Emergency ultrasound and urinalysis in the evaluation of flank pain.

OBJECTIVES To determine the sensitivity and specificity of limited emergency ultrasonography of the kidney in diagnosing renal colic. METHODS This was a prospective observational trial from December 2001 to December 2003 at a suburban emergency department. Patients who presented with flank pain suspicious for renal colic were enrolled. Exclusion criteria included fever, trauma, known current kidney stone, unstable vital signs, and inability to provide consent. All patients underwent sequential emergency ultrasonography and computed tomography of the kidneys and bladder. Data were analyzed using chi-square analysis. The primary outcome was the sensitivity and specificity of ultrasonography. Results were also stratified for presence of hematuria. RESULTS Fifty-eight of the 104 patients enrolled in the study were diagnosed with renal colic. The overall sensitivity and specificity of bedside ultrasonography for the detection of hydronephrosis were 86.8 (95% confidence interval [CI] = 78.8 to 92.3) and 82.4 (95% CI = 74.1 to 88.1), respectively. In patients with hematuria, hydronephrosis by emergency ultrasonography demonstrated a sensitivity and specificity of 87.8 (95% CI = 80.3 to 92.5) and 84.8 (95% CI = 73.7 to 91.9), respectively. In 55 of the cases, the initial computed tomograph was read by a resident and later re-read by an attending physician. Using the reading of the attending physician as the criterion standard resulted in a sensitivity and specificity of 83.3 (95% CI = 73.2 to 88.0) and 92.0 (95% CI = 79.9 to 97.6), respectively. CONCLUSIONS Emergency ultrasonography of the kidneys shows very good sensitivity and specificity for diagnosing renal colic in patients with flank pain and hematuria.