BACKGROUND: Center for Health Program Development and Management (CHPDM) was established to provide analytical and oversight services to Maryland's Department of Health and Mental Hygiene (DHMH) for quality assurance and use of the Medicaid program. These services affect the entire state's population and hinge on access to lots of data sources.

PLATFORMS: Sun SparcStation 1000E, Novel network, Pentium PCs.

PROBLEM SOLVED: We built a prototype in three months and a working warehouse in six months. A data mart maximized access to detailed characteristics of a wide range of subjects. For instance, policy analysts and researchers are interested in exploring data on individual doctors, dentists, labs, pharmacies, medical equipment providers, long-term care facilities, hospitals, drugs, etc. ­ a disparate and incompatible range of data sources. Researchers sing our praises because they are finally able to do something never before possible ­ access data!

PRODUCT FUNCTIONALITY: The data mart provides quick and easy access to the summarized data required to make accurate projections and achieve performance goals. It automates a manual entry process, makes scheduling much easier and helps ensure use of our latest and best programs. The most beneficial aspect of the Warehouse Administrator is how it handles change. When a small change has to be made to the data, the Warehouse Administrator is intelligent enough to carry changes everywhere. The success of the data mart is demonstrated regularly in the Maryland State Legislature where decisions are sometimes based solely on reports from CHPDM's data mart.

STRENGTHS: This project can be seen as innovative from several angles. Organizationally, no other state operates this way. The methodology used in the project included: resource analysis, business process studies, incorporating critical success factors, performance measures, entity definitions and relationships, individual interviews and a joint application design meeting. MDDBs allow for quick access to the data across a variety of dimensions. Our data is such that studies are conducted on a group or groups of recipients. We identified 16-20 identifiers for each recipient with several of those being "drill-down" hierarchies.

WEAKNESSES: Building the MDDBs was a learning process for everyone. We used the Warehouse Administrator to develop all the meta data for the system and to store all programs which were used in creating any preprocessing of files leading up to the creation of the final detail tables. The initial design was our best-guess feel for what was usable.

SELECTION CRITERIA: To make competent and educated choices to these changes, health care policy analysts need easy access to data. Studies of diagnoses, procedures, providers, etc., can assist analyses that can be run by the individual analyst at his or her desk, providing the same outcome no matter how the information is accessed. We were one of the first to use SAS's Data Warehouse Administrator to create our meta data. Even though it was new technology, we saw the value in a single point of control and automatic documenting of all the processes involved to deploy the data mart.

DELIVERABLES: To succeed, we needed commitment to incorporate business rules appropriately along the way. Though data experts were involved, we were dependent on business users for key usage information. Medicaid data comes from all over the state. Because of the non-standard data, the software had to deal with inherent inconsistencies. One MDDB needed to include recipient demographic and eligibility information and eight different types of claims files, each with different structures. This design phase was one of the most grueling ­ yet rewarding ­ work experiences some of us ever had.

VENDOR SUPPORT: SAS had data warehousing knowledge, database design expertise and vast experience with the software. It took education to help users understand what could and could not be provided through a data mart. It was a challenge to balance understanding user requirements and the ability to deliver on promises.

DOCUMENTATION: Data sources were easily identified, but what occurred during the design process was a narrowing of the scope. The Department of Health data existed from two different sources. Much of the new data had not been verified or validated so a decision was made to include only the earlier data sources in the data mart with phase 2 to include the newer data.

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

Don't have an account? Register for Free Unlimited Access