Clinical trials are critical for discovering new treatments for diseases and new ways to detect, diagnose, and reduce the risk of developing them. However, these trials can be notoriously expensive, dependent on myriad factors, such as disease type, the methods used, and the location of the trial. This holds true for malaria clinical trials.
In 2004, global funding for malaria R&D was a mere $323 million. This relatively low investment fosters an urgency to optimise the funding – a task for which a detailed understanding of the various costs involved in clinical trials can be instrumental. But how do we estimate the cost of a malaria clinical trial?
Understanding the cost necessitates untangling the factors contributing to it. The cost of a trial is majorly driven by the number of interventions involved, the types of these interventions, the number of subjects, the locations of the trials, and the number of sites and visits. More expensive interventions, such as gene and cell therapies, noticeably inflate the cost of the trial.
The stage of the trial also exercises a significant impact on the cost. Late-stage trials often involve larger patient cohorts, driving the cost up. The average cost of phase 3 trials, for instance, is reported to be around $30 million.
The geographical location of the trial site introduces another variable in the mix. High-income countries tend to be more expensive for trials. Meanwhile, countries like China and India allow trials to happen at a cost that is 30-40% lower than in western countries like the United States.
Fast Data Science’s Clinical Trial Risk Tool utilises the power of machine learning to predict the cost of a clinical trial from the protocol text. The tool works by analysing the schedule of events in the protocol and applying sophisticated algorithms that have learned from a vast corpus of historical data.
The ML model can predict which interventions are likely to be needed and how many, factoring in available information about the type of the disease, the stage of the trial, and other relevant details. This allows it to output a projected cost for the trial.
The Risk Tool also provides insights into how modifications in the trial design, such as changes in the number of subjects or site locations, could impact costs. These predictive capabilities are invaluable not only for budgeting purposes but also for strategic planning of clinical trials.
By predicting the cost of a malaria clinical trial through the protocol, we increase our capacity for strategic use of limited resources and catalyse the development of novel treatments and malaria control tools—an ongoing mission in areas where drug resistance has decreased the efficacy of commonly used treatments. The world of clinical trials is complex, expensive, and often unpredictable. Machine learning, however, is giving us a new way to understand and manage these challenges.
Fast Data Science’s Clinical Trial Risk Tool is a promising solution for those engaged in critical research and development activities, particularly for diseases like malaria that urgently need increased investment and resource optimisation. In such a setting, having a tool that can effectively predict trial costs from the protocol is more than a convenience; it’s a game changer.
