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.
Claims data is an overview of care activities for a given patient population. The scope is limited to the services provided to a patient and the billable interactions between insured patients and the healthcare system. While claims are a good reflection of data that's required to facilitate payment by an insurance company - such as tests, procedures, and services provided - they lack information on some key characteristics that affect patient outcomes.
Lab results, clinical notes, vital signs, and allergy lists are some of the most crucial elements not reported on claims. Valuable insights such as lifestyle factors and habits (i.e. smoking and alcohol use), behavioral health, functional and cognitive impairment, non-prescription drug use, immunizations, patient surveys, biomedical and genetic analysis, and consumer-generated health data are also not included on claims.
Take vital sign data for example. LOINC is the coding standard for vital signs. But most provider organization do not use LOINC codes to capture vital signs in their EHRs as it is not mandated by the Meaningful Use program, so this data is often missing from claims. Abnormal vital signs can indicate that respiratory, cardiovascular or neurological systems may be failing - which would require the soonest possible intervention to avoid acute events, hospitalizations and potentially life-threatening consequences. Monitoring these datasets over time is also helpful for early intervention efforts - patients with blood pressure readings that have been trending upward, for example, can be contacted to coordinate further evaluation and treatment.
The details contained within clinical data can also be used to ensure risk score accuracy. Coding errors are relatively common in risk-adjustment arrangements, especially in hierarchical condition category (HCC) models. The most common errors are related to lack of specificity, downcoding the severity, or using a code that doesn't map to an HCC even when an HCC code is applicable. Sometimes a condition has not been diagnosed at all, despite evidence contained in the clinical data. By incorporating clinical data, plans can identify missed diagnosis codes and coding errors that otherwise would have been missed due to inadequate documentation.
The other key distinction between clinical and claims data is availability. Due to the claims processing cycle, it often takes a month or more for member activity to be reported. The process is notoriously fraught with delays starting with the initial submission, as many facilities and provider offices submit batches of claims every week or every few weeks rather than at the time of service - rendering the information out-of-date by the time the claim is even received by payers, let alone acted upon. Incorrect or incomplete coding resulting in requests for revision or denial further delays the process by months - even a year in some cases.
In contrast, clinical data is generated constantly, often in real-time, or near real-time, and therefore offers a much more relevant and up-to-date depiction. It can also be aggregated and shared as it's generated, which creates opportunities for payers to intervene proactively, respond quickly, and plan ahead more effectively.
While the lag time associated with claims data affects revenue cycle management in fee-for-service arrangements, the consequences are much more significant in a value-based environments. The time period between a care event and when the claims data is available occurs when care coordination and guidance is needed the most. The typical lag time between a hospital visit, for example, and initial claims submission is 3-6 weeks. If payers conduct follow-ups on hospital visits based on incoming claims, the outreach will occur at least a month after the actual encounter, when any guidance they can offer is usually no longer relevant.
Payers who incorporate clinical data, such as ADT alerts, into care management workflows, can be notified in real-time or shortly members are admitted, discharged, or transferred to or from hospitals. With this knowledge, payers can reach out to members within days of the visit to guide them through next steps and facilitate follow-up appointments with in-network providers.
More timely outreach based on real-time clinical data increases the likelihood that the patient receives the necessary follow-up care and minimizes the chances of readmission. Connecting with members during a care event, when there’s an opportunity to add value to care coordination efforts, can also make all the difference in terms of member satisfaction.
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.