One of the primary goals for Medicare Advantage (MA) and other value-based care (VBC) programs is to effectively manage patients’ chronic conditions and avoid complications, including hospital admissions and re-admissions, and costly interventions. To accurately compensate providers for the care of these patients, VBC programs rely on clinical risk scores based on the diagnoses for which a patient is receiving care.
However, while it may seem like a straightforward process, finding and documenting the information required for accurate VBC risk assessment is not always easy.
Consider Medicare Advantage, which has defined approximately 80 categories of hierarchical condition categories (HCC) covering about 9,500 ICD-10-CM diagnosis codes. A risk adjustment factor (RAF) is assigned for each HCC category, with additional risk factors assigned based on gender, age, living situation, and other factors. These risk factors determine a patient’s overall RAF score; providers qualify for higher payments for patients with higher RAF scores.
To drive higher reimbursement, providers will often seek to identify diagnoses which will result in a higher RAF score. This requires focus on the coding of a patient’s diagnoses, and not necessarily on how well each condition is being managed. In other words, VBC models have created an overemphasis on coding, instead of on caring.
Antiquated review methods
Organizations use a variety of methods to increase RAF scores. One approach is to review a patient’s medical record, typically after the provider/patient encounter, to identify opportunities for “upcoding” a diagnosis to raise the RAF score. For example, if a diabetes diagnosis can be coded as having neurological manifestations, the risk score will be higher, as will the provider’s reimbursement. An organization might have staff review charts manually after a patient encounter to identify potential diagnoses that could qualify for higher risk. There are also a flurry of software products coming to market to help identify opportunities to “optimize” or “manage” risk which, in reality, is just a veiled way of saying “upcoding.”
Unfortunately, neither of these review methods provide the clinician with tools at the point of care to care for the patient; that is, properly manage, evaluate, assess, and treat a patient’s various conditions in accordance with the original intent of VBC: improved outcomes through better management of patient health. Particularly as healthcare shifts to more team-based care delivered across a variety of settings, including in-office, telehealth, home health and more, the need for tools that help clinicians efficiently and effectively deliver care has never been greater. All providers must have instant access to the information and data needed to manage and code each patient’s specific diagnoses – as well as to create documentation that supports the coding.
A comprehensive clinical care approach to VBC, as opposed to an approach that focuses on the transactional coding of diagnoses to maximize risk scores, requires a new set of solutions that enable any provider to quickly see the status of any known or suspected condition. Such technology empowers clinicians to act upon information immediately and helps them create accurate documentation that supports a patient’s diagnosis and satisfies potential quality care and MA audits.
Needed: Point-of-care decision support for VBC
Consider the potential benefits when providers are empowered with point-of-care decision support, diagnostic filtering, and documentation tools geared for VBC:
- Providers can select any patient condition and instantly see the symptoms, history, physical exam, tests, therapies, co-morbidities, and possible sequalae and co-morbidities for any of the more than 9,500 ICD-10 codes qualifying for risk adjustment, complete their work, and create the documentation needed to comply with MA requirements.
- Enhanced diagnostic interoperability between systems with FHIR-compliant, API-based web services. This facilitates the import of information from external systems using more than 1.5 million mappings between terminology standards such as ICD, SNOMED CT, LOINC, RxNorm, CPT, DSM5, etc., to the concepts in the clinical relevancy engine, which has more than 100 million clinical relevancy links.
- With the addition of a natural language processing (NLP) capability, providers can easily identify clinically relevant information in free-text and filter it diagnostically.
- With a single click, providers can see relevant coded or free-text clinical information related to a specific diagnosis, either at the point of care or post-encounter, using an automated chart reviewer. Users can quickly determine if documentation and care is adequate to satisfy an MA audit; ensure that applicable clinical quality measures are captured; and be prompted when additional information is required based on customized decision support needs or other enterprise requirements.
- An enterprise or vendor can easily build and maintain custom protocols to satisfy workflow and quality care requirements. The protocols can be condition-specific or specific to an enterprise or modality such as OASIS data for home care. Such tools do not require custom programming and can be easily designed and built by nurses and other providers, which is critical for scalability and deployment, especially in environments with complex requirements.
The traditional transactional approach to the coding of diagnoses to support billing in a fee-for-service environment is not adequate to meet the care requirements of VBC programs. Going forward, a better approach is required –– one that empowers providers with the information and tools they need to efficiently manage their patients’ care.
Photo: Hong Li, Getty Images
