Artificial intelligence applications to improve revenue cycle management in healthcare hold promise, but executives are concerned about the accuracy and reliability of the technology.
This is one of the results of Inovalon’s survey of more than 400 finance and revenue cycle executives and managers, of which 84% said they were optimistic about AI-enabled RCM in hospitals.
However, a third of respondents said they were concerned or skeptical about using AI in RCM, with concerns about accuracy and reliability (31%), lack of familiarity/understanding (17%), and AI being too new/untested (15%) being the key concerns.
Humans are better than AI
Twenty percent of respondents said they believe human performance — at least at this point — is superior to AI.
Julie Lambert, president and general manager of suppliers at Inovalon, said Healthcare IT News That applies to RCM as a whole, but there are certainly areas that could benefit more from AI.
“Rather than thinking about this as an either/or scenario, I want to challenge us to think about it more as expertise is a critical foundation for creating AI/ML models that work and are continually refined,” she said. “When technology and expertise are combined, the potential for the best outcomes is there.”
In her view, the areas where AI can have the biggest impact on RCM are the ones that are most painful and manual for providers today.
Of these areas, denials, pre-approvals and eligibility rank near the top for all providers, and she says it’s no coincidence that all of these items are related.
“Mistakes early in the application process will lead to rejections later on,” Lambert said.
What caused the rejection?
Knowing what situations are causing denials and how to detect or predict those denials before they happen is the perfect opportunity to use AI with expertise and claims outcome data to build and train models.
“There are opportunities within these processes themselves and within the overall connectivity to use ML and AI to improve the quality of service for providers,” she said.
Lambert added that an important factor to consider is that AI is not static and should never be treated as such.
“Designing a model that is capable of continuous learning is a core principle of AI – models will continuously learn from data and feedback loops will naturally arise from the results,” she said.
External factors
It is also important to be aware of and take into account knowledge of external factors that may impact the model. This could mean regulatory changes that impact the data structure, data elements in the response, or other factors that may introduce anomalies in the data.
“Make sure everyone is aware of any changes that impact the model so that interpretation of results doesn’t introduce false assumptions or correlations,” advises Lambert.
She added that it’s important to make sure people understand that AI isn’t just for senior leadership or data scientists — it’s for everyone, and that’s what will make AI successful.
“AI needs the involvement of actual people who are managing the data, running the operations, and managing the workflows that support the creation of the models,” she said.
Nathan Eddy is a healthcare and technology freelancer based in Berlin.
Email the author: [email protected]
Twitter: @dropdeaded209
The HIMSS Healthcare AI Forum is scheduled for September 5-6 in Boston. Learn more and register.
