Government incentives are underscoring the push for a national transition to health information exchanges. Our experts weigh in on some of the issues surrounding the implementation of HIEs.
With all the talk about electronic medical records and meaningful use, it's easy to put health information exchanges (HIEs) on the back burner. But that's not an option. Communicating in a digital healthcare world without HIEs is like trying to navigate a highway system with no freeways.
Karen Friedrich, vice president of strategic alliances for Carefx, says, “HIEs are designed to provide a complete end-to-end view of patient information for clinicians [who] are caring for these patients and making decisions affecting their health.”
In a perfect world, HIEs would facilitate the sharing of data between doctors' offices, hospitals and laboratories; everyone would share records, and security levels would be determined by the level of data needing to be exchanged. These free-flowing, robust sharing systems would allow proprietary LIS, PACS, EHRs — and more — from different vendors to communicate with one another.
But if there's one thing upon which we can all agree: This world is far from perfect. Motivated by a governmental push, HIEs are happening, and it's up to the vendors and end-users to figure out how to best facilitate the process.
According to the Department of Health and Human Services, $548 million was granted under the State HIE Cooperative Agreement Program, disbursed through cooperative agreements and awarded on a rolling basis in February and March 2010. $386 million was given to 40 states, territories and SDEs on February 12, 2010. The remaining funds were allocated to 16 states, territories and SDEs on March 15, 2010. The cooperative agreements are for four years.
For our July issue, Health Management Technology reached out to several experts, vendors and end-users to get their take on some of the major issues involved with HIE implementation.
The true value of HIEs
It's easy to get caught up in all the data-centric techno-speak, but, says Pam Arlotto, president and CEO of Maestro Strategies and author of the book, “Beyond Return on Investment: Expanding the Value of HIT,” it's important to keep in mind the reason for transitioning to HIEs in the first place: “The value will come not from the technology but from the use of the information to improve outcomes, design new care-delivery models and reduce the cost of patient care.
“As the 'information cloud' becomes more pervasive during stage two and stage three of meaningful-use deployment, applications will be created to sit on the 'cloud,' extracting information content and context based on the specific needs of the end-user — and innovation will flourish,” Arlotto relates. “Both technology vendors and healthcare providers will have to reinvent themselves to stay relevant.”
So, although it doesn't make all the logistical issues any easier to deal with, it may help to focus on the eventual goal of HIEs: a more efficient, streamlined healthcare system with improved outcomes and better patient care.
To translate, or not to translate: That is the question
As anyone even tangentially involved with the implementation or operation of proprietary IT solutions knows, interoperability between systems has become a huge issue. But will that issue continue to worsen, leading to an even larger number of “translators” on the HIE horizon?
Elliot Menschik, MD, PhD, general manager of provider markets for MEDecision, says no: “EHR/EMR and other clinical applications are currently on a rapid path to enable their users to achieve federal meaningful use (MU) standards, which we expect will provide a baseline of native interoperability around the continuity of care document (CCD). Furthermore, nearly all major healthcare IT vendors have made investments in the IHE/HITSP approach to cross-enterprise exchange (e.g. XDS and related integration profiles), which are increasingly reaching the field and enabling plug-and-play interoperability,” he says. “Today, proxies/translators are an unfortunate reality, but they will give way to native interoperability.”
Earl Jones, vice president and general manager of GE Healthcare's eHealth Solutions, doesn't think an HIE will require a large number of translator products. “A spoke-and-hub model based on a document registry and repository (or multiple repositories) concept is the most sustainable and efficient; this results in one platform managing connectivity and translation,” he says. “Our system uses a powerful integration platform as part of our implementation to connect systems with and without standards support. It includes a growing library of adapters that can be used with products that are not yet standards-compliant, to accelerate implementation. It can also be used to rapidly develop new adapters for proprietary or custom systems as needed.”
Translators will not proliferate, agrees Chuck Christian, director of information systems and CIO at Good Samaritan Hospital in Vincennes, Ind., agrees. “The obstacles to data sharing will be overcome (as they have been by the few successful growing HIEs),” he says. “For the foreseeable future, improved application of existing standards, interface engines and knowledgeable humans will be the keys.”
Nonetheless, it is a foregone conclusion that some major vendors will be reluctant to make it easy for the data stored in their systems to be shared with other systems. According to Hamid Tabatabaie, president and CEO of lifeIMAGE, “This is a self-preservation strategy of sorts. However, the standards are getting more mature; just as there were and are dozens of HL7 interface engine vendors today, I foresee there to be many exchange gateway providers whose systems will be tuned, to various degrees, to deal with a broad range of vendor-specific installations.”
Communicating between systems
To a large extent, the existence of translators depends on the federal certification process, says Joel Vengco, director and chief applications officer, Boston Medical Center (BMC). “If certification can address the EHR market and broaden to other areas of clinical information types soon, then this translator period may be short. If certification doesn't take hold immediately, then middleware companies will be in high demand for a bit longer,” he says. “I would predict that most of the major vendors are going to look at more open architecture and services and thus enable communication through the HIE over the next several years. My organization is using translators in the meantime while our vendors catch up.”
Mary Anne Leach, vice president and CIO, The Children's Hospital, Aurora, Col., says, “Absent widely adopted clinical content and data standards, translators (or ETL-type functions) may be required in the interim to facilitate data exchange, data aggregation, trending/analytics, and to support the most effective use of the system by clinician end-users.”
According to William A. Spooner, senior vice president and CIO at San Diego's Sharp Healthcare, “Numerous vendors are developing translator or gateway toolsets, based on the emerging standards (CCD/CDA/XDS).”
Daniel Morreale, CIO and president, Infoshare, AtlantiCare Health System, Atlantic City, agrees that translators, in relation to normalizing codes, will be a significant endeavor.
Says John Reifenberger, vice president of RHIO development, Axolotl: “Translators may start to emerge in the market; however, part of the service provided by Axolotl is to enable proprietary systems to communicate with Elysium Exchange. We have 15 years of experience doing this, and so far, though we have created templates and acquired knowledge that simplifies the process, each proprietary system is still somewhat unique. Until proprietary systems begin utilizing standards, a universal translator, a product that requires minimal manual configuration, seems unlikely.”
Organizations, such as ISO and IHE, are leading the way, relates Mary Kasal, executive director, Franciscan Health System (an Axolotl customer). “The great increase of other standards bodies has now developed the base knowledge to allow the development of extensible translation tools,” she says. “I believe this is fundamental infrastructure that, as of yet, lacks an organizational platform to support the use of the translations. I see that area as the next natural place for development. Metaphor is the banking industry's conversion from paper to electronic over a 20-year span.”
Michelle Blackmer, senior director of healthcare marketing for Initiate (an IBM company), sees integration as a primary objective of HIE; her product has established adapters for most EMR/EHR systems, and she expects to do the same for PACS and LIS. “We also see document sharing and exchange playing an increasingly [important role],” she relates.
ROI and patient benefits
Many IT systems are infamous for their soft return on investment. But Vengco has some hard facts to share regarding HIEs: Since implementing the Carefx portal solution in May 2009, BMC has improved the rate of scheduled referrals from 30 to 60 percent, reduced no-shows from 30 percent to less than 5 percent, and increased total number of incoming referrals by 100 a day (about 10 percent). When fully functional, Vengco estimates the portal system will generate an estimated additional $6.5 million each year for the client.
Reifenberger says that “ROI largely will depend on the customer's situation and environment. Much of the ROI might be affected by the decisions made on the topics in the first question,” he says. “Typically, tens of thousands are saved immediately in reduction of the administrative and material costs of paper records exchange. Patients benefit from a reduction in tests, easier care coordination, less time spent filling out forms and easy availability of their medical records.”
Kasal estimates her annual ROI is approximately $2 million — in paper handling alone.
According to Tabatabaie, redundant radiological exams account for $15-20 billion annually. “It will probably cost somewhere around $1 billion per year to operate a nation-wide system for access to prior imaging exams everywhere,” he says. “The ROI is pretty impressive.”
Focusing back on the big picture, Arlotto reminds us about the transformational nature of HIEs: “Goals for care coordination across disparate providers and clinical integration between hospitals and doctors cannot happen without HIE,” she says. “Quality measurement and improvement across provider entities is enabled through HIE. Ideally, patients will see improved efficiency and elimination of redundant provider services; they will have an improved patient experience; their providers should have more knowledge about their past health issues and problems; and they should enjoy more access to their own health information.”
We asked our experts the following question: What are the most important factors for an end-user to examine when considering the implementation of an HIE?
Reifenberger: “Governance. Are all the stakeholders in agreement on how data will be shared, security models, consent models, etc. Sustainability. Is there a business model, or are those deriving the most value footing the bill? Adoption. Successful HIE is dependent on widespread adoption of HIE services by the hospitals and physicians. Either a significant portion of the users should be interested in HIE in the region or a solid plan to drive adoption should be in place.”
Arlotto: “Business and clinical goals for the HIE and key stakeholders; flow of information from a patient perspective; how decisions will be made.”
Blackmer: How they will allow physicians to participate in information sharing without requiring them to give up ownership of their data; resources for physician outreach and marketing activities; and which portions they, as the sponsoring organization, will pay for vs. those that they will ask participants to pay for.”
Menschik: “How to use the HIE to enhance their business; specifically, to build patient/referrer affinity, build brand/differentiate and expand geographical reach; what will maximize success, scalability and sustainability: procuring traditional software and staffing up the organization to operate the HIE, or moving to a Software-as-a-Service model? Interoperability. Will the solution simplify vendor-neutral exchange among diverse systems, or does it depend on proprietary connections to 'like' systems?”
Tabatabaie: “Much of the success of an HIE depends on getting various healthcare organizations to agree on a comprehensive set of shared policies and principles that will help govern and operate an HIE for an expanded period of time. Conflicting or misaligned interests are probably the most important factor of an HIE's success. Will hospital A not want to share the oncology data on a patient because they may lose the patient to the newly formed oncology center of hospital B?
“Are there project-based approaches that can help facilities incrementally adopt and HIE? Boiling the ocean is not the way to go for an HIE. Starting with focused, achievable objectives and incrementally growing the number of use cases is most manageable. For instance, lifeIMAGE advocates an imaging HIE. A group of hospitals and their community physicians can all participate in sharing imaging results.
“Technology is often not the barrier, but it is expensive to acquire the capabilities of a master patient index and sharing information from diverse origins and formats. Therefore, there is such a thing as buying too much HIE which leads to inability to support the minimum requirements for the long term. That's why the lifeIMAGE solution is transaction based and allows facilities to pay for what they use, when they use it.”
Friedrich: “Anytime, Anywhere. Clinicians have to be able to access data from anywhere at anytime. Data needs to be viewed easily, with only Internet access and a browser required, versus only on a workstation where a piece of software has to be installed. The flexibility to look at information anywhere a clinician may need it provides value not only to the clinician but to the patient as well.
“Real-Time Data. Physicians need to be able to look at a patient summary and have access to information that is as current and up to date as possible. If a patient was in the Emergency Department and was told to follow up with his or her physician the following day, the corresponding clinician needs updated information to be able to provide the best care possible. If clinicians are working with data that is less than time sensitive because it is only uploaded once a week, for example, it provides much less of a value proposition for that patient's immediate care.
“What can the clinician end-users see? Not only is the type of data (medications, labs, radiology images, etc.) important to the clinical users, but the facilities from which they can see it is vital as well. Many HIEs are looking to implement a statewide system, but the likelihood of a patient who lives in one part of the state receiving care at a facility hours away are slim. Rather, a user who can see regional data will find what he or she needs quickly and completely, which increases the likelihood that he or she will use the HIE over and over, which, of course, is the goal.”
Jones: “When considering HIE implementation, the end-user should: Clearly define the purpose and value of HIE for the participating organizations. In other words, answer these questions: Why will we implement an HIE, who will use it and profit from it, and what work flows and/or processes will be enhanced? Next, document and get buy-in from the key stakeholders on the purpose and expected value of the HIE. This 'vision' statement can then be used to inform the 'how' (the actual selection and implementation of a solution) and to evaluate the success of the HIE once installed.
“Confront the issues of governance. This involves more than control and management of the HIE. It also centers on issues of trust. Are the participating organizations and/or stakeholders willing to share patient information? What barriers exist to data exchange and how can they be removed? Once the shared data is identified, determine whether its content and quality will be rich enough to bring value to the end-users.
“Identify how the HIE will deliver new capabilities to the targeted end-users and whether these will be valued by them. Will the users embrace these features, believing that they enhance the clinical work flow? Two factors influence end-user adoption: intuitive, ease of use and the delivery of filtered, highly-relevant information. The best technology in the world cannot make up for awkward, hard-to-use functionality that is disconnected from a deep understanding of the clinical workflows. Engage the end-users to ensure that they are committed to the changes required. Based on their feedback, it's possible to create a multi-staged implementation plan so that each stage brings value to the targeted user within a short period of time.”
Vengco: “Make sure to identify what data is the most useful to provide from the HIE out of the gate. While lab results may be fine as a first step, for our institution, our clinicians stated that labs alone weren't useful enough to make a trip to the HIE. So we added meds, problems, and allergies. That made it stickier. Think about the presentation of the data before you present any data. The CCD document structure can make for a wieldy work flow. Who's going to pay for it? [Other concerns] won't matter if the grant runs dry and there's no funding.”
Kasal: “Sustainability plan for support (ongoing operations) and for the needed expansions that always surface in technology tools in healthcare. Solid understanding of expected customers of HIE — specifically knowledge of possible competing HIE initiatives, culture of community towards sharing of data, knowledge of technology platforms in place in planned region. Knowledge of infrastructure requirements of HIE technology and readiness of customer base for use of HIE.”