Skip to main content
. 2018 Jan 2;28(6):2380–2388. doi: 10.1007/s00330-017-5230-0

Table 2.

Subgroup analysis

Subgroup No. of studies Pooled sensitivity (95 % CI) Pooled specificity (95 % CI) Pooled DOR Meta-regression RDOR Meta-regression (p-value)
Total 6 0.983 (0.976–0.989) 0.552 (0.542–0.562) 51.020
Publication year < 2016 2 0.976 (0.951–0.990) 0.883 (0.872–0.893) 75.907
≥ 2016 4 0.985 (0.977–0.990) 0.308 (0.296–0.321) 31.260 0.71 0.2701
No. of nodules < 1000 4 0.983 (0.958–0.995) 0.572 (0.535–0.608) 47.350
≥ 1000 2 0.983 (0.976–0.989) 0.550 (0.539–0.561) 53.742 0.98 0.9301
Korea vs. other Korea 3 0.985 (0.977–0.991) 0.261 (0.249–0.273) 22.237
Other 3 0.978 (0.958–0.990) 0.878 (0.868–0.887) 156.13 2.14 0.0037
Reference standard Cytology, histology 4 0.971 (0.961–0.980) 0.392 (0.379–0.405) 49.665
Cytology, histology and US follow-up 2 0.979 (0.959–0.991) 0.874 (0.863–0.884) 118.31 0.36 0.4241
Cut–off 3/4a 2 0.979 (0.959–0.991) 0.874 (0.863–0.884) 118.31
4/4b 4 0.971 (0.9610.980) 0.392 (0.379–0.405) 49.665 0.36 0.4241
Study design Prospective 2 0.972 (0.944–0.989) 0.911 (0.902–0.920) 361.91
Retrospective 4 0.985 (0.978–0.991) 0.263 (0.251–0.275) 23.023 78.61 <0.0001

CI confidence interval, DOR diagnostic odds ratio, RDOR relative diagnostic odds ratio