AI tools could ‘increase the scale, speed or sophistication’ of Medicare non-compliance, govt bodies warn

Government departments are quietly monitoring how AI is being used by practices to bill more on the MBS, according to a new report.

The technology is widely seen as a way of billing accurately for the services being provided, given the complexity of the system amid thousands of items, many with arcane item descriptors and unclear rules.

However, both Services Australia and the Department of Health, Disability and Ageing believe that AI may be used to “increase the scale, speed or sophistication” of non-compliance.

The glimpse into the concerns of officials was given this month in the Australian National Audit Office report called Artificial Intelligence and Medicare Benefits Integrity.

It says the health department’s Benefits Integrity Division began to pay more attention last year to practice management software offering AI features that suggest MBS items to bill for consults.

“Practice management and MBS billing software websites often refer to optimising billing processes, identifying missed billing opportunities, increased billing volumes, and in one case, suggesting the most lucrative combination of MBS item numbers.”

Last year, the division wrote an internal report saying the tools could lead to risks, including the billing of items that were incorrect or even clinically inappropriate, practitioner complacency, and full automation of the billing process without human input.

But back in May, it put together an internal draft report which expressed specific alarm about the potential use of AI to analyse clinical notes in order to suggest items to bill.

The audit office report does not name any specific programs.

The health department is apparently performing data and analytical modelling to “identify unusual item co-claiming combinations”, the report said.

However, the department apparently lacks information on which practices are using AI to assist with MBS billing, making detection of inappropriate use difficult. The report said more work was needed for “formal systematic identification and assessment of AI risks”.

“While artificial intelligence is not currently assessed as a demonstrated Medicare integrity risk in its own right, artificial intelligence-enabled technologies may increase the scale, speed or sophistication with which existing integrity risks manifest,” the health department said in response to the report’s recommendations.

“In relation to Medicare benefits integrity, the department’s focus will remain on preventing and responding to incorrect, inappropriate and fraudulent claiming.”