21 Aug 2015 — Daan Reid
We are proud to announce the sixth preview version of our web-based redevelopment of ADDIS, ADDIS 2 release 6.
The majority of this release was devoted to expanding the network meta-analysis functionality of ADDIS.
We have also made this functionality available as a separate web-application at gemtc.drugis.org.
Release 6 contains the following new features:
- Support 3 types of meta-analysis (network, pairwise, node-splitting).
- Allow setting of model parameters (fixed effect vs. random effects, likelihood/link function, run-length parameters).
- More model results (convergence diagnostics, relative effects plots for network meta-analyses, study effect forest plot for pairwise meta-analyses).
The functionality developed in release 6 is showcased in this video.
The focus of the next release will be divided between three subjects:
- Expanding Trialverse with features to include others’ extractions, and laying groundwork for automating tasks.
- Further expanding the NMA functionality, allowing inter-model comparisons, the specification of prior distributions, and the exporting of figures and data so they can be used in reports.
- User testing to improve the usability and stability of the platform.
15 May 2015 — Daan Reid
We are proud to announce the fifth preview version of our web-based redevelopment of ADDIS, ADDIS 2 release 5.
The majority of this release was devoted to the Trialverse component, which is intended to be the portal where users extract and share clinical trials data.
With this release, we will freeze Trialverse development (besides bugfixes) for the near future.
Release 5 contains the following new features:
- Creation of complete studies (general information, design, results)
- Creation of dataset-level concepts (variables, drugs) and their mapping to study-level concepts
- Versioning of datasets (all changes are tracked, previous versions always visible)
- Updated the MCDA component to the latest version
The functionality developed in release 5 is showcased in this video.
The next release will focus on expanding the analysis functionality of ADDIS 2, particularly the network meta-analysis component.
05 May 2015 — Gert van Valkenhoef
We are proud to announce the release of a new version of our web-based Multiple Criteria Decision Analysis (MCDA) tool.
The majority of this release was devoted to polishing the user interface, and introducing a new way of eliciting Partial Value Functions.
We invite everyone to try it out at mcda.drugis.org.
The release contains the following features and bug fixes:
- Remove the MacBeth Partial Value Function elicitation, in favor of a method based on bisection
- Add clear warnings when information will be destroyed
- Add a legend to the heat maps
- Report units of measurement more consistently, and display unbounded intervals correctly
- Add Gravatar support (profile pictures)
- Considerably improve client-side performance
Thanks to Joël Kuiper and Douwe Postmus for their work on this release, and the IMI PROTECT project for their support.
26 Mar 2015 — Gert van Valkenhoef
The ADDIS 1.16.6 release corrects the variance set for continuous outcomes in single-study benefit-risk analyses. It was incorrectly based on the standard deviation of the outcome, rather than the standard error.
06 Feb 2015 — Daan Reid
We are proud to announce the fourth preview version of our web-based redevelopment of ADDIS, ADDIS 2 release 4.
The majority of this release was devoted to the Trialverse component, which is intended to be the portal where users extract and share clinical trials data. Trialverse is not finished yet, but we hope this release gives a good picture of the direction we’re taking. We invite everyone to try it out at trialverse.org.
Release 4 contains the following new features:
- Creation of datasets
- Creation of studies within a dataset
- Definition of study structure, including arms, epochs, and outcome measures.
The functionality developed in release 4 is showcased in this video.
The next release will focus on finishing the Trialverse portal so that complete studies can be entered into the system, and so that users can collaborate to improve previously-entered studies.
