Health information exchanges (HIEs) are an appealing way of harnessing technology to improve care. Too often, doctors must make medical decisions based on incomplete clinical information. Electronic health records are part of the solution, but their full value won’t be achieved until they’re connected to provide a cross-institutional, comprehensive, timely and accessible view of a patient’s medical history.
Public and private HIEs are creating the connections and clearinghouses to make this possible. However, while everyone recognizes that better information can lead to better care, financing HIEs has proved to be very challenging. One hope has been that HIEs could pay off quickly in early cost savings. It stands to reason that doctors wouldn’t need to order as many tests if they knew the results of previous ones, and early studies led to very promising projections of costly tests averted.
Unfortunately, the intuitive link between HIEs and early cost reductions may not be so clear. For example, I recently co-authored a study of Quality Health Network (QHN) in Grand Junction, a leading HIE here in Colorado (http://jamia.bmj.com/content/early/2013/05/21/amiajnl-2012-001608.abstract). QHN has been a remarkably successful, nationally recognized model, providing robust information exchange since 2005 and achieving adoption by 90% of the practices in its community. Our study focused on laboratory and radiology testing by ambulatory providers over a five-year period, with the expectation that HIE adoption by providers would be associated with a decrease in the rate of testing. However, while HIE adoption was associated with a modest reduction in the number of laboratory tests ordered per encounter, it did not appear to affect rates of radiology testing, and no cost savings could be demonstrated for either type of test.
In general, while these results do not match the optimistic econometric projections of a decade ago, they are consistent with more recent results from other analyses of HIE in various settings: the value proposition for HIEs delivering early cost savings from reductions in testing isn’t as strong as we had thought.
These results are important, but I’m still convinced of the value of HIE. HIE isn’t just about, or even primarily about, more efficient test ordering. It’s about better care. I had the privilege of conducting an earlier qualitative study involving QHN, and I’ve been struck by the genuine enthusiasm for QHN by the Grand Junction community. Part of this reflects QHN’s excellent leadership, but provider enthusiasm wouldn’t be sustained if it wasn’t making a real difference in patient care. I think as our use and understanding of health information technology improves, these observations will increasingly be backed by evidence of the downstream benefits of better informed decisions, better coordinated transitions of care, and reductions in error.
As I’ve moved from informatics investigation to informatics development here at Health Language, I remain optimistic about the prospects of health information technology. Knowledge representation is a critical and challenging element of health information technologies, and it is gratifying to be engaged in work that I continue to believe will give providers and patients the tools they need for more effective care.