Shela Schemel in Modern Healthcare
We've all heard the gripe: Physicians are turning into glorified note-takers, as they spend time during patient visits staring at screens and typing frantically into electronic health records, only to revisit the files hours later touching up their notes.
The problem is, in part, the technology that doctors must use. But some providers are now adding even more technology — in the form of virtual assistants, like the one made by Notable — to make things easier.
“Doctors are now the most highly paid professional data-gatherers,” said Notable CEO Pranay Kapadia, whose company makes an AI-powered medical assistant app for the Apple Watch that captures notes and visits digitally and recommends billing codes. “The best part is access to data that had never been available before,” Kapadia said.
Notable isn't just a digital scribe. Instead, it translates what's happening in the patient-provider encounter into data points in a form suitable for inclusion in the EHR. It also can take voice dictation from the Apple Watch, turning the provider's voice-entered notes into the right format — and placing them into the right structured fields—in the EHR.
The software also learns from physicians' habits, predicting orders and billing codes and proactively inputting certain parts, like where labs should be sent, based on insurance requirements.
And when it has enough information, Notable can also predict billing codes.
Billing automation can only work, though, if the tool is integrated with health systems' other digital systems. Kapadia and his team aim to make Notable integrate easily into the EHR, with data transfer dependent on either robotic process automation or application programming interfaces, much as happens on Mint.com, the personal finance and budget manager made by Intuit, where Kapadia was head of product for Mint.com and Mint Mobile. Like Mint.com, Notable is relatively easy to set up, requiring at most a few weeks before a health system can start using it.
“We've seen enterprise technology be made or broken based on the integration,” he said. “It's only as good as the integration.”
But as with any technology solution, a certain level of human oversight could still be necessary. “I believe that the workflow would still require revenue integrity and coding professionals to either review the charges caught in the edits or fix any that are not passing edits that are important for specificity and modifier hierarchy,” said Shela Schemel, vice president of operations at Navigant.
Furthermore, the question of how voice recognition will select the procedural device if not spoken by the provider likely remains unanswered at this point. A potential downside could be that the recognition technology from AI could lead to more intervention in the prebill process, slowing down claims and billing via volume, as well as reimbursement and revenue-cycle operations.”