scholarly journals dsSurvival: Privacy preserving survival models for federated individual patient meta-analysis in DataSHIELD

2022 ◽  
Author(s):  
soumya banerjee
Keyword(s):  
Statistical Power ◽  
Cox Regression ◽  
Meta Analysis ◽  
Privacy Preserving ◽  
Survival Models ◽  
Aggregated Data ◽  
Individual Level ◽  
Manual Analysis ◽  
Analysis Workflow ◽  
Interactive Modelling

Abstract Objective Achieving sufficient statistical power in a survival analysis usually requires large amounts of data from different sites. Sensitivity of individual-level data, ethical and practical considerations regarding data sharing across institutions could be a potential challenge for achieving this added power. Hence we implemented a federated meta-analysis approach of survival models in DataSHIELD, where only anonymous aggregated data are shared across institutions, while simultaneously allowing for exploratory, interactive modelling. In this case, meta-analysis techniques to combine analysis results from each site are a solution, but a manual analysis workflow hinders exploration. Thus, the aim is to provide a framework for performing meta-analysis of Cox regression models across institutions without manual analysis steps for the data providers. Results We introduce a package ( dsSurvival) which allows privacy preserving meta-analysis of survival models, including the calculation of hazard ratios. Our tool can be of great use in biomedical research where there is a need for building survival models and there are privacy concerns about sharing data.

scholarly journals dsSurvival: Privacy preserving survival models for federated individual patient meta-analysis in DataSHIELD

2022 ◽  
Author(s):  
Soumya Banerjee ◽  
Ghislain Sofack ◽  
Thodoris Papakonstantinou ◽  
Demetris Avraam ◽  
Paul Burton ◽  
...  
Keyword(s):  
Statistical Power ◽  
Cox Regression ◽  
Meta Analysis ◽  
Synthetic Data ◽  
Privacy Preserving ◽  
Survival Models ◽  
Aggregated Data ◽  
Individual Level ◽  
Manual Analysis ◽  
Interactive Modelling

Achieving sufficient statistical power in a survival analysis usually requires large amounts of data from different sites. Sensitivity of individual-level data, ethical and practical considerations regarding data sharing across institutions could be a potential challenge for achieving this added power. Hence we implemented a federated meta-analysis approach of survival models in DataSHIELD, where only anonymous aggregated data are shared across institutions, while simultaneously allowing for exploratory, interactive modelling. In this case, meta-analysis techniques to combine analysis results from each site are a solution, but a manual analysis workflow hinders exploration. Thus, the aim is to provide a framework for performing meta-analysis of Cox regression models across institutions without manual analysis steps for the data providers. We introduce a package (dsSurvival) which allows privacy preserving meta-analysis of survival models, including the calculation of hazard ratios. Our tool can be of great use in biomedical research where there is a need for building survival models and there are privacy concerns about sharing data. A tutorial in bookdown format with code, diagnostics, plots and synthetic data is available here: https://neelsoumya.github.io/dsSurvivalbookdown/ All code is available from the following repositories: https://github.com/neelsoumya/dsSurvivalClient/ https://github.com/neelsoumya/dsSurvival/

Abstract P255: Association of Mild Chronic Kidney Disease with Venous Thromboembolism: Pooled Analysis of Prospective General Population Cohorts

Circulation ◽  
2012 ◽  
Vol 125 (suppl_10) ◽  
Author(s):  
Bakhtawar K Mahmoodi ◽  
Ron T Gansevoort ◽  
Inger Anne Naess ◽  
Pamela L Lutsey ◽  
Sigrid K Braekkan ◽  
...  
Keyword(s):  
Chronic Kidney Disease ◽  
Venous Thromboembolism ◽  
Kidney Disease ◽  
General Population ◽  
Cox Regression ◽  
Meta Analysis ◽  
Random Effect ◽  
Pooled Analysis ◽  
Individual Level ◽  
Increased Risk

Background: Recent findings suggest that mild chronic kidney disease (CKD) might be associated with increased risk of venous thromboembolism (VTE). However, results were partially inconsistent, which may be due to lack of power. We therefore performed a meta-analysis to investigate the association between mild CKD and VTE incidence. Methods: A literature search was performed to retrieve community-based cohorts with information on the association of estimated glomerular filtration rate (eGFR) and albuminuria with VTE. Five cohorts were identified that were pooled on individual level. To obtain pooled hazard ratios (HRs) for VTE, linear spline models were fitted using Cox regression with shared-frailty. Models were adjusted for age, sex, hypertension, total cholesterol, smoking, diabetes, history of cardiovascular disease and body-mass index. Random-effect meta-analysis was used to obtain adjusted pooled HRs of VTE with CKD versus no CKD. Results: The analysis included 95,154 participants with 1,178 VTE cases and 599,453 person-years of follow-up. Risk of VTE increased continuously with lower eGFR and higher ACR (Figure). Compared with eGFR 100 mL/min/1.73m², pooled adjusted HRs for VTE were 1.3 (1.0–1.7) for eGFR 60, 1.8 (1.3–2.6) for 45 and 1.9 (1.2–2.9) for 30 mL/min/1.73m². Compared with albumin-creatinine ratio (ACR) 5 mg/g, pooled adjusted HRs for VTE were 1.3 (1.04–1.7) for ACR 30, 1.6 (1.1–2.4) for 300 and 1.9 (1.2–3.1) for 1000 mg/g. There was no evidence for interaction between eGFR and ACR (P=0.22). The pooled adjusted HR for CKD (eGFR <60 ml/min/1.73m² or albuminuria ≥30 mg/g) vs. no CKD was 1.5 (95%CI, 1.2–2.1). Results were similar for idiopathic and provoked VTE. Conclusion: Both reduced eGFR and elevated albuminuria are novel independent predictors of VTE in the general population.

scholarly journals Long-term trends in blood pressure and hypertension in Russia: an analysis of data from 14 health surveys conducted in 1975–2017

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Elena Churilova ◽  
Vladimir M. Shkolnikov ◽  
Svetlana A. Shalnova ◽  
Alexander V. Kudryavtsev ◽  
Sofia Malyutina ◽  
...  
Keyword(s):  
Blood Pressure ◽  
Meta Analysis ◽  
Population Based ◽  
Mean Blood Pressure ◽  
Elevated Blood Pressure ◽  
Elevated Blood ◽  
Aggregated Data ◽  
Individual Level ◽  
Pressure Lowering ◽  
The Usa

Abstract Background Hypertension is recognized as an important contributor to high cardiovascular mortality in Russia. A comprehensive analysis of data from Russian studies that measured blood pressure in population-based samples has not been previously undertaken. This study aims to identify trends and patterns in mean blood pressure and the prevalence of hypertension in Russia over the most recent 40 years. Methods We obtained anonymized individual records of blood pressure measurements from 14 surveys conducted in Russia in 1975–2017 relating to a total of 137,687 individuals. For comparative purposes we obtained equivalent data from 4 surveys in the USA and England for 23,864 individuals. A meta-regression on aggregated data adjusted for education was undertaken to estimate time trends in mean systolic and diastolic blood pressure, the prevalence of elevated blood pressure (> 140/90 mmHg), and hypertension (defined as elevated blood pressure and/or the use of blood pressure-lowering) medication. A meta-analysis of pooled individual-level data was used to assess male-female differences in blood pressure and hypertension. Results During the period 1975–2017 mean blood pressure, the prevalence of elevated blood pressure and hypertension remained stable among Russian men. Among Russian women, mean systolic blood pressure decreased at an annual rate of 0.25 mmHg (p < 0.1) at age 35–54 years and by 0.8 mmHg (p < 0.01) at ages 55 and over. The prevalence of elevated blood pressure also decreased by 0.8% per year (p < 0.01), but the prevalence of hypertension remained stable. Mean blood pressure and prevalence of hypertension were higher in Russia compared to the USA and England at all ages and for both sexes. Conclusions In contrast to the generally observed downward trend in elevated blood pressure in many other countries, levels in Russia have changed little over the past 40 years, although there are some positive trends among women. Improved strategies to bring down the high levels of mean blood pressure and hypertension in Russia compared to countries such as England and the USA are important to further reduce the high burden of CVD in Russia.

scholarly journals metaCCA: Summary statistics-based multivariate meta-analysis of genome-wide association studies using canonical correlation analysis

2015 ◽  
Author(s):  
Anna Cichonska ◽  
Juho Rousu ◽  
Pekka Marttinen ◽  
Antti J Kangas ◽  
Pasi Soininen ◽  
...  
Keyword(s):  
Correlation Analysis ◽  
Canonical Correlation Analysis ◽  
Canonical Correlation ◽  
Statistical Power ◽  
Association Studies ◽  
Meta Analysis ◽  
Original Data ◽  
Genome Wide Association Studies ◽  
Summary Statistics ◽  
Individual Level

A dominant approach to genetic association studies is to perform univariate tests between genotype-phenotype pairs. However, analysing related traits together increases statistical power, and certain complex associations become detectable only when several variants are tested jointly. Currently, modest sample sizes of individual cohorts and restricted availability of individual-level genotype-phenotype data across the cohorts limit conducting multivariate tests. We introduce metaCCA, a computational framework for summary statistics-based analysis of a single or multiple studies that allows multivariate representation of both genotype and phenotype. It extends the statistical technique of canonical correlation analysis to the setting where original individual-level records are not available, and employs a covariance shrinkage algorithm to achieve robustness. Multivariate meta-analysis of two Finnish studies of nuclear magnetic resonance metabolomics by metaCCA, using standard univariate output from the program SNPTEST, shows an excellent agreement with the pooled individual-level analysis of original data. Motivated by strong multivariate signals in the lipid genes tested, we envision that multivariate association testing using metaCCA has a great potential to provide novel insights from already published summary statistics from high-throughput phenotyping technologies.

scholarly journals A versatile toolkit for molecular QTL mapping and meta-analysis at scale

2020 ◽  
Author(s):  
Corbin Quick ◽  
Li Guan ◽  
Zilin Li ◽  
Xihao Li ◽  
Rounak Dey ◽  
...  
Keyword(s):  
Statistical Power ◽  
Meta Analysis ◽  
Joint Analysis ◽  
Genetic Associations ◽  
Functional Variation ◽  
Individual Level ◽  
Biological Variables ◽  
Level Data ◽  
Data Files ◽  
The Uk

AbstractMolecular QTLs (xQTLs) are widely studied to identify functional variation and possible mechanisms underlying genetic associations with diseases. Larger xQTL sample sizes are critical to help identify causal variants, improve predictive models, and increase power to detect rare associations. This will require scalable and accurate methods for analysis of tens of thousands of molecular traits in large cohorts, and/or from summary statistics in meta-analysis, both of which are currently lacking. We developed APEX (All-in-one Package for Efficient Xqtl analysis), an efficient toolkit for xQTL mapping and meta-analysis that provides (a) highly optimized linear mixed models to account for relatedness and shared variation across molecular traits; (b) rapid factor analysis to infer latent technical and biological variables from molecular trait data; (c) fast and accurate trait-level omnibus tests that incorporate prior functional weights to increase statistical power; and (d) compact summary data files for flexible and accurate joint analysis of multiple variants (e.g., joint/conditional regression or Bayesian finemapping) without individual-level data in meta-analysis. We applied the methods to data from three LCL eQTL studies and the UK Biobank. APEX is open source: https://corbinq.github.io/apex.

Power-Enhanced Funnel Plots for Meta-Analysis

2020 ◽  
Vol 228 (1) ◽  
pp. 43-49 ◽  
Author(s):  
Michael Kossmeier ◽  
Ulrich S. Tran ◽  
Martin Voracek
Keyword(s):  
Statistical Power ◽  
Meta Analysis ◽  
Relevant Information ◽  
Selective Reporting ◽  
Graphical Display ◽  
Funnel Plot ◽  
Evidential Value ◽  
Graphical Displays ◽  
Funnel Plots ◽  
Meta Analyses

Abstract. Currently, dedicated graphical displays to depict study-level statistical power in the context of meta-analysis are unavailable. Here, we introduce the sunset (power-enhanced) funnel plot to visualize this relevant information for assessing the credibility, or evidential value, of a set of studies. The sunset funnel plot highlights the statistical power of primary studies to detect an underlying true effect of interest in the well-known funnel display with color-coded power regions and a second power axis. This graphical display allows meta-analysts to incorporate power considerations into classic funnel plot assessments of small-study effects. Nominally significant, but low-powered, studies might be seen as less credible and as more likely being affected by selective reporting. We exemplify the application of the sunset funnel plot with two published meta-analyses from medicine and psychology. Software to create this variation of the funnel plot is provided via a tailored R function. In conclusion, the sunset (power-enhanced) funnel plot is a novel and useful graphical display to critically examine and to present study-level power in the context of meta-analysis.

Replication of the Superstition and Performance Study by

2014 ◽  
Vol 45 (3) ◽  
pp. 239-245 ◽  
Author(s):  
Robert J. Calin-Jageman ◽  
Tracy L. Caldwell
Keyword(s):  
Task Difficulty ◽  
Statistical Power ◽  
Meta Analysis ◽  
A Priori ◽  
Significant Heterogeneity ◽  
Performance Study ◽  
Improve Performance ◽  
Research Designs ◽  
Series Of Experiments ◽  
And Performance

A recent series of experiments suggests that fostering superstitions can substantially improve performance on a variety of motor and cognitive tasks ( Damisch, Stoberock, & Mussweiler, 2010 ). We conducted two high-powered and precise replications of one of these experiments, examining if telling participants they had a lucky golf ball could improve their performance on a 10-shot golf task relative to controls. We found that the effect of superstition on performance is elusive: Participants told they had a lucky ball performed almost identically to controls. Our failure to replicate the target study was not due to lack of impact, lack of statistical power, differences in task difficulty, nor differences in participant belief in luck. A meta-analysis indicates significant heterogeneity in the effect of superstition on performance. This could be due to an unknown moderator, but no effect was observed among the studies with the strongest research designs (e.g., high power, a priori sampling plan).

Statistical Power in Meta-Analysis

2006 ◽  
Author(s):  
Guy Cafri ◽  
Michael T. Brannick ◽  
Jeffrey Kromrey
Keyword(s):  
Statistical Power ◽  
Meta Analysis

scholarly journals C-factor: a summary measure for systemic arterial calcifications

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Lieke M. Kuiper ◽  
M. Kamran Ikram ◽  
Maryam Kavousi ◽  
Meike W. Vernooij ◽  
M. Arfan Ikram ◽  
...  
Keyword(s):  
Cox Regression ◽  
Intraclass Correlation ◽  
Principal Component ◽  
Arterial Calcification ◽  
Standard Deviation Increase ◽  
Single Measure ◽  
Individual Level ◽  
Enhanced Ct ◽  
Study Participants ◽  
Overall Mortality

Abstract Background Arterial calcification, the hallmark of arteriosclerosis, has a widespread distribution in the human body with only moderate correlation among sites. Hitherto, a single measure capturing the systemic burden of arterial calcification was lacking. In this paper, we propose the C-factor as an overall measure of calcification burden. Methods To quantify calcification in the coronary arteries, aortic arch, extra- and intracranial carotid arteries, and vertebrobasilar arteries, 2384 Rotterdam Study participants underwent cardiac and extra-cardiac non-enhanced CT. We performed principal component analyses on the calcification volumes of all twenty-six possible combinations of these vessel beds. Each analysis’ first principal component represents the C-factor. Subsequently, we determined the correlation between the C-factor derived from all vessel beds and the other C-factors with intraclass correlation coefficient (ICC) analyses. Finally, we examined the association of the C-factor and calcification in the separate vessel beds with cardiovascular, non-cardiovascular, and overall mortality using Cox–regression analyses. Results The ICCs ranged from 0.80 to 0.99. Larger calcification volumes and a higher C-factor were all individually associated with higher risk of cardiovascular, non-cardiovascular, and overall mortality. When included simultaneously in a model, the C-factor was still associated with all three mortality types (adjusted hazard ratio per standard deviation increase (HR) > 1.52), whereas associations of the separate vessel beds with mortality attenuated substantially (HR < 1.26). Conclusions The C-factor summarizes the systemic component of arterial calcification on an individual level and appears robust among different combinations of vessel beds. Importantly, when mutually adjusted, the C-factor retains its strength of association with mortality while the site-specific associations attenuate.

Sign in / Sign up

Export Citation Format

Share Document