JONATHAN COHN NOVEMBER 25, 2011
[Guest Post by Darius Tahir]
Early this September Stanford Hospital discovered that somebody had posted personal data for nearly 20,000 emergency room patients online, so that anyone who happened across the page could look up everything from the patients’ names to the codes identifying their various diagnoses. Worse still, the data had been online for more than a year.
The tale of how the data ended up online involves the sort of slapstick you expect from a bad sitcom. It seems that a hospital staffer had given the data to a job applicant, as a test to see whether the applicant could manipulate it. The applicant then posted the data online, seeking advice on how to convert it to a bar graph. He didn’t get the job, but the damage had been done.
The episode may have wider relevance: The data came from electronic medical records. And if you have followed the health care debate, then you’ve probably heard reformers talk about how such records can transform medicine, by raising the quality of care while decreasing its costs. The idea, at least in theory, is that moving medical data from paper to digital sources will mean fewer mistakes, more sharing of crucial health information, and more opportunities to analyze and learn from a wealth of new data.
But, as the Stanford episode suggests, protecting privacy becomes a lot more essential when medical records go digital. And that raises a key question: Will the security measures necessary to protect against such breaches of privacy make them impractical to use?
By international standards, the U.S. appears to lag badly when it comes to the use of electronic medical records. Definitive statistics are tough to pin down, but a 2009 Commonwealth Fund report comparing primary care physicians worldwide found that 46 percent of American primary care physicians had electronic medical records, second-to-last among the ten countries Commonwealth compared. (Relatively speaking, the U.S. fared somewhat better on the question of whether physicians had “advanced capabilities.” Twenty-six percent did—or roughly in the middle of the countries the Fund surveyed.) The picture doesn’t seem to be much better for hospitals, with only one in five of them adopting basic electronic medical records.
The Obama administration has tried to hasten this shift to the digital age, by setting aside money in both the Recovery Act and the Affordable Care Act for the development and adoption of electronic medical records. Among other things, the new rules for the administration’s pilot program for bundling payments—Accountable Care Organizations—give points to hospitals that use electronic medical records. And most experts, left and right, agree that the case for wide use of such records is clear. Instead of pharmacists or nurses trying to decipher the handwriting of many doctors, which can look more like hieroglyphics than handwriting, they’d get their instructions from electronic text, vastly reducing errors. Doctors wouldn’t have to rely on patients, frequently in pain, trying to remember their entire medical history. They could just look up the digital records.
Research into public health would also become easier. In fact, that’s already starting to happen, as both the Food and Drug Administration and Department of Veterans Affairs have claimed that using electronic medical records to spot clusters of heart attacks helped them identify the risks of Vioxx, an anti-inflammatory drug recalled in 2004.
Better still, electronic records could improve care in ways that patients would quickly perceive. Christine Bechtel, Vice President of the National Partnership for Women and Families, explained to me what patients frequently say in focus groups: “I just want my doctors to talk to each other … . I’m tired of showing up in the exam room and being handed that clipboard for the umpteenth time as if I’ve arrived from another planet … . Technology could help me!”
All of these benefits, however, come with one potential risk: For this kind of coordination, records must be accessible by providers throughout the health care system (or be “interoperable,” as wonks call it). And that increases the possibility of information leaking out, by accident or by design—if, say, hackers were to figure out a way of breaking into a major hospital system’s records. A Wall Street Journal report estimated that more than 10 million Americans’ health data have been exposed by breaches of some kind.
For promoters of electronic medical records, the hard part isn’t coming up with security to thwart such leaks. It’s coming up with security that doesn’t limit the usefulness of the technology itself.
Aaron Carroll, a professor at the Indiana University School of Medicine and influential health blogger at The Incidental Economist, told me he worries about an overreaction to the possibility of breaches, “like with airport security,” with software vendors trying to keep up with the demands to lock down the security threats. The problem, he noted, is that “there’s a direct correlation between a lockdown and doctor unhappiness [with the system].” In other words, if security software becomes unwieldy, doctors might get so fed up that they stop using the technology.
Phillip Longman, a senior research fellow at the New America Foundation, notes that the Veterans Administration has struggled with this dilemma. The VA won acclaim for its integrated information systems, which both reduced costs and improved quality of care. But after a staffer lost a laptop in a burglary in 2008, the VA tightened security on its system, turning a vibrant, open-source network into a centralized one in which changing things to suit the needs of individual hospitals was difficult. “There’s a cost to privacy,” Longman concluded. “If you’re going to put up firewalls for proprietary software, you give power to the vendor and lose interoperability.”
Not all experts share this fear, however. Intermountain Hospitals is an integrated health system in Utah that wonks constantly praise for its quality and efficiency. It’s had an electronic medical record system since 1972. Dr. Chris Wood, Medical Director for Information Systems at Intermountain, told me that staff can now track every access of a medical record on a single board, in real time. While there are regular attempts to access data from outside—“there are a lot of savvy people out there [trying to hack into the system],” he concedes—he feels relatively optimistic that they’ve contained such intrusions.
That leaves breaches from the inside—to which the solution is education (to make sure doctors and caregivers aren’t being careless) and strict discipline of wrongdoing. Wood also opted for an airline metaphor in explaining the need to balance protection and functionality: “In an airline, you try to be as safe as possible—but the purpose is to get from point A to point B.” Christine Bechtel, too, thinks the market is making headway in that direction. “The software is becoming more usable and customer-friendly,” she says, and she predicts big changes in the future of the market as it adapts to a model with more care-coordination.
And how do the patients feel about electronic records? Bechtel shared the preliminary data from a study her organization did comparing patients whose doctors used electronic medical records to patients whose doctors didn’t. Both groups, it turned out, trust their physicians about equally. And three-quarters of those patients who didn’t have electronic medical records wanted them.
The patients who did not yet have electronic medical records were slightly more likely to worry about data exposure or theft—66 percent, versus 60 percent among those in practices that already used the records. But Bechtel suggested these results might indicate that patients will become more comfortable with electronic medical records as their use becomes more widespread and doctors get a chance to converse with their patients about the meaning and import of the changes.
If nothing else, comfort with electronic records may grow simply because younger people are more comfortable with the digital world. As Longman told me, “There’s a tremendous generational change—members of the Facebook generation just scratch their heads [when they hear about this].”
Darius Tahir is an intern at The New Republic.