Clinical trial data analysis provider Beaconcure is looking to expand its global footprint. While the Israel-based company has staff in Poland and Canada, CEO Yoran Bar is planning to relocate to Boston in August to expand the company’s commercial coverage in the US.
Beaconcure’s primary asset is the stand-alone platform called Verify, which is designed to review clinical trial datasets and pinpoint inconsistencies and errors. At present, a notable chunk of the process of study data validation done by pharma companies and CROs is done manually, Bar notes.
Data inconsistence can lead to significant regulatory delays, and there is high-risk of inconsistencies as data reviews are managed manually, he adds. The clinical trial space is running behind in digitisation in this regard because of sensitivity of data, necessary regulatory requirements, and high security standards, he explains.
Clinical trial data analysis developed in partnership with Pfizer
Beaconcure started six years ago and as a start-up. Its first target was to develop a platform to validate medical research. However, this approach was not as fruitful due to limited access to data, Bar explains.
But then, in conversations with Pfizer, it suggested Beaconcure should pivot towards validating clinical trial data output. Pfizer provided the start-up initially with historical data to develop its technology centred around machine learning. By the time Covid-19 hit, Beaconcure’s platform was ready to the point that it helped Pfizer validate data from its Covid-19 vaccine clinical trials, Bar noted.
Clinical trial data in tables and lists is uploaded to Verify, which it then converted into a digital format to find data inconsistencies. For example, if a trial is designed to dose a participant with 1mg of the trial drug, but the data shows the participant received 6mg, this can be investigated for accuracy, Bar explains.
In general, a limitation of machine learning is that it has a very limited capacity to learn, Bar notes. This means that for the platform to expand its offering, more data, case studies, and scenarios need to be fed into it. While Verify can analyse numeric data in tables, accuracy surrounding free text is still being improved, Bar adds.