Downstream systems. Pooled data. A firehose of information. Being swamped, flooded, or deluged with data. It’s striking how often information technology is discussed in terms of water metaphors. But a fitting image for the IT challenges looming over laboratories might be something more akin to a wave—a digital one.
That’s because the $19 billion federal program providing incentives to physicians and hospitals who can demonstrate “meaningful use” of electronic health records is entering a sweeping new phase in which laboratory information is a central part. Regulations for stage two of meaningful use were finalized in August.
What does it mean for laboratories? The most important implication of the meaningful use program for laboratories and pathologists, says Walter H. Henricks, MD, medical director of the Center for Pathology Informatics at the Cleveland Clinic, is this: “As more physicians implement EHRs, there will be a dramatic increase in expectation for electronic interfaces between laboratory information systems and the EHRs.”
The meaningful use incentives have spawned hundreds of new electronic health record systems, says Pat Wolfram, vice president of marketing for Ignis Systems, Portland, Ore. Care providers have to order lab tests within the EHR to qualify for meaningful use stage two incentives. “And many of these newer EHRs haven’t built a comprehensive orders module yet that can create clean and complete requisitions for a lab. So expect to see requisitions coming to your lab in a variety of ways until those EHRs have matured that part of their products.”
Under the 2009 American Recovery and Reinvestment Act, in 2010 the HITECH (Health Information Technology for Economic and Clinical Health Act) program started offering physicians and hospitals financial bonuses for moving to electronic health records. Physicians, for example, could receive up to $44,000 each through Medicare and up to $63,750 through Medicaid by shifting their paper records to an electronic format. Stage one, targeted for completion as early as 2011, required data capture and sharing, while stage two will require advanced clinical processes by 2014. Stage three, improved outcomes, will have a deadline to be set later.
Meaningful use (MU) refers to using certified EHR technology to improve health care quality, safety, and efficiency and to reduce health disparities, engage patients and families in their health care, improve care coordination, and improve population and public health, while maintaining privacy and security. Most significant for laboratories are two stage two MU requirements: By 2014, 30 percent of laboratory orders have to be entered using computerized physician order entry, and 55 percent of laboratory results will need to be reported in structured format.
The HITECH program’s first wave was the Centers for Medicare and Medicaid Services’ rule defining meaningful use and establishing incentives for meeting its criteria—the “how” of using an EHR to qualify for incentive payments. Now, with new regulations issued by the CMS and the Office of the National Coordinator for Health Information Technology (ONC), the stage two wave of the meaningful use program is about to sweep over the country’s 5,754 hospitals and 700,000 physicians. The ONC rule establishes certification criteria that EHR technology will need to meet in order to support meaningful use—in other words, “what” an EHR must be able to do, says Dr. Henricks.
The CMS reported in June 2012 that 55,000 physicians have earned MU incentives through Medicare and about 34,000 through Medicaid. But, as they have done with hospitals, the incentives are expected to draw more physicians into the MU program. Between 2010 and 2011, the percentage of hospitals and institutions reporting in the 22nd annual HIMSS Leadership survey that they had a fully operational EHR across the whole organization rose from 22 percent to 27 percent, while those saying they had begun to install an EHR in one facility rose from 32 percent to 34 percent. Those who had “not yet begun to plan” dropped from five percent to two percent. In the two years since the stimulus package was passed, an estimated 30 to 40 percent of physicians have adopted an EHR.
People are pretty responsive to incentives, says Bruce Beckwith, MD, chief of laboratory medicine at North Shore Medical Center in Salem, Mass. “And it’s enough money that physicians are paying attention.” But soon, the stakes will rise even more. For both hospitals and physicians, incentives to adopt MU will be phased out and by 2015 change to penalties for not adopting MU. Without EHRs and participation in meaningful use, individual physicians’ Medicare fee schedule payments will be cut by one percent in 2015, two percent in 2016, three percent in 2017 and after, and hospitals too will face cuts.
The permanent ONC HIT certification program became effective Oct. 4, 2012. Under it, certification of complete EHR systems and EHR modules is enabled, and an LIS may be certified as an EHR module. “With the emergence of EHRs, especially in physician practices, there’s a greater expectation that the lab deliver results through some type of electronic interface,” says Dr. Henricks. “Number one, physicians have made an investment in EHRs and are not going to want to enter lab results themselves or have them on paper. They may also still want to receive paper. But they’re going to expect electronic records. Second, the meaningful use requirements incorporate laboratory data into the EHR.”
Meaningful use stage two is creating a surge of EHR activity, Wolfram says. “In stage one, importing structured lab results to the EHR was only an option for meeting the incentive criteria. Although I don’t know any EHRs that are successful without a good lab results interface, it wasn’t a requirement for stage one. But structured lab results import will be a requirement for stage two, as is the placing of lab orders. So for a doctor to get reimbursed for MU, they’re going to have to bite the bullet and actually place orders now in the EHR.” With the computerized physician order entry requirements of stage two, says Dr. Henricks, the reality is that “the computer screen is now the requisition.”
Physicians will need to have implemented an orders module for stage two, and proved its use by the end of 2014, to get reimbursed, and Wolfram believes a lot of doctors are going to struggle to do so. “Adopting an EHR is a workflow project and it takes a while for a practice’s move from paper to a computer screen to mature and incubate.” Furthermore, “the EHR software developers will be tuning up the orders and results modules, where it usually takes two or three releases to build a decent workflow. Many of the newer EHRs are just starting their first release of orders.” Small one- or two-physician practices will find it especially difficult to evaluate all the EHRs that are out there, Wolfram believes, and they may choose one that drops off the map in the competitive environment.
In his area of Massachusetts, Dr. Beckwith has seen only a small uptick in demand, but he thinks it’s because his system mandated hospital EHRs and EMRs in the office settings already. “Our particular system was a little bit ahead of the curve. They basically said by this date you must install an electronic system—otherwise you’ll be out of our network. And there was minimal resistance. There were people who decided they didn’t want to, but it was a small number and from what I heard, typically physicians very close to the end of their practice career or people in a very highly specialized area where they didn’t feel it made sense. In most of the rest of the country, a lot of practices are now trying to implement EMRs for the first time in order to get the meaningful use incentive payments.”
But it’s still not a plug-and-play environment, Dr. Beckwith says. “Your version of an EMR is not the same as the office practice’s down the road, even if you’re running the same software. You may have set it up in a different way, you may have an in-office lab, you may have set up your test catalog differently, so there are lots of variables there. Maybe as a laboratory you are dealing with a receiving system that dealt with a different lab, and they use 50-gram glucose challenge instead of 75-gram or they use four time points instead of three time points in a glucose tolerance test. They’re not totally cookie cutter, so you may have to customize the result stream you send to another practice.”
Synchronizing test catalogs can also be a challenge, he adds. “If you brought in a new test in your lab that practices aren’t used to seeing, do you have a system to let them know, ‘Hey, we started doing serum porcelain levels; you might start receiving them now, and you have to build something on your end, a test definition, to accept those’? Or, for example, if you need to change requirements for a test, such as when your reference lab decided it would only accept plasma separator tubes from now on, and no serum separator tubes.”
That different systems may be relying on the same communication standard, typically HL7, does not solve the problem, Dr. Beckwith notes, because HL7 is flexible and there are multiple ways to implement it. “It really depends on how complicated the system is, whether these particular systems interfaced before, do the people who are working on it know the ‘gotchas’ of it, and where the subtleties are.”
“Practices expect that lab-to-EHR interfaces are similar to plumbing, and you just plug it in and results will flow from the lab to the EHR,” Wolfram says. “But the differing HL7 dialects and differing test codes cause each EHR-to-lab connection to have unique properties.”
He sees his company Ignis Systems, which offers a universal physician-centric connectivity product called EMR-Link, as a facilitator of EHR-LIS integration for physician practices. “Just moving from a paper system to an EHR is a huge cultural shift,” he says, and the laboratory is just one component of it, sometimes an underestimated component. “Laboratory integration is frequently the least well analyzed part of an EHR purchase.”
“It will be an EHR world eventually,” Wolfram says, “and orders and results workflows will start and end in that EHR. Standalone lab ordering and results systems that are not part of the EHR workflow will slowly lose ground to the EHRs. Because the doctor is going to say, ‘I bought an EHR, that’s my clinical cockpit, and that’s where all my patient care and my practice workflow is centered. Don’t make me or my staff log in to another system.’”
There are mechanisms to help ensure that systems are speaking a common language. LOINC (Logical Observation Identifier Names and Codes) is one of the standards the federal government has endorsed and it is embedded in the ONC EHR standards and certification criteria, Dr. Henricks says. It aims to serve as a “universal translator” between systems. LOINC has more than 30,000 codes for labs, but LOINC mapping in the LIS can be complex and is not easily implemented. “It also turns into a challenge for several reasons. Number one, it’s not easy to map the LOINC codes to the current lab database because there’s not a one-to-one mapping between the test dictionaries. What we’re seeing labs do sometimes is go to third parties to ‘LOINC’ their database. But LISs vary widely in their capabilities, and some don’t even have a place in their system to put a LOINC code.”
In addition, Dr. Henricks says, “Not every site may map the same test to the same LOINC code, and unless they do that, it’s not a universal translator.” Over time, solutions may evolve, but for now, “we have to overcome the inclination to think, ‘Oh, we’ll just use LOINC and the problem is solved.’” Another problem, he points out: Many existing LIS interfaces are on the older HL7 version, 2.3.1, and will require conversion to v 2.5.1.
Aseparate issue is how laboratory information is managed in the EHR department. “We now report our tests through the EHR,” says Dr. Henricks, “so those results will be presented on the screen according to the screen ‘real estate’ and formatting of the EHR, and that can present issues in terms of how different elements of reports are presented. Are all the elements there? The reference ranges, the layouts, where the text wraps, how you see comments formatting, and the display of lab results are determined by the EHR, but the lab has a role to play in calling attention to issues making displays appropriate for patient care.
“It’s kind of a new role for EHR systems as the landscape is shifting, and it becomes a role for someone in the lab. That’s also true on the order side,” Dr. Henricks says.
Despite advances in network technology, a secure electronic connection can be a problem, Dr. Beckwith says. In his large health system, the centralized health information security department is overwhelmed with requests for connectivity. “If you put in a request, it might be three to six months before they get to it. They’re very professional and require that there be a certain level of firewalls, that servers have to be patched regularly, and so on. But just recently we were trying to connect our lab system to a middleware program, and they basically gave up. They said there is some problem here we can’t figure out; use another method.” This incident was one reason Dr. Beckwith turned to a laboratory interfacing provider, a company that serves as an interface engine for results.
Differing results displays introduce even more variables, Dr. Beckwith notes. “I have seen examples where the display was either extremely misleading, or information was very hard to see. We want to actually see what the doctor is looking at, so often that’s a matter of their doing screen prints and e-mailing them to us. So if we send a result that’s out of range, how does that appear, and where is the comment? If there is a hemolyzed specimen, is that something they’re going to be able to see, or is it buried two screens down? The tricky part is that we as laboratorians didn’t buy their EMR system, so you may be limited in some situations as to what you can do, but this is not something you can ignore.”
EHR displays are traditionally character-based displays, he notes. “But these days, a lot of labs like complex results with images and charts. These are things that would look good as a PDF, but in HL7 you have to convert everything to text. Tables, for example, are usually not properly reconstructed. You can get around that by actually creating a PDF and sending a link, but if the receiving system is only processing text, you’re out of luck.”
Information passed between multiple EHRs can create additional problems—for instance, if a practice requests “copy to” results. If the specialist and primary care doctor are in different practices, can you handle such a request electronically? Some systems can and others cannot, Dr. Beckwith says. One practice he deals with has an in-office lab information system. “So the information we’d generate to our lab interface provider was passed to an in-office LIS, then passed to an EMR that doctors actually used, and it turned out results weren’t being received correctly. We made a custom workaround, but it was a non-trivial problem to fix.”
One of the Office of the National Coordinator for Health IT programs, called the Laboratory Reporting Initiative, is part of the standards and interoperability group, Dr. Henricks says, and it’s a program laboratories should be aware of. “This is an initiative that has aimed to establish a guideline for what interfaces between the lab and EHR look like technically. And this published guideline is still undergoing testing, but there was a draft finalized in July, and it’s part of the requirements for stage two. It’s called the ‘S & I LRI implementation guide.’”
If laboratories haven’t followed this, it can be a challenge to know all the testing that has to go into the interface, he says. “It’s more than just sending bits and bytes back and forth. It’s making sure the message received was the same one that was sent. When you’re looking at a display, there are challenges, because how can you tell what the display looks like in the receiving system? You need help from the receiving system to do the test.” Laboratory report elements subject to variations in EHRs and interfaces include reference range handling, explanatory comments and footnotes, abnormal result flags, preliminary reporting and updates, corrected result reporting and documentation, unsolicited results, and reflex test orders and results.
Different patient identifiers remain a challenge, Dr. Henricks says. “Even within health systems, hospitals may be using legacy medical numbers which can sometimes assign different identifiers to patients when a specimen hits the lab. An enterprise master patient index that assigns an internal identifier linking a patient to all other identifiers can help, but that’s not always useful to all the downstream systems.”
In theory, a national patient identifier would help, Dr. Henricks says, but he can see two sides to the debate over such an identifier. “On the one hand, it would be a huge boon if we had an identifier for wherever you went. But once it happened, what do you do with all the existing records? Who is going to map those into the new system?” Then, too, many people have legitimate concerns about privacy. “We want interoperability of data and to have patient information we can share anywhere, but the privacy laws have never been stronger, and we want things to be shared only when they should be. So those are two conflicting challenges. It’s kind of an interesting paradox.”
Regarding formatting, he stresses that laboratory results in the EHR are more than just columns of numbers. Anatomic pathology, molecular pathology, microbiology, electrophoresis, and transfusion medicine are a few of the areas that present unique considerations in data display. “We already have some tests that maybe would be better served with using graphics or tables, and not just words on paper, and EHRs may not be able to incorporate those reports. PDFs are an option, but then that becomes a single document within the EHR—if the EHR can even accommodate a PDF of a single document. So that’s one of the challenges as we go forward into molecular and sequencing: How do we best represent that data?”
An additional problem that laboratories and physician offices will have to solve is patient connectivity, Wolfram says. A Department of Health and Human Services proposed rule will give patients electronic access to their lab results. This will be difficult for the lab, he says, creating workflow and HIPAA challenges. “Does the lab have the staffing to facilitate distribution of labs to patients, or to answer the questions that patients will invariably ask? And does the lab staff really know who the person is that’s making the request for the lab results?” One approach is to authenticate and register the patient for secure messaging at the point of draw. “Then there’s the issue with the patient’s interpretation of the results. Our physicians are telling us lab result communiques to patients need to be physician guided; it really should come from the practice, not the lab. There hasn’t been an easy product for that, and EHR vendors need to gear up to make that more efficient.”
Amid the many problems of execution, there have been charges that the EHR was overhyped as a way to bring major savings to the health care industry. A Sept. 17 Wall Street Journal article titled “A major glitch for digitized health care records” suggests the touted savings have proved chimerical. But, says Wolfram, “There are a lot of reasons why EHRs don’t achieve their promise, depending on how they are implemented at the practice, what kind of training is given, and so on. Ultimately, if it’s a good EHR and focuses on the workflow of the practice and on ease of use, then you can curb costs.” That can also occur through better proactive monitoring of critical patient lab results, or through such measures as an EHR querying a repository whenever a physician orders a test, to alert them if another physician recently ordered the same test.
“There are studies showing EHRs do a lot of good and there are studies showing unintended consequences,” Dr. Henricks says. Targeted aspects of EHRs, such as preventing medication-related mistakes, have shown big patient safety benefits. “But there are hidden costs of implementing these EHR systems that, in my opinion, are tremendously underestimated. A lot of health care data are unstructured, and it takes quite a bit of effort to be able to work with other systems in an enterprise. There are a lot of hidden costs, despite the efficiencies gained.”
“Having said that, we can’t run a health care system on paper. This idea that it’s going to save a massive amount of money, though, is open to question,” Dr. Henricks says.
What can be done to improve LIS-EHR integration? Dr. Henricks would order up a couple of reforms. “One, I would ensure somehow that laboratory domain expertise is inserted in the EHR design process, both locally and at the policy-making level, so that laboratory information is handled correctly and optimally within the EHR. Whether that could be done more effectively through more requirements for certification or in developing more or less standard system displays of lab results, I don’t know.”
“And I think there needs to be some recognition that this has become part of the laboratory’s activities that needs some compensation or reimbursement. It’s not as though there’s a fee code for EHR informatics. So maybe there are innovative ways that labs could be compensated for managing some of the most important information in the patient health record.”
Dr. Beckwith agrees that the startup costs of installing an EHR are significant. “However, from the point of view of a lab,” he says, “the requests for interfaces are not going to stop, unless the meaningful use program is abandoned, which I don’t think is likely. Keeping track of lab results is a very beneficial function of EMRs, and once doctors have electronic access to lab results in their office systems, I don’t think they will be likely to go back to paper.”
He believes it’s important for labs to know if they’ve seen an increase in demand for interfaces. “That will probably increase as we move on to stage two of meaningful use. Certainly there are solutions and tools out there to help, but you have to realize this is a process that takes time and should be done correctly, both technically and in ongoing validation.” Laboratory directors should try to take as much control as they can of how results are reported to the doctor, he says. “It’s really within your purview, and it’s important to view this as like any other important project. These are the laboratory results your clients depend on, so it’s really critical. And this is how people look at the lab.”
Anne Paxton is a writer in Seattle.