(English) Through my career at PMDA, AMED, and the MHLW, I would like to look back on the efforts from the initial introduction of Electronic Study Data (CDISC) to the present, and also consider the goals that were aimed for at the time and the current situation. Furthermore, I would like to outline the future direction for promoting data standardization　etc. in our country where Healthcare DX (Digital Transformation) is promoting

(Japanese) PMDA、AMED、厚生労働省におけるこれまでの経歴を通じ、申請電子データ（CDISC）導入当初から現在までの取組みを振り返り、当時目指していたものと現状を考える。また、医療DXが進められている我が国において、データ標準化の推進など、今後の方向性について概説する。

TransCelerate has been conducting workshops with their members and CDISC to identify and prioritize use cases leveraging Digital Data Flow and USDM to enable AI. This session will provide an overview of those discussions and where companies have identified the most potential benefit. 

CDISC will provide an update on the 360 initiative including objectives and activities for 2025, work completed to date, and opportunities for the Japan community to engage with the project. 

CDISC will provide an overview of the CDISC Technology Landscape and how we are leveraging technology, models, and new methodologies to deliver the 360i connected standards and automated data flows. This will include an overview of USDM, BCs, Dataset-JSON, and analysis concepts, and how they fit together as well as how AI can be leveraged through the connection of the standards. 

CDISC will provide a demonstration of the work to date with 360i including leveraging the digital protocol within the USDM model linked to concepts to delivery machine readable specifications. This will be a live demonstration of the ability to use connected standards and the advantages it provides.

Background: Tohoku University Hospital Clinical Research Data Center has been developing CDISC standard implementation capabilities following their experience with in-house SDTM and ADaM implementation for regulatory submissions. Despite challenges in applying CDISC standards to existing operations due to diverse disease areas, the center has gradually implemented sustainable, non-personalized CDISC standard utilization by adopting CDASH-based templates provided by an EDC vendor.
 

Objective and Methods: To evaluate the feasibility and effectiveness of utilizing CDISC standards for central data monitoring report creation, two approaches were compared: (1) converting EDC data to SDTM and creating reports using BI tools, and (2) developing R programs utilizing EDC reporting functions. Data managers, statisticians, and medical informatics specialists promoting CDISC standards participated in this evaluation.
Results and Discussion: Findings of the comparison revealed that CDASH-based CRF standardization was sufficient for standardization and efficiency improvement within the center's current operational scope. Consequently, we have decided to strengthen CRF standardization using CDISC Terminology and implement reusable report output programs across multiple studies for the time being.
 

Conclusion: This approach, while fundamental, represents a highly effective strategy that can be implemented as part of existing EDC construction processes even with limited resources, leading to downstream process standardization. In addition, we will continue exploring opportunities for SDTM utilization within the overall clinical research process , while also continuing to evaluation and review of the reporting process.

Background: Recent activities through both CDISC (Clinical Data Interchange Standards Consortium) Open-Source Alliance (COSA) and PHUSE’s pharmaverse have led to the development of various R packages for CDISC standards implementation. While the pharmaceutical industry has begun to adopt these R packages, adoption in academia remains limited. 
Objective: We aimed to assess R packages recognized by COSA and pharmaverse to identify those that can be used in academia. We also aimed to share experiences using selected packages.
 

Methods: This research was conducted by two CDISC Japan User Group members. One was an industrial expert leading R utilization, who reviewed the functionality and development status of R packages based on public information and working experience to screen packages suitable for academia. The other was a researcher who had experience in CDISC standards implementation and reviewed the results considering academic situations. After their discussion, candidate packages were identified, and their overviews were summarized. Then, the researcher tested selected packages to assess their practical utility using the Study Data Tabulation Model (SDTM) and the Analysis Data Model (ADaM) data.
 

Results: We identified {admiral} for developing ADaM datasets and {teal} for interactive exploratory data analysis with Shiny web applications, and considered them currently available. We also identified {sdtm.oak} for developing SDTM datasets, though with limited functionality in its current version. We then confirmed that some of these packages could be useful in research.
 

Discussion: This collaborative research revealed that packages through COSA and pharmaverse could be utilized in academic settings. This utilization would contribute to minimizing costs for CDISC standards implementation and their statistical analysis, increasing the quality of deliverables, and enhancing both industry-academia and inter-academic collaborations. To promote broader adoption in academia, more publicly available documentation would be desirable. 

Despite Japan being one of the top three countries globally in terms of CDISC usage. However, efforts to create CDISC standards and develop new standards in Japan remain at the individual level, lacking organizational challenges. 
Therefore, the CDISC Japan User Group (CJUG) has embarked on a new initiative to develop a Therapeutic Area (TA) standard in the field of ophthalmology as a new step towards establishing CDISC standards from Japan. Ophthalmology is a field that requires the standardization of data related to visual function and eye diseases. However, international standardization has not yet progressed, and the development of standards is desirable to promote international data exchange and the advancement of clinical research.

This initiative, which officially started in January 2025, aims to expand the scope of CDISC standards and promote the development of data standards originating from Japan. To date, the project has established an overall development schedule and proceeded systematically and incrementally. In fact, the basic concepts and design for the ophthalmology TA have been constructed, and the examination of data collection items and mapping example have been made, steadily advancing towards TA development.

However, we have faced numerous challenges throughout this process. These include the specialization required for the target disease area, the difficulty of standardizing diverse data formats, and ensuring compatibility with existing CDISC standards. To overcome these challenges, we have strengthened our collaboration with experts and continuously explored solutions through ongoing discussions and trial and error.

This presentation aims to share the knowledge we have gained from our experiences so far, providing valuable guidance for those involved in future TA development. We hope it will contribute to the enhancement of CDISC standards and the advancement of international data standards.

Since the ICH M11 guideline reached Step 2 on September 27, 2022, the pharmaceutical industry has been abuzz with discussions about its implementation. The ICH M11 guideline promises to revolutionize the clinical trial landscape by modernizing the entire clinical trial life cycle through the use of structured study metadata. This modernization is expected to bring a plethora of benefits, including enhanced efficiency, improved data quality, and streamlined regulatory processes.
However, the implementation of ICH M11 is not without its challenges. The scope of its impact is vast, involving numerous stakeholders across the pharmaceutical industry. The initial stages of implementation will require significant investments in time, resources, and costs, as new systems and processes need to be developed and integrated.

In Japan, the Japan Pharmaceutical Manufacturers Association (JPMA) Pharmaceutical Evaluation Committee's Data Science Subcommittee Task Force 4 has been diligently investigating the implementation of ICH M11. The effort in 2024 have focused on understanding the potential changes and benefits that the full-scale implementation of ICH M11 will bring to pharmaceutical companies. We are also exploring the necessary steps that companies should take to prepare for this transformative shift.

As we look ahead, it is crucial for pharmaceutical companies to stay informed and proactive. Embracing the changes brought by ICH M11 will not only ensure compliance but also position companies to leverage the full spectrum of benefits, from improved trial efficiency to enhanced data integrity. The journey towards full implementation may be demanding, but the rewards promise to be substantial, paving the way for a more innovative and effective pharmaceutical industry.

Both ICH M11 and USDM (in the context of DDF: Digital Data Flow) have been frequently discussed in the industry since last year, but their differences are often not clearly articulated. For example, while M11 is expected to become a regulatory requirement in the near future, the implementation of USDM will remain optional due to differences in the granularity of information defined in the ICH M11 Technical Specification and USDM.

This presentation will compare the structure and controlled terminology of USDM and M11 Technical Specification, using examples to illustrate key distinctions. Additionally, the presenter will discuss the primary drivers behind both standards, positioning M11 as a regulatory requirement and USDM as a tool for business process optimization. Finally, the presenter will explore potential technical implementation approaches based on his experience deploying DDF solutions in the industry.

The CDISC Analysis Results Standard (ARS) provides a structured framework for integrating result data with metadata, promoting transparency, traceability, and consistency in clinical trial analysis. Traditionally, accessing and interpreting analysis results requires specialized statistical knowledge and programming expertise. However, advancements in AI technology present opportunities to make result data more accessible through human-like interactions, allowing broader audiences to engage with clinical findings.

This presentation explores how local Small Language Models (SLMs) can enhance the usability of ARS by enabling natural language queries to retrieve and interpret structured analysis results. By allowing clinical research professionals to interact with result data conversationally, SLMs can lower technical barriers, improve efficiency, and facilitate real-time insights without requiring extensive programming skills. However, deploying AI in regulated clinical environments raises important considerations, particularly regarding data privacy, security, model reliability, and the accuracy of generated responses.
A key advantage of local SLM is their ability to operate within secure, local PC environments, mitigating risks associated with cloud-based AI solutions. Unlike large cloud-based models, which require external data transmission, local SLMs process information entirely within a local PC and an organization’s infrastructure, ensuring compliance with data protection regulations and maintaining confidentiality of sensitive clinical data. Additionally, locally deployed SMLs offer better control over customization and model tuning to align with CDISC standards and specific analytical workflows. This approach reduces dependency on external providers and enhances long-term data governance strategies.

This session will examine the potential benefits and challenges of integrating SLMs with ARS, discussing key considerations such as model accuracy, validation requirements, regulatory acceptance, and practical implementation barriers. We will also highlight potential use cases where AI-driven exploration of result data can enhance decision-making in clinical research and regulatory submissions. By fostering discussion on the intersection of AI and standardized clinical data, we aim to provide insights into how SLMs can support secure, efficient, and user-friendly exploration of analysis results within the clinical data analytics workflow.

CDISC validation has evolved to ensure standardized data quality in clinical trials. CDISC Open Rules Engine (CORE), an open-source validation tool, enhances more chance of implementing and customizing conformance checks in our own systems. It centralizes executable validation rules and aligns with regulations, improving transparency and accessibility.

In Japan, interest in CORE is more and more growing and waiting for its adoption. The presentation highlights domestic discussions made to date as well as overview of CORE, introduces use cases of web-based or standalone application based on CORE, and provides future vision in pharmaceutical clinical development with CORE. Users expect CORE to evolve into a robust tool, with better rule coverage and usability. An active supports from ones who would benefit from CORE in the future is needed for continuous development and improvement. Collaboration, from an open-source standpoint, among stakeholders will shape its future, making it essential for clinical data validation.

The CDISC Conformance Rules Engine (CORE) Project is a significant initiative aimed at improving data quality and compliance in clinical trial. This presentation focuses on two key aspects: extending the CORE functionality to develop applications, and the process of rule editing for better conformance and usability.
Managing clinical trial data effectively and tailoring CORE to specific client needs can greatly improve efficiency. We will discuss how additional functionalities, such as user-friendly interfaces and automation, can be incorporated to streamline the validation process.

In addition, the session will cover rule editing within CORE, which is essential for maintaining accurate and up-to-date conformance rules. We will explain the process of modifying and refining rules to align with evolving CDISC standards, regulatory requirements and organisation's own management rules.
This includes considerations for defining rule logic, ensuring consistency, and validating changes before implementation. By enabling efficient rule customization, organizations can optimize their data validation workflows while ensuring alignment with global standards.

By the end of this presentation, attendees will gain insights into how the CORE Project can be extended and customized to meet organizational needs. They will also understand the significance of rule editing in maintaining high-quality, compliant datasets. This session is particularly valuable for professionals involved in clinical data management, regulatory compliance, and software development within the clinical research industry.