Abstract
Nurses often need to make quick decisions with incomplete diagnostic information while they are under time pressure. The use of a data-driven, computerized decision support approach in daily work activities has great potential to facilitate precise and context-sensitive use of the information implicit in nursing diagnoses. This study explored optimal information amounts and levels of information content for designing and implementing a diagnostic nursing decision support system. Specifically, the use of probability data for likely nursing problems and the preferred number of displayed nursing problems were presented to expert and novice nurses. The study used a counterbalanced, repeated-measures, and factorial design. The authors developed two scenarios: (1) a pneumonia patient with diabetes mellitus complications and (2) a patient with controlled diabetes who also had a bone fracture. A previously developed prototype for a diagnostic nursing decision support system was used to display the information. Eighteen novice and expert nurses from two hospitals in Korea participated. Results for the differing levels of content did not differ significantly with level of expertise, but the preferred amount of information was significant for the two groups of nurses. The differing information needs of novices and experts should be considered when designing future computer-based decision support.