The uncomfortable truth about where our healthcare dollars actually go
ThetaRho Team · April 2026 · 4 min read
If you’ve ever felt like American healthcare costs too much while delivering too little, you’re not alone—and you’re not wrong.
Here’s a number that should make you pause: the United States spends 17.5% of its entire GDP on healthcare. Compare that to countries like Canada, Germany, Sweden, and Australia, which spend between 10–12%. We’re paying one and a half times what similar developed nations pay.
But here’s what’s truly frustrating: despite this massive spending, we’re not getting better results.
The One Thing We Actually Do Well
There’s a silver lining in this story, though it comes with its own complications. When it comes to the actual delivery of care—the skill of nurses, doctors, and clinicians—America leads the world. Our healthcare professionals are innovative, highly trained, and genuinely excellent at what they do.
The problem? Everything around the actual caregiving is where we fall short. Access to care, equity in care, administrative overhead—in these areas, we rank near the bottom among developed nations.
Technology Was Supposed to Fix This
Remember when Electronic Health Records (EHRs) were going to revolutionize healthcare? The HIPAA act of 1996 kicked off the EHR era, and by 2015, 96% of healthcare provider organizations had adopted these systems.
The theory was simple: digitize patient records, make information flow seamlessly, and watch efficiency soar.
Reality had other plans. Healthcare spending as a percentage of GDP has continued to increase even after widespread EHR adoption. The technology that was supposed to streamline care has, in many cases, created new layers of complexity.
The Real Culprit: A System Built for Billing, Not Healing
The fundamental problem isn’t technology—it’s incentives. America’s healthcare system is built on a fee-for-service model where money only changes hands when you physically see a doctor. Add in roughly 1,000 payers, 8,000 large hospitals, and 100,000 clinics, and you have a system optimized for transactions, not outcomes.
Then there’s the insurance structure: private insurers covering you until age 65, then Medicare takes over. The uncomfortable truth? Private insurers aren’t financially incentivized to keep you healthy long-term. They’re incentivized to spend the minimum necessary until you become the government’s responsibility.
Where AI Actually Has a Shot
Here’s where things get interesting. AI has been hyped as healthcare’s savior for years—remember IBM’s Watson Health launch a decade ago? But most AI efforts have focused on clinical diagnosis, where the FDA has approved over 1,000 algorithms (777 in radiology alone).
The problem with this approach? A radiologist now needs help deciding which of those 777 algorithms to use. We’re solving complexity with more complexity.
The real opportunity lies elsewhere: administrative tasks. Between 25–30% of healthcare costs are pure administrative overhead. This isn’t care—it’s paperwork, scheduling, information retrieval, and inbox management.
A typical doctor sees 24 patients per day in 15-minute appointments. But they’re not spending those 15 minutes focused on you. They’re scrambling to find your information in a system that wasn’t designed to surface it quickly.
The Path Forward
The solution isn’t replacing doctors with AI. It’s using AI to handle the administrative burden that’s burning out healthcare workers and inflating costs.
Imagine a system where your doctor walks into your appointment already knowing your complete history—not because they spent an hour digging through records, but because AI surfaced the relevant information automatically. Imagine eliminating the redundant questions (“What medications are you on?” when they prescribed those medications).
This isn’t science fiction. It’s achievable with technology we have today. The question is whether the healthcare system can overcome its structural incentives to actually implement it.
The beef in AI and healthcare isn’t about replacing human judgment. It’s about giving humans the time and information they need to actually use that judgment well.
This post is part of The Clarity Protocol, ThetaRho’s ongoing series on AI, clinical workflow, and healthcare data. Next up: we digitized healthcare with a $36 billion federal bet on Electronic Health Records — and spending kept climbing anyway. The story of why is more uncomfortable than most people want to admit.