I recently attended the Massachusetts Health Data Consortium’s (MHDC) conference on Health Information Exchange (HIE), modestly titled “The Key to Integration and Accountability.” Although I’m a health IT geek, I felt I needed help understanding life outside the electronic health record (EHR) world. So, I roped in Char Kasprzak, statistical data analyst at Massachusetts Health Quality Partners, to give me a better picture of the quality implications of HIE (and to help me write this post).
John Halamka, CIO of Caregroup/Beth Israel Deaconess Medical Center, took the stage first and blasted through all the progress being made establishing the necessary frameworks for HIE to occur in Massachusetts. The takeaway message from John’s talk was that there have been many changes since September 2011 in the financial, technical, and legal structures involved in building health information exchange. The lessons learned from the initial pilot should enable Massachusetts to be ready for the first stage of statewide HIE.
HIE development in Massachusetts
Health care providers historically thought of HIE as a large institution run by a state or a major EHR vendor. It carried out the exchange of patient records in the crudest and most heavyweight way, by setting up one-to-one relationships with local hospitals and storing the records. (Some of the more sophisticated HIEs could link together hospitals instead, rather like Napster linked together end-users for file exchange.) These institutions still dominate, but HIE is now being used in a much broader sense, referring to the ability of institutions to share data with each other and even with patients over a variety of channels.
Despite the push for the health IT industry to use “HIE” as a verb rather than a noun, there was quite a lot of discussion at the event surrounding the structures and applications involved. Although HIE should be conceptually identified as a process (verb), having the structures and organizations (nouns) necessary to facilitate exchange is a challenge facing health care entities across the country. This conference did a good job of articulating these organizational challenges, and it presented clear plans on how Massachusetts is addressing them.
In Massachusetts, the model moving forward for phase one of HIE will be based on the Direct Project, with one central Health Information Service Provider (HISP) that will focus on PKI and S/MIME certificate management, maintaining a provider/entity directory, creating a web portal for those not ready for Direct, and maintaining an audit log of transactions. The concept of HISP was created in the Direct Project Implementation and Best Practices workgroups, and was designed to be an organizational and functional framework for the management of directed exchange between health care providers. The statewide HISP will consist of several existing HISP organizations, including Berkshire Health, Partners, Athena Health, and the New England Health Exchange Network. No small task, but not insurmountable.
I remain skeptical about the ability of providers and even hospitals to install EHRs capable of sending Direct-compliant messages conforming to the XDR/XDM IHE Profile for Direct Messaging. Not that it doesn’t work or because it’s some Herculean task, but essentially because it hasn’t been mandated. That may change, though, with the inclusion of Direct Messaging in the transport standards for Meaningful Use Stage 2. In Massachusetts, the creation of a health information highway (phase 1) is set to go live on October 15, 2012. Phase 2 will include analytics and population health, and Phase 3 is set to have search and retrieve, which will include a governance model for an Electronic Master Patient Index (EMPI) and Record Locator Service (RLS). Phase 2 and 3 will set a framework for querying patient data across entities, which is one of the biggest technical barriers to HIE. Currently, one of the best methods for this process is the Patient Identifier Cross-Referencing (PIX) profile, but few organizations are using this tool to its full potential.
What are the challenges?
When experts talk about exchanging health information, they tend to focus on the technology. Micky Tripathi, CEO and executive director of the Massachusetts eHealth Collaborative, pointed out at the event that the problem isn’t the aggregation or analysis of data, but the recording of data during the documentation process. In my experience, this is quite accurate: Having exchange standards and the ability to analyze big data is useless if you don’t capture the data in the first place, or capture it in a non-standard way. This was highlighted when the Massachusetts eHealth Collaborative ran the same reports on 44 quality measures, first using popHealth data, then again with Massachusetts eHealth Collaborative data, and received conflicting results for each measure. There are certainly lessons to be learned from this pilot about the importance of specifying numerators, denominators, vocabularies, and transmission templates.
Determining what to capture can be as important as how the data is captured. Natasha Khouri elaborated on the challenges of accurate data capture during her presentation on “Implementing Race and Ethnicity Data Collection in Massachusetts Hospitals — Not as Easy as It Sounds.” In 2006, Massachusetts added three new fields and 33 categories to more accurately record race and ethnicity information. The purpose of this is to address health disparities, which is something I’m very excited to see discussed at a health IT conference.
With accurate data in hand, direct interventions in communities can be more targeted and effective. However, the largest barrier to this seems to have been getting providers to ask questions about race and ethnicity. This was due to high training costs, staff resistance, and workflow changes necessary for collecting the demographic data. This problem was particularly interesting to me, having worked with the Fenway Health Institute to craft their Meaningful Use Stage 2 comments regarding the inclusion of gender identity and sexual orientation in the demographics criteria. Recording accurate data on vulnerable populations is vital to improving public health campaigns.
What about patients?
For a conference with no patient speakers, there was a surprising amount of discussion about how patients will be involved in HIE and the impact EHRs have on patients. Dr. Lawrence Garber,who serves as the medical informatics director for Reliant Medical Group, examined issues of patient consent. The research he discussed showed that when given the choice, about 5% of patients will opt out of HIE, while 95% will opt in. When patients opt in at the entity/organizational level, this enables automated exchange between providers, entities, care teams, and patients. Organizations utilize a Data Use and Reciprocal Support Agreement (DURSA) to establish a trust framework for authenticating entities that exchange data (presumably for the benefit of patients). DURSAs will likely play an important role as organizations move toward Accountable Care Organization models of care.
Information exchange should also lead to more patient satisfaction with their medical visits, where they will be able to spend more time talking to their doctors about current concerns instead of wasting time reviewing medical history from records that may be incomplete or inaccessible.
Dana Safran, VP of performance measurement and improvement at Blue Cross Blue Shield, explained at the conference that patients can expect better quality of care because quality improvement efforts start with being able to measure processes and outcomes. With HIE, it will be possible to get actual clinical data with which to enhance patient-reported outcome measures (PROMs) and really make them more reliable. Another topic that can be better measured with HIE is provider practice pattern variation. For example, identifying which providers are “outliers” in the number of tests they order, and showing them where they stand compared to their peers, can motivate them to more carefully consider whether each test is needed. Fewer unnecessary tests means cost savings for the whole system, including patients.
Toward the end of the conference, Dr. Nakhle A. Tarazi gave a presentation on his Elliot M. Stone Intern Project on the impact of EHRs on patient experience and satisfaction. The results were quite interesting, including:
- 59% of patients noticed no change in time spent with their provider.
- 65% of patients noticed no change in eye contact with their provider.
- 67% of patients noticed no change in wait time in the office.
The sample size was small, interviewing only 50 patients, but the results certainly warrant a larger, more in-depth study.
In Massachusetts, it seems like the state of the HIE is strong. The next year should be quite exciting. By this time in 2013, we should have a statewide HISP and a web portal service that enables exchange between providers. Halamka has promised that on October 15 the walls between Massachusetts health care orgs will begin to come down. If it is successful in Massachusetts, it could be a valuable model for other states. We also have the opportunity to involve patients in the process, and I hope organizations such as The Society for Participatory Medicine and Direct Trust will be involved in making patients active partners in the exchange of health data.
OSCON 2012 Healthcare Track — The conjunction of open source and open data with health technology promises to improve creaking infrastructure and give greater control and engagement to patients. Learn more at OSCON 2012, being held July 16-20 in Portland, Oregon.
Related: