Research Shows that Problems with Health Information Exchange Resist Cures (Part 1 of 2)

Given that Office of the National Coordinator for Health Information Technology (ONC) received 564 million dollars in the 2009 HITECH act to promote health information exchange, one has to give them credit for carrying out a thorough evaluation of progress in that area. The results? You don’t want to know.

There are certainly glass-full as well as glass-empty indications in the 98-page report that the ONC just released. But I feel that failure dominated. Basically, there has been a lot of relative growth in the use of HIE, but the starting point was so low that huge swaths of the industry remain untouched by HIE.

Furthermore, usage is enormously skewed:

In Q2 2012, for example, three states (Indiana, Colorado, and New York) accounted for over 85 percent of total directed transactions; in Q4 2013, five states (Michigan, Colorado, Indiana, New York, Michigan, and Vermont) accounted for over 85 percent of the total. Similarly, in Q2 a single state (Indiana) accounted for over 65 percent of total directed transactions; in Q4 2013, four states (California, Indiana, Texas, and New York) accounted for over 65 percent of the total. (p. 42)

This is a pretty empty glass, with the glass-full aspect being that if some states managed to achieve large numbers of participation, we should be able to do it everywhere. But we haven’t done it yet.

Why health information exchange is crucial

As readers know, health costs are eating up more and more of our income (in the US as well as elsewhere, thanks to aging populations and increasing chronic disease). Furthermore, any attempt to stem the problem requires coordinated care and long-term thinking. But the news in these areas has been disappointing as well. For instance:

  • Patient centered medical homes (PCMH) are not leading to better outcomes. One reason may be the limited use of health information exchange, because the success of treating a person in his own habitat depends on careful coordination.

  • Accountable Care Organizations are losing money and failing to attract new participants. A cynical series of articles explores their disappointing results. I suspect that two problems account for this: first, they have not made good use of health information exchange, and second, risk sharing is minimal and not extensive enough to cause a thoroughgoing change to long-term care.

  • Insurers are suffering too, because they have signed up enormous numbers of sick patients under the Affordable Care Act. The superficial adoption of fee-for-value and the failure of clinicians to achieve improvements in long-term outcomes are bankrupting the payers and pushing costs more and more onto ordinary consumers.

With these dire thoughts in mind, let’s turn to HIE.

HIE challenges and results

The rest of this article summarizes the information I find most salient in the ONC report, along with some research presented in a recent webinar by the Agency for Healthcare Research and Quality (AHRQ) on this timely topic. (The webinar itself hasn’t been put online yet.)

The ONC report covers the years 2011-2014, so possibly something momentous has happened over the past year to change the pattern. But I suspect that substantial progress will have to wait for widespread implementation of FHIR, which is too new to appear in the report.

You can read the report and parse the statistics until you get a headache, but I will cite just one more passage about the rate of HIE adoption in order to draw a broad conclusion.

As of 2015, the desire for actionable data, focus on MU 2 priorities, and exchange related to delivery system reform is in evidence. Care summary exchange rates facilitated through HIOs are high—for example, care record summaries (89%); discharge summaries (78%); and ambulatory clinical summaries (67%). Exchange rates are also high for test results (89%), ADT alerts (69%), and inpatient medication lists (68%). (p. 34)

What I find notable in the previous quote is that all the things where HIE use improved were things that clinicians have always done anyway. There is nothing new about sending out discharge summaries or reporting test results. (Nobody would take a test if the results weren’t reported–although I found it amusing to receive an email message recently from my PCP telling me to log into their portal to see results, and to find nothing on the portal but “See notes.” The notes, you might have guessed, were not on the portal.)

One hopes that using HIE instead of faxes and phone calls will lower costs and lead to faster action on urgent conditions. But a true leap in care will happen only when HIE is used for close team coordination and patient reporting–things that don’t happen routinely now. One sentence in the report hints at this: “Providers exchanged information, but they did not necessarily use it to support clinical decision-making.” (p. 77) One wonders what good the exchange is.

In the AHRQ webinar, experts from the Oregon Health & Science University reported results of a large literature review, including:

  • HIE reduces the use lab and radiology tests, as well emergency department use. This should lead to improved outcomes as well as lower costs, although the literature couldn’t confirm that.

  • Disappointingly, there was little evidence that hospital admissions were reduced, or that medication adherence improved.

  • Two studies claimed that HIE was “associated with improved quality of care” (a very vague endorsement).

In the next section of this article, I’ll return to the ONC report for some clues as to the reasons HIE isn’t working well.

About the author

Andy Oram

Andy Oram

Andy Oram writes and edits documents about many aspects of computing, ranging in size from blog postings to full-length books. Topics cover a wide range of computer technologies: data science and machine learning, programming languages, Web performance, Internet of Things, databases, free and open source software, and more. My editorial output at O'Reilly Media included the first books ever published commercially in the United States on Linux, the 2001 title Peer-to-Peer (frequently cited in connection with those technologies), and the 2007 title Beautiful Code. He is a regular correspondent on health IT and health policy for He also contributes to other publications about policy issues related to the Internet and about 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.