Supplementary Materials

Supplementary Material for:

Targeted clinical control of trauma patient coagulation through a thrombin dynamics model

Amor A. Menezes, Ryan F. Vilardi, Adam P. Arkin,* Mitchell J. Cohen*

*Corresponding author. Email: aparkin{at}lbl.gov (A.P.A.); mitchell.cohen{at}dhha.org (M.J.C.)

Published 4 January 2017, Sci. Transl. Med. 9, eaaf5045 (2017)
DOI: 10.1126/scitranslmed.aaf5045

This PDF file includes:

  • Materials and Methods
  • Fig. S1. Summary of our model development process.
  • Fig. S2. Lack of correlation between trauma patient characteristics and outcomes.
  • Fig. S3. Overlap of normal and trauma patient classes.
  • Fig. S4. Model-free machine learning techniques applied to measured factor concentrations.
  • Fig. S5. Lack of correlation between trauma patient characteristics and principal components.
  • Fig. S6. Linearity, justification, and fit of a thrombin dynamical system model.
  • Fig. S7. Model control-theoretic and temporal parameter effects.
  • Fig. S8. K-means classification using control-theoretic parameters.
  • Fig. S9. Independence of factor addition to normal plasma.
  • Fig. S10. Parameter effects caused by increasing the concentration of factor II, VIII, or X in a sample of normal plasma.
  • Fig. S11. Parameter effects caused by increasing the concentration of factor II in five normal plasma samples.
  • Fig. S12. Parameter effects caused by increasing the concentration of factor VIII in 10 normal plasma samples.
  • Fig. S13. Parameter effects caused by increasing the concentration of factor X in 10 normal plasma samples.
  • Fig. S14. Proposed clinical workflow.
  • Table S1. Summary of trauma patient characteristics.
  • Legend for table S2
  • Table S3. Closeness of trauma patient factor concentrations to values from control mean plasma.
  • Table S4. Closeness of trauma patient factor concentrations to means from 20 normal plasma samples.
  • Table S5. Unsuccessful k-means classification of normal and trauma using four CAT dimensions.
  • Table S6. Unsuccessful k-means classification of normal and trauma using seven concentrations, equivalent to using the first two and the first three principal
    components.
  • Table S7. Conversion factors used for Hockin-Mann model prediction of trauma CATs.
  • Legends for tables S8 and S9
  • Table S10. Mean of the SD of nonspiked factor concentrations across multiple normal plasma samples.
  • Table S11. Maximum nonspiked factor concentration in normal plasma samples.
  • Legends for tables S12 and S13
  • References (1821)

[Download PDF]

Other Supplementary Material for this manuscript includes the following:

  • Table S2 Normal and trauma patient plasma sample data (provided as an Excel file).
  • Table S8 Thrombin dynamical system linearity testing data (provided as an Excel file).
  • Table S9 Factor addition independence testing data (provided as an Excel file).
  • Table S12 Factor perturbation data (provided as an Excel file).
  • Table S13 Ancillary factor perturbation data and model validation data (provided as an Excel file).