Unspecified vs. Specified ICD-10 Codes: Why "Close Enough" Costs You Money

Aug 5, 2026

Cortney Swartwood

An unspecified code isn't technically wrong, but it's imprecise and incomplete, and incomplete is where a lot of home health reimbursement quietly leaks out of an agency. Here's what that actually looks like on a real chart, and why the more specific code is worth the extra effort every time.

Why unspecified codes happen in the first place

Nobody sits down and decides to under-code a patient on purpose. Unspecified codes usually happen because the documentation in front of the coder doesn't clearly support a more specific option, or because a clinician's note used general language and nobody went back to clarify it before the claim went out. Under time pressure, the unspecified code is the fast option, and it's defensible on paper, but it's still leaving money on the table.

Example one: diabetes with kidney involvement

Say a patient has type 2 diabetes and documented chronic kidney disease. If the note only says "diabetes" without connecting it clearly to the kidney disease, the coder may end up defaulting to E11.9, type 2 diabetes mellitus without complications.

The more specific code, E11.22, type 2 diabetes mellitus with diabetic chronic kidney disease, reflects a meaningfully more complex patient. That specificity affects comorbidity adjustment under PDGM, which affects the case mix weight, which affects the payment for that period. The patient's actual clinical picture didn't change between those two codes. Only what got captured changed, and that gap is real reimbursement, not a rounding error.

Example two: heart failure

A similar pattern shows up with heart failure. I50.9, heart failure unspecified, is a common default when documentation doesn't clarify the type. But a patient with documented chronic systolic congestive heart failure should be coded to I50.22, which is far more specific and reflects a higher level of clinical complexity.

Again, the patient's condition is the same either way, and what changes is whether the documentation and the code actually capture how complex that patient's care really is. Unspecified codes systematically understate that complexity, and PDGM's case mix adjustments are built to reward accurate specificity, not penalize it.

Why this adds up faster than agencies expect

A single unspecified code on a single chart looks small, and it rarely triggers alarm on its own. But unspecified coding tends to happen in a pattern, not a one off. If your documentation habits default to general language in one area, they usually default to general language across a lot of charts, so the gap between what you're coding and what you could accurately be coding shows up consistently across your full census, not just one patient.

That consistency is exactly what makes it expensive. A small per chart gap, multiplied across every relevant chart in a year, adds up to a big revenue loss number, and it's a number most agencies never actually calculate because no single chart looks like a problem by itself. As an example, Olli found a customer that was unintentionally undercoding. Prior to becoming our customer, they were losing roughly $160 in reimbursements per chart. By partnering with Olli, they were able to 1) identify this undercoding pattern, 2) provide more precise coding, and 3) capture $1.5M more in revenue per year that was previously being left on the table.

What actually closes the gap

Closing this gap almost never requires better clinicians. It requires documentation that clearly supports the more specific code when the more specific code is true, and a coding review step that catches the pattern instead of relying on any one coder to remember to double check every note.

That's a genuinely solvable problem, and it doesn't require asking clinicians to write longer notes. It requires the right prompts at the point of documentation, and a second layer that catches unspecified coding before the claim goes out, not after a payment audit finds it.