Diagnostic accuracy of non-radiologist performed ultrasound for abdominal aortic aneurysm: systematic review and meta-analysis.

ABSTRACT

BACKGROUND: Ultrasonography is increasingly used by clinicians to identify abdominal aortic aneurysms (AAA). We performed a systematic review and meta-analysis comparing the accuracy of non-radiologist performed ultrasound (NRPUS) for AAA disease to the 'gold standard' of radiologist performed aortic imaging (RPI), intra-operative findings or postmortem findings. METHODS: Cochrane Library, MEDLINE, EMBASE, SCOPUS-V.4, trial registries, conference proceedings, and article reference lists were searched to identify studies comparing NRPUS with RPI as the reference standard. Data abstracted from eligible studies was used to generate 2 × 2 contingency tables allowing calculation of pooled sensitivity and specificity values. RESULTS: 11 studies (944 patients) evaluated NRPUS for AAA detection. NRPUS had a pooled sensitivity of 0.975 [95% confidence interval (CI), 0.942-0.992] for AAA detection and a pooled specificity of 0.989 (95% CI, 0.979-0.995). CONCLUSIONS: Non-radiologist performed ultrasound achieves acceptable sensitivity and specificity for both detection and measurement of AAA. There was no evidence of significant heterogeneity with respect to pooled sensitivity or specificity.