Thrombelastographic pattern recognition in renal disease and trauma

Area of Science:

• Hemostasis and Thrombosis
• Clinical Pathology
• Medical Diagnostics

Background:

• Thrombelastography (TEG) is a viscoelastic assay measuring hemostasis.
• End-stage renal disease (ESRD) and trauma-induced coagulopathy (TIC) present unique TEG profiles.
• Distinct TEG patterns may differentiate these conditions from healthy controls.

Purpose of the Study:

• To determine if specific TEG patterns can accurately distinguish between ESRD, TIC, and healthy controls.
• To develop a classification system for interpreting TEG data in these patient populations.

Main Methods:

• TEG analysis was performed on blood samples from 54 ESRD patients and 16 trauma patients requiring massive transfusion.
• Independent TEG parameters were plotted to identify disease-specific patterns.
• A decision tree classification model was constructed using R-Project statistical software based on maximum amplitude (MA) and activated clotting time (ACT).

Main Results:

• Distinct clusters of TEG results were observed for ESRD, TIC, and control groups when plotting MA versus ACT.
• A classification tree using ACT (103 s) and MA (60.8 mm or 72.6 mm) achieved a 93.4% correct classification rate.
• ESRD showed prolonged clotting with normal/high clot strength, while TIC exhibited prolonged clotting with weakened clot strength.

Conclusions:

• ESRD and TIC exhibit unique and distinguishable TEG patterns.
• The developed taxonomic categorization provides a rigorous, algorithmic approach to TEG interpretation.
• This system has the potential to form the basis for clinical decision support software for viscoelastic hemostatic assays.