Operative versus nonoperative management of ureteropelvic junction obstruction in children.

OBJECTIVES: To describe and validate our strategy for treating patients with ureteropelvic junction obstruction operatively or nonoperatively according to ultrasonography and nuclear renal scan findings.
METHODS: A retrospective analysis of 243 patients from 1999 to 2006 with grade 3-4 hydronephrosis was performed. Depending on the grade of hydronephrosis with parenchymal thinning on ultrasonography and renal function on the nuclear renal scan, patients were treated with immediate pyeloplasty, pyeloplasty after a period of observation, or observation only.
RESULTS: Of 243 patients, 116 were found to have UPJO as determined by a half-life >20 minutes. The mean follow-up was 24.0 months (range 3-69). Immediate pyeloplasty was performed in 32 children, and 84 were treated conservatively. Crossover from observation to surgery occurred in 47 children. In the immediate pyeloplasty group, the mean pre- and postoperative differential function was 30.4% and 38.8%, respectively (P < .0001). In the observation-only group (n = 37), the initial mean renal function was 41.4% and stayed stable throughout the follow-up period, at a mean of 43.2% (P = .2764). In the delayed pyeloplasty group (n = 47), the initial mean renal function was 35.9% and increased to a mean of 41.6% after intervention (P = .0003). The median improvement of hydronephrosis on ultrasonography for those who underwent immediate surgery from before to after the intervention was from grade 4 to grade 3 (P < .0001). For those not undergoing surgery, it improved from grade 4 to grade 2.25 (P = .0026) and for those who underwent delayed surgery, from grade 4 to 3 (P = .0003).
CONCLUSIONS: According to our findings, the serial ultrasonography findings and initial renal function on nuclear renal scan are better indicators than the half-life alone for determining whether pyeloplasty is indicated.