PO-0333 MUSCLE THICKNESS ULTRASONOGRAPHY IN CRITICALLY ILL CHILDREN: ANALYSIS OF ACCURACY

Background and aims Critical illness results in muscle wasting, typically within the first week of admission. However, its quantification is difficult. Ultrasonographical evaluation has already been tested in adult patients. Here we aim to assess of the accuracy of this methodology in critically ill children. Methods Two independent investigators made ultrasonographical assessments of muscle thickness. A linear array commercial real time ultrasound scanner (Vivid S6) was used with a 12-MHz transducer. The transducer was placed perpendicularly to the long axis of the tight on three fifths of the distance from the anterior superior iliac spine to the superior patellar border. An excess of contact gel was applied to minimise image distortion. During the first week of critical illness we hypothesised a reduction in muscle size of 30% based on literature of critically ill adult patients. Results A group of 43 patients (newborns to 4-years old children) were included in the study. The average muscle thickness was 1.67 cm (SD: 0.52 cm). Hence 30% reduction would equal 0.50 cm. The intra-observer variability, as expressed by the limits of agreement (± 1.96 SD, containing 95% of the samples for normally distributed samples), was 0.61 cm for operator 1 and 0.83 cm for operator 2. Both distributions are visualised in a Bland-Altman plot in Figure 1. Combining both operator data revealed an observer variability of 0.72 cm. The median absolute inter-observer variability was 0.085 cm [IQR 0.040–0.20 cm]. Conclusions Although the inter-observer variability among the two operators is acceptable, the intra-observer variability may be too large with respect to the limited expected muscle reduction. Therefore, ultrasonographical analyses of muscle wasting need more standardisation for use in critically ill children.