For years now we have been talking about HIE as if health information exchange (the verb) is happening across health information exchanges (the noun). This is wrong for the simple reason that true, full health information exchange is not occurring.
Then what is being exchanged across these HIE networks that public and private entities are spending a fair amount of money to make happen?
Typically, it’s fairly limited clinical data-sets (lab results, discharge summaries, referral documentation) delivered as an HL7 message, ala CCD, or more recently CDA – data sets that rarely embed themselves into clinical workflow where they are most likely to be used. This is not to say that the HIEs that are being stood up today do not provide some level of value – certainly they do and numerous studies point to decrease in duplicative tests, identifying frequent flyers to ED, etc. What I am saying is that the healthcare industry is in a rapid state of transformation and the future need to transmit an ever wider range of health-relevant information across a distributed network will be critical to the success of a healthcare organization.
Some examples:
Advocate Health in Chicago has been working closely with Cerner on the development of “Smart Registries.” In speaking with their CMIO last fall I asked him if he was using their HIE to assist with the data aggregation and information distribution needs that are required to power Smart Registries (their instance taps over 50 data sources). He said no, the HIE simply does not/cannot provide sufficient data nor can it deliver the information they want to present to a physician practice.
In a briefing with one of the largest population health analytics vendors, Optum, who also happened to have acquired an HIE vendor (Axolotl) a few years back, I asked if they were using any of their HIE capabilities to aggregate data to power the new Optum One analytics platform? They said no, the HIE does not deliver rich enough data to do the type of analysis and ultimately deliver the value that Optum One offers.
Last week I met with a vendor CEO to discuss our latest research on CNM. He pointed out that they are seeing a huge need for HCOs to meld clinical and transactional data (claims, PBM, prior auths, etc.) to drive better care coordination across a community. They are now working with a payer and HIE vendor to enable such capabilities to better support ambulatory practices within that payer’s provider network. The HIE alone was not enough.
The currently limited capabilities of many an HIE will not serve the needs of tomorrow. This belief led us to our research on Clinician Network Management (CNM), which is a visionary look forward on how healthcare organizations need to rethink their dated, and now misnomer of an “HIE strategy” a strategy that has very little to do with the need for a more complete information payload delivered at the point-of-care to enable population health management.
As we found in our CNM research, exceedingly few organizations have come to this realization, though we are seeing glimmers of hope with those on the leading (bleeding) edge such as the aforementioned Advocate. Will others follow? Yes, of course as the transition from fee-for-service to value-based reimbursement will demand it. The real question is when and when they do, will they have the leadership to make it happen? This is a journey not for the feint of heart.
The problem with the HIE approach is that it leaves the current institution-centric data silos intact, and tries to connect them via various ties. A step in the right direction, but an inadequate longer-term strategy. In my view, actual patient-centered, cross-institutional *aggregated* data is the center of the next level of health IT. More thoughts here: http://blog.flowhealth.com/aggregating-clinical-data-beyond-interoperability/
Robert, thanks for the comment and link to your post. While philosophically I agree with your overall thesis, a patient-led data aggregation platform/repository with APIs to allow other systems to tap (eg EHRs, PHRs, etc.), this is something that is still quite a few years (10+) off into the future. The majority of patients aren’t ready for it, HCOs will be very reluctant for fear of losing customers and of course EHR vendors have no incentive to move in this direction as it commoditizes their product. There is also the issue of data integrity and normalization – not sure your typical consumer is ready to take that on. And then with have that minor annoyance of creating a master patient index/personal identifier.
No simple answers here in the short to medium term horizon.
Is the failure to launch a surprise? I often hear the HIE model rationale compared to the SABRE system for travel. But SABRE had a business case for its users – make booking tickets easier and faster so airlines could make more money. While the predominant business cases in health care are still “fill a bed” or “fill an appointment”, HIEs yet to match those. What’s worse is that some providers are hesitant to join HIEs for risk of LOSING patients to other participants.
You hit the nail on the head, at least for public HIEs, which have very few meaningful use cases that would make a provider want to use one. Private, enterprise HIEs are a different story. Many a healthcare organization has adopted such with the primary objective to drive referral flow (lab/radiology orders) back to the mothership, typically a hospital. But even here, the market is rapidly changing as to what is needed for the future – continuing to label it as HIE is no longer enough – it needs to extend far beyond that limited definition.