ABSTRACT
BACKGROUND: Critical care information systems (CCIS) are computer-based systems designed to process the growing amount of complex medical data in intensive care units (ICU). Previous studies have shown that CCICs can increase the quality of patient care by reducing errors and improving work efficiency; however, other studies have shown that CCISs can also cause harmful effects by disrupting workflow, facilitating medication errors or increasing charting time. The factors that decide whether a CCIS has a positive or negative impact on patient care are summarized under the term "usability". This article summarizes the results of three previously published papers on this topic. OBJECTIVE: The aim of the study was to identify which CCIS functions were considered useful by clinical ICU staff and how well these functions are implemented in the CCISs currently used in German ICUs. MATERIAL AND METHODS: An online survey was performed targeting nurses and physicians working in German ICUs using a previously validated questionnaire. The questionnaire included a list of functions (36 for physicians/31 for nurses) that were preselected by experts based on a comprehensive model of ICU work processes. Each of these functions was rated by the study participants on a Likert scale ranging from 0 (worst rating) to 5 (best rating) with respect to the usefulness to identify which functions of CCIS can truly be considered as useful by clinical ICU staff. Furthermore, the participants rated how well these functions were implemented in the CCIS currently in use on the ICU, also using a Likert scale of 0-5. Further questions were provided to rate specific technical usability aspects of the CCISs currently in use. In addition, to capture possible confounders the questionnaire recorded 18 individual and workspace characteristics which might influence the ratings. RESULTS: A total of 171 nurses and 741 physicians participated in the survey of which 535 used CCISs. Of the functions 33 were rated as useful for doctors and 28 functions for nurses with median scores between 4 and 5. Participants currently using CCISs gave higher ratings compared to participants not using CCISs. The quality of the functions was rated relatively lower than the usefulness and the availability. Furthermore, currently used CCISs in Germany differ greatly in their technical and task-specific usability. Of the CCISs investigated, the system ICUData had the best overall rating and technical usability followed by the systems ICM and MetaVision. The same three CCIS were rated best in task-specific functions without significant differences between them. CONCLUSION: Those functions that were identified as useful based on the ratings of clinical ICU staff should be implemented in current CCIS. The list of these functions might be regarded as a first step towards providing a catalog of functional requirements for CCISs. Furthermore, as the results show that the quality of the available functions was rated lower than the availability of the functions, manufacturers should shift more of the effort away from the development of new features and focus on improving the user-friendliness and quality of existing functions.
