This post originally appeared on
I recently met with leaders of provider organizations representing a range of specialties and geographies. One of the themes that came through loud and clear was their ongoing struggle to obtain claims data from their payers. As providers continue their efforts to navigate alternative payment models (APMs), their need for global claims data becomes a more pressing issue.
Why Claims Data Is Important to Providers
Many providers are still in a position with little or no downside risk. Even in this early stage, claims data are important. Many of the quality metrics require claims data in order for providers to fully understand their performance. It's common that services not performed by primary care doctors will not be fully captured in their own electronic health record (EHR). Colonoscopy, mammography, and diabetic retinal screening are good examples of this phenomenon.
Perhaps less obvious is the need for claims data in the context of gain-sharing arrangements, even when there is no downside risk. Providers in this setting need to manage costs to maximize their upside potential and can't really do that without a timely, robust claims feed from the payer. Learning how to identify and act on opportunities apparent in claims data will maximize current opportunities and prepare provider organizations for downside risk in the future.
Limitations on Its Value
Claims data, at best, is generally 3 months behind the point of care in terms of giving a clear picture of utilization and cost. This is because it takes that long for an adequate percentage of the claims from a given month to be submitted to the health plan. As a result, claims data are useful for planning and for retrospectively evaluating performance, but have limited value in real time. For providers entering contracts with downside risk, a claims database that shows utilization and cost trends, as well as performance against regional and national benchmarks, can be extremely helpful in identifying opportunities and prioritizing resource allocation for planned interventions.
This same database can also serve as a critical supplement to real-time data that providers obtain through their EHR. A population health platform enables an integrated view that brings together claims data, EHR data, and other information sources, such as a health information exchange (HIE). That approach is foundational to any organization that is committed to APMs that include downside risk.
Obstacles to Obtaining the Data
Given the obvious value of claims data to providers entering into APMs, it's not surprising that they are frustrated by challenges related to obtaining these data. Two categories of issues explain the majority of obstacles. First is the willingness (or unwillingness) of payers to share data. In the absence of regulatory requirements, it's not surprising that, across commercial payers, there is significant variability. Even for payers willing to share, a minimum number of their members must be attributed to that provider for them to be amenable to do this.
This issue of attribution is the second major obstacle. In most APMs, attribution is either based on primary care utilization only, or for a specialist participating in bundled payments, is based on the specialty provider that performs the index service for that bundle. The payers have employed methodologies that studiously avoid the same patient being attributed to two different providers. In this context, payers are generally unwilling to provide full claims feeds to any provider other than the attributed primary care organization. The result is that even if a provider is rendering services to a patient, if that patient is not attributed to them, some data may be excluded and the provider has only a limited view to the breadth of care the patient is receiving outside their office.
Behavioral Health (BH) Is Challenging the Status Quo
This unwillingness to duplicate attribution is particularly problematic for BH providers attempting to both manage and demonstrate their value. The current Medicaid waiver proposal in New York State may help overcome this critical logjam. It represents an important attempt to break the mold and make attribution reflect the integrated nature of health services. This includes specific language describing "... differential attribution methodologies utilizing a member's primary behavioral health provider (e.g., Article 31, 32, 36, or integrated clinic) or Health Home focusing on individuals with behavioral health diagnoses, rather than a primary care only attribution methodology." If this proposal holds, managed care organizations will be required to attribute members with BH services to both primary care and BH providers. That opens the door for BH providers, perhaps for the first time, gaining access to all claims data on their patients, enabling a truly integrated view of their patients' healthcare.
As providers face the challenges of caring for their patients in the midst of rapid change in both the way care is delivered and the way it is paid for, obtaining claims data often represents a frustrating distraction in their day-to-day struggles. That said, if there is a sustainable future for APMs, enabling providers to integrate robust claims feeds with their EHR and other sources of patient information will be foundational to success.
Martin Lustick, MD, is a physician and senior vice president and principal of .
This post appeared on