From HCC to CDPS: A Primer on Common Risk Adjustment Models

With multiple models and complicated formulas for calculation, healthcare risk adjustment is not a subject for the faint of heart. If you’ve ever been stumped by this difficult topic, we’re here to help. Read on as we dive into the realm of risk adjustment and break it all down into understandable terms.

So, what is a risk adjustment model? It's a statistical tool used in healthcare to forecast and assess the potential healthcare expenses associated with beneficiaries. This predictive model calculates the relative actuarial risk of enrollees, taking into account various factors such as demographics, health status, and pre-existing conditions.

By looking at these factors, the risk adjustment model aims to predict how much healthcare might cost for each individual. This helps ensure that money is distributed fairly among various health plans. The main aim of the model is to reduce the effects of different risk levels among enrollees and create a more even financial system in healthcare. (1) In the following paragraphs, let’s dive into various types of risk adjustment models for the Medicare, Medicare Advantage, Medicaid, and ACA markets.

Decoding the HCC Model

Think of hierarchical condition categories (HCCs) as the coding language of risk adjustment. The HCC list is like a menu of these codes, each with its own specific value. These codes group together based on similar health patterns. So, if you've got a condition that's likely to impact your healthcare spending, the HCC list makes sure it's accounted for. There are two types of HCC models — the CMS-HCC model and the HHS-HCC model — that we will explain in further detail below.

Medical coders use the ICD-10-CM coding system to figure out how much risk a certain diagnosis might bring. They can find HCC info using spreadsheets or coding software. When they’re coding a medical visit, they check the relevant diagnosis on the HCC list. If the diagnosis is considered important for HCC, they link it to a specific risk number.

Understanding Medicare Risk Adjustment Models

Now, let's shift our focus to the domain of Medicare Advantage, where the practice of risk adjustment takes center stage. Health plans provide historical medical data from the previous year and CMS mixes in demographic factors to derive a risk score for each beneficiary. The higher the score, the more support a plan might receive to cover that individual's healthcare costs. (2)

Let’s dive into a specific example to see how this works. (2)  Let’s say a health plan’s base monthly payment is $1,000. A senior woman aged 77, dealing with diabetes and congestive heart failure, attains a risk score of 1.344. In contrast, a 65-year-old man devoid of chronic ailments might carry a risk score of 0.4. The Centers for Medicare & Medicaid Services (CMS) modifies the monthly payments for plans to reflect the health status of each member, taking into account these calculated risk scores. Hence, the monthly payment for the woman would be $1,344 and the man would amount to $400.

Among the array of risk adjustment methods, those for Medicare are the most widely employed, particularly finding its stronghold in Medicare Advantage Organizations (MAOs), also known as Part C. A prospective Medicare beneficiary can opt for traditional Medicare Part A or B or to explore the realm of MAOs based on eligibility criteria.

Nuts and Bolts of the CMS-HCC Model (3 & 4)

The CMS-HCC model, used in Medicare Advantage, aims to calculate payments to health plans based on the health condition of enrolled beneficiaries. This model is a specialized adaptation of the broader HCC framework, tailored to suit the Medicare demographic. In addition to employing HCC codes, it leverages key demographic features to predict the potential healthcare costs of Medicare beneficiaries.

Recent Changes to the CMS-HCC Model

On March 31, 2023, CMS issued the Notification of Calendar Year (CY) 2024 Medicare Advantage (MA) Capitation Rates and Part C and Part D Payment Policies. This notification, commonly known as the "Final Rule" within the MA industry, highlights modifications from the V24 to V28 HCC mapping.

Notable changes include:

• V24 encompasses a total of 86 distinct HCCs, whereas V28 expands this spectrum to encompass 115 exclusive HCCs.
• Within this transition, 2,297 diagnostic codes that were previously aligned with an HCC in the V24 model no longer find correspondence with any HCC in the V28 model.
• Conversely, 268 diagnostic codes now exhibit an association with an HCC in the V28 model, although such a linkage was absent in the V24 counterpart.