ICD-10 specificity for family medicine: hypertension, diabetes, depression.
These three conditions sit on more chronic-care panels and CCM enrollments than anything else a family practice bills, and all three share the same failure mode: a diagnosis that's technically correct passes the first claim and then quietly under-supports everything that follows. This guide goes past the code list and into the combination-coding rules, the "use additional code" requirements, and the documentation elements a reviewer actually checks for each.
Key takeaways
- Hypertension and CKD are presumed related by ICD-10-CM's own rules — no need to write "due to" for the combination code to apply, and most practices under-code this without realizing the presumption runs the opposite way from what they'd expect.
- Diabetes complication codes need a second code, not just a better first one. A Z79.- medication-status code and the specific N18.- CKD stage are both required additions the E11.- code alone doesn't cover.
- Depression coding has no presumption to lean on. Episode pattern, severity, and remission status all have to be documented explicitly, or the claim defaults to F32.A with none of that detail on record.
- The claim that denies isn't usually the one that exposes an unspecified code. It's the CCM enrollment or the higher E/M level three months later that actually needed the specificity the original visit skipped.
Why these three conditions specifically
Hypertension, type 2 diabetes, and depression are the three chronic diagnoses most likely to appear on a CCM enrollment, drive an E/M level under the MDM pathway, and get carried forward, unchanged, across dozens of visits — nobody revisits the diagnosis code once it's on the problem list, so whatever specificity the first documentation captured tends to stay there for years. Each follows a different coding logic. Hypertension and diabetes use combination and complication codes governed by ICD-10-CM's causal-relationship conventions — get the convention right and the code is often mechanical. Depression has no such convention: episode count, severity, and remission status all have to be pulled from the note by name, or the default is an unspecified code with none of that detail behind it.
Hypertension: I10 through I13 and the presumed-causal-link rule
I10 (essential hypertension) is correct and complete for uncomplicated hypertension with no heart or kidney involvement. The moment either shows up in the chart, ICD-10-CM's Official Guidelines direct a combination code instead — and the rule that trips practices up is that the guidelines presume hypertension caused the heart or kidney disease whenever both are documented in the same patient, with no need for the physician to write "due to" anywhere in the note. That's the reverse of how most conditions link in ICD-10-CM, where causation has to be stated. Here it's assumed unless the record documents a different, unrelated cause.
| Presentation | Code | Notes |
|---|---|---|
| Hypertension alone | I10 | No heart or kidney involvement documented |
| Hypertensive heart disease, no heart failure | I11.9 | Causal link presumed — any documented heart disease qualifies |
| Hypertensive heart disease with heart failure | I11.0 | Add a heart failure type code (I50.-) when known |
| Hypertensive CKD, stage 1–4 or unspecified | I12.9 | Add the specific N18.- stage code |
| Hypertensive CKD, stage 5 or ESRD | I12.0 | Pair with N18.5 or N18.6 |
| Hypertensive heart and CKD, no heart failure, stage 1–4/unspecified | I13.10 | Never report I11 and I12 separately here |
| Hypertensive heart and CKD, no heart failure, stage 5/ESRD | I13.11 | Pair with N18.5 or N18.6 |
| Hypertensive heart and CKD, with heart failure, stage 1–4/unspecified | I13.0 | Add a heart failure code and the N18.- stage |
| Hypertensive heart and CKD, with heart failure, stage 5/ESRD | I13.2 | Add a heart failure code and N18.5/N18.6 |
The CKD stage code has its own trap. N18.3 ("chronic kidney disease, stage 3") is a category header, not a billable code — it exists only to organize N18.30 (stage 3 unspecified), N18.31 (3a), and N18.32 (3b) beneath it. A claim built on the bare header fails code-validity edits before medical necessity is even evaluated, and it's an easy mistake when a superbill was set up against an older, less granular list.
Sequencing rule: when heart disease and CKD are both present with hypertension, code I13 — never I11 and I12 together for the same patient. I13 already accounts for both; reporting I11 and I12 alongside it duplicates what the single code covers and reads as a coding error, not extra specificity.
Run a problem-list audit for patients coded I10 who also carry an active CKD or heart failure diagnosis elsewhere in the chart. That combination is the single most common sign the presumed-causal-link rule was missed — and it's a mechanical fix once found, since no new documentation is required.
Type 2 diabetes: complication coding and the required second code
E11.9 (without complications) is correct only when no diabetic complication is documented anywhere in the encounter. The moment one shows up — kidney, eye, nerve, or circulatory — ICD-10-CM requires the specific complication code, and in several high-volume scenarios, a second code as well.
| Category | Code | Additional code required |
|---|---|---|
| No documented complication | E11.9 | None |
| With hyperglycemia | E11.65 | None |
| With diabetic nephropathy | E11.21 | None |
| With diabetic chronic kidney disease | E11.22 | Yes — the specific N18.- stage code |
| With diabetic polyneuropathy | E11.42 | None, unless pain is documented (add G89.-) |
| With diabetic peripheral angiopathy, no gangrene | E11.51 | None |
| With diabetic peripheral angiopathy, with gangrene | E11.52 | None |
The complication codes only capture the disease process — they say nothing about how the diabetes is managed, which is where the second required addition comes in. On long-term insulin, ICD-10-CM requires Z79.4 as an additional code; on long-term oral hypoglycemic therapy, it's Z79.84. Neither replaces the E11.- diagnosis, and both matter to E/M leveling, since prescription drug management supporting moderate-complexity MDM is exactly what the additional code confirms.
The same causal logic that applies to hypertension and CKD applies here: diabetic kidney disease is coded as a manifestation of the diabetes (E11.22, plus the N18.- stage) rather than two unrelated conditions. A CCM claim built on E11.9 and I10 alone, with no complication codes and no Z79 status code, reads as two stable, low-acuity conditions even when the actual picture — nephropathy, insulin dependence, retinopathy — is considerably more complex.
Depression: episode, severity, and remission — no presumption to lean on
Depression coding works differently from the two conditions above. There's no combination-code convention doing the work — every element of specificity has to come from what the physician documents. The F32/F33 split carries real meaning on its own: F32 is a single episode, F33 is recurrent, and using the wrong family misstates the patient's psychiatric history, not just its severity.
| Presentation | Code |
|---|---|
| Not yet established as MDD, or episode pattern unknown (e.g., positive screen, first documentation) | F32.A Depression, unspecified |
| Single episode, unspecified severity | F32.9 |
| Single episode, mild / moderate / severe without psychotic features / severe with psychotic features | F32.0 / F32.1 / F32.2 / F32.3 |
| Single episode, in partial or full remission | F32.4 / F32.5 |
| Recurrent, unspecified severity | F33.9 |
| Recurrent, mild / moderate / severe without psychotic features / severe with psychotic symptoms | F33.0 / F33.1 / F33.2 / F33.3 |
| Recurrent, in partial or full remission | F33.41 / F33.42 |
F32.A is a legitimate, billable code — not a placeholder to avoid — but it's meant for a narrow moment: a positive depression screen (billed with Z13.31, encounter for screening for depression, before diagnosis is confirmed) or an initial visit where "depression" is documented without yet establishing episode count or severity. A chart that stays on F32.A for years after the diagnostic picture is clear is under-coding the same way an unspecified hypertension or diabetes code would be.
Remission status matters operationally, not just diagnostically. F33.42 (recurrent, full remission) supports a materially different treatment picture than F33.1 (recurrent, moderate) — carrying a severity code forward on a patient who's actually stabilized overstates acuity as much as an unspecified code understates it. Update remission status at follow-up visits as the clinical picture supports it.
Do and don't
- Use I12/I13 whenever hypertension and CKD or heart disease coexist — the causal link is presumed, no "due to" language required.
- Add the specific N18.- CKD stage code every time I12, I13, or E11.22 is billed; never report the N18.3/N18.4/etc. category header alone.
- Append Z79.4 or Z79.84 whenever a diabetic patient is on long-term insulin or oral hypoglycemic therapy.
- Update depression severity and remission status at follow-up visits as the clinical picture changes, not just at initial diagnosis.
- Don't report I11 and I12 separately for a patient who has both hypertensive heart disease and hypertensive CKD — use the I13 combination code.
- Don't leave a diabetic complication off the code just because it wasn't the visit's chief complaint — if it's documented anywhere in the note, code it.
- Don't leave F32.A on the problem list indefinitely once episode pattern and severity are clinically established.
- Don't assume a symptom-only or unspecified code will hold up on a CCM or chronic-condition E/M claim the way it might on a first visit.
Not sure your family medicine diagnosis coding is specific enough?
We'll audit a sample of your hypertension, diabetes, and depression charts for combination-code and specificity gaps, and show what's recoverable on E/M and CCM claims.
Frequently asked questions
Does hypertension have to be documented as "due to" chronic kidney disease to use the I12 or I13 combination code?
No — ICD-10-CM's official coding guidelines direct coders to presume a causal relationship between hypertension and chronic kidney disease whenever both conditions are documented in the same patient, even without the physician linking them explicitly with "due to" or "secondary to." That's the opposite of how most combination coding works, and it's the single most commonly missed rule in family medicine hypertension coding. The presumption only breaks if the record explicitly documents a different cause for the kidney disease; otherwise, code I12 (hypertensive CKD) or I13 (hypertensive heart and CKD) rather than I10 and N18.- reported separately, and add the specific N18.- stage code every time.
What's the difference between F32.A and F32.9 for a depression diagnosis?
F32.9 (major depressive disorder, single episode, unspecified) is a fully qualified MDD diagnosis missing only its severity — it belongs in the F32 single-episode family and implies the physician has already established this is a first, single depressive episode rather than a recurrent pattern. F32.A (Depression, unspecified) is a broader code for when the record doesn't yet establish whether the presentation is MDD at all, or whether it's a single or recurrent episode — the common real-world case is a positive depression screen or an initial visit where the physician documents "depression" without yet confirming DSM criteria or episode history. Use F32.A when the diagnostic workup isn't complete; move to the specific F32.- or F33.- code once episode pattern and severity are documented.
Do we need Z79.4 or Z79.84 on every diabetes claim?
Only when the patient is on long-term insulin or oral hypoglycemic therapy, and only as an additional code alongside the E11.- diagnosis — not a replacement for it and not required on every diabetes claim. Z79.4 (long-term use of insulin) and Z79.84 (long-term use of oral hypoglycemic drugs) tell a reviewer the treatment intensity behind the diagnosis, which matters most on E/M leveling and CCM claims where medication management is part of what justifies the code. Omitting it doesn't usually deny a claim outright, but it's one more piece of complexity documentation a reviewer expects to see on an audited chronic-condition claim.
Verify before billing. CPT is a registered trademark of the American Medical Association; codes here are paraphrased, not reproduced from the CPT Professional edition. CPT, HCPCS and ICD-10 codes, coverage policy, and bundling edits change, including annual code-set updates. This page reflects standard industry practice and is provided for general education — it is not a substitute for your own compliance review, your current payer contracts, or the current-year code sets. Confirm requirements against your specific payer mix before submitting claims.