A computer was programmed to collect, store, analyze, and display blood gas data in a newborn intensive care unit. Data were displayed if they were (1) markedly abnormal or (2) represented a worsening trend. A controlled study demonstrated that, with the display, the markedly abnormal blood gas values were followed by normal values in a shorter period, and fewer worsening trends progressed. However, with the computer-generated display, there were more overcorrections of both the markedly abnormal blood gas values and the detected worsening trends. The occurrence of pneumothoraces was associated with these overcorrected blood gas values. There were no significant differences in duration of supplemental oxygen administration, duration of tracheal intubation, or mortality between the infants cared for during the time of the computer-generated display and those cared for during the control period. This study demonstrates both benefits and risks of computer-generated displays and emphasizes the need for thorough evaluations of such systems.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/s0022-3476(83)80495-8 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!