Advanced Revenue Cycle Analytics: An Overview

Medicare Advantage plans were the first to use HCC for Risk Adjustment in 2004.  The way the Risk Adjusted program worked was relatively simple: the Medicare Advantage plan would assume both upside and downside risks for their patient population, and CMS would have a fixed price for these patients.  However, the critical element to making this work and avoiding “cherry picking” was HCC Risk Adjustment. 

Without HCC Risk Adjustment, Medicare Advantage plans would receive a fixed capitated per patient rate.  The strategy in this scenario would be very clear, avoid the sickest most expensive patients.  Without Risk Adjustment, Medicare Advantage plans would work hard to find and enroll the healthiest Medicare patients possible, leaving only the oldest and sickest patients in the Medicare risk pool.  For the Medicare Advantage program to successfully reduce Medicare expenses for CMS and improve quality of care for some of the sickest Medicare beneficiaries, Risk Adjustment was essential.

Medicare Advantage plans have responded by covering annual wellness visits for their members, where primary care physicians capture all patient problems.  Another strategy Medicare Advantage plans use is chart audits, where they ask for the medical records of their members to review documentation for diagnosis that are documented but not coded.  Medicare Advantage plans spend tremendous resources making sure to accurately reflect patient acuity and HCC scores, because they would be upside down on Risk Based contracts without this data.

The HCC model would assign patients an HCC risk score based on demographics and diagnosis history.  This average HCC score would always be 1.0 (this is critical to understanding HCC, I will explain more later).  The risk score would be multiplied by the average cost per Medicare beneficiary. The final outcome would be the amount that Medicare would pay the Medicare Advantage plan.  This system properly aligned incentives for Medicare Advantage plans to maximize profits by finding patients with higher HCC scores, where the Medicare Advantage plan thought they could take better care of these seniors, keep them healthy and out of the hospital.

The average Medicare patient has an HCC score of 1.0.  If a patient has an HCC score of 1.10, that patient is expected to have 10% higher healthcare expenses over the next twelve months, compared to the average patient.  Similarly, a patient with an HCC score of 0.90 is projected to cost Medicare 10% less than the average patient.  The demographic portion of the HCC is fixed; there is nothing that a practice or Medicare Advantage plan can do to change it.  However, the diagnosis portion of the HCC score will vary based on how accurately all diagnosis and comorbidities are coded.

Medicare-Advantage--HCC-Coding.PNG

In the example above, the patient’s demographic portion (top two rows) of the HCC score is fixed. However, notice the difference to the HCC score when changes are made to the diagnosis coding and disease interaction portions.  In this example, only three diagnosis codes differ between “Some Conditions Coded” and “All Conditions Coded”: 1) Diabetes with complications instead of Uncomplicated Diabetes, 2) Vascular Disease, and 3) CHF.  While this may seem like a small difference, the final outcome shows that there is significant risk to the Medicare Advantage plan if the provider does not code these diagnoses.

In order for Medicare Advantage plans to survive in a Risk Adjusted reimbursement model, it is essential to accurately capture all diagnosis coding and comorbidities. This patient has conditions that could complicate his diabetes, and these conditions will ultimately result in higher healthcare expenses over the next twelve months. Correct coding ensures that this patient receives the correct risk adjusted dollars per beneficiary from Medicare.

The HCC model is very accurate in predicting future healthcare expenses. The amount that Medicare will pay for this patient will differ greatly based on which diagnosis codes and comorbidities are captured. In this example, the model predicts that the patient will cost $16,380.52 in the next year. That is the amount that Medicare would pay the Medicare Advantage plan for that specific patient.  However, if some of the patient’s diagnoses are not recorded, Medicare would only pay the Medicare Advantage plan $6,434.56 for that patient.  The failure to accurately capture all of the diagnoses makes it impossible for the Medicare Advantage plan to make any profit. 

If you are interested in learning more about how your practice can improve coding HCC comorbidities, please contact hcc@whiteplume.com