Implementation of Complementary Model using Optimal Combination of Hematological Parameters for Sepsis Screening in...

Area of Science:

• Clinical Medicine
• Biochemistry
• Computational Biology

Background:

• Early detection and treatment significantly improve sepsis patient outcomes.
• Current sepsis scores (SIRS, SOFA, LODS) use limited hematological parameters, despite broader associations of clinical pathology parameters with sepsis outcomes.
• There is a need for complementary models to enhance existing sepsis screening criteria.

Purpose of the Study:

• To develop and validate a complementary model for sepsis screening using a wider range of hematological parameters.
• To compare the performance of the new model against existing sepsis-related clinical scores.

Main Methods:

• Statistical analysis of multiple clinical pathology parameters from sepsis and fever patient groups.
• Development of a complementary model using stepwise parameter selection and machine learning.
• Performance evaluation using Area Under the Curve (AUC) metrics.

Main Results:

• The developed complementary model demonstrated superior performance with an AUC of 0.86.
• This performance was significantly better than models based on specific hematology parameters within existing sepsis scores (AUC 0.74-0.51).

Conclusions:

• A novel complementary model based on an optimal combination of hematological parameters offers improved sepsis screening in patients with fever.
• This model provides a valuable tool to augment existing clinical criteria for early sepsis detection.