Transforms of the computerized patient record data model: from transactions to analytical processing

The design of Computerized Patient Record databases needs to be different depending on whether the focus of the database is transaction processing or analytical processing. Transaction processing includes the operational use of databases for single patient update during clinical encounters. Analytical processing includes cross-patient querying for research or clinical management. These different design needs are defined. A case study is presented for the methodology that transforms transaction databases to analytical ones. The resulting analytical design is argued to be successfully evaluated according to specific criteria of useability and maintainability. This methodology is easily replicated in any relational database environment.