Prognostic of modified early warning score for identification of deteriorating patients on general wards

Abstract

Prognostic of modified early warning score for identification of deteriorating patients on general wards Objective: Most in-hospital adverse events do not happen without warning but are preceded by some period of physiological instability and clinical deterioration. MEWS have been introduced despite limited high quality evidence to demonstrate their sensitivity, specificity and usefulness. SOS score is a MEWS that is used in Phitsanulok, Thailand. This study assessed the predictive ability of SOS score at 4, 8, 12, 24 hours before adverse events (T0). Materials and Methods: We conducted a nested case-control study of adult patients who had adverse events in a general ward and died during June-July 2015 matched 1:2 with control patients who stayed in the same ward, same date and time and survived after discharge. Data were obtained from administrative databases and retrospective chart review. Discrimination of the SOS score at each time was assessed within receiver characteristic (ROC) analyses for admission SOS score and SOS score at 4, 8, 12, 24 hours before adverse events and corresponding area under the curve (AUC). The sensitivities and specificities of different cutoff thresholds were investigated. Results: 41 patients who died were selected to be the case group and 82 patients who survived were selected to be the control group, all from 5666 adult patients. Acute respiratory failure (68.3%) was the most common adverse event. More medical patients were enrolled in study than surgical patients (85.4% vs 14.6%). The SOS score at 4 hours before adverse events is the best predictor for adverse events with an AUC of 0.972 (95% CI, 0.949-0.995). However, the SOS score at 8, 12, 24 hours before adverse events are still good predictors for adverse events (AUC 0.906, 0.915, 0.860 respectively). The SOS score ≥ 4 at 4 hours before adverse events is the best cut-off value for adverse events with a sensitivity 82.9%, a specificity 95.1% and a diagnostic effectiveness 91.1%. The SOS score ≥ 4 at 8, 12, 24 hours before adverse events are still good cut-off values for adverse events with a specificity 95.1%, 96.3%, 92.7%, respectively. However, sensitivity fell when the time before adverse events was increased if the SOS score ≥ 4 to be the cut-off value was used. Conclusions: The SOS score at 4, 8, 12, 24 hours before adverse events is a good predictive ability for patients who had adverse events in a general ward. The SOS score ≥4 is reasonable for using this value to be cut-off point of trigger threshold to initiate action for worsening adverse events. The SOS score ≥4 had a good predictive ability regardless of the time intervals leading up to 24 hours before adverse events.