Can the use of clinical decision-support (CDS) software significantly improve the appropriateness of imaging study orders? Yes, it can, particularly for physician trainees, according to research published online May 25 in the Journal of the American College of Radiology.
Healthcare providers who received feedback from CDS software on the appropriateness of their imaging requests ordered 50% fewer low-utility studies. What's more, their rate of ordering appropriate exams rose to more than 80%, noted researchers from the University of Virginia.
Physician trainees -- residents and fellows -- benefited the most from the CDS software, wrote lead author Dr. Timothy Huber and colleagues (JACR, May 25, 2018).
ACR Select
In 2014, the institution adopted version 6 of the ACR Select CDS platform (National Decision Support) and integrated it into its electronic health record software (Epic Systems). For the first six months, the CDS software was run only in the background, generating appropriateness scores for orders but not displaying this information to the providers. After this initial period, ordering providers were given real-time feedback on the utility of their orders.
After referring providers began receiving feedback when entering orders, the frequency of low-utility studies dropped significantly, while the frequency of indicated studies increased significantly.
| Frequency of imaging appropriateness score categorization | ||
| Appropriateness | Period without CDS feedback to provider (6 months) | Period with CDS feedback to provider (24 months) |
| Low utility | 11% | 5.4% |
| Marginal | 24.5% | 12.6% |
| Appropriate | 64.5% | 82% |
Physician trainees derived the greatest benefit from receiving CDS feedback. They had the largest decrease in the frequency of low-utility orders as well as the highest increase in the frequency of indicated orders, according to the researchers.
| Frequency of low-utility studies by ordering provider type | ||
| Ordering provider | Period without CDS feedback to provider (6 months) | Period with CDS feedback to provider (24 months) |
| Physician trainee (residents and fellows) | 10.8% | 4.8% |
| Midlevel provider (physician assistants and nurse practitioners) | 13.8% | 10.2% |
| Attending physician | 10% | 6.7% |
On a modality basis, the researchers found significant decreases in the proportion of low-utility studies and increases in the proportion of indicated studies for CT, MRI, and ultrasound. For reasons that are unclear, there was no improvement with nuclear medicine studies, however.
Huber and colleagues speculated that the nuclear medicine content in the CDS application at the time may not have been robust enough to make a difference. The software has since been updated to version 13, though, and includes input from the Society of Nuclear Medicine and Molecular Imaging (SNMMI).
"Locally developed or institution-specific CDS software that has been implemented at several institutions across the United States has demonstrated that ordering patterns can change, with reductions in the number of low-utility examinations ordered after implementing such systems," they wrote. "Our study demonstrates that similar results can be achieved with commercially available CDS and that the impact of these efforts can vary by provider type and imaging modality."
