Using abdominal computed tomography (CT) imaging, the investigators will estimate total body muscle mass at two time points in Intensive Care Unit (ICU) by assessing cross-sectional muscle areas at the L3 vertebral body level. This allows for a determination of the rate of sarcopenia development in the ICU.

With this information, the investigators propose to test if the rates of the development of sarcopenia differ in critically ill subjects with sepsis compared to a reference group of critically ill subjects with trauma (without sepsis).

The investigators propose to compare rates of development of sarcopenia in critically ill, septic subjects with a reference cohort of critically ill, trauma subjects who are not septic. The null hypothesis is that the rates of muscle loss in subjects with sepsis (a condition associated with high levels of inflammation) are comparable with the rates of muscle loss in subjects without sepsis. As secondary outcomes, the investigators will evaluate whether rates of muscle loss correlate with clinical assessments in the ICU (body mass index, frailty score, illness severity score), with clinical course through the hospital (ICU length of stay (LOS), duration of mechanical ventilation, hospital LOS, ICU readmissions, hospital readmissions), and with outcomes (discharge destination, ICU mortality, hospital mortality and six-month mortality).

To answer this question, the investigators propose to retrospectively analyze existing abdominal CT imaging of subjects treated in the ICU. Muscle mass estimations will be performed by determining skeletal muscle cross-sectional area for muscles at the level of the L3 vertebra using previously validated techniques. The CT images will have been gathered as part of routine care for subjects admitted to a mixed medical-surgical ICU. These data will be correlated with anonymized clinical and demographic information. Study participants will be grouped into two cohorts - subjects admitted to the ICU for sepsis, and subjects admitted to the ICU for trauma. People who experience both trauma and sepsis will be excluded. Only subjects with at least 2 or more abdominal CT imaging event will be included.

Results from this study will be assessed using a generalized linear model to determine if the two groups are different. From earlier studies, the investigators estimate that 73 subjects will be required in each group to detect between-group differences for the primary outcome.

The results of this study will add to the general understanding of the rates of muscle loss in a critical care setting for subjects admitted to ICU with sepsis, as compared to subjects with trauma. Secondary outcome information will be presented descriptively to show sarcopenia rates for the two cohorts, in the context of select clinical, demographic and administrative variables.

Inclusion Criteria:

• Admitted to ICU with sepsis or trauma
• Two abdominal CT imaging tests (separated by at least 24 hours, within the first 3 weeks of their ICU admission) performed as part of their routine clinical care.

Exclusion Criteria:

• Children (age < 18 years)
• Repeat admission to ICU within the previous 6 months
• Medical or neurological diagnosis that may plausibly affect the muscle bulk of the muscles at the L3 vertebra body level (e.g. achondroplasia, spinal muscular atrophy, spinal cord injury, etc.)

We are comparing muscle mass estimations based on previously ordered serial abdominal CT imaging. We are comparing sepsis to a reference group of trauma without sepsis. We anticipate that the sepsis group will contain subjects with abdominal sepsis primarily (as this would be a trigger for clinically testing patients with repeat abdominal CT imaging). We anticipate the the trauma cohort will consist of primarily splenic lacerations that are being monitored clinically, although this group will also likely contain patients with liver, renal, and intestinal injuries related to trauma.