Why Medical Risk Adjustment Matters
Medical Risk Adjustment (MRA) is the process of modifying payments and benchmarks to predict health plan spending based on enrolled beneficiary/patient population health status.1 MRA allows the Centers for Medicare & Medicaid Services (CMS) to estimate future spending for health plans, while allowing providers to understand the health characteristics and disease burden of their managed population.2 MRA is used in several value-based care models to reward physicians for providing high-quality, effective care. Initially employed by CMS to adjust capitation payments to Medicare Advantage (MA) plans, MRA is now used to calculate expenditure benchmarks for Medicare Shared Savings Program (MSSP) Accountable Care Organizations (ACOs) as well. This toolkit presents two distinct MRA models. Although both are similar in purpose and structure, they each have unique characteristics that address different patient populations. This toolkit references the CMS-HCC model more frequently to feature Medicare patient population best practices and highlight relative chronic conditions that “risk adjust” under the CMS-HCC model.
The two MRA models covered are:
- CMS-HCC (Hierarchical Condition Categories) risk adjustment model. MA plans and MSSP ACOs use this model to set contract rates. Generally, this model estimates higher expenditures for sicker beneficiaries and lower expenditures for healthier ones.3 • HHS (Department of Health and Human Services)-HCC risk adjustment model. Commercial health plans began using this model in 2014, influenced by the Affordable Care Act. CMS payments to contracted health plans are adjusted by patient demographics (such as age, gender, Medicaid eligibility and disabled status), diagnosis claim data, medical chart notes, and financial status.
- HHS-HCC characteristics also include disease interactions and diagnostic sources, and they are predictive of future medical anddrug spending.
- Medical Risk Adjustment – Medical Risk Adjustment methodology aims to offset the cost of providing health coverage for patients who are chronically ill and represent a relatively high-risk population. Medical Risk Adjustment use patients’ demographic data (age, sex, etc.) along with diagnoses to determine risk scores.
- Hierarchical Condition Categories (HCCs) – HCCs are used to calculate risk scores. ICD-10 diagnoses are ranked and grouped into categories based on similar cost patterns. Higher-weighted HCCs represent higher predicted healthcare costs. For example, diabetes with complications is ranked higher than diabetes without complications. Patients
may be included in more than one HCC grouping.
- Risk Adjustment Factor – Risk scores, also known as Risk Adjustment Factor (RAF) scores, are used to calculate a patient’s projected care and associated costs using
both retrospective and prospective models in medical risk adjustment that compares chronic conditions that a patient was treated for in the previous calendar year, compared to the current calendar year.
- A patient’s specific demographic values plus HCC values equals their RAF score. The sicker the patient, the higher the score will be above a normalized or healthy RAF score of 1.0 (Example: A patient with a 3.75 score denotes chronically ill status).
- Each patient’s RAF score is erased on January 1 of the current year. This helps to ensure funding is available from the Centers for Medicare and Medicaid (CMS) and improves health plan premiums per member per month (PMPM).
- Risk Score – A risk score is a prospective metric for patients that help indicate whether a particular outcome may occur. Typically, these scores are based on utilization events, which include hospital admissions and emergency department visits, and can sometimes indicate a clinical state for a patient, such as heart disease, diabetes, cancer, etc.
- A risk score is different from a medical risk adjustment factor score, which is based on algorithms used to compare patients exhibiting similar clinical patterns to help predict what outcome may occur. Risk scores are often included in reports generated from health insurance payers, electronic health records (EHRs), and through the manipulation of claims data.
- Risk scores can also allow payers and providers to define challenges and benchmarks for a given population. For example, a risk score for cardiovascular disease may include metric data such as age, diabetes status, smoking history, body mass index (BMI),etc. This data can help to develop preventive strategies and effective clinical treatment.
- Prospective Review – Value-based models, healthcare providers, ACOs and health plans must attempt to use this medical chart review process prior to patient visits to improve accuracy of documenting diagnoses at the point of care throughout the calendar year. This review process should result in significant impact on patients’ RAF scores and improved benchmark targets under risk and value-based arrangements.
- Concurrent Review – This medical chart review process has a “clean as you go” approach and allows clinical staff to review medical record documentation on an ongoing basis
as patients are seen throughout the year. This process helps to streamline billing, reduce denials and decrease inaccurate claims. A patient’s RAF score is derived solely based on documentation of diagnoses and other social determinants of health elements not always listed in the medical chart.
- Retrospective Review – This medical chart review process identifies relevant encounters and patient assessments after patients have been seen in the previous calendar year. It allows missed diagnosis codes that were not submitted on claims in the previous year to be documented during the current calendar year. This process focuses on patients who have already had an encounter during the previous year and current calendar years but whose conditions have not been documented between the two periods.
- Medical Risk Adjustment Models – Medical Risk Adjustment models are developed by CMS and use HCCs to reimburse healthcare organizations differentially based on disease burden and demographics. CMS uses two models: CMS-HCC, most commonly used by original Medicare and Medicare Advantage organizations, and HHS-HCC, developed through the passage of the Affordable Care Act (ACA) and including additional categories for all patient demographics.