Table 4.
Univariate and multivariate analysis to predict pathological subtype using optimal cut-off values.
| Variable | Cut-off value | Univariate | Multivariate | ||||
|---|---|---|---|---|---|---|---|
| OR | 95% CI | P value | OR | 95% CI | P value | ||
| Age | >57 years | 6.135 | 2.421–15.545 | <0.001∗ | 1.307 | 0.209–8.175 | 0.774 |
| Radiological findings | >1 pure solid nodule | 35.217 | 4.420–280.601 | 0.001∗ | 3.304 | 0.317–34.440 | 0.318 |
| Serum CEA level | >2.12 ng/mL | 4.822 | 1.909–12.181 | 0.001∗ | 0.957 | 0.158–5.788 | 0.961 |
| Serum miR-183 level | >1.233 (2−ΔΔCt) | 2.900 | 1.183–7.112 | 0.020∗ | 0.428 | 0.046–4.006 | 0.457 |
| Tumour size | >11 mm | 60.167 | 14.822–244.236 | <0.001∗ | 3.106 | 0.253–38.068 | 0.375 |
| Combined five factors | >0.499 | 159.900 | 29.287–873.024 | <0.001∗ | 60.033 | 2.462–1463.959 | 0.012∗ |
∗Statistically significant p value
OR, odds ratio; CI, confidence interval; mGGO, mixed ground-glass opacity; CEA, carcinoembryonic antigen.