The most recent Sentinel Event Alert¹ issued by the Joint Commission formally put the medical world on notice that the expanding use of diagnostic imaging will require more stringent oversight to ensure patient safety. Whether that oversight will be self-imposed or enforced from without, the Joint Commission recommends that practitioners be held to more clearly defined standards concerning the overall radiation body burden of their patients, and it lays out 21 specific recommendations.
When the ACR® gently chided the Joint Commission for some of the wording in its alert, including mixing up the names of its registries, there was clearly a bit of bristle in the response.² This is an issue on which the college has been proactive, both in launching the Image Gently campaigns and in adding a Dose Index Registry (DIR) to the other active registries that compose the National Radiology Data Registry (NRDR), launched in 2008.
The DIR was officially launched on March 14, 2011, as the newest component of the NRDR. In addition to the DIR, the NRDR warehouses seven other specific registries—covering oncologic PET, CT colonography, mammography, IV contrast extravasation, night coverage, pediatric CT quality improvement, and general radiology improvement—that help facilities benchmark outcomes and process-of-care measures and develop quality-improvement programs, according to the ACR.
Despite the fact that the DIR has been in operation for just seven months, its chair, Richard Morin, PhD, says that the seeds for the registry were planted years ago, when he chaired the ACR Commission on Medical Physics. The commission came up with the idea of a database that would capture dose information, and the ACR funded the development of software that would automatically extract the amount of radiation associated with a study from the scanner and produce a structured dose report.
A pilot involving one vendor resulted in proof of the concept, so other CT manufacturers were invited to participate. “We repeated the pilot with a much larger audience and demonstrated that this was a viable idea,” Morin recalls.
Already, the DIR has gathered information from 214 participating institutions, representing more than 150,000 exams and 250,000 individual series of CT studies. The largest single group participant contributes reports from 30 CT devices. Eschewing manual data entry, the DIR software automatically extracts imaging information from DICOM headers using the Integrating the Healthcare Enterprise Radiation Exposure Monitoring profile. By compiling those data, it can help institutions establish guidance on their dose indices for specific exams, relative to national benchmarks.
“It’s a way for the institution to see where it stands compared to everybody else,” Morin says. “We designed it, from the beginning, to be operator independent—no human interaction necessary.”
Of course, reports generated by the system are only as good as the data that have been entered. There’s no selection bias, Morin says, because every exam performed, on every participating machine, is logged in the system. The democracy of this process leaves its mark in two of the biggest challenges involved with DIR trending: nonstandardized naming conventions for CT studies and differences in patient size.
“Sometimes, it can be very difficult to try to match up your exam with someone else’s,” Morin says, particularly when seeking apples-to-apples comparisons of studies of the chest, abdomen, and pelvis. Likewise, body-habitus differences can cloud the accuracy of registry values because CT-beam modulation means that “big, thick people have higher values than small, thin people,” Morin says. “In our pilot, we saw some rather large variances—in some exams, a factor of four to six median values.”
Some larger institutions also have shown great internal variances among common studies because their protocols haven’t been harmonized across different devices and software iterations. “We hope, over time, that people will begin to home in on the optimal radiation for an exam. The community has responded very well; the staff is talking to people daily who want to sign up,” Morin says.
George Segall, MD, president of the Society of Nuclear Medicine, says that the DIR represents a natural evolution of medical informatics and can be a helpful tool for physicians, but that it should not serve as a standard of judgment—or the basis upon which a determination to perform