Assessing the Impact of Electronic Health Records
Researcher: David J. Ballard, MD, MSPH, PhD, FACP
Occupation:Senior Vice President and Chief Quality Officer, Baylor Health Care System, Dallas
Focus: Health Care Delivery
ADA Research Funding: Clinical Translational Research Award
Electronic health records have been touted as the future of medicine: Computerized systems keep track of patient progress over time while keeping tabs on doctors and making sure the latest medical research gets used in examining rooms.
The concept goes beyond computerized charts. With your patient records in digital form, it's easier for doctors to spot what has changed since you were last in their office, see your medications and medical history at a glance, and target areas you need to work harder on.
Such records promise to be a powerful—if controversial—tool for keeping tabs on doctors, too. Having records in searchable digital form means that patient outcomes would be easy to compile and calculate by doctor, by clinic, or by hospital. Doctors whose patients consistently had good blood glucose control, for example, might be more likely to get pay raises; doctors whose patients struggled might be asked to take remedial courses in diabetes care. (Think of the way teachers and schools are rated based on the results of standardized tests taken by students.)
In 2006, Baylor Health Care System in Texas began rolling out a new electronic health record (EHR) system in the practices of almost 300 primary care physicians in Dallas, Plano, Fort Worth, Waxahachie, and other cities in the Dallas–Fort Worth area. In addition to handling record-keeping, the system was designed to offer doctors reminders and prompts. If records showed that a patient had above-average cholesterol, for example, doctors were automatically asked if they wanted to prescribe a cholesterol-lowering drug. The latest medical research could be incorporated into the system, letting thousands of doctors know what techniques worked the best without having to reeducate each physician individually.
But for all the talk, there's scant data so far on whether digital records make a difference. "Outside of a few academic settings, we really have very little information about whether electronic health records actually help patients," says David Ballard, MD, MSPH, PhD, FACP, a physician and the executive director of Baylor Health Care System's Institute for Health Care Research and Improvement in Dallas. "Literally, there are a handful of places that have done EHR evaluations, usually on homegrown systems."
Baylor's system was a combination of commercially available ones, and covered tens of thousands of patients. Across the nation, community-based primary care groups like Baylor's—typically trying to control costs while still providing good care—are ground zero for the new record-keeping networks. An ADA grant is letting Ballard gather data that may help other health care groups decide how best to use electronic records. "This kind of work usually doesn't happen without support from groups such as the ADA," Ballard says. "Health care organizations are struggling to get this stuff deployed and typically don't commit the resources and time to study it."
To do his study, Ballard essentially took "before" and "after" snapshots of diabetes patient outcomes at 33 clinics. The study included 25,642 people. "We had the numbers for the clinics before, and then we were able to observe them through this process," he says.
Comparing the numbers at clinics before and after the EHR was put in place, Ballard says, showed that when the electronic system was used, more patients met standards deemed "optimal" by the health care center and its doctors. But when the researchers broke the results down further, they showed that electronic health records aren't a panacea. Improvements tended to be limited to quick fixes: With a prompt to add aspirin to a list of medications for a patient with diabetes, or a reminder to prescribe a drug to lower cholesterol or blood pressure, the electronic system got results. "It improves outcomes that are easily achieved by pharmacological therapy," Ballard says.
Yet when it came to things that couldn't simply be fixed with the stroke of a pen—say, getting someone to quit smoking or to closely manage blood glucose—the system made little difference. "Educating people requires physician time," says Cliff Fullerton, MD, a Baylor Health Care System doctor who helped design and implement the Baylor EHR. "Getting an EHR … doesn't buy that extra 15 or 20 minutes to talk to the patient."
In the future, EHR systems will become highly interactive, doing much more than deliver prompts to doctors. An automated record system could e-mail reminders to patients at home, for example; next-generation blood glucose monitors could connect to a hospital or clinic's record system over the Internet, updating a patient's chart on a daily basis. But there may be no replacement for conversations between people and their doctors. "The take-home message is it looks like EHR improves diabetes treatment processes, particularly things that are relatively easy to improve based on medication," Ballard says. "But patient behavior is still hard to change."
Help support diabetes science:
Join the Summit Circle, ADA's society of individuals who make a planned gift. Please call 1-888-700-7029 or go to diabetes.org/giving.