Input Variable Sepsis Prediction Algorithm Clinical Practice Age 60 years (Karlsson et al[@116269]) Incidence 14.1% (Sakr et al[@116249]) Sensitivity/specificity 80.0%/85.1% (NAVOY® Sepsis) 79.2%/78.5% (Lengquist et al[@116267]) Total time to sepsis treatment (compared with time of diagnosis for SOC) True positives     -3 hours True positives    0 hours False negatives     +3 hours False negatives     +3 hours Proportion septic shock when detection is set 3 hours earlier than SOC 40% (assumption) Proportion septic shock SOC (at 0 hours) 56.6% (Sakr et al[@116249]) In-hospital mortality True positives (-3 hours)    Septic shock, 39.5%    No septic shock, 0% False negatives (+3 hours)    Septic shock, 46.4%    No septic shock, 1.6% (Ferrer et al[@116257]) True positives (0 hours)    Septic shock, 42.8%    No septic shock, 0% False negatives (+3 hours)    Septic shock, 46.4%    No septic shock, 1.6% (Ferrer et al[@116257]) Postdischarge mortality first year (applied for patients with septic shock) 17.5% (Karlsson et al[@116269]) Readmission first year (all sepsis patients) 20.6% (Zilberberg et al[@116268]) ICU days Patients with sepsis     Septic shock, 7.4 days (Ferrer et al[@116257])    No septic shock, 2.0 days (Ferrer et al[@116257]) Patients without sepsis     True negatives, 1.0 days (Swedish registry data[@116272])     False positives, 2.0 days (assume same as “no septic shock”) Ward days Patients with sepsis     Septic shock, 17.0 days (Ferrer et al[@116257])    No septic shock, 5.7 days (Ferrer et al[@116257]) Patients without sepsis     True negatives, 5.7 days (assume same as “no septic shock”)    False positives, 5.7 days (assume same as “no septic shock”) Long-term consequences for patients with sepsis (frequency) Impaired kidney function (14.1%) Amputation (8.5%) Depression (2.8%) PTSD (9.9%) (York Health Economics Consortium[@116271]) Long-term survival Patients with sepsis     RR vs general population (Linder et al[@116270]) Patients without sepsis     General population survival Discount rate 3.0%