The utility of fetal renal pelvis dilation in distinguishing between obstructive and nonobstructive postnatal hydronephrosis

Abstract

BETWEEN OBSTRUCTIVE AND NONOBSTRUCTIVE POSTNATAL HYDRONEPHROSIS JEFFREY M. DICKE, DOUGLAS COPLEN, Washington University in St. Louis, Obstetrics & Gynecology, St. Louis, MO OBJECTIVE: To determine the efficacy of current thresholds for fetal renal pelvis dilation (RPD) in identifying significant postnatal urinary tract obstruction. STUDY DESIGN: A retrospective analysis of prospectively collected measurements of fetal RPD. The entire study population consists of 287 fetuses who underwent ultrasound (U/S) examination in the Department of Obstetrics and Gynecology with isolated RPD of $4 mm at 33 wks and known postnatal follow-up at St. Louis Children’s Hospital. All cases had postnatal U/S and evaluation by Ped Urology. RESULTS: Postnatal diagnoses included nonobstructive hydronephrosis (NOH) (40%), ureteropelvic junction obstruction (UPJO) (19%), megaureter (9%), reflux (7%),multicystic dysplastic kidney (7%), ureterocele (7%), normal (NL) (5%), and other (7%). One-way ANOVA indicated no difference in fetuses with normal postnatal ultrasound and those with NOH, defined as RPD that resolved or did not progress and did not require surgery. Receiver operating characteristic (ROC) analysis was performed comparing the maximum RPD in the NL/NOH (n = 128) group and the UPJO group (n = 53). ROC analysis revealed that the renal pelvis diameter that correctly classified the highest percentage (85%) of fetuses with respect to obstructive or nonobstructive postnatal hydronephrosis was 17 mm, with a sensitivity of 70% and specificity of 91%. Area under the ROC curve equals 0.859 with a standard error of 0.032 and a 95% confidence interval of 0.795 to 0.923. CONCLUSION: The current threshold for fetal pathologic RPD utilized at this institution was devised to ensure 100% sensitivity with less regard for specificity. When fetal RPD is associated with 100% sensitivity for detection of postnatal obstructive hydronephrosis, the false-positive rate is >90%. If maximum sensitivity is desired, consideration should be given to a single postnatal U/S rather than serial prenatal U/Ss in cases of mild ( < 10 mm) RPD.