Modified early warning score in assessing disease conditions and prognosis of 10,517 pre-hospital emergency cases

Objectives: To study the values and feasibility of modified early warning score (MEWS) in the disease condition assessment and death prediction in pre-hospital emergency cases (PHECs). Methods: 10,517 PHECs were collected as the study subjects, the data were collected on the scene for MEWS; 90-day was set as the end point and the patient's outcome was set as the observation indicators. The counting data were expressed as percentage and performed the X-2 test; the measurement data were expressed as mean +/- standard deviation ((X) over bar +/- s) and performed the group t test, with P < 0.05 considered as statistically significant. The corresponding predictors of the area under curve (AUC) of the MEWS characteristic curve were then calculated. Results: MEWS of the dead patients was significantly higher than the patients survived (P < 0.05), with the sensitivity as 88.45%, specificity as 71.28%, accuracy as 87.13%, Youden index as 0.5973, and AUC as 0.883. MEWS >= 4 points was the cut-off point to distinguish the critical conditions and predict the risk of death. Conclusions: MEWS had certain distinguishing abilities towards the disease condition assessment and death prediction in PHECs, and could be used as an effective tool to manage PHECs.