Ultrasonographic and computed tomography findings in renal suppurations: performance indicators and risks for...

The main purpose of the study was to identify and quantify clinical variables which predisposed incorrect diagnosis based on ultrasonography (US) and computed tomography (CT) findings in the patients with suppurative renal infections.

A retrospective review of radiological records of patients at 3 tertiary Serbian clinics of urology from 1999 to January 2006 was conducted. The series consisted of 49 women and 26 men, mean age = 56.7 years, ranging from 21 to 78. All patients underwent US examinations. Forty-one patients had undergone CT examinations, on 42 renal units (1 bilateral perirenal infection). The patients were categorized into 4 groups according to the type of suppurative infection: (a) unilocular renal or perirenal abscesses; (b) with pyonephrosis only; (c) with multiple renal or pyonephrosis infections along with extension on perirenal space, and (d) emphysematous infections.

Incorrect diagnosis based on US record only depended on the suppurative entity. In the presence of pyonephrosis only, the risk of incorrect diagnosis was reduced for 4 times [odds ratio (OR) = 0.25], and in presence of emphysematous infections the risk increased 19.5 times (OR = 19.49). Perinephric abscesses which were not seen on US were those associated with pyonephrosis, abscesses smaller than 6 cm and gas-forming abscesses. The diagnosis of abscesses considerably varied in the cases of unilocular purulent collections compared to complex purulent collections (chi(2) = 8.177, p = 0.004). Overall, the risk for incorrect diagnosis is about 14 times higher using US only (OR = 14.5), while CT reduced the risk of it for about 37 times (OR = 0.027).

CT was much more reliable than US to correctly identify the pathological entity of renal suppuration. Guiding treatment on US findings only seems hazardous as omission of some pathological processes is very likely.