Medical Records: We Need a National System

If you're like most Americans, you like to take responsibility for your own health, even the aspects of it that you find boring, incomprehensible or just icky. You keep detailed records of every doctor's visit and medication, just like you do for every tax deduction, maybe even in the same shoebox. You know that your blood pressure is the one with two numbers, right? and that your hay fever cleared up with those yellow pills from Dr. What's Her Name, with the accent, and that you had your last checkup for colorectal cancer right after that series on the "Today" show. If that's you, you can stop reading this right now. Just remember to take the shoebox with you to the hospital.

Actually, it's not that simple. A major change is occurring in medical record keeping, driven by the embarrassing realization that until now the information systems that keep track of Americans' cancer treatments have mostly lagged behind the ones they use to buy movie tickets online. "Eighty percent [of small practices], which provide more than half the medical care in the country, do not have computerized clinical record keeping," says Dr. David Kibbe, a leading consultant on health-care technology. They keep patient records in file drawers; the doctors scribble prescriptions on pads of paper and communicate with other health-care providers by picking up the phone and calling. The Obama administration's economic-stimulus package allocates almost $20 billion to help move this jury-rigged system into the 21st century, including direct subsidies to physicians for purchasing health-records systems … as soon as the nation figures out what the system should be.

In health-care think tanks there is frustration bordering on panic over the danger that the nation will miss a historic opportunity if millions of American doctors adopt a hodgepodge of stand-alone systems that don't readily communicate with each other. "Whatever is done has to be accompanied by a whole series of other changes," says Shannon Brownlee, Schwartz senior fellow for the New America Foundation, which is in the forefront of studying this issue. "There are a whole series of good little ideas in health care now, but if each is implemented separately it won't add up to an improvement. We'll end up digitizing a really bad system."

At a minimum, experts say, a national electronic health-records system should do the following:

• Permit immediate electronic information exchange between doctors, saving time on taking patients' history and money on tests or X-rays that may have already been performed.

• Replace handwritten prescriptions with an electronic network linking doctors and pharmacies. This would reduce mistakes, save time wasted on phone calls back and forth and enable automated warnings of drug interactions and drug sensitivities.

• Facilitate "data mining" for information about new (or existing) treatments. A new drug undergoes elaborate trials for years before it goes on the market, involving hundreds or perhaps thousands of subjects—and then gets dispensed, potentially, to millions of patients. How it affects them is potentially lifesaving information that now gets reported anecdotally and spreads by word of mouth, if at all. But a computer that aggregated the findings of large numbers of doctors could detect rare problems, or even unexpected benefits. That was how the problems with Vioxx surfaced, in computer data from the Department of Veterans Affairs and large private HMOs that have their own electronic records.

• And, finally, establish standards of care for disease against which actual treatments can be measured—or, to put it plainly, keep an eye on doctors, which may be the real reason why some of them are leery of it. "All doctors think they're practicing good medicine," says Len Nichols, director of the health-policy program at New America. "The data suggests not necessarily." Treatments are constantly evolving, and some doctors adapt more easily than others. "The best electronic-records systems are embedded in an application that helps the doctor decide, 'Do we do an MRI for this patient?' 'Is drug A better than B?' " When that kind of information becomes available, it is probably only a matter of time before the current fee-for-service model is supplanted by a system in which doctors are paid for performance, according to a metric of patient outcomes, adjusted for how sick they were when they went in. It is still unclear what the new landscape of American medicine will look like, and how electronic records will interface with the other changes being proposed by various players. But with luck, the day will come soon when you won't have to go looking in the shoebox for the paper on which you wrote down that you're allergic to ampicillin … no, wait, amoxicillin?