By Scot L. Harper, Ph.D., president, SLH Group, LLC
The pharmaceutical industry as we know it today has roots that reach back to the apothecaries and pharmacies of the Middle Ages, when drug discovery largely involved sourcing plants and herbs for natural remedies and drug development and testing was an unstructured concept that relied on the hit-or-miss reactions patients experienced.1 Failure was arguably more common than success and, unfortunately for the patients, treatment courses often did not yield optimal results.
The era of modern drug discovery and development arguably began with the breakthrough discoveries of insulin by Frederick Banting and colleagues in 19212 and penicillin by Alexander Fleming in 1928.3 In almost a century since, and especially from the 1950s forward, clinical development has proceeded at breakneck speed, guided by rigorous and specified scientific protocols and aided by breakthrough technologies that enable and realize previously unimaginable progress.
AI Makes Research Faster, More Efficient
One of the most exciting advances in clinical development is the relatively recent use of artificial intelligence (AI)-informed virtual assistants to guide and accelerate drug discovery. Until recently, analyzing the massive amounts of information generated in any clinical trial required a considerable amount of labor-intensive, resource-consuming effort. Site- and patient-level data tracking was a cumbersome process, often requiring days to secure resolutions of queries. Needless to say, this belabored route had huge potential to negatively impact timelines and costs of clinical development.
The advent of advanced data analytics tools and solutions that leverage AI has changed the clinical development landscape. Leading data analytics companies have engineered cutting-edge, AI-informed virtual assistants that are changing the rate at and efficiency with which the life sciences industry conducts clinical research and propelling drug discovery and development to new heights. Such virtual assistants — think of them as the clinical development cousins to Amazon’s Alexa, Apple’s Siri, and IBM’s Watson Assistant — do for drug development what those other chatbots do for our personal and professional lives: access information faster, enable progress, and accelerate productivity.
In today’s life sciences industry, such advantages are critical. Historically, pharma and biotech have had difficulty achieving clinical trial milestones, often experiencing a large chasm between targeted outcomes and actual results. The majority of global leaders in clinical operations rank patient recruitment (95 percent), site productivity (75 percent), and patient compliance (65 percent) as very important, but they are often unable to achieve these milestones. Only 47 percent report successful enrollment, 22 percent say their sites are productive, and 25 percent consider patient compliance efforts successful.4
AI-based virtual assistants are poised to change those outcomes. Virtual assistants are shifting the human-computer interaction paradigm, enabling clinical operations professionals to obtain more specific insights, faster than ever before, about the clinical trials they are running. Such knowledge goes a long way toward alleviating the planning, feasibility, and conduct challenges inherent in and pervasive throughout the drug development continuum.
The Ability To Ask Data For Answers
Through AI-powered virtual assistants, researchers can essentially “speak” to their data, asking questions and receiving responses about various aspects of study conduct. Because these virtual assistants are designed to focus exclusively on the domain, or subject area, of study conduct, they will not respond to extraneous queries about unrelated topics, such as the weather. They will, however, provide specific and detailed responses about a particular clinical trial. To properly answer a question, virtual assistants establish the context of a query, which is a combination of interpreting what a researcher wants to know (the intent of their question) with the full set of possible parameters or entities, such as names of people, organizations, locations, expressions of time, monetary values, etc. Once armed with such an understanding of the question at hand, which takes only seconds, virtual assistants mine available data about a trial to respond. As an example, if a clinical operations director wants to start their day by assessing the health of their portfolio of studies, instead of reaching out to a number of trial leads, he can simply ask his virtual assistant to display the portfolio risk chart. Follow-up questions about a particular study might include the following:
- What is the approved budget?
- What is the actual spend?
- How many subjects are enrolled?
These unique conversational experiences are much more dynamic than the traditional approach relying on canned analytics dashboards, and today’s virtual assistants can remember the context of previous inquiries and seamlessly enfold new entities into the discussion to provide rapid clinical operations insights.
Benefits Throughout The Development Spectrum
The benefits to clinical operations professionals and the impact on trials are enormous. Delays in starting up a study can have major cascading downstream effects, so being able to monitor startup progress in real time can avoid multiple complications. Additionally, key performance indicators (KPIs) can be tracked and managed to mitigate operational and financial risks, identifying obstacles and enabling decisions and course corrections before trial delays occur. Features designed specifically for tracking drug efficacy and patient safety enable simultaneous analysis of up to 50 variables, rendering the need for manual data analysis obsolete and contributing to significant savings in clinical trial staff time and effort. Increased efficiencies and cost savings can be brought to bear on critical outcomes such as patient recruitment, protocol adherence, prediction of study success, continuous process improvement, timely and accurate analytics insight, patient data privacy, and the ability to leverage previously untapped sources of data.
AI-informed virtual assistants let pharma and biotech talk the talk — literally. Such unprecedented conversational experiences with clinical trial data are shifting the clinical development paradigm for good and for the better, overcoming obstacles historically associated with clinical development and enhancing the life sciences industry’s ability to deliver safe and effective therapies.
AI-based technology was likely unfathomable to Banting and Fleming — imagine their reactions to actually being able to talk with their scientific data. And we’ve only just begun to scratch the surface of this amazing technology. We don’t need to think about fast-forwarding 100 years to imagine or anticipate the next astounding technological advance for clinical development. The future is happening now, and it is more promising and exciting than ever before.
About The Author:
Scot L. Harper, Ph.D., received his undergraduate degree from DePauw University and his doctoral degrees from Indiana University School of Medicine. He began his industry career with GlaxoSmithKline in medical affairs and new product development. He developed Eli Lilly and Company’s oncology investigator-initiated clinical trial program, became U.S. oncology medical director, then led U.S. clinical research operations. Dr. Harper headed up Novartis Pharmaceuticals’ North America clinical research operations and served as CEO of Novartis Clinical Operations, Inc. Upon retiring, Dr. he became the head of clinical development at oncology biotechnology company Endocyte. He is currently president of SLH Group, LLC, a life sciences consulting company.