Validation of Inter-Rater and Intra-Rater Reliability of Remote Dielectric Sensing Measurement
Masakazu Hori, Teruhiko Imamura, Atsuko Fukuo, Takuya Fukui, Takatoshi Koi, Yohei Ueno, Hiroshi Onoda, Shuhei Tanaka, Ryuichi Ushijima, Mitsuo Sobajima, Nobuyuki Fukuda, Hiroshi Ueno, Koichiro Kinugawa
Abstract
Remote dielectric sensing (ReDS) is a recently introduced non-invasive electromagnetic-based device used to quantify lung fluid levels. Nevertheless, its inter-rater and intra-rater reliability remain uncertain. In 10 healthy volunteers, ReDS values were measured three times successively by the officially trained expert examiner to validate intra-rater reliability. Similar measures were performed by a total of three examiners to validate inter-rater reliability. Intra-class correlation (ICC) was applied to validate each reliability. Ten healthy volunteers [median 34 (32, 40) years old, 10 men, body mass index 23.0 (21.2, 23.9) ] were included. Median ReDS value was 28% (25%, 31%). For the intra-rater reliability, ICC (1, 1) and ICC (1, 3) were 0.966 and 0.988, respectively (P < 0.001). For the inter-rater reliability, ICC (2, 1) and ICC (2, 3) were 0.683 and 0.866, respectively (P < 0.001). Given almost perfect intra-rater reliability, an examiner does not need to repeat ReDS measurement. Given substantial inter-rater reliability, ReDS measurements had better be measured by multiple examiners if possible.