Skip to main content
. 2017 Feb 10;12(2):e0172031. doi: 10.1371/journal.pone.0172031

Table 2. Optimal Ki-67 cutoff values for predicting tumor recurrence.

Cutoff value* Sensitivity Specificity Relative risk 95% CI P value
Average method
<17% vs ≥17% 67.86% 60.65% 1.7242 1.3047–2.2786 0.0001
<18% vs ≥18% 67.86% 63.44% 1.8561 1.4005–2.4599 < 0.0001
<19% vs ≥19% 60.71% 65.81% 1.7756 1.2848–2.4539 0.0005
Hot spot method
<22% vs ≥22% 75.00% 60.00% 1.8750 1.4733–2.3862 < 0.0001
<23% vs ≥23% 64.29% 62.37% 1.7082 1.2656–2.3054 0.0005
<24% vs ≥24% 64.29% 64.52% 1.8117 1.3394–2.4506 0.0001

*Cutoff values were obtained based on ROC curve analyses.

Abbreviation: CI, confidence interval