Skip to main content
. 2018 Mar;10(3):1522–1531. doi: 10.21037/jtd.2018.02.58

Table 3. Relationship between PLR and the clinicopathologic features.

Variable No. of studies No. of patients Effect of model OR (95% CI) P value Heterogeneity Egger’s P value
I2 (%) Ph
Tumor invasion (T3/4 vs. T1/2) 6 1,848 Fixed 1.543 (1.269, 1.876) 0.000 0.0 0.629 0.022
Random 1.543 (1.269, 1.876) 0.000
Lymph node metastasis (yes vs. no) 7 2,165 Fixed 1.427 (1.195, 1.705) 0.000 17.9 0.294 0.341
Random 1.461 (1.175, 1.816) 0.001
Differentiation (poor vs. well/moderate) 6 2,128 Fixed 1.196 (0.978, 1.462) 0.081 0.0 0.477 0.677
Random 1.196 (0.978, 1.462) 0.081
Vascular invasion(yes vs. no) 3 567 Fixed 1.104 (0.709, 1.717) 0.663 0.0 0.961 0.749
Random 1.104 (0.709, 1.717) 0.663
TNM stage(stage III/IV vs. I/II)) 7 2,612 Fixed 1.459 (1.235, 1.724) 0.000 29.6 0.202 0.345
Random 1.504 (1.198, 1.889)
Tumor length (>3 vs. ≤3 cm) 4 1,061 Fixed 1.810 (1.331, 2.461) 0.000 12.3 0.331 0.607
Random 1.774 (1.255, 2.507) 0.001

Ph, P value for heterogeneity; No., number; OR, odds ratio; CI, confidence interval; TNM, tumor node metastasis; T, depth of tumor invasion.