This Just In on EMRs: Workflow Really Matters | Mark Hagland | Healthcare Blogs Skip to content Skip to navigation

This Just In on EMRs: Workflow Really Matters

December 20, 2010
| Reprints
UC Davis study shows productivity depends on the level of customization

A new study carried out by management and computer science researchers at the University of California-Davis is confirming what thought-leaders in healthcare IT have long known on an instinctive level: that the degree to which EMR implementation can help improve physician productivity depends on the level of customization around specific types of workflow demands and working requirements of different medical specialties.

Hemant Bhargava, associate dean and professor of management and computer science at the UC Davis Graduate School of Management, and his colleagues, studied a multi-million-dollar IT project installed at six primary care offices from 2003 to 2006. The offices were part of a large primary care physician network affiliated with an academic medical center. Bhargava and his fellow researchers found, based on data collected from 100 physicians at six clinics, and spread across internal medicine, pediatrics, and family practice, that the initial EMR implementation resulted in a 25 to 33 percent drop in physician productivity. However, recovery of physician productivity following implementation. On the one hand, “internal medicine units adjusted to the new technology and experienced a slight increase in productivity. In contrast,” the study’s authors reported, “pediatricians and family practice doctors did not return to their original productivity levels and experienced a slightly lower productivity rate.”
Bhargava’s conclusion? “These differences by unit suggest that there is a mismatch between technology design and the workflow requirements and health administration expectations for individual care units.”

Now, here’s the really interesting part. Bhargava and his colleagues note that there are two different categories of activity when it comes to interacting with an EMR—information review and information entry. Information review includes patient history, notes from previous visits, charts of test data and radiological images; and the USE of an EMR tends to make such tasks more efficient. Further, those features are useful to internists, who tend to see a greater proportion of ill patients. In contrast, pediatricians’ work tends to involve more information entry and documentation, both of which are more time-consuming tasks.
I think more studies like this are needed, as they will continue to provide new insights into some of the specific and particular challenges facing doctors implementing EMRs. The fact that this study uncovered key differences between and among primary care specialties underscores how complex this whole enterprise of EMR implementation really is at the physician office level, and why broad generalizations about doctors’ responses to EMRs tend to be so superficial. I certainly hope consultants working in the physician-office implementation sphere are paying attention.



Remids me of the piece you published (and I wrote) last year
Looks like were both 'thought leaders'. See: COPE and the Doc Dilemma.

Frank Poggio
The Kelzon Group

Mark, this is why throwing money in an effort to rapidly automate a complex safety-critical industry like health care is so scary. The goal here should not be to implement EMRs, but to effectively automate workflow. And anyone who's done it knows that effective workflow automation is a slow and careful process this study confirms that.

Many IT folks are guilty of a similar problem, medical device connectivity. This connectivity is the automation of workflow through the integration of medical device and information systems, yet many focus on the connection rather than the workflow to their detriment.

I'd love to see a study similar to this one, that deals with medical device connectivity. There are many potential subjects: EMR documentation, smart infusion pumps, alarm notification...

Great article amd point! I have always contended that although most people don't seem to make a distinction, I see there is a significant difference between work process and work flow. I (we - my company) believe before any EMR is considered for implementation a thorough base line analysis / assessment and workflow study is completed. Then proposed work process (how it is actually done - handwritten versus keyboard or Dragon) changes are discussed.

The "workflow" (the order in which process (steps) are completed) should not change unless deemed it is for the better.

One example, w/o mentioning any names, is my physician. Used the big name EMR that the hospital "suggested / assisted with" and it takes too much time to use during a visit and is hard to discuss lab results. . . looking at LT computer screen. And what if a patient wanted a copy of the report - printing was not convenient? Proper workflow mapping, work process analysis / HR assessments would have identified critical success factors outside just getting the EMR working and training the professionals to use it.