Design features for clinical decision support systems for detecting clinical deterioration in hospitals: A scoping review.

Abstract

Background: Clinical Decision Support Systems (CDSS) can support clinicians in the timely detection of patients’ clinical deterioration in hospitals. Despite this, less than half of CDSS implemented for clinical deterioration are used by clinicians, with poor usability and bad design as contributing factors. Objective: The aim of this study was to identify design features and factors influencing the acceptance of CDSS for real-time detection of clinical deterioration.   Methods: We conducted a scoping review and searched three databases: Medline, Scopus and CINAHL. Studies describing the design features and factors influencing acceptance of CDSS for detecting deterioration in hospitals (including hospital-in-the-home) were included. A qualitative narrative synthesis was undertaken.   Results: Of 22 eligible articles, 17 described CDSS in the form of a dashboard and 5 described alerts. Of the 17 dashboards, 70% (12/17) used colour to signal hazards, with 42% (5/12) using a traffic light colour-coding system to classify patient risk level. 18% of dashboards used tables to display vital signs (3/17). Clinicians preferred dashboards that included both graphs and tables. Factors such as manual recording of vital signs (3/22) and high alarm rates (4/22) were shown to decrease acceptance of CDSS.   Conclusions: Most dashboards for detection of patient deterioration use a colour-coding system to classify risk. High acceptance of CDSS by end-users is likely when both tables and graphs are used to display vital signs, vital signs are continuously monitored, and when alarm frequency is low.