Queensland Adult Deterioration Detection System observation chart diagnostic accuracy in detecting patient deterioration: A retrospective case-control study

Background: Queensland Adult Deterioration Detection System (Q -ADDS) is the acute -care vital sign observation chart used widely throughout Queensland, Australia. The diagnostic accuracy of the chart in detecting patient deterioration is unknown. Aim: This study aims to assess how accurately the Q -ADDS observation chart predicts patient deterioration in acute -care hospitals and the contribution of each vital sign in predicting patient deterioration. Methods: This multi -centre retrospective case-control study compared vital sign data of 1152 patients that suffered a deterioration event and 1088 demographically and diagnostically matched non -deterioration patients. The efficacy of the Q -ADDS chart was determined by calculating the Area Under the Receiver Operator Characteristic Curve (AUROC) after logistic regression and Random Forest (RF) classification using the individual vital signs, and the aggregated Q -ADDS score. Findings: Q -ADDS predicted patient deterioration with above -chance accuracy 6 hours before the deterioration event (AUROC = 0.690), comparable to an optimised RF model using the same vital sign data. At the time of the deterioration event, the Q -ADDS performed at parity with the optimised model (AUROC = 0.907). The aggregated Q -ADDS score was a better predictor of patient deterioration than any individual vital sign. Discussion: Q -ADDS predictive validity is weaker than several other Early Warning Systems. However, its ability to discriminate between deteriorating and non -deteriorating patients is above the level expected by chance. No individual vital sign is a strong predictor of patient deterioration, but the aggregated weighted Q -ADDS score is a good deterioration predictor. Conclusion: The multivariable Q -ADDS score efficiently predicts clinical deterioration in acute -care hospitals, with the tool's discriminatory capacity increasing with proximity to the deterioration event. (c) 2023 Australian College of Nursing Ltd. Published by Elsevier Ltd. This is an open access article under the CC BY -NC -ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).