Ask Away Blog: Medicare Risk Adjustment

Risk adjustment is complex, but the right health plan can master three critical fundamentals: accurate documentation, effective member engagement, and the right technology.

Risk adjustment is used to determine Medicare Advantage enrollees' projected medical costs. This helps level the playing field for payers and providers and promotes health plan innovation and efficiency.

What is Medicare Risk Adjustment?

Many are still unaware of what is Medicare risk adjustment. For a certified medical coder, Medicare risk adjustment is about ensuring the diagnoses you capture on a claim accurately reflect the conditions the member is living with. For a health plan chief financial officer, it's about providing the CMS reimbursement for a member that matches the expected costs of the individual's clinical burden.

To calculate a Medicare Advantage beneficiary's Medicare risk score (their risk-adjusted factor or RAF medical abbreviation), CMS compares the patient's clinical conditions to those of a cohort of patients with similar demographics and healthcare needs. The resulting RAF represents the beneficiary's predicted cost to CMS for the upcoming year.

A higher RAF score indicates that the patient's medical records show a heavier-than-average disease burden. This is sometimes a result of poor chart documentation but can also be the effect of HCC coding that doesn't comply with best practice guidelines. Some experts suggest that using data from enrollees' encounters with physicians – rather than solely their diagnoses in a health risk assessment – could make Medicare Advantage risk adjustment more accurate and less prone to gaming. Others note, however, that this would be a costly change to the current system and could drive more people away from plans with lower premiums.

How is Medicare Risk Adjustment Calculated?

As you can imagine, it's complex. But the goal is to provide accurate information about enrollees to CMS so that it can properly compensate health plans for their risk-adjusted costs. That information consists of clinically related diagnosis codes submitted by providers on medical claims. The data is collected in two ways: prospective and retrospective.

Prospective risk adjustment involves predicting future healthcare costs based on the data of current-year diagnoses. This determines the per-member per-month (PMPM) payment CMS will make to an MAO. This payment is a portion of the Medicare reimbursement beneficiaries receive from CMS. To calculate a member's risk score, CMS uses a regression model that predicts an individual's expected future costs to the system based on an array of factors such as age, sex, and health conditions. This prediction becomes the member's risk score and is then used to transfer funds between insurers based on their relative average claims risk.

Health plans use this information to offer members valuable benefits such as enrollment in exercise programs, case or disease management, and transportation to medical appointments. An engaged partnership between the plan and its members is key to ensuring these programs are effective and deliver value. To do so, health plans must ensure that members' healthcare needs are accurately assessed and documented at the point of care and captured in medical records and claims data. This requires a high-quality, consistent cadence of encounters and supplemental submissions with strong oversight and quality control measures in place.

What are Hierarchical Condition Categories (HCCs)?

Hierarchical condition categories are a grouping of clinically related diagnoses that each has an associated cost to the healthcare system. They are the basis for risk adjustment models used to reimburse Medicare Advantage Organizations based on expected costs. HCCs are identified through submitted claims and encounter data. Only diagnosis codes that map to an HCC will be used in the model, which is why medical billing and coding are so important for healthcare organizations, especially those working with a risk-based payment model like an ACO or bundled payments. Without risk adjustment, Medicare Advantage plans would receive a fixed payment rate that doesn't take a patient's individual health status into account. This is why it's so important for physicians to document and code their patients correctly, especially when their disease state progresses or leads to serious complications. It's equally important for medical records to be reviewed annually to ensure that chronic diagnoses are documented and coded properly.

For example, HCC 17 Diabetes is a chronic disease that requires a lot of healthcare resources to manage. It's more expensive than HCC 18, which includes diabetic retinopathy, neuropathy, and kidney disease, which are less chronic and do not require extensive healthcare resources. It's also critical to capture all of the complications a patient is experiencing through documentation and coding, as a single diagnosis code may not be enough to qualify a patient for an HCC.

How is a Hierarchical Condition Category Calculated?

The risk adjustment model filters ICD codes into groups (or "hierarchies") of conditions that have similar associated costs to the healthcare system. Each of these HCCs then carries a numeric value that maps to a specific risk factor score. For example, patients with multiple chronic conditions may have HCCs grouped and assigned an HCC score of 2. This risk factor score represents the likelihood that this patient will generate high utilization and costs over time. The health plan receives payment based on the number of HCCs documented in each calendar year. It is vital that the health plan accurately captures and records all HCCs during each coding period. Otherwise, it will miss out on important payments. This is why training staff to understand the nuances of HCC coding is so important.

In addition to accurate HCC coding, the health plan must collect and report ICD-10-CM diagnosis codes related to a person's socioeconomic circumstances or psychosocial factors. Payers often use a subset of ICD-10-CM diagnoses called Z codes for this purpose, but whether and how these code sets can be used in risk adjustment is still being determined.

Some experts argue that CMS' current models undercompensate for selection by allowing insurers to avoid higher-risk enrollees. However, economic theory and empirical evidence suggest that this is unlikely to be the case. If anything, it seems likely that CMS' existing methods are underestimating differences in claims risk between plans that attract lower- and higher-risk enrollees and that a two-stage estimation procedure would make these differences even more pronounced.