Leveraging Augmented Intelligence to Improve Health Equity

Healthcare inequities in the U.S. have a staggering economic impact: approximately $320 billion, according to Deloitte. That number could eclipse $1 trillion in annual spending by 2040 if healthcare inequality remains unaddressed. For healthcare organizations that are eager to launch a health equity strategy but are discouraged by efforts that aren’t providing meaningful outcomes, or for those already seeing impactful change, knowing the tools available to drive measurable improvement will accelerate an effective strategy.

Health equity programs that leverage innovative technology, such as augmented intelligence (AI), have produced promising results. AI identifies and quantifies disparities across measures to pinpoint contributing characteristics for intentional goal-setting. A framework to achieve health equity success includes AI technology applied across an organization’s data foundation, commitment to quality and collective leadership effort. 

AI Opportunities Reduce Barriers to Overcoming Equity Challenges

Healthcare leaders are increasingly aware of the power of AI technology to influence clinical behaviors and make improvements on measurable benchmarks; however, perceived technical barriers can overcome health equity challenges through AI.

1. Is my data foundation sufficient? Gaps in a health organization’s data foundation do not prevent the use of AI. Common data challenges can include the following:

• Inconsistent collection of personal characteristic data (i.e., race, ethnicity, and language)
• Multiple source systems that categorize and store characteristics differently
• Lack of specific analytics for health equity
• Missing measurable goals, or the inability to accurately measure and analyze results

2. Are we aware of our unintentional or unrecognized bias? AI algorithms are trained on biased data, and the results so far for healthcare organizations in a Health Catalyst pilot program show that AI removes bias early on in the equity journey. Reducing bias should create consistency in data collection and access while providing other benefits:

• Establishing a single source of truth by enabling aggregation and consolidation of equity data, creating consistent and standardized definitions, training on the definitions, and tying data and efforts to clinical improvement work.
• Removing assumptions by quantifying disparities across many measures and dimensions.
• Identifying where inequalities may have been overlooked for targeted and customized care planning.

3. Do we have the right data measures and characteristics to analyze? When leaders see the financial and clinical results of data collection and documentation efforts, capturing social factors or additional race or ethnicity features becomes a priority. AI uses an organization’s data to remove unintentional bias and guide providers toward care that is unique and valuable for patients.

Quality Improvement Steps Drive Measurable Outcomes

Taking a measured, quality improvement approach to enhancing clinical outcomes includes five actions for health equity stakeholders:

1. Find the opportunity that is demonstrating poor clinical outcomes.
2. Organize the people and data necessary to clarify the root-cause drivers of the adverse outcome needing improvement.
3. Use data and AI to assist with clarifying causal patterns, unraveling bias, etc.
4. Understand what interventions are called for based on data analysis.
5. Intervene to improve the outcome and track performance with continued feedback and measurement. 

ChristianaCare leveraged a sophisticated operating platform as the organization’s single source of truth for personal characteristic data and key performance measures. AI was applied to identify and quantify disparities across measures and pinpoint contributing characteristics and affected sub-groups.

Through AI, ChristianaCare has identified specific opportunities to improve health equity.

• Age: readmissions for chronic obstructive pulmonary disease and heart failure
• Race and gender: heart failure readmission
• Race and geography: COVID-19 testing in Black/African American patients in parts of Wilmington

By initiating projects with focus, intentionality, deliberation and data-driven insights, organizations can expect considerable improvement in removing healthcare inequities. 

AI Eliminates Healthcare Inequities

Using AI to conduct systematic equity analysis will guide interventions and help organizations address the root cause driving disparities. There are three takeaways that will define success.

First, be intentional in planning and decision-making by pulling together the right stakeholders to support and enable equity efforts. Second, it is critical to create a quality process improvement methodology designed to follow the standard and work to iterate and improve. And third, AI technology can accelerate health equity strategies and quantify results, measure results by patient lives saved, and analyze financial returns.

When healthcare organizations have the tools to advance health equity, moving the needle will require not only quality but also improving relationships with patients. This means eliminating bias and opening the door to meaningful change.

Additional Reading

Would you like to learn more about this topic? Here are some articles we suggest:

• Health Equity: Why it Matters and How to Achieve it
• Advancing Health Equity: A Data-Driven Approach Closes the Gap Between Intent and Action
• How to Improve Health Equity Through Patient Engagement Technology
• Advancing Health Equity: An Analytics Framework for Measuring and Managing Disparities