"It's hard to get people to take an active role in their healthcare," said Jesse C. Crosson, an assistant professor at the UMDNJ-Robert Wood Johnson Medical School in Somerset, New Jersey.
So it's "very encouraging" to see some benefits in this study, said Crosson, who was not involved in the work but has studied the impact of electronic health records.
In the U.S., there has been a huge push to get doctors to switch from old-fashioned paper to electronic records. That's because digital records can, among other things, allow doctors, hospitals and other providers to communicate more easily - and hopefully cut down on errors, while getting more patients the tests and treatments they need.
Congress has authorized up to $27 billion in government incentives to get doctors and hospitals to put electronic records to "meaningful use." And by 2015, providers will face penalties if they don't switch.
"Meaningful use" means steps like having up-to-date medication lists for each patient, and electronically prescribing drugs.
But there hasn't been much evidence yet that electronic records are improving Americans' care.
In a recent study of 42 medical practices, Crosson found that switching to digital records did not seem to improve diabetes care. Patients at offices that made the switch were no more likely to be getting recommended tests and treatments than patients whose doctors had stuck with paper records.
But the new study, published in the Annals of Family Medicine, took electronic records a step farther.
Researchers randomly assigned 4,500 primary care patients to either stick with their normal care or have the chance to access personalized health records on a secure Web site, MyPreventiveCare.org.
The system automatically pulled information from patients' electronic records at their doctors' offices, then gave each patient a "tailored list" of preventive services they should get - like cancer screenings and immunizations. It also gave them links to educational materials on those services, and why they're recommended.
"What we tested is a higher level of functionality than exists in current practice," said lead researcher Dr. Alex H. Krist, of Virginia Commonwealth University in Richmond.
And it did seem to make a difference. Overall, patients who used the system were more likely to be up-to-date on their preventive care 16 months later: 25 percent were, which was up from less than 14 percent at the start of the study.
In contrast, there was little change among patients given standard care: Less than 13 percent were up-to-date on preventive care by the study's end, which was up from 11 percent.
The problem, though, was that most people who were offered personalized health records didn't choose to use them.
Of the 2,250 patients offered the chance, only 17 percent had done so 16 months later.
Krist said he thinks that's largely a product of the controlled clinical trial design: People were "invited" by mail to set up online health records, and that may not have cut it.
"We didn't field it in a way that a real medical practice would," Krist said.
If the personal records were actually promoted at the doctor's office, they would probably be more popular, according to Krist.
Crosson agreed that the constraints of the clinical trial were probably an important factor. "Sending something in the mail might not be the best way to get people to go online," he noted.
Right now, the MyPreventiveCare system is in use in 14 U.S. primary care practices. But the researchers are hoping to "field" it in 300 practices over the next couple years. (The system is currently a "non-commercial" product; the research is being funded by the U.S. Agency for Healthcare Research and Quality.)
To work, the personal health records have to be integrated into doctors' existing electronic records systems.
Crosson said he didn't think the logistics of doing that will be the challenging part; instead, he said, the "human factor" might be.
E-records, though, are not going to magically make us healthier.
Krist pointed out that people who used the online records were more likely to get recommended cancer screenings and vaccinations. But they weren't any more motivated to get advice on diet, exercise, smoking or weight loss, if they needed it.
That type of "health behavior change," Krist noted, is more complicated than getting a test or a shot. And people tend to need a lot more help in making those changes.
"Technology alone isn't the fix," he said.