Rethinking healthcare delivery amid Medicaid policy shifts

The news about the Trump administration planning massive cuts to Medicaid hit hard. The proposed changes are expected to cause $80 billion in economic losses across the healthcare industry, including nearly half a million job losses in healthcare delivery alone.

Let me be clear: I believe cutting Medicaid is a tragedy. And I strenuously disagree with it.

There is ample evidence from the past decade showing that people in states that expanded Medicaid have experienced important benefits compared to people in states that did not. A 2020 Kaiser Family Foundation review of over 400 studies found that Medicaid expansion was consistently linked to gains in coverage, improved access to care, enhanced financial security, and even improvements in health outcomes.

Medicaid is a vital safety net – one that protects the most vulnerable among us from being one unexpected medical bill away from homelessness, bankruptcy, or worse. Far too many people in this country live with the quiet terror that a single bill – maybe just a few hundred dollars – could send their entire lives into a spiral. Medicaid helps ensure that pregnant mothers, children, and people with disabilities get the care they need. Make no mistake, without access to care, people who depended on Medicaid will suffer.

I strongly advocate for continued funding of Medicaid as a top national priority.

At the same time, we face another painful truth: healthcare is consuming nearly 20% of our GDP, and the cost of care continues to skyrocket.

One major contributor to these unsustainable costs is the industry’s deep-rooted dependence on labor as the approach to solve every problem. Despite two decades of introducing new technologies, healthcare delivery is no more efficient than it was 15 years ago.

Take this example: most large hospitals still employ 50 to 100 people solely to read through and abstract information from charts into structured fields for various reporting systems such as quality registries. And that’s just one small example. There are hundreds of other roles that exist simply to compensate for broken or outdated information systems, for example dozens of people at every health system who transcribe information from faxed documents into the EHR.

As I’ve written before, I’ll never forget what then-UCSF CEO Mark Laret said in 2020: “We have a once-in-a-lifetime opportunity to reshape healthcare delivery. Let’s not waste it.”

That message feels more urgent now than ever.

Yes, there will (and should) be intense debate at the federal policy level. I want to offer a perspective that I believe important no matter what happens with healthcare funding and reimbursement. Because, even if healthcare reimbursement remains stable, we should not ignore the need to improve costs of care through delivering quality at better efficiency. We live in a moment where artificial intelligence is unlocking unprecedented productivity gains—and this creates a very real opportunity to reimagine how we deliver care.

Hospitals facing budget pressures today often believe they are facing one of two choices:

• Do they succumb to financial pressure by cutting services and closing departments?
• Or do they continue trying to patch the gaps using traditional methods that rely on ever-growing labor costs?

I believe they actually have the ability to choose a third path. Agentic AI is increasingly giving them a scalable, digital workforce that allows them to actually improve the services they offer, while simultaneously reducing labor costs.

Let me give you one example: the call center.

A mid-size health system often has a call center with 600 to 800 staff. The patient experience is typically adequate, but certainly not ideal. Long hold times. Reps who can’t always resolve your issue. Limited hours – Monday to Friday, 9 to 5 – right when most patients are working and unable to call.

Looking back on my time as VP Digital Health at UCSF when we enabled newly referred patients to digitally self-schedule, one thing always stood out to me: there was a huge spike in patients logging in to book appointments at 1 or 2 a.m. Why? Because that’s when busy working parents finally have time – after the job, the dishes, the laundry, the bedtime routines – to sit down, catch their breath, and manage life’s logistics.

And yet, when they finally sit down to engage with the traditional Mon-Fri 9a-5p healthcare system, it’s closed. No one is available to help.

Now imagine a different world. What if AI could enable that same parent to speak with a kind, informed, responsive agent – at 2 a.m. on a Saturday – even if their preferred language is not English? What if 100 such patients could call at the same time, and every single call could be handled, at the same time?

AI now allows us to redirect our precious human resources toward more direct patient care activities, away from manual, repetitive tasks like transcribing faxed documents from a PDF into the EHR or abstracting information from chart notes into discrete quality registry fields. 

We can already today shrink the army of people needed to perform these tasks or to chase prior authorizations and instead reinvest those precious resources into caregivers who greet, care, and directly connect with patients.

This shift allows us to both lower the cost of care and improve the quality of the experience and outcomes.

This isn’t a fantasy. This isn’t some distant future. It’s already happening.

When Medicaid cuts force hospitals into painful decisions – closing emergency departments, shutting down unprofitable specialties like endocrinology, prioritizing commercially insured patients just to cross-subsidize charity care – what if they had another option? AI will enable them to unlock new levels of productivity, letting them not just keep these services, but improve them, while also lowering costs of care.

This is our once-in-a-generation opportunity. Let’s not waste it.

‍