New Collaboration Aims to Use Advanced Data Analysis to Speed Trials

February 4, 2019

A Michigan CRO is teaming with MIT’s Computer Science and Artificial Intelligence Laboratory and the University of Maryland’s Center for Transitional Medicine (CTM) to create what organizers are calling “a supergroup” that will help speed drugs through the clinical trials pipeline.

Officials at MMS Holdings of Canton, Mich., MIT and Maryland’s CTM call their new venture the Health Analytics Collective. Its goal is to help use data — particularly real-world data — to help cut down on the time spent in trials, to augment label claims and to support new drug applications.

The group wants to merge real-world evidence and observational evidence from routine clinical practices with patient healthcare databases. The venture will rely on MIT’s Julia programming language, which is designed to whip through complicated computational problems quickly.

Julia, organizers hope, will help the group sift through massive amounts of data quickly to help compare effectiveness claims of a given drug against similar compounds, to help put together efficacy or safety data on a drug or device, to assess treatments already on the market, to sniff out treatment gaps and to help evaluate patient risk quickly. It’s all to, as organizers say, “guide a company’s business decisions with insights into a potential asset’s life-cycle management and due-diligence efforts.”

The team will be led by the Julia Lab, but housed in CTM.

The project comes at a time when the clinical trials industry is focusing on Big Data analytics in the hope that it can revolutionize trials the way it has other business sectors. Healthcare and pharma investments in Big Data are expected to reach $5.8 billion by the end of next year, MIT estimates.

The new collective was announced within days of a new trade group called Align Clinical CRO — consisting of most of the largest CROs in the market — offering up new operational data-sharing standards that the group hopes will simplify a traditionally thorny problem among sponsors, sites and CROs.

A typical trial spends between $50,000 and $200,000 just creating data extracts and data uploads — translating terms from site to sponsor and back again. Even a seemingly simple matter — such as determining when a site is “active” in a trial — can create complications.


-By Bill Myers