Alternative Payment Models support the need for data analysis—right now

July 10th, 2017 / By Barbara Aubry, RN

As a regulatory analyst, I do a lot of research. For years I’ve read a lot about pay-for-performance, quality initiatives and outcomes, Alternative Payment Models (APMs) and every payment method CMS has come up with in all their acronym glory. Early on I read repeatedly about the need for more data; cleaner, more accurate information to power the healthcare system. Admittedly, I always checked the authors of the articles and found many were IT professionals who of course, want more data—their job is data! After all, ICD-10 was going to give us even more accurate data—if we could just get those pesky clinicians to use it properly.

Articles authored by the healthcare industry (clinicians, hospitals and payers) demanding more data were few. Many view data generation, collection and storage as a burden that prevents them from spending more time with their patients—or paying claims.

Clinical colleagues I’ve spoken with find data to be the bane of their existence; they feel an inordinate amount of time is wasted creating it. I mostly agree with their lament since I’ve seen reams of poorly documented medical records that coders are hard pressed to assign a proper ICD-10 code to. Plus, healthcare is something like an ocean liner—it cannot make quick turns—so I thought all the talk about data was somewhat premature and there was still plenty of time to deal with it. Or at least there was still time for clinicians to decide what data was of real value to them since it seemed all data was of some value to someone in IT. Much data has been generated but many are not at all sure what to do with it. That was my notion—until about a week ago when I read a white paper from CAPG.

CAPG 2017 White Paper findings

CAPG (originally known as the California Association of Physician Groups) changed its name in 2013 to CAPG: The Voice of Accountable Physician Groups, since it now represents providers from all states involved in APMs. I do not know enough about this group to endorse them, but I did find their white paper titled “CAPG’s Guide to Alternative Payment Models: Case Studies of Risk-Based Coordinated Care 2017” quite interesting. The paper presents case studies based on provider experience while currently participating in multiple APMs. What I found especially intriguing was the idea that the case studies were reported by clinicians in the trenches doing their job(s) of providing care. Real life scenarios that support the type and scope of data important to the payment model.

The report looked at cases from:

  • A bundled payment model for commercial and Medicare Advantage lives
  • Next generation Accountable Care Organization for traditional Medicare
  • Managed fee-for-service for a Part C Medicare Advantage population
  • Global Risk model for Medicaid managed care
  • Subcapitation for commercial and Medicare Part C
  • Commercial pay for performance

Each model represents participation by a different provider group or network focusing on differing patient populations and services, (e.g. bundled payment for a joint replacement episode of care). What I found highly interesting were their successes and challenges. For example, in the bundled payment model they reported length of stay (LOS) reduced by 50 percent for commercial lives and 32 percent for Medicare beneficiaries. In this model they realized better value for patients and improved efficiency for physicians but they listed “improve access to data” as a key suggestion required to improve delivery of treatment in the model. Specifically, “data is critical in identifying where the highest percentage of costs reside and in assessing variation. However, additional data is needed to reach the level of detail necessary to provide individual physicians with their shared savings. For example, reconciling information by individual patient is a challenge but it’s necessary to establish the bundled price and assess physician performance.” Hmmm; so providers involved in this type of model right now can clearly see the value of accurate analysis of unstructured data? Interestingly, much data is already available but unstructured and not yet easily quantified. (Unless of course you were to work with some of my really smart 3M HIS colleagues who are expert at capturing and analyzing unstructured data).

It’s all about different data needs

The advanced ACO model folks reported the need to “receive more information directly from CMS to help us assess our performance and analyze population health.” Interesting; so it’s not just providers who need analyzed data but there is also a clear need for more sophisticated, analyzed and structured data from CMS.

The managed fee-for-service Part C provider group internally creates highly specific population health data based on documented, clinical patient risk. Their data is then transmitted to their payer partners who then send it on to CMS. But the provider group has no idea what their data actually looks like when it’s received by CMS. Do the payers share all the data the physicians provide? They want to “improve (data) transparency and consistency across health plans.”

“One area of improvement is speed up and improve the accuracy of the data transmitted from Medicaid to payers to the medical group” was a challenge for the pediatric group involved in the global risk payment model for Medicaid. The doctors in the group need accurate patient data ASAP from Medicaid and they are concerned with what they receive. They also want standardized data transmission formats which makes the data easier to work with on the clinical side.

Sixty-six percent of the participants in the programs above need more from their data. Whether the focus is accuracy, transparency, feedback or specific detail, data is critical to the continued development—and success—of alternate payment methods. So, if you are like me and believe you can pay attention to this in the future I hope I’ve been able to open your eyes—as mine were. This need cannot be put off “‘til next year”, or “wait until next year’s budget planning.”

Maybe we do need more data in the future, but let clinicians be involved in the conversation. They can determine what is most important based on their workflow. The future is here right now; clinicians are telling us what they need—and it’s not necessarily a new EHR. Better analysis and use of the unstructured information available now is required today in order to support where we hope to take healthcare in the U.S. I believe this is something the industry needs to accept and embrace if the new payment models are to realize the hoped for quality and savings.

Barbara Aubry is a regulatory analyst for 3M Health Information Systems.