Du Y, Bara M, Katlariwala P, Croutze R, Resch K, Porter J, Sam M, Wilson MP, Low G. Effect of training on resident inter-reader agreement with American College of Radiology Thyroid Imaging Reporting and Data System. World J Radiol 2022; 14(1): 19-29 [PMID: 35126875 DOI: 10.4329/wjr.v14.i1.19]
Corresponding Author of This Article
Yang Du, BSc, FRCPC, MD, Doctor, Staff Physician, Department of Radiology and Diagnostic Imaging, University of Alberta, 2A2.41 WMC, 8440-112 St NW, Edmonton T6G 2B7, Alberta, Canada. yang.du@usask.ca
Research Domain of This Article
Radiology, Nuclear Medicine & Medical Imaging
Article-Type of This Article
Retrospective Cohort Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
World J Radiol. Jan 28, 2022; 14(1): 19-29 Published online Jan 28, 2022. doi: 10.4329/wjr.v14.i1.19
Table 1 Pooled inter-reader agreement with the reference standard
Pre-training, k
Post-training, k
P value of the difference
Composition
0.46 (95%CI: 0.37 to 0.54), moderate
0.52 (95%CI: 0.44 to 0.61), moderate
0.32
Echogenicity
0.36 (95%CI: 0.29 to 0.44), fair
0.44 (95%CI: 0.37 to 0.52), moderate
0.30
Shape
0.09 (95%CI: 0.02 to 0.21), slight
0.67 (95%CI: 0.56 to 0.78), substantial
< 0.001
Margins
0.03 (95%CI: -0.14 to 0.08), slight
0.05 (95%CI: -0.05 to 0.15), slight
0.71
Echogenic Foci
0.28 (95%CI: 0.19 to 0.37), fair
0.45 (95%CI: 0.36 to 0.53), moderate
0.004
TI-RADS Level
0.14 (95%CI: 0.08 to 0.20), slight
0.36 (95%CI: 0.30 to 0.42), fair
< 0.001
Recommendations
0.36 (95%CI: 0.27 to 0.45), fair
0.50 (95%CI: 0.41 to 0.59), moderate
0.02
Table 2 Percentage reader agreement with the reference standard for sonographic features
Sonographic feature
RS
R1 pre
R1 post
R2 pre
R2 post
R3 pre
R3 post
Composition
n
n (%)
Spongiform
4
0 (0)
1 (25)
1 (25)
1 (25)
3 (75)
4 (100)
Cystic or almost completely cystic
11
3 (27.3)
5 (45.5)
7 (63.6)
8 (72.7)
10(90.9)
10(90.9)
Mixed cystic and solid
12
9 (75)
6 (50)
5 (41.7)
7 (58.3)
5 (58.3)
6 (50)
Solid
27
26 (96.3)
26 (96.3)
25 (92.6)
26 (96.3)
18 (66.7)
19 (70.4)
Echogenicity
Anechoic
11
3 (27.3)
5 (45.5)
5 (45.5)
5 (45.5)
9 (81.8)
8 (72.7)
Hyperechoic or isoechoic
27
23 (85.2)
23 (85.2)
19 (70.4)
21 (77.8)
19 (70.4)
20 (74.1)
Hypoechoic
12
2 (16.7)
4 (33.3)
9 (75)
8 (66.7)
4 (33.3)
4 (33.3)
Shape
Wilder than tall
42
38 (90.5)
39 (92.9)
7 (16.7)
39 (92.9)
41 (97.6)
40 (95.2)
Taller than wide
8
7 (87.5)
7 (87.5)
7 (87.5)
7 (87.5)
6 (75)
4 (50)
Margins
Smooth or ill defined
47
36 (76.6)
35 (74.5)
35 (74.5)
33 (70.2)
43 (91.5)
45 (95.7)
Lobulated or irregular
3
1 (33.3)
2 (66.7)
1 (33.3)
2 (66.7)
0 (0)
0 (0)
Echogenic foci
None or large comet tail artifact
41
20 (48.8)
36 (87.8)
29 (70.7)
39 (95.1)
29 (70.7)
29 (70.7)
Macrocalcification
3
1 (33.3)
1 (33.3)
0 (0)
2 (66.7)
2 (66.7)
2 (66.7)
Punctate echogenic foci
6
5 (83.3)
4 (66.7)
2 (33.3)
5 (83.3)
3 (50)
3 (50)
Table 3 Percentage reader agreement with the reference standard for American College of Radiology Thyroid Imaging Reporting and Data System levels
ACR TI-RADS level
RS, n
R1 pre, n (%)
R1 post, n (%)
R2 pre, n (%)
R2 post, n (%)
R3 pre, n (%)
R3 post, n (%)
1
11
1 (9.1)
5 (45.5)
1 (9.1)
7 (63.6)
10 (90.9)
8 (72.7)
2
9
3 (33.3)
4 (44.4)
0 (0)
4 (44.4)
3 (33.3)
3 (33.3)
3
9
4 (44.4)
5 (55.5)
1 (11.1)
6 (66.7)
4 (44.4)
6 (66.7)
4
13
4 (30.8)
5 (38.5)
5 (38.5)
9 (69.2)
5 (38.5)
5 (38.5)
5
8
7 (87.5)
4 (50)
6 (75)
5 (62.5)
3 (37.5)
3 (37.5)
Table 4 Percentage reader agreement with the reference standard for American College of Radiology Thyroid Imaging Reporting and Data System recommendations
Recommendations
RS, n
R1 pre, n (%)
R1 post, n (%)
R2 pre, n (%)
R2 post, n (%)
R3 pre, n (%)
R3 post, n (%)
No follow up
25
13 (52)
17 (68)
10 (40)
19 (76)
21 (84)
22 (88)
Follow up
5
3 (60)
1 (20)
1 (20)
3 (60)
3 (60)
3 (60)
FNA
20
17 (85)
15 (75)
18 (90)
17 (85)
11 (55)
13 (65)
Table 5 The relative sensitivity, specificity, positive predictive value, and negative predictive value per Thyroid Imaging Reporting and Data System Level on the pre-training assessment compared to the reference standard
Pre-training, Statistics
TI-RADS 1, %
TI-RADS 2, %
TI-RADS 3, %
TI-RADS 4, %
TI-RADS 5, %
Sensitivity
R1
9.1 (0.2-41.3)
33.3 (7.5-70.1)
44.4 (13.7-78.8)
30.8 (9.1-61.4)
87.5 (47.4-99.7)
R2
9.1 (0.2-41.3)
0 (0-33.6)
11.1 (0.3-48.3)
38.5 (13.9-68.4)
75 (34.9-96.8)
R3
90.9 (58.7-99.8)
33.3 (7.5-70.1)
44.4 (13.7-78.8)
38.5 (13.9-68.4)
37.5 (8.5-75.5)
Pooled
36.4 (20.4-54.9)
22.2 (8.6-42.3)
33.3 (16.5-54)
35.9 (21.2-52.8)
66.7 (44.7-84.4)
Specificity
R1
100 (91.0-100)
90.2 (76.9-97.3)
92.7 (80.1-98.5)
62.2 (44.8-77.5)
76.2 (60.6-88)
R2
100 (91-100)
97.6 (87.1-99.9)
80.5 (65.1-91.2)
81.1 (64.8-92)
50 (34.2-65.8)
R3
66.7 (49.8-80.9)
97.6 (87.1-99.9)
95.1 (83.5-99.4)
89.2 (74.6-97)
90.5 (77.4-97.3)
Pooled
88.9 (81.8-94)
95.1 (89.7-98.2)
89.4 (82.6-94.3)
76.6 (67.6-84.1)
72.2 (63.5-79.8)
Positive predictive value
R1
100
42.9 (16.8-73.6)
57.1 (26.4-83.2)
22.2 (10.3-41.6)
41.2 (27.7-56.1)
R2
100
0
11.1 (1.8-46.8)
41.7 (21.5-65.1)
22.2 (14.8-32.1)
R3
43.5 (32.2-55.5)
75 (26-96.2)
66.7 (30.1-90.3)
55.6 (28.3-79.8)
42.9 (17.1-73.2)
Pooled
48 (31.8-64.6)
50 (25.9-74.1)
40.9 (24.8-59.2)
35 (23.9-48)
31.4 (23.5-40.5)
Negative predictive value
R1
79.6 (76.4-82.5)
86.1 (79.4-90.8)
88.4 (80.8-93.2)
71.9 (62.2-79.9)
97 (83.5-99.5)
R2
79.6 (76.4-82.5)
81.6 (80.9-82.4)
80.5 (75.8-84.5)
79 (70.4-85.6)
91.3 (75.3-97.3)
R3
96.3 (79.8-99.4)
87 (80.7-91.4)
88.6 (81.2-93.4)
80.5 (72.6-86.5)
88.4 (81.5-92.9)
Pooled
83.2 (79.2-86.6)
84.8 (81.9-87.3)
85.9 (82.3-88.9)
77.3 (72.5-81.5)
91.9 (86.5-95.3)
Table 6 The relative sensitivity, specificity, positive predictive value, and negative predictive value per Thyroid Imaging Reporting and Data System Level on the post-training assessment compared to the reference standard
Post-training, Statistics
TI-RADS 1, %
TI-RADS 2, %
TI-RADS 3, %
TI-RADS 4, %
TI-RADS 5, %
Sensitivity
R1
45.5 (16.8-76.6)
44.4 (13.7-78.8)
55.6 (21.2-86.3)
38.5 (13.9-68.4)
50 (15.7-84.3)
R2
63.6 (30.8-89.1)
44.4 (13.7-78.8)
66.7 (29.9-92.5)
69.2 (38.6-90.9)
62.5 (24.5-91.5)
R3
72.7 (39-94)
33.3 (7.5-70.1)
66.7 (29.9-92.5)
38.5 (13.9-68.4)
37.5 (8.5-75.5)
Pooled
60.6 (42.1-77.1)
40.7 (22.4-61.2)
63 (42.4-80.6)
48.7 (32.4-65.2)
50 (29.1-70.9)
Specificity
R1
92.3 (79.1-98.4)
97.6 (87.1-99.9)
90.2 (76.9-97.3)
70.3 (53-84.1)
81 (65.9-91.4)
R2
94.9 (82.7-99.4)
97.6 (87.1-99.9)
95.1 (83.5-99.4)
73 (38.6-90.9)
90.5 (77.4-97.3)
R3
66.7 (49.8-80.9)
95.1 (83.5-99.4)
97.6 (87.1-99.9)
86.5 (71.2-95.5)
90.5 (77.4-97.3)
Pooled
84.6 (76.8-90.6)
96.8 (91.9-99.1)
94.3 (88.6-97.7)
76.6 (67.6-84.1)
87.3 (80.2-92.6)
Positive predictive value
R1
62.5 (32-85.5)
80 (33.6-96.9)
55.6 (29.4-79)
31.3 (16.3-51.5)
33.3 (16.5-56)
R2
77.8 (45.8-93.6)
80 (33.6-96.9)
75 (41.8-92.6)
47.4 (32.2-63.1)
55.6 (29.9-78.6)
R3
38.1 (25.8-52.2)
60 (22.6-88.5)
85.7 (45.1-97.8)
50 (25.6-74.4)
42.9 (17.1-73.2)
Pooled
52.6 (40.1-64.8)
73.3 (48.6-88.9)
70.8 (52.8-84.1)
42.2 (31.5-53.8)
42.9 (29-57.9)
Negative predictive value
R1
85.7 (77.6-91.2)
88.9 (81.7-93.5)
90.2 (81.6-95.1)
76.5 (66.8-84)
89.5 (80.7-94.5)
R2
90.2 (80.8-95.3)
88.9 (81.7-93.5)
92.9 (83.7-97)
87.1 (74.5-94)
92.7 (83.7-96.9)
R3
89.7 (76.3-95.9)
86.7 (80.3-91.2)
93 (84.1-97.1)
80 (71.9-86.2)
88.4 (81.5-92.9)
Pooled
88.4 (83.2-92.1)
88.2 (84.5-91.1)
92.1 (87.6-95)
81 (75.5-85.4)
90.2 (85.9-93.2)
Citation: Du Y, Bara M, Katlariwala P, Croutze R, Resch K, Porter J, Sam M, Wilson MP, Low G. Effect of training on resident inter-reader agreement with American College of Radiology Thyroid Imaging Reporting and Data System. World J Radiol 2022; 14(1): 19-29