90 Days In: The World of AI, Clinical Research, and Medtech
Three months ago I walked into the Deep 6 AI office, got my laptop, found a couch (we had outgrown our space and didn’t have a desk for me!), and started my journey into the AI and medtech space. I had no idea the crazy whirlwind I was getting myself into and now, 90 days in, I’m taking a moment to step back and reflect on what I’ve learned in a very short and very intense time with this amazing team.
Artificial intelligence and the healthcare space were both new areas for me, so we started with a crash course. I learned about natural language processing and machine learning, while also wrapping my head around oncology research, genomics, and precision medicine. The team demoed, diagramed, and whiteboarded for hours, stopping for my endless questions and note taking. I’m so grateful for the passion and team spirit that motivated so many of my colleagues to take time from their work to help me understand the true value and impact of what Deep 6 is doing.
Within a week I was on the floor at MAGI West, one of the major industry conferences for clinical research, listening to those devoting their careers to research share their successes and struggles in this exceedingly complex industry. I started noticing themes — nervousness about the rapid rise of unproven tech, the need to balance patient privacy with interoperability and data completeness, and the importance of keeping patient care at the center of everything they do in a rapidly changing ecosystem.
In the weeks and months since, I’ve talked with users as they shared their successes and hopes for our platform, read every industry report I could get my hands on, scoured the web for best practices on trial management (big thanks to Dan Sfera there), and spent countless hours learning everything I could about this space and how our software serves it.
The conclusion I’ve come to is straightforward: the need is huge, the opportunity is now, the challenges are real, and I’ve never been so excited to work on anything in my life.
Every new drug, device, procedure, or treatment has to go through the clinical trial process before it can be approved and used on the public. On average it takes at least ten years for a new medicine to complete the journey from initial discovery to the marketplace, with clinical trials taking six to seven years on average to complete.
But consider this: nearly 80% of all clinical trials fail to finish on time (link), with 20% delayed for six months are more. Over two-thirds of sites fail to meet their patient recruitment for a given trial. It’s easy to see why. Each clinical trial has dozens of criteria every patient needs to meet to be considered for the trial. And often that information is time-sensitive and buried in complex, unstructured medical records, leaving research staff with little choice but to comb through files one-by-one or wait for the right patients to come in for appointments.
The medical industry isn’t short on ideas on how to improve the healthcare we receive every day, but rather the clinical trials process (while crucial to ensure only proven treatments make it market) is acting as a bottleneck for innovation and improvement. In the meantime, millions of people are impacted by life-threatening illnesses, while life-saving treatments are delayed for years, or even permanently. For time-sensitive conditions like cancer, patients don’t have months or years to wait.
As someone coming to this space with a tech background, where speed is paramount and friction is the enemy, this cries out as an opportunity for us to do better. And we can.
The opportunity comes in the form of AI. Artificial intelligence is an often used, but rarely understood, term in technology. It’s thrown around for everything from how we advertise shoes to how we pick stocks to how we listen to music (links), but most people don’t understand what it is or how it works. I won’t try to explain the technical depths of AI here, but if you’re looking for some primers, drop me a line, I have some good ones to recommend!
The truth is that there’s immense opportunity in artificial intelligence, but we’re still in the early stages of harnessing its potential. Artificial intelligence (and its siblings, natural language processing and machine learning), give us the ability to do things with data we couldn’t do before, at a speed that will blow your mind if you think about it for too long.
For example, the Deep 6 AI platform can query millions of unstructured patient records on even the most complex criteria in seconds. This isn’t because we have massive server farms working behind the scenes, it’s because of the power of our platform to understand what it’s looking at and draw connections beyond demographics and diagnosis codes.
But AI is just part of the puzzle. AI by itself isn’t a product. It’s how you use AI to solve real problems for real users in an intuitive, meaningful way that creates the true magic. And that’s where Deep 6 is really shining. Our platform allows researchers to build queries and find patients that fit complex criteria in a matter of seconds, then takes them straight to where that evidence is in the record, so they can identify and recruit patients for clinical trials in a matter of minutes. But even that’s just the beginning.
We’re working with physicians and research staff on the frontline battling our toughest medical challenges to understand how they think about medical conditions and how they use medical records to support patient care. From there we’re building a platform that empowers them to see and understand their patients’ data in ways they never could before.
Real-time access to query their patient population opens up an entirely new world for them to dig into medical data, understand their patients, look for trends, and design their research. I don’t want to give away any of the details of the groundbreaking research these doctors are working on — but it will knock your socks off when it comes out. I’m in awe of them every day.
This technology allows us to harness the medical data we have and change the way we conduct research, bringing new treatments to the people that need it most. But that doesn’t mean it’s going to be easy.
Everything worth doing has its challenges, and we certainly have some big ones. We’re dealing with pulling disparate data from dozens of different sources at a hospital, often from antiquated, siloed systems with very different data structures.
Also, we’re working with largely unstructured data (think doctors notes and pathology reports). Unstructured medical data is inherently messy. Doctors write notes in shorthand or use voice to text. They use acronyms and misspell things. They have a thousand different ways to describe the same medical concept. They include notes about treatments they considered, questions the patients asked, and family histories.
In addition, this is sensitive, personal data discussing the most intimate details of people’s lives. It needs to be treated with the upmost respect and protected with the highest levels of security to ensure that information remains private. At the same time the right people need to be able to leverage that data to provide and improve patient care. It’s a delicate balance.
Our medical records are a reflection of our lives, and life is messy and tough. But it gives us this amazing chance to take that mess and turn it into something we can learn and grow from.
What Comes Next
As we tackle the challenges in front of us, I can confidently say this team is aligned and laser-focused in a way that’s rare to find, even in tech. Our mission of accelerating innovation in healthcare to help bring new treatments to patients faster is at the core of every person’s decisions, every day. From implementing our new customers, to building new features, to spreading the word, every member of the team is moving faster than I thought possible to make the most of this time and opportunity.
I don’t know exactly where Deep 6 will be a year from now, but I have some ideas, some hopes, and some pretty good bets. With what I’ve seen so far and the people I’ve met that are backing it, it’s going to be pretty amazing. I can’t wait to share it with all of you, every step of the way.