Shunt infection prevention practices in Hydrocephalus Clinical Research Network–Quality: a new quality improvement network for hydrocephalus management

Mandeep S. Tamber Division of Neurosurgery, University of British Columbia, Vancouver, British Columbia, Canada;

Search for other papers by Mandeep S. Tamber in
jns
Google Scholar
PubMed
Close
 MD, PhD
,
Hailey Jensen Department of Neurosurgery, University of Utah, Salt Lake City, Utah;

Search for other papers by Hailey Jensen in
jns
Google Scholar
PubMed
Close
 MStat
,
Jason Clawson Department of Neurosurgery, University of Utah, Salt Lake City, Utah;

Search for other papers by Jason Clawson in
jns
Google Scholar
PubMed
Close
 BA
,
Nichol Nunn Department of Neurosurgery, University of Utah, Salt Lake City, Utah;

Search for other papers by Nichol Nunn in
jns
Google Scholar
PubMed
Close
 MBA
,
John C. Wellons III Department of Pediatric Neurosurgery, Vanderbilt University Medical Center, Nashville, Tennessee;

Search for other papers by John C. Wellons III in
jns
Google Scholar
PubMed
Close
 MD, MSPH
,
Jodi Smith Department of Neurosurgery, Peyton Manning Children’s Hospital, Indianapolis, Indiana; and

Search for other papers by Jodi Smith in
jns
Google Scholar
PubMed
Close
 PhD, MD
,
Jonathan E. Martin Department of Neurosurgery, Connecticut Children’s Medical Center, Hartford, Connecticut

Search for other papers by Jonathan E. Martin in
jns
Google Scholar
PubMed
Close
 MD
, and
John R. W. Kestle Department of Neurosurgery, University of Utah, Salt Lake City, Utah;

Search for other papers by John R. W. Kestle in
jns
Google Scholar
PubMed
Close
 MD, MSc , on behalf of the HCRNq Investigators
Free access

OBJECTIVE

Knowledge-based tools used to standardize perioperative care, such as the shunt infection prevention protocol of the Hydrocephalus Clinical Research Network (HCRN), have demonstrated their ability to reduce surgeon-based and center-based variations in outcomes and improve patient care. The mere presence of high-quality evidence, however, does not necessarily translate into improved patient outcomes owing to the implementation gap. To advance understanding of how knowledge-based tools are being utilized in the routine clinical care of children with hydrocephalus, the HCRN-Quality (HCRNq) network was started in 2019. With a focus on CSF shunt infection, the authors present baseline data regarding CSF shunt infection rates and current shunt infection prevention practices in use at HCRNq sites.

METHODS

Baseline shunt surgery practices, infection rate, and risk factor data were prospectively collected within HCRNq. No standard infection protocol was recommended, but site use of a protocol was implied if at least 3 of 6 common shunt infection prevention practices were used in > 80% of shunt surgical procedures. Univariable and multivariable analyses of shunt infection risk factors were performed.

RESULTS

Thirty sites accrued data on 2437 procedures between November 2019 and June 2021. The unadjusted infection rate across all sites was 3.9% (range 0%–13%) and did not differ among shunt insertion, shunt revision, or shunt insertion after infection. Protocol use was implied for only 15/30 centers and 60% of shunt operations. On univariable analysis, iodine/DuraPrep (OR 0.57, 95% CI 0.37–0.88, p = 0.02) and the use of an antibiotic-impregnated catheter in any segment of the shunt (or both) decreased infection risk (OR 0.53, 95% CI 0.34–0.82, p = 0.01). Iodine-based prep solutions (OR 0.56, 95% 0.36–0.86, p = 0.02) and the use of antibiotic-impregnated catheters (OR 0.52, 95% CI 0.34–0.81, p = 0.01) retained significance in the multivariable model, but no relationship between protocol use and infection risk was demonstrated in this baseline analysis.

CONCLUSIONS

The authors have demonstrated that children undergoing CSF shunt surgery at HCRNq sites share similar demographic characteristics with other large North American multicenter cohorts, with similar observed baseline infection rates and risk factors. Many centers have implemented standardized shunt infection prevention practices, but considerable practice variation remains. As such, there is an opportunity to decrease shunt infection rates in these centers through continued standardization of care.

ABBREVIATIONS

HCRN = Hydrocephalus Clinical Research Network; HCRNq = Hydrocephalus Clinical Research Network–Quality; IV = intravenous.

OBJECTIVE

Knowledge-based tools used to standardize perioperative care, such as the shunt infection prevention protocol of the Hydrocephalus Clinical Research Network (HCRN), have demonstrated their ability to reduce surgeon-based and center-based variations in outcomes and improve patient care. The mere presence of high-quality evidence, however, does not necessarily translate into improved patient outcomes owing to the implementation gap. To advance understanding of how knowledge-based tools are being utilized in the routine clinical care of children with hydrocephalus, the HCRN-Quality (HCRNq) network was started in 2019. With a focus on CSF shunt infection, the authors present baseline data regarding CSF shunt infection rates and current shunt infection prevention practices in use at HCRNq sites.

METHODS

Baseline shunt surgery practices, infection rate, and risk factor data were prospectively collected within HCRNq. No standard infection protocol was recommended, but site use of a protocol was implied if at least 3 of 6 common shunt infection prevention practices were used in > 80% of shunt surgical procedures. Univariable and multivariable analyses of shunt infection risk factors were performed.

RESULTS

Thirty sites accrued data on 2437 procedures between November 2019 and June 2021. The unadjusted infection rate across all sites was 3.9% (range 0%–13%) and did not differ among shunt insertion, shunt revision, or shunt insertion after infection. Protocol use was implied for only 15/30 centers and 60% of shunt operations. On univariable analysis, iodine/DuraPrep (OR 0.57, 95% CI 0.37–0.88, p = 0.02) and the use of an antibiotic-impregnated catheter in any segment of the shunt (or both) decreased infection risk (OR 0.53, 95% CI 0.34–0.82, p = 0.01). Iodine-based prep solutions (OR 0.56, 95% 0.36–0.86, p = 0.02) and the use of antibiotic-impregnated catheters (OR 0.52, 95% CI 0.34–0.81, p = 0.01) retained significance in the multivariable model, but no relationship between protocol use and infection risk was demonstrated in this baseline analysis.

CONCLUSIONS

The authors have demonstrated that children undergoing CSF shunt surgery at HCRNq sites share similar demographic characteristics with other large North American multicenter cohorts, with similar observed baseline infection rates and risk factors. Many centers have implemented standardized shunt infection prevention practices, but considerable practice variation remains. As such, there is an opportunity to decrease shunt infection rates in these centers through continued standardization of care.

In Brief

Iterative quality improvement methodology and standardization of care can close the gap between best clinical practices and common clinical practices. Hydrocephalus Clinical Research Network-Quality (HCRNq) investigators documented potentially important baseline practice variations related to perioperative shunt infection prevention practices, implying that shunt surgery is not standardized and that an opportunity to improve outcomes using knowledge translation interventions exists. Established HCRNq infrastructure will systematically evaluate practice variation and the effects of care standardization on outcomes in pediatric hydrocephalus patients.

The care of children with hydrocephalus demands significant healthcare resources. Analysis of an administrative data set from 2003 demonstrated that there were 40,000 hydrocephalus-related admissions in the United States, accounting for over 400,000 hospital days and total hospital charges of over $2 billion (2003 dollars).1 A significant proportion of the burden of hydrocephalus care relates to the management of CSF shunt infections. Reasonable unadjusted cumulative estimates of the 24-month incidence of CSF shunt infection after a child’s initial shunt procedure are in the vicinity of 12% per patient and 7% per procedure.2 In this analysis, patient, surgeon, and hospital factors contributed to the wide variation in the CSF shunt infection rates observed across centers.2

The tools used to standardize perioperative care have demonstrated their ability to reduce surgeon-based and center-based variation in outcomes and to improve patient care.3 There are a handful of examples of well-designed knowledge-based instruments, such as surgical checklists, being adopted in a wide variety of clinical settings and proving themselves to be promising strategies for fostering improvements in the culture of patient safety and the quality of perioperative care.4 Likewise, data from the Hydrocephalus Clinical Research Network (HCRN) have shown that standardizing the perioperative care of children undergoing CSF shunt surgery with the implementation of knowledge-based perioperative protocols can decrease the incidence of CSF shunt infection, an effect that is in part attributable to reducing surgeon-based and center-based effects on the outcome.57

The mere presence of high-quality evidence, however, does not necessarily translate into improved patient outcomes. Within the health professions, an implementation gap exists between the best current research evidence and translation of this evidence into clinical practice.8 Pockets of excellence exist, but knowledge of these better ideas and practices often remain isolated and unknown to others. Knowledge translation efforts seek to advance sound medical evidence into evidence-based practice, thereby closing the implementation gap.9 In other words, knowledge translation interventions seek to close the gap between best clinical practices and common clinical practices.

Surgeons participating in the care of children with hydrocephalus offer a unique opportunity to study the effect of the implementation of evidence-based best practices into routine clinical care. The HCRN, established in 2008, currently consists of 14 high-volume, free-standing academic pediatric neurosurgical centers across North America. Surgeons participating in this network have generated high-quality research-based evidence related to the care of children with hydrocephalus, including multiple iterations of a perioperative protocol designed to reduce the incidence of CSF shunt infection.57 To advance our understanding of how this knowledge is being utilized in routine clinical care, a parallel network—HCRN-Quality (HCRNq)—was started in 2019. HCRNq currently consists of 33 centers across North America that embody a mixture of practice types and environments. Despite this heterogeneity, surgeons at these sites are uniformly dedicated to improving the care of children with hydrocephalus, and they feel that they can achieve this by implementing the evidence-based clinical protocols generated by HCRN and others. The commitment of the surgeons participating in HCRN and HCRNq to this aim, together with the shared infrastructure between these two networks, should facilitate the rapid dissemination of innovation and new ideas so that advancements in hydrocephalus care can impact the greatest number of children in as short a time as possible.

As a means of decreasing the incidence of shunt infection across the continent, iteratively developed knowledge-based CSF shunt infection prevention protocols will be disseminated to HCRNq centers and their surgeons. To determine the effectiveness of these new protocols, it is necessary to understand the current landscape of CSF shunt surgery at HCRNq-participating sites. In this paper, we present baseline data regarding CSF shunt infection rates and current shunt infection prevention practices in the use at HCRNq sites.

Methods

Establishment and Operation of HCRNq

HCRNq was established in July 2019. At present, 33 centers from across North America contribute data related to CSF shunt surgery and other initiatives to the HCRN Data Coordinating Center at the University of Utah. As a matter of standard practice for quality improvement initiatives, frequent regular conference calls are used as a venue for training surgeons and data coordinators, as well as for the vital purposes of audit and feedback.

Most sites required a submission to their local research ethics board or institutional review board in order to participate in this quality improvement registry. Case report forms were carefully designed to capture only essential information without burdening participating surgeons and coordinators; data quality was prioritized over data quantity. To evaluate the existing practices designed to prevent CSF shunt infection, a single one-page form was used to collect relevant patient and procedural data, as outlined below.

Data Collection

During this observational phase, procedure type (shunt insertion, shunt insertion after external ventricular drain placement [not infected], shunt revision, shunt insertion at the end of infection treatment), baseline shunt surgery infection prevention practices, relevant risk factor data, and infection outcomes were prospectively collected within HCRNq. Sites were instructed to employ HCRN criteria for documentation of a shunt infection, which would have been deemed to occur if any of the following had been noted: 1) identification of organisms on culture from CSF, wound swab, or pseudocyst fluid; 2) shunt erosion (defined as wound breakdown with visible shunt hardware); 3) abdominal pseudocyst (even in the absence of positive cultures); or 4) positive blood cultures in a child with a ventriculoatrial shunt.5

Procedural data were collected between November 1, 2019, and June 29, 2021, with an additional 6-month follow-up to capture any infections related to shunt procedures occurring toward the end of the study period.

Definition of an Implied Protocol

In this study, no standard infection protocol was recommended, but the shunt infection prevention practices in use for each CSF shunt operation at each site were recorded. We were aware that some sites may have followed an infection prevention protocol prior to joining HCRNq. Therefore, we asked the sites if they had used a published protocol, a locally designed protocol, or no specific protocol. We then looked at their observational data and found that their behavior was often not congruent with their stated protocol. Therefore, we used the observational data to define an implied protocol, meaning the consistent use of several infection prevention steps in most of their cases. Specifically, this was defined as "use of at least 3 of 6 common shunt infection prevention practices [1) administration of prophylactic intravenous (IV) antibiotics or continuation of antibiotics for the treatment of a CSF shunt infection; 2) patient skin preparation with chlorhexidine as the last step before draping; 3) those participating in surgery perform surgical scrub using soap and water, not alcohol-based scrub solutions/creams; 4) double gloving of all surgery participants; 5) Ioban applied to surgical field; and 6) vancomycin/gentamicin injected into the shunt reservoir] in >80% of shunt surgeries" (with compliance considered a process measure). Use of antibiotic-impregnated catheters during surgery was recorded separately.

Statistical Analysis

Categorical data were summarized using frequencies and percentages; mean ± SD were used to summarize continuous data. Univariable and multivariable analyses of shunt infection risk factors were performed using logistic regression, with adjustment for within-patient correlation where necessary. For the purposes of this risk factor analysis, we chose relevant covariates associated with the patient (age, etiology, interval between index shunt surgery and prior shunt surgery, interval between index shunt surgery and prior shunt infection, and other demographic variables), surgeon (CSF shunt surgery volume), and center (CSF shunt surgery volume and implied protocol use).

ORs with appropriate 95% CIs were chosen as the measure for reporting associations. All analyses were performed using SAS version 9.4 (SAS Institute). A p value of < 0.05 was deemed statistically significant.

Results

Distributions of Patients, Procedures, and CSF Shunt Infections in HCRNq

From November 1, 2019, to June 29, 2021, 2437 CSF shunt procedures accrued across 30 HCRNq sites. The mean ± SD age at CSF shunt surgery was 6.3 ± 6.6 years; only 25% of CSF shunt procedures occurred in children less than 6 months of age. As demonstrated in the HCRN cohort, the most common etiology recorded on the CSF shunt surgery procedure form was postintraventricular hemorrhage secondary to prematurity (27%), followed by myelomeningocele (16%).10,11 Although we are still in the process of validating the etiology data provided by the sites in HCRNq, Table 1 provides provisional evidence of the comparability of the etiology profiles between the HCRNq cohort and the most comparable cohort within the HCRN, which consisted of children whose hydrocephalus treatment included only CSF shunt–related procedures performed outside of an HCRN center prior to their first CSF shunt operation at a HCRN site.

TABLE 1.

Etiology profile comparison between HCRNq and HCRN

EtiologyProportion of HCRNq Cohort (%)Proportion of Comparable HCRN Cohort (%)*
Intraventricular hemorrhage of prematurity2733
Myelomeningocele1624
Posterior fossa cyst43
Posterior fossa tumor53
Supratentorial tumor64
Aqueductal stenosis66
Postinfectious44
Congenital communicating86
Other congenital52

Includes children whose hydrocephalus treatment included only CSF shunt–related procedures performed outside of an HCRN center prior to their treatment at an HCRN site.

Of the 2437 procedures conducted in HCRNq over the study period, 791 (33%) were new CSF shunt insertions, with an additional 165 procedures (7%) categorized as placement of a new shunt after insertion of an external ventricular drain in a situation where the CSF was not infected. CSF shunt revision procedures represented 58% of the total assembled procedures (1403 procedures), and there were 78 (3%) CSF shunt insertion procedures after treatment of a shunt infection.

Ten of the 30 sites that contributed data conducted 100 or more CSF shunt procedures during the study period; an additional 10 sites contributed 50 or fewer procedures to the data set. The mean number of CSF shunt procedures conducted by each eligible surgeon at an HCRNq site during the study timeframe was 21 ± 19.

The unadjusted infection rate across all procedures across all sites was 3.9% (range 0%–13%). Most CSF shunt infection diagnoses were made on the basis of a positive CSF culture or documentation of an abdominal pseudocyst. There was no relationship between the occurrence of infection and the type of CSF shunt procedure performed (p = 0.41), as has been demonstrated previously (Table 2).7

TABLE 2.

Procedure type and risk of CSF shunt infection

Infectionp Value*
YesNo
Procedure type0.410
 Shunt insertion24 (3.0)767 (97.0)
 Shunt insertion after EVD (not infected)6 (3.6)159 (96.4)
 Shunt revision63 (4.5)1340 (95.5)
 Shunt insertion at the end of infection treatment3 (3.8)75 (96.2)

Values are shown as number (%) unless indicated otherwise.

Determined with the chi-squared test.

Practice Variation in Perioperative Procedures to Prevent CSF Shunt Infection

The sites participating in HCRNq demonstrated variable adherence to common CSF shunt infection prevention practices that have been validated in the literature.57 Nearly 90% of CSF shunt surgery procedures during the study period were performed under suitable prophylactic antibiotic coverage, and 95% of cases were performed using an iodine-impregnated barrier drape (Ioban) (Table 3). There was significant heterogeneity with respect to the techniques for patient skin preparation, with alcohol, iodine/alcohol, or other iodine products and chlorhexidine being used, variably alone or in combination, in two-thirds to three-quarters of procedures; chlorhexidine prep application as the last step before draping was recorded in 52% of cases. A formal water/soap-based scrub by surgical participants was recorded for 50% of cases, and double gloving by surgical participants happened in only 49% of procedures. Seventy-five percent of procedures had a new antibiotic-impregnated catheter placed (proximal, distal, or both), and a considerable minority of procedures (18%) were performed with vancomycin/gentamicin injected into the shunt reservoir after placement or had other forms of topical antibiotics applied (e.g., vancomycin powder left in the wound in 11% of cases) (Table 3).

TABLE 3.

Procedural variables and risk of CSF shunt infection

UsedNot UsedInfectionp Value*
UsedNot Used
Skin prep included DuraPrep or other iodine products1646 (67.5)791 (32.5)53 (3.2)43 (5.4)0.017
Chlorhexidine was last step before draping1248 (51.6)1170 (48.4)57 (4.6)36 (3.1)0.049
Antibiotic compliance (compliance to prophylactic antibiotics or IV continuation of treatment for infection)2177 (89.3)260 (10.7)89 (4.1)7 (2.7)0.264
Scrub compliance1209 (50.4)1192 (49.6)43 (3.6)52 (4.4)0.355
Double glove compliance1158 (48.5)1230 (51.5)41 (3.5)54 (4.4)0.377
Ioban applied2305 (95.3)114 (4.7)91 (3.9)3 (2.6)0.429
New antibiotic-impregnated catheter placed (proximal, distal, or both)1831 (75.2)604 (24.8)62 (3.4)34 (5.6)0.013
Vancomycin/gentamicin injected into shunt448 (18.4)1989 (81.6)15 (3.3)81 (4.1)0.481
Irrigation of wounds before closure (saline, vancomycin, Ancef, gentamicin, or other)1936 (79.4)501 (20.6)74 (3.8)22 (4.4)0.490
Vancomycin powder left in wounds224 (10.8)1844 (89.2)7 (3.1)73 (4.0)0.573
Implied protocol >80% on any 3/6 steps1466 (60.2)971 (39.8)59 (4.0)37 (3.8)0.773
Implied protocol >70% on any 4/6 steps840 (34.5)1597 (65.5)36 (4.3)60 (3.8)0.481
Implied protocol >90% on any 5/6 steps114 (4.7)2323 (95.3)4 (3.5)92 (4.0)0.867
Site-reported protocol use1684 (69.6)737 (30.4)63 (3.7)33 (4.5)0.393

Values are shown as number (%) unless indicated otherwise.

Results are based on a univariable model, adjusted for within-patient correlation.

At the beginning of data collection, each site was asked whether they had a formal shunt infection prevention protocol already in use at their site; 18 sites reported using an established protocol and 12 sites reported not having a protocol in place. Procedural data submitted by each site were reviewed and site behavior was not always congruent with their stated protocol. The procedural data were therefore used to define an implied protocol as described in the Methods. Three of the 12 sites that reported not using a protocol appeared to demonstrate a consistent approach to CSF shunt infection prevention, whereas 6 of the 18 sites that self-reported having a protocol in place did not actually exhibit a stereotyped approach to CSF shunt infection prevention. Overall, our data demonstrated implied shunt infection protocol utilization in 15/30 sites and 60% of CSF shunt surgery procedures (Table 3).

Univariable Analysis of CSF Shunt Infection Risk Factors

There was no association between the occurrence of shunt infection and patient age, hydrocephalus etiology, prior CSF shunt operation less than 26 weeks before the recorded procedure, or any other patient-level covariate (Table 4).

TABLE 4.

Univariable analysis of patient-related risk factors for CSF shunt infection

InfectionOR (95% CI)p Value*
YesNo
Age at shunt procedure0.443
 ≤6 mos26 (4.3)575 (95.7)Reference
 >6 mos70 (3.8)1766 (96.2)1.21 (0.76–1.93)
Etiology of hydrocephalus0.923
 Post-IVH secondary to prematurity29 (4.3)638 (95.7)Reference
 Myelomeningocele17 (4.3)376 (95.7)1.00 (0.54–1.86)
 Aqueductal stenosis4 (3.1)124 (96.9)0.74 (0.26–2.11)
 Other etiology45 (3.8)1129 (96.2)0.92 (0.55–1.52)
 Unknown1 (1.3)74 (98.7)
Prior shunt surgery0.261
 No surgery w/in 26 wks68 (3.4)1909 (96.6)Reference
 <26 wks28 (6.1)432 (93.9)1.54 (0.93–2.54)

IVH = intraventricular hemorrhage.

Values are shown as number (%) unless indicated otherwise.

Results are based on a univariable model, adjusted for within-patient correlation.

An individual surgeon’s CSF shunt surgery volume was not associated with CSF shunt infection risk, nor was the volume of CSF shunt operations conducted at any given HCRNq site.

Two individual procedural variables had a significant association with the development of CSF shunt infection (Table 3). As demonstrated previously, use of antibiotic-impregnated catheters during CSF shunt surgery reduced the risk of shunt infection.12,13 When antibiotic-impregnated shunt components were implanted at either the proximal end, distal end, or both, infection risk was observed to be 50% lower (OR 0.53, 95% CI 0.34–0.82, p = 0.01). Additionally, use of iodine/alcohol or other iodine-based skin prep solutions, either alone or in combination with other agents, also appeared to protect against the occurrence of CSF shunt infection (OR 0.57, 95% CI 0.37–0.88, p = 0.02). It appeared that using iodine/alcohol or other iodine-based skin prep solutions as a component of patient skin preparation was protective specifically when antibiotic-impregnated catheters were not used. More specifically, in the subset of cases where antibiotic-impregnated shunt components were not used, the use of iodine-based prep solutions lowered the infection rate from 8.6% (not used) to 4.3% (p = 0.03). No other individual procedural steps designed to prevent CSF shunt infection were associated with the outcome (selected measures presented in Table 3).

Site-reported protocol use was not associated with the occurrence of CSF shunt infection (p = 0.39) (Table 3). Use of an implied protocol (as defined in the Methods) was also not associated with CSF shunt infection (Table 3). Acknowledging that our chosen definition could be seen as arbitrary, we conducted a sensitivity analysis to serially define the threshold for implied protocol use as reasonable combinations of > 70%, > 80%, and > 90% of cases that were performed at any given institution with application of 3, 4, or 5, respectively, of the 6 common CSF shunt infection prevention strategies listed in the Methods. Once again, no relationship between implied protocol use and CSF shunt infection was found (selected analyses presented in Table 3).

Multivariable Analysis of CSF Shunt Infection Risk Factors

Based on the variables that were found to be significant on univariable analysis, we conducted a stepwise multivariable logistic regression model using data from 2435 procedures (2 procedures were excluded because the use of antibiotic-impregnated catheters was reported as "unknown"). Iodine-based prep solutions (OR 0.56, 95% CI 0.36–0.86, p = 0.02) and the use of antibiotic-impregnated catheters (OR 0.52, 95% CI 0.34–0.81, p = 0.01) retained significance in our multivariable model (Table 5). Once again, no relationship between implied protocol use and infection risk was demonstrated in this baseline analysis, even when various thresholds for implied protocol use were forced into the multivariable model.

TABLE 5.

Stepwise multivariable analysis of CSF shunt infection risk factors

Infection*
OR (95% CI)p Value
New antibiotic impregnated catheter placed (proximal, distal, or both)0.012
 NoReference
 Yes0.52 (0.34–0.81)
Skin prep included DuraPrep or other iodine products0.015
 NoReference
 Yes0.56 (0.36–0.86)

Results are based on a multivariable model, adjusted for each of the predictors in this table and for within-patient correlation.

Discussion

This is the first description of data from HCRNq, a representative multicenter network of pediatric neurosurgical centers across North America assembled for the purpose of advancing knowledge translation initiatives related to the care of children with hydrocephalus. The overarching rationale for the creation of HCRNq was to ensure that important scientific findings, distilled into easily accessible instruments such as care pathways and perioperative protocols, can influence patient care as widely and as quickly as possible. In addition to facilitating data collection regarding process and outcome measures, the established infrastructure will also allow for the study of the effectiveness of different knowledge translation interventions.

The baseline shunt infection data reported here provide important insights into infection prevention practices and infection risk across a large number of centers spanning different practice types and environments. We have demonstrated that children undergoing CSF shunt surgery at HCRNq sites share similar demographic characteristics (e.g., age and etiology) with those of other large North American multicenter cohorts, such as those reported by the HCRN.10,11 Moreover, the observed baseline infection rates and risk factors were also largely similar between the two networks. Accordingly, we expect that the institution of a formal shunt infection protocol at HCRNq sites may result in a similar relative risk reduction as observed when prior iterations of a shunt infection prevention protocol were implemented at an expanding number of HCRN sites.5,6 A formal protocol was implemented at HCRNq sites beginning in July 2021, and we are currently tracking process and outcome measures related to this new intervention.

Ten years after the first description of the HCRN shunt infection protocol5 and after publication of a revised protocol 5 years later,6 a striking amount of variation remains around the shunt infection prevention practices in use at different sites. When we examined the baseline data, we observed discordance between site-reported protocol use and actual evidence that shunt infection prevention techniques were stereotypically utilized during CSF shunt surgery. Only 15 of the 30 sites that contributed data to this analysis demonstrated regimented use of infection prevention practices, and only 60% of procedures across the network were performed under the cover of what would be considered a shunt infection prevention protocol, explicit or implicit. These findings are consistent with the existence of a gap related to the implementation of standardized perioperative procedures designed to minimize the incidence of CSF shunt infection. Although it would be optimistic to expect that 100% of cases would occur under the cover of a shunt infection prevention protocol, recent HCRN experience suggests that perfect protocol compliance should be attainable in at least 80% of procedures, with an incrementally greater proportion of cases being covered if one allows for less than perfect compliance with the steps of a protocol. Herein lies an opportunity, as further standardization of shunt infection prevention practices may decrease CSF shunt infection rates.

In this baseline analysis, we could not demonstrate a relationship between implied (or passive) site protocol use and the occurrence of CSF shunt infection. Nevertheless, it is still possible that formal institution of a protocol at HCRNq sites, as is being done currently, may demonstrate an anticipated effect similar to that observed when a formal shunt infection protocol was implemented in other practice environments.5,6 One must remember that implementation of an intervention, such as a standardized perioperative protocol, is an active process that goes well beyond simply providing a protocol or checklist for use. Engagement of the entire perioperative team, with all members invested in the success of the program, as well as periodic educational campaigns directed at key personnel and near real-time audit and feedback of process (compliance) and outcome (infection) measures, round out some of the key elements of a successful implementation campaign.9 The infrastructure behind HCRNq is well positioned to provide these important supports.

One must also be cognizant of potential barriers to the implementation of evidence-based best practices, be they at the organizational, systems, contextual, or human resources level. The observed discordance between site-reported protocol use and actual evidence that the shunt infection prevention techniques were stereotypically utilized during CSF shunt surgery hints at the possible existence of surgeon-based factors affecting the utilization of knowledge-based instruments at the point of care. We are in the process of exploring how certain attributes and behaviors of pediatric neurosurgeons may affect their utilization of new clinical evidence; this work may provide some insight regarding how future knowledge translation interventions could be better tailored to the intended audience.

The streamlined data collection procedures inherent to the design of HCRNq do pose some limitations to the kinds of analyses that can be performed. The decision to prioritize data quality over data quantity has meant that HCRNq sites are not collecting information regarding potentially important covariates that could be associated with the outcome of interest. For example, HCRNq sites are not collecting data regarding complex chronic conditions, a variable which may be associated with CSF shunt revisions and infections.14 Additionally, omission of data regarding key risk factor variables may make it difficult to directly compare patients within HCRNq to other assembled cohorts.

Conclusions

HCRNq was created with the desire to spread innovation and new ideas quickly so that advancements in hydrocephalus care could impact the greatest number of children in as short a time as possible. Its establishment seeks to close the gap between best clinical practices and common clinical practices.

In this first examination of HCRNq data, we sought to explore the current landscape of CSF shunt surgery in centers participating in this network. We were able to demonstrate that children undergoing CSF shunt surgery at HCRNq sites share similar demographic characteristics (e.g., age and etiology) with those treated at other large North American multicenter cohorts. Moreover, we observed that the baseline infection rates and infection risk factors were also largely similar between the two networks. On this basis, we expect that institution of a formal shunt infection protocol at HCRNq sites may result in a similar relative risk reduction as observed when previous iterations of a CSF shunt infection protocol were sequentially implemented at HCRN sites.

In addition, a striking amount of variation remains around the shunt infection prevention practices in use at different sites. In our examination, only 15 of the 30 sites that contributed data to this analysis demonstrated regimented use of infection prevention practices, and only 60% of procedures across the network were performed under the cover of what would be considered a shunt infection prevention protocol, explicit or implicit. Herein lies an opportunity, as further standardization of shunt infection prevention practices may decrease CSF shunt infection rates going forward. After implementing a new shunt infection protocol at HCRNq sites in July 2021, we are in the process of tracking process (compliance) and outcome (infection) measures and updated data will be forthcoming. We hypothesize that steps taken to implement an evidence-based standardized protocol should decrease practice variation and improve infection rates.

Acknowledgments

We acknowledge the contributions of the investigators and research staff of HCRNq.

Appendix

HCRNq Investigators and Staff

Hal Meltzer1, Kimberly Hamilton1, Patricia Dekeseredy1, Benjamin Warf2, Weston Northam2, Daniel Weber2, Adam Porter2, Joanna Papadakis2, Amanda Mosher2, Joseph Piatt3, Gregory Heuer4, Lina Lopez4, Todd Maugans5, Patricia Clerkin6, Vivek Mehta7, Jenny Souster7, Cameron Elliott7, Wendy Beaudoin7, Heather Burton7, Sudeshna Bhattacharya7, Sydney Gnenz7, Kyle Halvorson8, Katherine Ingram8, Victoria Nguyen8, Jodi Smith9, Erin Delaney9, Heather Cero9, Emily Vance9, Ruth Bristol10, Jeremy Gaiser10, Toba Niazi11, Kari Bollerman11, Jonathan Benitez11, David Gonda12, Vijay Ravindra12, Aida Alvarez12, Nalin Gupta13, Shivani Mahuvakar13, Catalina Pen13, Andrew Foy14, Irene Kim14, Amy Nader14, Allison Gonzales14, Jeffrey Raskin15, Robin Bowman15, Klaudia Dziugan15, James Botros16, Heather Spader16, Aaron Lujan16, Jaquelyn Brown16, Raheel Ahmed17, Jenna Bock17, Stephanie Wilbrand17, Maggie Oimoen17, McKenzie Endres17, Renee Reynolds18, Joanna Gernsback19, Michael Omini19, Michael Muhonen20, Bianca Romero20, Ann Ritter21, Carolina Sandoval-Garcia22, Emma Venteicher22, Leah Kann22, Andrew Kobets23, Samuel Ahmad23, Ashley Castillo23, Simon Walling24, Daniel McNeely24, Sarah Szego24, Jonathan Martin25, David Hersh25, Petronella Stoltz25, Kelly Mahaney26, Anthony Bet26, Adrian Valladrez26, Gabriella Morton26, Michael Partington27, Dante Kyle27, Brett Whittemore28, Jignesh Tailor29, Marissa Lowe29, Mariah Shirrell29, Matty Vestal30, Beth Perry30, Hazani Benitez-Rosas30, Amayrani Salvario Salgado30, Mark Souweidane31, Francis Villamater31, Peter Morgenstern32, Leslie Melo32, Scott Wait33, Danielle Shears33, Samer Elbabaa34, Greg Olavarria34, Yazandra Parrimon34, Alice Williamson34, Casey Madura35, Sarah Wiersema35, and Sarah Westenbroek35.

1WVU Medicine Children’s Hospital, Morgantown, WV; 2Boston Children’s Hospital, Boston, MA; 3Alfred I. DuPont Hospital for Children, Wilmington, DE; 4Children’s Hospital of Philadelphia, Philadelphia, PA; 5Children’s Hospital of Georgia, Augusta, GA; 6Children’s Hospital at Dartmouth-Hitchcock, Manchester, NH; 7Stollery Children’s Hospital, Edmonton, Alberta, Canada; 8Children’s Hospital of Minnesota, Minneapolis, MN; 9Peyton Manning Children’s Hospital, Indianapolis, IN; 10Phoenix Children’s Hospital, Phoenix, AZ; 11Nicklaus Children’s Hospital, Miami, FL; 12Rady Children’s Hospital, San Diego, CA; 13UCSF Benioff Children’s Hospital, San Francisco, CA; 14Children’s Hospital of Wisconsin, Milwaukee, WI; 15Lurie Children’s Hospital of Chicago, Chicago, IL; 16University of New Mexico Children’s Hospital, Albuquerque, NM; 17American Family Children’s Hospital, Madison, WI; 18Oishei Children’s Hospital, Buffalo, NY; 19The Children’s Hospital at Oklahoma University Medicine, Oklahoma City, OK; 20Children’s Hospital of Orange County, Orange, CA; 21Children’s Hospital of Richmond, Richmond, VA; 22Masonic Children’s Hospital, Minneapolis, MN; 23The Children’s Hospital at Montefiore, Bronx, NY; 24IWK Health Centre, Halifax, NS, Canada; 25Connecticut Children’s Medical Center, Hartford, CT; 26Lucile Packard Children’s Hospital, Stanford, CA; 27Children’s Mercy Hospital, Kansas City, MO; 28Children’s Medical Center of Dallas, Dallas, TX; 29Riley Children’s Hospital, Indianapolis, IN; 30Duke Children’s Health Center, Durham, NC; 31Weill Cornell Medicine, New York, NY; 32Mount Sinai Hospital, New York, NY; 33Carolina Neurosurgery & Spine Associates, Charlotte, NC; 34Arnold Palmer Hospital for Children, Orlando, FL; and 35Helen Devos Children’s Hospital, Grand Rapids, MI.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Tamber, Wellons, Smith, Kestle. Acquisition of data: Tamber, Clawson, Wellons, Smith, Martin, Kestle. Analysis and interpretation of data: Tamber, Jensen, Clawson, Nunn, Kestle. Drafting the article: Tamber, Jensen, Kestle. Critically revising the article: Tamber, Jensen, Nunn, Wellons, Smith, Kestle. Reviewed submitted version of manuscript: Tamber, Clawson, Nunn, Wellons, Smith, Martin, Kestle. Approved the final version of the manuscript on behalf of all authors: Tamber. Administrative/technical/material support: Clawson, Nunn, Martin, Kestle. Study supervision: Clawson, Kestle.

Supplemental Information

Previous Presentations

This work was presented as an abstract at the 2022 AANS/CNS Joint Section on Pediatric Neurosurgery Annual Meeting, Washington, DC, December 1–4, 2022.

References

  • 1

    Simon TD, Riva-Cambrin J, Srivastava R, Bratton SL, Dean JM, Kestle JR. Hospital care for children with hydrocephalus in the United States: utilization, charges, comorbidities, and deaths. J Neurosurg Pediatr. 2008;1(2):131137.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Simon TD, Hall M, Riva-Cambrin J, et al. Infection rates following initial cerebrospinal fluid shunt placement across pediatric hospitals in the United States. Clinical article. J Neurosurg Pediatr. 2009;4(2):156165.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Treadwell JR, Lucas S, Tsou AY. Surgical checklists: a systematic review of impacts and implementation. BMJ Qual Saf. 2014;23(4):299318.

  • 4

    Haynes AB, Weiser TG, Berry WR, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med. 2009;360(5):491499.

  • 5

    Kestle JR, Riva-Cambrin J, Wellons JC III, et al. A standardized protocol to reduce cerebrospinal fluid shunt infection: the Hydrocephalus Clinical Research Network Quality Improvement Initiative. J Neurosurg Pediatr. 2011;8(1):2229.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Kestle JR, Holubkov R, Douglas Cochrane D, et al. A new Hydrocephalus Clinical Research Network protocol to reduce cerebrospinal fluid shunt infection. J Neurosurg Pediatr. 2016;17(4):391396.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Chu J, Jensen H, Holubkov R, et al. The Hydrocephalus Clinical Research Network quality improvement initiative: the role of antibiotic-impregnated catheters and vancomycin wound irrigation. J Neurosurg Pediatr. 2022;29(6):711718.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Morris ZS, Wooding S, Grant J. The answer is 17 years, what is the question: understanding time lags in translational research. J R Soc Med. 2011;104(12):510520.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Bauer MS, Damschroder L, Hagedorn H, Smith J, Kilbourne AM. An introduction to implementation science for the non-specialist. BMC Psychol. 2015;3(1):32.

  • 10

    Tamber MS, Kestle JRW, Reeder RW, et al. Temporal trends in surgical procedures for pediatric hydrocephalus: an analysis of the Hydrocephalus Clinical Research Network Core Data Project. J Neurosurg Pediatr. 2020;27(3):269276.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Tamber MS. Insights into the epidemiology of infant hydrocephalus. Childs Nerv Syst. 2021;37(11):33053311.

  • 12

    Mallucci CL, Jenkinson MD, Conroy EJ, et al. Antibiotic or silver versus standard ventriculoperitoneal shunts (BASICS): a multicentre, single-blinded, randomised trial and economic evaluation. Lancet. 2019;394(10208):15301539.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Klimo P Jr, Thompson CJ, Baird LC, Flannery AM. Pediatric hydrocephalus: systematic literature review and evidence-based guidelines. Part 7: Antibiotic-impregnated shunt systems versus conventional shunts in children: a systematic review and meta-analysis. J Neurosurg Pediatr. 2014;14(suppl 1):53-59.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Riva-Cambrin J, Kestle JR, Holubkov R, et al. Risk factors for shunt malfunction in pediatric hydrocephalus: a multicenter prospective cohort study. J Neurosurg Pediatr. 2016;17(4):382390.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • Collapse
  • Expand

Figure from Gupta et al. (pp 127–136).

  • 1

    Simon TD, Riva-Cambrin J, Srivastava R, Bratton SL, Dean JM, Kestle JR. Hospital care for children with hydrocephalus in the United States: utilization, charges, comorbidities, and deaths. J Neurosurg Pediatr. 2008;1(2):131137.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Simon TD, Hall M, Riva-Cambrin J, et al. Infection rates following initial cerebrospinal fluid shunt placement across pediatric hospitals in the United States. Clinical article. J Neurosurg Pediatr. 2009;4(2):156165.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Treadwell JR, Lucas S, Tsou AY. Surgical checklists: a systematic review of impacts and implementation. BMJ Qual Saf. 2014;23(4):299318.

  • 4

    Haynes AB, Weiser TG, Berry WR, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med. 2009;360(5):491499.

  • 5

    Kestle JR, Riva-Cambrin J, Wellons JC III, et al. A standardized protocol to reduce cerebrospinal fluid shunt infection: the Hydrocephalus Clinical Research Network Quality Improvement Initiative. J Neurosurg Pediatr. 2011;8(1):2229.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Kestle JR, Holubkov R, Douglas Cochrane D, et al. A new Hydrocephalus Clinical Research Network protocol to reduce cerebrospinal fluid shunt infection. J Neurosurg Pediatr. 2016;17(4):391396.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Chu J, Jensen H, Holubkov R, et al. The Hydrocephalus Clinical Research Network quality improvement initiative: the role of antibiotic-impregnated catheters and vancomycin wound irrigation. J Neurosurg Pediatr. 2022;29(6):711718.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Morris ZS, Wooding S, Grant J. The answer is 17 years, what is the question: understanding time lags in translational research. J R Soc Med. 2011;104(12):510520.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Bauer MS, Damschroder L, Hagedorn H, Smith J, Kilbourne AM. An introduction to implementation science for the non-specialist. BMC Psychol. 2015;3(1):32.

  • 10

    Tamber MS, Kestle JRW, Reeder RW, et al. Temporal trends in surgical procedures for pediatric hydrocephalus: an analysis of the Hydrocephalus Clinical Research Network Core Data Project. J Neurosurg Pediatr. 2020;27(3):269276.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Tamber MS. Insights into the epidemiology of infant hydrocephalus. Childs Nerv Syst. 2021;37(11):33053311.

  • 12

    Mallucci CL, Jenkinson MD, Conroy EJ, et al. Antibiotic or silver versus standard ventriculoperitoneal shunts (BASICS): a multicentre, single-blinded, randomised trial and economic evaluation. Lancet. 2019;394(10208):15301539.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Klimo P Jr, Thompson CJ, Baird LC, Flannery AM. Pediatric hydrocephalus: systematic literature review and evidence-based guidelines. Part 7: Antibiotic-impregnated shunt systems versus conventional shunts in children: a systematic review and meta-analysis. J Neurosurg Pediatr. 2014;14(suppl 1):53-59.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Riva-Cambrin J, Kestle JR, Holubkov R, et al. Risk factors for shunt malfunction in pediatric hydrocephalus: a multicenter prospective cohort study. J Neurosurg Pediatr. 2016;17(4):382390.

    • PubMed
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 2503 2503 200
PDF Downloads 1019 1019 244
EPUB Downloads 0 0 0