Using elastic net prediction modeling and tree classification, we identified cytokines that were able to classify subjects as having CRS or sepsis accurately.

This combination of IFNγ and IL1β was able to categorize subjects as having CRS or sepsis with 97% accuracy.