The Petitio Principii of ACOs

Dan Morreale, IT Strategist

The serious lack of infrastructure to support an expansive and meaningful ACO strategy is missing from the healthcare technology toolkit. Some may view this as a non critical issue but with so many organizations engaged in some form of ACO activity, we have to ask, “Have we begged the question?”

It seems to me we have concluded an accountable care organization can be enabled on a premise that lacks support. Given the current state of HIT, we should be struck first by the fact that information systems on the provider side are fragmented and do not facilitate the type of exchange of data needed to drive performance up and cost down. With few exceptions, we do not have information systems that create a functional sharing of medical information sufficient to assure that patients get the appropriate care without duplication and in a common manner.

Some systems like DocSite and Axsys Health, as well as a few others, are starting to be incorporated in EMR implementations and, by so doing, move the mission forward. But there is still much work to be done. Let’s look at a few of the challenges and opportunities that our health systems and vendors could embrace to advance the effort.

First, the lack of consistent standards in health care, for me at least, is call number one. Anita Golderba, in a presentation to the International Federation of Library Associations in 2008 stated, “Standards are like toothbrushes, a good idea but no one wants to use anyone else’s.” Sound familiar? We need real standards. Items like the CCD which most of us have embraced are a start, but when we realize that the CCD is mostly optional, it is less a standard and more of a starting point. The work at the IHE in creating both a framework and profiles takes us a bit further, but adoption is slow and costly. As customers, we must turn to our vendors and insist on more engagement around standards, then work with them to “operationalize” the results. I think the certification process will help, but we still have more work to do.

Second on my list is the lack of definition. We need a vocabulary, syntax and translation process which is effective and adopted. We need a Lexicon. Oh, we have vocabularies. ICD-9-CM, ICD-10-CM, SNOMED, RXNORM, LOINC, CPT, HCPCS are all vocabularies, but they each have a different file format. Syntax can help. HL7, ANSI, ARDEN have all been adopted by most of us, but the coordination of vocabulary and syntax is still missing. XML brings us a bit closer. Translation of data allows for the extraction, transformation, and exchange of data, but most of this is still very proprietary. The combination of vocabulary, syntax and translation produces a Lexicon of care which is still in its infancy and needs to grow to become effective.

Third is the lack of connectivity. This takes two forms. One is the lack of electronic health records in physician offices, and second is the tools by which we share data, i.e. health exchanges.

Although on the rise — the efforts of ONC and CMS to promote electronic health records are in part a testament to this — there is still very low penetration of EHRs in physician offices. Full use of an EHR for orders, notes and results seem to be under 32% (my best guess reading several 2010 surveys and not offered as a statistical sampling). This low adoption rate will derail the sharing of information at the heart of most ACO activity. Additionally the implementation of health exchanges to facilitate the movement of data between and among care givers is low. In my state, New Jersey, where there is a monumental effort to build a state wide health exchange, there are at least 12 different HIE projects underway at a variety of health systems, and only four are sharing a limited amount of clinical data. None of the efforts are sharing data between exchanges at this time.

Fourth we must address the lack of identity integrity. We do not have in this country a solid means of identifying an individual in relation to a health record. Identity integrity represents the accuracy and completeness of health data associated with an individual patient, as well as a reliable and valid means of denoting that the patient presenting is in fact the individual whose data is before the clinician. Although the 1996 HIPPA regulation mandated a healthcare identifier, the political climate around privacy caused Congress to enjoin HHS from moving on final regulations. The value of a national healthcare identifier is well documented and, as care givers and care receivers, we must implore Congress to readdress the issue. Barring that, we must look at a voluntary patient identifier such as the idea promulgated by the IEEE-USA Voluntary Health Care Identifier which helps to better identify the individual and the data associated with care.

Fifth the lack of an episodic grouper will hinder treatment. There are few episodic groupers that will allow health care providers and payers to review the treatment of a patient across inpatient, outpatient, pharmaceutical, and diagnostic events. This type of tool is imperative for those ACOs with a high risk model, so as to best determine the distribution of payments. It is also high on the radar of any ACO looking to extend quality and reduce cost. We need to know the entire course of care for a single illness or disease state. Effective disease management models will require a comprehensive grouper tool to provide a better understanding of treatment, cost and treatment quality. The need for clinical metrics, severity adjusted analysis, and structured data mining is imperative in the cost-quality paradigm powering so many ACOs.

It time for us to get to work and look at putting in place the tools, process and infrastructure needed to make ACOs and healthcare delivery effective. Without this type of infrastructure, data sharing will be difficult, not timely, and prone to errors. If however we can improve the infrastructure, we will be well on the way to building an appropriate foundation for care and can start talking about contextual EMR which will help to deliver the salient information for a provider at the point in care where it is most effective.