CATEGORY: Data Quality
REVIEWER: Lois Jones, information technology project manager for Kaiser Permanente.
BACKGROUND: Kaiser Permanente is America's leading integrated healthcare organization. Founded in 1945, it is a nonprofit, group- practice health maintenance organization (HMO) with headquarters in Oakland, California. Kaiser Permanente serves the healthcare needs of members in nine states and the District of Columbia. An integrated health delivery system, Kaiser Permanente organizes and provides or coordinates members' care, including preventive care (such as well-baby and prenatal care, immunizations and screening diagnostics), hospital and medical services, and pharmacy services.
PLATFORM:: IBM Mainframe running OS/390.
PROBLEM SOLVED: Kaiser Permanente has more than 18 million records regarding its insured members stored in a DB2 database. We needed the ability to better identify a person who called for assistance or with other general inquiries when only a name or address was provided. It was difficult to find members when searching by a member's name when the assumption was made that it had always been spelled correctly. The existing search process, a combination of wild-card searches and Soundex, was not doing the job well enough to provide quality service. Also, when adding new members, duplicates were allowed into the database when the member was already known to the organization. Basically, the search sub-system was not compensating for misspelled names and other variations in the names or addresses. The ability to quickly compensate for variations was seen as a primary requirement of the application. The problem had become rather significant and required immediate attention.
PRODUCT FUNCTIONALITY: Integration of SSA-NAME3 from Search Software America (SSA) into our current application has solved the various name search issues. SSA-NAME3's sophisticated searching and matching capability was able to overcome even the most significant problems with the data involved. It did so while still providing the ability to scale to a rather large data population. In addition, the SSA-NAME3 API was easily integrated into both the mainframe CoolGen application and on the client side to provide a client/server implementation. Our technical support team was able to implement the product following the instructions provided and by calling the SSA support line.
STRENGTHS: SSA-NAME3's strengths include its capability to find matches despite variations such as nicknames, misspelled names, hyphenated and multi-part names, transposed first and last names, and names from different cultural backgrounds. Even with 18 million records, there have not been any performance problems. In fact, performance has improved by 33 percent or more by denormalizing this highly used data. SSA- NAME3 provides full compatibility with the mainframe environment and was seamlessly added to the application. The support we received from SSA was excellent and consistently available via multiple channels.
WEAKNESSES: The "tuning" of the product required more time than its implementation. Adjusting the parameters to our population required some testing and retesting after each change. Search Software America has disclosed that this process will be easier in the next version of SSA- NAME3.
SELECTION CRITERIA:We were looking for a tool that would assist with real-time searching and matching, and SSA is a leading vendor in this space. Apart from the quality of the results achieved from SSA-NAME3, SSA was selected because of its stability, worldwide support and experience with large-scale critical systems.
DELIVERABLES: SSA-NAME3 delivered the keys that we needed to store in our database index and the search strategies that helped our applications find candidates for a particular search. The candidates were then scored and ranked by SSA-NAME3 so that our end users could easily pick from the most likely matches.
VENDOR SUPPORT:The support was one of the major criteria for our selection and consistently proved to be excellent. During the entire process we were able to either directly speak with a technical support individual or receive a return call within a short time. In addition, the account representative called on a frequent basis to ensure everything was going well and that we were getting the support we needed.
DOCUMENTATION: The documentation is quite thorough; but because it is thorough, it is quite large. We implemented the product using only the documentation and the telephone technical support. After implementation, we sent the people who would be maintaining the application to a formal training class.
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