Evaluation Of The Effect Of Statins On The Treatment Of Chronic Heart Failure Based On The Marginal Structure Model

Objective:This study aimed to evaluate the effect of statins on N-terminal prohormone of brain natriuretic peptide(NT-proBNP),New York Heart Association Class(NYHA)and ej ection fraction(EF)in patients with chronic heart failure.And verify the function of the marginal structure model(MSM)used in the evaluation of the interventionMethods:The subjects were hospitalized patients with chronic heart failure admitted to the First Affiliated Hospital of Shanxi Medical University and Shanxi Provincial Cardiovascular Hospital from January 1,2014 to January 1,2019.The t-test,chi-square test,and logistic regression analysis were used to select for confounding factors.The MSM is used to evaluate the effectiveness of statins.During the probability estimation stage of the MSM,logistic and random forest(RF)are respectively performed.In the inverse probability weighting stage,stabilized weight and truncated weight are used to establish virtual population;in the effectiveness estimation stage,compared with traditional statistical analysis methods,verify whether the MSM can effectively deal with time-dependent confounding and treatment conversionResults:1.Using traditional logistic regression analysis,it was found that after excluding the effects of baseline confounders,statins did not significantly reduce NT-proBNP levels in patients with chronic heart failure[OR(95%CI):0.941(0.693?1.280),P=0.700];Using the MSM to exclude the effects of baseline confounding factors,time-dependent confounding factors,and treatment conversion,it was found that statins have a significant effect on reduce NT-proBNP,in logistic-based MSM,the OR(95%CI)is 0.642(0.488?0.845),P=0.002,in RF-based MSM,the OR(95%CI)is 0.699(0.528?0.926),P=0.012?In this study,statins have not been found to significantly impact NYHA and EF2.The MSM can effectively deal with the time-dependent confounding and treatment conversion in the evaluation of the effectiveness of statins in patients with chronic heart failure.In the probability estimation stage of the MSM,the probability obtained by the RF are more stable than the logisticConclusion:1.Statins can effectively reduce the concentration of NT-proBNP in patients with chronic heart failure,and provide a new research direction for the development of treatment plans for patients with chronic heart failure with elevated NT-proBNP.Statins have not been found to have an improving effect on NYHA and EF2.The MSM can effectively resolve the bias caused by time-dependent confounding factors and treatment conversion in evaluating the effectiveness of the intervention.The weight distribution of the MSM based on RF probability estimation is more stable than based on logistic.