Over the past three years, there has been a significant shift towards decentralised clinical trials (DCT), propelled by the COVID-19 restrictions. New research has revealed clinical trials with components of decentralisation are estimated to rise 17% by the end of 2023, surpassing the peak activity observed in 2021 during the pandemic. Mark Lambrecht, PhD, Director, EMEA & APAC Health and Life Sciences Practice at SAS explores.
DCTs are clinical trials that use remote monitoring or testing for at least one aspect of the study but typically are almost entirely remote.
Remote monitoring or testing methods include using wearables to monitor oxygen levels from home, blood samples being taken at local health centres, or reporting diary entries from mobile devices.
The benefits of DCTs are vast. They reduce financial and research facility space requirements and enable access to a broader demographic of people, including those with limited mobility. Unlike traditional studies, collecting results isn’t limited to when participants visit the research facility, meaning gathered information is much more representative.
However, DCTs also pose new logistical challenges for researchers. Monitoring geographically distributed participants, and collecting results from a range of devices means that data comes from all sides, and in many forms.
DCTs are only possible because of the internet of things, but the huge amount of data that is created would be impossible to analyse effectively without artificial intelligence (AI). Meaning many of the methods’ benefits would be outweighed due to overwhelmed administrative processes.
Leveraging the power of AI and cloud analytics has enabled researchers to streamline each trial stage, and unlock the full potential of DCTs.
A significant benefit of DCTs is patient centricity – by reducing the need for travel, more diverse groups are able to participate. For clinical trials, removing participation biases is a vital step to ensuring that medicine is equitable for all people, without bias.
Pre-trial is a vital stage for ensuring that the results that are gathered will be conclusive and representative of communities, but it is an administratively intensive step.
Here, AI can assess most application forms and instantly reject those who would not be suitable to go ahead with the trial. Automated systems do this much faster than manually possible and don’t require staff oversight.
In making first contact, telehealth communications enable small numbers of research staff to interview a large number of participants without travelling. Cutting down on time requirements for this process means those who require a more in-depth assessment before participating can be visited by staff.
Wearable technology, diaries, and visual data recording equipment can be sent out to participants and virtually explained, once again hugely reducing the required time spent by research staff. Similar to assessment processes, at this point those who require extra guidance can be met by patient management staff, instead of this work being done for every individual.
During this initial stage, AI can instantly detect those who may not be using the equipment correctly, or at all, and researchers can be alerted to investigate why and assist.
During the monitoring and data collection stages, AI and machine learning (ML) come into their own to tap into hidden efficiency in DCTs.
Tracking participants in small or large-scale trials is a resource-intensive task. And when those involved are spread geographically it can be easy for them to slip between the cracks.
Artificial intelligence can monitor data in near real-time, with much higher capacity than manual processes could allow. If a participant begins to drop off from results, or devices stop feeding in data, the system will detect it quickly and steps can be taken to find out what’s going on.
Signs of abnormal results are also logged and reported to medical professionals to review. Should there be a problem, data can be retroactively assessed, preventing harm and safeguarding participants.
AI and ML also support in approving remotely collected data by verifying the identity of the participant submitting results. Automating this stage speeds up the collection process and reduces the chance of invalid data, and is invaluable in studies where a variety of data is being assessed.
Leveraging the insights from past data, AI and ML can strike the balance between maximising the impact of each notification and overall involvement in the study, without becoming overly intrusive in the participant’s daily activities, which can be frustrating and lead to withdrawal.
As results are collected, analytics works alongside AI to enhance insights and dramatically improve the time to visualise and present results.
Throughout the trial, results may be gathered from an array of different sources, including those which are image-based, text scripts and health diagnostic data – such as blood pressure, blood sugar readings, or oxygen levels. Having this many data inputs typically adds complexity to collating the results.
AI is able to extract the key data from any format and normalise it so that analytics platforms can translate this normalised data into reports containing visual graphs and written explanations of trends, or a combination of the two.
Cloud analytics enable geographically distributed trials to take place, often more efficiently and effectively than their traditional counterparts.
Cloud-based storage acts as a central hub where all the data is brought together in a coherent and consistent way. By only working from a single data source conclusions are clear, explainable, and arrived at in much shorter timeframes.
With these tools, researchers and stakeholders can reach conclusions on a trial much sooner than in non-AI-enhanced trials. Analytics provides a portal for discovering deep insights and can pull up results instantly at any point in the trial.
As IoT technology advances, clinical trials will be able to become even less invasive while improving their impact. With their ease of use meaning more people than ever can get involved with DCTs, driving unprecedented medical progress. Continued innovation in life sciences will further increase the benefits of DCTs, opening the door for combined decentralised and local trials, or even more, yet-to-be-imagined, designs. As these trials become even more prominent AI, analytics, and ML will become cornerstones, driving efficiency at all stages of trials.
There’s also plenty of other medical editorial at IoT Insider’s sister publication, Electronic Specifier. And you can always add to the discussion at our comments section below or on our LinkedIn page here.