Assessment of muscle mass using ultrasound with minimal versus maximal pressure compared with computed tomography in critically ill adult patients.

Background: Preserved skeletal muscle mass identified using computed tomography (CT) predicts improved outcomes from critical illness; however, CT imaging have few limitations such that it involves a radiation dose and transferring patients out of the intensive care unit. This study aimed to assess in critically ill patients the relationship between muscle mass estimates obtained using minimally invasive ultrasound techniques with both minimal and maximal pressure compared with CT images at the third lumber vertebra level.

Methods: All patients were treated in a single Australian intensive care unit. Eligible patients had paired assessments, within a 72-h window, of muscle mass by ultrasound (quadriceps muscle layer thickness in centimetres, with maximal and minimal pressure) and CT axial cross-sectional area (cm). Data are presented as mean (standard deviation), median (interquartile range), and frequencies [n (%)].

Results: Thirty-five patients [mean (standard deviation) age = 55 (16) years, median (interquartile range) body mass index = 27 (25-32) kg/m, and 26 (74%) men] contributed 41 paired measurements. Quadriceps muscle thickness measured using the maximal pressure technique was a strong independent predictor of lumbar muscle cross-sectional area. Within a multivariate mixed linear regression model and adjusting for sex, age, and body mass index, for every 1 cm increase in quadriceps muscle layer thickness, the lumbar muscle cross-sectional area increased by 35 cm (95% confidence interval = 11-59 cm). Similar univariate associations were observed using minimal pressure; however, as per multivariate analysis, there was no strength in this relationship [8 cm (95% confidence interval = -5 to 22 cm)].

Conclusion: Ultrasound assessment of the quadriceps muscle using maximal pressure reasonably predicts the skeletal muscle at the third lumbar vertebra level of critically ill patients. However, there is substantial uncertainty within these regression estimates, and this may reduce the current utility of this technique as a minimally invasive surrogate for CT assessment of skeletal muscle mass.