Sunday, October 17, 2010
HIE Guide for the Perplexed
The Government, through ONC, has awarded over $547 million to various States to create regional HIE (organizations). The fledgling new State HIEs (the organizations) are busy screening and purchasing HIEs (the platforms) and defining the rules of their local HIE (the act). There are several HIE (platform) vendors, notably Medicity and Axolotl (recently acquired by Ingenix), but even Microsoft and IBM are trying to make inroads into this fairly new market. In a parallel process, ONC is busy defining national standards and regulations for HIE (the act).
There are two basic models for any information exchange and HIE (the act) is no different.
The Centralized Model – All information creators/editors/contributors push their content to a centralized repository, preferably in real time, and all users/readers pull the information on demand from said centralized repository. This is the infamous “database in the sky” which houses every American’s medical records. Conceptually, this is the simplest model to understand. The Government will buy enough hardware to set up clusters upon clusters of databases, define the exact data elements and documents to be stored, assign a national identifier to all of us (physicians too) and finally publish specifications for pushing and pulling data securely. Every EHR vendor and medical information supplier (such as labs and pharmacies) will build the necessary web services and integrate them in their technology and we will all live happily ever after. However, other than the obvious monumental technology challenges involved in maintaining such infrastructure, Americans tend to experience significant discomfort with the concept of Uncle Sam having unfettered access to so much personal information and the obvious privacy issues it raises.
The Federated Model – Health information is maintained where it is created, be it physician offices, hospitals or other agencies of care. Contributors push content to users if they want to and readers pull content from contributors when they need to. The best way to think about this is to compare it to telephony. You call me when you need information from me or when you want to share with me information you have, and I do the same. This of course implies that we speak the same language and know each other’s phone number and if we don’t, we have a way to look it up somewhere. It also implies that we both have an agreed upon method of identifying the people we are discussing, e.g. cousin Jamie is my aunt Sarah’s daughter, not your uncle Bernie’s oldest boy. This model requires something akin to White and Yellow pages for providers, as well as unique identifiers for patients, and an agreed upon method to “make phone calls” or push and pull information.
And then, of course, you can have multiple hybrid models, which are somewhat centralized and somewhat federated. The Government funded State HIEs (the platforms) could be regional databases in the sky, or regional telephony networks with associated directories, both serving as federated super nodes on a national network of networks. The advocates of patient controlled medical records would suggest that PHRs should be the basic units of a federated model and nobody gets to gossip about cousin Jamie except cousin Jamie, who at my request will tell you as much as he/she chooses to tell you.
So what are we building out there? We are definitely not building the centralized national database, that’s for sure, but we seem to be engaged in building a little bit of everything else, and we are not starting from scratch.
Pre-HITECH HIE (the act)
Before Meaningful Use became a capitalized term, there where all sorts of HIEs (the acts) going on. Providers with halfway decent practice management systems (most providers) were sending out almost all claims electronically, receiving electronic remittance and checking health plan eligibility for patients in real time. Most accomplished this wondrous HIE (the act) through claim clearinghouses, such as Emdeon, RelayHealth or Gateway EDI, which have built very sophisticated national HIEs (platforms). On the clinical side, anyone with a Surescripts certified ePrescribe module could send prescriptions to most any pharmacy in the nation, receive refill requests and complete medications lists for insured patients, all electronically with a click of a button. Just like the claim clearinghouses, Surescripts created a national HIE (the platform), complete with provider directories and algorithms to identify patients. The national reference laboratories were not far behind and most physicians with an average EHR could send lab orders and receive results electronically from LabCorp and Quest. A host of regional labs were quickly following suit and hospitals were not far behind. The newest trend consisted of electronic clinical summaries in CCR or CCD standard. These were starting to gain some traction as a means of communication with PHRs, such as Google Health and Microsoft HealthVault and here and there files were exchanged between EHRs. It is important to note that clearinghouses charged physicians only a nominal fee (some were free), Surescripts was free to doctors and so were reference lab interfaces.
Enter HIE (the organization)
Although HIE (the act) was occurring, it wasn’t spreading fast enough to meet the 2014 goals, so the Government decided to sow the seeds for organized regional HIE (the act). The first order of business for these State HIE (organizations) is to enable HIE (the act) of prescriptions, structured lab results and patient summaries, which are all requirements for Meaningful Use. Why would they need to do that when physicians are expected to adopt certified EHRs which already have those abilities? For a while now I’ve been puzzled by the National Institute of Standards and Technology (NIST) requirements for ONC certification of EHRs. Unlike the comprehensive CCHIT certification, NIST does not require that a certified EHR be connected to Surescripts and it does not require standard capabilities to interface with national labs. Why not? Because the Government funded State HIEs (organizations) are intended to insert themselves between the physician EHR and Surescripts or LabCorp and later probably claim clearinghouses as well. What used to be a simple point-to-point web service, just turned a bit more complicated, and what used to be free to physicians, will be so no more because HIEs (the organizations) must become sustainable businesses and will charge fees for transactions.
Perhaps it is all worthwhile if these newfangled HIE (organizations) can somehow foster comprehensive national HIE (the act). While State HIEs (organizations) are selecting and deploying various HIE (platforms), using various privacy and security policies and a multitude of directory structures (which may or may not be shared) and various standards of communications, it is expected that all will become federated nodes in the overarching Nationwide Health Information Network (NHIN) which is itself under construction. [I understand that the NHIN is just an incomplete set of specifications, but for some reason, I have this vision of thick, black rubber cables crisscrossing the country.] In the meantime, a more nimble sibling of NHIN, namely NHIN Direct, is also taking shape by proposing to use a completely different and much simpler addressing and transport construct, similar to email, which will allow physicians to send unstructured information to one another, presumably outside the HIE (organization) scope. Concurrently, the historical clearinghouses and Surescripts networks continue to flourish and function as before. In addition to the official State HIE (organizations), there are other private HIE (the act) efforts sprinkled around here and there, which may or may not be inclined to connect to State HIE (the organizations), and some large EHR vendors are starting to run their own proprietary peer-to-peer networks across organizations. Every HIE (platform & organization) is promising “seamless” integration of all these conflicting initiatives.
Long ago and far away, in the miraculous country of Shinar, all of humanity decided to build a tower so tall that it will reach heaven. Recognizing that people united by good communications are sure to succeed, the Almighty, who opposed this particular venture, disrupted work by fragmenting the one accepted standard of communications. Since the StarTrek Universal Translator was not in existence at that time, humanity became confused and the mighty tower was never completed. Thus, to this day, humanity is wondering around this earth perplexed at their inability to build tall enough towers.