Predictive models of disease burden at diagnosis in persons with adult-onset ulcerative colitis using health administrative data

Table 2 Performance of Logistic Regression Models

MODELS	Optimism-Corrected C-Statistic Values^a	Unadjusted C-Statistic Values
Colitis Extent
Model 1: Extensive Colitis (vs. Left-Sided Colitis/Proctitis)	0.695 (0.669,0.754)	0.710
Model 2: Extensive OR Left-Sided Colitis (vs. Proctitis)	0.716 (0.687,0.781)	0.737
Colitis Activity
Model 3: Severe Colitis (vs. Mild-Moderate Colitis)	0.670 (0.637,0.733)	0.665
Model 4: Severe OR Moderate Colitis (vs. Mild Colitis)	0.663 (0.628,0.737)	0.675
Total Disease Burden
Model 5: Severe AND Extensive Colitis (vs. all other categories)	0.729 (0.690,0.796)	0.734
Model 6: High Colitis Burden (vs. all other categories) ^b	0.729 (0.702,0.789)	0.735

^aC-statistic estimates and confidence intervals adjusted following bootstrap internal validation of regression models (1000 iterations)
^b[severe AND extensive] OR [severe AND left-sided] OR [moderate AND extensive] vs. [severe AND proctitis] OR [moderate AND left-sided] OR [mild AND extensive] OR [moderate AND proctitis] OR [mild AND left-sided] OR [mild AND proctitis], based on the endoscopic mayo subscore of disease activity and Montreal classification of disease extent

ISSN: 1471-230X