30 min read

> "Everything in the book is a convention. The Guidelines are the rules about the conventions. Nobody

In This Chapter

Exercises Quiz Case Study 01 Case Study 02 Key Takeaways Further Reading

Chapter 9: ICD-10-CM Official Guidelines: The Rules That Decide Which Code Is Right

"Everything in the book is a convention. The Guidelines are the rules about the conventions. Nobody tells you that on your first day and it explains almost everything that confused you in your first month." — constructed

Overview

Chapter 8 taught you the conventions printed in the code book. This chapter teaches the document that governs them.

The ICD-10-CM Official Guidelines for Coding and Reporting are approved by the four Cooperating Parties, republished every October 1, and required — they accompany the code set as a national standard under HIPAA (Chapter 5 §5.7), which means a code assigned in violation of them is not unconventional but incorrect.

They are also free, about a hundred pages, and most working coders have never read them front to back.

That combination — binding, free, and unread — is one of the more remarkable facts about this profession, and it creates a genuine opportunity. A coder who spends one afternoon with the Guidelines knows things that people with five years of experience do not, because those people learned the conventions from a book and the rules from colleagues.

This chapter reads the parts that decide real claims. Not all of it — Sections II and III belong to inpatient coding and Chapter 33 returns to them, and the chapter-specific guidance in Section I.C belongs to Chapters 10 through 12. What is here is the machinery that applies to nearly everything: which diagnosis goes first, what to do with a diagnosis nobody has confirmed, when a symptom is coded and when it disappears into the disease, what the word "with" is permitted to assume, and how to resolve two rules that appear to contradict each other.

And one promise, deferred three times, is kept here. Chapter 4 asked whether a coder may link diabetes to chronic kidney disease when the record does not say so. Chapter 7 asked it again. Chapter 8 pointed at §9.7. Section 9.7 answers it.

In this chapter, you will learn to:

  • Describe the four sections and what each governs
  • Select the first-listed and the principal diagnosis
  • Apply the uncertain-diagnosis rule in both settings
  • Decide when a symptom is integral and when it is coded
  • Apply the "with" convention, exactly
  • Report acute and chronic conditions and sequence them
  • Sequence codes and say what sequencing communicates
  • Resolve conflicts between conventions

Learning Paths

🎓 Certification — All of it. §9.3, §9.5, §9.6, and §9.7 are the four most examined ideas in diagnosis coding, and the uncertain-diagnosis rule appears on every credential because the two settings behave in opposite ways.

💼 New Coder — All of it, and §9.7 twice. It is the one convention that permits something Chapter 4 §4.7 appeared to forbid, and understanding why is what separates a coder who applies rules from one who understands them.

💵 Biller / AR — §9.3 and §9.9. A denial for a diagnosis that "does not support the service" is frequently a sequencing problem rather than a code-selection problem.

🏥 Practice Manager — §9.1's observation that the Guidelines are free, binding, and unread is the whole chapter for you. One afternoon of staff time buys a measurable amount of accuracy.

9.1 What the Guidelines are and why they are free

They are a set of rules for assigning and sequencing ICD-10-CM codes, approved jointly by NCHS, CMS, the American Hospital Association, and AHIMA — the Cooperating Parties (Chapter 7 §7.10).

Their status. The Guidelines accompany and complement the conventions and instructions in the code set itself. Adherence to them is required when assigning ICD-10-CM codes, and the Guidelines state so in their own introduction. The HIPAA Transactions and Code Sets rule made ICD-10-CM a national standard, and it brought the conventions and guidelines with it.

Their limits, stated honestly. They do not answer every question. They are written at a level of generality that leaves genuine ambiguity, which is why Coding Clinic exists (Chapter 7 §7.10) and why the answer to a hard question is sometimes "this has been addressed and you have to look it up."

And the sentence that matters most, which appears in the Guidelines' own introduction: the importance of consistent, complete documentation cannot be overemphasized, because without such documentation, accurate coding cannot be achieved. The Guidelines say, in their first page, that they are downstream of Chapter 4. They are a set of rules for coding a record, and they cannot compensate for a record that does not say enough.

Why free matters

Because it removes every excuse.

The CPT book costs money (Chapter 6 §6.1) and that is a real barrier. The Guidelines cost nothing. They are published as a PDF, they are searchable, they are reissued annually, and any coder anywhere can have the current version in ninety seconds.

Which means the gap between a coder who knows the rules and one who does not is entirely a gap of attention. That is worth saying plainly, because it is unusual — most professional advantages are bought, and this one is not.

9.2 The four sections and what each governs

THE OFFICIAL GUIDELINES — four sections

  SECTION I     Conventions, general coding guidelines, and
                chapter-specific guidelines
                ├─ I.A  CONVENTIONS ......... Chapter 8's material,
                │                              in the source's words
                ├─ I.B  GENERAL RULES ....... most of THIS chapter
                └─ I.C  CHAPTER-SPECIFIC .... Chapters 10, 11, 12
                                               (and the largest part
                                                of the document)

  SECTION II    Selection of PRINCIPAL DIAGNOSIS
                └─ INPATIENT only .......... Chapter 33

  SECTION III   Reporting ADDITIONAL DIAGNOSES
                └─ INPATIENT only .......... Chapter 33

  SECTION IV    Diagnostic coding and reporting for
                OUTPATIENT services
                └─ physician offices, clinics, hospital outpatient,
                   emergency departments, ambulatory surgery

  ────────────────────────────────────────────────────────────────
  THE STRUCTURAL FACT THAT MATTERS:

  Sections II and III are INPATIENT. Section IV is OUTPATIENT.
  They are separate sections BECAUSE THEY DISAGREE — and §9.5 is
  the place they disagree most sharply.

Section I applies everywhere. Sections II and III apply to inpatient admissions. Section IV applies to outpatient encounters.

The separation is the point. A newcomer reading the Guidelines linearly hits Section II, learns a rule about principal diagnosis, and applies it in a clinic — where a different section governs and says something different. Know which setting you are in before you look for a rule.

Finding a rule quickly

The document is long and the numbering is the navigation. Three habits make it usable:

Search the PDF, do not scroll it. The Guidelines are published as a searchable file. Searching the condition name, or the word the classification uses, lands you in the relevant chapter-specific subsection in seconds.

Learn the section numbering as an address. I.C.1 is chapter-specific guidance for ICD-10-CM chapter 1 — infectious disease. I.C.9 is circulatory. The number after I.C. is the ICD-10-CM chapter number, which means Chapter 7 §7.2's range table doubles as an index to the Guidelines. That mapping is worth knowing; it turns "where is the rule about hypertension" into "I.C.9."

And cite the address when you record a decision. "Per Section I.C.1.d" is a defensible note. "Per the guidelines" is not, because nobody can check it — including you, in two years.

9.3 First-listed diagnosis in the outpatient setting

Section IV. The rule:

The first-listed diagnosis is the condition, problem, or other reason chiefly responsible for the services provided at this encounter.

Determined by the encounter, not by severity, not by chronicity, and not by what is most interesting about the patient.

What that means in practice

The reason for the visit governs. A patient with heart failure, diabetes, and hypertension who comes in for a wart removal has the wart first. The heart failure is more serious and it is not why they are here.

Chronic conditions may be reported as additional diagnoses where they are addressed, affect treatment, or require management at that encounter. They are not automatically reported because they exist. Section IV addresses this, and it is where Chapter 36's argument will start.

Codes for signs and symptoms are acceptable when no definitive diagnosis has been established. §9.6.

"First-listed," not "principal." Section IV deliberately uses a different term, because the inpatient concept (§9.4) has a different definition and applying it in an outpatient setting produces wrong answers.

⚠️ Where Claims Die

The diagnosis that supports the service is not first.

Sequencing is not cosmetic. Many payer edits and coverage policies evaluate the first-listed diagnosis, and a claim whose supporting diagnosis is in the fourth position can deny for medical necessity while carrying, on the same claim, the code that would have justified it.

This is one of the most maddening denials in the field, because everything needed is on the claim — in the wrong order.

Chapter 25 §25.5 covers diagnosis pointers, which is the mechanism by which a professional claim links each service line to the diagnoses supporting it, and which resolves much of this. But the first-listed diagnosis still carries weight, and the habit is: before submitting, ask whether the first-listed diagnosis is the reason this encounter happened.

Additional diagnoses in the outpatient setting

Section IV's rule on what else gets reported is short and is the source of a great deal of disagreement.

Report additional diagnoses that affect patient care in terms of requiring clinical evaluation, therapeutic treatment, diagnostic procedures, extended length of stay, or increased nursing care or monitoring.

And, correspondingly:

Do not report conditions that were previously treated and no longer exist, or conditions that have no bearing on the current encounter.

The operative question is whether the condition was addressed, not whether the patient has it.

That distinction does the work in three places this book cares about:

A problem list is not a diagnosis list for this encounter. Electronic records carry every condition a patient has ever had, permanently, and a great many of them have nothing to do with today's visit. Coding the problem list is not coding the encounter.

Chronic conditions are reportable when they are managed — evaluated, treated, monitored, or affecting the treatment given. A patient's hypertension addressed and medication continued is reportable. The same hypertension, mentioned nowhere in the assessment, is not.

And this is precisely where Chapter 36's argument begins, because risk-adjusted payment asks a question Section IV does not: not what did this encounter address but what does this patient's documented burden of illness look like across the year. Those are different questions with different correct answers, and the tension between them is real rather than a failure of one rule or the other.

⚠️ Where Claims Die

Coding the entire problem list.

It is the easiest error to make in an electronic record, because the problem list is right there, it is structured, and it looks authoritative. And it produces claims carrying eight or twelve diagnoses where two were addressed.

Three consequences: the first-listed diagnosis becomes harder to identify and may be wrong; the claim describes an encounter that did not happen; and — the one that costs money — a coding pattern of reporting unaddressed conditions is exactly what a risk-adjustment audit looks for (Chapter 36).

The discipline is a question asked of every candidate diagnosis: "where in the assessment or plan was this addressed?" If you cannot point at a line, it does not go on this claim.

9.4 Principal diagnosis in the inpatient setting

Sections II and III. The definition is different and the difference is load-bearing:

The principal diagnosis is the condition established after study to be chiefly responsible for occasioning the admission of the patient to the hospital for care.

"After study." That phrase is the whole difference. The outpatient rule asks what the encounter was for; the inpatient rule asks what — once the workup was complete — turned out to be the reason the patient needed to be admitted.

A patient admitted for chest pain, worked up, and found to have a myocardial infarction has a principal diagnosis of the infarction, not the chest pain. The chest pain is why they came. The infarction is what, after study, occasioned the admission.

This chapter states the definition and stops there. Principal diagnosis selection is a substantial discipline with a body of rules for competing diagnoses, symptoms followed by contrasting conditions, complications, and admissions from outpatient surgery. Chapter 33 §33.2 owns it, because the consequences are a DRG and the DRG is that chapter's subject.

What you need here is the contrast, because it sets up §9.5.

9.5 The uncertain-diagnosis rule, and the two settings that treat it oppositely

Here is the cleanest example in ICD-10-CM of two rules that are both correct and point in opposite directions.

Outpatient — Section IV

Do not code diagnoses documented as "probable," "suspected," "questionable," "rule out," "compatible with," "consistent with," or "working diagnosis." Code instead the documented signs, symptoms, abnormal test results, or other reason for the visit.

Inpatient — Section II

Code a diagnosis documented at the time of discharge as "probable," "suspected," "likely," "questionable," "possible," or "still to be ruled out," or with similar terms indicating uncertainty, as if it existed or was established.

THE SAME DOCUMENTATION, TWO SETTINGS

   The note says:  "Probable pneumonia."

   ┌──────────────────────────────┐   ┌──────────────────────────────┐
   │  OUTPATIENT (Section IV)     │   │  INPATIENT (Section II)      │
   ├──────────────────────────────┤   ├──────────────────────────────┤
   │  DO NOT code pneumonia.      │   │  CODE pneumonia — as if it   │
   │                              │   │  existed.                    │
   │  Code the documented signs   │   │                              │
   │  and symptoms instead:       │   │  (Documented at the time of  │
   │  cough, fever, shortness     │   │  DISCHARGE.)                 │
   │  of breath.                  │   │                              │
   └──────────────────────────────┘   └──────────────────────────────┘

   BOTH RULES ARE CORRECT. THEY ARE ANSWERING DIFFERENT QUESTIONS.

Why they differ — and this is worth understanding rather than memorizing

The outpatient rule protects the patient and the record. An office visit is a snapshot. A "probable" diagnosis coded as established puts a disease on a person's permanent record on the strength of a guess — with consequences for their insurability, their future care, and the accuracy of population data. And the outpatient encounter usually has documented symptoms to code instead.

The inpatient rule captures the workup. A patient admitted and investigated for five days consumed enormous resources pursuing a diagnosis. If the workup ends inconclusively and the record says "probable," coding only the symptoms would describe a five-day admission as though nothing had been determined — and would systematically understate what hospitals do. The rule captures the diagnostic intent and the resources it consumed.

Notice the shared logic: each rule codes what the setting actually establishes. The clinic establishes symptoms; the admission establishes a diagnostic conclusion, however hedged.

🎓 Exam Watch

This is on every diagnosis credential, and the question is always built the same way: a scenario with a hedged diagnosis, and the setting stated somewhere the candidate may skim past.

Read the setting first. Before you read the clinical facts, find out whether this is an office, a clinic, an emergency department, an ambulatory surgery center — all outpatient, Section IV — or an inpatient admission.

The trap that catches the most people: the emergency department is outpatient. A patient seen in the ED and sent home is an outpatient encounter regardless of how sick they were, and a hedged diagnosis there is not coded.

And a second trap: the inpatient rule says "documented at the time of discharge." An uncertain diagnosis appearing in a day-two progress note and abandoned by discharge is not the same thing.

9.6 Signs and symptoms: when they are integral and when they are coded

The rule has two halves and people remember only the first.

When a definitive diagnosis is established

Signs and symptoms that are integral to the disease process should not be coded separately, unless the classification instructs otherwise.

Cough with pneumonia. Chest pain with a myocardial infarction. Fever with an acute infection. These are part of the disease; coding them adds nothing and clutters the record.

And the half people forget

Signs and symptoms that are NOT routinely associated with the disease process SHOULD be coded when present.

A patient with pneumonia who also has hematuria — not a symptom of pneumonia — has both coded, because the second one is telling you something the first does not.

When no definitive diagnosis is established

Code the symptom. This is not a fallback or a failure; it is the correct answer, and it is explicitly endorsed by Section IV.

And it is exactly what happened on Account 10-4471. The March 14 assessment reads: "Right knee pain — new complaint this visit… no definitive diagnosis established today." The provider declined to name a disease, and the classification's answer to that is a symptom code.

   THE DECISION

   Is a definitive diagnosis established?
     │
     ├─ NO  ──────────────► CODE THE SYMPTOM.
     │                      Correct, endorsed, not a failure.
     │
     └─ YES ──────► Is the symptom INTEGRAL to it?
                      │
                      ├─ YES ──► DO NOT code the symptom.
                      │
                      └─ NO ───► CODE IT TOO. It is telling you
                                 something the diagnosis does not.

⚠️ Where Claims Die

Coding the symptom and the confirmed diagnosis.

A coder codes pneumonia and, wanting to be thorough, also codes the cough and the fever. Both are integral. The claim now carries three codes where one was correct, the extra two add nothing, and in a facility setting the pattern is visible in data as a coder who does not distinguish integral symptoms from reportable ones.

And the opposite error, which is worse: a patient with a confirmed diagnosis has a symptom that is not integral, and it goes unreported because the coder learned only the first half of the rule. That one loses information — and under risk adjustment (Chapter 36) it can lose money.

The test is not "is this a symptom." It is: would you expect this with this disease? If yes, it is integral. If no, it is a separate finding and it is reportable.

9.7 The "with" convention and the linkage it assumes

The promise, kept.

Chapter 4 §4.7 established the rule that a coder may not infer a causal relationship the record does not state. Chapter 7 §7.8 noted that a combination code asserts such a relationship. Chapter 8 §8.8 said "with" means "associated with" or "due to" and deferred the rest to here.

Here is the rest.

What the convention says

The Guidelines' conventions section addresses the word "with" (and "in") as it appears in a code title, in the Alphabetic Index, or in an instructional note in the Tabular. The substance:

The word "with" is to be interpreted to mean "associated with" or "due to" when it appears in a code title, the Alphabetic Index, or an instructional note in the Tabular List.

The classification presumes a causal relationship between the two conditions linked by these terms in the Index or Tabular. These conditions should be coded as related even in the absence of provider documentation explicitly linking them, unless the documentation clearly states the conditions are unrelated.

Read that middle sentence again, because it is a genuine exception to everything Chapter 4 taught.

The classification presumes the relationship. Not the coder — the classification. And that is the resolution of the apparent contradiction: the coder is not inferring anything. The coder is following an instruction that the Cooperating Parties wrote into the classification, which has already made the causal judgment as a matter of classification design.

And the boundaries, which matter as much

1. The presumption applies where the terms appear in the Index or in a Tabular instruction — not everywhere the word "with" appears in a clinical note. A provider writing "diabetes with a knee injury" has not triggered anything. The convention is about the classification's use of the word.

2. The presumption is defeated by documentation stating the conditions are unrelated. If the provider says the kidney disease is not diabetic, that governs. Explicit documentation always beats a presumption.

3. The presumption does not extend to conditions the classification has not linked. You may not generalize it. Two conditions the Index does not join under "with" are not presumed related, and linking them is the inference Chapter 4 forbids.

4. Where the Guidelines require an explicitly documented linkage, they say so. Some chapter-specific guidance requires provider documentation of a relationship, and where it does, the "with" presumption does not override it. Section I.C is where those live, and Chapters 10 and 11 walk the important ones.

   THE "WITH" CONVENTION — what it does and does not permit

   ┌─────────────────────────────────────────────────────────────┐
   │  THE INDEX or a TABULAR INSTRUCTION links two conditions    │
   │  with the word "with" or "in"                               │
   │                        │                                    │
   │                        ▼                                    │
   │  ► The CLASSIFICATION presumes a causal relationship        │
   │  ► Code them as related, EVEN WITHOUT explicit provider     │
   │    documentation linking them                               │
   │  ► UNLESS the documentation states they are unrelated       │
   └─────────────────────────────────────────────────────────────┘

   ┌─────────────────────────────────────────────────────────────┐
   │  Two conditions the classification has NOT linked           │
   │                        │                                    │
   │                        ▼                                    │
   │  ► NO presumption. Chapter 4 §4.7 governs.                  │
   │  ► A link requires PROVIDER DOCUMENTATION.                  │
   │  ► Inferring one is the error, not the convention.          │
   └─────────────────────────────────────────────────────────────┘

   THE DIFFERENCE IS NOT CLINICAL. IT IS WHETHER THE
   CLASSIFICATION HAS ALREADY MADE THE JUDGMENT FOR YOU.

🎓 Exam Watch

Two forms.

Form 1: apply it. A scenario documents diabetes and a condition the Index links under "with," with no explicit linking statement. The candidate must recognize that the combination code applies anyway.

Form 2: refuse to apply it. A scenario documents two conditions the classification has not linked, and the distractor is a combination code. The answer is two separate codes.

The discriminating question is always the same: has the classification linked these, in the Index or in a Tabular instruction? If yes, presume. If no, you need documentation. This is not a question about the clinical facts and candidates keep trying to answer it clinically.

⚖️ Compliance Check

The "with" convention is a presumption, not a license.

It is the most-abused convention in diagnosis coding, and the abuse has a shape: a coder or a risk-adjustment vendor treats it as general permission to link conditions that the classification has not linked, on the theory that the relationship is clinically plausible.

It is not. The presumption exists only where the Index or a Tabular instruction creates it, and extending it beyond that is the coder supplying a clinical judgment — which is Chapter 4 §4.7's prohibition and, at scale in a risk-adjustment context, an enforcement exposure (Chapter 36).

The safe formulation: "Did the classification link these two, or did I?" If the answer is "I did," stop.

The convention's exact language has been revised and clarified more than once since 2015. Read the current Section I.A.15 rather than any description of it, including this one.

9.8 Acute, chronic, and both

Short and frequently gotten wrong.

When the same condition is described as both acute (or subacute) and chronic, and separate subentries exist in the Alphabetic Index at the same indentation level, code both, and sequence the ACUTE (or subacute) code FIRST.

Three conditions on that rule, and all three matter:

  1. The same condition described both ways
  2. Separate subentries existing in the index
  3. At the same indentation level — §8.2's point that indentation is meaning

Where those hold: two codes, acute first.

Where they do not — where the index does not provide separate entries, or where a combination code exists — the ordinary rules govern instead.

A related item: an impending or threatened condition. The Guidelines address these specifically: if the condition occurred, code it as a confirmed diagnosis. If it did not occur, look up the main term for "impending" or "threatened" and see whether the index provides subentries; if it does, use them. If it does not, code the underlying condition or signs and symptoms.

Borderline diagnoses get their own treatment: a diagnosis documented at the time of discharge as "borderline" is coded as a confirmed diagnosis unless the classification provides a specific entry for a borderline condition. Note this is not the same as an uncertain diagnosis — "borderline" is a statement about a value, not about the physician's confidence — and the rule is different.

The order of precedence, which the rule does not state

The acute-and-chronic rule is frequently applied where it does not belong, because coders learn it as "acute and chronic means two codes" and stop.

A combination code, where one exists and fully describes the condition, comes first in the order of operations. Chapter 7 §7.8: where the classification provides a single code carrying both the acute and the chronic state, that code is used and the two-code rule does not arise. J44.1 — COPD with acute exacerbation — is exactly this: one code carrying a chronic disease and its acute state.

So the sequence is:

   "ACUTE AND CHRONIC" — the order of operations

   1. Does a COMBINATION CODE fully describe it?
        YES ──► use it. Done. One code.
        NO  ──► continue

   2. Does the INDEX provide separate subentries for acute and
      for chronic, AT THE SAME INDENTATION LEVEL?
        YES ──► two codes, ACUTE FIRST.
        NO  ──► continue

   3. Code what the index and Tabular actually provide,
      under the ordinary rules.

Step 1 is the one that gets skipped, and skipping it produces two codes where one was correct — which is a small error on a professional claim and a larger one in a facility setting, where the coded picture drives the grouper.

And a documentation note worth carrying: "acute on chronic" is a phrase clinicians use constantly and the classification does not universally provide for. Where the index does not offer separate subentries at the same level, you do not manufacture two codes because the phrase contained two words. Read what the index gives you.

9.9 Sequencing: what goes first and what it communicates

Sequencing is not a formatting preference. It is a statement, and three different audiences read it.

Who reads it What they take from it
The payer's edits Whether the first-listed diagnosis supports the service
The grouper (facility) Which condition drives the DRG or APC assignment — Chapters 33, 34
Anyone reading the record later What this encounter was about

The rules that fix sequence

Several conventions dictate order and they override the general rule:

  • "Code first" — the underlying condition precedes the manifestation (§8.6)
  • Etiology/manifestation pairs — always etiology first (§7.8)
  • Acute and chronic — acute first (§9.8)
  • Chapter-specific sequencing rules — Section I.C, and there are many: sepsis, obstetrics, HIV, neoplasms with chemotherapy, poisoning, and others. Chapters 10, 11, and 12 cover the ones you will meet.
  • Section IV — the first-listed diagnosis is the reason for the encounter (§9.3)
  • Sections II/III — the principal diagnosis is the condition established after study (§9.4)

When nothing fixes it

Where two conditions are equally responsible for the encounter and no rule or instruction directs otherwise, either may be sequenced first. The Guidelines say so explicitly for certain inpatient situations, and the practical outpatient equivalent is: sequence the one that best explains why the patient was seen.

Do not agonize over a genuinely free choice — but do check first that it is genuinely free, because most of the time something fixes it and the coder simply has not looked.

What sequencing does not mean

Three things people read into code order that are not there.

It is not a severity ranking. The first-listed diagnosis is the reason for the encounter, and the patient's most serious condition may be third or absent entirely. A reader who infers severity from position has misread the claim.

It is not a chronology. Codes are not ordered by when conditions began or by when they were diagnosed.

And it is not a statement about which condition the provider cares most about. It is a statement about what this encounter was for, made in a format with room for only one first position.

Why this matters practically: clinicians sometimes object to a sequencing decision on the grounds that the diagnosis listed first is not the patient's main problem. They are frequently right about the medicine and wrong about what the field means. The conversation goes better when a coder can say what the position actually asserts — which is a claim about the encounter, not about the patient.

🔍 Check Your Understanding

  1. A patient with metastatic cancer, heart failure, and diabetes is seen for a routine wart removal. What is first-listed, and what does that position assert?
  2. A problem list carries eleven conditions. The assessment addresses two. How many diagnoses go on the claim?
  3. A clinician objects that your sequencing puts a minor problem ahead of a serious one. What do you say?

(Answers: 1. The wart. The position asserts that it is the condition chiefly responsible for the services at this encounter — not that it is the most serious thing wrong with the patient. 2. Two — report conditions addressed at the encounter, not the problem list. 3. That the first position states what the encounter was for, not what matters most about the patient — and that if the serious condition was addressed, it belongs on the claim as an additional diagnosis, which is a different and legitimate question worth checking.)

9.10 Coding conventions in conflict, and how to resolve them

The hierarchy, which is the most useful single page in this chapter.

WHEN TWO RULES SEEM TO CONFLICT — the order of authority

   1.  ★ A SPECIFIC INSTRUCTION beats a GENERAL rule.
       A chapter-specific guideline in Section I.C governs over a
       general guideline in Section I.B. An instruction printed at
       the code governs over one at the category. The narrower rule
       wins.

   2.  THE OFFICIAL GUIDELINES beat the code book's CONVENTIONS.
       Chapter 8, Case Study 1: an Excludes1 note is a convention;
       the Guidelines' exception to it governs.

   3.  THE TABULAR beats the INDEX.
       The index locates; the Tabular classifies. §8.4.

   4.  EXPLICIT PROVIDER DOCUMENTATION beats a PRESUMPTION.
       §9.7 — a documented statement that two conditions are
       unrelated defeats the "with" presumption.

   5.  AND WHEN NOTHING RESOLVES IT:
       Coding Clinic. Then your MAC. Then a documented decision
       with your reasoning recorded.

   ────────────────────────────────────────────────────────────
   WHAT IS NOT ON THIS LIST: what pays more. It is not a
   tiebreaker and it is not a consideration.

Rule 1 is the one that does the most work, and it is the general form of the lesson in Chapter 7's Case Study 1: a specific instruction beats a general convention, and the only way to know a specific instruction exists is to read the guidance for the code you are assigning.

Rule 5's last clause deserves emphasis. When a question is genuinely unresolvable from the available authorities, the professional answer is not to guess silently. It is to make a decision, write down the reasoning, and be able to produce it. Chapter 8 §8.1 called this the path. An ambiguity you documented is defensible; the same ambiguity undocumented is indistinguishable from carelessness two years later.

🔍 Check Your Understanding

  1. A clinic note reads "probable pneumonia." What do you code, and what would change if this were an inpatient discharge summary?
  2. A patient with confirmed pneumonia also has documented hematuria. Code the hematuria or not?
  3. The Alphabetic Index links two conditions under "with," and the provider has not documented a causal relationship. May you code them as related?
  4. A chapter-specific guideline in Section I.C appears to contradict a general guideline in Section I.B. Which governs?

(Answers: 1. Code the documented signs and symptoms — cough, fever, whatever the record supports — not pneumonia. Inpatient at discharge, you would code the pneumonia as if established. 2. Code it. Hematuria is not integral to pneumonia, and a non-integral symptom is reportable. 3. Yes — the classification presumes the relationship where the Index links the terms, unless documentation states they are unrelated. 4. The chapter-specific guideline. Specific beats general.)

🗂️ The Encounter

🗂️ The Encounter

What this chapter contributes: the rules applied, and one question finally answered.

First-listed diagnosis. Section IV: the condition chiefly responsible for the services at this encounter. Account 10-4471 has four diagnoses and the encounter had two purposes — a scheduled chronic-disease follow-up and a new knee complaint with a procedure.

The claim lists M25.561 first, and here is the reasoning: the procedure performed at this encounter was the knee injection, and the diagnosis pointers (Chapter 25 §25.5) link line 2 to the knee. A defensible argument exists for sequencing the diabetes first, since the visit was scheduled for chronic-disease follow-up — and this is one of the genuinely free choices §9.9 describes. What is not free is that the diagnosis supporting each service line must point to that line.

The symptom code is correct, and now you know why. §9.6: no definitive diagnosis was established, so the symptom is coded. The March 14 assessment says so explicitly. This is not a failure of specificity and not a documentation gap — it is the classification's correct answer to a provider who honestly declined to name a disease.

And Q2, three times deferred, is now answerable.

```text THE DIABETES LINE — E11.9, and what §9.7 does and does not do

WHAT THE RECORD SAYS Assessment: "Type 2 diabetes mellitus - stable. Continue metformin. A1c ordered today." Problem list: "Chronic kidney disease, stage 3a" Linkage: NONE documented, in either direction.

WHAT §9.7 PERMITS The classification links diabetes and chronic kidney disease under "with" in the Alphabetic Index. So the presumption is AVAILABLE: a coder may report them as related WITHOUT an explicit provider statement.

► SO WHY IS THIS FILE CODED E11.9?

Because the "with" convention answers a question about
LINKAGE, and this encounter has a prior question about
WHETHER THE CKD WAS ADDRESSED AT ALL.

Section IV reports additional diagnoses that are addressed,
that affect treatment, or that require management AT THIS
ENCOUNTER. The assessment addresses diabetes, hypertension,
and hyperlipidemia. It does not address the kidney disease.
The problem list carries it; the encounter did not treat it.

A coder who reports E11.22 + N18.31 here has answered the
LINKAGE question correctly and skipped the ADDRESSED question
entirely.

► AND WHY DOES CHAPTER 36 COME BACK FOR IT?

Because "was it addressed at this encounter" and "does this
patient's record support reporting this condition sometime
this year" are DIFFERENT QUESTIONS, and risk adjustment asks
the second one.

That is Chapter 36's whole subject. The March 14 code is not
wrong. It is incomplete as a description of the patient, and
the difference between those two statements is worth money.

```

What this settles. That E11.9 was correct on March 14, and why — which is a different and better answer than "because that is what the assessment said."

What it does not settle. Whether the practice should have done something else — sent a query, addressed the CKD, structured the visit differently. Chapter 36 §36.7 takes that up, and the answer involves the word MEAT.

Open questions, updated:

  • Q2 — Is the diabetes code right? Partially resolved. It is correct for this encounter. Whether it is sufficient is a different question, and Chapter 36 owns it.
  • Q1 (modifier 25), Q5 (the knee), and Q6 (the \$185.00) remain open.

Conclusion

The Guidelines are binding, free, and mostly unread, and that is an opportunity rather than a complaint.

What was decided in this chapter. That the Guidelines govern the conventions and that a code assigned against them is incorrect rather than merely unconventional — and that they say, on their own first page, that they cannot compensate for a record that does not say enough. The four sections, and the structural fact that Sections II and III are inpatient and Section IV is outpatient because they disagree. First-listed diagnosis: the reason for this encounter, not the most serious problem. Principal diagnosis: the condition established after study — deferred to Chapter 33. The uncertain-diagnosis rule, opposite in the two settings, and why each is right for its setting. Signs and symptoms: integral ones are not coded and non-integral ones are — the half people forget. The "with" convention, and the resolution of a contradiction three chapters old: the coder is not inferring, the classification is presuming, and that presumption exists only where the Index or a Tabular instruction creates it. Acute and chronic, and the three conditions that must hold before you report both. Sequencing as a statement read by three audiences. And a hierarchy for resolving conflicts, on which what pays more does not appear.

What remains open. The chapter-specific guidelines — Section I.C, the largest part of the document — which is where most of the hard rules live.

The bridge to Chapter 10. Section I.C is organized by ICD-10-CM chapter, and it contains the rules that decide the cases that actually get argued about: sepsis sequencing, the neoplasm table and its six columns, HIV, the diabetes conventions in operation, and the pain codes. Chapter 10 works the first six body-system chapters with those rules in hand — infectious disease, neoplasms, endocrine, blood, mental health, and the nervous system. It is the first chapter of this book where you code charts rather than examples.

Key Terms

Official Guidelines for Coding and Reporting — the rules for assigning and sequencing ICD-10-CM codes, approved by the four Cooperating Parties, reissued annually, free, and required. (Ch.9)

Section I — conventions, general coding guidelines, and chapter-specific guidelines. Applies in all settings. (Ch.9)

Section II / Section III — selection of the principal diagnosis and reporting of additional diagnoses. Inpatient only. (Ch.9)

Section IV — diagnostic coding and reporting for outpatient services, including physician offices, clinics, hospital outpatient departments, emergency departments, and ambulatory surgery. (Ch.9)

First-listed diagnosis — outpatient: the condition, problem, or other reason chiefly responsible for the services provided at this encounter. (Ch.9)

Principal diagnosis — inpatient: the condition established after study to be chiefly responsible for occasioning the admission. (Ch.9)

Uncertain diagnosis rule — outpatient: do not code probable, suspected, or rule-out diagnoses; code the signs and symptoms. Inpatient: code them at discharge as if established. (Ch.9)

Integral sign or symptom — a sign or symptom routinely associated with a disease process, which is not separately coded when the definitive diagnosis is established. (Ch.9)

Non-integral sign or symptom — one not routinely associated with the documented disease, which is separately coded. (Ch.9)

"With" convention — where the Alphabetic Index or a Tabular instruction links two conditions with "with" or "in," the classification presumes a causal relationship, and they are coded as related absent documentation that they are unrelated. The presumption exists only where the classification creates it. (Ch.9)

Acute and chronic — where the same condition is documented both ways and separate index subentries exist at the same indentation level, code both and sequence the acute first. (Ch.9)

Impending or threatened condition — code as confirmed if it occurred; if not, check the index for "impending" or "threatened" subentries, and otherwise code the underlying condition or symptoms. (Ch.9)

Borderline diagnosis — coded as confirmed unless the classification provides a specific borderline entry. Distinct from an uncertain diagnosis. (Ch.9)

Sequencing — the order of codes on a claim or record, which is a statement read by payer edits, groupers, and future readers. (Ch.9)

Spaced Review

  1. A patient is seen in an emergency department and sent home. The note reads "rule out appendicitis; abdominal pain and nausea." What do you code, and what is the setting trap in this question?

  2. (Chapter 8) A code carries an Excludes1 note. The Guidelines contain an exception to Excludes1. Which governs, and where does that sit in the hierarchy in §9.10?

  3. State the "with" convention in your own words, including both what it permits and the two boundaries that limit it.

  4. (Chapter 4) Account 10-4471's assessment addresses diabetes, hypertension, and hyperlipidemia. The problem list also carries chronic kidney disease, stage 3a. Explain, in two sentences, why the encounter is coded E11.9 — and be precise about which of the two questions involved is doing the work.

  5. Name the three conditions that must all hold before you report both an acute and a chronic code for the same condition.