Fig. 3

ROC curves. Predictive value of multivariate models for AEs (A), ADRs (B), thrombocytopenia (C) and bleeding (D) in Chinese STEMI patients receiving PCI and bivalirudin as anticoagulant. The ROC curve analysis was conducted based on the probability of the incidence to the events in each patient. The probability was calculated from the formulars generated by the multivariate logistic regression analysis. The formulars were listed as follows: \(P\left(\mathrm{AEs}=\left.1\right|\mathrm{x}\right)=\frac{{e}^{g\left(x\right)}}{1+{e}^{g\left(x\right)}}\). \(g\left(x\right)=-1.721+0.925\times \mathrm{History\, of\, cardiac\, surgery}+1.067\times \mathrm{History\, of \,critical\, respiratory\, disease}+0.465\times \mathrm{CRUSADE\, risk\, stratification}-0.997\times \mathrm{Operative \,timing}+1.601\times \mathrm{Combined \,with \,glycoprotein \,IIb}/\mathrm{IIIa \,inhibitors}\). \(P\left(\mathrm{ADRs}=\left.1\right|\mathrm{x}\right)=\frac{{e}^{g\left(x\right)}}{1+{e}^{g\left(x\right)}}\). \(g\left(x\right)=-4.902+1.444\times \mathrm{History \,of \,cardiac\, surgery}+1.425\times \mathrm{History\, of \,renal\, function \,impairment}+0.747\times \mathrm{CRUSADE\, risk\, stratification}+1.732\times \mathrm{Combined\, with \,glycoprotein \,IIb}/\mathrm{IIIa \,inhibitors}\). \(P\left(\mathrm{thrombocytopenia}=\left.1\right|\mathrm{x}\right)=\frac{{e}^{g\left(x\right)}}{1+{e}^{g\left(x\right)}}\). \(g(x)=-7.215+1.762\times \mathrm{History\, of \,cardiac\, surgery}+1.483\times \mathrm{History\, of\, renal\, function\, impairment}+0.911\times \mathrm{Operative\, timing}+2.201\times \mathrm{Combined \,with\, glycoprotein\, IIb}/\mathrm{IIIa \, inhibitors}\). ROC, receiver’s operating characteristics; AUC, area under curve; CI, confidence interval; AEs, adverse events; ADRs, adverse drug reactions; STEMI, ST-segment elevation myocardial infarction; PCI, percutaneous coronary intervention