Hey friends, If you're new here, the last two issues covered medical billing 101 (how a $20,000 ER bill becomes a $2,400 bill) and health insurance 101. The throughline is Biscuit, a corgi who sprained his ankle chasing a squirrel and has been learning how American healthcare works ever since. Today he finds out that the same insurance company that paid for his ER visit just said no to his MRI, and the reason is buried in a document almost nobody reads. Whether you just started a job and your benefits packet is making your eyes glaze over, whether you've been picking the same plan for fifteen years and never really understood it, whether you're a CEO trying to pick a plan that won't make your team miserable, or whether you're just curious about how this all works. You're in the right place. By the end of this issue you'll know a word most people never learn, and that word will explain about 80% of the weird letters you've ever gotten from your insurance. -Shivang
tldr: The word is payer policy. It's the rulebook your insurance company writes for itself. It exists for real reasons (fraud prevention, evidence-based care, cost control) and creates real friction along the way. By the end you'll know what's in one, why it exists, and the one question to ask before any non-emergency procedure that can save you a denial letter.
A note on names: Since this newsletter is told through a cast of dogs, the big private insurance companies in our stories get dog-friendly stand-in names too. Let's call them Pawferred Health, Furtna Mutual, Houndmark Health Group, and Bonewell Plan. Medicare and Medicaid are real US government programs, so those keep their real names. BHC is our generic stand-in payer, and it's the one denying Biscuit's MRI today.
Basics first
cheatsheet
Payer: The company paying for your care. Your health insurance company. The word "payer" is what people in healthcare actually use, because not all payers are insurance companies (the government is a payer, and so is your employer if they self-fund their plan). When you hear "payer," think "whoever's writing the check."
Payer policy: A document the payer writes that says when it'll pay for a specific procedure, drug, or test, and under what conditions. There are several kinds, and the one we care about today is the clinical policy bulletin. It does three things: it spells out the coverage criteria (what's covered, what isn't), it shows the "why" behind those criteria (the clinical guidelines it leaned on and, often, what Medicare pays for the same thing), and it ties everything to codes (the ICD diagnosis codes and the CPT procedure codes). Then, halfway through, it points you to three other policies you also have to read. That last part is why it gets confusing.
Medical necessity: The phrase the payer uses to decide if care is needed. In practice it means "the doctor's documentation matches what the policy requires for this procedure," which we'll get into in a minute.
Prior authorization: A permission slip. Some procedures require the doctor to call the payer in advance and get a yes before doing the thing. We'll do a whole issue on this one down the road.
The letter that didn't make sense
A month after the ER visit (the one for the ankle he rolled chasing that squirrel, see Medical billing 101), Biscuit's ankle still hurts. He goes back to his doctor, Dr. Singh, who orders an MRI. Two weeks later, a letter arrives from BHC, Biscuit's insurance company.
Denied. Reason: not medically necessary.
Biscuit reads it twice. A doctor, a real one, with a degree, just said the MRI was necessary, so how can BHC, who has never met him, say it isn't?
Here's what nobody told Biscuit. BHC didn't decide his MRI wasn't necessary in the everyday sense of the word. They decided his paperwork didn't match a rule in a 47-page document on their website. That document spells out, in advance, the situations where an MRI for an ankle injury is the right call versus the situations where less expensive options should be tried first. Dr. Singh's note was clinically accurate, but it didn't happen to include the specific details the document was looking for, so the claim got flagged.
That document is the payer policy, and it's running the show.
Why payer policies exist (the honest version)
It's easy to assume payer policies exist to deny care, but that's not really why they were invented. The actual history is more interesting.
In the 1960s, when Medicare launched, the federal government suddenly had to pay for healthcare for tens of millions of older Americans. They needed a way to answer a basic question: how do we know the care we're paying for actually happened, was actually needed, and matched the standard of medicine at the time? The answer was to write it down, including what they cover, when they cover it, and what the documentation has to look like.
Private insurers followed suit through the 1970s and 80s for three specific reasons.
Fraud prevention. Without a rulebook, anyone could bill for anything. Real fraud (phantom procedures, upcoded visits, fake patients) has cost US healthcare billions over the decades. Policies give payers a way to flag claims that don't fit normal patterns of care.
Evidence-based care. Medicine changes fast, and a procedure that was state of the art in 1995 might be obsolete in 2010, replaced by something safer or cheaper. Policies are how payers say "we'll cover what the evidence supports right now." When a new clinical study comes out showing physical therapy works better than surgery for a certain back injury, the policy gets updated to reflect that.
Cost control. This is the part nobody loves talking about, but it's real. Healthcare costs in the US rise faster than inflation almost every year, and payers use policies to steer patients toward less expensive options first (the physical therapy before the MRI, the generic before the brand name). Sometimes this is good medicine, and sometimes it adds friction for patients who would have been better off skipping straight to the more expensive option.
All three of these are doing real work. The system isn't a villain, it's a system designed to do three jobs at once (catch fraud, follow evidence, manage cost), and the friction patients feel is what happens when those three jobs run into a real human with a real injury.
The catch is that policies change constantly. Medicare updates its rulebook quarterly, and commercial payers (Pawferred, Furtna, Houndmark, Bonewell, and the rest) update theirs whenever new evidence comes in or budgets shift. Last year's rule might not be this year's rule, and nobody emails your doctor about it. Hence, we started Converus, where you can track all relevant policies in one place.
So what is a payer policy, actually?
Strip the jargon and a payer policy is simply a set of rules a payer writes for itself, published so its own claims department knows what to pay for and under what conditions. The trap is assuming there's a single rulebook, when in reality there are several of them living in different places, and the one your denial letter points to is rarely the one that actually holds the answer you're looking for.
Medicare NCDs and LCDs: Medicare's own rulebook comes in two layers, where a National Coverage Determination (NCD) is a rule CMS sets for the whole country and a Local Coverage Determination (LCD) is written by the regional contractor that processes claims in your area, which is why the same procedure can be covered in one state and denied in the next. These determinations reach well beyond Medicare itself, because many commercial insurers borrow them as the starting point for their own policies, and Medicare Advantage plans are required to follow them outright.
Clinical policy bulletins (also called coverage policies or medical policies): This is the commercial version and the document this issue cares about, since it defines medical necessity for a specific procedure by spelling out who qualifies and what has to be tried before the payer will pay. Every company brands it differently, where Pawferred calls it a clinical policy bulletin and Houndmark calls it a coverage policy, but underneath the labels it's the same animal.
Administrative and reimbursement policies: These never ask whether you needed the care and instead govern only how the claim gets billed and paid, through the bundling and modifier rules that decide what a payer will actually reimburse, which means a claim can clear medical necessity and still die right here on a technicality nobody thought to check.
Pharmacy and formulary policies: This is the drug rulebook that sets which medications are covered and at what tier, and it's where step therapy lives, the rule that forces you to fail the cheaper drug before the payer will cover the one your doctor actually wanted.
All four matter, but we're staying with the clinical policy bulletin for the rest of this issue, because that's the document sitting behind Biscuit's "not medically necessary" letter and behind most of the ones that have ever landed in your mailbox.
The maddening part is that even the clinical policy bulletin doesn't tell you the whole story, because whether Biscuit's MRI needs prior authorization, and how to actually obtain it, is almost never written in that policy and instead lives in a login-only portal or in the rulebook of an outside vendor the payer hired to make the call. Many insurers route their MRI approvals to a company like eviCore, which means the criteria Dr. Singh has to satisfy are sitting on eviCore's website while BHC's own policy never mentions them at all. One document tells you the MRI is covered, and a second document in an entirely different website decides whether Biscuit gets it this week, which is a bit like a menu that lists the steak while someone you can't see in the back decides whether the kitchen will actually cook it tonight.
So the next time someone tells you to just go read the policy, the honest reply is to ask which policy they mean, and whether they've managed to track down all four of them yet.
What's inside a payer policy
Crack one open and you'll find the same skeleton every time.
What it covers: The specific procedure, by code (remember CPT codes from medical billing 101? this is where they live).
Who it covers: Age, diagnosis, medical history.
The criteria: "We'll pay for an MRI, but only after six weeks of physical therapy didn't work." This part is called criteria and it's where most denials come from.
Exclusions: Stuff they will never pay for, no matter what.
Why does the rule exist? Underneath the criteria sits the evidence: the medical studies and guidelines the rule is built on, often citing the same Medicare determinations we just walked through.
For Biscuit's MRI, the killer was the criteria section. BHC's policy said the doctor had to document either "six weeks of failed conservative treatment" or "a specific finding on physical exam called a positive anterior drawer test." Dr. Singh's note, written for clinical purposes, didn't include those exact details, so the mismatch became "not medically necessary."
This is the trick to that phrase. Medical necessity in payer language isn't a doctor's gut call, but a contract term that means "the documentation in the chart satisfies the criteria the policy lays out for this procedure." Those are two very different things, and the whole game lives in the gap between them.
The four flavors of payer (& how their policies differ)
If you've ever wondered why your friend at a different company has totally different coverage, this is part of the answer. There are four kinds of payers, and each one writes policies under different pressures.
Commercial insurers (the big private companies, like Pawferred Health, Furtna Mutual, Houndmark Health Group, and Bonewell Plan in our universe). They write their own payer policies and can update them when new evidence comes in or when costs need rebalancing. They're accountable to shareholders and to state regulators.
Medicare is the federal plan for people 65 and up. Policies are public, slow to change, and require formal evidence reviews. The pressure here is statutory, since Congress and CMS define what "reasonable and necessary" means, and the rulebook moves slower because the process is more formal.
Medicare Advantage is the plot twist. These are private insurers (yes, the same dog-name companies) paid by the government to run your Medicare benefits. They have to cover what Medicare covers, but they get to bolt on their own prior auth and their own criteria. In practice that often means more hoops than traditional Medicare, which is why a lot of the "my MA plan denied me" stories you've read follow the exact pattern Biscuit just lived through.
Medicaid is the state-by-state plan for lower-income people, and it looks different in every state. Florida's Medicaid rulebook can look nothing like California's, because each state's budget and politics shape its policies.
Self-insured employer plans. If you work at a company big enough to "self-fund" its health plan, your employer technically writes the rules but hires a commercial insurer to administer them day to day. The rulebook usually looks like that insurer's standard one with a few employer-specific tweaks on top.
If you're a CEO picking a plan for your team, here's the part to pay attention to. Two plans from the same commercial insurer can have completely different rulebooks, and the cheaper plan often has stricter criteria, which means more denials, which means more time your employees spend on the phone instead of working. The cost of that friction shows up in productivity, not on the premium line.
How doctors actually deal with payer policies
For most doctors, dealing with payer policies is genuinely hard.
In a small clinic, Dr. Singh has maybe eight different payers he bills (Pawferred, Furtna, Houndmark, Bonewell, Medicare, Medicaid, plus a few smaller ones), and each one has its own payer policy for every procedure he orders. There's no good way to keep up, so he writes his notes the way he's always written them and finds out about new criteria when the denials show up three weeks later. He isn't doing anything wrong, the system has just moved underneath him. And that's precisely how Biscuit got denied: the "positive anterior drawer test" line BHC added wasn't on Dr. Singh's radar, so it wasn't in the note, so a perfectly good MRI bounced.
To be clear, the fix is never to write something that didn't happen. The fix is to document everything that did, in the language the policy can recognize. Knowing what the policy is asking for helps doctors and billing teams capture the full picture of the visit, and the same exam ends up fully documented instead of partially documented.
In a big hospital, there's usually a team whose entire job is reading payer policies and updating the templates doctors use. When Furtna Mutual updates a rule, a new field appears on the screen, the doctor sees a new question to answer, and the denial doesn't happen.
We heard this from hundreds of doctors and billing folks, which is why we built Converus.ai: every payer policy in one place, updated in real time, matched to your specific situation. That's the pitch, and it's the last you'll hear of it this issue.
What you can actually do with this
If you're a patient. Before any non-emergency procedure or test, ask the front desk one sentence:
"Has prior authorization been obtained, and does the payer policy require any specific documentation?"
That single question, asked before the appointment, kills the most common denial there is: the kind where everyone agreed the care was right and the paperwork just missed a detail. The billing team will know exactly what you mean, and honestly, they'll be relieved you asked.
If you're picking a plan during open enrollment. Compare more than premiums and deductibles, because the rulebook costs you as much as the price tag. Pull the payer policy for one or two things you actually use (an MRI, a specific drug, therapy). Google "[insurer name] medical policy [thing]" and read the PDF. Stricter criteria today is friction you'll feel in March.
If you're a CEO or in HR. Ask your broker for the denial rate on the plans you're weighing. They have it. A plan with a 12% denial rate burns more employee time than a slightly pricier plan at 5%, and that cost is real even when it never touches the premium line.
Biscuit got his MRI. Everything the policy wanted had already happened in the exam room: the six-week timeline, the treatments that failed, the physical exam finding. The second note just put all of it on paper. Same visit, same dog, same ankle. Fully documented this time, by the book.
That's the whole game. Know what the policy is looking for, document accurately, and the system does what it's designed to do.
Hit reply. I read everything.
- Shivang
