How COVID-19 Is Heightening Demand for AI-based HCC Coding Technology
We recently spoke with Derek Kren, VP of Business Development at Inferscience, to understand how COVID-19 will likely affect long-term health and what that means for Medicare Advantage reimbursements and provider funding. Get a brief understanding of HCC risk adjustment and coding technology in 2021, or you can check out the full webinar here.
What is Risk Adjustment & HCC Coding?
Risk adjustment refers to the variety of methods used by the federal government, such as Centers for Medicare & Medicaid Services, and other health plans to offset the cost of providing health insurance for individuals who represent a relatively high risk for insurers. While there are many types of risk adjustment systems, one of the most commonly used models is CMS-HCC model, which is a way to determine how the provider gets paid.
More specifically, hierarchical conditional category (HCC) codes are used by CMS to reimburse Medicare Advantage providers and predict the cost to care for each patient. HCCs are mapped to an ICD-10 code, and along with demographic information such as age and gender, a risk adjustment factor (RAF) score is assigned.
While the system is almost two decades old, the continuous push towards value-based care signals a potential adoption by private payers in the near future as well. In addition, the number of Medicare Advantage enrollees climbed from 24.4 million in 2020 to 26.9 million in 2021, and CMS projects that amount to increase to 41 million by 2030.
How COVID-19’s Impact on Long-term Health Affects Practices with Medicare Advantage Patients
Video visits and overall telehealth adoption has certainly boomed during the pandemic, but patient utilization - the overall use of healthcare services to prevent or treat conditions- has decreased. Non-COVID related emergency department visits decreased by 42% in the early periods of last year’s stay-at-home orders, and urgent referrals for cancer detection have dropped, in some cases by 84%.
Not only does that have grave implications for long-term health, but it also means fewer opportunities to track HCC codes and receive the necessary levels of funding to care for patients that need it most. For example, mandated annual wellness visits will be that much more critical, and with the decrease in preventative care, the eventual burden on acute care could drive up costs and place more pressure on practices.
Democratization of Natural Language Processing & New HCC Coding Technology
Natural language processing is a subset of artificial intelligence that refers to technology’s ability to understand human speech in oral and written forms. While it’s used in a myriad of industries, its impact on healthcare has started to unfold in recent years, as the analysis of large amounts of patient information in EHR platforms can be used for clinical decision support, medical coding and more. Even speech-to-text assistants leverage NLP to cut down on clinical charting time and tedious data entry tasks. Beyond these administrative burdens that physicians often face with their EHR systems, though, there’s an additional layer of potential burnout that accompanies intricate medical billing and coding responsibilities. With NLP, however, a patient’s chart can analyze the necessary information to extract and identify codes in seconds.
Previously, this type of technology was only available to behemoth insurance companies, but over the past several years, the democratization of AI-based tools are giving providers and staff the opportunity to spend more time with patients and lift a large part of their medical billing and administrative burden.
Accurate and efficient HCC coding isn’t important just for the sake of funding, but it’s critical to receive the resources needed to care for patients that need it most. And without the use of technology, maximizing payments for Medicare Advantage patients will only become more difficult. For more information, listen to the DrChrono & Inferscience webinar here, or check out the Inferscience partner page.