Infections of central nervous system shunts.

Approximately 1 out of every 10 ventricular shunts for hydrocephalus will become infected. This represents a three- to fivefold increase in infection rate compared with other neurosurgical procedures. Most often, the infection results from colonization of the shunt device by normally nonpathogenic skin flora at the time of surgery. Properties of the foreign body itself may contribute to this increased risk of infection, and most infections are clinically apparent within the first 6 months following surgery. Meticulous surgical technique remains one of the most important variables in reducing shunt infection while the role of prophylactic antibiotics is still unclear. When a shunt infection is suspected, percutaneous needle aspiration of the shunt reservoir is most always diagnostic. Fluid specimens should be sent for Gram smear, culture, and susceptibility testing. Appropriate systemic antibiotics should be started. In most instances, the infected patient should be treated by external ventricular drainage and intraventricular antibiotics until the CSF is sterilized. The drain should then be removed and a new shunt placed. Immediate removal of the infected shunt with simultaneous replacement by a new shunt in a different location also has a high cure rate but carries an increased risk of subsequent infection. With few exceptions, antibiotic therapy alone or partial shunt revision will be unsuccessful, and these therapies carry increased morbidity and mortality rates due to prolonged shunt infection. Despite the frequency of shunt surgery and its high rate of infection, a well-organized investigational approach to this problem has been lacking. Both laboratory and clinical investigation needs to explore new shunting techniques, biomaterials, the role of antibiotics, and microbial factors involving host defenses and the shunt apparatus itself.