Agreement between emergency medicine residents and a radiologist in ultrasonographic detection and grading of hydronephrosis in patients with suspected renal colic: a prospective study.

Abstract

Background: Suspected renal colic is a frequent emergency department presentation. Bedside hydronephrosis assessment may support early triage, consultation, and disposition decisions.

Objectives: This study evaluated whether emergency medicine residents can reliably detect hydronephrosis and classify its severity using bedside urinary ultrasonography in adults with suspected renal colic.

Methods: This prospective observational agreement study included 272 adults with suspected renal colic. Bedside urinary ultrasonography was performed by 16 emergency medicine residents with at least two years of residency training and structured ultrasonography education. Before enrollment, residents completed standardized theoretical, hands-on, and supervised radiologist-guided renal ultrasonography training focused on hydronephrosis assessment. Radiologist-performed ultrasonography was completed within 2 h of emergency department admission by an experienced radiologist blinded to residents' findings. Both groups used the same severity categories: absent, mild, moderate, and severe. Diagnostic performance, agreement, and severity-level concordance were analyzed at the patient level. Computed tomography (CT)-based analyses were exploratory and evaluated CT-detected hydronephrosis rather than stone detection.

Results: Hydronephrosis was detected in 81 patients by both residents and the radiologist. Overall agreement was 91.18%, with a Cohen's kappa of 0.789. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 85.19%, 93.72%, 85.19%, 93.72%, and 91.18%, respectively. Exact severity-level agreement was 87.87%, with a linear weighted kappa of 0.813. In the CT subgroup, residents' ultrasonography for CT-detected hydronephrosis showed 81.63% sensitivity, 86.05% specificity, and 83.70% accuracy.

Conclusion: Emergency medicine residents with structured ultrasonography training demonstrated substantial agreement with radiologist-performed ultrasonography for detecting hydronephrosis and strong agreement for severity classification. These findings apply to trained residents and may not be generalizable to clinicians without similar training or supervision.