In the fading era of ICD-9, you can probably get along with a minimal understanding of the ICD-9 system – at least as long as you have a good billing staff and a superbill (or the electronic equivalent) on which you can just check a code. But the coming era of ICD-10 seems likely to pull you more fully into the process of diagnosis coding. Even with a billing staff good enough to deduce the correct code from your note, you’ll need to know what documentation they need. And to know that, you’ll need to be familiar with the structure of the codes and the rules for using them, which is the subject of this article, the second in our orientation to ICD-10 (see “Other articles in FPM’s ‘Countdown to ICD-10’ series”).

The current version of ICD-10-CM is available from the National Center for Health Statistics website (look for a PDF of the “Guidelines” and a compressed archive of the other files). Browsing through the files can help orient you. It may also be helpful to look up the codes described throughout the article.

The ICD-10 manual begins with “ICD-10-CM Official Guidelines for Coding and Reporting” and is then divided into two main parts: first, the alphabetic index of terms with corresponding codes (subdivided into an index of diseases and injuries, an index of external causes, and tables of drugs and neoplasms) and second, the tabular list of codes divided into chapters based on body systems or conditions. You’ll want to become familiar with the instructions at the chapter and section levels, since these apply to numerous code categories.

As with ICD-9, ICD-10 has three basic instructions:

1. Consult the alphabetic index first and then refer to the tabular list.

2. Read all pertinent instructions in both the alphabetic index and the tabular list.

3. In the tabular list, check for instructions at the three-character category level in addition to the code level.

Coding sequelae

Many instructions in ICD-10, such as “Code first” and “Use additional code,” may be familiar from ICD-9. Others are new. Late effects are handled somewhat differently. They are referred to as sequelae in ICD-10; they are reported in some cases with codes specifically intended to describe sequelae and in others by adding a seventh character to the code for the condition that caused the sequelae. An example of each follows:

I69 Sequelae of cerebrovascular disease. Codes in category I69 are used to identify sequelae of cerebrovascular diseases, which are themselves coded in categories I60–I67. For instance, I69.022 is dysarthria following nontraumatic subarachnoid hemorrhage, while codes for nontraumatic subarachnoid hemorrhage itself are found in a different category, I60.

S83 Dislocation and sprain of joints and ligaments of knee. This is one of the code categories that requires the addition of a seventh character – one of three:

• A: initial encounter,

• D: subsequent encounter,

• S: sequela.

So, for instance, S83.104S would code unspecified dislocation of right knee, sequela.

Exclusions

ICD-10 offers two different types of exclusions. Both indicate that the current category or code excludes some specified condition, which is described elsewhere in the code set. “Excludes1” indicates that the code for the excluded condition should not be reported with the current code, while “Excludes2” indicates that the code for the excluded condition may be reported with the current code, assuming that both conditions are present. For example, category C43, malignant melanoma of skin, specifies exclusions of both kinds:

• Excludes1: melanoma in situ (D03.-),

• Excludes2: malignant melanoma of skin of genital organs (C51-C52, C60.-, C63.-), Merkel cell carcinoma (C4A.-), and sites other than skin.

A code from category C43 should not be used to code melanoma in situ, nor should a code from category C43 be submitted with a code for melanoma in situ. On the other hand, both a category C43 code and a code listed in the Excludes2 instruction could be reported on the same date, for instance when multiple sites are involved.

The placeholder: X

Like the code from category S83 that we looked at earlier, many codes are seven characters long, with the seventh character representing status or additional clinical information (see “Code construction: ICD-10 style”). With the specified seven-character length comes the biggest structural change from ICD-9 to ICD-10, the use of placeholders. The placeholder character, X, stands for a character that has not been assigned a meaning but must be filled in to achieve the specified code length.

For example, codes for fatigue fracture of vertebra, in subcategory M48.4, require a placeholder for the sixth character because the base codes are only five characters long and require one of the following as a seventh character:

• A: initial encounter for fracture,

• D: subsequent encounter for fracture with routine healing,

• G: subsequent encounter for fracture with delayed healing,

• S: sequela of fracture.

To code the initial encounter for a fatigue fracture of a cervical vertebra (M48.42), then, a placeholder X must be added to arrive at code M48.42XA. If you miss the instruction regarding the seventh character and report only M48.42 or M48.42A, your claim will likely be rejected because your code is incomplete.

Placeholders are also built into some codes at the fifth character, as in S06.0, the subcategory for concussion. Here, the sixth character specifies the duration of loss of consciousness, and the required seventh character is A, D, or S – but the fifth character has not been assigned a meaning. So, for instance, S06.0X0A codes concussion without loss of consciousness, initial encounter.

Another example where the placeholder is needed is code W04, fall while being carried or supported by other persons. In this case, the three-character category is tantamount to a code, because it has no subdivisions. But since the instructions require that the appropriate seventh character (A,D, or S) be added, to make a complete code of W04 requires multiple placeholders (e.g., W04.XXXA).

It’s complicated

Clearly, ICD-10 codes can convey more specific information than ICD-9 codes. If you aren’t directly involved in code selection, and even if your electronic health record system suggests ranges of possible codes, you’ll need to make sure that your documentation includes all the information needed to determine a code – information such as which side of the patient’s body is involved, whether this was the first or a subsequent encounter for the problem, what condition gave rise to a sequela, whether healing is progressing as expected, and so on. Now is the time to consider whether your current documentation is sufficient and whether your coding resources will support complete ICD-10 code selection. If either is lacking, you’ll have time to consider options for achieving an optimal result.