Risk factors for cerebrospinal fluid shunt infection in pediatrics: A meta-analysis

Abstract

Background: Placement of cerebrospinal fluid (CSF) shunt for diversion remains a primary treatment for patients with hydrocephalus despite its surgical complications, including shunt infection, that remain high and become a medical and social problem. The meta-analysis was conducted to investigate risk factors of shunt infection in pediatrics.

Methods: Literature was searched on PubMed, Scopus, and the Cochrane Library. The methodology used for this investigation was preferred reporting items for systematic reviews and meta-analysis.

Results: This meta-analysis included five publications. The only significant results were found in ages <6 months with relative risk (RR) of 33.06 (95% confidence interval [CI] 9.27-117,99; P < 0.01), Caucasian race with RR of 15.24 (95% CI 6.77-34.34), and African-American race with RR of 2.37 (95% CI 2.07-2.70). The other results provided were not significant, such as intraventricular hemorrhage (IVH) of prematurity as the etiology of hydrocephalus with RR of 4.71 (95% CI 1.07-20.82), presence of gastrostomy during shunt insertion with RR of 3.80 (95% CI 0.91-15.88), and comorbidity of respiratory diseases with RR of 0.22 (95% CI 0.11-0.43).

Conclusion: Younger age during the shunt placement procedure, Caucasian race, and African-American race have a significantly higher risk of CSF shunt infection. The previously reported higher risk of shunt infection in cohort studies, such as IVH of prematurity and the presence of gastrostomy, were not significant in this study. Primary studies regarding shunt infection are advocated to be performed in a more extensive population with further risk factors included in the analysis.

Keywords: Hydrocephalus; Pediatrics; Risk factors; Shunt infection.

Figure 1:

Literature search flow.

Figure 2:

Bias analysis summary.

Figure 3:

Forest plot of the RR on age groups (a) <6 months, (b) 6-12 months, and (c) more than 12 months. MH: Mantel-Haenszel analysis, CI: Confidence interval, RR: Relative risk.

Figure 4:

Forest plot of the RR on etiology of (a) intraventricular hemorrhage of prematurity, (b) congenital hydrocephalus, and (c) tumor. MH: Mantel-Haenszel analysis, CI: Confidence interval, RR: Relative risk.

Figure 5:

Forest plot of RR of gastrostomy status. MH: Mantel-Haenszel analysis, CI: Confidence interval, RR : Relative risk

Figure 6:

Forest plot of RR of comorbid conditions, (a) malignancy and (b) respiratory. MH: Mantel-Haenszel analysis, CI : Confidence interval, RR: Relative risk.

Figure 7:

Forest plot of RR regarding the race of (a) Caucasian, (b) African-Americans, and (c) Latinos. MH: Mantel-Haenszel analysis, CI: Confidence interval, RR: Relative risk.