Abstract

Background: Muscle ultrasound is a valid tool to monitor muscle mass loss in critically ill patients. The level of experience is essential to the accuracy of the measurements. Aim: To evaluate the interobserver reliability of experienced and novice raters measuring muscle thickness and echo intensity of the quadriceps and tibialis anterior. Material and Methods: Cross-sectional observational study. Twenty-four critical care physiotherapists participated (5 experienced and 19 novice). Following a standardized ultrasound protocol, each rater measured the thickness (centimeters) of the quadriceps and tibialis anterior of 10 healthy and young models using linear and convex probes of portable devices. The Intraclass Correlation Coefficient and the Minimal Detectable Change (95% confidence interval) were calculated. Additionally, the novices scored the echo intensity of 19 muscle ultrasound images of critically ill patients using the Heckmatt score (qualitative assessment). The agreement with experienced raters was evaluated (Spearman Rho). Results: 960 muscle thickness measurements were performed (experienced = 200 and novice = 760). The mean thickness of the quadriceps and tibialis anterior was 4.4 +/- 0.77 and 2.4 +/- 0.35 centimeters for the experienced and 4.2 +/- 0.80 and 2.2 +/- 0.39 centimeters for the novices, respectively. Quadriceps' and tibialis' anterior reliability were 0.82 and 0.86 for experienced and 0.76 and 0.41 for novices, respectively. The Minimal Detectable Change ranged from 0.14-0.33 centimeters. The mean Heckmatt score was 2.6 +/- 0.83 points, with a reliability of 0.68 and an agreement with the experimenters of 0.78 [p < 0.001]. Conclusions: Interobserver reliability was excellent for experienced raters and moderate to good for novice raters. The level of experience could determine the reliability of the results.