Less is more: the design of early-warning scoring systems affects the speed and accuracy of scoring

Aim: To evaluate the effect of early-warning scoring system design on the speed and accuracy of scoring. Background: Despite the widespread implementation of early-warning scoring systems in hospitals, the speed and accuracy with which chart-users determine patients' early-warning scores has received minimal research attention. Design: Within-subjects, with scoring-system design as the independent variable. Methods: Forty-seven novice chart-users were presented with realistic vital sign observations recorded on charts with three different scoring-system designs. The rows for recording individual vital sign scores were either: (1) grouped together beneath all of the vital sign rows; (2) separated, with each row presented immediately below the corresponding vital sign row; or (3) excluded altogether. Participants' response times and error rates for determining the overall scores were measured for 54 time-points per design. Data were collected in December 2012-January 2013. Results: Contrary to predictions, participants responded fastest and made the fewest errors when using the chart design without individual vital sign scoring-rows. For the other two designs, participants were faster when the rows for scoring individual vital signs were separated (vs. grouped), but accuracy did not differ. For both of these designs, significantly more time-points were affected by scoring errors compared with adding errors. Finally, data for patients with more serious derangements yielded greater response times and error rates on all three charts. Conclusion: Early-warning scoring systems may be more effective without individual vital sign scoring-rows. Even when charts are designed by multi-disciplinary teams of human factors specialists and clinicians, empirical evaluations are essential