Skip to main content
. Author manuscript; available in PMC: 2011 Jan 1.
Published in final edited form as: Clin Cancer Res. 2009 Dec 22;16(1):195–202. doi: 10.1158/1078-0432.CCR-09-0948

Table 3.

A. Predictability of GEMCaP score with clinical outcome (n = 54)

Threshold method

Fixed Floating Integrated
Sensitivity 43% 80% 67%
Specificity 75% 50% 63%
Positive predictive value 68% 67% 60%
Negative predictive value 51% 67% 69%
Accuracy 57% 67% 65%
B. Frequency of identifying cases based on lymph node status

No. of unfavorable cases identified

Lymph node negative (n = 15) Lymph node positive (n = 15)

Nomogram 2 10
GEMCaP:
 Fixed 6 7
 Floating 12 12
 Integrated 10 10

NOTE: Sensitivity is the proportion of cases who are correctly classified based on the GEMCaP score (≥20%) as being at high risk of recurrence; specificity is the proportion of controls who are correctly classified as being at low risk of recurrence. The positive predictive value is the proportion of patients classified by the GEMCaP score to be at high risk of recurrence who are cases, and the negative predictive value is the proportion of patients classified to be at low risk of recurrence who are controls. The limitations of these last two estimates for a case-control study are known, but they are included only as a reference for future studies. Accuracy is the overall proportion of correctly classified cases and controls. Cases were identified by values >20% for GEMCaP and <40% for the nomogram.