A poster, titled “Outpatient Facility-Based Order Variation in Combination Imaging of the Abdomen” was presented at the AcademyHealth 2019 Annual Research Meeting held in Washington, D.C. from June 2-4.
The study’s lead author and presenter, Adam C. Powell, PhD, and his team, sought to examine the variation between healthcare facilities in ordering and receiving orders for abdominal computed tomography (CT) exams combining images taken with and without contrast as a proportion of overall abdominal CT orders. It is noted that although combination exams are sometimes clinically necessary, they expose patients to additional radiation. Identification of outlier facilities may be useful in targeting quality improvement initiatives.
The study analyzed orders for abdominal images that had been authorized through the prior authorization process of a non-denial prior authorization company. The study population consisted of all authorized outpatient orders for abdominal CT, issued from 2013 to 2017, pertaining to people with commercial and Medicare Advantage health plans, extracted from the database of a prior authorization company.
Analysis was performed in two ways: orders were grouped by ordering facility and by designated rendering facility. For each facility, the number of orders authorized for abdominal CT with contrast, without contrast, and both concurrently were counted. The ratio of combination orders to overall orders was calculated for each facility, and the mean of the ratios across all facilities was also calculated.
There were 1,052,294 orders meeting inclusion criteria, and were 16,972 ordering and 8,203 rendering facilities with >10 orders. More orders pertained to low-volume ordering facilities than rendering facilities. The mean ratio was 0.306 for ordering facilities and 0.311 for rendering facilities. There was a significant difference (P<.001) between the ratio for orders pertaining to hospital (0.260) versus non-hospital (0.309) ordering facilities, and also (P<.001) between the ratio for orders pertaining to hospital (0.292) versus non-hospital (0.334) rendering facilities. Rural ordering facilities had a lower ratio, 0.294, than urban facilities, 0.308 (P=0.005). Rural rendering facilities likewise had a lower ratio, 0.289, than urban facilities, 0.316 (P<.001).
Based upon these findings, substantial variation between facilities was observed. The distribution of the abdominal combination imaging ratios was positively skewed when orders were grouped according to ordering or rendering facility. The ratios decreased over time, and varied significantly according to urbanicity and facility type. Orders pertaining to hospitals and rural areas were significantly less likely to be combination orders.
In response to the findings, Dr. Powell remarked, “There is large base of research which suggests that different sites of service have differing practice patterns. Our study helped characterize the extent of variation that has occurred in the use of combination abdominal imaging across facilities.”
For information pertaining to the AcademyHealth 2019 Annual Research Meeting, visit: https://www.academyhealth.org/events/site/2019-annual-research-meeting