Some general points in estimating heterogeneity variance with the DerSimonian-Laird estimator

Böhning, Dankmar; Malzahn, Uwe; Dietz, Ekkehart; Schlattmann, Peter; Viwatwongkasem, Chukiat; Biggeri, Annibale

doi:10.1093/biostatistics/3.4.445

Cited by 98 publications

(92 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Herew i = 1/ˆ 2 i , and Q 2w is obtained by replacing w i byw i in Q 2 w (see also Reference [8]). The estimatorˆ 2 MM may also yield a negative estimate for the heterogeneity variance, and hence the truncated version max{0,ˆ 2 MM } is usually used.…”

Section: The Methods Of Moments Estimator (Mm)mentioning

confidence: 99%

“…, k as in practice, and we denote the estimate for the ith study byˆ 2 i . Strictly speaking, such a practice is problematic because of uncertainty in the estimatedˆ 2 i (see References [8,17,22]). Nevertheless, it is common to treat the estimatedˆ 2 i as fixed and known values when combining summary statistics from several studies.…”

Section: The Heterogeneity Variance Estimatorsmentioning

confidence: 99%

See 1 more Smart Citation

A comparison of heterogeneity variance estimators in combining results of studies

Sidik

Jonkman

2006

Statistics in Medicine

294

316

View full text Add to dashboard Cite

For random effects meta-analysis, seven different estimators of the heterogeneity variance are compared and assessed using a simulation study. The seven estimators are the variance component type estimator (VC), the method of moments estimator (MM), the maximum likelihood estimator (ML), the restricted maximum likelihood estimator (REML), the empirical Bayes estimator (EB), the model error variance type estimator (MV), and a variation of the MV estimator (MVvc). The performance of the estimators is compared in terms of both bias and mean squared error, using Monte Carlo simulation. The results show that the REML and especially the ML and MM estimators are not accurate, having large biases unless the true heterogeneity variance is small. The VC estimator tends to overestimate the heterogeneity variance in general, but is quite accurate when the number of studies is large. The MV estimator is not a good estimator when the heterogeneity variance is small to moderate, but it is reasonably accurate when the heterogeneity variance is large. The MVvc estimator is an improved estimator compared to the MV estimator, especially for small to moderate values of the heterogeneity variance. The two estimators MVvc and EB are found to be the most accurate in general, particularly when the heterogeneity variance is moderate to large.

show abstract

Section: The Methods Of Moments Estimator (Mm)mentioning

confidence: 99%

Section: The Heterogeneity Variance Estimatorsmentioning

confidence: 99%

A comparison of heterogeneity variance estimators in combining results of studies

Sidik

Jonkman

2006

Statistics in Medicine

294

316

View full text Add to dashboard Cite

show abstract

“…On the other hand, if the heterogeneity χ 2 test results in a P<0.05, it indicates diversity and heterogeneity between studies. Therefore, the use of random or random effect tests is recommended, such as the DerSimonian-Laird tests (4,6,26,91) . Global association tests were then used to assess the significance of the correlation between the polymorphism of gene GSTM1 and gastric cancer for all studies combined.…”

Section: Discussionmentioning

confidence: 99%

Genotype Association Gstm1 Null and Gastric Cancer: Evidence-Based Meta-Analysis

Ribeiro

Nascimento

Silva

2017

Arq. Gastroenterol.

View full text Add to dashboard Cite

-Background -Gastric cancer is the fourth most common cancer in men and the sixth among women, except for non-melanoma skin tumors, in Brazil. Epidemiological evidences reveal the multifactorial etiology of this cancer, highlighting risk factors such as: infection by the bacterium Helicobacter pylori, advanced age, smoking, chronic alcohol abuse, eating habits and genetic polymorphisms. Considering the context of genetic polymorphisms, there is the absence of the GSTM1 gene. The lack of GSTM1 function to detoxify xenobiotics and promote defense against oxidative stress leads to increased DNA damage, promoting gastric carcinogenesis. This process is multifactorial and the development of gastric cancer results from a complex interaction of these variables. Objective -The aim of this study was to investigate the association of GSTM1 null polymorphism in the pathogenesis of gastric cancer. Methods -A meta-analysis was conducted from 70 articles collected in SciELO and PubMed databases, between September 2015 and July 2016. In order to evaluate a possible association, we used the odds ratio (OR) and confidence interval of 95% (CI 95%). To assess the heterogeneity of the studies was used the chi-square test. Statistical analysis was performed using the BioEstat ® 5.3. Results -This study included 70 studies of case-control, including 28,549 individuals, which were assessed for the null polymorphism of the GSTM1 gene, and of which 11,208 (39.26%) were cases and 17,341 (60.74%) were controls. The final analysis showed that the presence of the GSTM1 gene acts as a protective factor against the development of gastric cancer (OR=0.788; 95%CI 0.725-0.857; P<0.0001). Positive statistical association was found in Asia (OR=0.736; P<0.0001) and Eurasia (OR=0.671; P=0.05). However, statistically significant data was not obtained in Europe (OR=1.033; 95%CI 0.873-1.222; P=0.705) and America (OR=0.866; 95%CI 0.549-1.364; P=0.534). Therefore, the results can not be deduced around the world. Conclusion -This meta-analysis concluded that the presence of the GSTM1 gene is a protector for the emergence of gastric cancer, especially in Asian countries, but this result was not found in Europe and America.

show abstract

“…If P>0.10, the studies were considered to lack heterogeneity and the pooled ORs were calculated using the fixed-effects model according to the Mantel-Haenszel method (15). Otherwise, the random-effect model was used according to the DerSimonian-Laird method (16). All the statistical analyses in the present study were performed using Review Manager 5 (http://tech.cochrane.org/revman) and Stata software ( version 8.2;StataCorp LP, College Station, TΧ, USA).…”

Section: Meta-analysismentioning

confidence: 99%

p16 promoter hypermethylation is associated with increased risk of nasopharyngeal carcinoma

Shao

Jiang

et al. 2014

Molecular and Clinical Oncology

View full text Add to dashboard Cite

Abstract. The present study aimed to investigate the effects of p16 hypermethylation on the risk of nasopharyngeal carcinoma (NPC) quantitatively, through a meta-analysis of available case-control studies including malignant and normal NPC tissue samples. Crude odds ratios (ORs) with 95% confidence intervals (CIs) were extracted and pooled to assess the strength of the association between p16 hypermethylation and NPC risk. A total of 9 studies, including 406 NPC and 376 control cases, were identified for meta-analysis. Statistically significant ORs of p16 hypermethylation were obtained from the NPC and control groups (OR=19.53; 95% CI: 9.54-39.97; P=0.685). The meta-analysis indicated that p16 hypermethylation significantly increases NPC risk.

show abstract

Some general points in estimating heterogeneity variance with the DerSimonian-Laird estimator

Cited by 98 publications

References 16 publications

A comparison of heterogeneity variance estimators in combining results of studies

A comparison of heterogeneity variance estimators in combining results of studies

Genotype Association Gstm1 Null and Gastric Cancer: Evidence-Based Meta-Analysis

p16 promoter hypermethylation is associated with increased risk of nasopharyngeal carcinoma

Contact Info

Product

Resources

About