Trend Displays to Support Critical Care: A Systematic Review

Abstract

An important aspect of designing information displays to support monitoring and decision-making in critical care is the representation of change of patient data over time. We systematically reviewed articles to identify novel alternatives to tabular and single variable plots of values over time to convey information about change in multiple related variables. Following screening of 5,119 articles, 28 met our inclusion criteria. They described 26 unique displays evaluated in 31 experiments. Methods for representing change varied widely. We classified these methods as enhanced graphical displays (enhanced plots of quantitative data over time), other displays (novel object and metaphoric displays), small multiples displays (multiple co-presented small graphic displays), and simple change indicator displays. Overall, findings support the value of an explicit display of trend information using many different approaches. Few studies directly compared different methods for displaying trend information in ways that would support broader conclusions about which approaches may be preferred for specific applications. The studies suggest that, for displaying patient data trends, it is feasible to develop electronic displays that will outperform both historical paper-based flowcharts and current electronic health record (EHR)-based tabular approaches. There is evidence to suggest that even minor improvements to current approaches such as the automatic presentation of simple line plots of trends on EHRs or the addition of simple graphical indicators of trend direction on patient monitors could lead to clinically meaningful improvements in diagnostic accuracy and efficiency.