Change log
Version 2.6
This is the current version of the data package. Introduced on 23 September 2026 with the following changes:
- Allocation method was added to the study information csv and in the study-data.json to describe how participants were allocated. Options are: RCT, QRCT, NRSI, NOT_SPECIFIED.
- In the summary of findings pivot format, an outcome's study-type is set to NON_RANDOMIZED if its analysis includes any NRSI studies, otherwise RANDOMIZED.
- Analysis files now includes data for other data type analyses.
Version 2.5
Introduced on 25 August 2026 with the following changes:
- A meta-data file containing information about the review was added
- The data package for Overviews of reviews now contain included and excluded reviews
- in a folder called CDnumber-review-references
- files within the folder called CDnumber-other-references
- References awaiting classification are no longer included in the data package for any review type.
- Analysis ids are added to the summary of findings pivot file.
- Data package info file updated to only list files relevant to the review type
Version 2.4
Introduced 22 January 2026 with the following changes:
Structured data (.json) included for summary of findings data.
Version 2.3
Introduced 8th January 2026 with the following changes.
Structured data (.json) included for analysis data.
Structured data (.json) and tabular data (.csv) included for study categories: excluded, awaiting classification and ongoing.
The direction of outcomes is now included in the structured data for included studies.
The tabular data file Overall estimates and settings and subgroup estimates is updated with the following additions:
- Added the column 'Filter criteria'
- Added columns for forest plot settings
- Sort studies by
- Graph label (left)
- Graph label (right)
- Graph scale
- Show risk of bias
Changes to packages for Diagnostic Test Accuracy reviews were included from 22 January 2026.
Structured data (.json) included for analysis data.
Structured data (.json) and tabular data (.csv) included for study categories: excluded, awaiting classification and ongoing.
The tabular data file for parameters is updated to include paired data.
Version 2.2
Introduced on the 19 February 2025 with the following changes.
Overall estimates and settings and subgroup estimates:
- Added the column "Effect T" for the t-test statistic for pooled estimates when the HKSJ confidence interval is used.
Version 2.1
Introduced on 21 October 2024 with the following changes.
Overall estimates and settings:
- The column for confidence level is relabelled to CI/PI level
- Added columns to display if Tau2 CI or prediction interval are included in forest plot for the analysis:
- Tau2 CI
- Prediction interval
Version 2
Introduced on 24 September 2024 with the following changes.
Overall estimates and settings:
- CI fields ('Study CI' & 'Estimate CI') consolidated to one column 'CI level' (there is now only one setting in RevMan for confidence level and it is also used for the prediction interval)
- Added columns:
- Heterogeneity estimator
- CI method
- Tau² CI start
- Tau² CI end
- PI start
- PI end
Subgroup estimates:
- Added columns
- Tau² CI start
- Tau² CI end
Version 1
The initial version of the data package. Introduced on 25 April 2023.
RM5 files
Prior to introduction of the data package, data from Cochrane reviews were published as RevMan 5 (RM5) files. These legacy files can be converted here: https://revman.cochrane.org/rm5Converter.