Benefits of Artificial Intelligence for Clinical Research Will Require Human Touch Along the Way

Leigh Householder, Executive Vice President, Managing Director, Technology and Data Solutions, Syneos Health®

The clinical research enterprise may be barreling toward a future in which artificial intelligence (AI) will play a major role in decreasing workloads and increasing efficiencies for study professionals, but a whole lot of traditional human touch will be needed along the way, one industry observer said as part of some big picture predictions for 2023.

Earlier in December, Syneos Health®, a biopharmaceutical solutions organization, released its collection of 2023 Health Trends: Personally, Purposely Building What’s Next. “From AI and machine learning producing an era of rapid new pathways, to an industry-wide push for equitable representation in clinical trials, 2023 will be a catalytic year for biopharma clinical development and commercialization,” the company notes in its press release on this, its 13th annual trends report.

Among the dozen trends it spotlights, Syneos Health expects AI and machine learning will help drive strategic decisions to accelerate the commercialization of new therapies to patients. In response to questions from ACRP, Leigh Householder, executive vice president and managing director of technology and data solutions for the company, elaborated on what this could mean for clinical research professionals in the near future.

“We’re seeing AI’s impact across an incredibly varied landscape in clinical research,” Householder said. “That may, in some cases, affect clinical trial management roles and expectations. [For example, we’re] testing robotic process automation (RPA) tools with clinical trial sponsors. RPA can upgrade the role of [clinical research associates (CRAs)] by automating some of the work done offsite in order to amplify ‘eyes and hands’ onsite. In that sense, RPA is like an exoskeleton, only it’s supporting cerebral tasks rather than physical feats.”

According to Householder, the task to be automated might be something as routine as sending reminder e-mails to trial sites for expiring or missing documents. For clinical research coordinators (CRCs), she said, “the starting point might be utilizing a tool that helps convert electronic health records into enterable data sources to streamline electronic data capture entry.”

However, Householder noted, with any process involving AI or machine learning, “it’s important to keep the project goals tightly focused and the training data transparent, while building in necessary assessment and validation.” That’s why, in such early days as these in a new technology’s growth, a human being is always watching, she added.

“[M]aybe at first, all we’re doing is taking a process that costs a human 20 clicks on a mouse and reducing it to five, or just one,” Householder explained. “But in a very short time, we’re looking at much more sophisticated automation, with all the implications for job upskilling, consolidation, work efficiencies, and more.”

In the meantime, the trends report notes how burnout continues to be rampant in medicine, and 2023 represents “an important reset—a chance to reinspire [healthcare professionals] and create the flexible choices needed for life, work, and mental health.” Further, the report states that “[a]fter two years of talking about a war for talent, in 2023, organizations will become more refined in how they think about designing and creating value for the individual.”

When asked how these trends might translate to the daily jobs of CRCs and CRAs, Householder said that, with staffing being the biggest challenge confronting sites, “the situation, including retention issues and burnout, has gotten much worse during the pandemic. In part, that’s because of outsized sponsor expectations and the sheer volume of portals, electronic items, new systems, and requests” involved in running trials.

“Think of this as a snowball that gets bigger and bigger,” Householder added. “It rolls downhill and lands on coordinators employed by the site. [CRAs], employed by the sponsor or contract research organization (CRO), may be one step removed, but they are onsite as well, sharing the pain. To the extent that AI arrives initially as work automation, it may ease the burden by reducing the volume of tasks, but I wouldn’t say we’re seeing that right now. Nor is AI changing the way CROs or sites define, recruit, or train staff at this moment.”

Before the biggest benefits of AI can be harvested, “there’s a need to build empathy and connectivity among people facing impossible workloads,” Householder cautioned. “This is true across healthcare, but especially so on the front lines, which include trial sites. Empathy, respect, admiration for overburdened staff—these very human responses will carry us a long way.”

The clinical research enterprise also needs new models for thinking about recruitment and retention, Householder said. “Thousands of healthcare workers left jobs during COVID-19, and thousands may now be ready to come back,” she observed. “Organizations, including CROs, sponsors, and sites, need to embrace this development and set up systems to welcome people back—maybe into the same roles, but with the rough edges softened by human understanding and empathy. It’s pretty low tech, but it may be what the nurse or doctor ordered.”

Edited by Gary Cramer