Ten-year Vision from ONC for Health IT Brings in Data Gradually

Posted on August 25, 2014 I Written By

Andy Oram is an editor at O'Reilly Media, a highly respected book publisher and technology information provider. An employee of the company since 1992, Andy currently specializes in open source, software engineering, and health IT, but his editorial output has ranged from a legal guide covering intellectual property to a graphic novel about teenage hackers. His articles have appeared often on EMR & EHR and other blogs in the health IT space. Andy also writes often for O'Reilly's Radar site (http://oreilly.com/) and other publications on policy issues related to the Internet and on trends affecting technical innovation and its effects on society. Print publications where his work has appeared include The Economist, Communications of the ACM, Copyright World, the Journal of Information Technology & Politics, Vanguardia Dossier, and Internet Law and Business. Conferences where he has presented talks include O'Reilly's Open Source Convention, FISL (Brazil), FOSDEM, and DebConf.

This is the summer of reformulation for national U.S. health efforts. In June, the Office of the National Coordinator (ONC) released its 10-year vision for achieving interoperability. The S&I Framework, a cooperative body set up by ONC, recently announced work on the vision’s goals and set up a comment forum. A phone call by the Health IT Standards Committeem (HITSC) on August 20, 2014 also took up the vision statement.

It’s no news to readers of this blog that interoperability is central to delivering better health care, both for individual patients who move from one facility to another and for institutions trying to accumulate the data that can reduce costs and improve treatment. But the state of data exchange among providers, as reported at these meetings, is pretty abysmal. Despite notable advances such as Blue Button and the Direct Project, only a minority of transitions are accompanied by electronic documents.

One can’t entirely blame the technology, because many providers report having data exchange available but using it on only a fraction of their patients. But an intensive study of representative documents generated by EHRs show that they make an uphill climb into a struggle for Everest. A Congressional request for ideas to improve health care has turned up similar complaints about inadequate databases and data exchange.

This is also a critical turning point for government efforts at health reform. The money appropriated by Congress for Meaningful Use is time-limited, and it’s hard to tell how the ONC and CMS can keep up their reform efforts without that considerable bribe to providers. (On the HITSC call, Beth Israel CIO John Halamka advised the callers to think about moving beyond Meaningful Use.) The ONC also has a new National Coordinator, who has announced a major reorganization and “streamlining” of its offices.

So what does the 10-year vision offer? It’s well worth reading, consisting of just 13 pages in fairly everyday language. I summarize the three stages as:

  • 3-year goal (2017): get health records to exchange data

  • 6-year goal (2020): bring patient-generated data into the health care system

  • 10-year goal (2024): use analytics

A few other themes run through the document, such as the ever-critical protection of patient privacy, but the three goals I just listed strike me as the three peaks we are striving to capture.

The 3-year goal reinforces the emphasis that the ONC has placed on interoperability for at least five years, since David Blumenthal served as National Coordinator and Stage 1 of Meaningful Use was designed. Because results of the efforts at interoperability and data exchange are so frustrating, it’s entirely reasonable for the ONC to concentrate on these basics for the first three years. But leaving aside patient-generated data and analytics temporarily also leads to risks and unrealized potential.

For years, manufacturers of popular fitness devices and other sites have created multiple, incompatible storage formats for data that is either collected from fitness and mobile devices or generated manually by patients. FDA-approved medical devices are also a Tower of Babyl. Unsurprisingly, computer companies such as Apple have finally moved in to innovate with platforms that the health care industry has failed to provide. But of course, Apple HealthKit, Google Fit, and Microsoft HealthVault are also incompatible, although HealthVault recognizes some industry standards and I expect the others to make similar steps toward standardization under pressure.

Interoperability should be a goal here at least as much as in current EHRs. All health reformers stress that personal behavior change is the key to controlling the chronic illnesses that take up most health care professionals’ time. In a few years, the real action will be in the patients’ daily lives, not in the sanitized clinics where EHRs are ensconced.

The ONC may not be ready yet to lead, but it could be playing a parental role in the current rush to satisfy Quantified Selfers and bring more patients into the quest to collect personal data. We should advise the platform developers to learn from the disturbing history of EHRs and work together before fragmentation becomes yet again the only real standard. Furthermore, leaders in health care have a perspective that goes beyond the focus of the computer vendors on consumerization: it is up to the health care, for instance, to ensure that standards encompass the medical devices used for treatment and monitoring, not just fitness devices.

How about analytics? These are among the thorniest technical issues in health care. Data quality varies so widely that it’s hard to trust the results of data crunching. Public health data sets are pretty clean and have been used succesfully in planning for quite some time–for instance, in this app for predicting health trends–but clinicians are wildly inconsistent in the types of data they record and where they record it (or whether it’s in a structured format at all). The quality CMS data seems to lie somewhere in between, depending on the data set. Statistical algorithms can handle a certain amount of outliers and untrustworthy samples, but they aren’t magic.

Unfortunately, we can’t wait 6 to 10 years for analytics to start being developed. Accountable care requires predictable, measurable results. Some ACOs are making headway, probably by picking off low-hanging fruit. (It doesn’t require much technology to phone patients a few days after a hospital discharge to ask whether they’re following their treatment plan, and the call can do wonders for their recovery.) But they aren’t using data effectively yet, which they eventually must do in order to make effective use of their resources.

The ONC paper has great goals. Their vision of a system that “allows patients to be active partners in their health and care” (p. 2) and spreads data throughout the system are right on target. But health care practice–whether patient self-monitoring or attempts to institute performance-based reimbursements–are running far ahead of the technical capabilities needed to support them. I don’t want to be making this same complaint 10 years from now.