The six factors: A shared framework to drive outcomes with retrospective risk customers

Paul Burke, Chief Product Officer, and Dave Meyer, Chief AI & Data Officer, Reveleer

Risk adjustment helps healthcare organizations receive fair compensation by accounting for the health status of their members in a changing industry. The integration of Artificial Intelligence (AI) into risk adjustment processes has already been shown to significantly enhance the accuracy and efficiency of these efforts. Merging human expertise with AI technology can streamline workflows, ensure compliance, and enhance staff efficiency. At Reveleer, our customers have been realizing the benefits of these efficiencies through our best-in-class integrated NLP solutions for years.

As AI advances, healthcare standards are increasing too. Additional scrutiny on the capabilities and the outcomes is to be expected and should be celebrated. With the additional administrative cost pressures and the continued increase in cost of care, it is critical for the buyer to understand and measure in the context of the end-to-end journey for risk adjustment programs.

The effectiveness of an AI-embedded risk adjustment technology is contingent upon several key input considerations that drive the accuracy scores for any solution.

Considerations Impacting Program Performance

‍

SaaS software is built to support customers in several scenarios to ensure it can meet the needs of customers and how they do business. Within risk adjustment, that flexibility enables customers to make use of software in the way that is best fit for their organizational constraints. It is important to note, however, that each decision in how a project or program is run can and will have an impact on performance. Customers and their solution providers must have a shared understanding of the following six factors to accurately observe and compare cohorts and continuously optimize.

The Six Factors Driving AI Risk Adjustment Performance

We created these six factors to help customers make better input decisions to ensure that our solutions and services meet or exceed performance targets:

1. Services vs. Self-Service Although both approaches are valuable for their own reasons, the choice between utilizing external services or adopting a self-service approach can influence the performance of the AI solution...	2. Project and Coding Guidelines/Behavior Employee adherence to standardized project and coding guidelines is essential for ensuring consistency and accuracy in data processing...	3. Coder Level and Domain Proficiency The proficiency of coders in the domain directly impacts the quality of data input...
4. Estimating Chart Quality Chart quality includes retrieval partners, retrieval types, and the density of medical charts...	5. Software Version and AI Model Software changes over time. Regular updates and advancements in both UX and AI models can enhance predictive capabilities...	6. Completeness and Accuracy of Data The completeness and accuracy of the input data are paramount...

And finally, ensuring the comparison data meets criteria like valid sample size, representativeness, data quality, sampling method, documentation, and relevance is essential for accurate results and meaningful comparisons.

Utilizing Productivity and Efficiency Measures in Value-Based Care
‍

Health plans, Accountable Care Organizations (ACOs), and providers rely on productivity and efficiency metrics to drive value-based care, risk adjustment, and quality abstraction. These metrics are essential for optimizing resource allocation, improving patient outcomes, and ensuring financial sustainability within their organizations. Productivity metrics help organizations track and enhance their performance in managing patient populations and meeting regulatory requirements in government markets (eg. Medicare, Medicaid, and ACA). For commercial lines of business, these measures provide insights into operational efficiency and cost-effectiveness, enabling better decision-making and competitive advantage.

Four Key Measures for Risk Adjustment Effectiveness & Efficiency

The following are key measures we work with our customers to realize effectiveness and efficiency of risk adjustment solutions:

1. Throughput & Processing Speed: Throughput refers to the amount of data the software can process within a given time frame. Processing speed is the rate at which the software can analyze and interpret data. In risk adjustment, high throughput and fast processing speed enable the software to handle large volumes of patient data quickly, ensuring timely and accurate risk score calculations ‍
   
   1. Charts processed per hour/day (e.g., AI-assisted chart retrieval vs. manual retrieval). ‍
      
   2. Time saved per review/extraction task (e.g. NLP-driven evidence extraction reducing manual abstraction time)

2. Accuracy & Quality Improvement: Accuracy in AI-backed risk adjustment software means the precision with which the software identifies and codes diagnoses, ensuring that all relevant conditions are captured. Quality improvement involves using AI to enhance the consistency and reliability of coding, reducing errors and improving the overall quality of risk adjustment ‍
   
   1. HCC/diagnosis capture rate (e.g., AI’s ability to find additional risk-adjusting conditions missed in manual reviews). ‍
   2. Error reduction rate (e.g., fewer miscodes or missing documentation due to AI validation).

3. Cost Reduction & Workforce Efficiency: Cost reduction refers to the decrease in operational expenses achieved by automating manual processes and reducing the need for extensive human intervention. Workforce efficiency is the improvement in productivity and effectiveness of staff, as AI tools streamline workflows and allow human resources to focus on more complex tasks ‍
   
   1. FTE reduction or reallocation (e.g., AI reducing manual workload, allowing coders/reviewers to focus on complex cases). ‍
   2. Operational cost savings (e.g., reduction in outsourced coding/review expenses through AI automation).

4. Turnaround Time & Workflow Acceleration: Turnaround time is the duration it takes for the software to complete a task, from data input to the final output. Workflow acceleration refers to the speeding up of various processes within the risk adjustment workflow, enabled by AI's ability to quickly analyze data and generate insights, thus reducing delays and improving overall efficiency. ‍
   
   1. Ave. time to complete a chart review or coding task before vs. after AI implementation. ‍
   2. Reduction in backlog processing time (e.g., AI enabling faster submission of risk and quality data).

Benefits of Investing in Measurable Software Solutions
‍

Organizations face several challenges in the healthcare risk adjustment and quality space, including the complexity of data management, the need for accurate and timely coding, and the pressure to reduce operational costs. Manual processes are often time-consuming and prone to errors, leading to inefficiencies and potential financial losses.

By investing in AI-embedded software solutions focused on the customer business outcomes, health plans, ACOs, and providers can overcome these challenges and achieve significant gains. Having reliable insights on the performance and effectiveness are how these solutions verify the ROI associated with accuracy and consistency in coding, streamlined workflows, and reduced administrative burden. Anyone can quote a statistic and back into the defense of that number. Picking a partner with a transparent measurement methodology is critical to ensure the results you seek are the same as the results you get—and most importantly you share an understanding of the methods and decisions driving those results.

As organizations evaluate their choices for risk adjustment software, it is imperative to consider the above factors and metrics that impact the performance of AI-embedded solutions. Organizations should expect and request the following from sales teams:  

Sales Team Expectations

• Access to knowledgeable professionals who can provide guidance and support throughout the implementation and utilization of the AI solution,  

• Solutions that offer customization options to meet the specific needs and workflows of the organization,  

• Commitment to continuous improvement and regular updates to the software and AI models to ensure optimal performance,  

• Training programs to enhance the proficiency of coders and other users in effectively leveraging the AI solution

• Measurement models to ensure the completeness and accuracy of data inputs, thereby enhancing the reliability of AI-driven insights

Investing in AI-embedded software solutions is essential for healthcare organizations aiming to improve risk adjustment and quality outcomes while reducing operational costs. By prioritizing transparency, customization, and ongoing support, organizations can find the right partner that will maximize the value of their technology investments and drive measurable business results.

To learn more about how Reveleer’s solutions can help your organization address these challenges and unlock new efficiencies, contact Reveleer today and speak with one of our experts.