Using AI to successfully appeal clinical claims denials
July 26, 2019
From HAL in 2001: A Space Odyssey to Skynet in The Terminator, artificial intelligence (AI) on the big screen has occasionally had a fairly sinister reputation. But in the real world, AI can offer you tangible fixes to expensive problems in the healthcare industry.
Rather than tackle the complexity of appealing clinical denials, understaffed back offices may simply write off denials as losses because they lack the resources, tools, processes, and expertise to mount successful appeals. However, in my experience, attacking this problem head on could save you millions of dollars each year. AI could be a new boon to providers to assist in this process.
Regardless of the availability of AI solutions in your organization, with so much money at stake it makes sense to dig deeply into these denials. By examining thousands of cases and tracking trends, we’ve identified the root causes of many clinical denials and how to overcome them. I often recommend several best practices providers can do at the front end to turn this red ink back to black:
Suggestions for back-office teams handling appeals include:
Here’s how AI can support your claims processes and help you avoid or recoup losses: By using automation to examine thousands of claims, you can define what a successful claim looks like compared to an unsuccessful one. Training the AI engine to look for missing documentation before the organization sends out the claim, and then communicating with the clinical and / or administrative teams to fill the gaps, can reduce your overall number of denials.
AI can help you clean up the back-end when denials begin to pile up. By training the engine to review denied claims and look for evidence of expected clinical care events that can support the claim in the patient’s chart, the tool can flag the claim for further action to resolve the denial. It can thus unearth the evidence needed to overturn a denial. AI can also flag claims, based on learned historical patterns of care, where there may be an opportunity to correct billed status and allow for a rebill or downgrade, thus capturing a portion of the denied claim (e.g., in-patient to observation). These actions can enable your staff to increase productivity significantly. For instance, clinicians can quality audit (QA) AI-reviewed charts and not have to comb through hundreds of pages manually.
Finally, robotic process automation (RPA) can follow the AI engine throughout the process. As it unearths information to help with the appeal letter, the RPA tool can pull that evidence into the letter on a real-time basis. The assigned registered nurse can then quickly review, sign and send this document to the payer as a formal appeal.
While front- and back-end robotics may have once had a distinctly Buck Rogers feel, today their use has gone mainstream. For example, Accenture’s recent survey of healthcare CFOs shows that 77 percent agree that performing routine and process-oriented tasks with tools like RPA gives finance departments the flexibility to provide insights to the rest of the organization – and recoup millions in denied costs.
AI and robotics can provide new support in the mission to offer patients the best healthcare services possible, cutting through complex processes and reams of data to make the case for reimbursement persuasively and quickly. And, they’ll never lock you out of your spaceship!