MH: The results that you wrote about in your report on topics such as optimizing user support, improving the ongoing management of order sets and clinician decision support, and so on, all seem very commonsensical.
ED: Yes, and that was the point. This was intended to provide practical advice for people considering implementing CPOE. We had done a previous report looking at preparations. So this really was intended for people doing it; it’s not policy-based or theoretical. This is about how it gets done in community hospitals.
MH: In general, one might make the assumption that implementing CPOE in a community hospital would be more difficult. Did your study tend to confirm that presumption?
ED: Yes. In the community hospital setting, you’ve got fewer resources to make this happen, and no residents to enter the orders.
JM: And no standing committees. We should point out that three of our hospitals in our six were part of a bigger health system. And nationally, many community hospitals are part of a system. And we tried carefully in the report where it made a difference, to point that out. One of the differences when a system-based hospital is involved is that some of the governance activities, such as building order sets, were accomplished at a system level, whereas at the true standalones, those activities were of course local. In any case, it’s important to keep in mind that if a multi-hospital health system is involved, some of the topics we discussed in the report get handled in a different way.
ED: Of course, the amount of support that individual hospitals get from their system organizations varies tremendously; some systems are systems in name only. Certainly, things like coverage from the help desk is a huge benefit. Community hospitals have an advantage in that doctors are more accepting of AMA standards of practice. And if you’re part of a system, that means that system-wide sets of protocols are more feasible in a community hospital, as are outside order sets. Because they might want to review them, but community hospital docs are more likely to accept such outside order sets and protocols. They have fewer resources but also in some ways, fewer barriers.
JM: And in terms of help desk, in our sample, the community hospitals that were part of systems still had very local mechanisms for user support. So it doesn’t mean you don’t have that same responsibility; it may be backed up by corporate support, but still…
MH: In other words, the more “system-y” a system is, the better?
ED: Yes, and it’s primarily with regard to resources; it means you don’t have to do it all alone.
MH: Were you surprised at the levels of challenges?
ED: It was about what we had expected. They also didn’t talk about it as though it was some monumental achievement, either. There used to be this theory that you can’t do it in community hospitals, and that doesn’t seem to be proven out by the experience of these hospitals.
JM: They just put their effort into it and made it happen.
ED: And there has to be a champion, but it doesn’t have to be the CEO, as some say. It could be a doctor or a nurse. They all had a doc leader, and the level of involvement of the CEO varied from passionate to nominal.
MH: What about the role of CIOs?
ED: Clearly in these places, the CIO and lead clinician had joint responsibilities, with the CIO more responsible for the technology stuff. The CIOs generally played a back role, enabling things to happen.
JM: Who were the people doing the heavy lifting, the folks who actually managed the application? Well, there’s a bit partnership at the staff level between the IT department and some who were not, between IT and the clinical departments.
MH: What’s your advice for CIOs?
ED: One, don’t start without a clinical leader who really wants this done. Two, don’t underestimate the amount of support that’s going to be required to get the physicians on board and keep them on board, and that requires 24/7 support. And three, worry a lot about the system going down, because it will.
JM: And that’s not just technology, it’s processes that alert users when you know, and that fill in during the downtime; it’s not just about a backup system.
MH: Any predictions, as CPOE becomes more universalized?
JM: I think one of Erica’s recommendations will go to the top of the list, and that is around this whole issue of downtime. All the hospitals we studied had an extensive list of practices for dealing with downtime, but I was surprised at how few of them really did have much of a backup system. And it’s not just CPOE that’s involved, of course; people are doing medication reconciliation electronically; nurses are recording vital signs electronically. So this whole issue of, it’s got to be there, is going to be sort of ‘less optional’ than in the past.
ED: And I think some of the people were surprised at how much unplanned downtime they had. There was a lot of variation by vendor product regarding the amount of downtime they had. And I think that needs to be put much higher on organizations’ selection criteria. It’s something objective you can track.
MH: When they had downtime when you were there, what happened?
ED: They went to paper. And they cued the latest info about the patient to print every hour or two, so when it went down, they went over and printed, and kept track on paper of stuff after that, and if it came up quickly, they reentered everything; and if it didn’t come up quickly, they were very selective about what they typed in. But none of them had the capacity to handle situations where they went down for more than 24 hours at a time.
JM: So people who are purchasing should have this much higher on their lists. And they should call real sites.
ED: Because nobody had a real-site, mirrored backup. It’s what you should do ideally. But for a community hospital, it’s better to have a system that has recovery mechanisms within itself and goes down as little as people. I also think that there will be more community sharing of order sets and community support. Many vendors now support that more.
