The Trials and Tribulations of Health Information Sharing: The Turbulent Rise of the Rhio
Article Outline
- Lessons Learned
- A Matter of Trust
- Information Exchange in the ED
- How To Measure the Gains
- The Future of Health Information Exchange
Any given emergency department (ED) patient can have dozens of medical record scattered across a city or a region, leading to inefficient health care resource use and general poor care, but efforts to fix the problem have failed time and again.
Information sharing, creating one medical record that follows the patient wherever he or she presents, is the technological holy grail sought by dozens of health information exchange projects around the country. Called Regional Health Information Organizations (RHIOs), they have different organizational structures, goals, funding sources and membership. But what they have in common, despite the enthusiasm for the potential good they could do, is a financial uphill battle.
They are the second generation of a trend that failed utterly in the 1990s, then referred to as Community Health Information Networks (CHINs). Over the past few years, improvements in health information technology, increasing costs of care and the quality movement have nourished resurgence in the trend, with the hope that growing conditions are better this time around.
There are a few exchanges that have found a business model to sustain them and an unusual level of cooperation among the parties in their markets, notably in Indianapolis, Spokane and Colorado.
But the headlines have been more focused on some highly public failures. An exchange in Santa Barbara, CA, was touted as a national model and had high-powered support from the California HealthCare Foundation and David Brailer, the former head of health information technology for the federal government. But the effort lost momentum and the exchange died. More recently, a health data exchange in Portland, OR, stalled, reportedly because participants balked at the price tag.
Meanwhile, at least 165 health information exchange projects exist in 49 states and the District of Columbia, according to a recent survey by the eHealth Initiative, a not-for-profit group that researches and supports health data exchange. Forty-five of the projects described themselves as in the implementation stage and 26 as fully operational, the survey found.
These efforts have a special resonance in the ED, and often start out with ED data sharing as their first project because it is expected to offer quick benefits to busy emergency physicians hungry for background data about their more complicated patients. Data exchange is also a goal for public health surveillance for catching epidemics early and helping the ED cope with disaster response.
To make a go of it, though, exchanges need to solve tricky funding, cooperation, technical and privacy challenges.
Lessons Learned
The positive side of the failed health information exchanges is that their leaders generally have been willing to discuss publicly what went wrong and provide some valuable insight. The post-mortem on the Santa Barbara experiment, for instance, was carried out in a Washington, DC, press conference and a series of papers on the Web site of the health policy journal Health Affairs. Brailer’s piece, for instance, argued that Santa Barbara, which got its start in 1998, was more academic exercise than a true health information exchange because it lacked many of the attributes a successful exchange would need, such as participants who had a financial stake in the exchange. Other weaknesses, Brailer said, included a top-heavy organizational structure with too much emphasis on high-tech gadgetry and a provider-driven model with too little involvement from other stakeholders. It was also expensive and led to widespread confusion about privacy issues.
Like Santa Barbara, many of the organizations began with grant money from state or federal government, or foundations. Eventually that money will run out, and health information exchanges will need to find a way to sustain themselves, either with contributions from their member organizations and/or by charging fees for providing data-related services. For instance, the Indianapolis exchange is funding itself largely from hospital fees for transmitting laboratory, radiology and other medical data among local doctors, hospitals, labs and other health care organizations. New York City’s exchange is looking at disease management as a potential revenue stream. “The successful ones have viewed themselves as businesses and saw the doctors, hospitals and other players as customers,” says Janet Marchibroda, chief executive officer of the eHealth Initiative.
The health care system’s tendency to financially reward doing more holds back progress with data sharing, Marchibroda says–hospitals have an incentive to do additional tests, so the exchange actually threatens their bottom lines.
A Matter of Trust
Another hurdle is overcoming the natural competition among players in a marketplace to develop enough trust–some use the term social capital–to open up their databases. Several competing hospitals in Spokane, WA, have managed to set aside their individual agendas to create Inland Northwest Health Services, through which they share a variety of information technology and emergency services.
For the Spokane hospitals, a separate not-for-profit organization is the model that works, and many exchanges do the same, hiring staff and a establishing a board including the local health care players. In Northeastern Pennsylvania, that model didn’t work, and the board of the RHIO there chose in May 2007 to shift to a more loose confederation that gave all parties more breathing room.
The New York Clinical Information Exchange (NYCLIX) got its start from the state hospital association, with a goal of improving clinical care by making more data available at the point of care. Because of its beginnings, the organization has been somewhat provider-driven, says board chairman Dr. Gil Kuperman, but he thinks a wider array of participants will provide a stronger base–health plans, nonacute provider organizations, state medical societies and consumers.
Dr. Kuperman says the group is relatively new but working hard to ensure that all the players feel like they own the process. He is optimistic that the traditional competition among New York hospitals will yield to an understanding that sharing data to improve quality of care is in everyone’s best interest. “The organizations are generally feeling, yes, we need to compete, but we don’t want to compete over access to data,” Dr. Kuperman says. “By making that a common good we can make life better for providers and patients.”
Beyond the issues of competition are technical hurdles–establishing common data standards and keeping information secure. Significant progress has been made on the federal level on data standards, but privacy remains a murky legal area and was one of the nails in the Santa Barbara RHIO’s coffin.
“The jury’s still out about how precisely this will happen,” says Marchibroda. “Hospitals in some markets will work together and we’ll see doctors working with health plans in other markets. It will come in different shapes and sizes.”
Information Exchange in the ED
Many of the emerging HIEs are focusing first on the ED, in some cases offering access to patient records from visits to other area providers and sometimes offering patient medication information from the databases of insurers.
Among emergency physicians there appears to be strong interest in access to the information. NYCLIX surveyed emergency physicians in 12 New York hospitals; 216 responded for a response rate of 58% to the Web-based survey tool. The results were published in August in an unedited online preprint to the Journal of the American Medical Informatics Society.
The gist of the survey was that outside information is difficult to obtain (NYCLIX is not yet providing shared information) but access to it would benefit at least some of their patients. Eighty-five percent of the respondents said it is difficult or very difficult to get patient data from outside their own hospitals, and 72% said their efforts fail at least half the time. Nearly two-thirds said that at least a quarter of their patients would benefit from easy electronic access to data, and 39% said at least half their patients would benefit.
Interestingly, 76% of the respondents said they had never heard of health information exchange before the survey, but 97% said they thought the way care is delivered would benefit.
Other research and experience indicates that the data from outside the hospital would have to be easy to access, says Dr. Jason Shapiro, the Mt. Sinai Hospital emergency physician who headed the survey. If a doctor has to take time to open a new browser window and punch in yet another login and password, it is much less likely he will bother to do so.
Investigators at New York-Presbyterian Hospital/Columbia University Medical Center looked at ED use of the hospital’s electronic medical record. They found that the record was accessed less than half the time–for between 20% and 50% of patients seen in the ED, according to their paper, published in the March/April issue of JAMIA.
“Is the glass half empty or half full?” asks Dr. George Hripcsak, professor of biomedical informatics at Columbia University and lead author on the paper. “Instead of concluding it wasn’t used as much as you might think, you could say that half the time they did look at it.”
Also, older history isn’t required for every patient–those with very simple or very severe injuries are unlikely to benefit, he notes. This raises the question of what the ideal percent of use would be to make the investment and trouble of setting up an exchange worthwhile.
Dr. Marc Overhage, a former emergency physician who runs the Indianapolis data exchange, says it is hard to pinpoint a percent of the time that it is useful. “There are something like 20 to 25% of patients who come in the door with chronic illness that probably benefit from a more detailed record review. We know it’s not 100% and it’s not 10% so it’s somewhere in the middle.”
Even if the system is accessed less than half the time, it takes the avoidance of only a few inappropriate admissions or expensive radiology workups to make it worthwhile, argues Dr. Shapiro.
“In emergency medicine we’re working in the dark right now and hopefully these health information exchanges will figure out what the business model is and how to be sustainable and be linked together in a national health information network,” he says. “It’s very primitive the way we are practicing today. In ten or 15 years, or hopefully less, we won’t be able to believe we practiced without it.”
Along with knowing allergies and medications, it can also be useful in the ED to know whether a patient has a regular source of care in case they need referral to primary care. The 3 most common requests the Indiana exchange gets from EDs are the patient’s primary care doctor, allergies and medication lists.
How To Measure the Gains
Another open question is how to quantify the benefits of health data exchange. NYCLIX is struggling with the question of quality metrics, notes Dr. Shapiro. The simplest may turn out to be monitoring medication use and the avoidance of drug interactions, allergic reactions and the spread of antibiotic-resistant organisms.
In Indiana, Dr. Overhage says the simplest gauge of success has been dollar savings, though that’s not the best metric to reflect quality of care. “It’s very difficult to create a measure of quality of care in the ED. Who’s going to fund a study where you have to have 2 ED docs review the entire record?”
Because of the difficulties, the system relies for support more on the perception among doctors that it is useful, Dr. Overhage says. And they have found it so. The exchange has worked on providing busy emergency physicians with a concise screen of relevant information such as recent diagnoses, care providers, blood types, and a few laboratory tests.
“ED doctors are working in a vacuum, spending a lot of effort extracting information from the patient, hunting and gathering data,” Dr. Overhage observes. “The information they get is often wrong or incomplete, so I’ve got to think that having information that is well organized, carefully delivered, and highly thoughtfully condensed that a clinician can use when appropriate seems like it ought to be a huge win.”
The availability of the data has changed practice in some EDs. Some have had to rewrite protocols for issues such as whether a test needs to be redone if it has been done already in the past 24 hours.
The Future of Health Information Exchange
There have been calls for a national information infrastructure that would link all the regional efforts. That seems a long way off, but some are optimistic. “Even though there are technical challenges and challenges with financing, in five to 10 years we will see a more robust clinical information system infrastructure in this country,” predicts New York’s Dr. Kuperman.
Dr. Overhage expects the health information exchange movement will simmer along for another 3 or 4 years and then see a strong surge.
While there have already been high-profile failures in health information exchange, boosters such as Marchibroda argue it’s just a matter of time before these exchanges become commonplace. “Getting information mobilized outside organizations is going to take a lot of time and leadership and commitment, and it’s going to be hard, but we don’t have any choice,” she says.
Inefficiency in the health care system, a big rise in chronic conditions and the pressure for transparency are all pushing HIEs along, with the strongest argument in the ED.
PII: S0196-0644(07)01550-8
doi:10.1016/j.annemergmed.2007.09.013
