The accuracy of noncontrast helical computed tomography versus intravenous pyelography in the diagnosis of suspected acute urolithiasis: a meta-analysis.

Abstract

We determine the accuracy of noncontrast helical computed tomography (NHCT) compared with that of intravenous pyelography (IVP) in diagnosing acute urolithiasis. Computerized searches of MEDLINE and EMBASE were combined with hand reviews of major journals and of articles from reference lists. Articles were assessed according to a priori criteria for inclusion. Study eligibility was independently assessed by 2 reviewers in a blinded fashion. Test results were combined and analysis of log-transformed data was conducted by using general linear models. No disagreement was found between the 2 investigators in terms of articles that met the inclusion criteria or between the results of the studies. Four studies involving a total of 296 patients met all of the a priori criteria. The pooled positive likelihood ratios (LR+) for NHCT and IVP are 23.15 (95% confidence interval [CI] 11.53 to 47.23) and 9.32 (95% CI 5.23 to 16.61), respectively. The pooled negative likelihood ratios (LR-) for NHCT and IVP are 0.05 (95% CI 0.02 to 0.15) and 0.33 (95% CI 0.23 to 0.48), respectively. The differences between NHCT and IVP were statistically significant for both LR+ (P =.046) and LR- (P =.013). Differences among trials were not statistically significant in either analysis (P =.125 for LR+; P =.114 for LR-). The studies analyzed consistently demonstrated NHCT to be superior to IVP in accurately diagnosing acute urolithiasis, and differences between the 2 tests for both LR+ and LR- were statistically significant.