Objective: To explore the predictive ability of Tp-ec, Q-Tc, Tp-e/Q-T and HRV on malignant arrhythmia during hospitalization. Method: 100 patients with malignant ventricular arrhythmia were included as the experimental group, another 100 patients without malignant ventricular arrhythmia were included as control group. The differences of Tp-ec, Q-Tc, Tp-e/Q-T and HRV were compared between the two groups. Multivariate logistic regression analysis was used to study variables and establish prediction model. ROC curve was used to evaluate the predictive ability and best predictive value of each index for malignant ventricular arrhythmia in hospital. Result: Compared with the control group, Tp-ec, Q-Tc, Tp-e/Q-T and HRV in the experimental group were significantly increased, (P < 0.001), HRV was decreased significantly. Multivariate logistic regression showed that the increase of Tp-ec, Q-Tc, Tp-e/Q-T and the decrease of HRV were the risk factors of malignant ventricular ventricular arrhythmia in hospital (OR = 11.169, 1.788, 1.001, 0.780), and bulid prediction model Z = -254.827 + 0.203 * Tp-ec + 0.581 * Q-Tc + 878.066 * Tp-e/Q-T-0.248 * SDNN. ROC curve showed that the area under the curve (AUC) of TP EC, Q-Tc, Tp-e/Q-T, HRV and predictive model for the diagnosis of malignant ventricular ventricular arrhythmia in hospital were 0.988, 0.905, 0.973, 0.901, 0.993, the best critical values were 100.365 ms, 447.078 ms, 0.239, 100.500, 181.792. Conclusion: The decrease of Tp-ec, Q-Tc, Tp-e/Q-T and HRV were the risk factors of malignant ventricular arrhythmia, and has predictive value for malignant ventricular arrhythmia in hospital. The prediction model combined with Tp-ec, Q-Tc, Tp-e/Q-T and HRV can improve the prediction ability of variables on malignant ventricular arrhythmia in hospital.
