We will perform a retrospective cohort study to assess the construct validity and performance of days alive and out of hospital at day 90 (DAOH90) in cohorts of patients with sepsis and septic shock who have been included in recent clinical trials.

Study Design

We will conduct a retrospective cohort study. After obtaining all required ethics approvals, we will obtain data from the ARISE, CHEST, and ADRENAL studies. We will obtain from each dataset, for each trial participant with sepsis, data regarding; Demographic details, co-morbidities, pre-morbid living situation, baseline severity of illness, receipt of and duration of organ support during the index hospitalisation, Duration of ICU stay during index hospitalisation, ICU mortality, Hospital length of stay (LOS) and mortality for index hospitalisation, Discharge destination from index hospitalisation, readmissions to hospital up to 2 years, long term mortality, and quality of life as measured by the EQ5D at longest follow-up.

The primary objective of the study is to assess the association between Days Alive and Out of Hospital at day 90 and longer-term quality of life as measured by the EQ5D.

Secondary objectives include:

• To describe the characteristics of DAOH90 in patients with sepsis and septic shock with respect to the distribution and the pattern of variability

• To assess the construct validity of DAOH90, by testing the association between baseline variables and process of care variables and DAOH90, with the hypothesis that indicators for more "sickness" will be associated with fewer DAOH

  • Higher age
  • Higher APACHE II
  • Higher baseline lactate
  • Higher SOFA score at baseline
  • Premorbid accommodation in a long term care facility compared to independent living
  • More co-morbidities at baseline
  • Requirement for more organ support: requirement for 1, 2 or 3 (CVS, Resp, RRT) organ supports, and duration of organ support
  • Participants with septic shock compared to sepsis
  • Longer duration of admission for the index ICU admission
  • Longer duration of hospital admission for the index hospitalisation

• To assess the extent to which DAOH90 captures long term mortality and resource use, by comparing the DAOH at day 90 to the DAOH at day 30, day 180 and day 365.

Sensitivity analysis For the primary analysis we will calculate the Days Alive and Out of Hospital with DAOH=0 for all patients who die prior to day 90, regardless of whether the death occurred in hospital or subsequent to discharge.

We will conduct a sensitivity analysis with Days Alive and Out of Hospital = the number of days alive and out of hospital prior to death. For example of a patient had a one week hospital admission and then spent 5 days at home prior to death the number of days alive and out of hospital = 5.

We will conduct a subgroup analysis based on:

• The group of trial participants who were admitted to the ICU directly from the Emergency Department compared to those admitted from the Operating Theatre or ward
• Sex

Inclusion Criteria:

• We will include data from all participants in the ARISE and ADRENAL trials where there is data available to calculate the number of days alive and out of hospital to Day 90, and data available with regards to quality of life at 6 months following recruitment.
• We will include data from all participants in the CHEST trial who had a diagnosis of sepsis at baseline, and for whom there is data available to calculate the number of days alive and out of hospital to Day 90 and data available with regards to quality of life at 6 months following recruitment

Participants who have consented to enrolment into the ARISE, ADRENAL and CHEST trials

Data sharing would be considered by the respective management committees of the constituent trials on a case by case basis