Number of public HIEs drop, bringing viability into question
Despite federal funding that aided their creation, the number of community and state health information exchanges is declining as HIEs struggle to remain financially viable now that seed money has dried up.
Those are among the results of a new national survey published in the July issue of Health Affairs that tracked community and state HIE efforts soon after federal funding ended.
“We found 106 operational HIE efforts that, as a group, engaged more than one-third of all U.S. providers in 2014,” states the study’s authors. “However, the number of operational HIE efforts is down from 119 in 2012, representing the first decline observed since the tracking of these efforts began in 2006.”
Specifically, the study reported an 11 percent drop in the number of state and community exchanges from 2012 to 2014. In addition, the survey results showed that only half of operational HIEs reported being financially viable, while all of them reported a variety of barriers to continuing operations.
The most important takeaway from the study is that funding from the Office of the National Coordinator for Health IT was not a panacea for HIEs, says Julia Adler-Milstein, co-author of the article and assistant professor in the School of Information and School of Public Health at the University of Michigan.
The State HIE Cooperative Agreement Program, launched by ONC in March 2010 with $547 million to develop HIEs, was intended to facilitate and expand the secure electronic movement and use of health information among organizations based on nationally recognized standards.
“The State HIE Cooperative Agreement really gave HIE efforts a big cash infusion,” says Adler-Milstein, who adds that more than $600 million was awarded to various community and state organizations to organize the exchanges. “However, the unfortunate reality is that money did not prove to be enough.”
The problem, she argues, is that HIEs face daunting challenges critical to their survival, such as identifying sustainable business models, securing the participation of stakeholders and establishing information governance. “The HIE effort is just incredibly hard because there are so many things that they have to get right,” adds Adler-Milstein. “Once the funding went away, what we’re seeing is that some of these efforts are finding that there isn’t a path forward.”
The authors of the study conclude that their findings “raise important questions about whether the current vision for HIE efforts will allow for the broad exchange of clinical data, or whether alternative approaches would be more successful.”
Micky Tripathi, CEO of the Massachusetts eHealth Collaborative, contends that the study “quantifies what we all know to be true,” namely that most public HIEs have a “shaky value proposition and less-than-stellar execution.
“As with any nascent industry, we see a large number of early suppliers followed by an inevitable industry shakeout as economies of scale and scope take over,” says Tripathi, who acknowledges that he has not read the actual study, only the press release reporting the results. “At the end of World War II, there were over 50 U.S. brand-name auto manufacturers—now we have four, depending on how you count. The public HIE sector is starting to go through the same type of industry rationalization.”
Likewise, John Halamka, MD, CIO of Beth Israel Deaconess Medical Center, says it’s “completely expected to see the decline of state HIEs,” especially as the private sector steps up its own initiatives such as the development and adoption of HL7’s emerging Fast Healthcare Interoperability Resources (FHIR) standard through efforts such as the Argonaut Project.
“We’re in the process of moving from government-led initiatives to private sector-led initiatives, using the rapidly maturing FHIR standards,” Halamka says.
He believes the next two years will see a “significant uptick of interoperability in most areas of the U.S. because the incentives in MACRA/MIPS depend upon data sharing in order to coordinate care and reduce total medical expenses.”
A separate but related study—also published this month in Health Affairs and co-authored by Adler-Milstein—examined the impact that EHR vendors have had on hospital engagement in HIE.
The authors examined whether more than 2,900 hospitals had adopted an EHR from the vendor who controlled the most market share. What they found was hospitals that used the dominant EHR vendor in their market engaged in 45 percent more information exchange than hospitals that used other vendors.
However, when the dominant vendor controlled a small proportion—under 20 percent—of the market, the study found that hospitals using the dominant vendor engaged in 59 percent more HIE activities than hospitals using a different vendor. Conversely, when the dominant vendor controlled 80 percent of the market, researchers found that hospitals using that vendor engaged in only 25 percent more HIE activities than hospitals using a different vendor.
According to Adler-Milstein, the key finding is that “if you are using the dominant vendor in your market you are doing more HIE than the hospitals in the market that are using the non-dominant vendor.” For instance, she said in Epic-dominated markets, those customer sites are engaging in far more HIE, effectively making it difficult for non-Epic sites to share information.
“If we let EHR vendors govern HIE, we just need to be really sure that they are connecting providers who share patients, as opposed to connecting providers who are strategically advantageous for those vendors,” argues Adler-Milstein.
She also surmises that the increase in EHR vendor activity in the HIE space and their dominance in certain markets may explain the decline of public HIEs—a situation that could justify federal intervention.
“In markets with low vendor dominance, hospitals may engage in less HIE with hospitals using other vendors’ systems, compared to markets with high vendor dominance, because of high costs and competitive barriers,” concluded the article. “Policies designed to promote cross-vendor HIE may need to take local market competition into account.”
Nonetheless, while Tripathi calls this HIE-EHR vendor study an “interesting and creative” approach, he says he is “very wary of high-level analyses of such a fragmented market where we have really thin market data.” According to Tripathi, whose comments were based on a press release about the study not the actual text of the study, there is insufficient data on vendor markets, provider markets, or the quality and quantity of HIE transactions.
“We should not have confidence in inferences made from analysis of bad data,” he contends. “Further, the study seems to define HIE activity as participation in a public HIE. The vast majority of health information exchange happens outside of public HIEs, and as noted by the first study, that trend will continue. Thus, it is a misconception to suggest that lower engagement with a public HIE is a measure of vendor interoperability activities in any given market.”
Tripathi also makes the case that “by focusing purely on the market share of vendors as the key explanatory variable, the study seems to ignore at least half of the issue, namely, provider market dynamics, which would introduce a lot of error into the modeling.”
In addition, he charges that the study makes the explicit assumption that public HIEs are connecting needed providers in ways that are best for the patient. “This is unsubstantiated empirically, and indeed, the latest data on information blocking from CMS points to a large number of complaints against public HIEs in Missouri,” concludes Tripathi.
However, Adler-Milstein disagrees.
“We do not focus on exchange only through public HIEs, we control for provider factors such as provider competitiveness, and we limited to markets in which we had high response rates,” she says. “There are definitely legitimate limitations of the study—but those aren’t them.”
When it comes to information blocking, Halamka argues that there is no such thing as information blocking when there is a business need and a competent vendor.
“It’s very reasonable that Epic has enabled a significant amount of interoperability because they are competent and have been focused on building interoperability features into their products,” he asserts, noting that in Boston “there are no issues with vendors or technology being the rate-limiting step to interoperability—it’s all about culture, politics and incentives.”
Nonetheless, Adler-Milstein references an unpublished survey of providers in which 50 percent of respondents indicate that EHR vendors are routinely engaging in information blocking. “I think it’s a real problem, and we need to get serious and figure out how we’re going to address it.”