How data analytics can increase the efficiency of clinical trials
Clinical trials are capital-intensive and complex with no guarantee of success. However, they're also of enormous value as continuous indicators of initial clinical validation. After that, there's continual testing of efficacy and tracking of progress with very experimental new drugs and treatment regimens which have reached an advanced stage of approval.
Clinical trials are becoming more expensive because of tighter regulations, the growing complexity of drugs themselves, the need to target them to smaller populations for personalized medicine, and a hike in salaries for clinicians. As a result, the number of trials is gradually decreasing yet worldwide health remains a continual challenge.
Big data and analytics can play a critical role in reducing costs and risks of failure so that pharma make the most of their clinical trials, sharing and reusing data produced to better assess opportunities.
R&D is a series of high-risk, high-investment decisions and the industry is facing a considerable productivity challenge in terms of identifying, testing, and bringing new drugs to market, especially in the context of the highly innovative therapies we seek today.
Pharmaceutical clinical trials and R&D create an astronomical amount of information to analyze. These are data lakes filled from IoT devices in the field which monitor participants, internal historical databases and external databases. Then, there's data reported directly from trial participants, or statistical data from research destined for analysis, in order to pull insightful conclusions from raw numbers. This data deluge can be overwhelming and identifying the right information at the right time becomes a key success factor.
The high costs of clinical trials are well known and shift almost daily. It takes approximately 10 to 15 years to bring a new drug from the laboratory to the pharmacy shelf. Over the three main phases, a single clinical trial costs an average of $1.1 billion[1] and there's only a 10% likelihood of success.
From 2005 to 2015 the number of the US National Institutes of Health-funded clinical trials registered annually fell from 1,580 to 930, a drop of more than 40 percent[2]. Research found that today's registered trials aren't any larger, and therefore, this represents an overall decline in funding.
According to the Digital R&D report by McKinsey, “Integrating internal company data (clinical trial management system, electronic data capture) with external data sets, including real world data and publicly available trial data, we can develop algorithms that are significantly more predictive of site-level performance. One client was able to improve enrollment rate 20 percent by applying these techniques[3].”
This is critical in today's world where some medical conditions have developed drug treatment immunity, and widespread viruses such as influenza are mutating at an accelerated rate, threatening a situation where clinical trials cannot keep up.
Decisions on whether to start and pursue a trial are based on cost and also the likelihood of the drug becoming profitable if it is found to be efficacious. If the market is limited and the cost required for development is high, such drugs may never make it to the market.
However, the optimization and digitalization of clinical trials can limit the investment, saving budget to do more trials, and helping the drug to hit the market quicker. Ultimately, this improves the health of patients more rapidly and increases the likelihood of profit for the pharma industry.
This is starting to happen in the market already. Many organizations are exploring new concepts driven by data that show a number of tantalizing possibilities.
An example here is a so-called site-less contract research organization (CRO). These organizations are designed to better manage costs for companies developing new medicines and drugs in niche markets. It enables pharma companies to outsource clinical trials to focus on the R&D.
This relies on technology to ensure that data is shared no matter where it originates, and volunteers can participate from wherever they live, saving a great deal of expense and improving the quality of the participant pool.
80% of trials fail to meet enrolment timeline[4]. New digital technologies plays a role in motivating the patient to enroll in a clinical trial, comply with the trial protocol and stay for the duration, so the trial accurately reflects the actions of the drug, device or procedure in a real-life situation.
This improves the speed and quality of trials, helping build the case to win the market stamp of approval. Ultimately, the sooner drug efficacy and safety are proven, the sooner the drug can start delivering results out in the market.
Randomized controlled trials (RCTs) are often described as the bedrock of research because they remove selection bias and randomize many elements of a trial that could bias the results.
But the increasing cost of these trials may make this bedrock crumble largely because the need for statistical validity calls on a large number of participants; it's the most thorough of research tools, but it's also the 'Rolls Royce' and costs the most.
"The emerging use of Big Data, including information from electronic health records and expanded patient registries, presents new opportunities to conduct large-scale studies with many of the benefits of RCTs but without the expense," says STAT, the global life sciences journal[5].
Use of big data can limit the need for RCTs by providing its own statistical evidence, and also by limiting the number of full RCTs that need to be conducted to achieve the best result.
[2] https://www.jhsph.edu/news/news-releases/2018/nih-funding-fewer-cinical-trials-study-suggests.html