Twenty-four studies (n=479) were included in the review. Only six studies were judged to be of high quality. Studies generally scored poorly on inclusion of a representative sample of patients, reporting a sufficiently large estimate of intra-examiner reliability and use of appropriate measures for calculating reliability and precision. It was generally unclear whether a representative sample of examiners was used.
Studies that reported intra-rater reliability using intra-class correlation (ICC) coefficients reported values ranged from 0.48 to 1.0 (number of studies unclear). Two studies reported ICC coefficients for inter-rater reliability on different images of 0.91 and 1.0. All ICC estimates of intra-rater and inter-rater reliability for the repeated measurement of the same rehabilitative ultrasound images were greater than 0.90.
When analysis was restricted to the four high-quality studies that assessed intra-rater reliability, ICC coefficients ranged from 0.62 to 0.97. One high-quality study assessed inter-rater reliability and reported good reliability and ICC estimates that ranged from 0.91 to 1.00.