"Transparency around how you define engagement is key to help your partners understand how you actually impact patients," Sara said. "However you define engagement, you should define that publicly and often." Organizations operating at Level 5 discuss their engagement definitions candidly with health plan partners, state agencies, and providers, ensure a consistent internal understanding of that definition, and regularly validate that definition against research and best practices.

The Path to Level 5

Engagement measurement standardization does not require every organization to measure engagement identically. Different interventions and populations may appropriately emphasize different dimensions. But advancement does require specific commitments.

First, organizations must explicitly define what they mean by engagement before building measurement systems. This definition should articulate which dimensions matter, be they behavioral, cognitive, affective, and/or social, and why those dimensions connect to desired outcomes. The definition exercise itself often reveals gaps between what an organization claims to do and what it actually measures.

Second, organizations must identify the outcomes they aim to drive and work backward to determine which engagement metrics predict progress toward those outcomes. This typically includes clinical outcomes, utilization patterns, cost management, patient activation, and social determinants of health. Organizations should be prepared to discuss these definitions publicly with partners and stakeholders, treating measurement transparency as an accountability mechanism rather than a competitive vulnerability.

Finally, organizations should assess their current maturity level honestly and build a roadmap for advancement. Moving from Level 2 to Level 3 requires codifying explicit definitions. Moving from Level 3 to Level 4 requires standardizing measurement across multiple dimensions. Moving from Level 4 to Level 5 requires establishing clear outcome linkages and validation that engagement metrics actually predict the outcomes that matter.

The research provides both a diagnostic of current industry practice and a framework for evolution. Whether the digital health ecosystem embraces that framework will determine whether engagement remains a source of confusion or becomes what it should be: a standardized, meaningful measure of how effectively healthcare interventions support patients toward sustained health improvement. For organizations willing to do the definitional work, the path forward is clear. For health plans and providers evaluating partners, the questions to ask are now equally clear.