Utility of the plain abdominal radiograph for diagnosing ureteral calculi

Abstract

Our study was designed to measure sensitivity, specificity, predictive value, and accuracy of the plain abdominal radiograph (PAR) for diagnosing ureteral calculi (UC) in our emergency department (ED), and to determine if obtaining this test in the ED provides clinically useful diagnostic information. Records of 206 adult patients (147 men, 59 women) who underwent emergency excretory urography as part of an ED evaluation for possible ureteral stones were reviewed. A retrospective scoring index was devised to estimate the clinical probability of a ureteral stone for each patient. All patients had a PAR shortly before the excretory urogram (EXU) was performed. Each PAR was separated from the corresponding EXU, and both were independently reviewed for radiographic evidence of ureterolithiasis. The results of the PAR and EXU interpretations were compared. There were 92 true-positive PARs (45%), 19 false positives (9%), and 39 true negatives (19%), yielding a sensitivity of 62% and a specificity of 67%. In patients with a high estimated clinical probability of a ureteral calculus, the PAR had a positive predictive value of 86% and a negative predictive value of 22%. In the remaining patients, the positive and negative predictive values of the PAR were 69% and 72%, respectively. Each patient's clinical index score was compared to his or her EXU results. For detecting UC, the sensitivity of our scoring index was 82%, and the specificity was 51%; the positive predictive value of the scoring index was 82%, and the negative predictive value was 52%.(ABSTRACT TRUNCATED AT 250 WORDS)