Table 5.

Results of meta-regression analysis of interobserver agreement of Liver Imaging Reporting and Data System in contrast-enhanced ultrasound

Covariate and subgroup	Number of included studies	Kappa value (95%CI)	P value
Study design			0.26
Retrospective	7	0.78 (0.69, 0.85)
Prospective	1	0.61 (0.45, 0.73)
Enrollment method			< 0.01
Consecutive	7	0.73 (0.62, 0.80)
Selective	1	0.94 (0.89, 0.97)
LI-RADS version			0.66
2016	4	0.74 (0.58, 0.85)
2017	4	0.78 (0.62, 0.88)
Number of interobservers			< 0.01
2	7	0.79 (0.71, 0.85)
3	1	0.39 (0.14, 0.59)
Number of US systems			0.93
1	4	0.77 (0.65, 0.85)
> 1	4	0.76 (0.51, 0.89)
Race of patients			< 0.01
Westerners	1	0.79 (0.71, 0.85)
Easterners	7	0.39 (0.14, 0.59)
Number of FLLs			0.16
< 100	5	0.72 (0.48, 0.86)
> 100	3	0.82 (0.76, 0.87)
Reference standard			0.58
Pathology	1	0.69 (0.52, 0.81)
Pathology and SCRS	7	0.77 (0.67, 0.84)

FLL: Focal hepatic lesion; LI-RADS: Liver Imaging Reporting and Data System; SCRS: Synthesized clinical reference standard; US: Ultrasound.