Evaluation of risk equations for prediction of short-term coronary heart disease events in patients with long-standing type 2 diabetes: the Translating Research into Action for Diabetes (TRIAD) study

Table 3 Discrimination and calibration of UKPDS and Framingham CHD Risk Equations

		N (Number of CHD events)	Discrimination Harrell’s c-index (95%CI)	Calibration GOF Chi-Square Statistics
				Unadjusted^a	Adjusted^a
UKPDS	Incident	5914 (120)	0.63 (0.58, 0.68)	55.64	2.18
	Duration	5914 (120)	0.64 (0.59, 0.69)	446.73	7.65
Framingham-Initial	Men	2560 (97)	0.61 (0.55, 0.67)	24.28	0.88
	Women	3354 (121)	0.59 (0.54, 0.64)	17.27	5.85
Framingham-Secondary	Men	1317 (209)	0.55 (0.51, 0.59)	19.45	8.52
	Women	1072 (165)	0.54 (0.49, 0.58)	77.59	5.23

^aUnadjusted and adjusted refer to the GOF chi-square statistics before vs. after recalibration, respectively.

ISSN: 1472-6823