EMRs don't guarantee quality care, a review of 50,000 patient records shows
■ On 14 of 17 measures, physicians using paper records did equally well as those using EMRs. They even outperformed electronic record users in one area.
Physician offices using electronic medical records systems and doctors still stuck on pen and paper deliver about the same quality of care, according to a major retrospective study released last month.
The study, in the July 9 Archives of Internal Medicine, is based on data collected in 2003 and 2004 as part of the Centers for Disease Control and Prevention's National Ambulatory Care Survey. Examining more than 50,000 patient records from more than 2,500 physician offices, researchers found no statistically significant difference between EMR practices and non-EMR practices on 14 of 17 guideline-based quality metrics.
Physicians using EMRs were less likely to order unnecessary, routine urinalyses or inappropriately prescribe benzodiazepines to treat depression. Researchers described the differences, though, as "clinically insignificant," because the adherence to guidelines was so near 100%. Pen-and-paper doctors were 14 percentage points more likely than their tech-savvy counterparts to prescribe statins appropriately for patients with high cholesterol.
The published findings surprised researchers.
"We are all fans of electronic health records," said lead author Jeffrey A. Linder, MD, MPH, describing the study's five co-authors. "We'd hoped that there would be some association with electronic health records and quality. It turns out we didn't find any."
The news comes amid increasing pressure on physicians from the government and insurers to implement EMRs to cut costs and improve quality and patient safety. Only a quarter of office-based physicians have gone digital, according to the most recent CDC data.
The study's finding sheds light on how little is known about the efficacy of electronic systems as quality improvement tools, said Ross Koppel, PhD, an investigator at the University of Pennsylvania School of Medicine's Center for Clinical Epidemiology and Biostatistics.
"EMRs seem to make sense, but like almost all of health IT, right now we're in the period of faith," said Dr. Koppel, who authored a 2005 JAMA study that found that computer systems for physician order entry actually can facilitate medication errors.
"We have to pay more attention to the details" of how physicians interact with electronic systems in real-world practice, he added.
Dr. Linder said one potential explanation for EMRs' poor showing in his study is that the CDC survey asked whether doctors were using an electronic system but did not delve any further. He speculated that some physicians might believe they are using an EMR when they are not. He said others could be using lower-end systems that do not offer the clinical decision support that health IT advocates say can make the biggest impact on quality. Yet other doctors may not be taking full advantage of the functionalities of their systems.
The ability of digital systems to drive quality, Dr. Linder added, depends on how well physician practices implement the technology.
EMRs "are part of the solution, but they are not sufficient for improving health care quality," said Dr. Linder, assistant professor of medicine at Harvard Medical School. "You can't just take all the paper in doctors' file cabinets and pour it into a computer and watch quality magically improve."
Learning proper IT use is key
The authors of the Archives paper write that their "findings are not a refutation of previous studies." And prior research has demonstrated that EMRs can improve quality. The Annals of Internal Medicine, in its May 16, 2006 issue, presented a systematic review of 257 studies. The review found many physicians increased their adherence to care guidelines, committed fewer medication errors and monitored patients more closely.
But much of this evidence came from randomized controlled trials at benchmark institutions such as Brigham and Women's Hospital in Boston, Intermountain Healthcare in Salt Lake City and the Veterans Health Administration. These organizations implemented internally developed systems, while most physician practices adopt off-the-shelf commercial software.
The Annals review concluded that the benchmark institutions improved quality and efficiency with health IT, but "whether and how other institutions can achieve similar benefits, and at what costs, are unclear."
Jon White, MD, said the new study gets it right.
"Just having an EHR does not by itself float the quality boat," said Dr. White, health IT director at the Dept. of Health and Human Services' Agency for Healthcare Research and Quality, which funded the new Archives study. "Making the information digital doesn't automatically let you do things with it that you haven't done before. It enables you to do it."
Steven Waldren, MD, director of the American Academy of Family Physicians' Center for Health Information Technology, agreed that assisting doctors to get the most out of their EMRs is the next critical step.
"In the past, we were all about getting physicians to adopt the technology," Dr. Waldren said. "Now we are worrying about helping them learn how to use that new functionality and get comfortable with it in their offices."
Even when physicians and hospitals successfully adopt systems with the ability to warn them about potential drug interactions or prompt them to render guideline-based care, many health professionals develop "alarm fatigue," said Penn's Dr. Koppel.
Robert Lamberts, MD, an Evans, Ga., internist and pediatrician whose practice uses General Electric's Centricity EMR, said the phenomenon is real. "It is just sensory overload when you think about how many recommendations go with each patient and how many things we could do, whether it's putting them on an ACE inhibitor or putting them on aspirin," he said. "It does get tiring, but I guess the difference is that with an EMR, if you're ignoring an issue, at least you know you're ignoring it."
Dr. Lamberts, who has received $1,000 speaking fees from the Healthcare Information and Management Systems Society for lectures on how to adopt an EMR, said he sees the quality impact of his EMR every day. For example, the system encourages him to target a 130 mmHg systolic blood pressure in his patients, where before he was satisfied with a 150 reading. Moreover, he said, it is long past time for doctors to go digital.
"Would you go to a bank that kept all their records on paper and said it was not going to go to a computerized system because it is too expensive?" he asked.
At the 2006 AMA Annual Meeting, delegates adopted policy saying that public and private payers should not require physicians to use EMRs. A bipartisan bill in the Senate, the Wired for Health Care Quality Act, includes $278 million in matching grants to help doctors purchase health information technology. At press time in late July, supporters of the bill were hoping it would come up for a vote on the Senate floor before the Aug. 6 summer recess.