. 2021 Aug;148(2):495–505.e14. doi: 10.1016/j.jaci.2021.02.029

Table E5.

Pairwise comparison of ROC analyses in the same sample cohort (n = 100)

Characteristic comparisons	Diagnostic performance (ROC analyses)						Pairwise comparison 8of AUCs (P value)
Characteristic comparisons	AUC	Lower bound (95% CI)	Upper bound (95% CI)	AUC	Lower bound (95% CI)	Upper bound (95% CI)	Pairwise comparison 8of AUCs (P value)
Peanut-sIgE (kU_A/L)-Peanut SA (%)	0.752	0.658	0.845	0.862	0.793	0.932	.0038
Ara h 1 sIgE (kU_A/L)-Ara h 1 SA (%)	0.668	0.559	0.777	0.716	0.613	0.820	.0898
Ara h 2 sIgE (kU_A/L)-Ara h 2 SA (%)	0.916	0.857	0.975	0.933	0.885	0.981	.5415
Ara h 3 sIgE (kU_A/L)-Ara h 3 SA (%)	0.672	0.563	0.780	0.738	0.639	0.837	.0664
Ara h 6 sIgE (kU_A/L)-Ara h 6 SA (%)	0.908	0.846	0.970	0.930	0.878	0.981	.2420
Ara h 8 sIgE (kU_A/L)-Ara h 8 SA (%)	0.624	0.511	0.737	0.618	0.507	0.730	.8743
Ara h 9 sIgE (kU_A/L)-Ara h 9 SA (%)	0.468	0.354	0.581	0.465	0.351	0.578	.9552
WD-PD	0.628	0.517	0.739	0.898	0.831	0.964	<.0001

Comparison of AUC paired data ROC curves as described by deLong et al.^E1