Accuracy of Ultrasound in Identifying Renal Scarring as Compared to DMSA Scan

Abstract Objective To assess the accuracy of renal ultrasound (RUS) in detecting renal scarring (RS). Methods All initial DMSA scans performed from 2006 - 2009 for history of urinary tract infection (UTI) or vesicoureteral reflux (VUR) in patients under 14 years old were identified, and clinical history obtained via chart review. Patients who had RUS within 4 months of DMSA scan and no documented UTI during that interval were included. Decreased uptake of tracer associated with loss of contours or cortical thinning defined a positive DMSA study. Increased echogenicity/dysplasia, cortical thinning, atrophic kidney and/or abnormal corticomedullary differentiation defined a positive RUS. The sensitivity and specificity of RUS in identifying RS were calculated using DMSA scan as the gold standard. Results A total of 144 patients had initial DMSA scans performed for UTI or VUR, with a RUS within 4 months, and no UTI between the two studies. 95/144 (66%) had RS on DMSA and 49/144 (34%) did not. Patients with or without RS on DMSA were not different in gender (p=0.073), age (p=0.432), insurance (p=1.000) or VUR grade (p=0.132). Only 39/144 (27.1%) patients had positive RUS. The sensitivity of RUS for RS was 35.8% and the specificity was 89.8%, leading to an accuracy of 54.2% (95%CI; 45.7-62.5%, p=0.999). Conclusions RUS demonstrated poor sensitivity for RS visualized on DMSA scan. This suggests that RUS is a poor screening test for RS or indicators of future renal scar. A normal ultrasound does not rule out RS or risk of future renal scar. Specificity of RUS was excellent.