Study of clinical experiments of novel sedative drugs based on chi-square test and logistics regression model

Abstract

Local sedation and analgesic drugs are often indispensable in medical trauma surgery. They can not only assist doctors to complete the operation, but also greatly reduce the pain of patients. Therefore, constantly looking for better effective sedative drugs is an extremely important field in medical research. For the new sedative drug "R drug", this paper uses the data analysis of patient intraoperative and postoperative adverse reactions, patient vital signs, patient satisfaction and related problems to establish an appropriate mathematical model and evaluate its clinical efficacy. First, Data cleaning was performed on the data[1], Quantification and normalized dimensionless pretreatment; Next, We calculated K2 values using the chi-square test and Yates's correction, By comparing the K2 values to the standard values, Then to judge whether there are significant differences between each index; Then, When predicting adverse reactions, we used a logistics regression model to predict the probability that each reaction may occur, The accuracy of the model is tested by the heat map of the confusion matrix and the ROC curve; Last, We performed the data analysis of each index in 0-5 minutes, By mapping the nuclear density of the different indicators under the B and R drugs, Mann-Whitney U test using independent samples, P values were calculated to determine whether the new drug group and the original drug group showed significant differences in vital sign data.