Detecting critical on-line information: The relationship between nurse characteristics, computer screen designs, and computer interaction measures during laboratory results retrieval tasks

Abstract

This study focused on the match between nurse characteristics and computer interface designs during critical information detection tasks and examined: (1) which of three computer screen density levels promoted the fastest, most accurate target detection performance and greatest subjective screen satisfaction among clinical nurses, and (2) which of a set of selected cognitive and demographic nurse characteristics were related to nurses' practiced performance speed, accuracy, and subjective screen satisfaction. A conceptual framework for research in nurse-computer interaction was developed by integrating concepts from human-computer interaction, nursing informatics, and developmental psychology. The study sample was 110 randomly selected clinical nurses from an East-coast university medical center. A two-factor within subjects repeated measures analysis was used for the study's screen design section. Overall and for practiced tasks, nurses found information targets significantly more quickly on high density than moderate or low density screens during practiced tasks. Also, nurses found information more quickly on moderate versus low density screens. Nurses' mean accuracy and subjective screen satisfaction scores were essentially the same for all screen types. These results suggest increases in screen information density can result in faster performance speeds without sacrificing nurses' accuracy or screen satisfaction. Using step-wise multiple regression analyses, nurses' age, spatial memory, and spatial visualization explained 35.9% of overall performance speed and 21.5% of accuracy variance. Younger nurses with higher spatial memory and visualization abilities had faster performance speeds and higher accuracy rates. Nurses' previous computer experience represented a basement effect. Nurse characteristic predictors explained 27.8% of high density screen performance speed, 10.9% for moderate, and 22.5% for low density screens. During step-wise regression, age emerged as a significant predictor for all three screen types, and spatial memory was a significant predictor for high and low density screens. Spatial visualization emerged as a significant predictor for performance speed on high and moderate screens but not low. The set of predictors explained little accuracy during practiced tasks, 12.4%, and virtually none of the screen design satisfaction variance, 6%. These results suggest nurse educators and system designers may need to design visual aids and/or computer training to accommodate nurses' age and cognitive individual differences.