DL: When you proposed Epic to the board, did you have to demonstrate any hard ROI?

SJ: When we went to our board in early 2004, after a substantial planning effort, one of the things they were imposing on any capital investment, certainly any one of the size we were talking about, was a return. And as you probably know, the returns on EMRs are sometimes difficult to outline and certainly difficult to capture. But the board was looking for diligence and thoughtful analysis on where we were expecting to get benefits. So we identified 23 or so measures of various kinds that would be impacted as a result of the EMR investment. About 11 or so had real tangible dollars attached, and there were others that we knew were valuable, but we couldn't attach a number to.

Nevertheless we had expectations, for example, that nursing satisfaction would improve and that would have an impact on retention and turnover. We see nurses coming out of school expecting to use an EMR and we think that will grow over time, but we can't measure that. But on the tangible side, we expected to save real money in forms, paper and printing costs. We forecast on a $62 million investment, $9.2 million projected per year return. That was for 1,500 beds in five hospitals. I don't want to call it a bargain, but it's less than you might typically spend and it's what we felt was believable. If we were careful, we could get it done for that, and that was what we thought we could afford. And we thought that with that investment, we'd have a return of a little less than $10 million a year. And though that resulted in a negative return, we would be on the hook to at least try and capture that savings, be it through case mix or anything else. We continue to look at those costs. In the case of transcription costs, for example, it's fallen way down and we've saved a dramatic number in labor.

DL: Are you finding other cost savings you hadn't thought of?

SJ: For one thing, given the stimulus package, we had no idea that this would qualify us for federal funding. We believe, as the rules exist today, that we could qualify. It is predicated on HIE, and frankly, that infrastructure isn't in place. We participate in a local RHIO, but if you think about the government's intent, with everybody plugging into a backbone, we're exchange-ready. We have a certified system and we exchange data now with the local RHIO, so we're ready for 2011 and what other rules they may impose. We do CPOE, reporting out of the system - it's meaningful use by any stretch of the imagination, so we're pretty sure we'll qualify. That's not savings per se; that's a benefit that will probably come in one door as the Medicare reimbursement rate goes out the other. But it's better than losing by not having anything.

I think the whole point is there has been the tendency to say that EMRs are a cost of doing business, there really is no return. And so unless somebody really forces you down the path to look for value, you won't look for value. You'll worry about getting it implemented and not necessarily look for savings the way we had to. We had to come back and talk about our promise to the board.

DL: Your board sounds pretty unique.

SJ: Well, I think they wanted management to be disciplined, and we took it there. We believed we were doing this for a reason, and we should be able to get a return. We wrote in the budget expected savings. And that put some pressure on the CFO and local management teams to go back and take a look at where we expected to get value and look for other opportunities. It's just a matter of the thought process. We have not gotten all the value we thought we were going to get, and we've gotten some of it in unexpected areas.

DL: Such as?

SJ: The paper savings were more than expected, almost $3 million. And we don't have transcribing errors anymore because we do very little transcription. We are a community hospital and these are physicians that are not our employees. So it's a lot about enticing, encouraging them, and having them lead a lot of these efforts. Because, as you know, they can vote with their feet.

DL: Are you using Stark to put product in their offices?

SJ: We are. It's a relatively slow uptake process. We have that for probably slightly more than 40 physicians. One of the clinics is a children's clinic, and the bulk of our admissions to our children's hospital come via the clinic. The clinicians there are just so wowed by their ability to feed the inpatient stay, and for that matter, as patients make their way from the clinic for whatever reason, the clinicians on the inpatient side are able to get a complete sense of that patient. So we have that continuity of care.

DL: Well that's the goal.

SJ: Absolutely, but not only that, there's an independent medical group that happens to have Epic. They had it in place before us and they do most of their admissions with us. There are about 140 physicians associated with it. So even though they're an independent company, they're using the same system, the Epic Care Everywhere exchange component that's inherent to the Epic software. When those patients come from the ambulatory environment into our facility, the physician in our facility, be it a hospitalist or anyone else, can go back into the Epic record, pull it forward and see Epic's PatientOne talk to Care Everywhere. It's unbelievably cool and is exactly what the government is intending for health information exchange - this free flowing information, complete picture, continuity of care. It makes so much sense for the patient and for the clinician. And in this case, it's across organizational boundaries.

DL: Are you live in all of your five hospitals?

SJ: The last hospital we expect to go live with is this fall; we're bringing in the revenue cycle. At first it was just the clinical app, the ED, critical care and physician order entry, nursing documentation, and pharmacy - the basic bulk of clinical applications. But in our next hospital we're bringing all of that plus revenue cycle, so it will be seamless.

DL: What are you doing in the other hospitals for revenue cycle?

SJ: We're using Keane (San Francisco) and we're going to replace that. We already replaced TDS that we had inherited, but that was an ancient mainframe-based system. We'll replace the Keane at our next hospital and then the others, so we'll be all Epic.

DL: Any reason you picked this hospital to be the first for the financial app?

SJ: In this case, the hospital has our smallest AR. We want to make sure we can appropriately test in a live environment this new revenue cycle app. It poses a certain amount of risk and AR days may go up, and we want to contain that and get that perfected. Now that we're live at our hospitals with Epic, we'll take just the revenue app down to our very first clinical facility and bring it up, then come back to our two very large facilities, and then we'll be done with Epic on both the clinical and financial parts of the system.

DL: So the Epic RCM that you're going to do was not part of the original deal?

DL: Speaking of changes, will the current economic climate impact your IT strategy?

SJ: I can't speak for others. We're glad we started when we did, given how long it takes and considering that you need to be live in 2011 for the funds to flow and not later than 2013 to receive the full credit. So it's none too soon to start. I think with little capital, the subsidies that come later in time don't help you now. And people are trying to figure that out. A way to avoid a lot of capital expense can be software as a service, it's less up front. But people are going to move forward because it is going to be required. Unless you're willing to tolerate reductions in reimbursement by 2015 and beyond, then you better figure out a way to have meaningful use. And that doesn't mean you have to buy Epic - there's and other cheaper solutions, and it comes down to fit. So you can get this done for reasonable sums of money. My worry is that there will be a rush to put these things in and it's not just buying them - there are five or six critical success factors like cultural, exchange managements, physician leadership.

Apart from picking a vendor whose system works, which we think we did, we took the classic people process and technology triangle. When the technology works, what's left is the stuff you've never seen, the people and process. It's an ongoing effort to refine the application, pay attention to every single physician that works in very different ways, and to try and build an application under the covers. That's where the action is, that's where the value comes from, that's what takes so long, and that's what will make it work in the end. You overlook two or three of these things and you're going to have fires.