Research Article

Risk Prediction of One-Year Mortality in Patients with Cardiac Arrhythmias Using Random Survival Forest

Table 3

Cox proportional hazard model with comprehensive risk factors.

PredictorsCoefficient valueHR95.0% CI
LowerUpper

Demographics
 log age.633<0.0011.8831.6372.167
 log BMI−.979<0.001.376.313.451
Clinical risk factors
 Cardiac arrest.890<0.0012.4352.1642.741
 Slow arrhythmias−.420<0.0010.6570.5720.755
 CHF.1190.0021.1261.0421.216
 Myocardial infraction.1680.0011.1821.0691.307
 Stroke.340<0.0011.4051.2661.560
 Renal failure.243<0.0011.2751.1321.437
Laboratory risk factors
 log K−1.146<0.001.318.256.395
 log WBC.266<0.0011.3051.2101.408
 log RBC−.4430.001.642.494.834
 log BUN.562<0.0011.7541.6051.916
 log glucose.1870.0021.2051.0711.356
 log CKPK−.106<0.001.900.872.928
 log AST.379<0.0011.4601.3571.572
 log ALT−.255<0.001.775.722.832
 log PT−.447<0.001.640.523.782
 log INR.334<0.0011.3971.2281.590
 log BR.107<0.0011.1131.0621.165
Medications
 Class I agents0.3760.0021.4561.1471.849
 Class II agents−0.316<0.0010.7290.6700.793
 Class III agents−0.864<0.0010.4210.3520.505
 Class V agents−0.203<0.0010.8160.7540.883

HR: hazard ratio; CI: confidence level; log  indicates log of variables.