European Journal of Cardiovascular Medicine

Ventriculoperitoneal shunt (VPS) insertion is a common neurosurgical procedure and firstline treatment for chronic hydrocephalus.¹ Contamination of ventriculoperitoneal shunts (VPS) by cutaneous flora, particularly coagulase-negative staphylococci, is a common cause of shunt infection and failure, leading to prolonged hospital stay, higher costs of care, and poor outcomes. Glove contamination may occur during VPS insertion, increasing risk of such infections.¹ The placement of ventriculoperitoneal shunt remains the mainstay in the surgical management of hydrocephalus.² Ventriculoperitoneal shunting is a popular method of cerebrospinal fluid diversion.² It is a suitable technique for most patients with hydrocephalus from any aetiology such as myelomeningocele, aqueductal stenosis, post infective hydrocephalus, brain tumor with obstructive hydrocephalus and other acquired conditions.² One of the most feared complications of cerebrospinal fluid (CSF) diversion surgeries, using for instance, a ventriculoperitoneal shunt (VPS), is a surgical site infection (SSI) leading to a iatrogenic shunt infection. Presumably, this infection occurs as a result of the contamination of the shunt with bacteria during surgery.³

HCP affects all age groups and its incidence and prevalence rise with aging demography.⁴ It is estimated that hydrocephalus affects more than 380,000 new individuals annually.⁵ It results in neurologic compromise and if remain untreated coma and death may follow.⁶  Shunt infections are challenging to treat and are associated with significant morbidity, increased rates of mortality and higher healthcare costs.⁷

Hence prevention of infection following shunt surgery is of utmost importance. Mainstays to prevent shunt infection are meticulous surgical technique and prophylactic use of antibiotics. As an individual measure to prevent shunt infections, antibiotics have received the most attention.⁸ 

Standard protocols have been devised to prevent shunt infections. Use of prophylactic antibiotic reduces infection rates. But it has poor penetrance into CSF and levels may be inadequate to combat organisms directly inoculated during surgery. In this regard intraventricular (IVT) antibiotics can be more efficacious as they bypass the blood–brain barrier (BBB) and achieve high local concentrations and exert more reliable bactericidal action. ⁹

The use of prophylactic intraventricular and topical vancomycin along with standard intravenous antibiotic prophylaxis and it has been reported to lower shunt infection rates in adult and children. This study examined the efficacy of per-operative intraventricular and topical vancomycin in reducing early ventriculo-peritoneal shunt infection in children.

This quasi-experimental study was carried out in the Department of Neurosurgery, Bangabandhu Sheikh Mujib Medical University from September 2022 to February 2024. Total 30 participants (n=30) were enrolled in this study who fulfilled the selection criteria. After obtaining informed written consent from participants/legal guardians, participant's data were recorded in a predetermined data sheet. Participants were divided into two groups- Group A- who underwent VP shunt surgery with per-operative intraventricular and topical vancomycin and Group B- who underwent VP shunt surgery without per-operative intraventricular and topical vancomycin. Other steps of surgery were identical in both groups. All the participants were discharged on 7^th post-operative day and were followed up for one month after the day of surgery at outpatient department. During follow up careful history was taken and clinical examinations were performed to identify and record signs and symptoms related to shunt infection. Suspected participants were admitted for required investigations and treatment. Data were processed and analyzed using SPSS (Statistical Package for Social Sciences) version 26 software. Appropriate statistical tests (Unpaired t-test, Chi-square test and Fischer exact test) for data analysis were done. Statistical significance was set at p- value <0.05.

Table-1: Age distribution of the study participants (n=30)

p-value obtained by Unpaired t-test, p<0.05 was considered as a level of significance

Overall, nearly half of the participants were < 2 years of age, 53.33% of participants were aged between > 2yrs -18 years. The mean 9.26±7.2 years in vancomycin group and 11.9±6.5 years in control group. This variability of age distribution between the two groups was statistically insignificant (p=0.30) (Table-1). Per-operative CSF analysis exhibited a significantly higher mean WBC count (11.40±16.2) in control group than vancomycin group (2.67±3.28) with an overall mean of 7.03±12.32 (p = 0.044). All the participants had glucose levels more than 2.5 mmol/L with a mean of 3.82±1.47 mmol/L with no significant group-wise variation. Mean CSF protein was 1.23±1.22. None of the per-operative CSF samples were positive for gram stain and none of the CSF samples yielded growth of any organism on culture (Table-2).

Postoperative clinical features of early shunt infection during follow-up period within 1 month. Four participants (13.33%) had fever among whom one participant from group A- 6.7% and three (03) participants from group B (p- value 0.282). In Group B, headache was observed in 6.7% of participants, while none occurred in Group A (p = 0.309). Neck stiffness was observed in 13.3% of Group B participants compared to none in Group A (p = 0.143). 2 participants (one in each group) had malfunctioning shunt with erythema over the shunt tract (6.7%) (p=1.000). 6.7% participant had abdominal lump (p=0.309). None of the participant in both groups had any episode of convulsion (0%), altered mental status, purulent wound discharge, wound breakdown or abdominal tenderness, lump or shunt malfunction (Table-3). 

Table-2: Per-operative CSF analysis findings of the study participants by group (n=30)

p-value obtained by Unpaired t test and p-value (b) obtained by Fisher exact test, p<0.05 was considered as a level of significance

Table-3: Distribution of the study participants by postoperative clinical features of early shunt infection during follow up after or within 01 (one) month (n=30)

p-value obtained by Chi-square test, p<0.05 was considered as a level of significance

Analysis of postoperative cerebrospinal fluid (CSF) findings reveals that the all the samples in both groups had CSF Leucocyte count <100 with overall mean 18.75±15.5 cells/mm³ (mean 40 in Group A and 11.67±7.64 in Group B) with all the samples having neutrophil percentage >10%. Glucose level was <2.5 in 50% samples (mean 2.72±1.46 mmol/L) and >2.5 in rest of the samples. Protein level was high (>0.45) in all 4 samples (mean 1.89±1.85 mmol/L). Gram staining was negative in all 4 samples and none of them yielded any growth on culture sensitivity. Findings were not statistically significant in both groups (p>0.05) (Table-4). Among the participants, early shunt infection was reported in 6.7% of cases in both Group A and Group B, resulting in a total of 6.7%. Conversely, absence of early shunt infection was observed in 93.3% of participants in both groups. There was no statistically significant association between early shunt infection and the use of intraventricular (IVT) and topical vancomycin (p=1.000) and the study failed to reject the null hypothesis (Table-5).

Table-4: Distribution of the study participants by postoperative CSF study findings (n=4)

p-value obtained by Fischer Exact test and p-value (b) obtained by Unpaired t-test, p<0.05 was considered as a level of significance

Table-5: Association of early shunt infection with IVT and topical vancomycin (n=30)

p-value obtained by Fischer exact test, p<0.05 was considered as a level of significance

Per-operative intraventricular and topical vancomycin may have efficacy in reducing early shunt infection following ventriculo-peritoneal (VP) shunt surgery. This study was designed to examine the efficacy by comparing two group of subjects each comprising 15 participants who underwent first time ventriculo-peritoneal shunt surgery for hydrocephalus (n=30; Group A= 15, Group B=15). Participants in Group A received intraventricular and topical vancomycin during surgery while the participants in Group B did not. The study results were not statistically significant to prove efficacy of per-operative intraventricular and topical vancomycin in reducing early shunt infection following ventriculo-peritoneal (VP) shunt surgery.

In the present study, the mean 9.26±7.2 years in vancomycin group and 11.9±6.5 years in control group, which higher than that described by McGirt et al.¹⁰ (7.6 years) and much higher than that (2.38 ± 4.36 years) found by Lakomkin and Hadjipanayis.¹¹. Both of those studies enrolled participants under 18 years. This difference was due to inclusion of participants of all age in the current study. 

Most of the participants in this study were male (63.3%) and male: female ratio was 1:1.7. A recent study done in our country by Hamid.¹² on 83 children undergoing CSF diversion procedures also found similar sex distribution where 66.7% participants were male and 33.3% were female.

No participant in both groups had any episode of post-operative convulsion or altered mental status (0%). One of the major concerns regarding application of intraventricular vancomycin is the risk of convulsion. Numerous studies have examined the safety of administering vancomycin into the CSF. Ng et al.¹³ concluded that this agent was both effective and safe. Similar safety profiles have been reported by Lee et al.¹⁴

Post-operative fever was observed in four participants (13.3%) among whom one (01) participant was from group A (6.7%) and 3 participants were from group (B 20%) (p- value 0.282). According to Duhaime et al.¹⁵ fever is not absolutely necessary for diagnosis of a shunt infection still Ajler and Yampolsky.¹⁶ emphasized on fever, CSF leukocytosis purulent discharge from wound and isolation of organism as diagnostic features. Since the extent clinical manifestations greatly depends on virulence of the causative organism participants with no other symptoms except for fever may in future culminate into shunt infection and eventually may result in other features related to shunt malfunction.

Two participants (one in each group) had erythema over the shunt tract (6.7%) and one (01) participant (3.3%) had abdominal lump. No participants had shunt malfunction, convulsion and abdominal tenderness in both groups. Depending on these clinical features clinical suspicion of infection was made in four (04) participants, among them clinical features were strongly suggestive of shunt infection in two participants and their shunts were removed. These features were justified according to the criteria described by Steinbok et al.¹⁷ and the shunt infection criteria adopted by this study. Since their shunts were removed these two participants were confirmed to have shunt infection. Removal of shunt hardware was itself defined as a criterion for shunt infection by James et al.¹⁸ which was not incorporated in the definition adopted by this study. The other two (02) participants revealed evidence of concurrent respiratory tract infection, no other clinical features of shunt infection except fever and insignificant CSF study results.

Both groups had per-operative CSF WBC counts below 100 cells/mm³, with an overall mean CSF leucocyte of 7.03±12.32 (p = 0.044). All the postoperative CSF samples revealed a WBC count < 100 cells/mm³, with a mean of 18.75±15.5 cells/ mm³ (p=0.18), which was greater than per-operative CSF WBC count. One clinically infected participant had CSF WBC count 40 cell/mm³ fairly above the criteria (>10 cells/mm³) defined by Bhimraj.¹⁹ (2017) for CSF shunt associated ventriculitis. Farooqi et al. ²⁰ and Sweeney et al.²¹ found no significant correlation between intraoperative CSF sampling and subsequent shunt infections and hence routine per-operative CSF sampling during VP shunt insertion was discouraged. CSF Neutrophil percentage of Group B was increased in postoperative CSF samples than that of per-operative samples (mean 53.3±32.5 vs 11.67±7.64).

Per-operative glucose levels were consistently higher than 2.5 mmol/L (mean-3.82±1.47) whereas two (50%) of four post-operative CSF samples showed glucose levels <2.5 mmol/L, (mean- 2.72 ±1.46 mmol/L). This indicates that the postoperative CSF glucose was lower than the per-operative samples. According to Ajler and Yampolsky¹⁶ an absolute CSF glucose level below 2.5 mmol/L was a diagnostic criteria of shunt infection along with fever. According to Zhang et al.²² per-operative CSF glucose levels of <2.8 mmol/L were associated with an increased risk of infection following ventriculo-peritoneal shunt surgery. In this study 2 of CSF samples from clinically diagnosed individuals with shunt infection revealed glucose level <2.5 mmol/L which is a strong predictor of CSF infection contrary to Hawkes et al.²³. Normal glucose levels do not rule out CSF infection, because up to 50 percent of participants with bacterial meningitis may have normal CSF glucose levels Dougherty and Roth.²⁴Clinical features must be weighed against low CSF glucose levels before embarking on diagnosing shunt infection.

In per-operative CSF analysis 60.0% participants had protein level >0.45 g/L (mean 1.89±1.85 g/dL) with slightly higher mean value in Group B without statistical significance (p = 0.604). Other than this protein concentration can be falsely elevated due to the presence of red blood cells Seehusen et al.²⁵ Clinically symptomatic participants with high CSF protein levels along with low CSF glucose levels have long been considered as a significant CSF change in bacterial meningitis or CSF infection. Biochemical parameters of CSF need to be justified with other clinical features and CSF findings.

Gram staining of CSF demonstrated no organism and CSF culture did not isolate any organism in all four (04) participants with clinical suspicion of shunt infection. Two of the participants were confirmed as shunt infection and their shunts were removed due to wound breakdown in one participant and erythema and swelling over the shunt tract in the other. One of them received per operative intraventricular and topical vancomycin while the other did not. Postoperative early shunt infection rate was 6.7% in both groups with no statistically significant difference (p=1.000). Isolation of microorganism(s) from the CSF is confirmatory of shunt infection. But Kontny et al.²⁶ found that only 85% of CSF samples from clinically confirmed infected shunts yielded growth of organism and 15% was false negative. In our study CSF cultures took 48 hours of incubation. CSF cultures have a minimum waiting time of 24 hours but organisms of the skin flora usually take 3 days or more, hence a final negative result takes at least of 4 days Desai et al.²⁷ and Winn.²⁸ For this reason, a negative culture with typical positive clinical signs and CSF pleocytosis suggesting infection were considered a shunt infection as described by van Lindert et al.²⁹.

Shunt infections, especially in adults, often present with nonspecific clinical signs, and affected participants can have normal CSF WBC counts and even negative gram stains and culture of CSF samples Conen et al.³⁰ Therefore, a high index of clinical suspicion is required for diagnosing shunt infection.

Our study result resonates with the standardized shunt protocol adopted by HCRN that included prophylactic IVT antibiotics and reported a significant reduction of shunt infection from 8.8% to 5.7% following the adoption of this protocol Kestle et al.³¹ However, they found no significant relationship between antibiotic injection into the shunt reservoir and decreased rates of infection. As a result, the particular role of IVT antibiotics in reducing infection rate was unclear in their study. Our study also fails to demonstrate any efficacy of IVT and topical vancomycin in reducing early shunt infection.

Ragel et al.³² has examined the role of 10 mg of intraventricular Vancomycin and 4 mg of intraventricular Gentamicin in adult subjects at the time of surgery in an aim of reducing postoperative shunt infection. In their analysis of 802 procedures, the authors concluded that combined intraventricular antibiotic therapy resulted in a significant reduction in postoperative shunt infections to 0.4% (control 5.4%), without any increase in treatment-related complications.

The overall infection rate in this study (6.7%) is comparable to rates 6.0%-7.2 % reported by number of established neurosurgical groups in the last several years and it indicates an institutional performance on par with that standard Raygor et al.³³ and Kestle et al.³⁴ . The 6.7% shunt infection rate in control group seen in our study compares favourably with that reported by Raygor et al.³³ and van Lindert et al.²⁹ , whereas infection rate in vancomycin group fairly differs from those studies (3.2% and 3.0% respectively). This difference is principally due to their longer duration of study with larger samples as well as due to use of antibiotic impregnated shunt systems and sutures.

Early shunt infection rate of 6.7% is by far the lowest rate among available literatures in our country which varies from 7.1% - 23.3%.^35,36 Strict adherence to the standards of procedure and use of checklist during shunt surgery might have helped reduce the overall infection rate in our study.

Results of this study failed to demonstrate any statistically significant efficacy of per-operative intraventricular and topical vancomycin in reducing early shunt infection. Overall early shunt infection rate was 6.7% which is similar to the rates reported by number of established neurosurgical groups.

1. Walek KW, Rajski M, Sastry RA, Mermel LA. Reducing ventriculoperitoneal shunt infection with intraoperative glove removal. Infection Control & Hospital Epidemiology. 2023 Feb;44(2):234-7.
2. Olomo SA, Obande JO, Bot GM, Binitie PO. Randomized trial of shunt infection rates comparing intraoperative Vancomycin versus Gentamicin in ventriculoperitoneal shunt system preparation. Egyptian Journal of Neurosurgery. 2023 Nov 16;38(1):68.
3. van Bilsen MW, Volovici V, Bartels RH. Topical antibiotics to prevent shunt infections—sensible or dangerous in the absence of evidence?. Acta Neurochirurgica. 2022 Jul;164(7):1789-91.
4. Kahle KT, Kulkarni AV, Limbrick DD, Warf BC. Hydrocephalus in children. The lancet. 2016 Feb 20;387(10020):788-99..
5. Dewan MC, Rattani A, Mekary R, Glancz LJ, Yunusa I, Baticulon RE, Fieggen G, Wellons JC, Park KB, Warf BC. Global hydrocephalus epidemiology and incidence: systematic review and meta-analysis. Journal of neurosurgery. 2018 Apr 27;130(4):1065-79.
6. Bonow RH, Hanak BW, Browd SR. Hydrocephalus in Children. Young child. 2018 Jan 1;3:7.
7. Fuller AT, Haglund MM, Lim S, Mukasa J, Muhumuza M, Kiryabwire J, Ssenyonjo H, Smith ER. Pediatric neurosurgical outcomes following a neurosurgery health system intervention at Mulago National Referral Hospital in Uganda. World neurosurgery. 2016 Nov 1;95:309-14.
8. Lakomkin N, Hadjipanayis CG. The role of prophylactic intraventricular antibiotics in reducing the incidence of infection and revision surgery in pediatric patients undergoing shunt placement. Neurosurgery. 2021 Feb 1;88(2):301-5.
9. LeBras M, Chow I, Mabasa VH, Ensom MH. Systematic review of efficacy, pharmacokinetics, and administration of intraventricular aminoglycosides in adults. Neurocritical care. 2016 Dec;25:492-507.
10. 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 Apr 1;36(7):858-62.
11. Lakomkin N, Hadjipanayis CG. The role of prophylactic intraventricular antibiotics in reducing the incidence of infection and revision surgery in pediatric patients undergoing shunt placement. Neurosurgery. 2021 Feb 1;88(2):301-5.