We recently asked Alexa if she could code a few medical charts for us. Her reply? “Sorry I don’t know that.” We expected this. After all, the U.S. healthcare industry spends billions of dollars on 250,000 medical coders every year to do the job. This way of doing business might be error-prone, inefficient, and bound by constantly changing regulations, but hey, IT IS a solution. But you know what else is a solution? Autonomous medical coding. Tech company Nym.health just closed with $6 million in fresh financing and will implement their brand of autonomous medical coding right here in the USA. And who knows? It just might change the revenue cycle as we know it.
Romancing the Data
Nym Co-founder and CEO Amihai Neiderman is a veteran officer who served in unit 8200 of the Israeli Intelligence Corps. (He also hacked his country’s public WiFi just to prove he could do it, but that’s another story.) As a country, Israel may be much smaller than the U.S. However, their government has no shortage of incoming intelligence data. Technology plays a role in helping the country do more with less, helping officers like Amihai process the data accurately, and make decisions. After all, lives could be at stake.
And when decisions are made in healthcare, lives also hang in the balance. No one knows this better than Amihai’s wife, a physician and ENT surgeon. Early in their relationship, Amihai discovered that his wife had to pour through thousands of medical records to find relevant patients for her study. Astounded by how manual and time consuming the process was, Amihai decided to take what he learned in his own profession to fix his wife’s data problems. Their collaboration led to the development of CLU (Clinical Language Understanding) – the technology behind Nym’s autonomous coding.
How Autonomous Medical Coding Works
Source: Nym
Make no mistake: Nym’s autonomous coding is different from Computer-Assisted Coding (CAC) in that it requires no human intervention. And unlike other AI solutions, it understands unstructured physician language, and can show you how it arrived at its coding decisions.
Here’s how it works:
1. Doctors or medical scribes document the patient visit
2. Nym identifies any inaccuracies and assigns medical codes to the language
3. If any part of the medical chart isn’t understood, it goes into the coder’s queue for further action
Bye, Bye, AI Black Box
One of AI’s biggest barriers in health tech is that computers will arrive at a decision and there’s no way to see what’s going on behind the scenes. Nym AI lets you go back to review its coding thought process for auditing and compliance purposes – and it flags medical charts it doesn’t fully understand for human coding.
“We always know how we’re progressing, because at the end of the day, our goal is to get an accurate code we can explain,” explains Amihai. “That helps us stay focused on developing our technology and raise funds, which is important as Nym goes to market, and we review our analytics and feedback.”
“Explainability is also essential for startups who want to secure FDA-approval for healthcare technology that’s more than an assistant,” says Amihai.
Good to know.
Changing the Revenue Cycle Mindset
There are a lot of companies out there who will tell you that they have what it takes to streamline your revenue cycle or give better insight into the process as a whole. Nym wants you to stop thinking about the revenue cycle as a cycle. Period. After all, when you hear the word “cycle” you probably think of something with a lot of steps that could take hours, days, weeks, even months.
“It used to take our clients two weeks to have a medical chart coded, now it takes less than two seconds,” explains Amihai. “They get the results almost immediately, and the revenue cycle is collapsed into a single touchpoint.”
One of Nym’s clients just started to offer point-of-service collection, something that was previously unheard of. “Now they can get the results they want and estimate how much the patient has to pay immediately,” said Amihai.
Capturing the Inherent Beauty of the Medical Language
Did you catch the sarcasm in the heading of this section? Well, that’s one thing Nym’s coding doesn’t like: sarcasm. It also doesn’t like subtext, colloquialisms, literary devices, anything that gives language color and life, goes waaaay outside its realm of (medical) knowledge, or has a meaning that’s open to interpretation.
That’s why medical language, especially when it’s written by scribes who know how to write for billing, is a prime candidate for this technology, and the kind of language we use in everyday conversations is not.
“The way a physician expresses themselves in writing – in simple language – preferably without subtext, sarcasm, or questions, makes it easier to build a computational linguistics model that can interpret what they’re saying,” explains Amihai.
Believing in Computational Linguistic Unicorns
As Amihai explains, today’s AI models (which rely on probability and extracting meaning from groups of related terms) are great if you’re trying to build a search engine, but they just don’t cut it if you’re trying to analyze a piece of text, extract meaning from that text, and act upon it.
That’s why Nym’s computational linguistic unicorns are such an essential part of what Nym does. They understand linguistics and computer science. And most importantly, they understand the process the brain goes through when it’s learning a new language. As you can imagine, finding someone with these skills who can execute and apply their knowledge outside of academia is very, very rare. But these are the skills that have helped Amihai’s team perfect their CLU technology.
Securing the Right Domain
When we asked Amihai why Nym decided to go with a .health domain name, he told us “As of right now, our technology, our market, our actual domain (aside from our web domain) is healthcare, and we wanted to express that we’re a health and technology company,” said Amihai.
And we’re happy to have Nym on board!
Bringing Nym to Market
Nym recently partnered with QueueLogix and is doing a few pilots of their technology at QueueLogix-enabled medical centers. They’ve also partnered with ScribeAmerica (another Health Channels company) that offers QueueLogix as a software solution for revenue cycle management.
“ScribeAmerica wanted to offer coding services to all of their clients, so they needed to either hire 10,000 medical coders, or partner with a company like us that can automate it,” explained Amihai. “One of their needs was explainability for compliance issues, and we’re the only company that sold that specific solution and ability.”
Want to learn more about Nym and how autonomous medical coding works? Download their whitepaper or check out their website to request a demo.