From complexity to clarity: Harnessing the power of AI/ML and risk-informed strategies to streamline by eClinical Solutions

From complexity to clarity: Harnessing the power of AI/ML and risk-informed strategies to streamline clinical data management

Table of Contents 3 An increasingly complex data landscape ...................................................................................................................... 3 New challenges for data management ......................................................................................................................... 4 The importance of centralized data and efficient processes .......................................................................... 4 Data stewardship and clinical data science — an evolving responsibility .............................................. 5 Innovative AI and ML powered data management .............................................................................................. 6 Adopting a risk-based approach ........................................................................................................................................ 6 Innovative tools that power modern clinical trials ................................................................................................. 8 Enhance trial efficiency with the elluminate Clinical Data Cloud .............................................................. 9 Enable innovation across your data management team .............................................................................. 10 Conclusion ...................................................................................................................................................................................... 1 1 Introduction .........................................................................................................................................................................................

FROM COMPLEXITY TO CLARITY

Introduction In today’s fast-paced world, driven by demands for speed and efficiency, the field of clinical development has undergone a remarkable transformation. The way trials are being conducted has changed significantly with decentralized clinical trials (DCT) becoming mainstream and the collection of clinical data from wearables and other remote-monitoring devices becoming common practice. While these advances have made clinical studies more accessible and less burdensome to patients, the volume and variety of clinical data being captured per trial has grown exponentially — and at a faster pace than ever before. Managing this deluge of data has become particularly challenging. In response, roles across data management and clinical teams are evolving by adopting new approaches and technologies. In this eBook, we explore how artificial intelligence (AI), machine learning (ML), and risk-based methodologies can be leveraged by data management teams to meet the demands of modern clinical trials while ensuring data quality, increasing efficiency, and reducing cycle times.

An increasingly complex data landscape Over recent years, advances in clinical development have led to more complex clinical trials with resulting data becoming more complex too. DCTs and the advent of wearable technology have greatly amplified the volume of data generated. As a result, near real-time data collection, continuous patient monitoring, and the rapid gathering of valuable insights have become possible. Additionally, the COVID-19 pandemic has been an undeniable force in bringing sweeping changes to the way clinical research is conducted. Most significantly, it acted as a catalyst for moving away from inefficient practices and pushing new clinical trial

execution models to become standard practice. With remote and virtual monitoring models becoming mainstream, many organizations felt the urgency to expand risk-based approaches, further accelerating the initiatives set in motion by ICH E6 (R2 and R3). Furthermore, gone are the days when clinical trials solely focused on testing a single drug for one specific indication. Instead, we’re seeing a variety of trial designs, including adaptive, basket and umbrella trials, which allow for the investigation of multiple drugs and/or multiple indications within a single trial.

FROM COMPLEXITY TO CLARITY

New challenges for data management Collectively, these advances in trial design, trial execution, and data collection have contributed to the development of more robust trials and findings, however, the challenges for data management teams are increasing. ■ Volume of data – More complex trial designs, DCTs, and remote data capture technologies generate vastly more data. See Figure 1. ■ Velocity of data – Data is received faster than ever. For example, data from wearables can be uploaded automatically in real time. The challenge now is to make real-time analysis feasible. ■ Variety of data – As technology progresses, the types of patient data collected continues to grow and may routinely include imaging, biomarker, genomic, and electronic patient reported outcomes (ePRO) data – just to name a few. The collection of this data across various sources further drives the need for rapid standardization and harmonization of the data.

The importance of centralized data and efficient processes Despite the progress made in data generation, life science organizations vary greatly in how equipped they are to manage data efficiently and effectively. With vastly more data sources involved in today’s clinical trials, there is a risk that data can become siloed as traditional methods often require multiple systems for data review and downstream data delivery. Legacy manual processes make it incredibly difficult to detect missing information or to identify outliers and trends. There is a clear need for analytics-based modern data management that aligns people, processes and technology, alongside a centralized data system that facilitates the ingestion, standardization and harmonization of data.

Figure 1.

The scale of the data issue Traditional clinic-based trial

Trial using DCT components

260

=3.4

=>68

A manageable amount of data

An overwhelming amount of data — more than can be reviewed

subjects on a 6-month trial had their pulse measured and recorded every 2 weeks

thousand data points

subjects on a 6-month trial used wearable technology that transmitted data every minute

million data points

While the modern data landscape poses a lot of challenges, with the right infrastructure and processes in place, it also brings huge opportunities for generating valuable insights from trials and in a shorter timeframe, vastly increasing the potential for getting new drugs to patients faster.

FROM COMPLEXITY TO CLARITY

Data stewardship and clinical data science — an evolving responsibility As data becomes more abundant, complex, and digitized, the role of managing that data must evolve accordingly. Traditionally, data management focused on cleaning and querying listings of electronic data capture (EDC), and prior to that, paper forms. Now, it encompasses a much broader remit where teams could be considered data stewards, or clinical data scientists, responsible for the modern data strategy. Clinical data science, as defined by the Society for Clinical Data Management (SCDM), encompasses “domain, process, and technology expertise… data analytics skills and Good Clinical Data Management Practices.” 1 For example, with numerous external vendors involved in trials, data management professionals must manage all data work streams ensuring they are collected in a way that is going to meet the objectives of the protocol. As part of this ongoing evolution, data management teams are increasingly responsible for ensuring risk monitoring and management are completed accurately and efficiently. This crucial step includes monitoring and analytics tools to proactively identify trends and risks, helping to ensure patient safety and mitigate costly downstream errors or trial delays. The transformation of clinical data management towards data stewardship and clinical data science will enable the execution of more complex trials and data-driven approaches to drive faster, better decision making throughout the development lifecycle.

1 The Evolution of Clinical Data Management into Clinical Data Science” (An SCDM Position Paper on how to Create a Clinical Data Science Organization), Society for Clinical Data Management, September 2022, pg. 7 https://scdm.org/wp-caontent/uploads/2022/09/ SCDM-Position-Paper-Evolution-into-Clinical-to-Data-Science-V9.0.pdf

FROM COMPLEXITY TO CLARITY

Innovative AI and ML powered data management For years, AI and ML have been discussed as potential tools for improving efficiency and introducing new capabilities for a smarter world. Now, the promise has become a reality.

■ Artificial intelligence (AI) involves computer systems that can perform tasks that traditionally require human learning, reasoning, problem-solving, and decision-making skills.

■ Machine learning (ML) is a subset of AI that enables computers to learn and improve from data without explicit programming. Here machines can automatically analyze and interpret data, make predictions, and adjust their behavior based on experience. Today, these valuable technologies have moved beyond the hype with applications affecting many aspects of our everyday lives. In clinical development, the potential impact of AI/ML is vast. Trial design, patient recruitment, and data management and analysis

can all benefit. For data management teams, the integration of AI/ ML can address the limitations of traditional approaches that are simply not scalable enough to support the direction that clinical trials are headed in. There are many opportunities for using AI/ML in data management and clinical data review. One example of this use is for data quality review in which ML modules provide an initial analysis by examining extensive data from multiple perspectives and presenting it in a simplified format for humans to understand. Consequently, only a small percentage of the data requires manual review. Additionally, ML algorithms can evaluate data across multiple dimensions and flag anything with a concern with accompanying reasons. As a result, the data manager can focus solely on the dimensions that impact that specific data point. This effectively reduces complexity and better equips teams to manage the pace and volume of incoming data.

Adopting a risk-based approach Historically, data management has involved examining 100% of the data, meaning equal weight is being put on every data point. As we’ve explored, this is unsustainable as the volume of data grows to billions of data points per trial. AI/ML will contribute to more efficient handling of this proliferation. However, in tandem, adopting a risk-based approach will proactively mitigate risk to ensure a trial is conducted safely and efficiently. With a risk-based approach that includes a comprehensive risk assessment, data

management teams can determine which data are most critical and impart a monitoring strategy that provides the most value throughout the trial lifecycle. While risk-based monitoring (RBM) has been part of FDA guidance since 2013, it has since evolved into more expansive approaches, such as risk-based quality management (RBQM) and risk-based data management (RBDM). See Figure 2.

FROM COMPLEXITY TO CLARITY

Figure 2.

Risk-based approaches Risk-based monitoring (RBM) ■ Limited to monitoring activities ■ Identifies risks related to data collection ■ Reactive approach that addresses risks as they are identified/managed by the monitoring team

Risk-based quality management (RBQM) ■ Broad cross-functional approach across the trial lifecycle ■ Uses early, targeted risk assessment to identify critical data/processes where attention needs to be focused. Identification of these aids choice of monitoring strategy that provices the most value ■ Holistic approach to initial and ongoing risk assessment ■ Proactive identification of risks before they become issues ■ Added to GCP via ICG E6(R2) in 2018

Risk-based data management (RBDM) ■ Applies principles of risk-based monitoring to clinical data management to avoid reviewing all data manually ■ AI/ML can be used to run central statistical monitoring models on the non-critical data to aid identification of outliers or trends ■ Focus on the critical data and let automation manage the rest flagging to you any areas for review

Data is collected in a highquality and consistent manner, while minimizing the risk of errors or inconsistencies.

Use spans from protocol design to data analysis and reporting. Ensures that the trial is conducted safely, efficiently, and with the highest possible quality standards.

Offers improved data quality by focusing data review on the critical data that matters. Can be used to streamline and automate processes for lower risk data.

Benefits of risk-based strategies Reduce risk and cost n Improve data quality n Improve operational efficiency Fully scalable, futureproof processes n Compliant with industry guidance

FROM COMPLEXITY TO CLARITY

Innovative tools that power modern clinical trials The modern clinical data landscape brings both significant opportunities and challenges for today’s data management, driving the need for more advanced technologies to manage these challenges. Rather than working in an environment of disparate clinical data and fragmented technology, a future-forward clinical data and analytics infrastructure provides a centralized location where all data can be collected and reviewed in near real-time. Advanced clinical data and analytics platforms facilitate rapid data integration, with embedded capabilities such as AI/ML, RBQM, and advanced analytics that enable smarter, faster business decisions. There are numerous benefits to leveraging such technologies in your clinical trials.

Centralized clinical data platforms

AI-enabled data review

RBQM

Democratize your data

Ensure data quality

Proactively reduce risk

■ Optimize your data flow by eliminating silos

■ Data quality is improved—at scale—using

■ Implementing a continuous risk assessment

and multiple systems for data review and

anomaly detection models to identify

to identify, document, plan, and mitigate

downstream data delivery. The eradication

anomalous subjects or atypical values

study risks through an integrated RBQM

of silos and real-time access to data across

across all data sources. Issues are surfaced

solution enables a monitoring strategy

clinical teams facilitates cross-functional

that wouldn’t be detected with traditional

designed to optimize resources.

collaboration and more efficient, data-driven

approaches.

decision-making.

Optimize time to value ■ Existing APIs (Application Programming

Reduce efforts ■ Automate clinical review objectives to detect anomalous data, reducing the amount

Ensure patient safety ■ Detect eligibility issues, safety, and medical incongruencies early on by leveraging RBQM with analytics and embedded AI/ML.

Interfaces) and bi-directional interoperability

of manual review required and enabling

with other systems allow seamless

reviewers to focus on the most critical data

Achieve data synergy

integration of new data sources for more

domains and tasks.

■ When an RBQM solution is integrated within

efficient data review and cleaning. Builtin automation, standardization and

Decrease cycle times

interoperability allow your data to work for

■ The time from data entry to issue

you, right from the start.

Increase productivity ■ A unified end-user experience with

detection, and data entry to data cleaned

a centralized platform, analytics silos are removed and the need to access multiple applications for high-quality data is eliminated.

is significantly reduced. Because issues are surfaced that would not be detected in a traditional/manual approach, costly and

integrated data review capabilities enables

time-consuming downstream errors are

faster decision-making with easy-to-use

prevented.

analytics, workflow, and issue management — all in one place, increasing the speed and efficiency of your trial.

FROM COMPLEXITY TO CLARITY

Enhance trial efficiency with the elluminate Clinical Data Cloud® elluminate Clinical Data Cloud is a flexible clinical data platform that fits within existing clinical technology ecosystems. It is designed for rapid data integration, providing users with access to all trial data from a single centralized location, along with advanced analytics that enable smarter, faster business decisions.

DATA INGESTION, MAPPING & STORAGE

INTELLIGENT APPLICATIONS

DATA CONSUMERS Data Management

DATA SOURCES Genomics Labs EDC

EHR

Data Central

eSource

MDR eCOA

Claims

Clinical Analytics

Mapper

Clinical Operations

CTMS IVRS

Operational Insights

Importer

FORMATS RBQM

Excel CSV

Medical Monitors

Metadata & Standards

Statistical Computing Environment

MDR CDR ODR

elluminate IQ Review (eIQ Review) provides AI-enabled data review capabilities to support data stewards in ensuring data integrity in a more efficient, scalable way. Embedded within Data Central, eIQ Review surfaces issues that wouldn’t be identified in a traditional review approach, and manual efforts are reduced, freeing data management teams to focus on critical data. As a result, data quality is improved, productivity across teams is increased, and cycle times are reduced.

Biostatistics Clinical Programmers Executive Management

*AI-Enabled with elluminate IQ

elluminate RBQM provides clinical teams with a single platform for all study data sources to support a more focused, data-driven monitoring effort that enables active oversight and improves the compliance of a clinical trial. elluminate RBQM proactively identifies potential issues the moment they arise, facilitates cross-functional collaboration between teams, and ensures study data quality and integrity from start to finish.

FROM COMPLEXITY TO CLARITY

Enable innovation across your data management team With all these new developments, creating new data workflows and investing in new technologies can seem overwhelming, despite the clear benefits. While implementing an end-to-end process may be suitable for some, it’s also an option to introduce processes and tools that are at a suitable scale for your organization’s needs and available resources. It may be appropriate to outsource your data management work to a specialist team equipped with the necessary tools and expertise for your ongoing trials. By leveraging their knowledge and capabilities, you can ensure efficient data management and make the most of their technological resources.

How to be an advocate for innovation The growing challenge of navigating today’s clinical data complexity is clear, but how can you help drive innovation and gain support from within your organization? 1.

Be informed – Remain open to learning and keep abreast of industry challenges and advancements.

2. Be engaged – Identify your challenges and pain points and effectively communicate them to the teams developing innovative tools. Also, communicate within your organization to build support for change. 3. Be a skilled communicator – Effective communication is crucial for securing institutional buy-in and managing change effectively. 4. Be an early adopter – Gain efficiencies sooner. Additionally, your feedback and recommendations will actively influence further product development. 5. Be open and adaptable – Innovative tools are designed to create new, more efficient workflows, and by working together with your technology provider, can be aligned as closely as possible with your current practices.

eClinical Solutions’ Biometrics Services combine exceptional expertise with best-in-class technologies and processes. This integration allows for the delivery of high-quality clinical data solutions, spanning from data acquisition to insights. Empowered by advanced technology, including the elluminate platform, our expert biometrics strategists enable sponsors to access their most critical data quickly while ensuring standardization, efficiency, and streamlined oversight throughout the trial lifecycle. By partnering with eClinical Solutions’ Biometrics Services, users can confidently navigate the challenges of data proliferation, harness the power of innovative technology, and unlock valuable insights.

FROM COMPLEXITY TO CLARITY

Conclusion

Next steps

The trends discussed here will continue to shape the clinical development landscape, and the field of clinical data management will continue to transform as trials become increasingly complex and data intensive. Regardless of how much the industry changes, ensuring study integrity and patient safety will remain core priorities in clinical research. Leveraging innovative technologies and approaches will not only ensure that these objectives are achieved but will also provide opportunities to streamline processes and increase productivity across data management teams. With the elluminate Clinical Data Cloud and Biometrics Services experts, life sciences organizations can begin to maximize the value of their clinical data and reduce trial cycle times in order to help get treatments to patients, faster.

For more information on how eClinical Solutions can empower your data management team to efficiently navigate the changing data landscape and generate valuable data insights, visit https://www.eclinicalsol.com/ Or explore our range of eBooks and webinars, including: Webinar: Overcoming Barriers to RBQM Adoption Webinar: Introduction to elluminate IQ eBook: The New Clinical Biometrics Blueprint

FROM COMPLEXITY TO CLARITY

About eClinical Solutions eClinical Solutions helps life sciences organizations around the world accelerate clinical development initiatives with expert biometrics services and the elluminate Clinical Data Cloud® – the foundation of digital trials. Together, the elluminate platform and eClinical Biometrics Services give clients self-service access to all their data from one centralized location plus advanced analytics that help them make smarter, faster business decisions. Leveraged by 100+ biopharmaceutical companies, elluminate is a purpose-built clinical data cloud designed to accelerate digitization efforts that proactively manage the exponential growth of clinical data volume, variety, and velocity. eClinical’s Biometrics Services combine top talent with bestin-class technologies, rigorous processes and the power of elluminate, to deliver efficient, high-quality clinical data solutions fit for the new era of digital trials. From acquisition to submission, we enable sponsors to execute an endto-end data strategy that embeds standardization and streamlines oversight. Whether partnering with us for studybased biometrics services, functional service outsourcing, or specialized clinical data consultancy, sponsors of all sizes rely on eClinical Solutions for agile, expert support with the right skills and technology for today’s speed-driven, complex trials.

eclinicalsol.com

info@eclinicalsol.com

T: 877-355-8668 (877-ELLUMN8)