Brigham and Women’s Hospital cut subsequent diagnostic imaging by 17 percent and subsequent CT imaging by 16 percent after implementing CD import to PACS among emergency department (ED) transfer patients, according to a study published in the May edition of Radiology. The authors characterized the CD import system as a partial success and noted the need for more reliable, health IT-based image transfer mechanisms.

Approximately 2.2 million patients are transferred between EDs annually in the U.S., and diagnostic imaging is often a key factor in the evaluation of such patients, according to Aaron Sodickson, MD, PhD, of the department of radiology at Brigham and Women’s Hospital in Boston, and colleagues.

The researchers designed the study to test their hypothesis that CD import into PACS could lead to a decrease in repeat exams. “Physicians often order repeat imaging examinations that are available on CD because of issues related to poor navigation tools or because of inoperable, inaccessible or lost CDs,” wrote Sodickson.

The two-part study analyzed the impact of a CD import system on 1,487 consecutive ED transfer patients from February 1, 2009 to August 31, 2009 and compared those data with a previous historical cohort of 254 patients, transferred between August 2007 and January 2008, prior to the CD import system.

Among the 1,487 ED transfers, images were successfully imported into PACS for 1,161 patients, comprising 2,221 exams. Sixty-eight percent of the exams included CT imaging.

When the researchers calculated imaging utilization within the first 24 hours after CD import, they determined that the successful-import group had a mean rate of 2.74 exams per patient, compared with 3.30 exams per patient in the failed-import group. The difference, a 17 percent decrease, reached statistical significance, offered Sodickson and colleagues.

Similarly, successful CD import translated into fewer subsequent CT scans, at 1.19 and 1.41 scans per patient in the successful- and failed-import groups, respectively, with this decrease of 16 percent also reaching statistical significance.

Narrowing the focus to CT scans obtained in the ED, physicians ordered 0.84 scans per successful-import patient and 0.99 scans per failed-import patient, according to the authors. Compared with the historical control group, which had a mean CT utilization rate of 1.18 CT scans per patient in the ED, the successful-import group had a 29 percent decrease in CT utilization.

The researchers noted that the decrease in subsequent imaging was largely attributable to decreased repeat imaging, as prior to the CD import system 29 percent of outside CT studies were repeated in the ED. They estimated that 35 percent of the repeat studies could have been avoided and calculated a predicted rate of repeat ED CT studies of 19 percent. The study showed a repeat rate of 16 percent for CT images ordered in the ED, thus showing “a greater than expected reduction in CT utilization.”

Sodickson and colleagues acknowledged several limitations to the study. Specifically, the decreases in utilization observed from the historical cohort and the current successful-import group may have been related to an overall interval decrease in imaging utilization. In addition, the three cohorts were not identical, which might have influenced subsequent imaging orders.

Finally, the researchers pointed out that the single-institution study may not be generalizable to all institutions. Sodickson wrote, “However, if our results are generalizable to the approximately 2.2 million ED transfer patients per year, our observed reductions in post-transfer CT utilization (from 1.41 to 1.19 CT scans per transfer patient) would be estimated to result in a decrease of 484,000 CT scans per year in the U.S.”

The authors continued, “[I]t is imperative that imaging be used judiciously and that any unnecessary imaging be eliminated whenever reasonably possible.”

Sodickson and colleagues characterized CT import into PACS as “an imperfect interim solution,” with a failure rate of 22 percent in the study. They suggested universal electronic medical records, image repositories or robust image transfer networks could address the shortcomings associated with the CD import process.