Suspected ureteral colic: primary helical CT versus selective helical CT after unenhanced radiography and sonography

Abstract

Objective: The purpose of this study was to compare the accuracy of unenhanced helical CT with combined sonography and unenhanced radiography in patients with acute flank pain suggestive of ureteral colic.

Subjects and methods: From January 1997 to December 1999, 181 consecutive patients with acute flank pain underwent unenhanced radiography, sonography, and unenhanced helical CT (protocol A). From January 2000 to December 2000, 96 consecutive patients arriving at the emergency department with acute flank pain were alternately submitted either to primary unenhanced helical CT (protocol B, 48 patients) or to unenhanced radiography and sonography with the addition of helical CT in unclear cases (protocol C, 48 patients).

Results: When compared with the diagnostic accuracy for ureterolithiasis of the combined sonography and radiography in the same group of subjects (protocol A), CT had a greater sensitivity (92% vs 77%), negative predictive value (87% vs 68%), and overall accuracy (94% vs 83%). Among patients who underwent primary CT (protocol B), we found three false-negatives (all with spontaneous stone passage) and no false-positives. Among patients initially examined with unenhanced radiography and sonography (protocol C), we found one false-positive (leading to patient admission and needless repeated radiographic and sonographic studies) and six false-negatives (all followed by an uncomplicated course and spontaneous passage); CT depicted four of these stones but did not result in change in treatment. Fourteen percent of the patients in protocol C required invasive treatment, but combined sonography and radiography showed stones and hydronephrosis in all these patients.

Conclusion: Unenhanced CT was the most accurate modality for determining the presence of ureterolithiasis. The combination of abdominal radiography and sonography, however, yielded comparable results with no clinically important misdiagnoses and thus can be used as an alternative when CT resources are limited.