A retrospective observational study of sarcopenia and outcomes in critically ill patients.

Frailty assessment in patients admitted to intensive care is often limited using traditional clinical frailty assessment tools. Opportunistic use of contemporary computed tomography (CT) can provide an objective estimate of low skeletal muscle mass (sarcopenia) as a proxy for frailty. The aim of this study was to establish the prevalence of sarcopenia in an Australian intensive care unit (ICU) population and to examine the relationship between sarcopenia and clinical outcomes. We undertook a single centre retrospective study of 1085 adult patients admitted to a single ICU over 12 months. Patients with a contemporary CT scan including the L3 vertebral body were included. Patients were categorised as sarcopenic or non-sarcopenic using previously published data. A total of 279 patients with a mean age of 67 years had an eligible CT scan; 163 (58%) were male. Higher 30-day mortality was associated with the use of CRRT (continuous renal replacement therapy) during the ICU admission (OR 6.84,  < 0.001) and also associated with lower cross-sectional muscle area (odds ratio (OR) 0.98,  = 0.004). Sarcopenia was found to be highly prevalent in this particular Australian ICU population (68%) and associated with older age (68 versus 55 years,  < 0.001), lower body mass index (27 versus 32 kg m,  < 0.001), more comorbidities (3 versus 2,  = 0.009), and longer stays in hospital (279 versus 223 h,  = 0.043). As a continuous predictor, lumbar muscle mass was associated with 30-day mortality with and without adjusting for other covariates.