Getting to the point

With analysts predicting a surge in business spending on AI to reach $300 billion by 2026, it is becoming increasingly clear that the transformative power of AI will also have major implications for the future of healthcare and the wider pharmaceutical industry.

There is huge potential in the ability of AI to help streamline and quicken the processes involved in the running of clinical trials.

From trial design to monitoring, the potential here for better ecosystem integration cannot be overstated. General efficacy, efficiency and designing bespoke fit for purpose trials will all be greatly improved thanks to AI.

Before we delve into the wider parameters that can be reimagined by AI, we must consider how each stage of the trial process is impacted by its introduction. First, we have trial design, centred around the process of planning.

Trial design aims to ensure that the trial is conducted in a scientifically sound manner and that the data generated is reliable and credible.

It is also important to note that each trial will have very different determining factors.

For example, one trial’s priority will be speed to market, such as the sense of urgency around COVID-19 vaccines when faced with a global pandemic, while another may need to run to a slower time frame due to budget and resource constraints.

Ultimately, all trials will have, front and centre, the same high standards of patient safety and efficacy.

Subsequently, a variety of digital data platforms work in unison with key personnel to assist with the first stage of creating a trial.

Historically, this part of the value chain takes up a significant amount of time. It is here that the trial designers will decide on the protocol for the entire study.

It is at this point that the best location, what country or region, will be identified for the trial, as well as an assessment of the correct demographics for the trial to include.

For a ten-year trial, the design process can take up to six months. Tools powered by AI can help analyse the trial data, as well as provide easy access to past studies, and help with predictions of the best ways to run a trial based on past data, all with the goal of better streamlining these processes.

Trial design is notorious for its fragmented data and disconnected systems. Inputs for trial artifacts are often scattered across dozens of systems and formats.

Therefore, having technologies that can help structure, standardise, and digitise data elements from a range of inputs and sources is crucial.

For example, chief information officers can implement tools to automate data management across the trial life cycle.

These tools intelligently interpret data elements, feed downstream systems, and auto-populate required reports and analyses.

These tools can utilise existing systems to seamlessly integrate the data flow – providing a single, collaborative touchpoint for all interactions during a clinical trial known as a single source of truth (SSOT).

The centralisation of data within a SSOT is a critical step to ensuring AI can provide accurate, reliable data interpretation. Ultimately, AI is only as good as the data we provide it with – without strong data management and governance AI will fall short of its capabilities.

The application of AI enables the identification of potential next best actions, especially in the face of unexpected challenges.

This will lead to the early acquisition of more accurate data sets and the ability to facilitate more efficient trial monitoring.

There are several ways in which the use of AI can aid the streamlining of the planning and execution process for clinical trials.

AI can provide insights that will aid patient recruitment, as well as providing important insights into a range of other variables. The managing and monitoring of drug dosage is immensely important if any success is to be achieved through the trial.

Having live visibility of all sites and their inventory gives trial coordinators and clinical supply managers greater logical understanding of the locations they manage, equalling better integration through this one accessible data point – the single source of truth.

In this data-intensive phase of the trial, it is crucial to concentrate on consolidating and centralising the collected data.

Innovative tools that streamline data collection can significantly reduce the dependence on in-person trial sites, creating a more effective and less intrusive process.

The primary goal here is to bring all the data together in one place. Subsequently, AI becomes instrumental, offering enhanced insights derived from the consolidated data.

It is important to emphasise that while gathering data is a crucial first step, the real benefit of AI lies in its ability to uncover meaningful insights during study execution and in understanding patient outcomes.

Outcomes can be predicted by AI both during the trial and also during the planning stages of future trials.

The global news organisation Reuters has reported how companies such as Amgen, Bayer and Novartis are training AI to scan billions of public health records, prescription data, medical insurance claims and their internal data to find trial patients – in some cases halving the time it takes to sign them up.

In the article, German drugmaker Bayer explained it used AI to cut the number of participants needed by several thousand for a late-stage trial for Asundexian, an experimental drug designed to reduce the long-term risk of strokes in adults.

Ramji Vasudevan is Senior Engineering Leader at Altimetrik.com