March 09, 2012|Categories: All Things EMR
The buzz this week has been all about a new article in Health Affairs showing an association between electronic access to diagnostic radiology and laboratory studies and increased diagnostic test utilization.
The buzz was loud enough that Farzad Mostashari, MD, ScM, the National Coordinator for Health IT, felt compelled to set “some facts straight.”
The study, based on retrospective review of data from the National Ambulatory Medical Care Survey, would seem to fly in the face of conventional wisdom that electronic medical records (EMR) should reduce costs by providing doctors access to prior test results, thus reducing unnecessary and duplicative testing.
But it doesn’t.
It’s easy to misunderstand the study. However, it was not about EMR use, but rather the provision of electronic results and the images themselves. To quote the investigators: “other indicators of office computerization—computerized physician order entry and the use of an electronic medical records system—showed no association with test ordering.”
What the study did show is that among a host of potentially cofounding variables, the volume of test ordering is higher for practices and their practitioners if they got their results in an electronic form.
Why the paradox? Unfortunately, the study lacked the ability to actually slice and dice the data in a truly meaningful way.
- First, it is not clear that they controlled for the variability of practice types and locations. Back in 2008, the vast majority of physician’s offices that had already deployed electronic results reporting were large employed groups and academic medical centers. Don’t these docs order more studies than small offices?
- Second, the study doesn’t look at whether the diagnostic testing was appropriate. While the study specifically splits out MRI and CT (modalities that are rarely part of typical quality care gaps) it is difficult to really tell what’s going on without a deeper analysis of the indications for testing.
- Third, the study does not look at whether the differences of care were a result of differences in test ordering, or differences in test compliance. At athenahealth, we are sensitive to the fact that connecting the clinician’s intent to an outcome (i.e., the patient actually getting the test) is critical, because our real-time data show that a high percentage of all physician orders go unfilled, meaning a doctor can write for a completely appropriate test, and that patient, for whatever reason, just does not get it.
- Finally, the study provides no insight into the total economic picture. How many of these studies prevented death, disability or more severe complications? You can criticize test ordering, but not unless you know the whole story.
The real problem with the study is that it failed to recognize the effect of plain old e-commerce on the medical supply chain. They almost make the point in their discussion section:
“Perhaps physicians who order more imaging studies, for whatever reason, are more likely to acquire health information technology that facilitates the retrieval of imaging results or images.”
Say I’m a doctor who orders a lot of studies, a neurologist, an orthopedist, or a geriatrician who specializes in cognitive decline (think MRI). My fax machine is jammed up with results, so I ask that imaging center or lab for an interface, or just a way to log in online, rather than jam up my fax machine.
At athenahealth, this is what we do every day – take high volume transactions and make them electronic. It’s why we don’t charge for interfaces.
It’s what everyone was doing in 2008 – taking advantage of information technology to make their lives easier. And, so here we have a handful of folks trying to influence national policy on the basis of four-year-old data.
Believe what you want about the nation’s health IT strategy, but this study adds little to the dialogue.