My watch list  

ODM Data Analysis—A tool for the automatic validation, monitoring and generation of generic descriptive statistics of patient data

by Tobias Johannes Brix, Philipp Bruland, Saad Sarfraz, Jan Ernsting, Philipp Neuhaus, Michael Storck, Justin Doods, Sonja Ständer, Martin Dugas


A required step for presenting results of clinical studies is the declaration of participants demographic and baseline characteristics as claimed by the FDAAA 801. The common workflow to accomplish this task is to export the clinical data from the used electronic data capture system and import it into statistical software like SAS software or IBM SPSS. This software requires trained users, who have to implement the analysis individually for each item. These expenditures may become an obstacle for small studies. Objective of this work is to design, implement and evaluate an open source application, called ODM Data Analysis, for the semi-automatic analysis of clinical study data.


The system requires clinical data in the CDISC Operational Data Model format. After uploading the file, its syntax and data type conformity of the collected data is validated. The completeness of the study data is determined and basic statistics, including illustrative charts for each item, are generated. Datasets from four clinical studies have been used to evaluate the application’s performance and functionality.


The system is implemented as an open source web application (available at and also provided as Docker image which enables an easy distribution and installation on local systems. Study data is only stored in the application as long as the calculations are performed which is compliant with data protection endeavors. Analysis times are below half an hour, even for larger studies with over 6000 subjects.


Medical experts have ensured the usefulness of this application to grant an overview of their collected study data for monitoring purposes and to generate descriptive statistics without further user interaction. The semi-automatic analysis has its limitations and cannot replace the complex analysis of statisticians, but it can be used as a starting point for their examination and reporting.

Authors:   Tobias Johannes Brix; Philipp Bruland; Saad Sarfraz; Jan Ernsting; Philipp Neuhaus; Michael Storck; Justin Doods; Sonja Ständer; Martin Dugas
Journal:   PLoS ONE
Volume:   13
edition:   6
Year:   2018
Pages:   e0199242
DOI:   10.1371/journal.pone.0199242
Publication date:   22-Jun-2018
Facts, background information, dossiers
More about Public Library of Science
Your browser is not current. Microsoft Internet Explorer 6.0 does not support some functions on Chemie.DE