Why Specificity Matters When Documenting and Coding HCC’s

The Centers for Medicare & Medicaid Services (CMS) initiated the Hierarchical Condition Category (HCC) model in 2004, but it is now becoming increasingly common as the industry shifts to value-based payment models.

Under the HCC risk adjustment model, specificity in documentation and coding is key to providing a complete picture of patient health, establishing accurate benchmarks and ensuring appropriate compensation. HCC codes are used to calculate a patient’s risk adjustment factor (RAF) score, so that predicted cost of care will be more in line with the patient’s health status.

Attention to precise and accurate coding can significantly impact the RAF score, which determines appropriate compensation and enables better coordination of care for complex patient populations.

Here are some of the most common oversights when it comes to specificity in HCC coding:


Diabetes mellitus is one of the most inaccurately coded chronic conditions due to the complexity of diabetes coding and the prevalence of diabetes in the US. According to the Centers for Disease Control (CDC), 34.2million Americans—just over 1 in 10—have diabetes, and approximately 7.3million have been undiagnosed.

Diabetes mellitus is directly responsible for many acute and chronic complications in several organ systems, including neuropathy, retinopathy and skin conditions. Often, the cause-and-effect relationship between diabetes and these manifestations is not coded to the appropriate level specificity.

In general, chronic conditions with complications typically have a higher HCC weight than chronic conditions without complications. In some instances, the chronic condition without a complication does not have any HCC weight. In the case of diabetes, complete coding of complications has a particularly significant impact on RAF scores, since ICD-10 codes for diabetes with complications carry a RAF three times higher than a diagnosis of diabetes uncomplicated.

Diabetes mellitus codes are combination codes that describe the type of diabetes, the body system affected and the complications affecting that body system. If a patient has diabetes with a manifestation or complication that has its own code, that code should always be used instead of an unspecified diabetes code.


-      Code to the highest degree of specificity when documenting diabetes and assign as many codes as needed to describe all of the complications and associated conditions of the disease. 

-      When documenting all diabetic complications, ensure that the documentation links the complication to the diabetes and that the causal relationship between the conditions is clear.


Morbid Obesity

Morbid obesity is another illustrative example of the importance of specificity in HCC coding.

According to the CDC, more than one‐third of adults in the United States are obese. Among those, over 15 million are morbidly obese. Patients with a BMI equal to or above 40, or who are 100 pounds or more overweight, qualify as morbidly obese. Often the BMI for these patients is documented, but not the corresponding diagnosis of morbid obesity.

In this case, both diagnoses are essential for appropriate code and HCC assignments. If the patient has a BMI greater than 40, the relevant code for obesity plus a code for the BMI should be assigned. The BMI should be coded as a secondary diagnosis when documented in the presence of obesity, morbid obesity, or another clinically significant and related comorbidity (diabetes, hypertension, etc.) The patient’s BMI must be clearly documented as coders are not allowed to calculate BMI. For patients with a BMI over 40, morbid obesity must also be coded, in addition to the BMI, for the diagnosis to contribute to risk scoring.Without the additional morbid obesity code, the BMI diagnosis carries no HCC value.

For example, for a patient with a BMI of 45, both ICD-10 codes E66.01 (morbid obesity) and Z68.42 (BMI 45 - 49.9) should be assigned.

Also important to keep in mind that a BMI of 35 or greater can be diagnosed as morbid obesity if the patient had one or more-related conditions or comorbidities. Common comorbid conditions include sleep apnea, osteoarthritis, diabetes, hypertension and coronary artery disease.


-      When coding for obesity, it’s important to code for both the BMI as well as the obesity diagnosis.

-      The treating provider must document obesity, morbid obesity, or any other diagnosis-related code from a BMI measurement. Coders and billers cannot infer obesity from a BMI value or percentage.

- For all conditions, it’s important to link manifestations and complications. Coders can’t assume there is a connection with conditions listed in the medical record — the provider needs to make the link.Some terms that can be used to link conditions are “because of”, “related to”, “due to” and “associated with.”



Depression is also commonly underspecified in HCC coding and documentation. Major depressive disorder, or clinical depression, is a condition that affects roughly 15.7 million people in the United States over the age of 18.

When a patient has a diagnosis of major depression, if left uncategorized, the diagnosis has no HCC weight. The most frequently reported ICD-10 diagnosis code for depression - F32.9 (Major depressive disorder, single episode, unspecified) - does not map to an HCC code and therefore does not factor into risk adjustment scores.

More specific depression codes, however, do map to HCC codes. If any category is chosen, such as “mild” or “moderate”, the diagnosis does contribute to the RAF score.


-      Be as specific as possible when documenting depression. Avoid broad terms and unspecified codes such as “depression” (ICD-10 F32.9) if a more specific diagnosis is applicable

-      When documenting major depressive disorder, use terms that specify episodic frequency (single or recurrent), severity (mild, moderate, severe without psychotic features or severe with psychotic features) and the clinical status of the current episode (in partial/full remission) in the documentation and coding.