EMR Implementations: The Good, the Bad, and the Ugly

On May 19, 2010, the GSB's Program in Healthcare Innovation held its final brown bag discussion of the 2009-2010 academic year. Kevin Tabb joined a group of faculty with an interest in healthcare-related topics to talk about SHC's implementation of the Epic electronic medical record (EMR) system.

SHC spent north of $100 million over four years to implement what it called an organizational transformation, which included the implementation of Epic. It was a three-phase process, with inpatient functions (the hospital) going live in mid-2008; ambulatory (clinics) rolling out over 18 months through summer 2009; and all back-office functions coming online in fall 2009. Tabb described the 5,000+ person roll-out as a mid-sized Epic installation.

When SHC decided to adopt an EMR in 2005, it was in a somewhat defensive position relative to its local competitors. The Palo Alto Medical Foundation (PAMF), located less than 2 miles from the Stanford Hospital, began its implementation in 1999 and had a fully functioning EMR in place by 2002. Oakland-based Kaiser Permanente had launched a major Epic implementation in 2003. "Patients were curious as to why we weren't offering them access to their medical records online, why we weren't as advanced as some of our competitors, so there was a recognition that we needed to do something," Tabb said (1). SHC also needed a way to strengthen its ties to the community physicians who referred patients to Stanford's hospital and specialists. "As a complex care center that does less primary care than others, our survival depends on physicians within the community referring patients to us. We needed a way to make it easier for community physicians to refer to us, and that meant providing them with access to information about patients that they refer here and also making the interactions easier. We want to ensure that independent physicians … So we wanted a product that would help extend our reach into the community."

While Tabb admitted that the amount of money SHC spent on the implementation of Epic was "staggering," the organization invested heavily to ensure a high level of physician and staff involvement in designing the system, and enough training and at-the-elbow support during the go-live period to be sure that there would be no detrimental effect on patient care. In fact, more than 75 percent of SHC's implementation budget was spent on project staffing (2). SHC also viewed Epic as just one aspect of a larger organization transformation effort. As a result, it dedicated roughly the first year of the four-year effort to having cross-functional teams of internal staff members and external consultants redesign its processes to achieve greater efficiency and effectiveness in how work is performed.

So, what does SHC have to show for its sizable investment in adopting a comprehensive EMR system? According to Tabb, that still remains to be seen. On a positive note, SHC completed its Epic roll-out on time, within budget, and without any serious "meltdowns." All employees in the hospitals and clinics are using the system, along with SHC's network of community physicians There are fewer silos within the organization as a result of adopting an enterprise-wide system, and SHC now has access to a wealth of data that was never before available. "Five years ago, there were no real metrics being used to measure outcomes and the quality of clinical care…. Now we're in the top quartile, headed to the top decile," Tabb noted.

However, along with these positive outcomes, Tabb was candid about sharing the mistakes, issues, and problems that still need to be addressed. At the top of his list was the fact that only limited workflow changes had been realized to date from its process improvement efforts:

We thought it was really important to take a hard look at all the processes that we had in place. This was our opportunity to come up with ideal ways of working, not simply to replicate our very poor processes when we put in the new systems—because that's just a really expensive copy machine. So we spent a lot of time and money working with consultants on mapping out our processes and coming up with ideas. But, frankly, we didn't achieve more than 25 percent of what we thought we were going to achieve. This was partly because, even with all the maps and the thinking, we did not have a good handle on where our current processes were. We really only discovered where our current processes were when we disrupted everything and tore out the old system and put in the new one.

As a result, processes were still fraught with workarounds and had not become standardized, as SHC had hoped. Using Epic to perform many of these suboptimal processes also took longer Tabb elaborated: "It takes multiple, multiple steps to do anything in the system. So, for the time being at least, it's still faster to scribble something illegible on a piece of paper than it is to get it in the computer. There are lots of good reasons to get it in the computer, but we can't argue with people that we're making them more productive…. It is not faster. It is more cumbersome."

Additionally, while SHC now had access to "infinite" amounts of data from the system, Tabb asserted that the challenge was to transform the data into meaningful, actionable information. "We need to be able to take this data and figure out how to use it to really improve the quality of care. We have to learn how to get good information out of the system to do things better. We're only just starting to crack the nut on that," he said.

Reflecting on SHC's current position and its approach to the EMR implementation, Tabb shared this insight:

One of the mistakes that I think we made is that we budgeted and sold to leadership a huge investment to get to go-live, with very little thinking about what happens afterwards. So, yes, we have a maintenance team. But that's not what I'm talking about. There really needs to be an investment now in optimizing the system and getting the best out of it. But at this point, the organization has other investment priorities. I think we should have and could have prepared ourselves differently for this.

At some level, the adoption of a comprehensive EMR system is no longer optional for a major medical center like Stanford. However, Tabb is the first to agree that the organization must do more than absorb this significant expense as a cost of doing business. While Stanford has traditionally maintained a "product leadership" position within the medical community (i.e., for providing leading-edge, high-quality care), competitors such as Kaiser Permanente are seeking to disrupt the status quo by offering interesting new care paradigms that leverage the EMR as their foundation. In the coming 12 to 18 months, SHC's challenge will be to tune and tweak and really understand its new system so that it can begin to unlock innovative new possibilities for maintaining its leadership status. SHC's awareness of and honesty about the issues and hurdles uncovered through the EMR implementation process is a good first step. But more hard work still lies ahead for SHC to shift from a defensive to offensive position in this arena, and to realize the benefits that other healthcare environments have begun to achieve.

Notes:

(1) From an interview with Kevin Tabb conducted by the authors in spring 2009 for a case study about SHC's Epic implementation (see GSB No. OIT-101, "Building a Business Case for an Electronic Medical Record System at Stanford Hospital and Clinics," October 2010).

(2) Calculated based on other interviews and data provided by SHC for the same case study.