AI in clinical research

On the 20th of May Paolo Morelli, CEO of CROS NT, joined the Scientific Board of Italian ePharma Day 2020 to discuss the growing role of the new technologies in clinical trials. We have taken this opportunity to talk to him about one of the most debated technologies of the last few years – Artificial Intelligence (AI) – to understand how it is currently used in the industry and what future it has.

There are so many discussions around AI right now. What are our biggest doubts related to it?

There are a number of questions that we are asking ourselves now. Here are just a few examples:

  • Will AI change and automate processes, in clinical research, making it more efficient?
  • Will researchers have to change the way they do their work?
  • Will pharma companies have to reorganize their R&D departments?
  • Will the approach to the training of clinical research personnel change?
  • Will patient privacy be protected?

For sure companies performing clinical trials are more and more interested in AI. They want to have data of higher quality, perform recruitment faster and in general, be more efficient. Also, when it comes to the financial side of the projects, the journey has just begun, and a lot is yet to come.

What does the notion of AI encompass?

I suggest we start with what is not AI. For example, automation is not AI. Automation is the use of machines that follow a precise program defined by Human Intelligence. The artificial intelligence (AI), sometimes called machine intelligence, is opposite to the Human intelligence (HI). We can define AI as any device that collects information and takes actions that maximize its chance of successfully achieving its goals.

Is AI already used in clinical trials?

Yes, I can outline two of the biggest areas where it is already in use. The first one is the use of AI over eSource/mSource data to support decision making (for ex: feasibility, communication with the patient) and the second one is the use of AI to reduce data entry and the burden of paper documentation (for ex. eConsent, Regulatory document completion, SDV).

How will AI use be expanded in the future? Where else we can apply AI in clinical trials?

For sure AI will allow a continuous communication with the patient, even though the human touch cannot be replaced by technology. I also believe that, at some point, SDV will be fully automated thanks to NLP technologies. With the proper training data set AI will be more successful in detecting data issues than Human Intelligence.

But even if patients and the industry are typically open to the innovative solutions, regulators will have a big role in the spread of the innovation, and can support it by setting up a modern regulatory framework.

Once AI becomes an integral part of the clinical trials industry, what will be the potential benefits for the patients?

The combination of AI with the scientific content will support a better communication with the patient: eConsent will allow a better understanding of the texts compared to the paper version of the Informed Consent. AI will answer questions that the subjects may have about the study, and the patient will be more informed about how his data is treated. I believe that we will see a bigger presence and influence of the patients in clinical research. It will be important, though, to overcome privacy challenges before introducing technology and informing the patients about the use of all their data.

Do you think AI will replace the human professionals in clinical trials of the future?

The omnipresence of overly abundant data (both clinical data and project data) can be successfully triaged and managed with the assistance of AI. And yet, it is HI (human intelligence) that is needed to make the final decision from the options offered. Think about our air traffic control system, highly automated, advanced technology, and yet there remains the need for HI to make the final decisions regarding all these simultaneous active flights (projects).

What are the aspects that can determine success of AI in clinical trials?

AI in Clinical ResearchThere are three main aspects that will determine the success of AI in Clinical Trials. Availability of the technology is not one of them.

One critical aspect is the availability of System Integrators as a bridge between technology and researchers. Their goal will be to understand clinical processes and to shape all this new technology around the clinical trial activities.

The second important aspect will be the ethics, morals, and governance around AI and ML. Regulators and Life Science companies will need to setup a framework before going too far down the road of AI.

And the last thing is data. AI needs data. Accurate data. Lots of it. We need to start collecting organised data and train AI before it can successfully achieve the desired result of making the clinical trials faster, more efficient, and ensuring higher quality.

At CROS NT, did you already start exploring such innovative technologies?

When I founded CROS NT and other PM Holding companies, I identified Innovation as an important value and the heart of everything that we do. Looking at the impact of innovative technology on the Life Science Industry, I believe that investing in this field was the right choice.

At CROS NT we are already supporting our clients with the projects that focus on patient digital data and are planning to partner with our clients to conduct the so-called virtual trials more and more frequently.

Would you like to learn how CROS NT can support you in your virtual trials? Contact us by clicking here.