Skip to content

Study guides

ICD-10-CM Conventions: A Study Guide for Coding Students (Based on the FY2027 Official Guidelines)

Conventions are the general rules for using ICD-10-CM. They are built into the Alphabetic Index and the Tabular List as punctuation, abbreviations and instructional notes. When a convention conflicts with a guideline, the convention wins. HIPAA requires coders to follow both. Most certification exam questions test conventions directly or indirectly: a wrong placeholder, a missed 'code first' note or a misread Excludes note makes an answer wrong. Each section below explains one convention in plain language, lists the points to memorise, and works through a short example. The code numbers are illustrations. Always confirm the full code and its notes in your own FY2027 Tabular List.

1. Alphabetic Index and Tabular List structure

ICD-10-CM has two parts. The Alphabetic Index is an alphabetical list of terms with the code each one leads to. It has four parts: the Index of Diseases and Injury, the Index of External Causes of Injury, the Table of Neoplasms, and the Table of Drugs and Chemicals. The Tabular List is the numbered list of codes. It is organized into chapters by body system or type of condition and uses indentation to show which entries belong under which.

In the Tabular List, every category has 3 characters. A subcategory has 4 or 5 characters. A code can have 3, 4, 5, 6 or 7 characters, and the final level of subdivision is always the code. A 3-character category with no further subdivision counts as a code. A code that takes a 7th character is still called a code, but it is invalid without that 7th character. Only codes can be reported. Categories and subcategories cannot.

Key points

  • The Index has four parts: Diseases and Injury, External Causes, Neoplasm Table, and Table of Drugs and Chemicals.
  • Categories have 3 characters, subcategories have 4 or 5, and codes have 3 to 7.
  • Only full codes are reportable, including any required 7th character.
  • A 3-character category with no subdivisions is a complete code.

Worked example

Decoding S72.141D, which appears in the guidelines. S72 is the 3-character category for fracture of femur. S72.1 and S72.14 are subcategories. S72.141 is the code for a displaced intertrochanteric fracture of the right femur. The 7th character D means a subsequent encounter for a closed fracture with routine healing. Reporting S72.141 without the D is invalid. By contrast, I10, Essential (primary) hypertension, is a 3-character category with no subdivisions, so I10 alone is a complete, reportable code.

2. Locating a code

Always start in the Alphabetic Index by looking up the main term, which is usually the condition rather than the body site. Then verify the code in the Tabular List. Never start in the Tabular List, and never code from the Index alone. Read and follow the instructional notes in both places.

The Index often gives only part of a code. A dash at the end of an Index entry means more characters are needed. Laterality and 7th characters can only be selected in the Tabular List. Even when the Index entry has no dash, check the Tabular List to confirm that no 7th character is required.

Key points

  • Look up the term in the Index first, then verify in the Tabular List.
  • A dash in the Index means the code is incomplete.
  • Laterality and 7th characters are chosen in the Tabular List.
  • Instructional notes in both the Index and the Tabular List must be followed.

Worked example

Scenario: A patient is treated in the ED for a laceration of the scalp with no foreign body. Index: Laceration, scalp, which points to S01.01- (the dash means the code is incomplete). Tabular: S01.01 is Laceration without foreign body of scalp, and category S01 requires a 7th character. S01.01 has only 5 characters, so the placeholder X fills the 6th position. The patient is receiving active treatment, so the 7th character is A. Final code: S01.01XA.

3. Placeholder X and 7th characters

ICD-10-CM uses the letter X as a placeholder in certain codes to leave room for future expansion. The poisoning, adverse effect and underdosing codes in categories T36-T50 are the classic example. Wherever the classification shows a placeholder X, you must include it, or the code is invalid.

Some categories require a 7th character for every code in the category, or wherever the Tabular notes say so. The 7th character must always sit in the 7th position. If the code is shorter than 6 characters, fill each empty position with X before adding the 7th character. In most injury categories, A means the patient is receiving active treatment, D means routine care during healing or recovery, and S means sequela. Fracture categories have additional values.

Key points

  • A placeholder X shown in the code is mandatory.
  • The 7th character always goes in the 7th position; fill any gaps with X.
  • A code that requires a 7th character is invalid without it.
  • A means active treatment. A first visit to a new provider does not by itself make the encounter 'initial'.

Worked example

Scenario: A patient has urticaria from amoxicillin that was correctly prescribed and properly taken. This is an adverse effect. Code the nature of the effect first: L50.9, Urticaria, unspecified. Next, look up Amoxicillin in the Table of Drugs and Chemicals, Adverse Effect column, and verify the result in the Tabular List. Never code directly from the Table. Category T36.0 covers penicillins. In T36.0X5, the X is the 5th-character placeholder and 5 means adverse effect. Add A for active treatment: T36.0X5A. Final codes: L50.9, T36.0X5A.

4. Abbreviations NEC and NOS

NEC means 'not elsewhere classifiable.' It signals that the condition is specified in the record but has no specific code of its own, so it goes to the 'other specified' code. In the Index, an entry marked NEC sends you to an 'other' code. In the Tabular List, an NEC entry under a code marks that code as the 'other specified' code.

NOS means 'not otherwise specified' and is the same as 'unspecified.' Use the NOS or unspecified code when the record does not give enough detail for a more specific code.

Key points

  • NEC means specific information is documented, but no specific code exists, so use the 'other' code.
  • NOS means information is missing, so use the 'unspecified' code.
  • NEC appears in both the Index and the Tabular List with the same meaning.
  • Do not mix them up: NEC is a lack of a code; NOS is a lack of documentation.

Worked example

Scenario A: The provider documents only 'anemia,' with no type. Index: Anemia leads to D64.9. Tabular: D64.9 is Anemia, unspecified. This is an NOS situation because the type was not documented. Scenario B: The provider documents a specific type of anemia that has no dedicated code. The Index entry Anemia, specified type NEC leads to D64.89, Other specified anemias. Here the information exists, but the classification has no separate code for it.

5. Punctuation: brackets, parentheses, colons

Square brackets [ ] have two uses. In the Tabular List they enclose synonyms, alternative wording or explanations. In the Alphabetic Index they mark a manifestation code, which must be sequenced second.

Parentheses ( ) enclose nonessential modifiers. These are words that may or may not appear in the diagnosis without changing the code. In the Index, a main term's nonessential modifiers carry down to its subterms, with one exception: if a nonessential modifier and a subentry contradict each other, the subentry wins. For example, under the main term Enteritis, 'acute' is a nonessential modifier and 'chronic' is a subentry, so 'acute' does not apply to the 'chronic' subentry.

A colon : in the Tabular List follows an incomplete term that needs one of the modifiers listed after the colon before the code can be assigned. Commas in the Index can mean alternative wording, an essential or nonessential modifier, or 'and/or,' depending on context.

Key points

  • Tabular brackets enclose synonyms or explanations; Index brackets mark a manifestation code that goes second.
  • Words in parentheses do not change the code, whether present or absent.
  • If a nonessential modifier conflicts with a subentry, the subentry takes precedence.
  • A colon means the term is incomplete without one of the modifiers that follow it.

Worked example

Scenario: The provider documents 'acute pneumonia,' with no organism. Index: Pneumonia (acute) (double)... J18.9. Because 'acute' is in parentheses, it is nonessential, and the code is J18.9 whether or not 'acute' is written. Bracket examples: the Tabular title for B20 reads 'Human immunodeficiency virus [HIV] disease,' where the brackets hold the abbreviation. In the Index entry for Alzheimer's disease, the dementia code appears in brackets after the Alzheimer's code, which tells you the dementia code is a manifestation and is sequenced second.

6. "And", "with", "see" and "see also"

When 'and' appears in a code title, read it as 'and/or.' For example, tuberculosis of bones, tuberculosis of joints, and tuberculosis of bones and joints all classify to subcategory A18.0, Tuberculosis of bones and joints.

'With' and 'in,' when they appear in a code title, an Index entry or an instructional note, mean 'associated with' or 'due to.' The classification presumes the two conditions are related, so code them as related even if the provider does not link them. There are two exceptions: the documentation clearly states the conditions are unrelated, or another guideline requires an explicitly documented link, as the sepsis guideline does for organ dysfunction. If two conditions are not linked by such a term in the classification, the provider must document the link. In the Index, 'with' is listed immediately after the main term or subterm, not in alphabetical order.

'See' after a main term is mandatory: you must go to the referenced term to find the code. 'See also' is optional: it points to another main term that might help, but you do not need to follow it if the original entry already gives you the code.

Key points

  • 'And' in a title means 'and/or.'
  • 'With' and 'in' create a presumed causal link unless documentation states the conditions are unrelated or a guideline requires an explicit link.
  • In the Index, 'with' appears right below the main term or subterm, ahead of the alphabetical subterms.
  • 'See' must be followed; 'see also' is optional.

Worked example

Scenario: A patient has hypertension and stage 3 chronic kidney disease (CKD). The provider does not say whether they are related. Index: Hypertension, with, chronic kidney disease. Because 'with' links them, the relationship is presumed. Code I12.9, Hypertensive chronic kidney disease with stage 1 through stage 4 chronic kidney disease, or unspecified chronic kidney disease, plus N18.30, Chronic kidney disease, stage 3 unspecified, to identify the stage. Coding I10 plus N18.30 would be wrong. If the provider had documented that the CKD is not due to the hypertension, I10 and N18.30 would be correct.

7. Code first, use additional code and code also notes

'Use additional code' tells you to add a second code after the current one, if the information is known. 'Code first' tells you to sequence the underlying condition before the current code, if that condition is present. These notes appear with etiology/manifestation pairs. They also appear with other codes where a second code is needed to fully describe a condition, such as identifying the organism that caused an infection.

'Code, if applicable, any causal condition first' means this code may be listed first only when the cause is unknown or does not apply. If the cause is known, the code for the cause goes first.

A 'code also' note means two codes may be needed to fully describe a condition, but it gives no sequencing direction. The order depends on the circumstances of the encounter.

Key points

  • 'Use additional code' means the current code goes first and the added code follows.
  • 'Code first' means the underlying condition goes ahead of the current code.
  • 'Code, if applicable, any causal condition first' means the current code may be first-listed only when the cause is unknown or not applicable.
  • 'Code also' requires both codes but does not dictate their order.

Worked example

Scenario A: A urinary tract infection due to E. coli. N39.0, Urinary tract infection, site not specified, carries a 'use additional code' note to identify the infectious agent (B95-B97). Codes: N39.0, then B96.20, Unspecified Escherichia coli [E. coli] as the cause of diseases classified elsewhere. Scenario B: A patient with known essential hypertension presents in hypertensive urgency. Category I16 says to code also any identified hypertensive disease. Assign I16.0, Hypertensive urgency, and I10, and sequence them according to the reason for the encounter.

8. Etiology and manifestation convention

Some diseases cause effects in other body systems. ICD-10-CM requires the underlying cause, the etiology, to be coded first and the effect, the manifestation, second. The etiology code carries a 'use additional code' note, and the manifestation code carries a 'code first' note.

Most manifestation codes include 'in diseases classified elsewhere' in their titles. These codes can never be the principal or first-listed diagnosis and must always follow the underlying condition code. Some manifestation codes do not have that phrase in the title. When the paired 'use additional code' and 'code first' notes are present, the same sequencing rule applies to them. In the Index, the etiology code is listed first and the manifestation code follows in brackets.

Key points

  • Etiology first, manifestation second.
  • 'In diseases classified elsewhere' codes are never first-listed or principal.
  • In the Index, the bracketed code is always sequenced second.
  • Paired 'code first' and 'use additional code' notes enforce the order even when the title lacks 'in diseases classified elsewhere.'

Worked example

Scenario: A patient has dementia due to Alzheimer's disease. Severity is not documented, and there is no behavioral, psychotic, mood or anxiety disturbance. The Index lists the Alzheimer's code followed by the dementia code in brackets. Codes in order: G30.9, Alzheimer's disease, unspecified, then F02.80, Dementia in other diseases classified elsewhere, unspecified severity, without behavioral disturbance, psychotic disturbance, mood disturbance, and anxiety. Listing F02.80 first is never allowed. The guidelines' Parkinson's disease example follows the same pattern: a code from category G20 first, then the F02 code.

9. Includes notes and inclusion terms

An Includes note appears directly under a 3-character category title. It defines the category further or gives examples of what the category contains.

Inclusion terms are lists of conditions printed under some codes, and the code is to be used for those conditions. They may be synonyms of the code title. Under an 'other specified' code, they list the various conditions assigned there. The lists are not exhaustive: a term found only in the Alphabetic Index can still lead to that code.

Key points

  • Includes notes sit under 3-character category titles and define or illustrate the category.
  • Inclusion terms are synonyms or examples of conditions coded to that code.
  • Inclusion term lists are not complete; the Index may assign other terms to the code.
  • Under an 'other specified' code, inclusion terms show the different conditions grouped there.

Worked example

Scenario A: The provider documents psychogenic dysmenorrhea. It is an inclusion term under F45.8, Other somatoform disorders, so F45.8 is assigned. Teeth grinding is another inclusion term under the same code, which shows that an 'other specified' code can hold unrelated conditions. Scenario B: A patient has hypertension, heart disease and CKD. The Includes note at category I13 says the conditions in I11 and I12 are included together in I13. Use one I13 code, not separate I11 and I12 codes, and add an N18 code to identify the CKD stage.

10. Excludes1 vs Excludes2

An Excludes1 note is a pure exclusion that means 'NOT CODED HERE!' The excluded code should never be reported together with the code above the note, because the two conditions cannot occur together, such as the congenital and acquired forms of the same condition. There is one exception: if the two conditions are unrelated to each other, both may be coded. If you cannot tell whether they are related, query the provider.

An Excludes2 note means 'not included here.' The excluded condition is not part of the condition the code represents, but the patient may have both at the same time. When appropriate, report both codes together.

Key points

  • Excludes1 means never together, unless the two conditions are clearly unrelated.
  • If it is unclear whether conditions under an Excludes1 note are related, query the provider.
  • Excludes2 means both codes may be reported together when the patient has both conditions.
  • Both types mean the excluded codes are independent of each other.

Worked example

Scenario (from the guidelines): F45.8, Other somatoform disorders, has an Excludes1 note for sleep related teeth grinding (G47.63), because teeth grinding is also an inclusion term under F45.8. For teeth grinding alone, assign only one of the two codes. Now suppose a patient has psychogenic dysmenorrhea, which codes to F45.8, and also has sleep related teeth grinding, which codes to G47.63. These conditions are clearly unrelated, so the Excludes1 exception applies and F45.8 and G47.63 are reported together.

11. Default codes

The code printed next to a main term in the Alphabetic Index is the default code. It represents either the form most commonly associated with that term or the unspecified code for the condition. When a condition is documented with no further detail, such as acute or chronic, assign the default code.

The guidelines also set some defaults of their own. For example, if the type of diabetes mellitus is not documented, the default is type 2 (category E11). Always verify the default code in the Tabular List.

Key points

  • The default code is the code next to the main term in the Index.
  • Use it when the documentation adds no further detail.
  • Diabetes with no type documented defaults to type 2 (E11.-).
  • Look for subterms that match the documentation before settling on the default.

Worked example

Scenario A: The provider documents 'appendicitis' and nothing more. Index: Appendicitis leads to K37. Tabular: K37, Unspecified appendicitis, is a 3-character code with no subdivisions, so K37 is complete. Scenario B: The note says 'diabetes mellitus, on insulin,' with no type and no complications. The type defaults to type 2: E11.9, Type 2 diabetes mellitus without complications. Add Z79.4, Long term (current) use of insulin.

12. Combination codes

A combination code is a single code that classifies one of three things: two diagnoses, a diagnosis with an associated secondary process (manifestation), or a diagnosis with an associated complication. You find combination codes through subterms in the Alphabetic Index and through the inclusion and exclusion notes in the Tabular List.

When a combination code fully describes everything documented, or when the Index directs you to it, assign only the combination code and do not code its parts separately. If the combination code lacks needed detail about the manifestation or complication, add a secondary code. When a combination code includes a common symptom of the diagnosis, do not code that symptom separately.

Key points

  • One code can capture two diagnoses, a diagnosis plus manifestation, or a diagnosis plus complication.
  • Do not code the parts separately when the combination code covers them all.
  • Add a secondary code only when the combination code lacks necessary specificity.
  • Find combination codes through Index subterms and Tabular notes.

Worked example

Scenario A: Atherosclerotic coronary artery disease of the native coronary arteries with unstable angina. Assign only I25.110, Atherosclerotic heart disease of native coronary artery with unstable angina pectoris. Do not add a separate angina code. Scenario B: Pneumonia due to MRSA. Assign J15.212, Pneumonia due to Methicillin resistant Staphylococcus aureus, alone. Do not add B95.62 or a Z16.11- code. Scenario C: Type 2 diabetes with diabetic CKD, stage 4. Assign E11.22, which covers the diabetes and the kidney complication, plus N18.4 for the stage, because the combination code does not include the stage.

13. Laterality

Many codes specify right, left or bilateral. If the condition is bilateral and no bilateral code exists, assign separate right and left codes. If the side is not documented anywhere, use the unspecified-side code, but this should rarely happen.

Laterality is one of the exceptions to the rule that codes must come from the patient's provider. If the provider does not document the side, you may take it from another clinician's documentation, such as a nurse's note. If the documentation conflicts, query the provider.

Sometimes a bilateral condition is treated one side at a time in separate encounters. Use the bilateral code for the first encounter, because the condition still exists on both sides. At the later encounter, if the first side is resolved, use the unilateral code for the remaining side. If the first treatment did not resolve the condition on that side, keep using the bilateral code. Remember also that each unique code is reported only once per encounter.

Key points

  • No bilateral code available: report the right and left codes separately.
  • Laterality may be taken from other clinicians' documentation; query the provider if documentation conflicts.
  • The unspecified side is a last resort.
  • For staged bilateral treatment, use the bilateral code first and the remaining side's code later, if the first side is resolved.

Worked example

Scenario: A patient has age-related nuclear cataracts in both eyes. Encounter 1 is surgery on the right eye. Both eyes still have the condition at this encounter, so assign H25.13, Age-related nuclear cataract, bilateral. Encounter 2, weeks later, is surgery on the left eye. The right cataract has been removed, so assign H25.12, Age-related nuclear cataract, left eye. Using H25.13 again would be incorrect.

14. Syndromes

When coding a syndrome, follow the Alphabetic Index first. If the Index gives no guidance, assign codes for the manifestations documented for that syndrome.

If the syndrome has no unique code, you may also code manifestations that are not an integral part of the disease process. For conditions in the congenital chapter, the rule is similar. When a code specifically identifies the anomaly or disorder, do not code its inherent features separately, but do code manifestations that are not an inherent part of it.

Key points

  • Look up the syndrome in the Index first.
  • With no Index guidance, code the documented manifestations.
  • Do not separately code features that are an inherent part of a uniquely coded syndrome.
  • Do code manifestations that are not inherent to the syndrome.

Worked example

Scenario: The provider documents Down syndrome with no karyotype detail. Index: Syndrome, Down leads to Q90.9, Down syndrome, unspecified. Verify in the Tabular List and follow its notes. Features that are an inherent part of Down syndrome are not coded separately. If the provider documents another condition that is not an inherent component, code it in addition. If the documented syndrome had no Index entry at all, you would instead code each documented manifestation.

15. "Other" and "unspecified" codes

Codes titled 'other' or 'other specified' are for cases where the record gives specific detail but no specific code exists for it. Index entries marked NEC lead to these codes.

Codes titled 'unspecified' are for cases where the record does not give enough information to assign a more specific code. If a category has no unspecified code, the 'other specified' code covers both the other and unspecified situations.

Unspecified and sign or symptom codes are legitimate when they reflect what is actually known at that encounter. You must not choose a more specific code that the documentation does not support. Ordering unnecessary tests just to reach a more specific code is also inappropriate.

Key points

  • 'Other' means detail is documented but there is no specific code for it.
  • 'Unspecified' means the documentation lacks the detail.
  • When no unspecified code exists, the 'other specified' code covers both situations.
  • Code to the level of certainty known at that encounter. Never guess at specificity.

Worked example

Scenario: In an office visit, the provider diagnoses pneumonia, and no organism has been identified or is being tested for. Assign J18.9, Pneumonia, unspecified organism. This accurately reflects what is known. Choosing a bacterial pneumonia code because bacteria seem 'likely' would not be supported. If the provider had instead documented a specific type of condition that the classification lacks a code for, the Index NEC entry would lead you to the 'other' code in that category.

Common mistakes

  • Coding from the Alphabetic Index alone, or starting in the Tabular List, instead of looking up the term in the Index and then verifying it in the Tabular List.
  • Leaving out a required placeholder X or 7th character, or putting the 7th character in the wrong position, which makes the code invalid.
  • Reporting a category or subcategory, such as A18.0 or S72.14, instead of a complete code.
  • Sequencing an 'in diseases classified elsewhere' manifestation code first, or ignoring the bracketed Index code that must come second.
  • Reading 'code also' as an order to sequence; it requires both codes but leaves the order to the circumstances of the encounter.
  • Treating Excludes2 as 'never together,' or applying Excludes1 to conditions that are clearly unrelated.
  • Ignoring the presumed link created by 'with,' for example coding I10 plus an N18 code for hypertension with CKD instead of an I12 code plus the N18 code.
  • Adding separate codes for elements already captured in a combination code, such as an angina code with I25.110, or B95.62 with J15.212.
  • Applying a nonessential modifier to a subentry it conflicts with, as with 'acute' and 'chronic' under Enteritis.
  • Ignoring a 'see' instruction, which is mandatory, or wasting time following 'see also' when the original entry already gives the code.
  • Choosing 'unspecified' when the record supports a more specific code, or choosing a specific code the record does not support.
  • Mixing up NEC ('other': detail documented but no specific code) with NOS ('unspecified': detail missing).
  • Assigning the bilateral code at the second encounter after the first side has been treated and resolved.
  • Using the unspecified-side code when another clinician has documented laterality.
  • Treating inclusion term lists as complete, or reporting the same diagnosis code more than once for one encounter.

Written with AI and checked automatically against the ICD-10-CM FY2027 Official Guidelines and code set (2026-10-14). See our editorial policy.