ICD-10-CM Basics

Before You Start

This page is for learners meeting diagnosis-coding concepts for the first time, and for returning learners who want to check what has changed. It works as a first lesson and as a refresher: if you have studied this before, jump to the freshness habit and the safe research workflow, which are the sections people forget between attempts.

Read Medical Terminology — Recommended before — and Anatomy for Coders — Recommended before — first if decoding a term or the laterality rule is still unfamiliar; index entries and code descriptions are written in decodable terms, and specificity and laterality build directly on the anatomy page. Neither is required — this page defines its own terms — but they make it faster.

Difficulty: Beginner to Intermediate. The page is easy to read, but it asks you to hold two things apart that learners tend to merge: how the system is built (stable) and what the current rule says (changes on a schedule). Nobody memorizes the code set. Professionals look things up every working day, and needing to check a reference is the normal state of this work, not a gap in yours.

If you need a specific term while reading, use the Glossary — Definition.

What this is and why it matters

ICD-10-CM gives you a research method for describing why a patient was seen, not a list to memorize. You will learn what diagnosis codes are for, how a code is put together, how the two-part lookup works, why the official guidelines have the final say, and why a code only means something when you also know its year. That method is what lets you read a code you have never seen and know how you would confirm it — instead of trusting whatever a search box returns first.

Two habits fail almost everyone who skips this page. The first is trying to "learn ICD-10-CM" by memorizing codes from search results. The second is treating the first code-like result as the answer. Both break for the same reasons: codes and guidelines change every year, the alphabetic lookup is never the final step, and adding detail the record does not support is not accuracy — it is an error with compliance consequences. This page replaces memorization with a workflow, and it makes the difference between coding why and coding what was done concrete before you meet the procedure code set.

What you will be able to do

• Explain the purpose of diagnosis coding and tell it apart from procedure and service coding.
• Describe the general structure of an ICD-10-CM code — category, added characters, the placeholder character, and the seventh-character idea — without memorizing any specific code.
• Explain what the Alphabetic Index and the Tabular List each do, and why a code found in the Index is always verified in the Tabular List before use.
• Recognize the major convention types (instructional notes, inclusion and exclusion notes, abbreviations such as NEC and NOS, punctuation conventions) as signals the Index and Tabular List use, and know that their exact meanings live in the official guidelines.
• Explain why specificity and laterality are limited by documentation, and spot an unsupported inference in a fictional note.
• State that the code set and guidelines are updated on a published schedule, and name what a learner or professional verifies before relying on any code or rule.

The simple version

Think of an old library with a card catalog and shelves. To find a book, you first flip through the card catalog to get a rough location — that is the Alphabetic Index, the alphabetical list that points you toward a code. But the card is not the book. You walk to the shelf to confirm you have the right title and to read the notes taped there — that is the Tabular List, the chapter-organized list that carries the full descriptions and the instructions. The sticky notes on the shelf, and some on the catalog card too, are the conventions and instructional notes; the library's rulebook that says how to read them all is the set of official guidelines. And a category is like a shelf section — the broad group a code belongs to before you narrow down.

Here is where the analogy breaks, and the breaks are the point. A real library catalog barely changes; the code set and the official guidelines are updated on a schedule, so the shelf you trusted last year may have moved. The sticky notes can change which book is the right one, so you cannot skip them. Some notes sit on the catalog card itself — the Index carries cross-references you must read before walking to the shelf. And a library lets you pick any book you like, while diagnosis coding lets you pick only the "book" the provider actually documented. The record, not your guess, decides which one is right.

What diagnosis coding is for

ICD-10-CM — the International Classification of Diseases, Tenth Revision, Clinical Modification — is the diagnosis code set used to report why care happened in United States healthcare settings under federal standards. A diagnosis code stands in for a condition, a symptom, a sign, or another documented reason for an encounter, expressed in a standardized form so that many different systems — claims, public-health statistics, research, and quality measurement — can all read it the same way.

The word "coding" covers more than one job, and that is the first thing to separate. Diagnosis coding answers why the patient was seen. Procedure and service coding answers what was done. They use different code sets, and mixing them up is one of the most common beginner errors.

Text alternative: diagnosis codes (ICD-10-CM) say why care occurred; procedure and service codes say what was done. Professional services in any setting are reported with CPT; ICD-10-PCS is used for procedures on hospital inpatient facility claims, while the diagnosis on that same claim is still ICD-10-CM. This resource does not cover ICD-10-PCS beyond this orientation note.

The publisher and maintainer of ICD-10-CM and its guidelines, and the federal adoption behind them, are facts you confirm at the source rather than take on memory: the code set and Official Guidelines for Coding and Reporting are maintained by the Centers for Disease Control and Prevention's National Center for Health Statistics (CDC/NCHS) together with the Centers for Medicare & Medicaid Services (CMS). Confirm the current maintaining organizations and their federal role on those official pages before relying on any statement of who controls the code set.

Recall check: in one sentence each, say what a diagnosis code reports and what a procedure code reports. Which code set answers "why was the patient seen?"

How codes are structured — at a glance

You need enough structure to read a code, not to memorize one. ICD-10-CM codes are alphanumeric and vary in length within a fixed minimum and maximum; the first character is always a letter, and the first three characters together form the category — the broad grouping the code belongs to. Characters after the category add detail such as the affected site, severity, laterality, or the episode of care. Some categories require a seventh character in the final position, and when the characters needed to reach that position are not otherwise used, a placeholder character — the letter X — fills the empty spots so the seventh character lands where the structure requires. Chapters are organized largely by body system or by type of condition, which is why the anatomy you learned earlier helps you navigate.

Here is the shape of a code, with no real code attached to any condition:

Text alternative: a code begins with a letter, forms a three-character category, may add several characters of detail, may use the letter X only as a structural placeholder, and may require a seventh character in specific categories. The illustration shows positions and their jobs; it is not a real code and is not tied to any condition.

Treat this as the pattern to recognize, not a rule to recite. The exact structural description — minimum and maximum length, which positions carry which detail, and where the seventh character applies — is defined by the current official guidelines and the current code set; confirm the wording there rather than against memory or an older reference.

Recall check: what do the first three characters of a code form, and what is the letter X doing when it appears inside a code?

The Alphabetic Index and the Tabular List

Diagnosis lookup is a two-step move, and both steps are required. The Alphabetic Index lists terms alphabetically and points you toward a code. The Tabular List is the structured, alphanumeric list arranged by chapter (largely by body system or condition) that carries the full code descriptions, the instructional notes, and the conventions. The official guidelines direct you to find your term in the Index and then verify the code in the Tabular List — a code is never assigned from the Index alone.

The Index has more than one part, each with its own logic — a main index of terms, a separate index for external causes, and tables such as the Neoplasm table and the Table of Drugs and Chemicals. Name and use them as the current official guidelines describe them, and confirm that description at the source.

Here is the discipline on a fictional lookup, stopping before any code is chosen. Coder B reads Patient A's documentation and identifies the reason for the encounter. Coder B looks that term up in the Alphabetic Index and finds an entry that points toward a code. Coder B does not stop there: the next move is to turn to that location in the Tabular List, read the full description, and read every note attached — because the Tabular List may show that a different or more complete code is required, or that another code must be reported as well. Only after the Tabular List and its notes agree with the documentation does a code become a candidate. This walk-through stops here on purpose: the goal is the sequence, not a code.

One more trap catches people who use an encoder or coding software instead of a printed book. A software tool that returns a code has not replaced the verification step for you; the same rule applies. Read the Tabular-level description and notes the tool surfaces, and confirm they match the documentation, rather than accepting the first result because the screen looks finished. The tool speeds the lookup; it does not excuse you from confirming the code against its notes and the record.

Recall check: you found an entry in the Alphabetic Index. What is the next required step before that code can be used, and why is the Index never the last word?

Conventions and instructional notes — what they are, and where their rules live

The Index and Tabular List speak in a set of standardized signals called conventions. You should be able to recognize the major types; their precise meanings are defined in the conventions section of the official guidelines, which this page points to rather than restates.

Text alternative: each row names a convention type, gives the plain idea it signals so you can recognize it, and points to the official guidelines' conventions section for its exact meaning. The two kinds of exclusion note are the classic confusion — they look similar and mean opposite things, so their meanings are always confirmed in the guidelines, not guessed.

The takeaway is a boundary, not a memory list: recognize the signal, then read its defined meaning in the official guidelines. The exact wording of any single convention is look-up content, and it can change when the guidelines change.

The official guidelines and why they control

The Official Guidelines for Coding and Reporting are published together with the code set and are part of the standard. They govern how codes are sequenced, how much specificity is required, how conventions are read, and which rules apply in which setting. When the Index, the Tabular List, and the guidelines seem to disagree, the guidelines and the documentation control — not a search result, an old note, or a habit.

Some guideline rules differ by setting. The handling of an uncertain or unconfirmed diagnosis, for example, is not the same in every care setting.

Stop and Verify: This detail can depend on a current rule, code year, payer, employer, setting, or document. Check the controlling official source and the applicable policy before acting. Whether a rule applies the same way in an outpatient setting and a hospital inpatient setting — including how an uncertain diagnosis is handled and how codes are sequenced — is defined by the current official guidelines for that setting; this page does not state the rule.

For how to tell which source controls which kind of question, see Source Verification Standards — Standard.

Specificity, laterality, and the limits of documentation

This is where the anatomy page's rule becomes a coding rule. A code is assigned to the level of specificity the documentation supports — no more. Unspecified codes exist precisely for situations that are documented but not further specified, and laterality (which side) is reported only when the record documents it. Specificity means accuracy relative to the record; it does not mean "add the most detailed code you can imagine."

Compare two fictional notes (examples are fictional and simplified for coding literacy, not clinical education):

• Patient A's note documents a condition and clearly states the left side. Here "left" is documented, so laterality is supported.
• Patient B's note documents the same kind of condition but records no side at all.

The first note supports a laterality detail because the provider wrote it. The second does not — and no amount of anatomy knowledge, pattern-matching, or "it's usually that side" turns a blank into a documented side. When the record is incomplete for the code you would otherwise assign, the professional response is your employer's query process, or reporting as documented under the guideline direction — never inference. Neither note is taken to a final code here; the point is what each record supports.

For the underlying rule that laterality is never inferred, see Anatomy for Coders — Remediation. For why documentation integrity matters beyond coding accuracy, see HIPAA and Compliance — Related.

Recall check: a note documents a condition but no side. Can you report a side? What are the two allowed professional responses when the record is incomplete?

Symptoms vs. definitive diagnoses (concept only)

Learners often ask whether to code the symptom or the underlying condition. As a concept: symptoms and signs are coded when no definitive diagnosis has been established, and they are handled differently once a definitive diagnosis that explains them is documented. The exact rule — including how it differs by setting — is guideline content you locate in the official document.

Stop and Verify: This detail can depend on a current rule, code year, payer, employer, setting, or document. Check the controlling official source and the applicable policy before acting. Whether a symptom is reported, and how it relates to a documented definitive diagnosis, is defined by the current official guidelines; this page names the concept only and does not state the rule.

Annual updates and the freshness habit

A code or a rule means nothing without its year. The code set and the official guidelines are updated on a published cycle — an annual update, plus any additional updates the official sources announce during the year — with effective dates the official sources publish. Historically, the annual ICD-10-CM update takes effect on October 1; confirm the current cadence and effective dates at the source rather than trusting this or any fixed date, because the schedule and any mid-year updates are announced officially and can change.

The practical habit that follows: before you trust any reference, encoder, or study material, check that it matches the current code year. Old study notes are a well-known error source, because a code that was valid two cycles ago may have been revised, replaced, or retired, and a guideline rule may read differently now. When you write your own notes, record the code year on them, the way you would date any reference.

See Source Verification Standards — Standard — and Official Resources and Study Tools — Official resource.

Recall check: two references both describe the same code. What single fact decides which one you can trust, and where do you confirm it?

A safe research workflow (conceptual)

This is the checklist to keep. It is a learning model, not operational instruction; real work follows your employer's policy and the current official sources.

1. Read the documentation and identify the condition or the reason for the encounter.
2. Locate that term in the Alphabetic Index.
3. Verify the candidate code in the Tabular List — read the full description.
4. Apply the conventions and instructional notes attached in the Tabular List (and any cross-references from the Index).
5. Check the official guidelines for sequencing and any setting-dependent rules.
6. Confirm the code year of every reference you used.
7. If the documentation is insufficient for the code, query through your employer's process or report as documented under the guideline direction — never infer.

Text alternative (workflow diagram, described): the workflow runs documentation → Alphabetic Index → Tabular List → conventions and notes → official guidelines → confirm code year → query if the record is insufficient. It is drawn as a one-way sequence, labeled a learning model, so that the Index is never mistaken for the endpoint.

The rule underneath all seven steps: never rely on a search snippet, a forum post, or memory for a real decision. Those can start a question; they never end one.

Understand, memorize, look up, verify

Common misconceptions

Check yourself

• Say what diagnosis coding is for and how it differs from coding what was done.
• Reconstruct the safe research workflow in order, including the two allowed responses when the record is insufficient.
• Name the convention types you can recognize, and say where their exact meanings are defined.
• Sort three fictional documentation details into documented, unspecified, or inferred.
• Given four sources for the meaning of a convention, say which one controls.
• State in one sentence why the code year of a reference matters.

Teach it back: explain to a new coworker why a code is never taken from the Index alone. Include one distinction (Index vs. Tabular List), one professional-context point (the query process and the current code year), and one thing you would confirm in the official guidelines. There is no model answer; the goal is that your explanation names the verification step.

Ready to move on?

• I can say what diagnosis coding is for and how it differs from coding what was done.
• I can describe the Index-then-Tabular workflow and why the Index is never the final step.
• I can name the convention types and know the official guidelines define them.
• I know specificity and laterality come from documentation, and what a professional does when the record is incomplete.
• I check the code year of any reference before trusting it.

This page does not qualify anyone to assign diagnosis codes for real encounters. It teaches the concepts and the safe research workflow only.

If you got something wrong

If two ideas stay tangled, put them side by side and compare before more practice. If you relied on an outdated or non-authoritative source, review the source challenge below before continuing.

In the profession

Coders assign diagnosis codes from documentation under the official guidelines and employer policy. Billers and patient-accounts staff recognize diagnosis codes on claims and route questions to coders rather than editing codes themselves; that recognize-versus-assign boundary is real, and crossing it is not a shortcut. Auditors and compliance staff review whether specificity is supported by the record, and providers document. Employer query policy, the official guidelines, and the current code set control real work — no memory, forum, or search box does.

In certification

Diagnosis coding is a broad domain area in both facility-oriented and professional coding credentials, and billing credentials expect that you can recognize diagnosis codes on a claim. If you are researching credentials, the CCA Certification Review — Certification connection — carries a setting-to-code-set map (which settings pair ICD-10-CM with which procedure code set), and the CPC Certification Review — Certification connection — carries a "Foundations to build first" sequence and an allowed-references-and-editions verification item, which is the same code-year habit this page teaches. Domains are named as broad public-outline areas only; confirm each current exam-content outline with the issuing organization. This page does not align to any exam and does not predict questions.

In careers

Job postings phrase this as "ICD-10-CM proficiency" or "diagnosis coding experience," and some name a particular encoder. This page gives conceptual orientation only. Proficiency is built with current references, supervised practice, and employer training, and no employment outcome is implied by reading it. For role orientation, see Medical Biller vs. Medical Coder — Career connection.

For continuing learning

Returning learners refresh the workflow and the convention types first, then check what changed: the current code-year effective dates, the guideline version, and any convention changes. A six-month return needs the freshness habit plus the workflow checklist; a years-away return needs the full page, then CPT Basics. This resource provides no continuing-education or renewal credit, and none is implied.

Study options

• 5-Minute Review: diagnosis vs. procedure, and the never-from-the-Index rule.
• 15-Minute Study: read "How codes are structured" and walk through one synthetic lookup to the Tabular List.
• Full Lesson: read every section — structure, Index and Tabular, conventions, guidelines, specificity, the symptoms concept, updates, and the workflow.
• Refresher or Deep Dive: the freshness habit and the source challenge with four hypothetical sources; the separate parts of the Index are optional deeper reading.

There is no single best plan. Use the short review when time is limited, and return for the full pass when a section keeps tripping you up.

If you remember only five things: diagnosis codes report why, not what was done; you go Index then Tabular List, and the Index is never the last word; the conventions and the official guidelines control the answer; specificity is bounded by the documentation; and a code only means something with its year.

Check the source yourself

Not every source that shows a code controls what that code should be. The current official guidelines and code set control a diagnosis-coding answer. An encoder or software tool speeds the lookup but does not replace the verification step or the guidelines behind it. A study guide, a general website, or a coworker's memory may raise a good question, but none of them controls a real-work answer, and each can be out of date.

Freshness note: the code set and the official guidelines are periodically updated (verify the current cadence and effective dates); the workflow concept is stable; setting-dependent rules are context-dependent and defined by the current guidelines for that setting.

Source challenge: you need the meaning of a convention as it applies to a code in Patient A's record. You have four sources: the current official guidelines, an encoder's pop-up tip, a three-year-old study guide, and a forum thread. Which one controls the answer you can act on, which are acceptable only to raise a question, and what would you still confirm before relying on any of them? See Official Resources and Study Tools — Official resource.

Frequently asked questions

What is ICD-10-CM used for?

It is the diagnosis code set used to report why care occurred — conditions, symptoms, and other documented reasons for an encounter — in United States healthcare settings, in a standardized form used for claims, statistics, research, and quality measurement.

Do I have to memorize ICD-10-CM codes?

No. Nobody memorizes the code set. What you learn is the workflow — documentation, Alphabetic Index, Tabular List, conventions, official guidelines, code year — and you look up every specific code and rule. Professionals look things up every working day.

Why can't I code from the Alphabetic Index?

Because the Index only points you toward a code. The Tabular List carries the full description and the notes that may change the code or require another, and the official guidelines direct you to verify there. A code is never assigned from the Index alone.

What is the placeholder character X?

It is a structural character that fills an empty position so that a required later character — such as a seventh character — lands in the correct place. It is only structural; it carries no meaning of its own, and it is not a code.

How often does ICD-10-CM change?

On a published schedule — an annual update plus any additional updates the official sources announce, with effective dates they publish (historically the annual update has taken effect on October 1). Confirm the current cadence and effective dates at the official source rather than relying on any fixed date.

Can I code a diagnosis that is "probably" present?

That depends on the setting and is defined by the current official guidelines, which handle uncertain diagnoses differently in different settings. Treat it as a Stop-and-Verify point and check the current guidelines; this page names the concept only and does not state the rule.

Where to go next

• Recommended next: CPT Basics — Next lesson — where what was done completes the picture that this page started with why.
• If you struggled: Anatomy for Coders — Remediation — for laterality and do-not-infer, then retry the workflow recall prompts.
• If you already know this: Modifiers — Related — after CPT Basics.
• If your goal is billing: HIPAA and Compliance — Related, then Medical Biller vs. Medical Coder — Career connection.
• If your goal is coding: CPT Basics — Next lesson, then Modifiers — Related.

Sources to verify before relying on this page

• CDC/NCHS official ICD-10-CM pages and the Official Guidelines for Coding and Reporting — control who maintains the code set, the structure description, the conventions, the categories of guideline rules, the setting-dependent rules, and the effective dates and versions.
• CMS official ICD-10 resources — control the federal adoption statements and the update announcements.
• Official CMS ICD-10-PCS identification — controls the one-line note that inpatient facility procedures are reported with ICD-10-PCS while professional services in any setting use CPT.
• Issuing organizations' current public exam-content outlines — control the statement that the credential reviews name diagnosis coding as a domain and terminology and structure as foundations to build first.
• Source Verification Standards — Standard — for the learner source-literacy framing used in "Check the source yourself."