Every Dollar Goes Through Coding: The Case for Revenue Assurance

If you work in the healthcare revenue cycle, I don’t need to tell you that denials get a lot of attention. Especially in light of payers using AI for denials, possibly without human review. It makes sense that CFOs and revenue cycle leaders are spending a significant amount of time and money fighting denials on the back end. While in parallel, costs are climbing faster than revenue.

Denials may be the loud and overt symptom of revenue leakage; however, the bigger issue is that organizations are leaking revenue in places that they can’t see. Take undercoding: it’s much quieter and more difficult to pinpoint than denials, and most organizations don’t have the data intelligence to see where it’s happening.

I’m the director of product management at Arintra, where I lead strategy for our autonomous coding platform expansion across specialties, analytics, and clinical documentation improvement (CDI). I want to walk through what we’re seeing in our customer organizations now that they’re using autonomous coding at scale and what it means for revenue leakage.

Let me set this framework up. Every dollar a health system or provider group bills has to go through coding. But not every dollar goes through denials. Coding sits at the middle of the revenue cycle and impacts prior authorization upstream, CDI alongside it, and denials and audits downstream. Building on that framework, I want to make the case for a different way of thinking about all of this: revenue assurance is the proactive layer that catches leakage before it becomes a denial in the first place.

Denials Are the Loud Problem While the Quiet Ones Are Costing You More

Denials are very visible. They show up on analysis reports, they have a dollar value attached, and in most organizations, there’s a person or a team dedicated to fighting them. So they get a lot of attention, which they should.

Let’s turn to an example of a quieter form of leakage. Undercoding doesn’t get as much attention as denials. What usually happens is that the claim goes out, the payer pays it, and so it’s settled. There’s nothing to review. The only way you’d notice something was off is if you checked the chart and somehow noticed that the level of care the provider offered was higher than the level of care that got billed. Almost no organization is doing that systematically because manual chart review at scale isn’t feasible.

The structural reasons for undercoding aren’t difficult to fathom:

1. Providers aren’t coders. They are trained to take care of their patients, not convert clinical language into financial language.

2. Most charts, especially in high-volume settings, aren’t processed or reviewed by professional coders.

3. There’s no insight into documentation at scale. Most clinical documentation improvement (CDI) programs involve reviewing a small subset of charts periodically.

4. And then there’s fear. A lot of providers undercode because they’re worried about defensibility. If a payer pushes back on something they’ve coded, they don’t have a clear audit trail to point to. So they take 80 cents on the dollar rather than risk losing the whole dollar to a denial.

That fear-based undercoding is real, and it's costly. Vanova Health, one of our MSO customers, ran into exactly that pattern across their network of independent physician practices in Northern New Jersey. Their providers were delivering high-level care, documenting it thoroughly, and then coding conservatively (defaulting to a lower-level code) because they didn't feel confident that they could defend anything higher. 

Fear is a real issue, but only part of the problem. The bigger picture is that this kind of leakage is mostly invisible at the organizational level because the data to see it hasn't existed. Until you have oversight of every chart and can trace every code back to the documentation that supports it, you don't know where the leaks are.

Coding Sits at the Center of the Revenue Cycle

People tend to think of coding as a simple mid-cycle handoff, important but limited in scope. I'd argue that it's the foundation of the entire revenue cycle, and what happens at every other step depends on coding.

Upstream, prior authorization depends on coding. If the authorization doesn't match what gets coded after the visit, that mismatch becomes a denial.

Alongside coding, CDI depends on it as well. The whole point of CDI is to make sure that clinical documentation captures the full picture of the care a provider delivers and supports accurate coding. If you don't have visibility into how documentation maps to codes across hundreds of providers and departments, your CDI program is working from samples and educated guesses.

Downstream, denials and audits are deeply connected to coding. When a payer pushes back on a claim, the counterplay is grounded in the coding logic and the documentation that supports it. If you've got a clear, traceable record of why a code was assigned, you've got something to fight with. If you don't, the claim gets written off or sits in a queue waiting for someone with bandwidth to dig in.

That's why I think of coding as the heart of the revenue cycle. What gets authorized, what gets billed, how much gets paid, and what gets denied all depend on it.

Revenue Assurance: A More Proactive RCM Philosophy

Now, I want to talk in more detail about a term that I introduced earlier: revenue assurance.

Revenue cycle management, as historically practiced, is mostly reactive. It involves questions like, can we efficiently process what's being coded? Can we get paid for it? When it gets denied, can we respond to that quickly? That's the work of revenue cycle teams today, and they do it well.

Revenue assurance works a bit differently. This way of doing RCM is more proactive, and it involves:

• Capturing clinical detail upfront, at the point of care, so the documentation supports accurate coding from the start.
• Capturing the full value of care right off the bat, instead of catching what was missed weeks or months later.
• Preempting potential issues before they become denials, downcoding, or written-off dollars. 

The reason revenue assurance hasn't existed as a category until now is that the underlying data didn't exist. You can't preempt a denial if you don't have the clinical data structured well enough to know it's coming. You can't catch undercoding at scale if you can't see how every provider is documenting, not just the ones flagged for review. You can't automate appeals if you don't have the audit trail built in. Autonomous coding gives you all of that, along with the foundation for revenue assurance.

Coding is Where Revenue Assurance Starts

Because autonomous coding can read charts at scale, apply payer-specific rules, and produce a structured record of which clinical evidence supports which codes, you suddenly have the data to:

• Match authorizations to what's actually going to be coded
• Catch and prevent denials before claims go out
• Automate denial appeals with documentation that's already captured and structured
• Surface documentation patterns at the provider and department level so behavior changes
• Improve compliance because every code traces back to the documentation that supports it

I really want to focus on the compliance angle, particularly for CFOs and compliance officers. When your revenue uplift is supported by complete documentation that traces back to every code, that's compliant revenue. Yes, it’s more money, but it’s also auditable money, with the audit trail built in from the start instead of being reconstructed after the fact. For organizations worried about CMS audits or payer scrutiny down the line, that's a meaningful change in posture.

Look Past Denials: Every Dollar Goes Through Coding

If you're a CFO or a revenue cycle leader, you know that the math on denials isn't going to improve on its own. Payers are going to keep getting more aggressive, and the cost of fighting each denial keeps climbing. That work is real and it's not going away.

But there's a bigger question that doesn't get asked enough: how much revenue are you leaking before any claim ever hits a denial queue? How much are you settling for without knowing it?

Those are exactly the kinds of questions revenue assurance helps you answer. And the function that makes it possible is coding, done autonomously, at scale, with a structured record behind every decision. 

Every dollar has to go through coding. Not every dollar has to go through denials. The work to assure revenue starts where the dollar is first translated from clinical work into financial work, and that's the place to invest.

‍