Performance of statistical methods for meta-analysis when true study effects are non-normally distributed: A comparison between DerSimonian–Laird and restricted maximum likelihood

@article{Kontopantelis2012PerformanceOS,
  title={Performance of statistical methods for meta-analysis when true study effects are non-normally distributed: A comparison between DerSimonian–Laird and restricted maximum likelihood},
  author={Evangelos Kontopantelis and David Reeves},
  journal={Statistical Methods in Medical Research},
  year={2012},
  volume={21},
  pages={657 - 659}
}
However, due to computational limitations we did not include Restricted Maximum Likelihood (REML) estimator for the between-study variance. Lately, we have observed that the iterative REML approach has been increasingly replacing the non-iterative DerSimonian-Laird (DL) as the method of choice in published meta-analyses. Jackson et al examined the performance of the two methods in terms of coverage, for normally distributed effects only, and found that results for the two methods were similar. 
The Stata Journal
This article describes the new meta-analysis command metaan, which can be used to perform fixedor random-effects meta-analysis. Besides the standard DerSimonian and Laird approach, metaan offers aExpand
A comparison of heterogeneity variance estimators in simulated random-effects meta-analyses.
TLDR
The estimated summary effect of the meta-analysis and its confidence interval derived from the Hartung-Knapp-Sidik-Jonkman method are more robust to changes in the heterogeneity variance estimate and show minimal deviation from the nominal coverage of 95% under most of the simulated scenarios. Expand
Estimating the Heterogeneity Variance in a Random-Effects Meta-Analysis
In a meta-analysis, differences in the design and conduct of studies may cause variation in effects beyond what is expected from chance alone. This additional variation is commonly known asExpand
Estimation of an overall standardized mean difference in random-effects meta-analysis if the distribution of random effects departs from normal.
TLDR
This study examines the performance of various random-effects methods for computing an average effect size estimate and a confidence interval around it, when the normality assumption is not met, suggesting that Hartung's profile likelihood methods yielding the best performance under suboptimal conditions. Expand
Selecting the best meta-analytic estimator for evidence-based practice: a simulation study.
TLDR
A simulation study was conducted to compare estimator performance and demonstrates that the IVhet and quality effects estimators, though biased, have the lowest mean squared error. Expand
Methods to estimate the between‐study variance and its uncertainty in meta‐analysis†
TLDR
The aim is to identify known methods for estimation of the between‐study variance and its corresponding uncertainty, and to summarise the simulation and empirical evidence that compares them and recommend the Q‐profile method and the alternative approach based on a ‘generalised Cochran between‐ study variance statistic’. Expand
Evaluation of the Normality Assumption in Meta-analyses.
TLDR
This work presents a standardization framework for evaluation of the normality assumption and examines its performance in random-effects meta-analyses with simulation studies and real examples and illustrates the real-world significance of examining thenormality assumption with examples. Expand
A Re-Analysis of the Cochrane Library Data: The Dangers of Unobserved Heterogeneity in Meta-Analyses
TLDR
It is found that assuming homogeneity often results in a misleading analysis, since heterogeneity is very likely present but undetected, and one solution is to test the sensitivity of the meta-analysis conclusions to assumed moderate and large degrees of heterogeneity. Expand
When should meta‐analysis avoid making hidden normality assumptions?
  • D. Jackson, I. White
  • Medicine, Computer Science
  • Biometrical journal. Biometrische Zeitschrift
  • 2018
TLDR
It is concluded that statistical methods that make fewer normality assumptions should be considered more often in practice, and alternatives that make less use of the normal distribution are discussed. Expand
Meta-analysis in clinical trials revisited.
TLDR
A random-effects model to summarize the evidence about treatment efficacy from a number of related clinical trials and a discussion of repurposing the method for Big Data meta-analysis and Genome Wide Association Studies for studying the importance of genetic variants in complex diseases are reviewed. Expand
...
1
2
3
4
5
...

References

SHOWING 1-4 OF 4 REFERENCES
Performance of statistical methods for meta-analysis when true study effects are non-normally distributed: A simulation study
TLDR
The performance of the fixed-effects approach and seven frequentist RE MA methods are discussed, including DerSimonian–Laird, Q-based, maximum likelihood, profile likelihood, Biggerstaff–Tweedie, Sidik–Jonkman and Follmann–Proschan. Expand
Metaan: Random-effects Meta-analysis
This article describes the new meta-analysis command metaan, which can be used to perform fixed- or random-effects meta-analysis. Besides the standard DerSimonian and Laird approach, metaan offers aExpand
How does the DerSimonian and Laird procedure for random effects meta-analysis compare with its more efficient but harder to compute counterparts?
The procedure suggested by DerSimonian and Laird is the simplest and most commonly used method for fitting the random effects model for meta-analysis. Here it is shown that, unless all studies are ofExpand
Meta-analysis in clinical trials.
TLDR
This paper examines eight published reviews each reporting results from several related trials in order to evaluate the efficacy of a certain treatment for a specified medical condition and suggests a simple noniterative procedure for characterizing the distribution of treatment effects in a series of studies. Expand