Storage Management and Planning in the EMR Era
Frank ClarkHealth IT professionals have had their hands full in recent months, scrambling to understand how best to meet the provisions of meaningful use in order to capitalize on American Recovery and Reinvestment Act incentives—and, eventually, to avoid CMS payment penalties. One area affected by meaningful use that often goes unconsidered, however, is storage. “With the new meaningful-use requirements, widespread electronic medical record (EMR) adoption is going to generate even further demand on storage,” according to Frank Clark, PhD, vice president for IT and CIO at the Medical University of South Carolina (MUSC) in Charleston. “These changes will require a well–thought-out, well-defined strategy to meet the demands and needs that will be placed on storage.” These new demands, Clark says, come on top of the demands already being placed on health-care organizations’ storage infrastructures by medical images—not only those gleaned from radiology and cardiology, but also those coming from dermatology, ophthalmology, and other specialty areas that are relative newcomers to generating image data. “Surgeons are also wanting to use more images from radiology and cardiology in their operating rooms,” he says. “That’s where the real growth and demand for storage currently is—across the different medical-imaging modalities.” Changing Architectures In response to the growing demand placed on its archive by medical-imaging studies, three or four years ago, MUSC transitioned away from direct attached storage to a SAN. “We thought we could achieve some economies of scale,” Clark explains. “We feel that we’ve reduced our overall storage costs, and this is a more efficient way to store our data.” The new storage architecture has images and related data remaining easily and rapidly accessible on the fiber-channel SAN for three years; for the following four years, the data are stored on solid-state drives, before moving to tape, once they reach the ripe old age of seven years. Clark explains that these decisions were not made arbitrarily; rather, they were the result of an extensive analysis that MUSC conducted on how often data were accessed after they reached a certain age. “We have studies going back 15 or 20 years,” he notes. “We worked with the radiologists to determine that keeping the past three years of studies on the fastest storage would be sufficient. It’s worked out pretty well; we learned that they rarely go back and look at studies that are more than seven years old.” The next step, Clark says, is to prepare for the changes to come with implementation of modern health IT systems across the enterprise. “As you might expect, we run a lot of clinical point-of-care systems, administrative systems, and financial systems, in addition to the EMR,” he says. “We’ve tried to get our user community to begin to look at archiving their data the same way medical images are archived.” The challenge, he says, is getting organization-wide buy-in on the tiered-storage concept, which is tricky because not all users understand that storage is becoming more expensive. “I think users have historically thought that storage was almost free and that they didn’t have to worry about it,” he says. “We have to try to change the culture—to get them to think about what it is they need instantaneously and what they would be willing to wait a few minutes before getting. As we approach petabytes of storage, it begins to add up, so we’ve tried to approach the subject of storage-cost savings from that perspective.” Planning Ahead To anticipate future storage needs, Clark and his colleagues at MUSC make forecasts based both on historical usage and on clinical systems and modalities that will be coming online in the near future, generating fresh demand. “Our storage is close to a petabyte, across the organization, and is growing at a rate of 14% to 15% annually,” he says. “We try to work closely with key areas (like radiology and cardiology) to anticipate their storage needs. We pay close attention to new devices that might be coming in, such as CT and PET systems, so we stay as close to the work and activity as we can. We’ve been pretty good at staying ahead of the demand curve.” Ultimately, MUSC plans to achieve modular certification for meaningful use, which means bringing together flexible information systems from multiple vendors. Clark anticipates that the impending modernization of some of the organization’s older information systems will spur new demand. “We’ve started moving some of the bigger applications through the tiers, but unfortunately, some of the legacy systems don’t have the elegant archiving capabilities that more modern applications have,” he says. “We are replacing some of those systems with those that are more efficient and flexible (from a number of perspectives) and are easier to integrate into our overall IT and storage strategy.” That strategy, he says, might one day extend to emerging storage solutions such as those made possible by cloud computing. He notes that MUSC’s secondary data center is located relatively nearby, and both the primary and secondary data centers are in an area that is vulnerable to hurricanes. “Health care is probably not as quick to buy into the new, innovative ideas, and we’re certainly prioritizing the privacy and security of patient data,” he says. “We were also slow to come around to desktop virtualization, which has proven to be a great resource. Over time, we will look at cloud computing and cloud storage as well. From the disaster-recovery perspective, it’s certainly a possibility.” In the meantime, Clark advises his fellow health-care IT professionals not to underestimate the role that effective storage and archiving should play in an overall IT strategy. “We’re dealing with the historical perception that storage is inexpensive,” he says, “but there’s a huge demand coming from these new modalities and information systems. A well–thought-out and well-defined storage strategy is key to an efficient and effective IT operation.” Cat Vasko is editor of and associate editor of Radiology Business Journal.