How digital clinical data, AI and NLP can enhance RQ performance

Everything about health care is rapidly changing — how we work, how we live, and how we seek and get care. Technology influences the interaction between providers, patients and health plans. It has become vital to propel innovation in how we receive, interact with, and share data and insights — all in a timely manner. This will empower the multitude of recipients to improve access, outcomes, efficiency and performance for all.

Consider the following pre- and post-COVID telehealth statistics:

  • Before the COVID-19 public health emergency, ~15K Medicare members/week received telemedicine services, roughly 800K visits each year1
  • In the eight-month period from mid-March 2020 to mid-October 2020, more than 24 million received Medicare telemedicine service1
  • Before the pandemic, 56.40% did not believe they could receive the same level of care from telehealth compared to in-person care, but 79.85% now say it is possible2

Roughly a year later in March 2021, Americans said this about telehealth:

  • 85.52% get the care they need more easily 2
  • 64.05% prefer to have parts of their annual physical done via telehealth 2
  • 51.64% visit their physician more often2
  • 31.26% have decreased health care costs since using telehealth 2

Laying the groundwork for change

As perceptions and behaviors changed during the pandemic, challenges piled up. Patients still needed care when life and well-being were on the line. The one thing that didn’t change was our shared goals of quality outcomes and efficient delivery. We expanded technology, adapted, persevered, and had a little help from regulatory relaxation to help solve challenges.

Providers and health plans alike want to focus their time on providing the best care possible. How do we continue to think differently?

Next generation of offerings

The next generation of offerings must facilitate care with a refined focus on:

  • Emphasizing persistent approach to gap closure
  • Concentrating on continuity of care, not moments of care
  • Wrapping services to help improve member care
  • Supporting greater accuracy of member health records

Start with engagement

Engagement is key to creating better outcomes and efficient care delivery. Establish a flexible yet intensive approach that embeds more touch points — daily, weekly — between individuals and their providers. Build true, professional relationships and maximize engagement to drive performance and outcomes.

Customize your approach

You may have to customize for providers who have more resources and require less assistance.  Supplementing what they do today may be all they need.

Gather feedback

You must have feedback. Engagement and customizing your approach can’t be done effectively if you don’t track and measure progress. Reporting is critical to measuring success. Here are some tips:

  • Establish a roadmap toward high performance with pivot points
  • Review progress and outcome goals regularly
  • Provide continuous education to providers and care teams
  • Help with prioritization and provider trending reports

Digital integration

Complex patients need care throughout the year. Clinical data acquisition is important to the digital delivery and integration to address member care needs. It allows an iterative, year-round approach to risk and quality gap management. It also forms a more meaningful, longitudinal view of the patient.

This new view comes from sharing new and outstanding gaps as well as clinical activity from each visit. Better yet, acquiring that data in an automated manner from multiple modalities is far less laborious for providers.

Next-generation offering benefits

Standardizing data acquisition and embedding resources can resolve multiple provider pain points.  Using deep learning, machine learning algorithms and NLP (natural language processing) solutions provides more streamlined, informed approaches to driving value and efficiency. 

  • Technology — Release the power of clinical data acquisition by using common core EHR (electronic health record) capabilities intended to support data simplification. Artificial intelligence (AI) helps prioritize and allows you to focus resources focus where there is opportunity.
  • Support — Build provider relationships to expand beyond transaction and payment issues. Focus on their outreach opportunities with patients. This helps identify barriers to care and avoid situations of patients self-fragmenting their care.
  • Outcomes — It’s easier to get a provider to say yes to receiving data through digital means. Once they realize the benefits of getting updated information throughout the year, it becomes a necessity. They want to understand gap status, new gaps, new patients so they can care for their patients completely.  
  • Performance — The earlier and more complete view of a patient’s clinical picture helps inform early performance on quality measures and less end-of-year chaos for chasing down patients. It also creates a framework for the most accurate and complete risk adjustment efforts on these patients.

Data acquired from multiple programs feeds program-directive decisioning

Use a combination of technology, field assets and an end-to-end solution with multi-payer risk and quality analytics. Feed information into provider workflows to assist with pre- and post-encounter member management. Infuse data with interactive, refreshed member information to continue care for subsequent encounters.

AI empowers decisioning

AI is transforming business through self-driving cars, facial or image recognition, voice assistants and more. To make data relevant for business use cases, processes must learn from data and subject matter expertise. Large, complex sets of inputs and millions of samples help train AI to “learn” the correct way to classify the sample.

Health care has relied on humans to facilitate complex decisioning and AI is well suited to advance decisioning. That said, as more analytics organizations try their hand at health care, we must remember that not all theories translate easily to health care. It is very complex, especially risk adjustment. Actions recommended by AI should not be retrofitted for risk adjustment but rather built for risk adjustment.

A central hub can facilitate multi-channel chart acquisition so data can be used and reused. It creates an opportunity to address members in a different way. We can shift to focus less on timelines, generally referred to as “prospective” or “retrospective,” and rather more iteratively. You can also design multi-health plan, member-provider analytics to drive a member care journey through next best actions. This assists providers and results in improved program outcomes when member care needs are correctly documented on the claim and the chart.

AI directs processes

It helps to make action-based decisions centered on changing how we acquire data. The level of data received closer to point of care also changes the way we perform our processes in some of the following ways:

1. Intelligent chart review targeting goals are to:

  • Prioritize charts for retrieval
    • Identify charts pre-targeting to remove them from the workflow
    • Reduce waste
    • Decrease administrative burden
    • Limit provider abrasion

Using Optum AI, we have observed a reduction in chart retrieval requests of 3% to >10%.

Even achieving the goals above, you will still be able to:

  • Submit complete member health history
  • Ensure risk scores are accurate through more complete capture and submission

2. Intelligent chart review: coding. AI helps the chart review process to be more precise when capturing suspected but unreported diagnosis codes. Optum AI has enabled a 12% to 15% increase in conditions captured and a 3% to 6% increase in validation rates in our internal QA oversight process. This led to more accurate and complete coding by performing the following steps:

  • Step 1: Smart chart routing analyzes potential unreported diagnosis codes, makes a decision on further chart review, then may route based on coder expertise.
  • Step 2: AI-enabled coding assists coders with specific diagnosis code suspects or full chart-targeted condition review.
  • Step 3: Completeness review detects if a member’s health history may still indicate possible unreported diagnosis codes and can route chart for additional review, if needed.

3. Automated quality abstraction. Information received should address both risk and quality and the distinct processes that support both programs. Capabilities must direct digitally retrieved charts for abstraction. Use AI to identify quality gaps addressed in the document.

Key takeaways

  1. Patients are maturing as to how they want to participate in their care journey. We must change how we provide information to support providers with these changes.
  2. Addressing members care gaps isn’t a one-time process. It’s an iterative management action to keep members healthier longer.
  3. Interoperability is the wave of the future, but it needs to be less theory and more practical application.
  4. Health is complicated. As more analytics organizations try their hand at health care, we need to remember that not all theory can easily translate to health care. Look for health care and technology expertise vs. just technology.

To find out more about facilitating digital clinical acquisition to enhance your risk and quality programs, watch our on-demand session and read our white paper.

About the authors

Chris Corbin, VP, Growth, Strategy and Partnerships

Chris leads the Optum digital integration strategy with top health systems and technology partners, supporting prospective programming. His previous roles within the organization include leading strategy and solution architecture for the Optum suite of risk and quality solutions.

Karl Korn, Senior Director of Business Development and Strategy

Karl oversees regional teams driving digital integration and execution. He also supports development and implementation for proof-of-concept business models with a current focus on prospective risk adjustment. He has more than 10 years’ experience leading provider engagement solutions and value-based/risk-bearing strategy design and execution.

Kelly Tucker, VP of Risk Adjustment

Kelly Tucker is responsible for strategy and solutions transformation for Optum Risk and Quality Solutions. She has over 15 years of health care experience, spending over a decade leading day-to-day operations for Optum prospective and retrospective risk adjustment programs.

[1] CMS. Trump administration finalizes permanent expansion of Medicare telehealth services and improved payment for time doctors spend with patients. Press release. Dec. 1, 2020.

[2] Sykes. How Americans feel about telehealth: One year later.

Leave a Reply