Agreement between handheld ultrasonography and computed tomography for abdominal aortic diameter measurement in unselected emergency department patients: a prospective feasibility study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background Abdominal aortic pathologies (aneurysm, rupture, or dissection) are among the most critical and life threatening diagnoses in patients presenting to the emergency department (ED) with abdominal or back pain. This study aimed to evaluate the agreement between bedside handheld ultrasonography (HHUS) measurements of abdominal aortic diameter performed by ED physicians and the reference standard, contrast enhanced computed tomography (CT), while also assessing inter rater reliability between clinicians. Methods This prospective, single centre, observational, single blind study involved 69 patients who underwent contrast enhanced abdominal CT or CT angiography for clinical indications in the ED. Aortic measurements in both transverse and longitudinal planes were performed using HHUS by an emergency medicine specialist and a resident. Reference CT measurements were recorded by a third blinded emergency medicine specialist. Data were analysed using the intraclass correlation coefficient (ICC), Bland–Altman analysis, and Spearman’s correlation. Results Sixty nine patients (mean age 52.29 ± 21.12 years; 52.2% female) were analysed. Agreement between the two operators was high at every segment (ICC 0.898–0.985) and 0.968 (95% CI 0.955–0.976) when pooled across segments. Agreement between HHUS and CT varied by segment: it was excellent at the distal aorta (ICC = 0.911; 95% CI 0.781–0.956), moderate in the longitudinal plane (ICC = 0.681; 95% CI 0.329–0.836) and at the proximal segment (ICC = 0.542; 95% CI 0.142–0.751), and poor at the mid aorta (ICC = 0.408; 95% CI 0.195–0.585); pooled across segments the ICC was 0.654 (95% CI 0.469–0.765). HHUS underestimated CT at every segment, with a mean bias of − 0.12 to − 0.36 cm, all within the pre specified 5 mm margin; the limits of agreement nevertheless extended beyond 5 mm at every segment except the distal aorta measured by the specialist. Two patients had an aortic diameter of 3 cm or more on CT and both were identified by HHUS. Conclusions In unselected ED patients, HHUS derived aortic diameters showed moderate overall agreement with CT, with the best agreement at the distal aorta and the poorest at the mid aorta. HHUS systematically underestimated CT, although every mean difference lay within a pre specified 5 mm margin. Inter operator reliability was high irrespective of seniority. These findings indicate that bedside aortic measurement with HHUS is feasible and reproducible rather than a substitute for definitive sizing, and that its diagnostic performance for aortic pathology, including any screening or triage role, requires dedicated accuracy studies.