Hydrocephalus is a condition where there is an excessive accumulation of cerebrospinal fluid (CSF) in the cerebral ventricles. It can be due to inhibition of the circulation uptake of the fluid or it can be due to increased production of the fluid. This condition could be congenital or it can be acquired as a result of an abnormality [1]. The aqueductal stenosis accounts for the most common cause of congenital hydrocephalus. Congenital hydrocephalus can result from mutations in the L1CAM gene that leads to accumulation of the CSF due to the narrowing between the third and fourth ventricles. Other causes that lead to hydrocephalus include developmental disorders of the brain and/or spinal cord such as spina bifida and encephalocele. In other situations, it can be due to congenital brain tumors. On the other hand, acquired hydrocephalus can result from traumatic head injury, intraventricular hemorrhage, and subarachnoid hemorrhage or from other diseases like meningitis [2, 3]. Hydrocephalus is common and present worldwide. Munch TN et al. studied 1928683 live-born children in Denmark over a 30 year period and estimated a prevalence of 1.1 per 1000 with congenital hydrocephalus [4]. A study was conducted in 1997 in Saudi Arabia estimated 1.6 per 1000 live births had infantile hydrocephalus [5]. Untreated hydrocephalus might lead to increased intracranial pressure. As a consequence, many complicated conditions might occur such as herniation, intracranial hematoma, cerebral edema, or crushed brain tissue [6, 7] While in pediatric patients it can cause irritabilities, chronic headaches, learning difficulties, visual disturbances, and in some serious cases, severe mental retardation [7, 8]. The standard effective treatment is the surgical insertion of a cerebrospinal fluid shunt to drain the extra fluids to other regions of the body [9]. A cerebrospinal shunt is composed of a proximal catheter that is placed in the cerebral ventricle, a valve that regulates the pressure and the flow, and a distal catheter which drains the fluid to a body cavity. This cavity could be the atrium, bladder, or the peritoneal cavity [7, 10]. Although ventriculoatrial (VA) shunt is not the popular option, it could be used instead of ventriculoperitoneal (VP) shunt in some special cases that need revisions due to infection and obstruction [11]. The ventriculovesical (VV) shunt is the route of choice for pressure sensitive patients who require frequent VP shunt revisions due to recurrent obstructions.12 However, VA and VV shunts are associated with multiple life threatening complications [12]. VP shunt is an alternative and is the most commonly used. It provides a successful treatment for thousands of hydrocephalus cases annually [1].  VP has been used since 1950. Although all shunt surgeries are associated with complications such as; mechanical malfunction, hemorrhage, and infections; these complications are considered milder with VP shunt. Major complications that cause shunt failure can increase the need of shunt revision. In addition, it has been shown that complications can lead to drop of intellectual performance, seizures, and increase in mortality rate.8 Shunt infection is a major complication, and might lead to ventriculitis, meningitis, and other types of infections. It is defined as the presence of pathogenic organisms in the CSF specimen, in concurrence with CSF pleocytosis, neurological symptoms, fever and shunt failure.1 the association between infection and VP shunt and the risk factors of infection post VP shunt insertion have been identified in some regions around the world with variable types of organisms [9, 10]. A report from South Korea showed that 10.5% of the cases that underwent the VP shunt had infections [10].  It was found that patient’s age, duration of the operation, and history of shunt failure are risk factors that play a role in infection rate [8]. This study aimed at estimating the epidemiology of central nervous system infection post VP shunt surgery in pediatric patients at National Guard Health Affairs in Riyadh city and identify the possible risk factors contributing to VP shunt infections.

This study is a retrospective case series, where the incidence of infection post VP shunt surgery was estimated over 20 years. It was conducted at King Abdullah Specialized Children Hospital (KASCH), Riyadh, Saudi Arabia. KASCH is 600-bed hospital that has emergency, oncology diagnostic, and treatment departments. It also includes intensive care units, burn units, and a trauma center. Prior to the opening of KASCH, patients were followed in King Fahad National Guard hospital, later on; their medical records were transferred to KASCH. Pediatric patients who had the VP shunt surgery at KASCH were randomly sampled using the “Research Randomizer” website. The sample size was 700 cases, of witch only 172 met the inclusion criteria. Neonates and pediatric patient (from age 0 - 14 years old) of both genders who had their VP shunt surgery at KASCH from 1997 to 2017 were included. Patient who underwent any other neurological procedure one year prior to the VP shunt, patient with any missing preoperative or postoperative data, patient with congenital or acquired immunodeficiency, and patients who had the VP shunt surgery done in other hospital were excluded. A data extraction sheet was developed by the authors and validated by King Abdullah International Medical Research Center (KAIMRC) to gather the patient variables that were needed to estimate the incidence of infection post VP shunt surgery. A group of 10 medical students reviewed the patients’ charts at the medical record departments at KASCH and King Abdulaziz Medical City (KAMC) during the period between April and June 2019 and collected the needed data. All the collected data were entered in a Microsoft Excel sheet and the analysis was carried out using SPSS software version 23.  (SPSS Inc., Chicago, Illinois, USA). To avoid introducing potential selection bias, missing values were imputed using guidelines by Stuart and colleagues (2009). The proposed method provides valid inferences under the missing at random (MAR) assumption, which assumes that missing data are associated to observed variables and not to unobserved information. In multiple imputation, missing values are imputed stochastically several times. Imputing missing values several times allows the quantification of the uncertainty in results associated with imputation, and to account for this uncertainty in the final standard errors, confidence intervals and p- values were calculated. Multiple variables needing to be imputed was defined and imputed taking into consideration the whole dataset. A sensitivity analysis was conducted to interpret any significant difference between actual and imputed datasets. Assessment of normality was done using Shapiro-Wilk test. Data were expressed as means with standard deviations or as frequencies and percentages. Since, the study continuous variables failed normality assumption, evaluations of collected data were done using Mann-Whitney U test was used to compare means of the study independent samples, while Kruskal-Wallis test was used to determine whether there were any significant differences between means of more than two independent groups. Post hoc test was used for pairwise comparisons to further analyze these differences. Fisher’s exact test or chi-square test was used for categorical variables analysis. In an attempt to further investigate the associations between the study predictors and outcomes, different reported post-operative complications were combined to compute a new binary variable with subjects having either no reported complication or reported complication. This study outcomes was then used in several univariate models using the measured variables as potential predictors. Binary logistic regression analysis was then performed to calculate the odds ratio with their 95% confidence intervals. All the study hypothesis testing was carried out at the 5% (2-sided) significance level.

From the year 1997 till 2017, a total of 172 neonates and pediatric patients, who met the study inclusion criteria, were successfully included in this study. (Table 1) summarize the demographic and clinical characteristic of the study cohort. In brief, the cohort compromised of 96 (55.8%) female and 76 (44.2%) male patients, of which 114 (66.3%) were full term gestation and 58 (33.7%) were preterm. The majority of study cases were insertion (146 cases, 84.9%), while the remaining 26 (15.1%) cases were revision cases either due to malfunction or infection or both. Regarding the type of shunt inserted, 151 (87.8%) cases had Ventriculo-Peritoneal Shunt, while 18 (10.5%) cases had External Ventricular Drainage. Moreover, 112 (65.1%) of the study cases had multiple different comorbidities prior to their surgery. Additionally, antibiotics were prescribed for 136 (79.1%) of the cases preoperatively. The remaining were either poorly documented or missing data. (Table 1) illustrates demographic and clinical characteristics for the study cohort collected from the cases’ medical records.

Table 1: Demographic and Clinical characteristics summarized and reported for the study.

For the study cohort, the mean shunt insertion cases surgery time was 86.9 (±43.2) minutes, while revision cases recorded a mean surgery time of 90.9 (±41.0) minutes. For post-operative complications, 37 (21.5%) cases reported malfunction with mean onset time for symptoms of 5.1 (±11.6) months. On the other hand, 16 (9.3%) cases reported post-operative infection, and the mean onset time for symptoms was 1.3 (±2.5) months post-operative. The remaining post-operative complication ranged from obtrusions (7 cases, 4.1%), hemorrhages (1 case, 0.6%). A total of 12 (7.0%) cases passed away, two of them secondary to malfunction shunt and one case reported an infection due to shunt insertion surgery, and the remaining 9 cases died causes not related to shunt surgery. (Table 2) illustrates surgery outcomes for the study cases.

Table 2: Categorical and Continuous Variables cross tabulation against post-surgical complications (N = 172).

Among the six categorical variables, the type of surgery (revision vs. insertion) and gestation (term vs. pre-term) were significantly different across the reported post- operative complications, with p-values of 0.034 and 0.01 respectively (Table 2). We found that 31% of revision surgeries and 33% of the pre-term gestations had a malfunction shunt as a complication post operatively. Additionally, the pre-term gestation and revision surgeries reported a significant increase in infection incidence compared to other categories, with percentages of 8.6% and 10.3% respectively (Table 2). In addition, among the four continues variables, we found that age at surgery, weight at surgery and weight at birth were significantly associated with an increase in the incidence of reporting post-operative complications with p-values of 0.035, 0.043 and 0.022 respectively (Table 2). From the univariate logistic regression analysis, for each one-year decrease in age at surgery the odds of reporting post-operative complications increase by 21% (OR = 1.21, 95% CI: 1.04 – 1.41, p-value = 0.02). In addition, revision cases were significantly associate with a 33.3% increase in the odds of post-operative complications (OR= 1.33, 95% CI: 1.03 – 4.24, p-value = 0.04). The highest reported odds ratio was for preterm gestation, in which an increase of odds in reported post-operative complications was calculated to be 39% when compared to full term gestation (OR= 1.39, 95% CI: 1.24 – 4.61, p-value = 0.009) (Table 3).

Table 3: Univariate binary logistic regression analysis.

VP shunt is a well-used procedure done for patients with hydrocephalus. It has different types of complications, one of which is infection. In this study, infection was defined by the presence of clinical symptoms and positive CSF culture that led to revision to be performed in order to cure the infection. The incidence of infections post VP shunt was high, it was documented an overall CSF infection per procedure of 9.3% Most reported infections were in the first two months. However, it was consistent with the rate reported in other studies. The range of the infection per procedure that was reported by majority of international studies is 5 to 15%. [13-19] Two local studies reported rate of 25.9% and 8%. [20, 21]. Range varies between studies depending on the method and design of the study, definition of shunt infection, and the period patients were followed up. In previous international studies, the most common organism identified was coagulase negative Staphylococcus species [13]. Our study is consistent with it. Staphylococcus epidermidis, Staphylococcus aureus and a variety of gram- negative rods were also identified [13, 17, 19] One local study reported Pseudomonas aeruginosa as most common cause [20]. The association between two major variables and rate of infection were tested in this study, as well as previous studies done elsewhere [16, 19]. First variable was gestational age which was consistent with previous studies [19]. In this study, there was an increase in incidence rate of 10.3% among premature patients. This association might be due to prematurity immune system and their susceptibility to infection. In addition, higher rates of prior VP shunt insertions or revision were associated with higher infection rates. The association was found in different studies, too [16,19] The relationship between age at surgery and infection rate post VP shunt was controversial [13,15,16,18] In this study age at surgery, weight at surgery and weight at birth were significantly associated with an increase in the incidence of reporting post-operative complications. The major limitation of this study is the missing of some data such as precise gestational age and the use of prophylactic antibiotic; giving the fact that some medical records were electronic and some were paper charts. In addition, we were not able to assess some of the risk factors that contributed to infection rate such as etiology of hydrocephalus, immunity status of the patients, and preoperative measures that was taken. We did not assess surgeons’ experience and hospital factors. In National Guard hospital, surgeries used to be performed by adult surgeons and was changed to pediatric surgeons. This might have a role on the outcome of VP shunts. Furthermore, we included EVD insertion as well, which counted for 10.5% of the cases. Infections rate will probably be higher if we excluded EVD insertion from the study. We recommend for further studies, specifically controlled prospective studies that identify risk factors that make pediatric patients who undergo VP shunt susceptible for infections. Prospective studies can overcome some of the limitations of the retrospective studies. In addition, we recommend the implementation of further precautions to reduce the frequency of infection.

Shunt infections are common and complicate otherwise successful treatment of hydrocephalus, leading to increased healthcare costs and patient morbidity. The incidence of infections post VP shunt in this study was 9.3% with a mean symptoms onset time of 1.3. ±2.5 months post-operatively, which take a place in the high range of other reported studies around the world. Gestational age, age at surgery were identified as the major risk factors for post- operative infection. Further studies to identify factors that place pediatric patients who undergo VP shunt surgeries in a greater risk for infection are needed.

First and foremost, we have to thank our research supervisors Dr AlAlola, and Azzubi for their assistance and dedicated involvement. We would also like to acknowledge Dr. Nazish Masud for her comments that greatly improved the sampling method. Finally, we like to express a special appreciation and thanks to Mr. Abdullah AlKhaldi, Ms. Ghadah AlQuwaiee, Ms. Majd Binkhunain, Ms. Mai AlOtaibi, and Ms. Ghaida AlMutairi at the college of medicine, KSAU-HS, for their valuable efforts in the data collection.

1. Murgas GY, Snowden JN. Ventricular shunt infections. Immunopathogenesis and clinical management. J Neuroimmunol 2014; 276: 1-8.
2. Hydrocephalus Fact Sheet. NINDS. 2016
3. Haridas A, Tomita T. Hydrocephalus in children: Physiology, pathogenesis, and etiology. UpToDate. 2018
4. Munch TN, Rostgaard K, Rasmussen ML, Wohlfahrt J, Juhler M, et al. Familial aggregation of congenital hydrocephalus in a nationwide cohort. Brain. 2012; 135: 2409-2415.
5. Murshid WR, Jarallah JS, Dad MI. Epidemiology of Infantile Hydrocephalus in Saudi Arabia: Birth Prevalence. And Associated Factors. Pediatric Neurosurg. 2000; 32: 119-123.
6. Awad ME. Infantile hydrocephalus in the south-western region of Saudi Arabia. Annals of Tropical Paediatrics. 1992; 12:335-338.
7. Dulebohn SC, Mesfin FB. Ventriculoperitoneal Shunt. In: StatPeals Treasure Island (FL): StatPearls Publishing. 2017.
8. National Organization for Rare Diseases (NORD). 2007.
9. McGirt MJ, Zaas A, Fuchs HE, Timothy M, Kaye GK, Daniel J. Sexton; Risk Factors for Pediatric Ventriculoperitoneal Shunt Infection and Predictors of Infectious Pathogens, CID, 2003; 36: 858-862.
10. Treatment of hydrocephalus. National Hydrocephalus Foundation. 2014.
11. Karahan O, Yavuz C, Demirtas S, Caliskan A, Kamasak K, et al. Reasons, procedures, and outcomes in ventriculoatrial shunts: A single-center experience. Surgical Neurology International. 2013; 4:10.
12. West CGH. Ventriculovesical shunt. Journal of Neurosurgery. 1980; 53:858-560.
13. Pan P. Outcome Analysis of Ventriculoperitoneal Shunt Surgery in Pediatric Hydrocephalus [Internet]. Journal of pediatric neurosciences. Medknow Publications & Media Pvt Ltd. 2018
14. Vinchon M, Rekate H, Kulkarni AV. Pediatric hydrocephalus outcomes: a review. Fluids and Barriers of the CNS 2012; 9.
15. Braga MHV, Carvalho GTCD, Brandão RACS, Lima FBFD, Costa BS. Early shunt complications in 46 children with hydrocephalus. Arquivos De Neuro-Psiquiatria 2009; 67:273-277.
16. Simon TD, Hall M, Riva-Cambrin J, Albert JE, Jeffries HE, Lafleur B, et al. Infection rates following initial cerebrospinal fluid shunt placement across pediatric hospitals in the United States. Journal of Neurosurgery: Pediatrics. 2009; 4:156-165.
17. Gardner P, Leipzig TJ, Sadigh M. Infections of mechanical cerebrospinal fluid shunts. Current clinical topics in infectious diseases. U.S. National Library of Medicine. 1988.
18. Vinchon M, Dhellemmes P. Cerebrospinal fluid shunt infection: risk factors and long- term follow-up. Childs Nervous System. 2006; 22:692-697.
19. Mcgirt MJ, Zaas A, Fuchs HE, George TM, Kaye K, Sexton DJ. Risk Factors for Pediatric Ventriculoperitoneal Shunt Infection and Predictors of Infectious Pathogens. Clinical Infectious Diseases. 2003; 36: 858- 862.
20. Aly MB, Kamal H. Ventriculo-Peritoneal Shunt Infections in Infants and Children. Libyan Journal of Medicine. 2008; 3:20–22.
21. Anazi ARA, Nasser MJ. Hydrocephalus in Eastern Province of Saudi Arabia. Qatar Medical Journal. 2003.