Limited agreement between nutritional screening tools and CT-defined body composition alterations in hospitalized...

Nutritional screening tools are widely used in routine oncologic care to identify patients at nutritional risk. However, their ability to identify hospitalized patients with relevant alterations in body composition remains unclear. This study aimed to evaluate whether nutritional screening tools routinely used in clinical practice capture patients with CT-defined changes in muscle mass and quality.

This retrospective, cross-sectional study included hospitalized adult cancer patients treated at a tertiary cancer center in São Paulo, Brazil, between 2022 and 2023. Eligible patients underwent abdominal computed tomography (CT) and nutritional screening during hospitalization. Skeletal muscle index (SMI) and skeletal muscle density (SMD) were assessed at the third lumbar vertebra (L3) using CoreSlicer® software. Nutritional risk was evaluated using routinely implemented screening tools, and associations with CT-derived body composition parameters were analyzed.

Two hundred seventy-two patients were included. Low muscle mass was identified in 51.1% of patients and low muscle density in 52.2%, including individuals with normal or elevated body mass index. Nutritional risk identified by NRS-2002 was significantly associated with SMI but showed limited ability to identify CT-defined alterations when used in isolation. The CNS demonstrated high sensitivity but low specificity. Many patients with CT-detected alterations in muscle mass and quality were not classified as at nutritional risk by screening tools.

Widely used screening instruments may overlook clinically relevant alterations in body composition among patients. When imaging is already available as part of routine oncologic care, opportunistic CT-based assessment may complement screening and support more individualized supportive care interventions.