摘要
纵向数据在不同时间点存在内在相关性问题,而传统的Meta分析技术无法处理这种问题。基于多水平模型的回归系数可以充分考虑纵向数据各个时间点的相关性。本文利用SAS软件实现多水平回归系数模型的Meta分析,并提供简洁易操作的SAS代码。
Longitudinal data had intrinsic correlation problems at different time points, and traditional metaanalysis techniques cannot resolve this problem. Regression coefficients based on multi-level models can fully consider the correlations of longitudinal data at various time points. This paper uses SAS software to perform multi-level regression coefficient model meta-analysis and provides programming code which is simple and easy to operate.
引文
1Jackson D,Rollins K,Coughlin P.A multivariate model for the meta-analysis of study level survival data at multiple times.Res Synth Methods,2014,5(3):264-272.
2Arends LR,Hunink MG,Stijnen T.Meta-analysis of summary survival curve data.Stat Med,2008,27(22):4381-4396.
3Ishak KJ,Platt RW,Joseph L,et al.Meta-analysis of longitudinal studies.Clin Trials,2007,4(5):525-539.
4Connis RT,Evans RL,Hendricks RD.Meta-analysis with longitudinal studies:controlling for analytical bias.Psychol Rep,1996,79(3 Pt 2):1383-1386.
5Lopes HF,Müller P,Rosner GL.Bayesian meta-analysis for longitudinal data models using multivariate mixture priors.Biometrics,2003,59(1):66-75.
6Maas CJM,Hox JJ,Lensvelt-Mulders GJLM.Longitudinal Meta-analysis.Quality Quantity,2004,38(4):381-389.
7Idris NRN.Modelling repeated measures data in meta-analysis:an alternative approach.Available at:http://irep.iium.edu.my/5552/1/SKSM16.pdf.
8张天嵩,董圣杰,周支瑞.高级Meta分析方法-基于Stata实现.上海:复旦大学出版社,2015.
9Tan A,Liao C,Mo Z,et al.Meta-analysis of holmium laser enucleation versus transurethral resection of the prostate for symptomatic prostatic obstruction.Br J Surg,2007,94(10):1201-1208.
10Jackson D,White IR,Riley RD.A matrix-based method of moments for fitting the multivariate random effects model for meta-analysis and meta-regression.Biom J,2013,55(2):231-245.