Comparative Effectiveness of Surgical Treatments for Pediatric Hydrocephalus

Neurosurgery. 2018 Sep 1;83(3):480-487. doi: 10.1093/neuros/nyx440.


Background: Pediatric hydrocephalus represents a high health care burden in the United States (US). Surgery is the mainstay of treatment.

Objective: To perform a comparative effectiveness analysis for endoscopic third ventriculostomy (ETV) and cerebrospinal fluid shunt placement in pediatric hydrocephalus patients in the US using a large administrative claims database through the application of propensity scores matching.

Methods: The MarketScan® database (Truven Health Analytics, Atlanta, Georgia) 2003 to 2011 was used. Patients 19 yr or younger at first occurrence of ETV or shunt during the study period were included. The study outcome, surgery failure, was defined as further surgical treatment for hydrocephalus subsequent to initial ETV or shunt procedure. Age, etiology of hydrocephalus, and history of shunt were used to create matched samples for the ETV and shunt cohorts. Kaplan-Meier survival curves, stratified log-rank test, and Cox proportional-hazard models were used to analyze samples.

Results: There were 3231 eligible cases (478 ETV; 2753 shunt). Propensity scores matching produced 455 balanced pairs. For matched samples, 326 of 455 (72%) pairs were concordant, while 129 pairs were discordant in surgery outcomes within 3 mo. Among discordant pairs, ETV patients were more likely to experience surgery failure compared to patients receiving shunt (relative risk = 1.4, P value = .011). Furthermore, patients' age < 1 yr had lower ETV success rates than those with shunt (P value = .009). No similar pattern was found in patients' age ≥ 1 yr.

Conclusion: There was no significant effect on time to failure between patients undergoing ETV and shunt, except in infants' age <1 yr.

Publication types

  • Comparative Study

MeSH terms

  • Cerebrospinal Fluid Shunts / methods*
  • Child
  • Child, Preschool
  • Female
  • Georgia
  • Humans
  • Hydrocephalus / surgery*
  • Infant
  • Male
  • Outcome Assessment, Health Care
  • Propensity Score
  • Proportional Hazards Models
  • Treatment Outcome*
  • Ventriculostomy / methods*