Repeated tapping of ventricular reservoir in preterm infants with post-hemorrhagic ventricular dilatation does not increase the risk of reservoir infection

Abstract

Controlled removal of cerebrospinal fluid (CSF) by serial tapping of ventricular reservoir, such as the McComb reservoir, is an effective way to decompress the ventricular system in preterm infants with rapidly progressing post-hemorrhagic ventricular dilatation (PHVD) while awaiting optimal conditions for permanent CSF drainage through a ventriculo-peritoneal shunt. However, the data regarding the risk of infection from repeated invasive tapping of a ventricular reservoir over a prolonged period are scarce. The aim of this study is to determine the incidence of ventricular reservoir infection from repeated tapping and to evaluate how often reservoir infection accompanies blood culture-proven sepsis in preterm infants with PHVD. We reviewed the medical records of all infants with PHVD receiving serial reservoir tap at the University of Michigan from January 2000 through June 2007. Serial reservoir taps were carried out by the neonatology team using aseptic technique. Surveillance CSF cultures were sent on a weekly basis as per unit practice or when clinical sepsis was suspected. Reservoir-tapping-related infection was present if a reservoir became infected during serial tapping 7 days or more after surgical placement of the reservoir. During the study period, ventricular reservoirs were placed in 35 infants for management of hydrocephalus. Six infants (17%) received ventricular reservoirs secondary to post-meningitic hydrocephalus or congenital brain malformations and were excluded. In the remaining 29 infants (birth weight: 1070+/-639 g (range 525 to 3204); gestational age: 26.9+/-4.1 weeks (range 23 to 40)) with PHVD, serial tapping of the ventricular reservoir was performed on 681 occasions (average number of taps per infant: 24 (range 2 to 82)). There were no cases of CSF culture-proven reservoir infection related to repeated taps. Thirteen of the 29 (45%) infants with PHVD developed blood culture-proven late onset sepsis with the following organisms: coagulase-negative Staphylococcus (9), Candida albicans (2), Escherichia coli (1), and Staphylococcus aureus (1), but all of the accompanying CSF cultures from the reservoir were negative. Ventricular reservoir infection from serial taps neither occurred in this consecutive series of preterm infants with PHVD, nor did ventricular reservoir infection accompany blood culture-proven sepsis. Concern of reservoir infection from repeated tapping should not be a limiting factor against placement of reservoirs.