Table 5. Performance of different models on diabetes risk prediction.

Model	Accuracy (%)	Sensitivity (%)	Specificity (%)	Positive predictivevalue (%)	Negative predictivevalue (%)
CDP	68.4	82.7	67.5	21.5	97.3
CDP + 2 h-Glucose	67.6	90.7	66.2	22.4	98.5
CDP + Biomarker risk score	82.6	70.7	83.3	31.3	96.4

CDP: sex, age, waist circumference, fasting glucose, hypertension, dyslipidaemia, family history of diabetes, physical activity and smoking status.

Biomarker risk score: (as described in table 4).

Prevalence rate 9.7% [1].