From The Editor | July 18, 2012

Have Best-Of-Breed EMRs Lost The Hospital Battle?

Ken Congdon, Editor In Chief of Health IT Outcomes

By Ken Congdon, editor in chief, Health IT Outcomes
Follow Me On Twitter @KenOnHIT

Is a best-of-breed or enterprise/single vendor EMR implementation the best approach for your hospital? This topic has been the focus of much debate ever since EMR software came into existence, and both sides have compelling arguments. At a high-level, a best-of-breed approach to EMR implementation typically offers a greater level of functionality for specific specialties or departments. However, integrating disparate best-of-breed EMR systems within a facility can be a challenging and expensive undertaking. An enterprise (or single vendor approach), on the other hand, offers complete data integration out of the box (which is also expensive). However, an enterprise EMR often forces hospitals to sacrifice functionality in specific departmental applications.

Prior to Meaningful Use, I’d argue that most hospitals in the U.S. that had adopted some form of EMR opted (sometimes unknowingly) for a best-of-breed approach. At the very least, best-of-breed EMRs ran neck-and-neck with enterprise deployments. I say this because, more often than not, EMRs were installed to address specific departmental or facility needs prior to 2009. There wasn’t a nationwide push at the time to establish a universal electronic health record. If advantages could be realized by installing an EMR solution in the ER or cardiology wing, a system was installed in those departments and not in others, often without much regard to whether or not the two systems could “talk to one another.”

The Trend Toward Enterprise EMRs

Enter 2009. The HITECH (Health Information Technology for Economic and Clinical Health) Act was passed as part of ARRA (The American Recovery and Reinvestment Act), providing a whole slew of financial incentives for healthcare providers to adopt EMRs. Then, in 2011, Stage One Meaningful Use requirements were announced, basically providing healthcare providers with a roadmap on how to use EMRs in a “meaningful” way so that hospitals and other healthcare facilities could receive their incentive payments. Creating a universal electronic health record that provides a comprehensive view of the patient data from multiple systems and facilities is a focal point of Meaningful Use. All of the sudden, integration became much more important.

Integration standards and languages, such as HL7, have been developed to make it possible to integrate health data from best-of-breed EMRs. However, more and more, it seems like hospitals are abandoning this approach in favor of enterprise options. I began to form this perception last summer when I attended the NG Healthcare Summit in Scottsdale, AZ. More than half of the hospitals I spoke with at the event were in the process of ripping out their existing best-of-breed EMR systems and replacing them with an enterprise system — most were opting for Epic.

This trend has only seemed to gain traction based on my phone conversations with the technology leaders of hospitals and health systems over the past year. For example, I recently spoke with Bill Weyrick, senior manager of information systems at Dartmouth-Hitchcock Medical Center. During the call, Weyrick provided me with a detailed overview of his organization’s flash cutover to 15 Epic EMR modules. Prior to installing Epic, Dartmouth-Hitchcock’s EMR efforts were largely focused around a homegrown system it developed called CIS (Clinical Information System).

“We had 35 developers on staff to maintain and support our homegrown EMR,” says Weyrick. “However, to get our homegrown system to a place where it would meet Meaningful Use criteria would require us to add a great deal of functionality including barcode administration and CPOE (computerized physician order entry) system wide. We would have had to hire hundreds of additional developers to pull this off. Therefore, we opted to go with a commercially available EMR software package.”

In a separate conversation, Michael Mistretta, CIO of MedCentral Health System provided a different driver for his organization’s decision to deploy Siemens Soarian EMR solution enterprise-wide.  “Vendor management was a key consideration in our decision to use a single vendor approach to EMR implementation,” says Mistretta. “With a single vendor, I only have one finger to point at. It simplifies my environment because I don’t have Siemens telling me it’s McKesson’s problem and vice versa. Also, the built-in interoperability is key. There is a trade-off in the fact that the system does not provide prime functionality to certain departments or specialties within our health system, but at this point in time, it’s much more beneficial for our organization to have the ability to share data across the continuum of care quickly and easily.”

What’s This Mean For Best-Of-Breed?

The previous examples are but a few of the dozens of conversations I’ve had over the past year that seem to show a clear trend toward enterprise EMR adoption in hospitals. What does this trend mean for the future of best-of-breed EMRs in the hospital market? That depends. Not all conversations I had with hospital executives over the past year opted for an enterprise approach. Some hospitals, especially small to midsize community hospitals, opted for the best-of-breed approach. Interestingly enough, cost was a major factor in these decisions as well. Whereas most large health systems argued that the cost to integrate disparate systems was too overwhelming, community hospitals said the cost to rip out their existing EMR investments and replace them with an expensive enterprise solution was too prohibitive. Many community hospitals were determined to “make their existing investments work.”

Furthermore, many of the health systems I interviewed that said they opted for an enterprise approach, didn’t take a true single vendor approach. Instead, they relied on a single vendor for 80-90% of their EHR applications, but used other vendors for a handful of departmental applications. The most common example of this was the use of one vendor for most or all inpatient EMR applications, and other vendors for ambulatory EMR applications. Examples also existed where specialty departments, such as the ER, had a separate specialty EMR, while the rest of the hospital leveraged an EMR by a single vendor. I typically refer to this as a “hybrid” approach.

In addition, in other industries, best-of-breed versus enterprise implementation approaches tend to be cyclical. For a couple of years, one approach seems to have the advantage, but a few years later the other approach sees a resurgence of activity. The same could hold true for hospital EMRs. For example, many of the hospitals or health systems may come to realize that the functionality they are sacrificing in some of their departments is coming at too high a cost and look to replace enterprise modules with specialty EMRs targeted for specific applications. Time will tell.

Lastly, I’m not naive enough to think that the conversations I had over the past year are truly representative of the hospital market as a whole. There are approximately 6,000 hospitals in the U.S. I have only had conversations with a few dozen. If you have different perspectives or opinions to share, please comment on this article, or email them to me at