Why HCC Recapture is Not Enough
HCC code recapture of previously documented, or "known", conditions is a starting point - not a complete strategy - for assessing risk. Recapture updates a patient's risk profile based on codes submitted to payers in the previous calendar year, which does not factor in many common scenarios.
Incorporating the full picture of risk - including new and suspected diagnoses as well as previously documented conditions - into stratification, care management and RAF score calculations is vital to success in risk-based arrangements. Let's explore some of the scenarios that HCC recapture of known conditions does not take into account.
Obesity is a common example of this issue. Any patient who has a BMI greater than 40 qualifies as morbidly obese. However, often only the BMI for these patients is documented, and not the corresponding diagnosis of morbid obesity. Without the additional morbid obesity code, the BMI diagnosis carries no HCC value and therefore would not be logged by recapture tools. Bronchitis is another example where a risk-adjustable code is often applicable, but not used. If bronchitis is coded but not specified as acute or chronic - it does not map to an HCC code. Chronic bronchitis, on the other hand, does contribute to a patient's risk score.
Many previously documented chronic conditions persist for patients which are not regularly documented through claims systems, encounter forms or superbills. An amputation, for example, might not be evaluated unless the patient complains of skin issues. Or an old myocardial infarction, which causes permanent damage to the heart muscle, might not be discussed during an encounter years after diagnosis if the patient is being treated for an unrelated issue.
Many chronic conditions progress even when caught early and managed properly. If a condition was coded in a prior year, it's possible that the disease has progressed or the patient has developed related complications. For example, a patient with diabetes could have developed retinopathy or a renal manifestation. Recapturing a condition that has progressed or developed a complication can be a more significant case of undercoding than is often realized. Chronic conditions with complications usually have a higher HCC weight than conditions without complications, and, in some instances, the condition without a complication does not have any HCC weight, so recapturing in this scenario would almost always translate to an artificially lower RAF score. In the case of diabetes, ICD-10 codes for diabetes with complications carry a RAF three times higher than a diagnosis of diabetes uncomplicated.
Predictive modeling is a technique that can be used to detect the statistical probability of additional diagnoses based on a comparison to patients with similar backgrounds and conditions. These algorithms look for similarities among patients with a shared condition, and reviews the similarities in context of the patient’s profile to determine if the shared traits are predictive. The suspected diagnoses can then be presented to providers at the point-of-care for assessment. Due to recent advancements in AI and machine learning, predictive modeling has become an increasingly common and reliable method of identifying undiagnosed conditions for earlier, more effective care.
Payers in value-based contracts are expected to move beyond their traditional role as insurance purveyors and assume the broader responsibility of improving outcomes and care coordination. Claims data is not a sufficient foundation for successfully fulfilling this role. By leveraging the valuable insights and real-time availability that clinical data offers, plans can carry out the level of proactive outreach and crises prevention that value-based care management demands.
AI is no replacement for human judgement but it can go a long way towards simplifying and streamlining data management and analysis. AI-enabled chart review helps providers and plans that coordinate care for complex populations work smarter, not harder, by shifting the focus from volume of charts targeted to precision targeting of charts.
The Centers for Medicare & Medicaid Services (CMS) initiated the Hierarchical Condition Category (HCC) model in 2004 to adjust payments to Medicare Advantage Organizations (MAOs). The model has been more prevalent in recent years as HCCs are becoming more widely recognized as one of value-based programs' most important components. This heightened visibility is due in large part to the growth and success of Medicare Advantage.