RSNA: Hard-stop decision support reduces low-yield imaging orders
CHICAGO--Modification of the order entry workflow can change physician behavior and result in fewer low-yield imaging orders, shared Vartan Vartanians, MD, clinical research associate at Massachusetts General Hospital (MGH) of Boston during a scientific session at the Radiological Society of North American (RSNA) meeting.

Vartanians summarized a trial designed to evaluate the effect of preventing non-clinician support staff from completing orders for outpatient CT, MRI, and nuclear medicine studies that received initial low utility decision support scores. In 2007, MGH launched HSOR (hard stop on red) as a new feature in its existing radiology order entry system.

The MGH order entry system incorporates a numerical stoplight that assigns a value from 1 to 9 to provide information about the utility of all ordered imaging studies. HSOR prevents non-clinician users from ordering studies on low-utility exams with a score in the 1 to 3 range. In these cases, the responsible physician must log in to order or cancel the exam or change the indication.

The trial compared imaging studies ordered in the six months preceding HSOR (April to December 2006) to studies ordered six months post-HSOR (April to December 2007).

The percentage of orders placed by clinicians (vs. support staff) increased from 26 percent to 54 percent, said Vartanians. In addition, the percentage of clinician-based orders that were canceled after the system assigned a red score increased from 24 percent to 38.5 percent. The percentage of red, or low-yield studies, proceeding from order to completion dropped from 61 percent to 47 percent.

Results for studies ordered by support staff were equally promising. Stopped studies rose from 8 percent to 57 percent. After the HSOR launch, support staff proceeded with a mere 7 percent of low-utility orders, compared to 82 percent prior to HSOR.

Overall, the percent of low-yield exams ordered dropped from 5 percent to less than 2 percent, and cancelled exams increased from 11 percent to 44 percent.

This modest change in the business logic of the order entry system resulted in a substantial decrease in the rate of low-utility imaging, and a large increase in the percentage of tests personally ordered by clinicians. “We believe that these results demonstrate the importance of placing the decision support information in front of the responsible decision maker, [which can increase the appropriateness of outpatient imaging]," summed Vartanians.