||Critical care environments are difficult to computerize. Conceptual data management challenges include temporal intensity of the data, currency issues, simultaneity of multiple processes and the need to manage extensive human interactions while maintaining security. In critical care, where minutes count, 80-90% of collected data are temporally-related. Complex temporal, verbal, visual and contextual reasoning tasks are unconscious mental modelling strategies of expert critical care nurses and doctors. UtahNICU"" is a conceptual design project to address clinical experts' needs in Neonatal Intensive Care (NICU) by ""view integrating"" a legacy patient database with a relational, physiologic research database designed to support real-time NICU bedside data management. The retiring database. Legacy!, was a temporally linear (admission, interim, discharge), dataflow-oriented, non-relational database application with unidirectional links to lookup files to generate hardcoded reports offsite. The Legacy2 database was impossible for clinical personnel to maintain or query. Inadequate healthcare coding and missing time-stamping led to scattering of medication, procedure and results data in narratives throughout the database. We chose to develop a relational model with time-series capabilities to emulate the clinical experts' mental modelling of complex ICU data. Complex data contain time-dependent, highly variant, repeated measures information with missing values. We categorized Legacy2's 799 fields into six clinical data element categories, then internally remodelled each following the research database's relational, time-sensitive model. The final design added conceptual structures (family, social, procedures) missing from Legacy2. The new design focused on the entity, INFANT. All other included constructs related directly or secondarily to the infant patient. The new UtahNICU logical design can incorporate Legacy2 data, provide core and ancillary capabilities for a 1999-standard daily note for new data and is normalized to fourth normal form. UtahNICU will support interval summaries and user queries by bedside practitioners. The new database will meet the owners' specifications for data preservation, accessibility and institution independence with existing database software (OracleS) and hardware (high-end PC computers/servers) at acceptable cost.""