Skip to content Skip to navigation

Bridges to Somewhere New

July 26, 2011
| Reprints
Anticipating some of the revelations to come out of the emerging data superstructures

I found an article in the July 13 Journal of the American Medical Association to be fascinating, as it validates what many experts have been saying for some time now, and that is that, when it comes to the large data and information superstructures that are being built now in healthcare, there is great potential going forward for improvements to patient safety and care quality that sometimes can’t be fully appreciated at the outset of their development.

As the article by Tracy Hampton, Ph.D. notes, “Up to one million patients in the United States may be taking two medications that can lead to unexpected increases in blood glucose levels when used simultaneously.” Dr. Hampton reports that “Data mining techniques have revealed that the combination of the antidepressant paroxetine and the cholesterol-lowering medication pravastation may cause this adverse effect,” citing an article in the May 25 edition of Clinical Pharmacology & Therapeutics, that noted this finding.
Furthermore, the author writes, “’If a physician has a patient on these two medications, and their diabetes becomes harder to control, the physician may want to consider changing the medications,’ said principal investigator Russ Altman, M.D., Ph.D., professor of bioengineering, genetics, and medicine at Stanford University.”

What’s significant here is that this medical finding emerged out of data mining work. What happened in this case is that Nicolas Tatonetti, a doctoral candidate in the biomedical informatics training program at Stanford, and his colleagues, mined the Food and Drug Administration’s Adverse Event Reporting System (AERS) for side-effect profiles involving glucose homeostatis, and “found a strong signal for co-medication with pravastatin and paroxetine.” Using data they could cull from the EHRs of three geographically distinct sites, they retrospectively evaluated changes in blood glucose in 104 patients with diabetes and 135 patients without diabetes who had been on both medications, and assessed the mean random blood glucose levels before and after treatment with those drugs.

The implications of this discovery are quite remarkable, really. Think about it: as patient care organizations are collaborating across states and regions; as the Nationwide Health Information Network begins to seriously emerge as a reality in the next few years; as public health reporting systems evolve forward; as outcomes measurement and pay-for-performance collaboratives move towards maturity; and as such concepts as accountable care organizations, the patient-centered medical home, and others, blossom under federal healthcare reform incentives and mandates, connections are going to be made that involve “greater than the sum of their parts” information superstructures.

Inevitably, more discoveries will be made like this one, discoveries that will improve patient safety and care quality for patients, their families, communities, and society. It’s like the synapses in the brain of a growing child constantly making new connections. And, through the collaboration of healthcare IT leaders, clinicians and clinician leaders, clinical researchers, public health authorities, health alliances and associations, professional societies, and every other kind of individual and entity, we will make connections that hitherto have been impossible.

So when resisters to clinical transformation and other processes grumble, we will be able to point to small breakthroughs like this one, and say, justifiably, Just Wait. Because the whole we’re collectively reaching to really will be more than the sum of these various parts. And bridges will be built that will make everyone pause and say, gee, isn’t it amazing we didn’t know those things before?