Research focusing on the reproducibility of MSU

Research focusing on the reproducibility of MSU measurement of CET is limited, with varying fair to excellent intra-class correlation coefficient (ICC) outcomes (Lee et al. 2011; Miller et al. 2002; Poltawski et al. 2012). Furthermore, intra-rater reliability and the smallest detectable change (SDC) were not determined (Lee et al. 2011), ordinal scales were used (Miller et al. 2002) or an ICC value for ordinal data instead of the recommended weighted κ value was applied (Poltawski et al. 2012, Fleiss and Cohen, 1973). Consequently, more insight into all reproducibility aspects of MSU thickness measurement of the CET, using objective quantitative measurements, is needed.
Earlier research on the reliability of MSU in other structures has indicated that p2y inhibitor MSU seems to be a rater-dependent technique (O\’Connor et al. 2005; Rutten et al. 2006). This variation seems to be specifically dependent on the level of experience and standardization of techniques (O\’Connor et al. 2005; Rutten et al. 2006). Therefore, the use of a standardized measurement protocol seems very important. However, all the studies performed on reproducibility of CET thickness measurements use various protocols. The protocols contain insufficient description of important standardization aspects that can negatively influence reproducibility. The protocols use subjective thickness measurements, without clear agreement on terminology and practice of the raters, lack measurable reference points for exact location of thickness measurements and/or fail to describe the positioning of the patients, including positioning of the elbow and accompanying joints (Lee et al. 2011; Miller et al. 2002; Poltawski et al. 2012). More insight into the reproducibility of CET thickness measurement, addressing these aspects, is needed.


Seventy-three healthy individuals participated in the study (44% females) with a mean (± SD) age of 35.7 (14.9) y and a mean (± SD) BMI of 23.9 (3.6) kg/m2. Sixty-four participants were right hand dominant. The descriptive data for longitudinal and transverse thickness of the CET, for both raters separately, are summarized in Table 1. There were no missing values, and all data were normally distributed.
Inter-rater reliability for both the longitudinal and transverse planes was fair to good (ICCs of 0.67 and 0.49, respectively). ICC values for intra-rater reliability for both raters as well as for longitudinal and transverse planes were excellent (ICCs = 0.85–0.92), with the exception of the ICC for rater 2 in the transverse plane, which was fair to good (ICC = 0.73). All ICC values were statistically significant with p-values < 0.001. The SDCs for both raters, as well as for longitudinal and transverse planes, ranged from 0.50 to 0.78 mm and comprised 9.8%–16.3% of the mean thickness. The ICC values for inter- and intra-rater reliability, the SDC values for intra-rater agreement and the intra-rater agreement with 95% limits of agreement are listed in Table 2.
Jaen-Diaz et al. (2010), Toprak et al. (2012) and Ustuner et al. (2013) determined reference values of 4.02–5.30 mm, 4.57 mm (SD ± 0.63) and 4.60 mm (SD ± 0.65), respectively, for CET thickness of the dominant arm in healthy people, which are comparable to our findings.
This is the first study that investigated all components of reproducibility of the MSU thickness measurements of CET in a large group of participants in the longitudinal and transverse planes. The results of the present study are comparable to the results of Krogh et al. (2013), who reported that intra-rater reliability varied between 0.76 and 0.81 and intra-reliability between 0.45 and 0.65. One study determined the inter-rater reliability of objective MSU thickness measurements of the CET in transverse plane only and found excellent inter-rater reliability for MSU measurements in healthy individuals and patients with a clinical diagnosis of LET (ICC = 0.86 and 0.75, respectively). The ICC value for the transverse thickness measurement from the present study was lower (ICC = 0.49). The thickness measurement in the longitudinal plane was better specified by means of bony landmarks, instead of the less exact location for measurement in transverse plane. This could explain the lower ICC value for inter-rater reliability for the transverse measurements compared with the results of Lee et al. (2011) and with the ICC values of the longitudinal measurements in the present study. Another explanation could be that the raters had less experience in measuring the CET thickness in the transverse plane because they use the longitudinal measurement more frequently in daily practice. Even though both measurements exhibited sufficient levels of reliability because of the excellent and fair outcomes, use of the longitudinal thickness measurement rather than the transverse measurement is recommended in daily practice.