Data-driven prediction of adverse drug reactions induced by drug-drug interactions

Table 1 Operational characteristics of adverse drug reaction (ADR) models

Adverse drug reaction	AUC	Accuracy	Sensitivity	Specificity	PPV
Exemplar ADR models
Rheumatic heart disease	0.97	0.96	0.81	0.96	0.02
Heat stroke	0.90	0.84	0.83	0.85	0.02
Spontaneous abortion	0.84	0.89	0.63	0.89	0.05
Vestibular disorder	0.91	0.91	0.73	0.91	0.07
All 1,096 ADR models
Mean	0.86	0.89	0.63	0.90	0.09
Standard deviation	0.05	0.05	0.15	0.05	0.08

The performance of models to predict ADRs induced by drug-drug interactions (DDIs) were estimated from 10-fold cross-validation. The values were calculated by using a DDI score threshold of 0.0 for all ADRs. The calculated positive predictive values (PPV) are given for the developed ADR models

ISSN: 2050-6511