This study was a meta-epidemiologic review of the reporting quality of Mendelian randomisation studies which use individual-level UK Biobank data. The dataset consists of 116 articles rated as "Yes", "Partially", "No" or "Unclear" on 25 data-extraction items based on the STROBE-MR reporting guidelines. The articles were searched for on 15/07/2022.

Ethics approval was not needed for this study, and because the data were collected from reviewing publicly available articles anonymisation is not needed. The protocol for the study which was pre-registered at https://doi.org/10.17605/OSF.IO/NWRDJ.

This repository contains 5 folders and 13 files:

The 'Analysis' folder contains 2 file:

MRReportingReview_Code_R(v4.1.0)_v1.0_20221123.txt
MRReportingReview_CodeDependencies_20221123.txt

This is an Rscript to conduct data analysis and the environment/dependency information.

The 'Data' folder contains 2 files:

MRReportingReview_Data_v1.0_20221123.csv
MRReportingReview_ShortItemNames_v1.0_20221123.csv

These are the data and shortened item names to replace the full item names in the bar chart.

The 'Protocol' folder contains 1 files:

MRReportingReview_Protocol_v1.0_20221123.rtf

The 'Results' folder contains 5 files:

MRReportingReview_SecondReviewerMeans_v1.0_20221123.csv
MRReportingReview_ItemPercentages_v1.0_20221123.csv
MRReportingReview_RegressionResults_v1.0_20221123.csv
MRReportingReview_PercentageBarchart_v1.0_20221123.jpg
MRReportingReview_RegressionForestPlot_v1.0_20221123.jpg

These files contain the mean article completeness (%) for each second reviewer, the number and percentage of articles with each response for each data-extraction item, the regressions results of article characteristics with overall article completeness (%) and the figures which depict the latter two tables graphically.

The 'Supplemental' folder contains 3 files:

MRReportingReview_DataExtractionItems_v1.0_20221123.csv
MRReportingReview_ExcludedArticles_v1.0_20221123.csv
MRReportingReview_SearchStrategy_v1.0_20221123.csv
These files contain information on the data-extraction items, articles that were excluded at full text screening and the search strategy used.

All the files within this dataset are stored in TXT, CSV, RTF and JPG formats, and can be opened with a number of software packages.

This repository is also stored publicly as an interactive CodeOcean capsule at: 

https://doi.org/10.24433/CO.4457049.v4

The study was accepted for publication in the journal BMJ Evidence-Based Medicine under the title 'Reporting and methodological quality of studies that use Mendelian randomisation in UK Biobank: a meta-epidemiologic study'. The article is publicly available at: 

https://doi.org/10.1101/2022.04.25.22274252