Comparative Effectiveness for AI MedTech

"Outcomes are what matter to payers. If your AI tool can identify patients earlier or avoid unnecessary procedures, that’s where the value lies."

- Midwest UnitedHealthcare Medical Director

Let’s dive into a topic that’s a bit of a double-edged sword for AI Medtech developers: comparative effectiveness. If you’re new to the term, it’s basically about how your product stacks up against what’s already out there. In other words, if your AI tool is supposed to be the next big thing in diagnostics or decision-making, payers want to know: is it better than the current standard of care? Or is it just different?

From a payer’s standpoint, they have a job to do, and that job is to manage costs while ensuring patients get the best possible care. Payers are responsible for making decisions that impact millions of people, and they have to be judicious about where they allocate resources. If your AI tool offers a new way to help diagnose a condition but doesn’t demonstrably improve outcomes or lower costs compared to existing methods, payers are going to be skeptical about covering it. From their perspective, it’s not just about adding new technology; it’s about adding value.

This is why payers often demand comparative effectiveness studies, even though they’re not required by the FDA. They need to be convinced that your product is worth the investment—not just in terms of its upfront cost, but in terms of the overall impact it will have on patient care and the healthcare system as a whole. And let’s be clear: they’re going to hold your AI tool to the same evidence standards as any other medical intervention, even if that might not seem fair.

For AI Medtech developers, particularly those working on decision tools or diagnostics, this can feel like a tough pill to swallow. After all, diagnostics and decision-support tools aren’t the same as treatments or interventions. They’re meant to assist healthcare providers by giving them better information, faster results, or more accurate predictions. So why should they be held to the same standards as, say, a new drug or a surgical device?

At the end of the day, payers are focused on outcomes—how well a product improves patient health and whether it helps reduce the overall cost burden. Just because your AI tool doesn’t directly treat a condition doesn’t mean it gets a pass on proving its worth. Payers want to know that using your tool leads to better decision-making by clinicians, which in turn should lead to better outcomes for patients and lower costs for the system.

"If you don’t have the clinical utility of your product confirmed, nobody’s even going to look at your economic analysis. Payers are willing to pay and pay reasonably well for things that improve health outcomes, but they’re not willing to pay for things that are marginal."

- Northeast Blues Medical Director

Let’s say your AI tool helps radiologists detect early signs of lung cancer more accurately than traditional methods. The FDA might approve it based on its ability to consistently identify those signs in your studies. But from the payer’s perspective, the real question is: Does this tool actually help reduce lung cancer mortality rates? Does it lead to earlier interventions that improve patient outcomes? And does it ultimately save money by reducing the need for more aggressive, costly treatments down the line?

You might be thinking, “But it’s a diagnostic! It’s to help physicians do their job better. We are providing better information. Why are we being subject to treatment evaluation requirements? That makes no sense at all!”

Well, yes and no. While better diagnostics can certainly contribute to improved care, they also come with costs—both direct and indirect. If your AI tool leads to more testing, more procedures, or more interventions without a clear benefit in terms of outcomes, payers might see it as adding unnecessary costs to the system.

"Just because their AI works doesn’t mean it matters. They need to define what 'working' really means and who cares."

- East Coast Regional Blues Plan VP

This is why payers apply the same evidence standards to AI tools as they do to other medical technologies. They need to be sure that any new tool or method not only provides accurate and reliable information but also translates into real-world benefits. In other words, they’re looking for a clear link between the use of your AI tool and improved patient outcomes or cost savings.

For AI Medtech developers, this means that your evidence strategy needs to go beyond proving that your tool works. You need to show that it works better—that it leads to better outcomes, more efficient care, or lower costs.

If you'd like some insight into how payers would view comparative effectiveness for your particular tech, set up a time to chat