Cerebrospinal Fluid Shunt Infections in Children: Do Hematologic and Cerebrospinal Fluid White Cells Examinations Correlate With the Type of Infection?

BACKGROUND

Cerebrospinal fluid (CSF) shunt infections in children represent an increasing problem in clinical practice. However, comprehensive clinical, laboratory and microbiologic data are scarce in pediatric age.

METHODS

We conducted a 10-year retrospective study to (1) analyze clinical, laboratory and microbiologic parameters associated with infections in children; (2) analyze results according to the type of catheter (medicated or not), type of infection (first or relapses), type of hydrocephalus (acquired and congenital), presence or not of bacteriemia; (3) describe antibiotic susceptibilities and their evolution during the study period.

RESULTS

Eighty-seven children with shunt infection and 61 children with mechanical shunt malfunction were enrolled. Fever, vomit, leukocytosis and elevated C-reactive protein were more frequent in the infected group (P < 0.001), while neurologic symptoms developed more frequently in the noninfected group (10.3% vs. 27.87%; P = 0.006). Local signs of inflammation and abdomen distension were similarly reported in the 2 groups. Children with medicated shunts had lower cell count in the CSF (12/mm3) compared with those with nonmedicated shunts (380/mm3; P < 0.0001). Gram-negative bacteria were more common in the not-medicated catheters (90.91% vs. 50% of cultures; P = 0.04). Gram-negative bacteria were identified in 50.67% of CSF cultures, Gram-positive bacteria in 53.33% and fungi were observed in 5.33%. Sixteen children (18.4%) had also a positive blood culture. Enterococci isolation was associated with relapsed infections (37.50% vs. 15.25%; P = 0.05).

CONCLUSIONS

Our study shows that the diagnosis and management of children with shunt infections are challenging. Prospective studies with a comprehensive approach focusing on patient, medical, microbiologic and surgical risk factors for first infection are urgently needed.