Radiologists were twice as likely to recommend additional imaging in diagnostic reports in 2008 than they were in 1995, based on a single-center study in the November issue of Radiology.

Christopher L. Sistrom, MD, from the department of radiology at Massachusetts General Hospital in Boston, and colleagues sought to quantify the rates of recommendation for additional imaging in a large number of radiology reports of different modalities and to estimate the effects of 11 clinically relevant factors.

The researchers studied all diagnostic imaging exams (nearly six million) interpreted by radiologists between Jan. 1, 1995 through Dec. 31, 2008 at their facility. A total of 555 radiologists interpreted the studies, although 229 of them performed less than 1,000 each.

The investigators used natural language processing (NLP) technology to extract information about any recommendations from radiology report texts. Their analytic data set included three quantitative variables: the interpreting radiologist's experience; the year of study; and patient age. Categoric variables described patient location (inpatient, outpatient or emergency department), whether a resident dictated the case, patient gender, modality, body area studied, ordering service, radiologist's specialty division and whether the exam result was positive.

Of the 5.95 million exams reviewed, 10.54 percent had at least one recommendation for additional imaging in the dictated report detected by the NLP algorithm. The authors found that recommendations increased during the 13 years of study, with the unadjusted rate rising from roughly 6 percent to 12 percent.

After accounting for all other factors, Sistrom and colleagues reported that the odds of any one exam resulting in a recommendation for additional imaging increased by 2.16 times from 1995 to 2008. As radiologist experience increased, the odds of a recommendation for additional imaging decreased by about 15 percent per decade.

Studies that had positive findings were more likely to have a recommendation for additional imaging, the researchers found. The remaining categoric and quantitative variables also had significant effects on the tendency for a recommendation for additional imaging.

“This study shows that recommendations for additional imaging are made with different frequency depending on at least 11 relevant factors; any effort to measure and remediate variation among radiologists in recommendation rates must take these factors into account by using case-mix adjustment,” the authors concluded.