CDISC Enables Efficient Streamlining of Clinical Trial Safety Evaluation

by Geoffrey Mann, Thomas J Pedersen, Rebecca Lyzinski, Anisa Scott, Andrew J Foglia, John Cromer, Meichen Dong, Nora Varga, Sam Gardner, Christopher J Kirchberg, Byron A Wingerd, Russell D Wolfinger, & Wenjun Bao
JMP Statistical Discovery LLC



Wenjun Bao

Chief Scientist and Director of Advanced Analytics R&D

John Cromer

Principal Research Statistician Developer

Meichen Dong

Research Statistician Developer

Drew Foglia

Distinguished Software Developer

Sam Gardner

Senior Product Manager

Chris Kirchberg

Principal Systems Engineer

Rebecca Lyzinski

Senior Software Developer

Geoffrey Mann

Manager, Advanced Analytics R&D

Thomas Pedersen

Principal Technical Writer

Anisa Scott

Principal Staff Scientist

Nora Varga

Clinical Research Software Developer

Byron Wingerd

Principal Systems Engineer

Russ Wolfinger

Distinguished Research Fellow


The use of standardized data formats has increasingly facilitated both the submission of clinical trial data by pharmaceutical companies and its review by regulatory agencies. The Clinical Data Interchange Standards Consortium (CDISC) format has become the de facto worldwide standard for clinical trial data submissions and review. CDISC data are captured and organized in defined datasets, or domains, with predefined variable names and attributes. As a result, researchers and reviewers have a consistent, repeatable, and rapid way to load, review, and analyze data. The benefits provided by this standardization have culminated in the development of software that intuitively utilizes CDISC structures to quickly import and analyze the data as well as summarize the results for scientists and clinicians in a collection of readable and easy-to-understand reports.

In this paper, we will show how CDISC datasets with their required domains and respective variables can be utilized by JMP® Clinical to create clinical trial analysis reports easily and quickly with interactive graphs and tables for use in FDA New Drug Applications and Clinical Reviews. These reports are also ready for inclusion in regulatory submissions such as regulatory documents and clinical study reports to improve reviewability, interpretability, and efficiency of regulatory submission and review.


Clinical trials, in the context of this report, are experiments designed and performed to evaluate the safety and efficacy of putative drugs.1,2 Assessing and managing clinical trial data brings unique challenges that require input from multiple technical disciplines and organizations. Medical monitors, medical writers, medical reviewers, data managers and biostatisticians from sponsors, CROs and regulatory agencies must work together to ensure that the trial is successful, and that the data are analyzed and managed correctly. Historically, data have been recorded and presented in widely different formats and with definitions that significantly inhibited the exchange and reuse of clinical trial data. As such, lack of data interoperability and the inefficiency to which it contributes have been considered barriers in translating promising discoveries into health improvements.3 These same challenges in semantic and syntactic interoperability have complicated regulatory review, necessitating the development of solutions to improve clinical trial data analysis processes.

CDISC: Foundation for Quality and Standardization

CDISC4 was founded as a non-profit organization in 2000 and, in the years since, has worked to advance interoperability in clinical trials through open and consensus-based data standards.

Standardization of clinical trial data helps ensure the precision and quality of data, potentially leading to reduced time and expense needed for drug development. Kush et al.5 pointed out in the early stage of CDISC development that “Use of CDISC standards at project initiation can save 70 to 90% of time and resources spent prior to first patient enrolled and approximately 75% of the non-patient participation time during the Study Conduct and Analysis stages.” These efficiencies, combined with the increased ability to directly probe, reanalyze, and reuse submitted data, are the reasons why regulatory agencies around the world are beginning to require the submission of data in CDISC standard.6

JMP® Clinical: Clinical Trial Evaluation Based on CDISC Data

JMP® Clinical,7 a dedicated clinical trial software, was first released in 2010 in response to requests from the United States Food and Drug Administration (FDA) and the clinical trial industry. The CDISC standard makes software development much faster and increases the efficiency of the analysis procedure and results presentation because all the domains and variables share common definitions.

The CDISC standard not only modernizes clinical trial data submission, but also provides the foundation to streamline and standardize the clinical trial data collection, analysis, and management process. With well-defined variables, data formats, and domains in CDISC, both researchers and software systems know where, what, and how to utilize variables to assess the safety and efficacy of the drug more accurately, efficiently, and effectively. CDISC standardized data is a foundation for building intelligent software that supports the streamlined review of safety and efficacy clinical trial data. CDISC standardized data, when used in combination with other guidelines such as: International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) E3 guidelines,8 The Journal of the American Medical Association (JAMA) presentation rules for authors,9 RECIST guidelines,10 makes it possible for JMP® Clinical software, to summarize clinical trial data. This summary can include common demographic subgroups, protocol-defined subgroups and comparisons between groups for per-protocol and Intent To Treat (ITT) subsets of study subjects, without ad hoc programming, while retaining its unique interactive visualization capability. Modern software with interactive graphics and interactive filters allows researchers to dig deep into their results, to summarize information more quickly, and to more easily understand patterns and trends and their connection to drug safety easier.

Regulatory Agencies: Reassuring Standards for Enhancing Clinical Trial Reviewing

CDISC initially chose to use the SAS transport file as the storage format for their Submission Data Tabulation Model (SDTM, then called the Submission DataSet or SDS), Standard for the Exchange of Nonclinical Data (SEND), and Analysis Data Model (ADaM) as the use of SAS was nearly universal in the pharmaceutical industry and at the FDA, according to a short history of CDISC and SAS transport files on the CDISC website.11 At the same time, the FDA promoted the transition from paper to electronic data submission. FDA required submitted data to be in the SAS version 5 transport file format for each domain (e.g., demographics, adverse events), serving as a precursor to CDISC domains/datasets since 2004.12 The China National Medical Products Administration (NMPA)13 and the Japan Pharmaceuticals and Medical Devices Agency (PMDA)14 also required the SAS transport file as the format for data submission. Since 2016, the FDA15,16 and PMDA14,17 have required, and the NMPA18 and Europe European Medicines Agency (EMA)19 have recommended, CDISC format for clinical data submission.

In 2011, the FDA released a report with an evaluation of JMP®20 and JMP® Clinical as tools for FDA reviewers.21 This Assessment of the Impact of the Electronic Submission and Review Environment on the Efficiency and Effectiveness of the Review of Human Drugs – Final Report indicated JMP® 7.0 was already being used by the FDA at that time, and JMP® Clinical was listed as being under pilot testing at the Center for Drug Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER).21 In 2018, the FDA revised its procedures for CDER Medical Officer Conversion to Career-Conditional.22 This document lists the required training courses for reviewers to convert “from the medical officer temporary appointment of level 1, associate reviewer, to the career conditional appointment of level 2, reviewer”. The required training included the Medical Dictionary for Regulatory Activities (MedDRA)23 for medical terminology, CDISC for Data standards, and JMP® and JMP® Clinical training with multiple models for the standard analysis procedures. The PMDA also acknowledged that JMP® and JMP® Clinical were used in their review teams in 2015.24 Finally, the EMA announced in the regulatory session of CDISC Europe Interchange 2022 that they are evaluating JMP® Clinical as visualization software.25

Many reviewers and researchers prefer to analyze the data themselves, and this requires software that can perform the analysis and integrate with the terminology of the data standards. Manually reviewing the results can involve several tedious and repetitive steps, including summarizing the count and frequency of demographics, events, and interventions, calculation of risk differences across treatment groups, and examination of data over time for all participants and at the patient level. In addition, manual generation of Adverse Event (AE) narratives often requires significant effort on the part of the medical reviewers and writers. The CDISC data standard enables software, like JMP® Clinical and other tools to reduce the tedious and repetitive manual work and automatically generate the analysis results in a standardized presentation. The standardized presentation further improves efficiency, ensures quality, and provides for effective communication of the trial results and analyses.

In this paper, we discuss clinical trial summary information and follow the flow of FDA New Drug Application (NDA) submissions, Clinical Reviews (CR) and Biosimilar Multi-disciplinary Evaluation and Review (BMER) to reveal how the various domains of SDTM and ADaM are used to assess drug safety.

Note: You'll be redirected to The Journal of the Society for Clinical Data Management web site for download.


This article was originally published in the Journal of the Society for Clinical Data Management , Volume 3, Issue 1, Spring 2023.