Clear as mud is what I would hear after attempting to explain something I really understood to my family that they didn't have a clue about. I'm from a small town, and here in Tennessee, we still use that phrase today. I learned a lot about breaking things down so someone could understand them.
When I think about why healthcare is so expensive, the answer is pretty clear to me. I've spent the last 15 years in the front and back office of healthcare, and I'm the kind of person who can't leave a problem alone until I know why it works the way it does. For most people, the whole thing is clear as mud. This is my attempt to explain a portion of it. I want to find a root cause small enough that we can take action on it, because if enough of us fix the pieces we understand, healthcare becomes what it should be: a miracle that is affordable.
To get there, we have to look at some history.
When you could pay cash
Before the 1930s, most people paid their doctor directly. A visit cost a few dollars. A hospital stay was something you could put on a payment plan with the hospital itself. Health insurance barely existed because there wasn't much to insure. Medicine couldn't do that much yet, so the bills weren't that big.
The first real health plan in this country started in 1929 at Baylor Hospital in Dallas. A group of teachers paid 50 cents a month for up to 21 days of hospital care. That idea became Blue Cross. It was a prepayment plan between patients and a hospital. No employer in the middle.
Then came World War II. The government froze wages in 1942 to control inflation, but it let employers add fringe benefits. Companies competing for scarce workers started offering health insurance because they couldn't offer more pay. In 1954 the IRS made it official: an employer could pay for an employee's health insurance and nobody paid tax on it. From that point on, the cheapest way to get health insurance in America was through your job, and the cheapest way for an employer to give you a raise was to give it to you in benefits.
That one tax decision is the reason your health insurance is tied to your employment today. It wasn't designed. It was a wartime workaround that never got undone.
Medicare, Medicaid, and the standard everyone follows
In 1965 Congress created Medicare and Medicaid. For the first time, the federal government became a large buyer of healthcare, and a large buyer needs a standard way to be billed. Medicare defined the claim forms, the procedure and diagnosis codes, the rules for what counted as a covered service, and in 1983 it started paying hospitals a fixed amount per diagnosis instead of whatever the hospital said it cost.
Commercial insurers followed. It was easier to adopt the government's forms, codes, and coverage logic than to invent their own. Today, when a private insurer evaluates a claim, it is largely using a structure CMS built. The 1996 HIPAA law standardized the electronic claim and the electronic remittance on top of that, and now every provider in the country bills every payer in the same language.
That standard is a good thing. It is also the reason a payer can write one rule and apply it to millions of claims at once.
If McDonald's billed like a hospital
To see what that standard actually looks like, imagine McDonald's worked the way healthcare works.
You walk in and tell the kid at the counter you're hungry. He writes down a code for why you came in. That's the admitting diagnosis. In a hospital it's the reason you showed up: chest pain, broken arm, trouble breathing. For a five year old: it's the word for what's wrong with you, except instead of "tummy hurts" it's a code like R10.9.
Then the manager comes out, looks at you, and decides what you actually have. Turns out you're not just hungry, you're also dehydrated and tired. Each of those gets its own code. Those are diagnosis codes, and there are more than 70,000 of them. A broken arm alone has dozens, depending on which bone, which side, and whether this is the first time anyone's looked at it. Five year old version: these are the names of all the different ways you can be sick or hurt, and there's a different number for every single one.
Now they make your food. A burger, fries, and a drink. But they can't write "burger." They have to write a code for the work of making the burger. That's the CPT code, the code for what the doctor or nurse did to you. There are more than 11,000 of those. Five year old version: it's the number for what the doctor did, like a stitch, an x-ray, or a checkup.
The ketchup packets, the cup, the straw, the bag. Those are supplies, and supplies have their own code set called HCPCS. Drugs, wheelchairs, CPAP machines, crutches. Five year old version: it's the number for the stuff they gave you, not the thing they did to you.
Then there's the detail. You wanted the burger with no onions, and it was made by the manager instead of the kid, and it was the second burger in the same order. Each of those facts changes the price, and each gets a two character tag called a modifier. There are a few hundred of them. Five year old version: it's a little note on the number that says "this one was different."
Before the order gets paid, it has to pass the rules. Fries and a drink can't be billed separately when they came in a combo. Two burgers in one order need a note explaining why. Those are NCCI edits, the government's list of which codes can't be billed together, and there are hundreds of thousands of code pairs on it. Five year old version: it's the rulebook that says which numbers aren't allowed to be on the same receipt.
All of that is the drive-through. If you eat in, meaning you get admitted and stay overnight, the whole order gets rolled up differently. Medicare looks at your diagnosis codes and the big procedures, picks one code for the whole stay called a DRG, and pays a flat amount for it. Chest pain is a DRG. Heart attack is a DRG. A heart transplant is a DRG. The supplies, the room, the nursing, most of the little things that would each get their own line at the drive-through, get bundled into that one number whether the hospital used a lot of them or a little. Five year old version: instead of paying for every fry, they pay one price for "dinner," and the restaurant has to make it work. There are around 770 of those, and which one you land in can swing the payment by tens of thousands of dollars, so getting the diagnosis codes right on an inpatient stay matters more than almost anything else on the bill.
On top of all that, the hospital has to put the right place code for where it happened, the right revenue code for which department did it, the right dates, the right provider numbers, and the right insurance ID. Then the claim goes through a clearinghouse to the payer, the payer runs it against its own rulebook, which is not the same as Medicare's rulebook and not the same as the next payer's, and sends back a response with its own set of codes explaining what it paid and what it didn't.
So the McDonald's receipt for a burger, fries, and a drink now has an admitting code, three diagnosis codes, three procedure codes, four supply codes, a handful of modifiers, a place code, a revenue code, and it's been checked against hundreds of thousands of pairing rules and one payer's private policy. If any one of those is off, you don't get paid, and you find out 30 days later.
Nobody in this picture is being stupid. Every code exists for a reason, usually because somebody abused the system before it existed. But stack 60 years of reasons on top of each other and you get a billing process so dense that it takes a trained person to order a cheeseburger, and most of the mistakes that happen are not mistakes about care. They are mistakes about paperwork.
The money
In 1960, Americans spent about $27 billion on healthcare, around 5 percent of the economy. Today it's over $5 trillion, close to 18 percent. Adjust for inflation all you want, that is not the same dollar doing the same work. Part of the growth is real: we live longer and medicine can do things that were science fiction in 1960. Part of it is that almost nobody in the chain spends their own money.
Your employer pays the premium. The insurer pays the provider. You see a copay and maybe an explanation of benefits you don't read. The person receiving the care, the person paying for it, and the person deciding what it costs are three different people, and none of them can see the price. In every other market, that's where prices go up.
Where denials came from
Here is the part I know from the inside.
When I started in revenue cycle, the people who trained me had been doing the work since the 1980s and 1990s. In their telling, a denial was an exception. You sent a clean claim, the payer paid it. Utilization review existed, but the fight over whether a service was covered mostly happened before the service, not after.
That is not the world I've worked in. Today payers deny around 10 percent of claims on the first pass. Across hospitals alone that's hundreds of billions of dollars a year sitting in denied status. Most of those claims are for care that was already delivered and already covered. A large share of them get overturned if someone appeals, which tells you the denial was never about whether the care was appropriate.
The reasons it exploded are not mysterious.
First, the standard. Once every claim arrives in the same format, a payer can run it through automated rules and deny by category. Missing a modifier, a date outside a window, a code pairing the rule doesn't like. No human has to look at it.
Second, the incentive. A claim that doesn't get paid this month is money the insurer keeps this month. A claim that never gets appealed is money the insurer keeps forever. Providers don't have the staff to work every denial, so a meaningful percentage gets written off. Everybody in the industry knows this, and the math works for the payer.
Third, the arms race. Providers hired denial teams, bought software, and outsourced to revenue cycle firms. Payers hired more reviewers and bought better software. Both sides are spending real money fighting over claims for care that already happened. Depending on whose estimate you use, somewhere between 15 and 30 cents of every healthcare dollar goes to administration, and a chunk of that is this fight.
When a CFO tells me denials are a constant headache and it feels like there's no win, this is what they're describing. Not a medical problem. A process problem that grew out of a billing standard, a tax rule, and an incentive nobody chose on purpose.
Some of it is a miracle
I want to be fair about the other side of the ledger, because it's easy to read this far and decide the whole system is a scam. It isn't.
Some of what medicine does now is a miracle, and miracles are expensive for a reason. Think about a heart transplant. A team opens your chest, takes out your heart, puts in someone else's, and you go home and live a relatively normal life. Not many surgeons in the country can do that. The ones who can spent 15 years training for it and will do a few hundred in their whole career. What should that cost? How much is your life worth to you? How much is your kid's? There is a real, honest cost to having that surgeon exist at all, and the same goes for a drug that turns a death sentence into a chronic condition after a decade and a few billion dollars of failed attempts. I don't want to pretend otherwise.
The problem isn't that the miracle costs money. The problem is that we've wrapped the miracle in so much friction that we can't tell what the miracle costs and what the friction costs. The surgeon's time is in the bill. So is the three people in a back office fighting with a payer for four months about whether the transplant was authorized.
What we can actually do
Without getting too deep into economics, politics, and government, we now understand some of why healthcare costs so much and where the system is broken. Some, not all. If I tried to list every reason, which might be technically impossible, this would be a book. Maybe I should write one.
So what can we do about it?
The answer isn't in one place. It's spread across a dozen specialties, and it's going to take ambitious people who each want to solve a different hard problem. A pharmacist sees a piece of this I'll never see. So does a nurse running a floor, a hospital CFO, an actuary at a payer, a primary care doctor who spends half her day on paperwork. This is a massive system, and nobody understands all of it. But each of us understands our corner, and each corner can be made better by the person who knows why it works the way it does.
Nothing I've described gets fixed in one move, and I'm suspicious of anyone who says it does. Rewriting the tax code, unbundling insurance from employment, changing how payers are regulated, those are 20-year fights. I'll vote on them. I'm not going to wait on them.
So here's my corner. What I can do is fix the part I understand.
The friction in the back office is a solvable problem. The rules payers use to deny claims are written down. The deadlines are written down. The clinical criteria are written down. The reason providers lose is not that the rules are unknowable, it's that nobody has the staff to apply them to every single claim. That is a labor problem, and labor problems in 2026 have a different answer than they did in 2015.
If every denied claim got worked, on time, with the right evidence, two things happen. Providers stop writing off money for care they already delivered, which means they don't have to raise prices on everyone else to cover the loss. And payers lose the math that makes denying worth it. A denial that always gets appealed and usually gets overturned costs the payer more than paying the claim the first time. The incentive flips.
Same with prior authorization. If an authorization request shows up complete, with the criteria already met and documented, there's nothing to deny. The fight never starts.
That won't make healthcare cheap. It will make one expensive, stupid part of it smaller, and it will do it without waiting on Congress. I spent 15 years watching billing teams work these queues by hand. Now I'm building the thing that works them for you. That's the piece of the mud I can see through, so that's where I'm digging.