Modernizing the healthcare revenue cycle with integrity

Struggling to manage data within complex platforms and pipelines

Ensemble Health Partners set off to unlock the power of their revenue cycle intelligence engine, EIQ,® through unified, accurate enterprise data insights. Unfortunately, with legacy SSIS and SQL server–based processing, Ensemble struggled to reliably surface and sync data across various platforms and reporting repositories promptly. Pipelines were slow and unreliable across the ETL process. The result was disjointed data and complicated client workflows that obstructed enterprisewide insights.

Grant Veazey, CTO at Ensemble, explained, “Our previous approach required each client to have their own database, which inherently added layers of complexity and overhead, pushing up opportunity costs that could be better served driving innovation and value to clients. Sure, we were already harmonizing data at the client level — with half of our providers having multiple databases and potentially having multiple EHR instances — but data was still siloed in each client database.”

To simplify and foster confidence in their data, Ensemble wanted to modernize their underlying data infrastructure with lakehouse architecture. This would allow Ensemble to unify their data across different environments with a single source of truth to support standard enrichment activities to drive more-uniform reporting. Ensemble needed a scalable platform that could power upstream data-based decision-making and downstream client services — improving operations across the board — including cash collections, capturing revenue, optimizing reimbursements and personalizing patient interactions.

Data integrity and reliability on the Databricks Platform

Ensemble decided to standardize their data and AI on the Databricks Data Intelligence Platform, allowing them to easily integrate data across the revenue cycle, drive use cases and unlock efficiency with a holistic view of the entire claims transaction journey.

With Delta Lake as their core storage layer, Ensemble has improved the reliability and consistency of their extract, transform, load (ETL) pipelines — helping to keep data synced and accurate across environments using the same enrichment layer. Uniformity in data processing and unobstructed insights allows Ensemble to reconcile financial performance with each client’s host system accurately. This new capability ensures data integrity across Ensemble’s application, analytics and Power BI visualization layers, building trust and growing loyalty with clients.

From an efficiency and cost perspective, Ensemble is gaining speed without needing to increase labor. Mike Wilson, VP of Data Engineering at Ensemble, said, “The reusability of the pipelines is great for integrity. We know this pipeline works well so we’re going to reuse it. It also helps with our implementations so that we can onboard new clients much faster, allowing us to scale without having to add more people to support those new clients.”

For data governance and security, Ensemble relies on Unity Catalog to manage fine-grained access to different types of data with uniform standards. Databricks Notebooks help with cross-team collaboration as data users can explore and share data more efficiently and securely. Armed with these capabilities, Ensemble can break down data silos, expand payer insights and distribute analytics with understanding to fulfill their use cases.

Improving client outcomes with data-driven insights at scale

Since migrating EIQ to the Databricks Data Intelligence Platform, Ensemble has gained new insights and functionality to improve services at scale. Now, the company is moving faster with fewer resources, simple lakehouse architecture and access controls to become more data-driven with efficient results.

“We’ve seen a 2x improvement in performance and a 10x reduction in failure rates,” Wilson explained. “Now, our data teams are moving faster than ever before.” Rapid ETL processes accelerate the analysis of current and relevant data to provide accurate insights for clients. To further reduce tech debt, Ensemble is treating batch and real-time processing in a singular way to push consistent data streams to dependent downstream applications.

With scalability limits a thing of the past, Ensemble can efficiently consume more than 25 billion transactions and 65,000 payer updates and rule changes each year. The company is leveraging over 25,000 variables to continually inform and tune over 5,500 ML models supporting end-to-end revenue cycle management (RCM) intelligence. These insights have helped them improve revenue cycle operations for their clients including improving coding accuracy to prevent claim denials, accelerating the appeals process, prioritizing collections opportunities and identifying underpaid claims. Call center improvements have also been made using an API that surfaces specific information for associates to provide a personalized patient experience.

Veazey concluded, “As a data-driven business, Databricks has given us an advantage in the marketplace because we’re able to see insights across 30 health systems and growing. The value of that asset, not only for us and our clients but for the industry, is going to be vast.”

Moving forward, Ensemble will continue to harness the power of their enterprise data on the Databricks Platform and expand their use of generative AI to eliminate friction, accelerate reimbursement and improve revenue yield for healthcare providers across the country.