Designing and testing computer based screening engine for severe sepsis/septic shock

Area of Science:

• Critical Care Medicine
• Health Informatics
• Clinical Decision Support

Background:

• Sepsis remains a leading cause of mortality in intensive care units.
• Timely identification and treatment are crucial for improving patient outcomes.
• Electronic medical records (EMR) offer potential for automated sepsis screening.

Purpose of the Study:

• To evaluate the performance of an automated sepsis screening tool (sepsis "sniffer") for identifying patients with severe sepsis/septic shock.
• To assess the tool's accuracy in a real-world clinical setting within a medical intensive care unit.

Main Methods:

• Prospective implementation of the sepsis "sniffer" tool over a two-month period.
• The tool utilized electronic medical records for patient data analysis.
• Performance metrics including sensitivity, specificity, and positive predictive value were calculated.

Main Results:

• Out of 320 consecutive patients, 37 developed severe sepsis/septic shock.
• The sepsis "sniffer" achieved a sensitivity of 48% and a specificity of 86%.
• The positive predictive value of the tool was determined to be 32%.

Conclusions:

• The sepsis "sniffer" shows potential for early sepsis detection but requires further refinement.
• Current performance metrics suggest limitations for immediate clinical practice implementation.
• Additional research and development are needed to improve the tool's accuracy and utility.