WEDNESDAY, Aug. 31 (HealthDay News) -- New research suggests that using electronic medical records instead of paper files could greatly improve care for diabetic patients by boosting communication.

Diabetic patients "did better and improved faster" at various kinds of medical clinics that had switched to e-records, said study author Dr. Randall D. Cebul, a professor of medicine at Case Western Reserve University in Cleveland. "The differences were rather remarkable."

However, there are caveats to the study. The research doesn't prove that electronic records directly improved patient care; other factors could explain the difference. And it can be costly for clinics to convert to electronic records, although many may be able to get financial incentives from the federal government.

Cebul said the findings are strong. He acknowledged, however, that clinics often failed to fully follow guidelines about care for diabetics, even when they used electronic records.

Converting to electronic records may seem like a slam dunk when it comes to patient care. Proponents say they make it easier for doctors to communicate with patients and with one another. The records are also supposed to cut down on medical errors by doing things like providing warnings about medication allergies.

The Obama Administration is so confident that a move to e-records will improve care and cut costs that it made the shift a key part of health care reform efforts. Still, doctors seem slow to adopt the technology: A HealthDay/Harris Interactive poll conducted a year ago found that fewer than 1 in 10 adults used email to communicate with their physician.

And researchers have had trouble confirming that electronic records actually boost medical care. "In a lot of studies, there hasn't been enough change that they can say it's improved things," noted Richard Hillestad, a senior principal researcher who studies medical record-keeping at the non-profit research group the Rand Corporation.

In the new study, researchers analyzed the medical records of more than 27,000 adults who received care for diabetes at clinics in the Cleveland area. They wanted to see if there was any difference between results at clinics that had adopted electronic records and those that hadn't.

The study findings appear in the Sept. 1 issue of the New England Journal of Medicine.

After adjusting their statistics so they wouldn't be thrown off by factors such as differences in medical conditions among patients, the researchers found that people who visited clinics with electronic records were more than a third more likely to have received care that met all four clinical benchmarks. The benchmarks require that patients undergo eye examinations, pneumococcal vaccinations, kidney management and measurement of glycated hemoglobin (also known as HbA1c, a means of tracking blood sugar control).

Patients at clinics with electronic records were also 15 percent more likely to meet all five benchmarks for their personal health, including specific ranges for blood pressure, cholesterol and weight.

However, it's possible that the clinics with paper records simply provided worse care in general -- their decision to not move to electronic records could be a sign that they're behind the times in other ways. But Celub said the evidence suggests that is not the case.

He acknowledged that the study does not use the prospective approach that's considered the gold standard in medical research. Instead, it looked back at what happened to the patients instead of randomly assigning clinics to use electronic or paper records and then tracking their progress. "It's not a randomized controlled trial," he said, "but that will never happen."

Why? Because physicians won't want to be stuck with paper records. "How many doctors would like to be randomly assigned to stay in the dark for three years?" he asked. "Nobody."

The study findings also reveal how few patients met all the standards. Of those who went to clinics with electronic records, the unadjusted statistics show that just 51 percent met all four benchmarks for care, compared to only 7 percent of those at clinics that relied on paper. And just 44 percent of those at the electronic-record-using clinics met four out of five benchmarks for how they were doing medically, compared to only 16 percent for those at clinics that relied on paper.

"We have a long way to go," Cebul said.

Still, Rand Corp.'s Hillestad said the study does suggest that electronic records have value in the treatment of people with chronic illnesses. "If you can start to see evidence that electronic health records are improving the care and cost of chronic illness, that will be a really big step."

More information

There's more on personal health records at the U.S. National Library of Medicine.