Evaluating and communicating probability of acute cholecystitis consistently using a data-derived risk stratification algorithm synthesizing ultrasound and clinical parameters
摘要
The sonographic diagnosis of acute cholecystitis presents challenges. In our practice, we enhance diagnostic accuracy by determining a patient’s risk of acute cholecystitis using four non-image parameters and five imaging parameters. The non-image risk assessment is based on patient age, sex, leukocytosis, and the presence of a sonographic Murphy sign. The imaging risk stratification is derived from evaluating gallbladder (GB) distention, GB wall thickness, GB contents, pericholecystic irregular collections, and hepatic artery peak systolic velocity (HAv). By applying a standardized scoring framework, patients are stratified into one of four diagnostic categories for acute cholecystitis: (1) practically excluded (< 1% probability); (2) reduced risk (< 10% probability); (3) elevated risk (25–30% probability); (4) substantially elevated risk (with three subgroups having 50%, 75%, and 90% probability). This review outlines the methodology of our approach, provides supporting data from published cohorts, and explains the macro-enabled Excel tool we use to streamline the analysis. This tool generates standardized report and impression statements that can be directly incorporated into radiology reporting templates. The approach promotes consistency in reporting, particularly amongst trainees formulating preliminary interpretations, and offers consistent evidence-based probability estimates to emergency department physicians and surgeons for clinical decision-making using a framework that eliminates indeterminate assessments.
Graphical Abstract