Additional imaging often recommended, seldom pursued, in PET/CT reports

Nuclear medicine physicians and radiologists recommended additional imaging in approximately one-third of PET/CT reports, but more than half of these recommendations were unnecessary, according to a study published in the January issue of Clinical Radiology. Ordering clinicians followed recommendations for further imaging in less than one-third of the reports.

Previous research has indicated a nearly twofold increase in recommendations for additional imaging (RAI) since 1995. Estimates of RAI incidence range from 12 to 31 percent. PET/CT may be particularly susceptible to this phenomenon as both components of the exam can result in findings that are difficult to characterize completely.

A.B. Shinagare, MD, from the department of radiology at Brigham and Women’s Hospital in Boston, and colleagues devised a retrospective study to determine the incidence, appropriateness and outcomes of recommendations for RAI in oncologic PET/CT reports.  

Two blinded reviewers re-read the first 250 PET/CT reports acquired in 2008 to identify RAI. They characterized RAI as unnecessary if the results met any of the following criteria:

• PET/CT provided sufficient characterization of the finding for diagnosis or appropriate management;
• The finding lacked specific concerning features and was stable when compared with prior imaging performed at least 12 months earlier; or
• The finding, whether suspicious or not, would not be expected to impact patient outcome or management in the context of the disease.

The reviewers checked each patient’s medical record and imaging studies during a two-year follow-up period to confirm clinical significance.

Shinagare and colleagues reported 84 RAI for 88 PET/CT findings in 29.6 percent of the PET/CT reports. The reviewers characterized 51.2 percent of RAI as unnecessary, citing sufficient characterization and guidance by PET/CT in 62.8 percent of the cases. Further lesion characterization would not have changed clinical management or outcome for the 25.6 percent of patients with a finding of widespread metastatic disease. The PET/CT finding was stable based on prior imaging in 11.6 percent of patients.

Referring clinicians did not follow the RAI in 69 percent of the reports. “Ordering clinicians are more likely to know the entire patient history and other key clinical information, whereas radiologists and nuclear medicine physicians often are not provided or do not have access to all of these data,” wrote Shinagare and colleagues.

Nevertheless, imagers and clinicians should share responsibility for reducing unnecessary additional imaging exams, according to the researchers. Radiologists and nuclear medicine physicians can aid clinicians by sharing their knowledge of how additional imaging may or may not provide clinically useful information.

“Perhaps the most important observation was that no adverse impact on patient management or outcome was associated with the 51.2 percent of RAI deemed unnecessary by the readers, or with the 69 percent of RAI not followed by referring clinicians,” they wrote.

The researchers did not find statistically significant differences in RAI incidence or appropriateness among various PET/CT readers, nor did they note statistically significant differences in RAI incidence based on primary cancer type or organ involved.

Shinagare and colleagues offered a two-pronged strategy for reducing RAI without compromising patient care and outcomes: Consider the patient’s clinical circumstances and the anticipated relevance of the additional imaging results in the decision-making process.