Diagnosis of Ventricular Shunt Infection in Children: A Systematic Review.

BACKGROUND

No widely accepted gold standard for diagnosis of shunt infection exists, with definitions variable among clinicians and publications. This article summarizes the utility of commonly used diagnostic tools and provides a comprehensive review of optimal measures for diagnosis.

METHODS

A query of PubMed was performed extracting articles related to shunt infection in children. Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed, resulting in 1756 articles related to shunt infection, 49 of which ultimately met inclusion criteria.

RESULTS

Of the 49 articles included in the analysis, 9 did not define infection, 9 used culture alone, 9 used cultures and/or symptomatology, and 4 used a combination of cultures, cerebral spinal fluid (CSF) pleocytosis and symptomatology. The remainder of the studies used definitions from the Centers for Disease Control and Prevention (n = 2) and the Hydrocephalus Clinical Research Network (n = 2) or borrowed elements from these definitions. Variation in definition stems from the lack of sensitivity and specificity of commonly used signs, symptoms, and tests. Shunt tap alone is considered half as sensitive as hardware culture. Fever upon presentation was present in 16% to 42% of cases. CSF pleocytosis combined with fever has a sensitivity of 82% and specificity of 99%. CSF eosinophilia, lactic acid, serum anti-Staphylococcus epidermidis titer, procalcitonin, and C-reactive protein are non-specific and their utility is not well established.

CONCLUSIONS

The definition of shunt infection is variable across studies, with CSF culture and/or symptomatology being the most commonly utilized parameters.