The risks of omitting routine postoperative imaging after insertion of new ventricular shunts in children: a 10-year retrospective cohort study.

Abstract

Objective: Ventriculoperitoneal shunt insertion is the primary treatment for pediatric hydrocephalus and is one of the most common procedures performed in pediatric neurosurgery. However, the value of routine postoperative imaging remains controversial. Although some evidence suggests that routine postoperative imaging provides a low yield after shunt revision, its role following initial shunt insertion in children has not been reported. Additionally, the role of routine postoperative imaging as a baseline study in altering the rate of longer-term shunt revision rates remains unexplored.

Methods: A retrospective review was conducted using a prospectively maintained surgical database, encompassing 10 years of consecutive cases from a quaternary pediatric neurosurgical center serving a population of 5.3 million. Statewide electronic medical records including imaging and mortality data were reviewed. Only cases involving the insertion of a new ventricular shunt system were included. Based on longstanding physician preferences, approximately half the neurosurgeons in the department did not routinely perform postoperative imaging during the study period. The primary outcomes were shunt-related mortality during the study period, shunt revision or mortality within 7 days of surgery, or hospital discharge, with time to first revision as a secondary outcome.

Results: During the study period 1485 shunt procedures were performed, of which 427 involved the placement of a new ventriculoperitoneal shunt system. The mean age of the patients was 5.2 years (range premature to 18 years). A total of 206/427 (48%) underwent routine postoperative imaging with MR or CT, and 153/427 (36%) underwent routine postoperative shunt series radiographs. During the study period, 2 abdominal radiographs identified preperitoneal placement; however, in both cases a symptomatic subcutaneous fluid collection was present (2/427, 0.5%). Postoperative CT or MR studies identified 1 case of extraventricular catheter placement and 5 cases of a kink adjacent to the valve (6/427, 1.4%), of which 4 cases were detected on routine imaging (4/427, 0.9%). No significant differences were detected between patients with or without imaging in 7-day postoperative (14 [7%] vs 8 [4%], p = 0.25) or 7-day postdischarge (2 [1%] vs 1 [< 1%], p = 0.96) rates of noninfective shunt revision, shunt-related mortality (4 [2%] vs 5 [2%], respectively, p = 0.8) or Kaplan-Meier long-term shunt survival curves (all p > 0.05).

Conclusions: Routine postoperative imaging after pediatric shunt insertion rarely identifies clinically significant issues, and consideration could be given to its omission.