Meta-Analysis
Copyright ©The Author(s) 2020.
World J Clin Cases. Nov 26, 2020; 8(22): 5589-5602
Published online Nov 26, 2020. doi: 10.12998/wjcc.v8.i22.5589
Table 1 Characteristics of included studies
Ref.
Publication year
Country
Study design
Enrollment method
US system
Number of patients (male/female)
Age in yr, mean ± SD
Number of FLLs (benign/malignant)
LI-RADS version
Reference standard
Chen et al[25]2017ChinaRetrospectiveConsecutiveHitachi HI Vision Preirus; Siemens S3000377 (207/170)43.56 11.48429 (96/333)2016Pathology and SCRS
Schellhaas et al[20]2018GermanyRetrospectiveConsecutiveSiemens S2000; GE Logiq E9; Toshiba Aplio 50055 (44/11)65.90 (53-86)155 (5/50)2016Pathology and SCRS
Ling et al[28]2018ChinaRetrospectiveConsecutivePhilips IU2256 (44/12)52.5256 (10/46)2016Pathology
Liu et al[26]2019ChinaProspectiveConsecutivePhilips IU22; Toshiba Aplio 50082 (61/21)58.15 10.9782 (34/48)2017Pathology and SCRS
Zhou et al[27]2019ChinaRetrospectiveConsecutiveGE Logiq E9241 (166/75)56 10272 (44/228)2016Pathology and SCRS
Tan et al[21]2020SingaporeRetrospectiveSelectiveGE Logiq E9; Toshiba Aplio 50045 (32/13)63.1 (34-84)46 (9/37)2017Pathology and SCRS
Wang et al[6]2020ChinaRetrospectiveConsecutiveHitachi Hi Vision Avius258 (200/58)52 11355 (235/120)2017Pathology and SCRS
Cui et al32020ChinaRetrospectiveSelectiveHitachi Hi Vision Avius63 (47/16)56 884 (35/49)2017Pathology and SCRS
Table 2 Analysis of interobserver agreement for included studies
Ref.
Kappa value (95%CI)
Weight in fixed model, %
Weight in random model, %
Chen et al[25]0.81 (0.77, 0.84)31.4014.00
Schellhaas et al[20]0.39 (0.14, 0.59)3.8011.40
Ling et al[28]0.69 (0.52, 0.81)3.9011.40
Liu et al[26]0.61 (0.45, 0.73)5.8012.30
Zhou et al[27]0.87 (0.84, 0.90)19.9013.80
Tan et al[21]0.94 (0.89, 0.97)3.2010.90
Wang et al[6]0.77 (0.72, 0.81)26.0013.90
Cui et al10.67 (0.53, 0.77)6.0012.30
Table 3 Interobserver agreement of pooled included studies
Summary type
Kappa value (95%CI)
Z
P value
Fixed effects model0.79 (0.77, 0.81)39.68< 0.01
Random effects model0.76 (0.67, 0.83)9.99< 0.01
Table 4 Sensitivity analysis eliminating studies one by one
Ref.
Kappa value (95%CI)
I2 (95%CI), %
P value
Chen et al[25]0.75 (0.62; 0.84)92.42 (87.07; 95.57)< 0.01
Schellhaas et al[20]0.79 (0.71; 0.85)89.44 (80.83; 94.22)< 0.01
Ling et al[28]0.77 (0.67; 0.84)92.30 (86.72; 95.58)< 0.01
Liu et al[26]0.78 (0.69; 0.85)91.44 (84.80; 95.13)< 0.01
Zhou et al[27]0.74 (0.63; 0.82)89.78 (81.61; 94.36)< 0.01
Tan et al[21]0.73 (0.62; 0.80)90.21 (82.51; 94.60)< 0.01
Wang et al[6]0.76 (0.64; 0.85)92.41 (86.90; 95.59]< 0.01
Cui et al10.77 (0.67; 0.85)92.04 (86.02; 95.43)< 0.01
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 design0.26
Retrospective70.78 (0.69, 0.85)
Prospective10.61 (0.45, 0.73)
Enrollment method< 0.01
Consecutive70.73 (0.62, 0.80)
Selective10.94 (0.89, 0.97)
LI-RADS version0.66
201640.74 (0.58, 0.85)
201740.78 (0.62, 0.88)
Number of interobservers< 0.01
270.79 (0.71, 0.85)
310.39 (0.14, 0.59)
Number of US systems0.93
140.77 (0.65, 0.85)
> 140.76 (0.51, 0.89)
Race of patients< 0.01
Westerners10.79 (0.71, 0.85)
Easterners70.39 (0.14, 0.59)
Number of FLLs0.16
< 10050.72 (0.48, 0.86)
> 10030.82 (0.76, 0.87)
Reference standard0.58
Pathology10.69 (0.52, 0.81)
Pathology and SCRS70.77 (0.67, 0.84)