Disconnected, part 2: New interoperability approaches offer promise and challenges

Although the interoperability of electronic health records (EHRs) currently is very limited (see “Disconnected, part 1”), several moves are afoot to improve it. Some of these initiatives are well underway, while another one is in an early, experimental phase. They range from new combinations of existing techniques to a novel approach that applies widely used Internet standards to healthcare.

Most physicians are unlikely to see the fruits of these labors for some time, however. But CommonWell plans to expand to 5,000 sites by the end of 2015, and major developments in all of these projects can be expected in the next few years. Here’s a brief rundown on each of them, including how they might help you in the future.

CommonWell Health Alliance

CommonWell Health Alliance is a collaboration of EHR vendors that represent about 70% of the hospital market and 24% of the ambulatory care market. It is implementing an interoperability network that includes a common patient matching method, a nationwide record locator, and privacy and security safeguards. Launched early in 2013, CommonWell started testing at a few sites later that year.

Today, CommonWell includes the biggest EHR vendors in the acute-care market except for Epic, which is also the largest ambulatory care EHR vendor. As a result of Epic’s refusal to join CommonWell, the alliance represents less than a quarter of the ambulatory care EHR market.

Related:EHR's broken promise

CommonWell now has 29 members and expects to have 40 by year’s end, according to Jitin Asnaani, MBA, the alliance’s executive director. So far, only 73 provider sites have gone live on CommonWell, but rollouts by some big EHR vendors to their customer bases are expected to begin this fall. The 5,000-site goal includes practices and hospitals that will have the ability to connect with CommonWell but will choose not to initially, Asnaani admits. 

Three vendors-Cerner, McKesson, and athenahealth-have said they’ll make CommonWell available to their clients for free. A Cerner spokesman said the company would offer the service gratis for five years after a customer signs up for it “with the exception of a one-time setup fee.” 

It’s unclear whether the CommonWell vendors will charge usage or subscription fees later if the service catches on, Asnaani says. But for now, they’re paying fees to support the organization, and they’re bearing the cost of interfaces to RelayHealth, the McKesson subsidiary that provides the backbone for CommonWell. Cloud-based EHR vendors need write only one interface to RelayHealth; some suppliers of premise-based EHRs are connecting all their customers to a cloud-based solution that they link to RelayHealth through a single interface, Asnaani pointed out.

RelayHealth supplies peer-to-peer connectivity among participating sites, regardless of which CommonWell member’s EHR they use. CommonWell acts as a hub for patient matching, record location and retrieval, patient access, privacy, consent management, and trusted data access. 

NEXT: User experience with CommonWell

Asnaani estimates that the accuracy of CommonWell’s patient matching across sites is in the 80%-90% range, compared with the national average of 50%-60% for external data exchange. Besides using the customary demographic information, he says, CommonWell’s algorithm incorporates driver’s license numbers and answers to questions that patients are asked at provider sites.

According to Asnaani, CommonWell is in a “solid position” to provide interoperability in the long term. But he cautions, “For interoperability to work, you need to get technology in front of the user and build a great user experience so they want to use the technology.”

The user experience has not been so great to date at Palmetto Health in Columbia, South Carolina, says Tripp Jennings, MD, system vice president and medical informatics officer for Palmetto. In the first year of testing and using the system, he says, “We haven’t seen the use of CommonWell on the outpatient side like we would have hoped.”

Related:Mining EHR data for quality improvement

The CommonWell service is being used mainly by physician practices outside of Palmetto Health, many of which lack access to the organization’s private HIE, he explains. For Palmetto physicians, he notes, part of the challenge is that they have separate icons in their EHR for CommonWell, the HIE, and a U.S. Drug Enforcement Administration database, which forces the doctors to go to three places for patient data. Palmetto is working with its EHR vendor to improve the user experience. 

Nevertheless, Jennings is very bullish on CommonWell’s potential. Noting that many patients are treated both at Palmetto and at hospitals in south Florida, where they travel part of the year, he says that CommonWell makes it “feasible for us to have that transactional information about the care of the patient” in those other facilities. Moreover, he adds, CommonWell is much less expensive than participating in statewide HIEs, which lack economies of scale.

“It’s the best option I’ve seen to connect on a large scale from the viewpoint of a record locator service,” he said.

NEXT: eHealth Exchange

EHealth Exchange

The eHealth Exchange, the descendant of the federal government’s Nationwide Health Information Network (NwHIN) also has been trying to solve the interoperability puzzle. But despite its rapid growth since 2012, when it became a private sector entity, the exchange has made only modest progress in helping its participants communicate with each other.

Today, the eHealth Exchange includes about 100 participants encompassing 1,600 hospitals, 13,000 medical practices, 8,300 pharmacies, and 3,400 dialysis centers. The U.S. Departments of Defense and Veterans Affairs, the Centers for Medicare & Medicaid Services, and the Social Security Administration also participate in the eHealth Exchange. 

The eHealth Exchange has been managed by a nonprofit called the Sequoia Project. In October 2012, Sequoia partnered with the EHR/HIE Interoperability Workgroup (IWG). The objective of IWG, which includes 19 states and dozens of EHR and HIE vendors, was to narrow the specifications for the transport mechanisms and the content in health information exchanges. If the participants agreed to use narrowly defined specs, it was hoped, disparate EHRs would be able to exchange care summaries without specially designed interfaces.

Related:ONC aims for improved interoperability by 2017

Eventually, the Healthcare Management and Systems Society (HIMSS), which represents health IT professionals and EHR vendors, took over the eHealth Exchange/IWG project. HIMSS recently started an offshoot called ConCert by HIMSS that has taken on IWG’s testing and certification.

Meanwhile, the eHealth Exchange has continued to add provider organizations and networks that show the ability to exchange information using the exchange’s standards and meet other requirements. Mariann Yeager, MBA, executive director of the Sequoia Project, says that the leading EHRs comply much better than they did two years ago with underlying standards such as HL7’s Consolidated Data Architecture, which forms the basis for standardized care summaries that must be exchanged to meet a meaningful use requirement. Some of these products can be quickly tested for interoperability and have very consistent performance, she says.

“Any eHealth Exchange member can query and get authorization to access records from any other participant in the country,” she notes. A major drawback, however, is that the eHealth Exchange doesn’t have a record locator service. As a result, physicians have to query specific organizations where the patient for whom they’re seeking records has received care previously.

In addition, Yeager notes, software interfaces that connect disparate systems are still required. “It’s not plug and play yet. That’s a vision. That comes when we have more specific constraints and more mature health IT systems. We’re starting to see examples of close to plug and play, but there’s no testing program in our industry that can do that out of the box.”

An Intel case study of an eHealth Exchange implementation in Oregon was more critical of the exchange’s capabilities. Intel’s researchers looked at the exchange of patient data between two different versions of Epic EHRs at Kaiser Permanente and at Providence Health & Services, and a Greenway EHR used by Premise Health, a worksite clinic.

While the participants were able to pull and push records from and to one another, partly through Direct messaging, the Intel team found that the standards lacked specificity and varied greatly in their implementation. The customization of EHRs by healthcare organizations created a unique experience in setting up each data exchange, they note. Significant resources were required to resolve these differences, fill gaps in interoperability, and “incorporate the exchanged data into coordinated care workflows.”

NEXT: Concert by HIMSS

Concert by HIMSS

This relatively new organization tests and certifies EHRs, health information exchange (HIE) software, and HISPs. Its certification requirements are far more rigorous than those of the Office of National Coordinator for Health IT (ONC), but pertain only to interoperability.

ConCert tests and certifies the interoperability of EHR and HIE software and health information service providers (HISPs), which enable Direct messaging between healthcare providers. In this context, “interoperability” means the adherence of these products to specifications developed by IWG, Integrating the Healthcare Enterprise (IHE), and HIMSS. 

Sandra Vance, MHA, senior director of interoperability initiatives for HIMSS North America, says that the initiatives of ConCert by HIMSS and eHealth Exchange are “complementary.” However, the two organizations are still working out how to use their testing programs together, she notes.

Related:4 ways EHR vendors are building better systems

Unlike the eHealth Exchange and CommonWell, Vance says, HIMSS simply certifies that products meet interoperability standards; it’s not providing a network for information exchange. It’s also not guaranteeing that EHRs or HIEs can connect without interfaces, she says, although compliance with the specs will make it easier for vendors to write those interfaces.

In addition, she says, “much less” work will be required to exchange Consolidated Clinical Document Architectures (CCDAs)  between HIMSS-certified products. 

The ConCert certification for EHRs goes far beyond what’s required for Meaningful Use stage 2. Besides Direct messaging capability, EHRs must be able to query other EHRs for records. In addition, the certification is supposed to include support for a provider directory. But Vance says IWG is still working on the latter feature.

To date, Cerner, Greenway, and NextGen have all joined the ConCert initiative, and “another great big” vendor will be announced soon, Vance says. Ten companies are going through pilot testing, she adds.

NEXT: Fast healthcare interoperability resources

Fast Healthcare Interoperability Resources (FHIR)

As long as EHR vendors have to write interfaces to other systems and HIEs for each implementation, or “instance,” of their products, interoperability will continue to be limited. Moreover, interoperability currently focuses on sending and receiving  care summaries, rather than individual records. This form of information exchange is not very useful to physicians who need specific data about a patient quickly and don’t have time to search for it in a CCDA.

Even in advanced interoperability models like CommonWell, the record locator can only determine where records on a patient are stored, not what’s in those records, notes Micky Tripathi, Ph.D., president of the Massachusetts eHealth Collaborative. That doesn’t help providers who are looking for a particular piece of information.

Fast Healthcare Interoperability Resources (FHIR), a proposed standards framework from HL7, the leading healthcare standards organization, promises to solve all of these problems. By using snippets of data known as resources to represent clinical entities within EHRs, FHIR will enable developers to write plug-ins that can connect to any FHIR-enabled EHR through a standard application programming interface (API). Not only will this eliminate the need to write customized interfaces, but it will allow providers to query databases for individual records, FHIR proponents say.

The Argonaut Project, a collaborative of 68 software developers, EHR vendors, healthcare organizations, and consulting firms, is testing the first iteration of FHIR with the 16 data elements that are found in CCDAs, plus the OAuth Internet authorization standard. Tripathi, who is also the Argonaut Project manager, said he expects some EHRs to be FHIR-enabled within the next 12 to 18 months. Among the EHR vendors in the collaborative are athenahealth, Cerner, Epic, GE, McKesson, Meditech, NextGen, and Practice Fusion.

Related:Health information exchange: No 'one size fits all' solution

It is likely that FHIR will be used initially for purposes other than health information exchange. For example, it might expand the capabilities of EHRs by allowing outside developers to plug third-party applications into them. Consumers might also be able to use FHIR-enabled mobile health apps with their providers’ EHRs.

The mechanism for FHIR-enabled EHRs to exchange health information is unclear, partly because of privacy and control issues. Patients might download their records, store them in some kind of online repository, and send them to the providers they choose. Or providers in a region might form a cloud-based network that can share data through apps plugged into their EHRs. 

Tripathi believes that both of these kinds of FHIR-enabled information exchange will be used, depending on the circumstances. Provider-to-provider exchanges won’t disappear, he says, “because most patients don’t want to be the drivers [of data exchange]. Most patients are healthy, and they don’t want to be the ones responsible for it unless there are very easy solutions to do it.”

However, he predicts that as health data becomes available in standardized forms, consumers might store their health records online in health information “banks.” These services might be provided by pharmacies or perhaps an entity such as Microsoft HealthVault, he suggests. Then, if a patient were visiting a physician that didn’t have his or her records from other sources, the patient could authorize the provider to download them from the health information bank.

NEXT: Conclusion

Conclusion

Interoperability is far from a reality today. While it’s slowly becoming easier to interface different EHRs through networks that allow query-based as well as Direct exchanges, the implementation of these exchanges is still laborious and expensive, and manual work often is required to ensure that providers get the information they’re looking for. On top of that, patient matching is still relatively poor within and between most organizations-an issue that networks like CommonWell are just starting to address.

Despite the potential of FHIR, which is exciting to many in the industry, it is still in a very early stage, notes Sandra Vance of HIMSS. She expects that the incremental approach to improving interoperability will prevail for the foreseeable future. But it’s unclear that it will produce the kind of instant access to specific patient information that providers are craving. Until that arrives, physicians will have to make do with what they already have.

NEXT: Guiding principles for nationwide interoperability

Guiding principles for nationwide interoperability

Build upon the existing health IT infrastructure

Significant investments have been made in health IT across the care delivery system and in other relevant sectors that need to exchange electronic health information with individuals and care providers. To the extent possible, we will encourage stakeholders to build from existing health IT infrastructure, increasing interoperability and functionality as needed.

One size does not fit all

Interoperability requires technical and policy conformance among networks, technical systems and their components. It also requires behavior and culture change on the part of users. We will strive for baseline interoperability across health IT infrastructure, while allowing innovators and technologists to vary the usability in order to best meet the user’s needs based on the scenario at hand, technology available, workflow design, personal preferences and other factors.

Empower individuals

Members of the public are rapidly adopting technology, particularly mobile technology, to manage numerous aspects of their lives, including health and wellness. However, many of these innovative apps and online tools do not yet integrate electronic health information from the care delivery system. Electronic health information from the care delivery system should be easily accessible to individuals and empower them to become more active partners and participants in their health and care.

Leverage the market

Demand for interoperability from health IT users is a powerful driver to advance our vision. As delivery system reform increasingly depends on the seamless flow of electronic clinical health information, we will work with and support these efforts. The market should encourage innovation to meet evolving demands for interoperability.

Simplify

Where possible, simpler solutions should be implemented first, with allowance for more complex methods in the future.

Maintain modularity

Because medicine and technology will change over time, we must preserve systems’ abilities to evolve and take advantage of the best of technology and health care delivery. Modularity creates flexibility that allows innovation and adoption of new, more efficient approaches over time without overhauling entire systems.

Consider the current environment and support multiple levels of advancement

Not every individual or clinical practice will incorporate health IT into their work in the next three to 10 years and not every practice will adopt health IT at the same level of sophistication. We must therefore account for a range of capabilities among information sources and information users, including EHR and non-EHR users, as we advance interoperability.

Focus on value

We will strive to make sure our interoperability efforts yield the greatest value to individuals and care providers; improved health, health care and lower costs should be measurable over time and at a minimum, offset resource investment.

Protect privacy and security in all aspects of interoperability

It is essential to maintain public trust that health information is safe and secure. To better establish and maintain that trust, we will strive to ensure that appropriate, strong, and effective safeguards for electronic health information are in place as interoperability increases across the industry. We will also support greater transparency for individuals regarding the business practices of entities that use their data, particularly those that are not covered by the HIPAA Privacy and Security Rule, while considering the preferences of individuals.

Scalability and universal access

Standards and methods for achieving interoperability must be accessible nationwide and capable of handling significant and growing volumes of electronic health information, even if implemented incrementally, to ensure no one is left on the wrong side of the digital divide.

Source: ONC, “Connecting Health and Care for the Nation: A Shared Nationwide Interoperability Roadmap”