What "We've Always Billed It This Way" Really Costs

Short answer: ICD-10 codes answer "what's wrong with the patient," CPT codes answer "what did we do about it." You need both, correctly matched, for a claim to get paid.

I get some version of this question a lot from newer practice managers — usually right after their first denial that says something like "diagnosis does not support procedure." It's a confusing rejection to get if nobody ever walked you through how these two code sets actually work together. So here's the plain-language breakdown I wish someone had given me earlier.

The Core Difference

ICD-10 (International Classification of Diseases, 10th Revision) codes describe the diagnosis — the condition, symptom, or reason the patient came in. Something like "type 2 diabetes without complications" or "acute sinusitis" gets translated into a specific alphanumeric ICD-10 code.

CPT (Current Procedural Terminology) codes describe the service — what the provider actually did during that visit. An office visit, an X-ray, a minor surgical procedure, a lab test — each has its own CPT code.

Put simply: ICD-10 is the "why," CPT is the "what." A claim needs both, and — this is the part that trips people up — the diagnosis code has to logically justify the procedure code. If the ICD-10 code doesn't support "medical necessity" for the CPT code billed, the payer has grounds to deny it, even if everything else on the claim is accurate.

This is the foundation of accurate medical coding and it's also the single most common place where claims fall apart.

Where Practices Actually Get Tripped Up

Mismatched specificity. Payers increasingly push back on "unspecified" ICD-10 codes when the documentation supported something more detailed.

Medical necessity mismatches. The big one. If the ICD-10 code doesn't justify the CPT procedure billed, it's an automatic red flag.

Missed modifiers. Bilateral procedures, same-day repeats, multi-provider services — skip the modifier and even correct codes can get denied or mispaid.

Outdated code versions. ICD-10 updates every October, CPT every January. Coding past either date on last year's set means working from an outdated reference without realizing it.

If your team is running into denials tied to diagnosis-procedure mismatches specifically, it's worth reading through why medical coding accuracy matters so directly to reimbursement outcomes — it walks through the mechanics of how these mismatches actually trigger denials on the payer side.

Why This Matters for Medical Coding Billing, Not Just Coding

It's worth separating two things that get lumped together: getting the codes right and getting the claim right. Medical coding billing as a combined function is really about making sure accurate codes make it into a claim that's formatted, documented and submitted the way a specific payer expects. A coder can assign perfect ICD-10 and CPT codes, but if the billing team doesn't attach the right modifiers or format the claim to a payer's specific requirements, it still comes back denied.

That handoff between coding and billing is where a lot of avoidable denials actually originate — not from wrong codes, but from a disconnect in how they're submitted. There's a solid walkthrough of what medical claims processing and billing actually involves end to end if you want to see the full path a claim takes from coding to payment.

A Practical Reference Habit Worth Building

For practice managers who want a quick sanity check before claims go out, this is the short version worth keeping close:

Does the ICD-10 code reflect the most specific diagnosis supported by the documentation?

Does the CPT code accurately describe the service performed — not a close approximation?

Does the diagnosis code logically support medical necessity for that specific procedure code?

Are any required modifiers attached (bilateral, multiple procedures, distinct service)?

Are both code sets current for the date of service?

If your team is catching denials after the fact rather than before submission, it's usually a sign this check is happening too late in the workflow. A closer look at managing claim edits and rejections covers how to move that catch point earlier, before the claim leaves the building.

The Bigger Picture

ICD-10 and CPT aren't competing systems — they're two halves of the same sentence a claim is trying to tell the payer: "here's what was wrong and here's what we did about it." When those two halves don't logically connect, or when either one is outdated or imprecise, the payer has every reason to push back. Understanding the broader role coding plays across the revenue cycle makes it a lot easier to see why getting this specific pairing right has such an outsized effect on reimbursement.

Most practice managers don't need to become coding experts. But knowing this distinction — and knowing which questions to ask when a denial comes back — makes it a lot easier to have a useful conversation with your coding team instead of just forwarding the denial and hoping it gets fixed.

Quick question: does your front desk know a patient's coverage has lapsed before they check in, or after the claim gets denied three weeks later?

If it's the second one, you're not alone — but you're also leaving money on the table without realizing it.

There's a rough industry rule of thumb floating around RCM circles: verify insurance within 48 hours of the appointment, or risk eating the cost later. It sounds almost too simple to matter. It matters a lot.

Here's the thing about coverage — it changes more than people think

Plans lapse. Employers switch carriers. Patients change jobs and forget to mention it. None of that shows up on the intake form from six months ago. It only shows up when someone actually checks — close to the visit, not weeks before it.

Verify too early, and the information's stale by the time the patient walks in. Verify too late (or not at all), and you've already delivered care you might not get paid for.

What actually goes wrong without it

A few things, and they compound:

The claim gets denied for eligibility, not because anything was billed wrong

The practice either eats the cost or has to chase the patient for payment after the fact — which is a much harder conversation than having it up front

Staff time gets burned fixing something that a 5-minute check could've caught

That last denial usually doesn't disappear either — it just shows up later as a growing balance sitting in patient AR, getting harder to collect the longer it sits.

It's not just about coverage — it's about what's actually authorized

Verification confirms a patient has insurance. It doesn't automatically mean a specific procedure is cleared to be billed. That's a separate step — prior authorization — and skipping it (or assuming verification covers it) is one of the most common ways practices get blindsided by an otherwise "verified" patient's claim getting denied anyway.

So why 48 hours specifically?

It's close enough to the visit that the info is current, but early enough to actually fix a problem before the patient shows up — reschedule, collect payment upfront, or get authorization sorted, instead of finding out mid-visit or after.

Practices that build this into a non-negotiable step — not a "when we have time" task — consistently see fewer denials, faster payments, and a lot less awkward billing back-and-forth with patients who thought they were covered.

It's not glamorous work. But it's one of the cheapest fixes in the entire revenue cycle for the amount of money it protects.