Survey identifies orthopedists’ preferences for MR knee reports

November 30, 2004

Radiologists need to communicate with referring physicians more often and listen more closely to orthopedics surgeons’ reporting preferences, according to the results of University of California, San Diego survey that were announced Tuesday at the RSNA meeting.

Radiologists need to communicate with referring physicians more often and listen more closely to orthopedics surgeons' reporting preferences, according to the results of University of California, San Diego survey that were announced Tuesday at the RSNA meeting.

The survey found that orthopedic surgeons are frequently dissatisfied with the format and content of radiologist reports on MR knee examinations, said Dr. Kay Lozano, a radiologist at Radiology Imaging Association in Denver. Lozano conducted the survey this year at UCSD under the supervision of radiology professor Dr. Donald Resnick. Findings were based on 135 responses to a multichoice survey mailed to 2000 members of the Arthroscopy Association of North America.

About 44% of the orthopedic surgeons said that their radiologists are not aware of their preferences for reporting. About 70% of the respondents said they contact the radiologist when they see discrepancies between arthroscopy and the MR reports. About 47% said that radiologists never read their arthroscopy reports to find discrepant findings.

"Communication is the best way to fix this problem, by personal interaction, conferences, telephone calls, and review of arthroscopy reports," Lozano said.

Respondents said that radiologists fail to make the grade for report uniformity and cited uneven interpretative quality and a frequent lack of attention to articular cartilage.

About 71% of respondents said findings relating to articular cartilage defects are the most common abnormalities overlooked in radiologists' reports. About 12% said descriptions of abnormal plicae are most often ignored, 9% said that meniscal tears are most frequently missed, and 4% said that radiologists most often fail to describe ligament injuries.

In terms of report style, 35% of the respondents prefer a report organized by anatomic structure, while 27% prefer it organized by departments. Twenty-one percent said they want reports organized by abnormal and normal findings.

Relating to report format, a majority of respondents prefer a standard findings and impressions section in the report, while 28% would rather see the impressions listed before the findings. About 27% want reports organized in an outline format.

When asked which anatomic structure should be routinely covered on a normal MR knee report, all respondents listed cruciate ligaments and menisci. About 31% included articular cartilage, and 13% requested collateral ligament descriptions.

One of the respondents commented that reports should be limited to abnormal structures. This opinion may be widely held among orthopedic surgeons, Lozano said.

For characterization of meniscal tears, 99% of orthopedic surgeons want the radiologist to describe the tear type. About 58% want information about the magnitude of displacement. Approximately 47% request information on the distinction between red and white zone involvement, and 17% want to know about stability in the description of meniscal tears.

For descriptions of ligament injuries, a majority of surgeons prefer that the radiologist describes their exact location and grade.

About 67% of respondents said they always read the radiologist's report. About 22% frequently read the report, usually to assure that the MR results agree with their physical examination. Eight percent sometimes read the report, and 2% never read it.

Of the orthopedic surgeons who read radiologists' reports, 50% read the entire report, 39% always read the impression and refer to the findings only if abnormalities are mentioned in the impression, and 8% read the impression section only.

Among the 31% of orthopedic surgeons who do not routinely read reports, their decision to review those reports is based on ensuring that MR finding correlate with the physical exam or to ensure that they do not miss additional pathology. Surgeons who do not routinely read the reports often interpret the images themselves.

Lozano mentioned deficiencies in radiology residents' education as a possible cause of reporting problems. She cited a survey that found that 86 radiology programs provide less than one hour of didactic instruction in radiologic reporting for their residents. In another survey, 93% of residents expressed dissatisfaction with their involvement in apprentice-style instruction in dictation style.

"Each report should be tailored to answer the specific clinical questions that are relevant to that case, but the preference of the referring physician needs to also be addressed," Lozano said.