ORLANDO, Fla.—What does it take to improve performance? Some of the simplest strategies include competition and observation. Both angles apply to the American College of Radiology (ACR) dose index registry.
Registries are an important quality improvement tool that allow institutions to compare their performance with others, Richard L. Morin, PhD, of the Mayo Clinic Jacksonville in Fla., explained during a presentation at the Society for Imaging Informatics in Medicine (SIIM).
Since May 2011, more than 450 facilities have registered for the dose index registry and 200 are active participants. The registry transmits dose data from the scanner to the DICOM structured report. ACR software anonymizes the data and then it is uploaded to the ACR database. Facilities receive semi-annual reports, which can be sliced and diced to compare institutional, local, regional and national trends.
The most common use is as a discussion point in monthly CT meetings among radiologists, technologists and physicists, said Morin. “They’ll review the data related to a specific protocol and decide if they want consider any changes.”
It is critical to examine outliers and look beyond the data. It isn’t uncommon, for example, for facilities to observe a group of patients who received what it appear to be 10 times to average dose index for a non-contrast head CT. Typically, however, the exam had been changed to a perfusion scan. If the technologist does not close the first scan, it is incorrectly recorded as a non-contrast head CT.
In a little more than one year, the registry has overcome other growing pains as well. For example, it has managed to accommodate (but not standardize) the more than 1,400 terms for head CT, and it captures data from both new and old scanners.
Such data also may provide useful talking points to share with concerned patients, offered Morin. Given the national climate and the renewed hype in the wake of the June 7 study published in The Lancet, such quality improvement initiatives could prove valuable on multiple levels.