Empowering Organizations to Address Gaps in Care: Putting Health Equity Analytics Methods into Practice

July 30, 2025 · Guest Contributor

Everyone deserves the opportunity to achieve their best possible health. Identifying gaps in care, including health care disparities, is critical to acting to achieve this goal. Stratifying health care quality metrics by sociodemographic factors and non-medical drivers of health is a key tool for accomplishing this.

When applied with a person- and community-centered approach to care, stratification can help pinpoint where current efforts are leaving groups behind, identify opportunities for more efficient care delivery and support more effective quality improvement. But stratifying by only one characteristic at a time can result in missed nuances that impact a person’s access to care, experience of care and resulting outcomes. And by evaluating a single quality measure in isolation from other related measures, organizations might not see “bigger picture” gaps that could benefit from a coordinated approach to improvement.

New advanced data analytics methods tackle these challenges. In 2023, with support from the California Health Care Foundation, NCQA released an issue brief evaluating four analytic approaches that integrate multiple measures and stratification factors into composite scores that promote a holistic approach to evaluating health outcomes. What remained unclear was what the real-world implementation of these methods would look like. Would the juice be worth the squeeze? In other words, could organizations feasibly calculate the methods? Would data summaries align with organizational priorities?

Putting Advanced Health Equity Analytic Methods into Practice

Supported by The Commonwealth Fund and the California Health Care Foundation, NCQA tackled this challenge by evaluating the real-world implementation of four health equity measure compositing approaches:

• The Population Health Performance Index.
• The Within-Plan Improvement factor of the Health Equity Summary Score.
• The Health Equity Metric.
• Humana’s Health Equity Quality Measure.

The strength of these methods is in their flexibility: They can be tailored to an organization’s quality priorities. We wanted to understand which approaches are valid and meaningful for health care organizations, their usefulness in tracking and acting on health care disparities, the organizational factors that affect adoption, implementation and impact.

To do this, we partnered with three health care organizations representing different health care segments, populations and geographies: Inland Empire Health Plan, CareSource and ChristianaCare. Our partners brought their expertise and data to the project, informing the design, interpretation and evaluation of each analytic approach.

When asked why they participated in the project, one partner noted: “We currently have a disparity dashboard, but it’s somewhat flat. I’m not able to do robust trending and tracking. It was nice to be able to look at these scoring approaches to see what I could borrow from and incorporate into our upcoming dashboard redesign.”

Our Research Methods

NCQA worked with the partners to understand their clinical, quality and population health priorities. Each partner identified up to four quality measures and up to three sociodemographic stratification factors (e.g., language, geography, race, ethnicity), and provided a de-identified data extract of stratified quality performance. NCQA analyzed the data using the four equity analytic methods, which let us compare how each method summarized the same performance data, and provided detailed results to our partners.

Using the National Implementation Research Network’s Hexagon Tool as our evaluation framework, we conducted a semi-structured interview with partners to find out how they interpreted the results, including relative strengths and weaknesses of the approaches. We followed these interviews with a joint focus group, where the partners came together to discuss broader questions of trade-offs between methods, desirable (and undesirable) characteristics of different approaches and other implementation considerations.

Early Findings from Our Research

We’re still summarizing findings, but a few things stand out so far.

1. It’s all about choices.

A lot of little decisions go into implementing these methods, and cumulatively they can have a big impact. For example, when choosing measures, it’s important to consider how they relate to each other. If they’re too different, the final composite metric may create a summary that doesn’t make clinical sense. If they’re too similar, the correlation between performance measures could skew the results. Another example is the choice of reference (comparison) groups. As one partner shared, they currently use the largest group within their membership as a reference group, “but when you view it another way, where the reference group is your highest outcome group, it reveals a completely different statistic.”

2. Organizational culture matters.

Does your organization prioritize detail? Is “showing your work” important for leadership buy-in? Some methods do better than others in peeling back the layers. Is it important to keep your dashboards uncluttered and at a high level? Select a method that provides an aggregate number versus numbers for different sub-domains.

3. Link to a specific business use case.

There’s a wide range of uses for these metrics: facilitating quality conversations with state Medicaid agencies, tracking the impact of quality interventions on gaps in care over time, incentivizing internal improvement across business units. Different analytic approaches may be better suited to different use cases—and selecting the one that matches your business needs can support organizational buy-in and impact. One partner used these methods to compare health equity performance across business units: “This gave us an opportunity to compare equity scores among our three biggest markets. We never really looked at health equity that way before.”

4. Challenge assumptions.

If results contradict an organization’s prior analyses without a clear understanding of why that is, it can lead to mistrust. But sometimes these multi-measure, multi-factor approaches identify findings that challenge an organization’s current understanding of its populations and programs. For example, an intervention that demonstrated success in high-needs populations might have missed a community with different language needs. Remaining open to possibilities can present opportunities to identify new areas for quality improvement.

5. When in doubt, prioritize clarity and communication.

Some equity composite methods are complex, and can be challenging to explain. An approach with lots of technical bells and whistles may not be as well suited as one that has transparent calculations and a clear interpretation. As one partner explained: “If I understand it, I can speak to it, and then I can get buy-in.”

While not every organization is in a position to implement these methods today, there are opportunities to implement key elements, such as multi-factor stratification, or combining stratification with changes in trends over time. The bottom line is that we’re working to make a set of analytic approaches accessible to organizations, to help them set priorities in areas that matter to them and the populations they serve, without leaving groups behind.

What’s Next

To learn more about this work, and how it may be relevant to you, please join us for a webinar on Wednesday, August 27, at 1pm (ET). Speakers from NCQA will share more on current findings, and representatives from our partner organizations will speak about their experiences and lessons learned from this work.

Panelists:

• Erin Brigham-Gray, Assistant Vice President, Quality Operations, CareSource
• Jacqueline Ortiz, Chief Community Health Impact Officer, ChristianaCare
• Lorena Chandler, Vice President and Chief Health Equity Officer, Inland Empire Health Plan

Register for the webinar today.

There’s more information to come! For all you data lovers out there, we plan to publish quantitative results from our testing efforts to inform future methods development. Early next year, we’ll release a playbook of practical methods for organizations implementing these and similar approaches.

We thank the California Health Care Foundation and the Commonwealth Fund for partnering with us in this important and rewarding work.