The study analyzed the number and type of radiology exams ordered by 1,187 office-based physicians in 2008, after 28,741 patient encounters. The researchers found no evidence that office-based physicians with electronic access to imaging or laboratory tests ordered fewer additional exams or lab tests. Instead, patients who saw physicians with access to radiology reports and/or images through PACS or an electronic health record (EHR) portal had higher odds of having an exam ordered (Health Affairs, Vol. 31:3, pp. 488-496).
Physicians who had access to electronic imaging results ordered radiology exams for 18% of total patient visits, compared with 12.9% for physicians who did not have access. Electronic record access-enabled physicians were twice as likely to order CT scans and MRI exams. These physicians ordered both CT and MRI exams for 2.2% of patient visits, compared with 1.1% for physicians with no electronic access.
As might be expected, surgeons and other specialists were more likely to order imaging exams for their patients than primary care physicians, but not by much. While surgeons ordered imaging exams during 20.7% of patient visits, medical specialists weighed in at 15.6% and primary care physicians at 13.4%. These ordering-pattern statistics were not broken out by electronic access to radiology reports.
Dr. Danny McCormick, assistant professor of medicine at Harvard Medical School in Boston, and co-authors analyzed data from the 2008 National Ambulatory Medical Care Survey of a nationally representative sample of the offices of 1,187 nonfederal physicians. Hospital outpatient departments and offices of radiologists, anesthesiologists, and pathologists were excluded. The survey was conducted by the National Center for Health Statistics.
The authors acknowledged that ordering patterns could be affected by self-referral. Unfortunately, the survey data used in the study did not include information about whether a physician had any financial interest related to ordering exams. In addition, the sample size of orthopedic specialists and cardiologists -- two medical specialties associated with above-the-norm exam ordering patterns -- was too small to produce statistically meaningful results, co-author Dr. David Himmelstein, a professor at CUNY School of Public Health at Hunter College, told AuntMinnie.com.
"One possible explanation for our findings is that ready access to imaging results, or to the images themselves, reduces the time and effort required to review study results," the authors wrote.
They suggested that "the effect may be to provide subtle encouragement to physicians to order more imaging studies," due in part to the convenience of computerized access to do so. But there could be other reasons, they noted, and limited data were available on computerization's effects on imaging costs in ambulatory settings.
What is also unknown -- and shall remain so for several years if the National Ambulatory Medical Care Survey is utilized -- is the impact of the economy in the U.S. and escalating electronic availability of patients' records through EHRs, health information exchanges (HIEs), and Web-based PACS and cloud archive portals. Himmelstein said that he and his co-authors only had access to 2008 data; 2009 data became available after the article had been completed and submitted for publication. In response to AuntMinnie.com's inquiry, he said that an analysis of 2010 and/or 2011 data might produce different findings.
The authors pointed out, however, that estimates of healthcare cost savings have been based on incomplete data, relying on a few flagship healthcare institutions with cutting-edge systems, and not generalizable to current medical practice. Use of health IT remains unproved as an effective cost-control strategy for reducing the ordering of unnecessary tests, they concluded.
Dr. James Thrall, professor of radiology at Harvard and radiologist-in-chief at Massachusetts General Hospital, a flagship healthcare institution, did not disagree with the study's findings. He noted that the landscape of physician access to electronic medical records has changed between 2008 and 2012.
"Because computer access to medical data, including image data, is now a given in the U.S., the important question is no longer whether to use information technology, but how to best use it to achieve desired outcomes in the health system," he told AuntMinnie.com. "Providing evidence-based practice guidelines to physicians through an IT program at the point of care is a way that fits conveniently and efficiently into the existing physician work process."
Thrall referenced the success that Massachusetts General has had with office-based and hospital-based physicians using decision-support systems interfaced with computerized order entry (CPOE) of radiology exams -- though it is a pioneering initiative. Over a three-year period, primary care physicians reduced their orders per visit by 25%, with the largest reductions seen among the highest users of imaging.
Other radiologists contacted by AuntMinnie.com observed that after the technologies of RIS and PACS were first deployed at primarily academic and flagship healthcare institutions, their benefits eventually trickled down to community hospitals. Over the past two decades, RIS and PACS have become ubiquitous in radiology departments, and as speech recognition and dictation systems have improved and become easier to use, they have increasingly been adopted.
All of the radiologists contacted suggested it would be interesting to see if the survey results differed if 2011 or 2012 data were used.