Designing observation charts to optimize the detection of patient deterioriation: reliance on the subjective preferences of healthcare professionals is not enough

Aim: Observation charts are critical to detecting patient deterioration. Research suggests their design has a dramatic impact on user performance in terms of failure rates for detecting abnormal vital signs and how quickly users can interpret recorded observations. In this study, we examined the design preferences of professional chart users to assess their alignment with objective performance data. In addition, we tested the assumptions of prior knowledge that chart designers appear to have made about chart users.

Methods: We conducted an online survey of health professionals (n=347). Participants answered questions about their observation chart design preferences in general, and were randomly assigned to evaluate one of nine specific charts.

Results: Chart users' preferences for design features were not always consistent with objective performance data. While some views concurred with empirical findings (e.g., participants preferred to plot observations on a graph with graded colouring, where the colours corresponded with degrees of abnormality), others did not (e.g., participants preferred plotting blood pressure and pulse together on the same chart area, which the objective data suggest is problematic). Additionally, a substantial proportion of respondents were unfamiliar with some of the assumed knowledge required to interpret many charts (e.g., particular abbreviations).

Conclusions: It is dangerous to rely solely on subjective opinions - even those of experienced health professionals - when developing patient observation charts, as optimal design may be counterintuitive and some preferences may merely reflect familiarity. Objective performance data is also required. In addition, the level of assumed knowledge required to use a chart should be minimized.