Automated severity scoring in level I trauma patients

Area of Science:

• Critical Care Medicine
• Trauma Surgery
• Health Informatics

Background:

• Accurate outcome prediction is crucial for managing trauma patients in Surgical Intensive Care Units (SICUs).
• Existing scoring systems like the Trauma Score (TS) and Injury Severity Score (ISS) are used, but their efficiency can be improved.
• Computerized physiological scoring systems offer potential for automated and consistent patient assessment.

Purpose of the Study:

• To evaluate the sensitivity and specificity of a computerized Simplified Acute Physiology Score (SAPS) for predicting outcomes in Level I trauma patients.
• To compare the predictive performance of SAPS against the combined TS and ISS.
• To assess the relationship between severity scores, ICU length of stay (LOS), and patient survival.

Main Methods:

• A retrospective study of 1434 consecutive trauma patients admitted to a SICU over 3 years.
• Automated calculation of SAPS using electronic flowsheets for the most abnormal values in the first 24 hours of SICU admission.
• Manual calculation of TS and ISS by trained nurses.
• Logistic regression and Receiver Operating Characteristic (ROC) curve analysis to compare predictive power.

Main Results:

• Analysis included 1085 patients (995 survivors, 90 non-survivors).
• Survivors had significantly lower mean SAPS (7.0) and shorter LOS (2.5 days) compared to non-survivors (SAPS 20.2, LOS 4.9 days).
• ROC curve analysis showed no statistically significant difference between SAPS and combined TS/ISS in outcome prediction (P > 0.70).

Conclusions:

• The computerized SAPS is an effective tool for outcome prediction in trauma patients admitted to the SICU.
• SAPS demonstrates equivalent predictive performance to the combination of TS and ISS.
• Automated physiological scoring systems can reliably aid in prognostication for critically ill trauma patients.