Implementing a clinical decision support system (CDSS) had little effect on reducing inappropriate imaging orders by doctors in university hospitals, according to a study published February 10 in JAMA.
The finding is from a clinical trial involving 26 departments at three German academic medical centers that compared the proportions of inappropriate imaging requests between departments that used a CDSS with those that did not.
“At the completion of the current study, all participating departments chose to discontinue using the CDSS,” noted first author Stijntje Dijk, MD, a PhD candidate at Erasmus MC University Medical Center in Rotterdam, the Netherlands, and colleagues.
CDSSs are computer-based software applications designed to provide physicians with evidence-based recommendations and insights to assist them in making informed clinical decisions. Despite their promising potential, CDSS implementation faces challenges, the researchers noted, and hence they sought to understand the real-world effects of integrating one into the clinical workflow at their hospitals.
In this study, the CDSS was originally based on the American College of Radiology’s appropriateness criteria, which covers more than 15,000 appropriate use criteria across all modalities. The primary outcome measure was the proportion of inappropriate imaging requests made per department before and after implementation.
Between December 2021 and June 2024, 13 departments at three hospitals implemented the CDSS intervention and 13 departments did not. The CDSS provided ordering physicians with information as to whether their imaging requests were appropriate, appropriate under certain conditions, or inappropriate. In addition, alternative diagnostic tests, including the corresponding appropriateness score, were suggested by the system, after which physicians could choose to modify their imaging requests.
Prior to the implementation of the CDSS, there were 21,625 imaging requests from the departments that didn’t deploy it, of which 1,367 (6.3%) were categorized as inappropriate, and 13,338 imaging requests from departments that did deploy it, of which 1,007 (7.6%) were categorized as inappropriate.
After implementation, there were 10,055 imaging requests from the departments that didn’t use it, 518 (5.2%) of which were categorized as inappropriate, and 7,206 imaging requests from those that did, of which 461 (6.4%) were categorized as inappropriate.
According to the analysis, the departments that used the CDSS showed a similar reduction (mean difference, approximately -0.5%) in inappropriate imaging requests compared with those that did not (mean difference, -1.8%), with a difference-in-differences value of 1.3 percentage points, which was not statistically significant, the researchers reported.
“The clinical decision support system did not reduce the number of inappropriate imaging requests ordered by physicians in academic hospital settings,” the group wrote.
Notably, few physicians changed their initial imaging requests that were categorized as inappropriate, and, in rare cases, even switched their order to less appropriate diagnostic tests, the researchers added. Users expressed that the system lacked choice in indications, and the selection process was time-consuming and inefficient, they wrote.
Ultimately, when a study fails to show an effect on the primary outcome, several factors must be considered before concluding it is a negative trial, according to the group. For instance, because the current trial was conducted at academic hospitals, there was a low baseline percentage of inappropriate imaging requests, which left little room for reduction, they suggested.
“Other settings may demonstrate different results when the number of baseline inappropriate imaging requests is higher,” the group concluded.
The full study is available here.
