Predicting the serum digoxin concentrations of infants in the neonatal intensive care unit through an artificial neural network

Table 3 Area under the curve (AUC) of the receiver operating characteristic (ROC) curves to differentiate toxicity concentration (i.e., equal and above 1.5 ng/ml) or not for each ANN model on validation dataset

Model	No. of parameters	Parameters	AUC	SE	Sig.	95% CI
Model	No. of parameters	Parameters	AUC	SE	Sig.	Lower Bond	Upper Bond
1	11	All Variables	0.558	0.151	0.686	0.263	0.854
2	10	-Sex	0.658	0.152	0.273	0.361	0.956
3	9	-Sex-DCM	0.738	0.140	0.100	0.463	1.000
4	8	-Sex-DCM -PH	0.658	0.152	0.273	0.361	0.956
5	7	-Sex-DCM -PH -Captopril	0.575	0.147	0.603	0.286	0.864
6	6	-Sex-DCM -PH -Captopril -Furosemide	0.617	0.150	0.419	0.324	0.910
7	5	-Sex-DCM -PH -Captopril -Furosemide -VSD	0.633	0.141	0.356	0.357	0.909
8	4*	-Sex-DCM -PH -Captopril -Furosemide -VSD -ibuprofen	0.638	0.151	0.341	0.342	0.933

“- “in the column of parameters refers to “exclude” that specific variable from the model 1, which contain all variables. AUC area under the curve; SE standard error of AUC; Sig. significance of AUC finding
All 11 variables include: dose per total body weight, gender, postmenstrual age (PMA), Congestive heart failure (CHF), dilated cardiomyopathy (DCM), pulmonary hypertension (PH), Ventricular septal defect (VSD), with captopril, with furosemide, with ibuprofen
*Common variables used in population pharmacokinetics were dose per total body weight, PMA, CHF

ISSN: 1471-2431