CASE STUDY - Rapid Data Exploration with Surgical Precision
Commentary by Paul Smith, CIO, Provisio
Provisio is based in Nashville, Tennessee, and provides pharmaceutical and medical device companies with prequalified clinical trial candidates who match specific requirements. Its iTrials service, which conforms to the Health Information Portability and Accountability Act (HIPAA) delivers complex, critical information that speeds the planning, modeling, recruitment, and enrollment for clinical trials, and drives innovation in research and marketing for drugs and medical devices.
Provisio constructs and maintains comprehensive lifetime records for more than 80 million patients, as well as information on 500,000 physicians and healthcare providers. Because the data that Provisio and its iTrials service provide can play a key role in the success of groundbreaking healthcare research, and in the well-being of patients who agree to participate in clinical trials, highly efficient data management and analysis are critical.
A Database Emergency
However, with so much complex healthcare-related data accumulating rapidly, data analysis was a challenge for Provisio. Its database was distributed across several Microsoft SQL Server clusters, with about 230,000 database tables, and pinpointing the most relevant patient information was difficult and extremely time-consuming. In fact, processing complex, data-intensive queries in response to requests from customers—for example, how many people in a specific geographic area have Parkinson’s disease—could often take weeks.
Provisio knew that if it could not deliver faster performance when processing and refining complex queries, it risked delaying the important work of its customers and possibly missing out on new business opportunities. To meet its objective, Provisio turned to the ParAccel Analytic Database (PADB), the world’s fastest, most cost-effective platform for empowering analytics-driven businesses.
Technology Transplant Success
In late 2009, Provisio implemented PADB, which uses standard tools and hardware, into its existing platform; PADB runs on Linux, as well as the company’s HP servers. Provisio executives said they were impressed with ParAccel even before the company’s columnar-based, massively parallel processing (MPP) software solution started delivering results because it was one of only two providers that agreed to run a test query.
Paul Smith, CIO of Provisio’s iTrials service, said PADB from ParAccel has had a significant impact on the way the company does business, and allows Provisio to “gracefully manage” the health histories of more than 80 million Americans. And because data access times for large queries have been reduced from hours to seconds, iTrials can now offer real-time access to more than 250 billion points of healthcare data in the company’s platform. In addition, Provisio is able to provide new product lines that weren’t possible before.
More than a Full Recovery
ParAccel’s unprecedented ability to rapidly produce insight from very complex data exploration is helping Provisio’s executives uncover hidden revenue streams and launch new services. For instance, the speed and accuracy of PADB allows Provisio to deliver services on demand.
Provisio’s executive team expects that the ability to explore Provisio’s iTrials Data Universe—hundreds of billions of data points against an infinite range of criteria—on an ad hoc basis will provide greater insight into long-term patient outcomes and behaviors. Access to this type of information will also enable a broader range of users, including corporate, government, forensic, and academic organizations, to explore U.S. patient and provider data.
According to iTrials’ CIO Paul Smith, the “game-changing performance” of PADB positions iTrials and Provisio to seize new business opportunities and respond with speed to the processing-intensive data requests of customers.
This article originally appeared in the issue of .