Table 3.

Model performance measures with the addition of continuous hs-cTnT or NT-proBNP to Model 4 variables.

	C-statistic	NRI, categorical**	NRI, continuous	IDI
Model 4	0.884 (0.858 to 0.909)
hs-cTnT	0.892‡ (0.869 to 0.916)	0.001 (−0.037 to 0.040)	0.195† (0.067 to 0.322)	0.010† (0.003 to 0.018)
NT-proBNP	0.888‡ (0.864 to 0.913)	0.005 (−0.024 to 0.034)	0.148‡ (0.020 to 0.276)	0.004 (−0.001 to 0.009)
hs-cTnT and NT-proBNP	0.894‡ (0.871 to 0.917)	0.005 (−0.036 to 0.046)	0.176† (0.048 to 0.303)	0.012† (0.003 to 0.020)

^*P<0.001

^†P<0.01

^‡P<0.05

^**10-year risks of 10% and 30% were used as thresholds for categorical NRI.

Note: Data are given as value (95% confidence interval). All prediction statistics were based on 10-year predicted risk. Model 4 variables included: age, gender, race, systolic blood pressure, antihypertensive medication, smoking, alcohol intake, level of education, body mass index, total and high-density lipoprotein cholesterols, diabetes, history of cardiovascular disease, estimated glomerular filtration rate and albuminuria (albumin-creatinine ratio) at baseline

Abbreviations: hs-cTnT, high-sensitivity cardiac troponin T; NT-proBNP, N-terminal pro–brain natriuretic peptide; net reclassification improvement (NRI), integrated discrimination improvement (IDI)