When healthcare executives ask me about AI, there’s one question that comes up more than any other: “How do we get paid for this?”

It’s a practical question. AI systems cost money. If there’s no reimbursement pathway, the economics don’t work—especially for services operating on thin margins.

Let me walk through how AI intersects with Australian healthcare funding, focusing on MBS and PBS implications.

The Current State: Mostly Unfunded

Let’s be direct: most clinical AI applications currently have no specific MBS item numbers. You can deploy AI, but you won’t receive additional Medicare reimbursement for using it.

This creates a situation where AI has to justify itself through efficiency gains or quality improvements rather than direct revenue. For some applications, that works. For others, the economics are challenging.

How MBS Reimbursement Works (Brief Refresher)

The MBS funds medical services through item numbers that describe specific procedures, consultations, or diagnostic tests. To add a new item number, the Medical Services Advisory Committee (MSAC) must evaluate the service and recommend it to the Minister.

The MSAC evaluation considers:

• Clinical effectiveness (does it improve outcomes?)
• Safety (is it as safe or safer than alternatives?)
• Cost-effectiveness (is the cost justified by the benefits?)
• Population impact (who would use it and how often?)

New item numbers are added regularly, but the process takes 18-36 months from application to implementation. It’s thorough, which is good for ensuring value, but slow.

What’s Currently Covered

A few AI-adjacent applications have MBS coverage:

Diabetic retinopathy screening. AI-assisted retinal screening can be billed under existing screening item numbers. The AI isn’t specifically covered, but the screening service is. This makes AI deployment economically viable because it increases screening capacity without requiring additional specialist time.

Diagnostic imaging. Radiology AI runs in the background of studies that have their own item numbers. The AI cost is absorbed into practice overhead, offset by efficiency gains. This works when AI genuinely improves throughput.

Pathology. Similar to imaging—AI assists with tests that have existing item numbers. The AI cost is part of providing the service, not separately reimbursed.

What’s in the MSAC Pipeline

MSAC is evaluating several AI-related applications. I can’t predict outcomes, but here’s what’s in progress:

AI-assisted genomic analysis. Several applications propose using AI to accelerate genomic interpretation, particularly for rare diseases and cancer. If approved, this could create item numbers for AI-assisted genomic services.

Cardiac monitoring AI. Continuous cardiac monitoring with AI analysis for arrhythmia detection is under evaluation. This is adjacent to existing holter monitor item numbers but potentially broader.

Remote monitoring AI. AI-powered remote patient monitoring for chronic disease management is being evaluated. This connects to telehealth items but isn’t identical.

None of these are guaranteed. MSAC evaluations can take unexpected turns, and cost-effectiveness is often the sticking point.

The PBS Angle

PBS (Pharmaceutical Benefits Scheme) is primarily about medications, but there are AI implications:

Precision medicine. AI-guided treatment selection (choosing the right drug for the right patient) could influence PBS subsidisation. If AI can identify which patients will respond to expensive therapies, it might support more targeted PBS coverage—better outcomes at lower total cost.

Medication safety. AI systems that prevent adverse drug events or optimise dosing could theoretically justify PBS consideration, though this is more speculative.

Currently, PBS doesn’t specifically address AI. But as AI becomes integrated into prescribing decisions, the intersection will grow.

Making the Economics Work Today

Given limited reimbursement, how do you justify AI investment?

Efficiency gains that are real, not theoretical. If AI genuinely increases radiologist throughput, you can read more studies with the same staff. Calculate the value of that additional capacity.

Quality improvements with downstream savings. Earlier detection of conditions might reduce treatment costs later. This is harder to prove and capture financially, but it’s part of the value equation.

Competitive positioning. In private healthcare, AI capabilities can differentiate your service. Patients and referrers may prefer providers with advanced diagnostic technology.

Workforce risk mitigation. In areas with workforce shortages (pathology, radiology), AI that maintains service levels despite recruitment challenges has strategic value beyond immediate ROI.

Preparation for future reimbursement. If you believe MBS coverage is coming, early adoption positions you to benefit when it arrives.

What Private Health Insurance Covers

Private health insurers make their own decisions about what to cover. Some are experimenting with AI-related coverage:

Premium offerings. Some funds are trialling coverage for AI-enhanced diagnostics as part of premium products.

Value-based arrangements. Insurers interested in total cost reduction may support AI that improves outcomes and reduces downstream costs, even without explicit coverage.

Prevention programs. AI-powered health risk assessment and prevention programs are appearing in some fund offerings.

This is fragmented and variable. If you’re in the private sector, talk to your major insurance partners about their AI perspectives.

Advocacy for Better Reimbursement

If AI reimbursement matters to you, engagement with the policy process is worthwhile:

MSAC submissions. Applications can come from manufacturers, clinicians, or healthcare organisations. If you have compelling evidence for an AI application, consider a submission.

Industry associations. Medical colleges and healthcare industry associations provide input to government on MBS priorities. Ensure AI is on their agenda.

Government consultations. The Department of Health regularly consults on MBS review. Participate when opportunities arise.

Build the evidence base. MSAC decisions rely on evidence. Implementing AI with rigorous outcome measurement contributes to the evidence that future reimbursement decisions will need.

Looking Forward

I expect AI-specific MBS items within the next three to five years, starting with applications that have the strongest evidence base—probably in imaging and genomics first.

In the meantime, AI investment needs to be justified through efficiency, quality, and strategic positioning rather than direct reimbursement.

That’s a higher bar, but not an insurmountable one. The organisations implementing AI now are building capability that will serve them well when reimbursement catches up to the technology.

---

Dr. Rebecca Liu is a health informatics specialist and former Chief Clinical Information Officer. She advises healthcare organisations on clinical AI strategy and implementation.