Newtown, CT – June 3, 2016
Beyond the mind-numbing rules labyrinth that defines the proposed MACRA regs, few elements illustrate the implementation madness better than the CEGs, PCGs and PRGs – For those who have come to understand that US healthcare is as replete of arcane acronyms as the military (and likely outdoes the military in this domain as it does in spending), the three new gems in the lexicon refer to care episode groups, patient condition groups and patient relationship groups. Like many elements of MACRA, it stems from good ideas and observations of data deficiencies that can lead to poorly designed incentives. In a new payment world, understanding with greater precision why a service is rendered, the underlying need of the patient at the time the service was rendered, and the relationship to the patient of the provider delivering the service, all vastly improve the manner in which providers are held accountable. Today’s claims coding systems, which rely principally on placing certain diagnosis and procedure codes on each service delivered, leave much to interpretation on the back end when assembling these services into coherent units of accounting and accountability. In a fee-for-service payment environment, it simply doesn’t matter that much. But in an environment in which physicians are at financial risk for the management of a patient’s illness, injury or treatment, it does matter. Many of us who toil in this field daily have developed analytic tools that try and compensate for the lack of precision in the data. Our own tool, the PROMETHEUS Analytics, by most accounts, does a fairer job at compensating for the lack of precision than others, but we’re first to acknowledge it’s not perfect. And, seemingly, perfection is what MACRA aims for, which, in the hands of the government, rapidly turns into lunacy.
What this means to you – On October 1st, 2015, all of the claims billed by physicians, hospitals and other providers in the U.S. started using ICD-10 diagnosis codes instead of ICD-9 diagnosis codes. This conversion, which was delayed twice due to legitimate concerns about the implementation timeline, caused every health plan, every provider, every analytics company and all those who handle, store and distribute claims data to make changes to their information systems. The change was to accept, in a currently used field on a claim, one code value for another. It was a significant undertaking which is still rippling through. The total estimated cost to the industry of making this shift was billions of dollars, and the juice does not seem to have been worth the squeeze because the additional specificity of ICD-10 over ICD-9 is negligible. What is contemplated here, potentially, is the addition of three new fields with new codes. As mentioned last week it would make the implementation of ICD-10 feel like a gentle stroll in a pleasant park on a warm spring afternoon. And yet we need this new information if, and only if, it’s constructed the right way. Our friend Harold Miller has a report that provides good detail on what would be useful and is worth reading. But, as can now be expected by this Administration’s CMS, the direction proposed is not particularly useful, will create an additional burden on physicians, disrupt health care billing practices and add to the MACRA/MIPS lunacy. Having timely, reliable and actionable information feedback loops is an essential part of practice transformation. To achieve that goal, adding new information on claims is essential. For example, ICD codes don’t tell you which cancer stage or tumor type a patient has. Care episode groups could. ICD codes don’t tell you either whether the cancer patient is struggling with activities of daily living, but patient condition groups could. And finally, neither ICD nor procedure codes tell you whether the provider rendering the service is the oncologist primarily responsible for managing the patient’s cancer, or simply one that is consulting on the case. Patient relationship groups could. Clearly, adding three new fields on claims, agreeing on a code structure and nomenclature, and implementing the needed information technology changes in every claims billing system, claims data warehouse, analytics application, and many other systems would be a massively daunting effort. Would the juice be worth the squeeze? Maybe, but only if done right, in full and complete collaboration with the industry. Do we trust CMS to get it right on its own and impose its rule on all of us? No freaking way. But that’s the lunacy that is staring us in the face right now. And that’s what we have to stop before it’s too late. Because with what CMS is serving up now, not only is the juice not worth the squeeze, it’s toxic.
Francois de Brantes