Outdated technology, inefficient workflows and a heavy reliance on in-person interactions collectively make the administration of healthcare one of the costliest and most time-consuming aspects of the system. But it doesn’t have to be. Notable Health has developed AI-powered software to automate the myriad manual and cumbersome tasks like patient registration, billing and clinical note-taking, resulting in a new, virtual system of data gathering and management that offloads the busy work while improving the experience for both doctors and patients.
The company has seen an uptick in business now that Covid-19 has accelerated the digital transformation of healthcare, prompting many health systems to adopt technology that not only enables virtual visits, but provides better data visibility across all aspects of the healthcare continuum. Notable Health CEO and co-founder Pranay Kapadia – who previously worked at fintech startup Blend and financial services platform Intuit – sat down with Greymatter to discuss the role of automation and artificial intelligence in healthcare, his experience innovating legacy stack and enterprise software in heavily regulated industries, and strategies for scaling in a pandemic.
You can listen to the podcast here:
Below are key highlights from the podcast:
HOW DO YOU INNOVATE IN REGULATED, COMPLEX INDUSTRIES?
“What everyone’s trying to figure out in these verticals is how you can collect the highest fidelity data, in the most seamless of ways, and figure out which method is the most accurate either to manage risk or manage money.”
“We had to look at the entire data stream in the workflow to find out why [current methods of administration] exist in healthcare. And then we started to think about what kind of experience we wanted doctors and patients to have, what kind of technology could deliver it, and how we could use machine learning to solve problems. We thought of it that way, versus trying to create a point solution that we would just stick into the healthcare system and probably wouldn’t actually change anything.”
WHAT IS DIFFERENT ABOUT BUILDING ENTERPRISE SOFTWARE IN HEALTHCARE?
“You can fall in love with the AI. You can fall in love with the design. You could fall in love with the solution, but it’s only as good as the integration.There is no CIO who is going to rip out all their existing enterprise software and just replace everything, especially not in healthcare. It just doesn’t work that way. You have to figure out how to build something that becomes the system of engagement and the system of intelligence that actually sits on top of the system of record.”
“The road to purgatory in healthcare is driven by integration. It takes too long to roll things out and iterate in order to make sure you are delivering something of value. And it’s expensive – it costs a million dollars to integrate something in the healthcare system. I jokingly say every other industry might have an API as we know it to mean application programming interface. But in healthcare, the “P” stands for person not programming. Because every time you add a tool, you just add more people. They copy and paste between systems, and it just gets more complicated and inefficient.”
HOW DO YOU PROVE ROI IN HEALTHCARE AT SCALE?
“Almost every other industry actually functions on growth, whereas healthcare is one of containment. So when you think about actually building things at scale, you have to build things that truly drive out wastage and yet have a competitive, durable advantage.”
“We tend to roll things out with smaller groups to make sure that the platform is stable and then roll it out to the bigger groups. As we are building products, we ask ourselves how we can take childlike curiosity with a framework called “learn, de-skill, automate and enhance.” It’s very similar to how we learn as human beings, but we do that in rapid cycles. For example, we learn about a specific process, like the system to get the right patients in to see the right physicians at the right time. Can we take that on and learn how to do that ourselves? Can we now de-skill it to the point where someone with a $15 an hour job salary can actually do it because it’s just following systems? From there, can we then automate it? And then, can we sprinkle in ML or AI to really enhance it to the point where now it’s actually radically, profoundly different? Then we do that as fast as possible, as many times as possible. And that has been transformative in how quickly we’ve been able to build products in healthcare safely.”
HOW IS COVID-19 ACCELERATING HEALTHCARE’S OVERDUE DIGITAL TRANSFORMATION?
“Now is the time in healthcare to reset and rebuild. You have to ensure that your patients feel safe and ensure that your providers are safe, because if they aren’t, the whole system comes crashing down.”
“Everything has been fundamentally transformed where patients are now demanding virtual and touchless experiences. On one of our first contracts, the leaders of the healthcare system said that even if 5% of their patients used our virtual patient experience, it would be transformative. We are currently seeing 95% of patients actually checking in on their own device now. We grew 2.5x in the first quarter of this year. People want a touchless experience, and health systems say their staff wants touchless experiences. Nobody wants to share anything anymore.”
“It is not a one-year endeavor to take the trillion dollars’ worth of administration expenses in healthcare and drive it down by 95%. But it’s a mission worth going after. We see now why it has to happen, so it’s inevitable. We have to treat this like the catalyst that can make it possible faster than ever before.”
HOW DO YOU ASSEMBLE THE RIGHT TEAM TO WORK IN COMPLEX INDUSTRIES?
“Having a common vocabulary is tremendously important. You have to have an open and honest environment. In any new vertical, you have to be able to speak the language of the customer. You have to be explicit as a team in talking about what you are good at and where your knowledge gaps are. It’s more than finding people with the right title and brand recognition. They have to be good with ambiguity. They have to know they are the ones that will create these systems and playbooks for how we’ll actually roll things out.”
