We have improved the Clinical Trial Risk Tool in the last 6 months, making it more user friendly and taking on board the feedback that we’ve received. We’ve improved the accuracy of the machine learning components too.
The tool now outputs its key figures such as risk levels and estimated cost in easily readable cards, so you can see at a glance the key takeaways from your protocol:
The risk factors are now organised into collapsible categories, so you can explore them easily without an information overload.
The tool identifies endpoints and inclusion and exclusion criteria. In future, we are hoping to use this to retrieve trials with comparable endpoints or criteria from the registries such as ClinicalTrials.gov.
Check your trial design
We have an easily digestible set of recommendations for the user to improve the trial, so you can see what are the high priority actions that you need to take with your protocol.
Currently we’re working on a feature to allow the user to generate an itemised budget for the trial in Excel from the schedule of events. We expect this to take a few months and to be finished in Q4 of 2025.
We are also working on making the tool output the probability of success for the trial. A number of groups have calculated aggregate statistics around trials proceeding from Phase 1->Phase 2, or Phase 3->approval, etc, e.g. MIT’s Project Alpha or Chufan Gao, Automatically Labeling Clinical Trial Outcomes, so these statistics or machine learning models could be integrated into the tool, and in addition to outputting the “risk score” it could also output the “probability of success” or similar, based on past trials.
We hope to also produce an accountable budget range (where should a reasonable bid fall), based on past trials.
Gao, Chufan, et al. Automatically Labeling Clinical Trial Outcomes: A Large-Scale Benchmark for Drug Development. arXiv preprint arXiv:2406.10292 (2024).
Shomesh Chaudhuri, Joonhyuk Cho, Andrew W. Lo, Manish Singh, and Chi Heem Wong, Debiasing Probability of Success Estimates for Clinical Trials (2022)
You may have been tasked with creating a clinical trial site budget. This is a budget itemising all the costs that will be incurred at the study site. The site budget may be needed To estimate the total cost of that site running part of the trial, as part of a bid to the CRO or sponsor To identify who needs to be reimbursed for each cost item To ascertain whether or not it is feasible to run the trial at that site To ensure that the site is reimbursed for the costs that they incur while running the trial If you are building a site budget, the most important document is the study protocol.
Fast Data Science are pleased to announce that the Clinical Trial Risk Tool, has been accepted as a supplier on the UK Government’s G-Cloud 15 framework. The G-Cloud 15 framework allows public sector bodies to buy cloud-based computing services such as AI, hosting, software and support directly without lengthy, costly traditional tender processes. What does the Clinical Trial Risk Tool do? The Clinical Trial Risk Tool helps users to analyse clinical trial protocols and documents.
Estimating the total cost of a clinical trial before it runs is challenging. Public data on past trial costs can be hard to come by, as many companies guard this information carefully. Trials in high income countries and low and middle income countries have very different costs. Upload your clinical trial protocol and create a cost benchmark with AI Protocol to cost benchmark The Clinical Trial Risk Tool uses AI and Natural Language Processing (NLP) to estimate the cost of a trial using the information contained in the clinical trial protocol.
