Can a CDS tool help inexperienced physicians reduce imaging waste?

House staff physicians are particularly prone to overusing healthcare resources, including imaging. Researchers sought to find out if a clinical decision support (CDS) tool could help those clinicians with less experience reduce inappropriate imaging orders.

In the single-institution study, the team retrospectively analyzed more than 5,000 CT and MRI orders gathered before CDS implementation (April 2013), more than 7,000 orders from the first intervention period (post-CDS 1) and more than 14,000 order from the second intervention (post-CDS 2). Appropriate orders were based on the ACR’s Appropriateness Criteria and scored on a scale of 1-9, with higher scores signifying more appropriate orders.

Overall, implementing the CDS tool increased the scores of CT and MRI orders, but there was no significant difference between house staff and non-house staff, reported Jashvant Poeran, MD, PhD, of Mount Sinai’s Icahn School of Medicine in New York, and colleagues.

“Importantly, in this first study to examine the differential effects of CDS based on provider status, there is no significant difference in appropriateness of advanced imaging use between house staff and non–house staff providers,” Poeran, with Icahn’s Department of Orthopedic Surgery, and colleagues wrote. “This was seen in univariate analysis, as well as the quasiexperimental segmented regression analysis and the generalized linear model, which contributes to the robustness of our findings.”

The results, which Poeran et al. described as “unexpected,” go against current literature which, in at least one study cited by the researchers, found a significant association of increased advanced imaging orders during supervised house staff encounters compared to attending physician-only encounters. There may be several reasons for these surprising results, including the fact that the intuition’s radiology department involved in the study, offers lectures on appropriate imaging use at the start of each academic year. This may close the knowledge gap between more- and less-experienced physicians.

Specifically, the median appropriate use score prior to CDS implementation was 6.2. That score held steady during the post-CDS 1 period, which spanned April 2014 to March 2015. However, during the post-CDS 2 period, (April 2015 to June 2016) the mean score jumped to 6.7.

Although their results were surprising, the researchers wrote that they reflect the positive effect CDS tools can have in increasing the rate of appropriate imaging use.

“Although our study revealed no significant difference in the effect of CDS on house staff physicians compared with non–house staff physicians, the results likely reflect the multifactorial influences on ordering behavior in our health system,” the authors concluded. “CDS tools may thus serve as an equalizer in directing quality patient care from physicians across the training spectrum with varying familiarity with Appropriateness Criteria.

The full results of the study were published online Feb. 19 in the American Journal of Roentgenology.