. Author manuscript; available in PMC: 2019 Jan 1.

Published in final edited form as: Obes Surg. 2018 Jan;28(1):61–68. doi: 10.1007/s11695-017-2799-y

Table 2.

Area under the curve and bootstrap 95% confidence intervals

Group	N	Optimal cut-off point^*			AUC	Bootstrap 95% CI
Group	N	DiaRem	Sensitivity	Specificity	AUC	Bootstrap 95% CI
All subjects	520	8	0.62	0.91	0.825	[0.785, 0.859]
White	361	8	0.64	0.93	0.839	[0.798, 0.879]
Hispanic	130	7	0.59	0.85	0.790	[0.710, 0.861]
RYGB	340	8	0.69	0.91	0.858	[0.818, 0.898]
LSG	119	7	0.56	0.89	0.707	[0.604, 0.807]
LAGB	61	4	0.80	0.85	0.880	[0.775, 0.962]
White and RYGB	291	8	0.69	0.92	0.860	[0.814, 0.903]
White and LSG	36	8	0.47	0.94	0.714	[0.523, 0.876]
White and LAGB	34	4	0.71	0.80	0.833	[0.673, 0.956]
Hispanic and RYGB	37	15	0.69	0.90	0.863	[0.722, 0.961]
Hispanic and LSB	71	7	0.58	0.84	0.713	[0.582, 0.830]
Hispanic and LAGB	22	5	0.89	1.00	0.947	[0.833, 1.000]

Optimal cut-off point was defined as the DiaRem score that maximizes sensitivity and specificity