Unlocking the Untapped Potential of Patient Data with AI

Health plans are turning to a more advanced, technology-driven approach to prospective risk adjustment. Payers and providers are investing in solutions to unlock hidden insights in clinical data that provide a more complete view of patient health status at the point of care.

Fortunately, in healthcare, the data is out there today to create more accurate views of patient care journeys. Unfortunately for value-based care organizations, orchestrating that data to create a symphony of patient insights is complicated.  

Clinical data such as blood pressure control, A1C levels and medication adherence can be fragmented across multiple health systems, specialists and medical services providers. Finding, aggregating and analyzing all clinical data for every patient without technology is impossible given time, budgets and bandwidth.

This is where artificial intelligence (AI) is transforming prospective risk adjustment. AI can align disparate data sources and extract insights into patient health, such as undiagnosed conditions. Whether key information exists within clinical notes, claims information or pharmacy records, AI can find them and stitch together a holistic view at the patient level.

Value-based care organizations can use these insights to power prospective risk adjustment. By surfacing clinical insights such as suspected HCCs at the point of care, they can confirm patient risk factors and reduce uncertainty and inaccuracies in retrospective risk adjustment programs.

The challenge: data integration and ingestion

When a single patient chart can be hundreds of pages long, it seems counterintuitive to want more data.  In fact, the industry is built on a growing mountain of data. More than 73,000 ICD codes alone help paint the picture of patient health, and there are many other health data sources that can add color to a provider’s understanding including:

• Clinical data
• Network data
• Claims data  
• Pharmacy and lab data  
• Patient demographics data

The challenge for value-based care organizations is how to capture insights from patient data that extends beyond their health system or provider network. Healthcare is fragmented across tens of thousands of medical groups, thousands of dialysis centers and labs, and more than 75 HIEs. In the past, it’s been virtually impossible for providers and health plans to know everything about a patient’s care journeys.

Even if value-based care organizations access could tap this network, the next hurdle is data ingestion. Generating clinical insights from disparate data sources is more complicated than logging in and hitting “upload.” Where is it stored? How is it reconciled and organized? How do we unlock its potential?

The solution: AI-powered retrieval and insights

It’s easy to say start with more data. The question is how.  

First, companies will need a solution to ingest large amounts of data. Our Platform leverages Optical Character Recognition (OCR) technology to process 96,000 pages per hour. This capability extracts critical data from diverse documents such as medical records, claims, and lab results, maintaining full referenceability throughout the process.

Value-based care organizations also need solutions that can manage both structured and unstructured data from standardized EHR entries to more complex physician notes and scanned documents.  At Reveleer, we provide a holistic view of patient care previously unattainable without AI by consolidating data from disparate systems - including EHRs, billing records, and clinical notes.

Security is crucial to healthcare data management. Our Platform incorporates robust security protocols and compliance measures, including HIPAA, ensuring sensitive healthcare data is protected.

The result is real-time data access and visibility. Comprehensive patient information is available when it is needed – at the point of care. By bridging data gaps and providing actionable insights, our Platform empowers value-based care organizations to make informed decisions and improve patient outcomes.

The result: Influencing care and supporting risk accuracy

The value hidden in disparate data sources is real.  

In a recent study, we analyzed nearly 500 patients from New York with an average age of 81. By tapping clinical sources outside the customer’s existing data network, we found new clinical data for 96% of the population. Digging further we identified potentially undiagnosed risk adjustable conditions for 84%, creating hundreds of opportunities for more effective patient visits.  

In addition to the benefits this could bring to patients, these insights can improve the accuracy of retrospective risk adjustment and, ultimately, reimbursements. Conditions diagnosed, treated and documented correctly at the point of care can create efficiencies downstream in retrospective risk programs, because there’s no need to chase what’s already found.  

AI is already gaining adoption across industries. In finance, AI predicts fraud by combining data from multiple sources to generate highly accurate predictions. Marketing functions mine data sources to understand buyer preferences. 

In healthcare, AI can uncover clinical insights by integrating and analyzing disparate data points from across the healthcare ecosystem. With approaches such as Natural language processing (NLP) organizations can train AI systems on linguistic rules, clinical heuristics and coder behavior to assess potential undiagnosed conditions. 

The first step is creating a holistic view of patient health by integrating data from disparate, siloed sources and unlocking its potential with AI.

‍

‍