Missing value processing method based on multiple fill method and guided by Markov Chain Monte Carlo (MCMC) simulation under Meta analysis numerical value missing state

Abstract

The invention discloses a missing value processing method based on a multiple fill method and guided by Markov Chain Monte Carlo (MCMC) simulation under a Meta analysis numerical value missing state. The missing value processing method is characterized in that the Bayesian probability theory serves as the base point of the missing value processing method, the method is guided by the MCMC random simulation method, based on the multiple fill method, and applied to the Meta analysis data missing state of any missingness pattern, so that integrated integration of missing value processing and missing data sets in conventional Meta analysis is achieved. By means of the method, the missing data can be fitted to the maximum degree, so that integrity of the extracted data and feasibility of subsequent standard statistic are guaranteed as much as possible, statistical test effectiveness of Meta analysis is remarkably promoted, robustness and possibility of a system evaluation result are improved, and scientificity and comprehensiveness of system overview evaluation are enhanced.

Description

technical field [0001] The invention relates to the field of medical statistics, in particular to an MCMC simulation-oriented method for processing missing values ​​based on multiple imputation methods in the state of missing values ​​in Meta analysis. Background technique [0002] In the field of biomedicine, Meta-analysis is the systematic analysis and quantitative synthesis of the results of multiple independent small-sample clinical trials and basic experimental research with the same research purpose, in order to improve the efficiency of statistical tests and increase the accuracy of effect value estimation. Resolve inconsistencies in findings across studies and seek new hypotheses. Meta-analysis is used in the evaluation and optimization of the accuracy of clinical diagnostic techniques, the evaluation and optimization of clinical treatment effects, the evaluation of causal associations in etiology, the evaluation of disease prevention interventions, the cost-benefit ...

AI Technical Summary

Problems solved by technology

The application of missing value processing in the data processing of Meta analysis can ensure the integrity of the extracted data to the greatest extent, and can solve the relatively common problems in missing data, especially when the data is in any missing mode, the MCMC model can be used to deal with it Complicated missing data problems can improve statistical efficiency; it can also effectively avoid the distorted variable distribution caused by the mean filling method, and make the distribution after substitution closer to the true value

Images