REVIEWER: Dr. Patricia Cerrito, professor of mathematics at the University of Louisville.
BACKGROUND: The University of Louisville is a state-supported, metropolitan research university located in Kentucky's largest urban area. A part of the university as well as the largest concentration of healthcare providers in the state, Louisville Medical Center's member institutions are building a reputation for leading-edge medical research and breakthrough medical treatments.
PLATFORMS: Currently, we use the product locally on desktop PCs running the Windows 2000 and XP operating systems, and we are investigating licensing the product for use on a Windows NT server.
PROBLEM SOLVED: The team turned to SAS Text Miner to uncover potential areas for cost savings and to reveal new ways to improve patient care. By examining textual documents alongside large hospital databases, researchers have been able to discover relationships between physician practices and patient outcomes that affect overall care as well as healthcare costs.
PRODUCT FUNCTIONALITY: With SAS Text Miner, we can easily quantify and analyze text from various sources, including physicians' chart notes, billing records and medication orders to strengthen our research in cancer, genetics, heart disease and environmental health. The solution highlights relevant patterns in documents and quantifies text-based information, allowing us to solve patient-care problems that have existed for decades. Since implementing SAS, we've learned that prescriptions of certain medications resulted in prolonged hospital stays. Further, we have been able to use the product's automatic extraction of patient information from the pharmacy order database to enhance and, we hope, eventually replace manual extraction from patient charts. Using insights derived from SAS, we are also able to regularly present and publish research results that may reduce hospital costs and improve patient care worldwide.
STRENGTHS: SAS integrates its text mining capabilities into its data mining solution better than any other vendor. Further, SAS has the best data mining algorithms and the simplest interface for managing and importing data. Tight integration between tools allows us to move quickly between applications and explore structured data and unstructured text more deeply than ever before. Consequently, we have better patient care and more efficient reporting, which, in turn, improves the hospital's overall medical ranking. We've also found SAS to be a great tool for training researchers in investigating clinical data. In fact, we are currently preparing online courses to begin this fall that integrate data mining and healthcare outcomes.
WEAKNESSES: Overall, the product is helpful and easy to use. It would be beneficial, however, if macros specific to certain databases (such as Medline, which I use quite a bit) were built into the product. When I suggested this to SAS, they were responsive to my request and are currently working with me to meet my specific needs.
SELECTION CRITERIA: We chose SAS because it has the most statistically sound algorithms. No other software delivers the same depth and breadth of analytic functionality. We needed something with the ability to solve both extremely narrow and very broad questions; and with SAS, I am able to uncover relationships in the data that clinicians may not have initially considered, leading to improved and innovative patient care.
DELIVERABLES: SAS provides extensive statistical capabilities to meet specialized and hospital-wide analytic needs. SAS Text Miner allows my team to examine relationships between physician practices and thousands of patient outcome records. Coupling quantitative representation of text with traditional data mining techniques, SAS pulls relevant variables from patient charts, making it easy for us to find patterns in the data. With SAS, we have been able to save money and improve patient care - something many hospitals find difficult to achieve.
VENDOR SUPPORT: SAS was extremely helpful during both the pre- and post-implementation phases of the project. Three developers from SAS spent two days with us, pointing out ways to improve our documentation and walking us through each step of the text mining process so we could better understand how to achieve the best results. I've used SAS for many years, and I have yet to work with another vendor who works so closely with its customers to meet their needs and seek their input on future enhancements.
DOCUMENTATION: A long-time SAS user, I was able to learn to use the product without documentation. However, I find both the online help and the course notes for the product to be helpful when needed. I do have one area of improvement for the documentation: The documentation concentrates too heavily on the techniques of working with the product without a thorough treatment of the concepts behind those techniques and how those techniques can assist you in your daily activities. I am currently working with SAS to address that gap.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access