Skip to content Skip to navigation

Evidence-Based Order Sets: One Hospital’s Nuanced Experience

September 26, 2012
by Mark Hagland
| Reprints
At Northwest Hospital in Seattle, a chief medical officer shares his perspectives on what has worked in implementing evidence-based physician order sets

Clinical transformation is moving ahead at Northwest Hospital, a 250-bed facility in Seattle. Northwest Hospital, a community hospital, has for the past three years been a part of the University of Washington, under the umbrella of UW Medicine, along with University Medical Center, Harborview Medical Center, and Valley Medical Center (in nearby Renton).

Like its fellow affiliated hospitals within UW Medicine, Northwest Hospital continues to make progress with regard to the adoption of evidence-based medicine tools. The organization has been partnering with the Los Angeles-based Zynx Health to implement evidence-based order sets, and the progress of that work has been accelerated through its implementation of computerized physician order entry (CPOE) earlier this year. Chief medical officer Gregory Schroedl, M.D. spoke recently with HCI Editor-in-Chief Mark Hagland regarding his organization’s work in this area. Below are excerpts from that interview.

I had interviewed you for the HCI January 2011 cover story on evidence-based care. At that time, you hadn’t yet gone live on CPOE. Have you gone live on CPOE since then?

Yes, we went live on CPOE in April 2012, and that was with the Siemens Soarian EMR, and we did use Zynx as the underpinnings for building evidence-based order sets. We went live several units at a time, first in our childbirth and geriatric psychiatric units, and eventually across the enterprise. And it was very smooth and very successful.

Gregory Schroedl, M.D.

Were there any challenges in the CPOE go-live or in the go-live with the Zynx order sets?

One of the challenges that occurred between the time you and I spoke in 2011 and going live was, we were planning to build our order sets in the Zynx author space… They have a software program where you can go into a Zynx site and build your order sets and do your drafts, and it references the evidence, and then when you’ve completed all your order sets, they’re then transferred to your hospital EMR. And we found that we were having difficulty accomplishing that, and that we were going to have to duplicate work to use that methodology. And Zynx was realizing that, depending on the hospital EMR vendor, the ease of that method varied significantly. So that was part of the reason that Zynx developed their new value-based program, because what we were able to do, then, is to build our order sets in the Siemens EMR, and link it to the Zynx evidence; and we’re now arranging for Zynx to come back and do a gap analysis.

So you more or less had to rework that implementation process?

Well, we had used Zynx in the past to build paper-based order sets; our next step was to get those online so that physicians could print them out as PDFs. And if we had done it the way Zynx had originally planned it for us, we would have built the order sets in Zynx and then launched them in the EMR. But we were one of the institutions that turned to them and said, this is too cumbersome, based on the interaction between the Zynx order sets, and our EMR. So we built them in the EMR. And then the Zynx people adjusted their methodology to meet our needs by reviewing our order sets after we had built them.

And right now, we’re in the process of taking 35 of the most common order sets built based on Zynx order sets, and having Zynx do an analysis for gaps in key evidence. And we’ve actually sent them those order sets for analysis. Our next step is to allow them VPN [virtual private network] access to our system, so they can just actively go in, say on a quarterly basis, to help us make modifications. That will allow us to not only allow us to make sure we don’t have any gaps in our initially published order sets, but then if there are any needed changes, such as an FDA black-box warning, we would probably ask them to tell us about the modification, and then we would make it. But they would be able to look at the order sets and make recommendations for changes, on a regular basis.

When implementing evidence-based order sets for CPOE, how does a patient care organization get the best results along with the greatest buy-in from the physicians?

Well, the first phase is to have the physicians acknowledge that there are bodies of knowledge and sets of evidence that will help them with the care of the patients; in some cultures, that’s a little harder to do. You know, medicine has oftentimes been experience-based and authority-based, as opposed to evidence-based. It’s the notion that “I’ve always done it this way, I think it’s right,” as opposed to using careful analysis. We’ve been fortunate that we really began this whole process 14 years ago. But once you have that acknowledgement, which is the first step, the key is to make the process simple and make it easy for physicians to do the right thing.

So if I’m presented with an order set for a patient with pneumonia that already has all the CMS components in it, and the simple elements are pre-checked on the order set, and I’m given choices of appropriate antibiotics that are acceptable, including alternatives to penicillin, for example, and there’s some simple explanation for me as to when I should choose one item over another, then it’ll be faster and easier for me, and I’m more likely to do the right thing each time.