China
Africa
Explore chapters and articles related to this topic
Fixed Effects and Random Effects in Meta-Analysis
Published in Ding-Geng (Din) Chen, Karl E. Peace, Applied Meta-Analysis with R and Stata, 2021
Ding-Geng (Din) Chen, Karl E. Peace
As described in Wikipedia, “In statistics, a meta-analysis combines the results of several studies that address a set of related research hypotheses. This is normally done by identification of a common measure of effect size, which is modeled using a form of meta-regression. Resulting overall averages when controlling for study characteristics can be considered meta-effect sizes, which are more powerful estimates of the true effect size than those derived in a single study under a given single set of assumptions and conditions.” We thus begin introducing this effect size (ES) in Section 3.2.1.
Meta-Regression
Published in Christopher H. Schmid, Theo Stijnen, Ian R. White, Handbook of Meta-Analysis, 2020
Julian P.T. Higgins, Jose A. López-López, Ariel M. Aloe
Meta-regression analyses examine observational associations between effect size estimates and study-level covariates. Because the values of covariates are not randomized across studies, and typically do not follow processes comparable with randomization, causal inferences about the influence of study-level covariates on the effect sizes usually cannot be made with confidence. Associations observed may truly be due to the actual covariate included in the analysis, but frequently will be caused by confounding covariates that are correlated with both the covariate and the effect size. Nonetheless, the associations observed can sometimes be used to generate new hypotheses that can be tested in further primary studies (Baker et al., 2009).
Small Sample Meta-Analyses
Published in Rens van de Schoot, Milica Miočević, Small Sample Size Solutions, 2020
Researchers can account for between-studies differences by coding them as moderator variables, and controlling for their influence using meta-regression (Higgins & Thompson, 2004). Similar to classic regression, meta-regression posits that the outcome – in this case, the effect size of a study – is a function of the value of the moderators for that study. Both the fixed-effects and random-effects model can be extended to meta-regression. The advantage of coding between-studies differences as moderators, rather than using them as exclusion criteria, is that all studies can be included, as long as any differences are controlled for using meta-regression.
Meat Intake and the Risk of Bladder Cancer: A Systematic Review and Meta-Analysis of Observational Studies
Published in Nutrition and Cancer, 2023
Jinchuan Yu, Haigui Li, Zhengxiang Liu, Ting Wang, Fuding Zhou, Shaodi Ma, Baochun Chen, Wenjun Chen
Stata/SE15.1 and Revman5.3 software were used for data analysis. RRs, ORs, or HRs and their corresponding 95% CIs were extracted from each study for the meta-analysis. Given the differences in exposure categories in the study, we obtained a summary estimate by comparing the RRs of the highest with the lowest meat intake categories. The fixed-effects and random-effects models were adopted to pool RRs. Heterogeneity between studies was assessed using the Q and I2 statistics. For the I2 value, 25%, 50%, and 75% represented low, medium, and high levels of heterogeneity, respectively, while I2 >50% indicated that heterogeneity existed. Subgroup analyses were performed according to study design, region, publication year, and quality score to explore the sources of heterogeneity. Meta-regression was also adopted to explore the heterogeneity. Sensitivity analysis was performed to explore whether the inclusion of a study had a substantial impact on the results. A funnel chart was adopted to qualitatively evaluate publication bias. Egger’s test and Begg’s test were used to quantitatively evaluate publication bias, with P < 0.05 indicating statistical significance.
Efficacy and safety of PARP inhibitors in the treatment of BRCA-mutated breast cancer: an updated systematic review and meta-analysis of randomized controlled trials
Published in Expert Review of Clinical Pharmacology, 2023
Xiaoyu Sun, Suying Xu, Yiming Li, Xuemei Lv, Minjie Wei, Miao He
For dichotomous variables (ORR), the RR and 95% CI were calculated for each study. Analysis of event occurrence time variables using HR and 95% CI (PFS and OS). All data were expressed as the combination of HR or RR and 95%CI, and p <0.05 was statistically significant. We assessed the between-study heterogeneity by using the inconsistency index (I [2] statistic), which estimates the percentage of total variability across all studies [21]. I2 regarded an estimated value applied three fixed knots at 25%, 50% and 75% as an indicator of mild, moderate, and high heterogeneity. If the test showed I2 >50% or p <0.10, the data were calculated through a random‐effects (RE) model [22]. Otherwise, a fixed‐effects (FE) model was used to pool effect size [22]. To deeply explore the heterogeneity and its potential influence, subgroup analysis was performed. Meta-regression analysis was employed to examine which characteristics might be the possible source of heterogeneity. In addition to, publication bias was also estimated by Egger’s test and Begg’s test [23,24]. Sensitivity analysis, which examined the robustness of included trials to different aspects from methodological bias. All p-values were two-sided, and all statistical analyses were performed using Review Manager 5.3 and Stata 12.0 software.
Combined aerobic and strength training for fitness outcomes in heart failure: meta-analysis and meta-regression
Published in Disability and Rehabilitation, 2022
Geovana de Almeida Righi, Felipe Barreto Schuch, Tainara Tolves, Angélica Trevisan De Nardi, Natiele Camponogara Righi, Luis Ulisses Signori, Antônio Marcos Vargas da Silva
Increasing VO2 peak and quadriceps muscle strength may reduce mortality once both measures are independent predictors of mortality in people with HF [55,56]. Also, VO2 peak is directly associated with quality of life [55]. Besides that, improvement on quadriceps muscle strength is required. Sarcopenia is an independent predictor of 3-year mortality in elderly subjects [56] and it is prevalence 20% higher in HF patients compared with healthy controls with the same age [57]. Also, people with HF, similarly to people with chronic obstructive pulmonary disease and coronary heart disease, present muscle weakness which may influence the quality of life, exercise tolerance, and functional capacity [58]. The data demonstrate that CT is important for this population, once the same intervention will improve both outcomes. The limitations of our study are the low number of studies with low risk of bias, the lack of statistical power in the subgroup and the meta-regression analysis due to the low number of included studies (less than 10). Also the high number of cardiac rehabilitation protocols and the lack of consensus and detailed description of the exercise training made it difficult to characterize the moderators (in the subgroup analysis). It has been found a potential publication bias in the analysis of VO2 peak, CT versus control, but it was adjusted. Hence, we adjusted and corrected properly. Future studies should consider other relevant outcomes in HF, such as quality of life, hospital admissions, and mortality.