Applications of Bayesian Mixed Treatment Comparison in Antihvpertensive Agents’Drug Evaluations
|Course||Public Health and Preventive Medicine|
|Keywords||Hypertension Antihypertensive Agent Bayesian Mixed TreatmentComparison|
ObjectiveThe situation of hypertension is severe in China, and bring huge burden of disease. There are so many antihypertensive agents, and how to select the optimal drug becomes a problem. In this paper, the method of Bayesian mixed treatment comparison is simply introduced, and taking the drug evaluations of different antihypertensive agents as an example, to explore the application of Bayesian mixed treatment comparison in drug evaluations of antihypertensive agents, to provide methodological basis for further use in future’s antihypertensive agents’drug evaluation.MethodsAccording to its principle and procedure, Bayesian mixed treatment comparison was applied in the example of drug evaluations of different classes of antihypertensive drug evaluation. After collecting data from literature database in China, Bayesian mixed treatment comparison was implemented through the soft of WinBUGS, and the efficacy of different classes of antihypertensive agents was evaluated. ResultsIn the example of antihypertensive agents’ drug evaluation,31literatures was included after the selection, and then the extracted data was analyzed through Bayesian mixed treatment comparison. It was found that all classed of antihypertensive agents could be ranked by ranking the Bayesian probability of each efficacy index:for the reduction of SBP, compound hypotensor was the best, and ACEI was the worst; for the response rate, compound hypotensor was the best, and β-blocker was the worst; for the incidence of adverse events, ARB was the best, ACEI was the worst; for the reduction of10-year risk of ICVD, ARB was the best, and β-blocker was the worst.ConclusionsBayesian mixed treatment comparison is the development of the traditional Meta-analysis. In drug evaluations of antihypertensive agents, its proper application can make it possible comparing two drugs when there is no head-to-head trail, and also can work it out comparing several drugs in the same time and showing which drug is the best, and then provide useful information for clinical physician or policy decision makers in order to choose the optimal drug.