Usefulness of abdominal flat plate radiographs in patients with suspected ureteral calculi

Abstract

A total of 120 plain abdominal radiographs (PARs) from intravenous pyelograms (IVPs) were presented without patient history in a random manner to a staff radiologist and staff emergency physician (EP). Scout films were used as the PAR to minimize calculus movement. Each reader examined each film for the presence of ureteral calculi. The readers localized all calculi and rated the likelihood of their presence. Fifty-one PARs from IVPs demonstrated ureteral calculus disease; 40 of these films demonstrated 41 radiopaque calculi, while 11 films demonstrated obstruction without radiopaque calculi. These readings were then matched to the known IVP results. Together the readers were correct on 50% of the predictions (radiologist, 51%; EP, 48%). The EP and the radiologist had similarly correct percentages on PARs from positive IVPs (33% and 36%, respectively) and on PARs from negative IVPs (60% and 64%). True-positive (33% and 25%), true-negative (76% and 71%), false-positive (27% and 35%), and false-negative (36% and 55%) rates were constructed for the radiologist and EP, respectively. Our study demonstrates that PARs are unreliable and do not appear to add useful information in the evaluation of suspected ureteral colic.