Skip to main content
. Author manuscript; available in PMC: 2024 Jan 1.
Published in final edited form as: J Clin Gastroenterol. 2023 Jan 1;57(1):103–110. doi: 10.1097/MCG.0000000000001710

Table 4.

Percent of patients whose risk was among the top 20%, 15%, 10%, 5% and 2.5%, sensitivity, specificity, positive predicted Value (PPV), and risk fold increase for each of the three models.

Model 1 Model 2 Model 3
High risk patients High risk patients High risk patients
Top 20% Top 15% Top 10% Top 5% Top 2.5% Top 20% Top 15% Top 10% Top 5% Top 2.5% Top 20% Top 15% Top 10% Top 5% Top 2.5%
No. of eligible patients whose risk was above each risk threshold 2278 1708 1139 569 285 2267 1701 1134 567 284 2278 1709 1139 570 285
Sensitivity (%) 60.0 52.0 42.0 27.3 17.3 56.2 48.7 37.1 25.3 16.0 60.0 52.7 39.0 24.2 14.7
Specificity (%) 80.3 85.3 90.3 95.2 97.6 80.3 85.3 90.2 95.2 97.6 80.3 85.3 90.3 95.2 97.6
PPV (%) 2.5 2.9 3.5 4.5 5.7 2.3 2.6 3.0 4.1 5.2 2.6 3.0 3.4 4.2 5.1
Fold increase in risk* 3.0 3.5 4.2 5.5 6.9 2.8 3.2 3.7 5.1 6.4 3.0 3.5 3.9 4.8 5.9
*

Compared the incidence rate in the entire cohort.

Model 1: Age, weight change, HbA1c change, HgA1c change rate

Model 2: Age, weight change, HbA1c change, HgA1c in prior 6 months

Model 3: Age, weight change, HbA1c change, HgA1c in prior 18 months

All reported numbers were average across 20, 8 and 6 validation datasets for Models 1, 2 and 3, respectively, estimated in patients with complete 36 months follow up or those who developed PDAC in 36 months (n=11,379).