It can take up to 15 years to get a new drug through the regulatory-approval process before it becomes available to patients. One of the reasons is the time it takes to analyze the mountains of data generated by clinical trials about the efficacy and safety of the product. Drug companies perform quality-assurance tasks to verify that the steps carried out during clinical trials comply with government regulations. Regulatory agencies then check the data from the trials, as well as from the trial sponsors, to ensure that the safety of patients wasn’t compromised.
The way in which quality is assured hasn’t changed much in about 25 years, says Timothé Ménard, head of quality data science at pharmaceutical company Roche in Basel, Switzerland. Meanwhile, the number of clinical trials has increased, and their designs have become more complex, Ménard notes.
That makes it challenging for pharmaceutical sponsors’ quality teams to identify and correct issues in a timely manner.
Ménard and other drug-company quality leads and quality data scientists say advanced analytics could help speed up the QA process. The analytics programs use machine learning, statistical modeling, natural language processing, predictive modeling, and related techniques to evaluate information.
“It’s the application of some of those advanced analytic techniques that we have seen used in other industries, such as finance, that we are looking to bring to the quality-assurance space,” Ménard says.
Faster resolution to issues
There are two pain points that data analytics can solve, he says. One is the tedious, time-consuming process of auditing clinical-trial sites and processes. Every drug company has QA teams that conduct audits to ensure good clinical practices. The teams travel to a large number of clinical-trial sites around the world and manually check the data. Ménard estimates that at any given time, Roche trials are being conducted at thousands of sites.
“There’s so much data being generated that it’s nearly impossible to catch all the errors,” he says. “Even if we were able to hire an army of auditors, the likelihood we would miss something is high. The current situation is not sustainable.”
Advanced analytics would give a more comprehensive overview, he says, enabling problems with a study to be detected earlier and resolved more quickly. A reliable, standardized QA tool could augment or even replace traditional audits, he suggests. The tool, he says, could allow a holistic look at a clinical program, in near real time, while focusing fewer resources on issues that happened in the distant past.
“Improving the quality of clinical trials is something we all agree is for the greater good.”
Another benefit of analytics is the ability for audits to be conducted remotely. The technology could have been used during COVID-19 shutdowns, when auditors were unable to make in-person visits to clinical-trial locations, Ménard says.
“COVID really shone a light on what we had recognized many years ago: that something needs to change,” says Roshan D’Souza, a head of Roche’s data analytics and operations.
The other pain point is the timing of when regulators conduct their inspections. They take place after the clinical trials are completed but before a drug is released to the market. If regulators find a problem, it can take the company up to a year to fix it, D’Souza says.
“A consequence of these late-issue detections is that drug approvals could get delayed,” he says. “That’s a problem for the pharmaceutical company as well as patients.”
Advanced analytics could make it easier to investigate and resolve problems earlier, he says, adding, “That would be a complete change in the way we’re doing quality assurance now.”
The benefits of an industry group
To better understand how applying data analytics for QA processes would work, Roche and a handful of other pharmaceutical companies formed the Intercompany Quality Analytics industry group in 2019. The group, which goes by the moniker IMPALA, conducted several use-case studies and published its findings in a 2021 journal article that Ménard coauthored.
But, D’Souza says, IMPALA couldn’t hope to change the current QA processes on its own. The solution was to form the IEEE IMPALA Consortium, a program of the IEEE Industry Standards and Technology Organization, which provides its customers with legal and financial infrastructure and administrative support for standards development and industry adoption of emerging technologies.
Roche and 12 other drug companies including Biogen, Boehringer Ingelheim, Pfizer, and Sanofi are members of the consortium. Its mission is to work with regulatory agencies and other pharmaceutical companies to adopt advanced analytics and associated methodologies in the biopharmaceutical clinical-QA process, with a goal of bringing quality, life-saving medicines to patients faster.
“What we are trying to do is effect a paradigm shift,” says D’Souza, a deputy chair of the steering committee’s consortium. “The shift must happen across the entire industry—which is why a consortium is necessary. We wanted to make sure all the biopharma companies actively participate in shaping the future of how quality assurance is done using advanced analytics. These include technical standards and best practices.”
Algorithms, open-source documents, and workshops
The consortium is developing several work products. One is a scheduling plan for audits. The plan uses algorithms to determine the best times to make site visits based on considerations such as peak hours and holidays as well as the auditor’s workload and travel time.
The consortium is planning to hold a workshop this year to educate others about its efforts. It held one in November with industry peers and health authority delegates at the Research Quality Association’s international conference in Brighton, England, to gather ideas about other areas where data analytics could be used. An overview of the findings is now available.
“The benefit of the consortium is that we can tap into the members’ collective intelligence,” D’Souza says. “Improving the quality of clinical trials is something we all agree is for the greater good. It’s for the benefit of patients, society, and the health care system. That’s why we’re teaming up.”
Original Source: https://spectrum.ieee.org/big-data-accelerate-drug-approvals