Interobserver and Intraobserver Reliability of Hepatic Shear Wave Elastography and the Influence of Fasted Versus Nonfasted States in Healthy Volunteers

Objectives The primary objective of this study was to assess the effect of fasting versus ingestion of food and water on hepatic measurements by shear wave elastography (SWE) in healthy participants. The secondary objective was to assess inter- and intra-reader reliability of hepatic elastography in healthy participants. Methods Twenty healthy participants were enrolled in this prospective study and underwent quantitative SWE under fasting conditions and after the ingestion of water and food and water. Two blinded sonographers each independently performed a total of 6 sessions of hepatic SWE in each participant. Sessions 1 to 3 were performed on day 0 and sessions 4 to 6 on day 7. Statistical tests used included the Wilcoxon signed ranks test, the intraclass correlation coefficient, and Bland-Altman plots. Results There were no significant differences in hepatic SWE measurements after the ingestion of water versus the fasting state. Statistical significance was assessed asP< 0.05. The postprandial status had a statistically significant effect on hepatic SWE measurements at 1 hour (P= .04) but not at 3 hours (P= .08). By the intraclass correlation coefficient, there was poor-to-moderate inter-reader agreement and minimal-to-moderate intra-reader agreement. The median inter-reader difference in SWE measurements ranged from 0.66 to 0.96 kPa. The median intra-reader difference ranged from 0.43 to 0.55 kPa. Conclusions Our study shows that the ingestion of water has no effect on hepatic SWE measurements in healthy participants. The postprandial state had a significant effect on SWE measurements at 1 hour after ingestion but not at 3 hours. The inter-reader and intra-reader agreements were variable and moderate at best.

Automated summary

This study found that fasting versus ingestion of food and water had no significant effect on hepatic SWE measurements in healthy participants, but there was a significant effect at 1 hour postprandial. Both inter- and intra-reader agreements were variable and moderate at best.