Evidence Based Medicine - Life Saving. Cost Saving.

Company:

A leading health informatics and medical management services provider gathers clinical data, aggregates that data for each member, and applies rules drawn from the medical literature to identify health plan members whose care could be improved. The company covers health plan members nationwide in a range of health management and data analytics products and services – all to reduce medical errors and medical costs.

The system underlying all of their products compiles member data from a variety of sources such as medical and pharmacy claims, lab results and information provided directly from members through disease management programs, and analyzes this information against highly respected sources of evidence-based medicine to identify gaps in care, medical errors and quality issues.

While data is constantly being absorbed, the company continuously updates their database of the best standards in care through efforts of physicians, pharmacist, nurses and information technologists in their clinical development center.

Challenge:

The system as originally developed was not designed to scale to the volume and variety of data nor complexity of fact models and rule sets that became necessary as the company evolved, and was putting severe strains on their ability to maintain and grow their business. A particular challenge was that the underlying problem sets, while very similar in nature to those in underwriting, securitization and claims processing, were significantly more complex and high volume:

• Millions of records a day (versus thousands in most other circumstances)
• A fact model 10 to 20 times more complex than in similar problem spaces
• Over 7000 rules (versus a more typical 500 to 1500 rules)

Solution:

The company contracted Pathfinder to design and implement a new platform architecture that could help them grow their competitive advantage.

We analyzed the existing system, architected a new platform, piloted initial implementation and handed off to internal development teams for build-out and maintenance.

Some elements of Pathfinder’s solution included:

• SOA – Set up a Five Zone SOA Model with a centralized operational data store (ODS). All new development could now leverage the ODS and make use of common business services. Existing applications were tied into the new infrastructure via web services integration and periodic data feeds.
• ETL – Migrated data ingest from ad hoc to robust ETL using Informatica. Turnaround for onboarding new clients went from months to weeks. Overall quality of ingested data improved dramatically, which resulted in more effective application of the evidenced-based medicine analytics system.
• Business Rules Engine – Retooled evidence-based medicine analytics system from Ad Hoc to a Business Rules Engine (BRE). Most rules engines are not equipped for this data volume or complexity, and rigorous testing of potential rules engines for scalability and performance were required to find the subset that could handle these problems. Clinical staff was able to develop and maintain rules more efficiently, as well as implement far more complex medical rules.
• OLAP and Data Self Service – implemented an analytic and data/reporting distribution platform to enable plan account managers conduct additional analysis on their data sets, and also recreate analytics and reports for the participating employer groups.
• Self Reported Data Capture – also created a platform to facilitate the capture of self reported data by members, e.g. health risk assessments, OTC medications, health behavior, etc
• Messaging – enhanced the platform to generate and deliver targeted messages to providers about improvements to their plan of care, and to members to provide information needed to make provider encounters productive.

Impact:

The company was able to realize a number of operational and quality improvements:

• Able to redeploy operation staff from maintaining the analytics system to higher value tasks
• Able to onboard new clients in a matter of weeks rather than months
• Scaled the new platform to actively cover over 12 million members nationwide
• Developed new products and services, such as a Personal Health Record (PHR) system, more quickly using the architecture as a base

In 2005, the company was purchased by a major insurance company for a multiple of over 10 times revenue.