Abstract

Early diagnosis of muscle wasting in critically ill patients with acute kidney injury requiring continuous kidney replacement therapy (AKI-CKRT) may improve outcomes via timely rehabilitation and nutrition. Muscle ultrasound (MUS) has recently gained traction for assessing muscle atrophy in the intensive care unit (ICU) but requires training to achieve reproducibility. We evaluated the inter-rater reliability of MUS in patients with AKI-CKRT performed by multidisciplinary raters including nephrologists. Two blinded independent raters used portable ultrasound to acquire images of the rectus femoris (RF). All raters were clinicians routinely caring for patients with CKRT in the ICU and were initially novices in MUS. They underwent three two-hour teleconference training sessions in MUS led by an experienced physiotherapist. Inter-rater reliability was evaluated with intraclass correlation coefficients (ICCs) [95% confidence interval] using a two-way random-effects model. We analyzed 54 MUS images (27 pairs) from nine patients at baseline (n = 16), day 3 (n = 6), day 7 (n = 8), ICU discharge (n = 10), hospital discharge (n = 10), and 1-3 months after discharge (n = 4). The mean (+/- standard deviation) values of RF thickness, cross-sectional area, and echointensity were 1.7 +/- 1.4 cm, 4.6 +/- 2.7 cm(2), and 84.0 +/- 17.7 AU, respectively. Reliability was excellent for RF thickness (ICC = 0.96 [0.91-0.98], p < 0.001) and cross-sectional area (ICC = 0.92 [0.83-0.96], p < 0.001) but poor for echointensity (ICC = 0.41 [0.04-0.68], p < 0.05). These results demonstrate reliable assessment of muscle size in patients with AKI-CKRT using ultrasound performed by multidisciplinary novice sonographers trained via teleconference, suggesting that this methodology may be useful in future studies of muscle wasting in patients with AKI-CKRT.