The hospital used two Enterprise Resource Planning (ERP) products: Oracle for patients and Lawson for financials. There was no common warehouse for both sets of data. The hospital requires various-use cases in key areas. These include hospital readmission predictions, clinical classifications/prescribing errors/denials management, and natural-language processing. Their Clinical Data Warehouse (CDW) requires a nightly day-end process. Business goals included high-performance reporting; self-service, near-real-time, and predictive analytics; data discovery and exploration across structured, unstructured, and big data; infrastructure/implementation support; and integration with existing business intelligence.
TechD provided the answer—in the form of the IBM Pure Data System for Analytics. This open platform for analytic development allows users to leverage many technologies, and migrate to others as needed. Improvements over other SAS predictive model deployments include significantly reduced cost of ownership, various performance improvements, and greatly improved efficiency. With the improvements that TechD has implemented, users can scale-up their analytic strategies without an increase in cost or headcount.
When TechD deployed IBM Pure Data, we integrated important data functions across the organization. As a result of the upgrades TechD put in-place, the hospital now has a single data warehouse for both patients and financials. Administrators and personnel enjoy access to archived and unstructured data through standard reporting tools. Results include faster processing and the elimination of multiple standard tables. Now, data migration tables are produced much faster. In fact, TechD helped reduce the average speed for nightly end-of-day reporting from five hours to 45 seconds—a major victory.