Imaging requests for acute abdominal pain in the emergency department: a retrospective cross-sectional study

To evaluate whether the appropriateness of imaging requests (AIR) for acute abdominal pain (AAP) in two private hospitals in Santiago (2023) was associated with higher diagnostic yield, and to explore predictors through a multivariable model.

A retrospective cross-sectional study included patients aged ≥ 15 years presenting with AAP who underwent US, CT, or MRI. AIR was classified according to the ACR Appropriateness Criteria, and radiology reports were categorized as confirmatory or normal. A multivariable logistic regression model, incorporating interactions between AIR, imaging modality, and age, was applied to identify predictors of diagnostic yield.

A total of 189 imaging studies were analyzed (75 US, 62 CT, 52 MRI). Overall, 57.7% of requests were appropriate. Confirmatory findings occurred in 66.1% of cases, significantly more frequent in appropriate requests (84.4%) than in inappropriate ones (41.3%; p < 0.01). Inappropriateness reduced diagnostic yield from 78.3% to 17.4% in US and from 91.5% to 25.9% in CT, with no significant difference for MRI (86% vs. 81.3%). Older age was independently associated with higher probabilities of confirmatory findings. Epigastric pain and acute gastroenteritis were disproportionately linked to overuse, while hepatobiliary conditions showed predominantly appropriate use. The model demonstrated high discriminatory performance (AUC = 0.88).

Adherence to ACR criteria was strongly associated with improved diagnostic yield in AAP, particularly for CT and US. Age and local epidemiology, including high biliary disease prevalence, influenced modality performance. These findings support context-sensitive decision-support tools and prospective studies to refine imaging strategies and enhance patient safety in emergency care.