Ureteral calculi in patients with flank pain: correlation of plain radiography with unenhanced helical CT.

PURPOSE: To determine the sensitivity and specificity of plain radiography for the detection of ureteral calculi with use of unenhanced helical computed tomography (CT) as the standard of reference. MATERIALS AND METHODS: Plain radiographs and helical CT scans of 178 patients with acute flank pain were reviewed retrospectively. Three interpretations of plain radiographs were used: (a) Original reading. This was the report made at the time of the patient's evaluation before the patient underwent CT. (b) Blinded retrospective reading. Each plain radiograph was interpreted without knowledge of the CT findings. (c) Unblinded retrospective reading. The plain radiograph of each patient whose CT scan showed a stone was reviewed with the CT scan. RESULTS: The original reading had a sensitivity of 45% and a specificity of 77% for the detection of ureteral calculi. The blinded retrospective reading had a sensitivity of 59% and a specificity of 71%. The unblinded retrospective reading had a sensitivity of 59% (95% confidence interval: 47%, 70%). CONCLUSION: Plain radiography is of limited value for aiding the diagnosis of ureteral stones. All patients with acute flank pain for whom radiologic imaging is recommended can directly undergo unenhanced helical CT; plain radiographs need not be obtained first.