Interview: The major changes AI will bring to the clinical landsacpe in 2024

By Liza Laws

- Last updated on GMT

© Getty Images
© Getty Images

Related tags AI Artificial intelligence Clinical trials Data management Patient centricity

Over the last year, we’ve seen how rapid technology adoption, namely AI, can impact clinical trials.

As we head into 2024, Raj Indupuri, CEO at eClinical Solutions expects we’ll see major changes to the clinical trial landscape. We put some questions to him, to find out exactly where he thinks this part of the industry is headed.

How has rapid technology adoption, particularly AI, impacted clinical trials over the last year?

In 2023, we witnessed the rapid adoption of innovative technologies in life sciences; the biggest one being artificial intelligence (AI), which has proven to play a critical role in supporting modern data management. The volume and variety of clinical trial data increased significantly in recent years, with the average phase 3 trial now generating over three million data points. Even the most comprehensive data review strategies leave room for issues, leading to costly and time-consuming errors and necessitating the evolution of traditional data management approaches. AI-enabled data review capabilities for data managers and clinical data reviewers have helped to ensure data integrity in a more efficient, scalable way.

Other benefits of using AI in this way include increasing productivity – when many clinical teams face increased pressure to do more with fewer resources – and optimizing time to value. As the clinical data ecosystem evolves, detecting anomalous data becomes more complex and time-consuming. AI has become an excellent tool for optimizing data review processes and decreasing cycle times.

AI applications can also support patient recruitment (identifying potential trial patients for more targeted recruitment efforts), data analytics for decision-making (deriving insights from massive amounts of clinical trial data generated to help inform sponsors’ and researchers’ decisions), remote data collection and monitoring (enabling real-time data collection while patients participate remotely), as well as risk prediction and mitigation (proactively addressing challenges and improving the overall success rate of trials).

As AI continues to mature and companies define and refine their AI use cases, more adoption and benefits will be realized over time. Rapid tech advancements, macro trends, and the current industry state are putting more pressure on clinical development to be agile, fail fast, and be technology-first and data-driven. AI will play a critical role, and companies will need to adopt or be left behind.

What major changes do you anticipate will happen for clinical trials next year?

As we head into 2024, in addition to the continued acceleration of technology advancements, I expect to see changes in how digital tools are adopted for the conduct of modern trials.

As part of our continued life sciences digital transformation, I anticipate a move towards EDC-free digital trials. The potential of electronic health record (EHR) data directly integrating with other data acquisition applications and data infrastructure would phase out the need for EDC, enabling improved patient experiences and alleviating burdens on research sites.

Simultaneously, I see our industry moving beyond the hype surrounding DCTs. We’re recognizing that DCT components are really part of the natural evolution towards digitalization of trials, rather than their own separate segment.

Can you explain the concept of EDC-free digital trials and how EMR/EHR might play a crucial role in this shift?

When EDC was introduced, now over 20 years ago, the promise of EDC was to digitize and create better experiences. The intent was seamless acquisition and faster access to data for efficient analysis and improved outcomes. Industry adopted, and EDC did eliminate labor-intensive paper-based processes. It reduced errors associated with paper processes, making it better for data handling and data quality. However, EDC created other processes and overheads at the same time, adding new burdens for sites. The pain points were not completely solved. We ended up moving paper-based processes to electronic, web-based systems, but more inefficiencies impacted sites. The cost, resources and time for EDC processes are expensive and didn’t live up to the efficiency promise.

For the last decade plus our industry has been trying to solve that challenge. Meanwhile, eSource and eCOA have been increasing as data sources direct from patients, including external data, biomarkers, genomics, and more. The data environment is much more complex than when EDC was envisioned. The majority of data today comes from external data sources and not from EDC, which was never intended to handle such data diversity. Instead, we can bring those external data sources directly into a modern data infrastructure to eliminate silos and have data in a central source of truth for more rapid analysis and streamlined data handling, despite the increasing complexity. The reduction in data coming from EDC also lessens that double data entry workload for sites.

The shift towards direct integration of EHR data has the potential to bring meaningful and long-awaited efficiencies for clinical research. Because of recent advances in technology, we’re at a point where machine learning and AI can also be used to extract data and connect different systems, leveraging the underlying modern data infrastructure capable of ingesting, storing, and analyzing the complex data streams of modern trials. Bringing in EHR data and minimizing the dependency on EDC will help the industry realize the original streamlined vision of data handling while also tackling the mounting data challenges created by the industry’s rapid innovation.

I expect we’ll see EDC-free trials within 1-2 years, unleashing the potential for groundbreaking efficiency improvements through enhanced data sharing and reduced fragmentation. The past five years of innovation contributed to increased fragmentation of digital tools, but tech innovation has set the foundation for a new era characterized by streamlined processes, reduced friction, and the collaboration power of AI. Seamlessly connected data streams are essential to scale and power tomorrow’s breakthroughs in modern digital trials.

What challenges or obstacles might be associated with the transition to EDC-free digital trials, and how can these be addressed?

This is a pivotal time for our industry, and addressing the challenges that come with the transformation to EDC-free digital trials will be necessary to realize the desired benefits. Obstacles may include:

  • Data standardization – ​Diverse data formats and standards across different EMR/EHR systems may hinder seamless integration, not to mention that integrating disparate data sources and systems can be complex, time-consuming, and potentially pose interoperability challenges.
  • Data quality and accuracy​ – The ability to access vast, high-quality data in real-time for high performance of AI/ML models is imperative.
  • Infrastructure costs​ – Upgrading or establishing new digital trial infrastructure involves upfront costs and investing in the future.
  • Change management ​– Change management can be unintentionally minimized during new technology implementation, but having a strategy promotes successful adoption. This includes leadership buy-in, building champions, communication and upskilling. Focus on how experiences will change and the problems that will be solved. Change management is an ongoing effort that gains momentum as anticipated value is achieved.
  • Regulatory compliance – ​Adhering to evolving regulatory requirements for digital data in clinical trials and for AI will be paramount. Data security and privacy will continue to be top priorities.
  • Workforce knowledge and skills​ – Having the proper training can ensure more effective processes, as long as those working with these tools are knowledgeable in digital and AI technology.
  • ‘Human in the Loop’ strategy ​– While AI may alleviate a lot of the workload, human oversight is still imperative to ensure accuracy. Training and building trust in AI models is key.

Addressing these challenges requires collaboration from stakeholders across healthcare and research. Initiatives such as data sharing for AI applications and intelligence, regulatory changes and data standards can enable our industry to embrace technological advancement. The transition to EDC-free digital trials will pave the way for more efficient, patient-centric, and data-driven clinical research.

You mention there will be a move beyond the hype of decentralized clinical trials (DCTs) and that DCTs were just a natural evolution towards digital trials. Why do you suggest that it does not need its own category?

DCTs have been a trending industry topic, gaining momentum during the pandemic. While a lot was learned during this period of DCT hype, the term was unnecessarily a buzzword. Whether you call them DCTs, virtual trials or patient-centric trials, we have been talking about this for the last decade. The objective is to bring research to patients and collect data directly from them, whether that is data collection through apps or devices or reaching patients directly with mobile health units and telehealth. Then we started talking about this shift toward DCTs, and there has been a lot of hype around the term. In reality, this was bound to happen. The approaches categorized as DCTs were really part of the natural progression toward digital trials.

The modern clinical trial will connect and collaborate with patients in new ways. Biopharma companies are recognizing that DCT is not a separate initiative or strategy, and we’re seeing companies with recently created DCT teams dismantling or moving these DCT teams into their core clinical trial teams. Many of the beneficial aspects of DCTs we saw accelerate in recent years will continue and will accelerate further, but it will be as part of a modern clinical trial rather than a separate type of DCT trial implementation.

Patient-centric, tech-driven approaches must also factor in what patients and sites want from a digital trial. Hybrid models still prevail, and the industry is finding the right balance. The movement to modern trials is not only about data quality, cost and efficiency – but about taking advantage of digital technologies when and where it promotes our shared industry goals for patient-focused, inclusive research that reduces, not adds to, the demands on sites and patients.

Only then can we call modern trials successful.

Related news

Show more