Over half of emergency CT requests are considered 'inadequate'
New data are shining a light on the issue of inappropriate CT referrals among emergency departments.
Published in Insights Into Imaging, a new study suggests that as much as 72% of CT requisitions from emergency department providers are considered inadequate according to the Reason for exam Imaging Reporting And Data System (RI-RADS). This system categorizes requisitions based on three key elements—impression, clinical information and diagnostic question—to guide clinicians in submitting appropriate imaging referrals with sufficient clinical reasoning. It was developed to help standardize the requisition process and provide radiologists with more clinical information, thus improving the diagnostic process.
“Radiologists now rely more on the requisitions from clinicians to be informed about the patients. Adequate clinical information can improve the protocol selection as well as the interpretation accuracy,” Jingyu Zhong, MD, with the department of imaging at Tongren Hospital, Shanghai, China, and colleagues explained. “However, clinicians often fail to provide sufficient clinical information for radiologists, and radiologists usually cannot take enough time to extract relevant clinical information for interpreting the images from the vast amount of data in the electronic medical records.”
Researchers recently assessed the clinical reasoning of emergency department clinicians within their organization using RI-RADS. To do this, they tasked a radiologist with retrospectively scoring all CT scan requests submitted by emergency providers at the organization throughout an entire year. The radiologist evaluated the extent to which the differential diagnosis on the requisition form matched the CT diagnosis on the final report and filed the referrals under the following categories: A (adequate), B (barely adequate), C (considerably limited), D (deficient) and X (no information).
Of the nearly 1,300 requisitions, 22.2% and 5.5% were considered adequate and barely adequate. The majority—nearly 66%—were labeled as considerably limited, while another 4.1% and 2.2% were deemed deficient or lacking in adequate information. Referrals from internal medicine providers were more likely to be assigned to category A, as were requests for imaging of the chest and contrast-enhanced studies, while requisitions of the abdomen and head were of lesser quality. All RI-RADS category A referrals displayed greater clinical reasoning.
One of the likely culprits for low quality requisitions pertains to time restraints and the reason for the referral. For example, many orders from the organization’s fever clinic and thoracic surgeons often were submitted to quickly rule out things like pneumonia or rib fractures. These requisitions do not typically provide much clinical information beyond ruling out these suspicions, making the requests lower in quality based on RI-RADS.
There are numerous variables that impact the quality of requisitions, the team added. They suggested that integrating large language models (LLMs) into EHRs where imaging orders are submitted could help resolve the issue.
“LLMs could make it possible for clinicians to enter their radiology requisitions into the electronic system and instantly receive the corresponding RI-RADS grade,” the authors suggested. “This immediate feedback would allow them to modify their requisition to ensure requisition completeness and improve communication among radiologists and clinicians.”
Read more about the team’s suggestions for improving referrals here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.