GAO: Pentagon Health Records Don’t Compute

Even though the Veterans Administration and the Pentagon have been working to integrate their electronic medical records systems for over a decade, they still can’t fully communicate with each other, according to a report (PDF) by the Government Accountability Office. While the departments have made some progress, the GAO says they “continue to face challenges in managing the activities required to achieve this inherently complex goal.” Of the information the departments are able to share, much of it is not readable by computers. Instead, much of the sharing involves what are essentially electronic versions of paper documents, rather than fully sortable and analyzable information databases. Without fully computable data, the DoD and VA are “missing 95 percent of the potential benefits of an integrated system,” says Phillip Longman, a senior fellow at the New America Foundation and the author of Best Care Anywhere: Why VA Health Care is Better Than Yours.

“The really big payoff for health IT is when you have it truly integrated across large populations so that researchers can go back after the fact and study patient outcomes,” Longman explains. The VA’s VistA system, which has fully computable data, allowed VA doctors to be among the first to notice the dangers of Vioxx, an arthritis drug that increased the rate of heart attacks in patients. But if you’re dealing with paper records or with files that have simply been scanned, you can’t do that kind of research, Longman says.

And that’s exactly the problem. After a decade of effort, the DoD is still struggling to get its proprietary, black box software to work with the open-source VistA. The sharing they have now “essentially takes the place of a fax machine,” Longman says—it doesn’t hold a candle to the effectiveness of simply adopting VistA government-wide. Critics like Longman have long wondered why the Pentagon has spent $20 billion on a medical records system that military doctors hate because it barely works and is so hard to use. “Were it not for various bureaucratic wars, DoD could be wired for VistA in an afternoon,” Longman says.

Longman’s real worry, however, is that the kind of inefficiency and chaos that has plagued the DoD’s electronic medical records boondoggle will soon hit the private sector. He believes that if the government requires hospitals and other health care providers to adopt electronic medical records as part of health care reform legislation, then those providers should have to show cause if they want to use closed-source software. (VistA, because it’s free and extremely sophisticated, would presumably be the alternative for those who could not prove they needed to buy proprietary software.) “The GAO seems to be completely [oblivous] of these larger issues,” Longman says. He sees two different paths for health IT. “If most of American health care ends up being wired by proprietary software makers, we will have set ourselves on a path of development that will negatively affect how well we adopt health IT,” he says. Open-source software like VistA has the advantages of being developed with the direct input of doctors (its users), being constantly updated and refined, and being free. Longman is hoping that just because DoD decided that free, sophisticated, and effective wasn’t the route it wanted to go doesn’t mean the rest of the American health care sector has to follow suit.