Life sciences industry, it is time to start talking about robotics and cognitive analytics.
Actually, it’s time to start doing something about it. Many other industries are already there. Publishing, for example, or the automotive industry. Or financial services.
I hear you: “But we’re different. This is R&D in life sciences after all – not publishing.” It’s true – life sciences obviously operates within a web of regulatory oversight and complex science that makes for some unique challenges. And with people’s health involved, the stakes are even higher.
But that doesn’t rule out the potential for robotics and cognitive analytics in R&D – far from it. As an industry that actively embraces innovation and could use improved productivity, automation should be a top priority.
Consider clinical trials, for example. Many of the processes behind clinical trials are resource-intensive, relying heavily on highly trained scientists and doctors at virtually every step. And the costs add up: Studies put the overall cost of bringing a single drug to market at anywhere between $1.5-2.5 billion¹ and these costs continue to rise.
Now consider the actual work these valuable people are doing to support the trials. PhDs, for example, are frequently engaged in repetitive, low-level tasks – generating first drafts of documents like protocols, statistical analysis plans and data management plans; adverse event processing and follow-ups; and quality control of regulatory submissions for completeness. These jobs could be easily automated to free up time for other, more important activities related to the trials. Just as important, these additional activities and the use of automation could accelerate the entire clinical trial process significantly.
In concept, who could argue with using robotics and cognitive analytics tools to take on repetitive, low-value work in place of scientists, doctors, and others who have far better ways to spend their time? It’s a case that has been made before, certainly. But putting it into practice requires real confidence that it can work at the operational level in the context of life sciences R&D. Here are a few specific examples of how some life sciences companies are already bringing robotics and cognitive analytics to bear on clinical trials that may give you the information and confidence you need to take the lead on this important, potentially transformative development.
One of the most time-consuming aspects of the site initiation process is assembling the primary documents required by law – everything from signed and dated financial disclosure forms and investigator agreements to approval letters from ethics and Institutional Review Board (IRB) organizations. If you’ve ever been involved in this process, you know it can be deceptively complex, given the volume of interactions and compliance considerations. It doesn’t help that some of these documents, like the investigator CV and medical license, are generally good for only one calendar year, rather than the entire length of the clinical trial.
This process is typically executed on a trial-by-trial basis. But what if it were managed as a continuous process, constantly updated as documents expire and are renewed? This could be a prime target for automation:
- With robotic process automation, a virtual calendar of constantly updated upcoming documentation could help guide the entire process.
- Natural language processing capabilities could help identify which investigators were already known to the sponsor, and whether their CVs, medical licenses, and financial disclosures were already on file and up to date.
- Natural language generation tools could automatically draft emails to investigators to request updated documentation only as needed and on time.
Each of these examples and others – soliciting information where there’s a need, updating forms with minimal or no human intervention, communicating with relevant stakeholders, confirming information is updated, and many more – can be improved via automation.
Contract negotiations are another time-consuming aspect of clinical trials – and can be particularly frustrating when you consider that the contracts are usually between companies that have worked with one another many times before. “What were the terms of our most recent contract with this company?” “What information do we need in this year’s contract negotiations that was likely included in previous contracts?” The answers to common-sense questions like these often remain elusive because of the intense amount of manual effort they would require to answer.
But what if you could simply have a robotic tool comb through prior contracts to generate a clear view of the history of the relationship, including ample relevant details for prepopulating an updated contract? If this sounds like science fiction, it’s not. Tools such as Blue Prism, Automation Anywhere, and Narrative Sciences are making it happen today.
These examples are only scratching the surface of what’s possible in clinical trials, based on readily available technology today. It’s not difficult to imagine using robotics and cognitive automation to transform the process of patient recruitment, reviewing clinical records to find patients who might fit the criteria for recruitment, for example. Those who have already begun experimenting with these technologies tend to quickly move on to more complex tasks they hadn’t considered before. And maybe that’s the beauty of these types of capabilities: Once in place, they tend to spread into other parts of life sciences organizations. So often, that’s exactly how transformation happens. Isn’t it time to take that first step?
1Deloitte R&D ROI Study; Tufts Center for the Study of Drug Development