Emergency department (ED) physicians are inconsistent in how they order imaging exams, and this variation could lead to unnecessary imaging studies and unnecessary patient exposure to radiation, according to a study published in the September issue of the American Journal of Roentgenology.
The results show that perhaps ED physicians could use more education to standardize how they order imaging studies, wrote a team led by Vladimir Valchinov, PhD, of the Harvard Medical School in Boston.
"We found that significant variation in the use of low-cost and high-cost imaging in the ED was attributable to the ordering physicians, after controlling for multiple patient-, visit-, and physician-level covariates," the group wrote (AJR, September 2019, Vol. 213:3, pp. 637-643). "This ... variability in imaging utilization among ED physicians validates ongoing efforts to reduce [it] via the development, monitoring, and improvement of initiatives using well-defined imaging quality measure in the ED."
Many professional organizations have developed guidelines intended to reduce unnecessary imaging, including the American Board of Internal Medicine's Choosing Wisely campaign. These initiatives are based on the concept that the test-ordering behaviors of individual physicians contribute to variations in orders for imaging exams. But these variations may be attributable to other factors, according to Valchinov and colleagues.
To determine what factors influence trends in low- and high-cost imaging (MRI, CT, and nuclear medicine) in the ED, the researchers conducted a study that included a review of imaging exams ordered for 56,793 patients in a level I adult trauma emergency department between April 2013 to March 2014, controlling for patient, visit, and physician characteristics. During the study period, 51 attending physicians staffed the ED.
Of the total number of patients who presented to the ED during the study, 49.5% underwent imaging; of these, 38.2% underwent low-cost imaging and 21.9% underwent high-cost imaging.
The group found the following factors were associated with higher numbers of imaging orders:
| Factors that influence ED orders for imaging, by odds ratio | ||
| Low-cost imaging | High-cost imaging | |
| Patient characteristic | ||
| Female | Reference | Reference |
| Male | 1.08 | 1.08 |
| Patient age | ||
| 65 to 75 | 0.82 | 0.87 |
| No. of secondary diagnoses | ||
| 4 to 7 | 5.76 | 3.67 |
| Time of arrival in ED | ||
| Monday through Friday (night) | 1.03 | 1.31 |
| Saturday or Sunday (evening) | 1.17 | 1.28 |
| Saturday or Sunday (night) | 1.01 | 1.55 |
| No. of ED arrivals within previous 4 hours | ||
| > 40 | 1.01 | 1.12 |
| No. of years of physician experience | ||
| 4 to 7 | 1.21 | 1.03 |
| Annual physician volume of visits | ||
| 250 to 700 | 1.00 | 1.13 |
Of the 51 ED physicians, 47.1% had a higher-than-average odds ratio of ordering high-cost imaging, and 49% had a higher than average odds ratio for ordering low-cost imaging, the group found. When the researchers adjusted for physician variability, they found a 10.9% difference among physicians for low-cost imaging and a 14% difference for high-cost imaging.
"Despite significantly different patients, shifts, and physicians, physician decision-making appears to play a large role in the variation of imaging ordering," the group concluded.
