The
Part 1
Part 1
scale
automation for clinical programming
statistical analyses
Digital Statistical Computing Environment (SCE) Playbook for Clinical Development
Increasing
and
and
2 What’s Inside Introduction .................................................................................................................................. 3 What is a Statistical Computing Environment (SCE)? ................................. 4 Historical Approaches to Statistical Computing ................................................ 5 SCE Adoption in Today’s Market .................................................................................. 6 Taking a Modern Approach ............................................................................................. 8 Leveraging an Integrated SCE to Streamline Digital Trials ....................... 9 elluminate Statistical Computing Environment .......................................... 10
Introduction
As innovation in clinical development continues to accelerate and decentralized trial (DCT) models become more widely adopted, the volume of clinical trial data from a variety of sources continues to proliferate, creating both opportunities and challenges for life sciences organizations. While more advanced trial modalities such as DCT make trials more accessible and relieves patient burden, the volume and variety of study data collected adds to the complexity of managing clinical data workflows throughout the trial lifecycle. More complex multi-arm trials used for oncology products can also require enormous data and content stores, challenging the capabilities of legacy clinical data stores. Leveraging a modern technology solution is one way sponsors can get in front of these new clinical development challenges. This is the first in our series of e-books to highlight how life sciences organizations can utilize an integrated statistical computing environment to streamline the production of submission deliverables while increasing programming and analysis efficiencies.
SCE PLAYBOOK FOR CLINICAL DEVELOPMENT 3
What is a Statistical Computing Environment (SCE)?
A Statistical Computing Environment (SCE) is a set of tools for computational processing of clinical data that provides a foundation for demonstrating rigor — which requires transparency, reproducibility, and adequate documentation — in the analysis and reporting of clinical trial results.
Computational Processing of Clinical Data
There are a variety of factors driving the need for an advanced computing environment within life sciences organizations. As computational complexity continues to increase — driven by factors including larger data sets, more sophisticated methods and the emergence of data science techniques — a high-performance computing environment is critical to support output generation. Moreover, a modern statistical computing environment provides the flexibility required to support highly specialized analytic needs and language-agnostic approaches. Additionally, because many organizations support global cross-functional teams, cloud-based solutions are becoming table stakes to ensure the availability of, and access to, real-time clinical data.
4
COLLECTION SUBMISSION ANALYSIS TRANSFORM REVIEW AGGREGATION DATA CONVERSION REPORT GENERATION DATASET MAPPING STATISTICAL METHODS OUTPUT CREATION
Historical Approaches to Statistical Computing
For organizations that do not yet have a statistical computing environment, managing statistical analyses is a manual, disjointed, and highly regimented SOP-driven process for clinical programmers, statistical programmers, data scientists and statisticians. Historically, electronic content-driven approaches for statistical analysis relied upon a dedicated server folder structure, controlled by a network administrator. These approaches require manual sign-off, validation of programs in disparate systems, and often resulted in difficult, time-consuming audits and cumbersome rerun processes.
The first SCEs were able to remove part of the administrative burden. Access control, validation status and version management were very helpful, although advanced features such as programming run order, control of data source changes, and linkage to metadata sources and change history were not yet available in these systems.
In more recent years, large sponsors and CROs have implemented customized SAS-based tools to solve the challenges associated with statistical analysis. While these bespoke solutions can improve analysis and programming workflows, they have been heavily process-driven with long implementation timelines and do not have the flexibility to support additional programming languages — such as R & Python — that are continuing to gain traction among data scientists and statisticians.
SCE PLAYBOOK FOR CLINICAL DEVELOPMENT 5
Related Resource: 2022 Gartner® Market Guide for Life Science E-Clinical Platforms Download
SCE Adoption in Today’s Market
In a recent poll , 40% of respondents from mid-market and large sponsors have either purchased an SCE, are in the process of purchasing an SCE, or are building their own validated system. These results indicate that the value of a statistical computing environment is resonating within the market. As DCT and digital trials continue to evolve with larger and more complex data sets, we anticipate these dynamics shifting in the near future. Given the rapidly changing landscape, it seems likely that the 52% of organizations using standalone SAS and/or R will start moving towards purchasing an SCE in order to reduce administrative burden and keep up with the digital evolution of moder clinical trials.
an End-to-End Clinical Data Workflow with a Platform Approach
Automating
Watch Now
SCE PLAYBOOK FOR CLINICAL DEVELOPMENT 7 2022 eClinical Solutions Industry Poll: How sponsors are approaching statistical computing needs1 What’s your organization’s current solution for statistical computing? 8% Outsource tasks or have no internal statistical computing 52% Don’t have an SCE but have standalone SAS and/or R 16% Have either purchased or are purchasing a commercial SCE 24% Have built or are building their own validated SCE
Taking a Modern Approach
In order to modernize their approaches, organizations have recently been turning to consortia like Transcelerate Biopharma, in addition to stalwarts like CDISC. Transcelerate’s Modernization of Statistical Analytics Framework 2 aims to increase the use of modern statistical analysis technologies within the industry, enabling the faster delivery of innovative treatments to patients. Within this framework, three core principles for regulatory compliance are identified as follows: 3
Accuracy
Reproducibility
Traceability
The measure of correctness of software libraries that are used to generate results from a modern statistical analytics environment
The ability to recreate statistical analysis outputs from the original dataset along with the associated environment, including all artifacts and dependencies
■ Full version control and access management ■ Development lifecycle workflows ■ Customizable folder structure within and accross studies ■ Library of reusable code snippets, macros, and functions ■ Intelligent editor with auto-complete and automation of common taskes Execution ■ Integrated compute environments for SAS, R, and Python ■ Support for package/ library validation ■ Interactive real-time results or scheduled batch execution ■ Log checking and display of results ■ Full audit trail and reproducibility of results
■ Dependency tracking for both data and inherited code ■ Automated execution of programs based on data refresh ■ Impact analysis and notification of data structural changes ■ Auto-creation of programs based on configurable templates ■ Data flow visualization from source output 1 Leveraging a Statistical Computing Environment with the key capabilities and functionality outlined below will ensure your organization is in line with digital
Modern
can make sure that these three principles of regulatory compliance will be met, while increasing development efficiency and streamlining program execution. 8 2 3
The ability to trace inputs to outputs, providing evidence to connect e-data, code, and environment to the final output that is produced Program Development
Automation
approaches.
SCEs
Leveraging an Integrated SCE to Streamline Digital Trials
Demonstrating rigor in the analysis and reporting of clinical trial results requires transparency, reproducibility, and adequate documentation. With the increasing amount of clinical data available, more sophisticated methods, and an emergence of data science techniques, a high-performance computing environment is necessary in order to produce accurate and consistent results in support of regulatory submissions.
Automate & Accelerate Statistical Analyses with an Integrated Statistical Computing Environment Watch Now
Breaking silos with integrated workflows and interoperability
DATAREPOSITORY CLINICALDATA
AREPO T I O NAL
METADATA REPOSITORY
REP O S I YROT
Automation and standardization at scale to unlock value
More companies are considering leveraging an e-clinical platform as they optimize and modernize their analytics and infrastructure strategy. Gartner defines an e-clinical platform as “an integrated suite of technologies connected to a platform architecture that provides services and solutions to manage clinical trial planning and execution”. 4 Utilizing a Statistical Computing Environment within an e-clinical platform that provides access to metadata, clinical data, and operational data can provide even greater benefits for programmers and statisticians. With access to these key elements, data, standards, and mappings can be leveraged for maximum reuse and increased programming and analysis efficiencies.
SCE PLAYBOOK FOR CLINICAL DEVELOPMENT 9
Combined CDR, ODR and MDR enables use cases across the clinical development value chain
elluminate Statistical Computing Environment (SCE)
elluminate SCE provides a validated environment for demonstrating rigor in statistical analysis and reporting.
cloud computing environments, elluminate SCE provides one centralized place for all your data, metadata, programs, and results — increasing the efficiency, traceability, and visibility of your statistical analyses. As a component of elluminate, the statistical computing environment enhances the value of the platform approach by leveraging data, standards, and mappings to maximize reuse and increase programming and analyses efficiencies. DATA INGESTION, MAPPING & STORAGE INTELLIGENT APPLICATIONS DATA SOURCES FORMATS DATA CONSUMERS Data Central Analytics Operational Insights RBQM Mapper Importer Metadata & Standards Data Management Medical Monitors Clinical Operations Biostatistics Executive Management Genomics Labs CTMS EHR eCOA Claims EDC IVRS Excel CSV eSource One platform for clinical and operational data from ingestion to insights -across all data sources MDR Statistical Computing Environment CDR ODR MDR 10
Clinical
production
submission
analysis outputs
auditable,
eluminate SCE in action Watch Now
The
Unlike other
Fully integrated with the elluminate
Data Cloud, elluminate SCE enables the
of
or exploratory
— within and across studies — in a way that is transparent, reproducible, efficient,
automated, and secure.
About eClinical Solutions
eClinical Solutions’ products and services are helping life sciences companies adopt a more modern data infrastructure to break down data silos, automate processes and accelerate the speed of drug development. With the elluminate Clinical Data Cloud, life sciences organizations can begin to maximize the value of their clinical data, help their teams accelerate development processes, and serve patients more efficiently throughout the clinical trial lifecycle.
For more information or to request a demonstration of elluminate SCE, visit us at eclinicalsol.com/products/statistical-computing-environment/ or email us at info@eclinicalsol.com
Footnotes
elluminate SCE Fact Sheet Download
1 eClinical Solutions industry poll with 25 respondents from large and mid-size sponsors in February 2021 to gain illustrative insights.
2 Transcelerate, Modernization of Statistical Analytics Framework, June 2021 https://www.transceleratebiopharmainc.com/ initiatives/modernization-statistical-analytics/
3 Transcelerate, Modernization of Statistical Analytics Framework Infographic, June 2021 https://www.transceleratebiopharmainc. com/wp-content/uploads/2021/04/TransCelerate_MSA-Infographic_April-2021-1.pdf
4 Gartner, Market Guide for Life Science E-Clinical Platforms, 26 April 2022, By Jeff Smith
SCE PLAYBOOK FOR CLINICAL DEVELOPMENT 11
©2022 eClinical Solutions LLC. All rights reserved. elluminate® is a registered service mark of eClinical Solutions LLC. eclinicalsol.com info@eclinicalsol.com T: 877-355-8668 (877-ELLUMN8)