Do non-orthopedists order more MRI scans on normal knees than orthopedists do? An answer based on your anecdotal experience may well be correct -- if only for the city or center where you practice.
In an era of pervasive medical cost-consciousness, any tendency by general, non-orthopedist physicians to order unnecessary MRIs could be an unacceptable drain on limited healthcare resources. Thus, the perception of significant differences in imaging requests has generated ongoing interest and studies.
The latest study is by Dr. Paul Sherman, a radiology resident at Brooke Army Medical Center in Fort Sam Houston, TX. He was surprised two years ago by a published study from Albany Medical College that found no significant differences between the MRIs ordered by orthopedic surgeons and other physicians (Radiology, June 1998, Vol. 207, pp. 633-636).
"We didn't think our experiences at Brooke Army Medical Center supported these conclusions," Sherman noted. So he undertook a retrospective review of records to test his hypothesis that BAMC's physicians did differ in their MRI referrals on suspected knee injuries.
Sherman and colleagues examined the records of patients who underwent knee MR imaging from January 1995 through December 1998, eventually checking the charts of 754 patients. Of these, 374 were referred by orthopedic surgeons; 384 were seen by non-specialists.
Among the patients referred by non-orthopedists, 69.5% of the MR images showed normal anatomy. In patients sent by orthopedic specialists, only 30.5% had normal anatomy on their scans, a highly significant difference (p<0.00001).
The researchers also tabulated the number of studies showing meniscal and ligament tears, chondromalacia, osteoarthritis, and Baker's cysts. There were no significant differences in the abnormalities seen in the two groups -- except for lateral meniscus tears, which were seen in 40.4% of the patients referred by non-orthopedists and 59.6% of those referred by orthopedists.
The patients ranged in age from 12 to 93, with a mean age of 35. About 60% of the patients in the study were men. Sherman said the age ranges were similar in both groups. Overall, the patients underwent a total of 758 MR examinations.
Sherman speculated that non-orthopedists send out more "normal" knee MRIs in response to the types of injuries reported, primarily among young men. "The doctor expects to see some abnormality, and when that abnormality is not immediately seen, they refer the patient to more specialized care," he said.
But Sherman did not find gender-related treatment differences in his study population, which included 301 women. In anticipation of seeing differences in referral patterns by patient age, he also divided the population into seven age groups for statistical analysis. However, there did not appear to be any statistically significant differences among the age groups.
The findings will likely be used to educate BAMC physicians on knee MRI, Sherman said, as there is no indication that center administrators would favor a dramatic response such as denying direct referrals by generalist physicians.
However, some military medical facilities around the world do require orthopedic review of cases before knee MR imaging is ordered, Sherman said, in response to demand that had overwhelmed smaller radiological services.
During his presentation at the November meeting of the RSNA in Chicago, Sherman also acknowledged several limitations to his study. In some cases patient history data was missing, Sherman noted. Another obvious limitation, he said, was that the data reflect military medical procedures that could differ from civilian practice.
But the greater limitation may be the well-established phenomenon that practices vary from location to location.
"It could be a dangerous thing to try and generalize this study," said Dr. David Rubin, assistant professor of radiology at Washington University, St. Louis, and moderator of the RSNA session where Sherman presented his study. Noted Rubin: "Clearly there are very huge local differences in healthcare practices."
By Edward SusmanAuntMinnie.com contributing writer
January 31, 2000
Copyright © 2000 AuntMinnie.com
