scholarly journals Select relevant moderators using Bayesian regularized meta-regression

2021 ◽  
Author(s):  
Caspar J. Van Lissa ◽  
Sara van Erp
Keyword(s):  
State Of The Art ◽  
Meta Analysis ◽  
Predictive Performance ◽  
Small Sample ◽  
Training Data ◽  
Type I ◽  
Lower Ability ◽  
Meta Regression ◽  
Meta Analyses ◽  
Type Ii Errors

When analyzing a heterogeneous body of literature, there may be many potentially relevant between-studies differences. These differences can be coded as moderators, and accounted for using meta-regression. However, many applied meta-analyses lack the power to adequately account for multiple moderators, as the number of studies on any given topic is often low. The present study introduces Bayesian Regularized Meta-Analysis (BRMA), an exploratory algorithm that can select relevant moderators from a larger number of candidates. This approach is suitable when heterogeneity is suspected, but it is not known which moderators most strongly influence the observed effect size. We present a simulation study to validate the performance of BRMA relative to state-of-the-art meta-regression (RMA). Results indicated that BRMA compared favorably to RMA on three metrics: predictive performance, which is a measure of the generalizability of results, the ability to reject irrelevant moderators, and the ability to recover population parameters with low bias. BRMA had slightly lower ability to detect true effects of relevant moderators, but the overall proportion of Type I and Type II errors was equivalent to RMA. Furthermore, BRMA regression coefficients were slightly biased towards zero (by design), but its estimates of residual heterogeneity were unbiased. BRMA performed well with as few as 20 studies in the training data, suggesting its suitability as a small sample solution. We discuss how applied researchers can use BRMA to explorate between-studies heterogeneity in meta-analysis.

scholarly journals Cluster Wild Bootstrapping to Handle Dependent Effect Sizes in Meta-Analysis with a Small Number of Studies

2021 ◽  
Author(s):  
Megha Joshi ◽  
James E Pustejovsky ◽  
S. Natasha Beretvas
Keyword(s):  
Effect Size ◽  
Type I Error ◽  
Meta Analysis ◽  
Error Rates ◽  
Small Sample ◽  
Type I ◽  
Hypothesis Tests ◽  
Type I Error Rates ◽  
Meta Analyses ◽  
Small Sample Correction

The most common and well-known meta-regression models work under the assumption that there is only one effect size estimate per study and that the estimates are independent. However, meta-analytic reviews of social science research often include multiple effect size estimates per primary study, leading to dependence in the estimates. Some meta-analyses also include multiple studies conducted by the same lab or investigator, creating another potential source of dependence. An increasingly popular method to handle dependence is robust variance estimation (RVE), but this method can result in inflated Type I error rates when the number of studies is small. Small-sample correction methods for RVE have been shown to control Type I error rates adequately but may be overly conservative, especially for tests of multiple-contrast hypotheses. We evaluated an alternative method for handling dependence, cluster wild bootstrapping, which has been examined in the econometrics literature but not in the context of meta-analysis. Results from two simulation studies indicate that cluster wild bootstrapping maintains adequate Type I error rates and provides more power than extant small sample correction methods, particularly for multiple-contrast hypothesis tests. We recommend using cluster wild bootstrapping to conduct hypothesis tests for meta-analyses with a small number of studies. We have also created an R package that implements such tests.

Clinical outcomes of cancer patients with COVID-19: A systematic review and meta-analysis.

2021 ◽  
Vol 39 (15_suppl) ◽  
pp. e18600-e18600
Author(s):  
Maryam Alasfour ◽  
Salman Alawadi ◽  
Malak AlMojel ◽  
Philippos Apolinario Costa ◽  
Priscila Barreto Coelho ◽  
...  
Keyword(s):  
Mortality Rate ◽  
Cancer Patients ◽  
Clinical Outcomes ◽  
Meta Analysis ◽  
Small Sample ◽  
Prediction Intervals ◽  
Invasive Ventilation ◽  
Newcastle Ottawa Scale ◽  
Small Sample Sizes ◽  
Meta Analyses

e18600 Background: Patients with coronavirus disease 2019 (COVID-19) and cancer have worse clinical outcomes compared to those without cancer. Primary studies have examined this population, but most had small sample sizes and conflicting results. Prior meta-analyses exclude most US and European data or only examine mortality. The present meta-analysis evaluates the prevalence of several clinical outcomes in cancer patients with COVID-19, including new emerging data from Europe and the US. Methods: A systematic search of PubMED, medRxiv, JMIR and Embase by two independent investigators included peer-reviewed papers and preprints up to July 8, 2020. The primary outcome was mortality. Other outcomes were ICU and non-ICU admission, mild, moderate and severe complications, ARDS, invasive ventilation, stable, and clinically improved rates. Study quality was assessed through the Newcastle–Ottawa scale. Random effects model was used to derive prevalence rates, their 95% confidence intervals (CI) and 95% prediction intervals (PI). Results: Thirty-four studies (N = 4,371) were included in the analysis. The mortality prevalence rate was 25.2% (95% CI: 21.1–29.7; 95% PI: 9.8-51.1; I 2 = 85.4), with 11.9% ICU admissions (95% CI: 9.2-15.4; 95% PI: 4.3-28.9; I 2= 77.8) and 25.2% clinically stable (95% CI: 21.1-29.7; 95% PI: 9.8-51.1; I 2 = 85.4). Furthermore, 42.5% developed severe complications (95% CI: 30.4-55.7; 95% PI: 8.2-85.9; I 2 = 94.3), with 22.7% developing ARDS (95% CI: 15.4-32.2; 95% PI: 5.8-58.6; I 2 = 82.4), and 11.3% needing invasive ventilation (95% CI: 6.7-18.4; 95% PI: 2.3-41.1; I 2 = 79.8). Post-follow up, 49% clinically improved (95% CI: 35.6-62.6; 95% PI: 9.8-89.4; I 2 = 92.5). All outcomes had large I 2 , suggesting high levels of heterogeneity among studies, and wide PIs indicating high variability within outcomes. Despite this variability, the mortality rate in cancer patients with COVID-19, even at the lower end of the PI (9.8%), is higher than the 2% mortality rate of the non-cancer with COVID-19 population, but not as high as what other meta-analyses conclude, which is around 25%. Conclusions: Patients with cancer who develop COVID-19 have a higher probability of mortality compared to the general population with COVID-19, but possibly not as high as previous studies have shown. A large proportion of them developed severe complications, but a larger proportion recovered. Prevalence of mortality and other outcomes published in prior meta-analyses did not report prediction intervals, which compromises the clinical utilization of such results.

Statistical and measurement pitfalls in the use of meta-regression in meta-analysis

2017 ◽  
Vol 22 (5) ◽  
pp. 469-476 ◽  
Author(s):  
Frank L. Schmidt
Keyword(s):  
Social Sciences ◽  
Organizational Behavior ◽  
Meta Analysis ◽  
Social Science Research ◽  
Science Research ◽  
Content Type ◽  
Marketing Social ◽  
Meta Regression ◽  
Meta Analyses ◽  
Almost All

Purpose Meta-regression is widely used and misused today in meta-analyses in psychology, organizational behavior, marketing, management, and other social sciences, as an approach to the identification and calibration of moderators, with most users being unaware of serious problems in its use. The purpose of this paper is to describe nine serious methodological problems that plague applications of meta-regression. Design/methodology/approach This paper is methodological in nature and is based on well-established principles of measurement and statistics. These principles are used to illuminate the potential pitfalls in typical applications of meta-regression. Findings The analysis in this paper demonstrates that many of the nine statistical and measurement pitfalls in the use of meta-regression are nearly universal in applications in the literature, leading to the conclusion that few meta-regressions in the literature today are trustworthy. A second conclusion is that in almost all cases, hierarchical subgrouping of studies is superior to meta-regression as a method of identifying and calibrating moderators. Finally, a third conclusion is that, contrary to popular belief among researchers, the process of accurately identifying and calibrating moderators, even with the best available methods, is complex, difficult, and data demanding. Practical implications This paper provides useful guidance to meta-analytic researchers that will improve the practice of moderator identification and calibration in social science research literatures. Social implications Today, many important decisions are made on the basis of the results of meta-analyses. These include decisions in medicine, pharmacology, applied psychology, management, marketing, social policy, and other social sciences. The guidance provided in this paper will improve the quality of such decisions by improving the accuracy and trustworthiness of meta-analytic results. Originality/value This paper is original and valuable in that there is no similar listing and discussion of the pitfalls in the use of meta-regression in the literature, and there is currently a widespread lack of knowledge of these problems among meta-analytic researchers in all disciplines.

Abstract P260: Large-scale Meta-analysis of Diverse Ancestry Genome-wide Association Studies to Identify Pharmacogenomic Interactions of Sulfonylureas and ECG Phenotypes

Circulation ◽  
2016 ◽  
Vol 133 (suppl_1) ◽  
Author(s):  
James S Floyd ◽  
Colleen Sitlani ◽  
Christy L Avery ◽  
Eric A Whitsel ◽  
Leslie Lange ◽  
...  
Keyword(s):  
Repeated Measures ◽  
Qt Interval ◽  
Association Studies ◽  
Meta Analysis ◽  
Small Sample ◽  
Genome Wide Association ◽  
European Ancestry ◽  
Genome Wide Association Studies ◽  
Genome Wide ◽  
Meta Analyses

Introduction: Sulfonylureas are a commonly-used class of diabetes medication that can prolong the QT-interval, which is a leading cause of drug withdrawals from the market given the possible risk of life-threatening arrhythmias. Previously, we conducted a meta-analysis of genome-wide association studies of sulfonylurea-genetic interactions on QT interval among 9 European-ancestry (EA) cohorts using cross-sectional data, with null results. To improve our power to identify novel drug-gene interactions, we have included repeated measures of medication use and QT interval and expanded our study to include several additional cohorts, including African-American (AA) and Hispanic-ancestry (HA) cohorts with a high prevalence of sulfonylurea use. To identify potentially differential effects on cardiac depolarization and repolarization, we have also added two phenotypes - the JT and QRS intervals, which together comprise the QT interval. Hypothesis: The use of repeated measures and expansion of our meta-analysis to include diverse ancestry populations will allow us to identify novel pharmacogenomic interactions for sulfonylureas on the ECG phenotypes QT, JT, and QRS. Methods: Cohorts with unrelated individuals used generalized estimating equations to estimate interactions; cohorts with related individuals used mixed effect models clustered on family. For each ECG phenotype (QT, JT, QRS), we conducted ancestry-specific (EA, AA, HA) inverse variance weighted meta-analyses using standard errors based on the t-distribution to correct for small sample inflation in the test statistic. Ancestry-specific summary estimates were combined using MANTRA, an analytic method that accounts for differences in local linkage disequilibrium between ethnic groups. Results: Our study included 65,997 participants from 21 cohorts, including 4,020 (6%) sulfonylurea users, a substantial increase from the 26,986 participants and 846 sulfonylureas users in the previous meta-analysis. Preliminary ancestry-specific meta-analyses have identified genome-wide significant associations (P < 5х10–8) for each ECG phenotype, and analyses with MANTRA are in progress. Conclusions: In the setting of the largest collection of pharmacogenomic studies to date, we used repeated measurements and leveraged diverse ancestry populations to identify new pharmacogenomic loci for ECG traits associated with cardiovascular risk.

Associations between paraoxonase-1 and systemic lupus erythematosus

Lupus ◽  
2019 ◽  
Vol 28 (13) ◽  
pp. 1571-1576
Author(s):  
S -C Bae ◽  
Y H Lee
Keyword(s):  
Systemic Lupus Erythematosus ◽  
Confidence Interval ◽  
Lupus Erythematosus ◽  
Meta Analysis ◽  
Small Sample ◽  
Mean Difference ◽  
Paraoxonase 1 ◽  
Systemic Lupus ◽  
Standard Mean Difference ◽  
Meta Analyses

Objective The objective of this analysis was to explore associations between paraoxonase-1 levels, gene polymorphisms and systemic lupus erythematosus. Methods Meta-analyses of paraoxonase-1 levels and Q192R and L55M and polymorphisms in systemic lupus erythematosus were conducted. Results Nine articles were incorporated in our meta-analysis, which uncovered that the paraoxonase-1 level was decreased in systemic lupus erythematosus compared to control (standard mean difference = −1.626, 95% confidence interval = −2.829–−0.424, p = 0.008). Ethnicity-specific meta-analysis demonstrated a relation tendency between decreased paraoxonase-1 activity and lupus in Europeans (standard mean difference = −1.236, 95% confidence interval = −2.634–0.163, p = 0.083). Paraoxonase-1 activity was reduced in systemic lupus erythematosus in a single Arab and African population. Decreased paraoxonase-1 activity was found in a small sample of systemic lupus erythematosus patients (standard mean difference = −1.642, 95% confidence interval = −3.076–−0.247, p = 0.021). Ethnicity-specific analysis indicated a relationship between the paraoxonase-1 55 M allele in the Arab systemic lupus erythematosus population. However, a lack of association with systemic lupus erythematosus and the paraoxonase-1 192 R allele was observed. Conclusions Meta-analyses revealed reduced paraoxonase-1 activity in patients with systemic lupus erythematosus and found possible associations between systemic lupus erythematosus and paraoxonase-1 L55M polymorphism in a specific ethnic group.

scholarly journals Vitamin D2 and D3 Supplementation Effect on Serum 25(OH)D Levels: A Meta-Analysis

2020 ◽  
Vol 4 (Supplement_2) ◽  
pp. 1782-1782
Author(s):  
Meline Chakalian ◽  
Joyce Cao ◽  
Jiang Hu ◽  
Casey Vanous ◽  
Simon Sum
Keyword(s):  
Vitamin D ◽  
Meta Analysis ◽  
Small Sample ◽  
Relative Potency ◽  
Health Concern ◽  
Outcome Variable ◽  
Vitamin D2 ◽  
Baseline Serum ◽  
Meta Regression ◽  
The Impact

Abstract Objectives Vitamin D insufficiency is a global health concern that affects nearly 50% of the population worldwide. Growing demand for vegan/vegetarian products has aroused interest in the plant-sourced D2 form for use in dietary supplements. However, vitamin D2’s ability to raise serum 25(OH)D levels in relation to D3 among existing scientific literature is inconclusive. This study sought to compare vitamin D2 to D3 in increasing serum 25(OH)D levels in order to better understand the relative potency and dosage required to address vitamin D insufficiencies. Methods PubMed and Embase databases were searched through July of 2018. Randomized controlled trials comparing D2 and D3 supplementation of equivalent dosages and the resulting increase in serum 25(OH)D levels in adults were eligible for this meta-analysis. A meta regression was conducted to compare the impact of both vitamin D forms on serum 25(OH)D levels. The outcome variable evaluated was the serum 25(OH)D levels. Results Nine RCTs (n = 628) with vitamin D dose ranging from 10 mcg per day to 1250 mcg per week, and an intervention duration from 2 to 16 weeks were eligible. Subjects included healthy adults as well as those with chronic kidney disease. There was substantial heterogeneity among the studies (I2 = 78.07%). The meta-regression showed vitamin D supplementation regardless of form was effective in raising serum 25(OH)D levels (P &lt; 0.0001). The mean effect size expressed as the standardized mean difference (SMD) from baseline serum 25(OH)D levels was 1.16 [95% CI: 0.83, 1.49] for D2 and 1.52 [95% CI: 0.99, 2.04] for D3. While there was a trend of greater increase caused by D3 numerically, the difference between D2 and D3 was not statistically significant. When duration and frequency of supplementation were examined, similar trends of non-significant greater increases for D3 relative to D2 were observed. Conclusions This research shows both vitamin D2 and D3 supplementation can significantly increase serum 25(OH)D levels. Though the results did not reach statistical significance, there is a consistent trend of vitamin D3 offering additional effectiveness relative to D2. The high heterogeneity across studies and small sample size likely contributed to the non-significant results and limited the ability to identify a quantitative relative potency that can be used for a D2 dosage recommendation. Funding Sources None.

scholarly journals Epigenome-wide meta-analysis of blood DNA methylation and its association with subcortical volumes: findings from the ENIGMA Epigenetics Working Group

2019 ◽  
Author(s):  
Tianye Jia ◽  
Congying Chu ◽  
Yun Liu ◽  
Jenny van Dongen ◽  
Evangelos Papastergios ◽  
...  
Keyword(s):  
Dna Methylation ◽  
Nucleus Accumbens ◽  
Large Scale ◽  
Biomarker Discovery ◽  
Association Studies ◽  
Meta Analysis ◽  
Hippocampal Volume ◽  
Small Sample ◽  
Small Sample Sizes ◽  
Meta Analyses

AbstractDNA methylation, which is modulated by both genetic factors and environmental exposures, may offer a unique opportunity to discover novel biomarkers of disease-related brain phenotypes, even when measured in other tissues than brain, such as blood. A few studies of small sample sizes have revealed associations between blood DNA methylation and neuropsychopathology, however, large-scale epigenome-wide association studies (EWAS) are needed to investigate the utility of DNA methylation profiling as a peripheral marker for the brain. Here, in an analysis of eleven international cohorts, totalling 3337 individuals, we report epigenome-wide meta-analyses of blood DNA methylation with volumes of the hippocampus, thalamus and nucleus accumbens (NAcc)—three subcortical regions selected for their associations with disease and heritability and volumetric variability. Analyses of individual CpGs revealed genome-wide significant associations with hippocampal volume at two loci. No significant associations were found for analyses of thalamus and nucleus accumbens volumes. Cluster-based analyses revealed additional differentially methylated regions (DMRs) associated with hippocampal volume. DNA methylation at these loci affected expression of proximal genes involved in learning and memory, stem cell maintenance and differentiation, fatty acid metabolism and type-2 diabetes. These DNA methylation marks, their interaction with genetic variants and their impact on gene expression offer new insights into the relationship between epigenetic variation and brain structure and may provide the basis for biomarker discovery in neurodegeneration and neuropsychiatric conditions.

scholarly journals Is Sensory Loss an Understudied Risk Factor for Frailty? A Systematic Review and Meta-analysis

2020 ◽  
Vol 75 (12) ◽  
pp. 2461-2470
Author(s):  
Benjamin Kye Jyn Tan ◽  
Ryan Eyn Kidd Man ◽  
Alfred Tau Liang Gan ◽  
Eva K Fenwick ◽  
Varshini Varadaraj ◽  
...  
Keyword(s):  
Systematic Review ◽  
Risk Factor ◽  
Longitudinal Studies ◽  
Meta Analysis ◽  
Sensory Loss ◽  
Cross Sectional ◽  
Sensory Impairments ◽  
Subgroup Analyses ◽  
Meta Regression ◽  
Meta Analyses

Abstract Background Age-related sensory loss and frailty are common conditions among older adults, but epidemiologic research on their possible links has been inconclusive. Clarifying this relationship is important because sensory loss may be a clinically relevant risk factor for frailty. Methods In this systematic review and meta-analysis, we searched 3 databases for observational studies investigating 4 sensory impairments—vision (VI), hearing (HI), smell (SI), and taste (TI)—and their relationships with frailty. We meta-analyzed the cross-sectional associations of VI/HI each with pre-frailty and frailty, investigated sources of heterogeneity using meta-regression and subgroup analyses, and assessed publication bias using Egger’s test. Results We included 17 cross-sectional and 7 longitudinal studies in our review (N = 34,085) from 766 records. Our cross-sectional meta-analyses found that HI and VI were, respectively, associated with 1.5- to 2-fold greater odds of pre-frailty and 2.5- to 3-fold greater odds of frailty. Our results remained largely unchanged after subgroup analyses and meta-regression, though the association between HI and pre-frailty was no longer significant in 2 subgroups which lacked sufficient studies. We did not detect publication bias. Longitudinal studies largely found positive associations between VI/HI and frailty progression from baseline robustness, though they were inconclusive about frailty progression from baseline pre-frailty. Sparse literature and heterogenous methods precluded meta-analyses and conclusions on the SI/TI–frailty relationships. Conclusions Our meta-analyses demonstrate significant cross-sectional associations between VI/HI with pre-frailty and frailty. Our review also highlights knowledge gaps on the directionality and modifiability of these relationships and the impact of SI/TI and multiple sensory impairments on frailty.

Some Problems with the Investigation of Noninferiority in Meta-analysis

2004 ◽  
Vol 43 (05) ◽  
pp. 470-474 ◽  
Author(s):  
N. Victor ◽  
S. Witte
Keyword(s):  
Meta Analysis ◽  
Sensitivity Analyses ◽  
Intention To Treat ◽  
Protocol Population ◽  
Noninferiority Trials ◽  
Interval Approach ◽  
Bivariate Method ◽  
Full Analysis ◽  
Meta Regression ◽  
Meta Analyses

Summary Objectives: Noninferiority trials have become commonplace in recent years. Like individual clinical trials, meta-analyses can also investigate noninferiority. However, certain important topics have to be considered. Methods: The proposed methods in this paper have their origin in the framework of noninferiority trials and meta-analyses. This paper can therefore be seen as a combination of both fields. Two issues are highlighted in the paper; difficulties in the choice of delta for a noninferiority meta-analysis leading to different deltas and methods for meta-analyses with different analysis sets, based on the full-analysis set with the intention-to-treat principle or the per-protocol population. Analytical methods, sensitivity analyses, meta-regression, and a bivariate method are introduced. The proposed graphical presentations support the analytical results. Conclusion: The confidence interval approach using meta-regression or bivariate methods is appropriate using both analysis sets for meta-analyses investigating noninferiority.

Sign in / Sign up

Export Citation Format

Share Document