![](https://medcitynews.com/uploads/2021/07/GettyImages-1280895783-150x150.jpg)
Hospitals are facing a slew of obstacles that negatively impact their revenue cycle processes — from billing backlogs to labor shortages to burnt out clinicians who don’t have time to deal with hours of administrative processes. One startup is on a mission to alleviate hospitals’ coding and billing woes with machine learning, and investors appear to be on board.
CodaMetrix, an AI platform for hospital revenue cycle management, on Monday closed a $55 million Series A funding round led by SignalFire. The financing round also included investments from Frist Cressey Ventures, Martin Ventures, Yale Medicine, CU Healthcare Innovation Fund and Mass General Brigham physician organizations.
Initially, the company’s technology was developed at Mass General Brigham for in-house use at the health system. By the time CodaMetrix officially launched in 2019, the technology had been used at Mass General Brigham for about a decade, said CEO Hamid Tabatabaie in a recent interview.
Since its inception, CodaMetrix has been focused on addressing one major problem plaguing healthcare providers: the complicated process of translating patient records into codes that are used for either billing and revenue cycle or clinical and research purposes.
“As you would imagine, a human being has to read the handwriting and notes from a physician, and that can be error prone. It’s also very time consuming. We’re trying to reduce the cost of health care — this is a big meaty chunk that we’re attacking,” Tabatabaie declared.
To alleviate the problem, CodaMetrix has developed machine learning software designed to automate the medical coding process.
The company’s AI is trained on historical coding data from its health system customers. CodaMetrix cleans, curates and processes that data so that it can establish a dataset that appropriately trains a machine learning tool. That tool is designed to replicate that medical coding process that most hospitals still complete manually.
The tool works by intaking patients’ medical reports and making predictions about what was done to the patient and which procedure and diagnostic codes will be appropriate for the care interaction.
“When we’re sufficiently competent, we completely automate the case so it doesn’t need to be touched by human being. When we’re not, we can send it back to the customer for manual coding. And we have another product where we can send it back to the customer and actually show them the codes that we had predicted and the evidence associated with those codes so they have a good starting point to make their job faster,” said Javed Aslam, CodaMetrix’s chief data science officer.
CodaMetrix’s customers own and operate 111 hospitals across the country, Tabatabaie said. Some of these customers include Mass General Brigham, Yale Medicine, Henry Ford Health and University of Colorado Medicine.
The company is aiming to sell its software to the top 200 health systems in the U.S.
“We’re looking for the largest providers in any region that preferably include one or more academic medical centers. The reason for that is that rare cases tend to be treated in large academic medical centers. And when you’re in the data training business, those rare cases are an opportunity to learn. There’s also the fact that they have larger volumes. Our solution probably disproportionately benefits them, because the larger volume institutions have the higher coding costs,” Tabatabaie said.
CodaMetrix is certainly not the only company seeking to make the medical coding process easier with AI — others include Fathom, RCxRules and Nym Health. But CodaMetrix is the only provider of revenue cycle software that has come out of a healthcare provider setting with immense prior access to clinical data, Tabatabaie pointed out.
He also said that CodaMetrix is unique because the company believes that the control and configurability of its system has to be in the hands of the provider.
“They’re the ones who have to decide what the quality of output should be. They’re the ones who have to decide how to automate some things and not others. So no black box should show up and tell people how to code. It should be completely transparent, and the controls should be theirs,” Tabatabaie declared.
Another reason he claims CodaMetrix stands out from its competitors is its software can perform coding for more specialties than any other company.
Photo: sorbetto, Getty Images