• Comorbidities (CIRS-G scale),

• Functional status of the patient with

  • Clinical frailty scale (1 month before infection)
  • ADL score (1 month before infection)

• Biological data

  • Blood group
  • TP, D-dimers, CRP, creatinine level at the patient's entry, triglyceridemia, fibrinogen, ferritin
  • Parameters derived from the platelet formula count on D1 of the start of intensive care (lymphocytes, neutrophils, platelets, average platelet volume, red blood cell distribution index), SYSMEX data (IG: Immature granulocytes; HFLC: high fluorescent lymphocyte count)

• Resuscitation outcomes

  • LDH rate at the start of intensive care
  • PaO2 / FiO2 ratio at the start of intensive care
  • IGSII / SASPII score (simplified acute physiology score) on D1 of the start of intensive care
  • SOFA score (sepsis-related organ failure assessment): a posteriori estimate based on IGSII / SASPII
  • Delay between the appearance of the first signs of infection and admission to intensive care

First hypothesis: considering a single analysis variable (age), with expected mortality of 30% in patients under 70, and 70% in patients over 70 (with 40% of patients over 70), a total of 130 patients is necessary to show a statistically significant difference between these two groups with a power of 90% (bilateral alpha risk test of 5%). Since the multivariate analysis considers the integration of several factors, considering 15 factors, hoping for a coefficient of determination of 0.5 of the model, to achieve an optimism of less than 10%, it will be necessary to include 185 patients.

After the publication of data on mortality in ICU in Lombardy region, Italy in April 2020, it was considered that a stopping of the trial at 185 patients would impair its statistical power and induce a potential risk of patients' selection bias. As a consequence the scientific committee decided that all the patients admitted to ICU until May 7th would be proposed the study.