Article Text

Original article
In vitro assessment of bacterial colonisation rates of goat umbilical cord segments using three embodiments of a novel neonatal umbilical catheter protection device
1. Carl L Dambkowski1,
2. Eric F Chehab2,
3. Joseph D Shih2,
4. Ross Venook2,
5. James K Wall2,3
1. 1Department of Emergency Medicine, Stanford Health Care, Palo Alto, California, USA
2. 2Department of Bioengineering, Stanford University, Stanford, California, USA
3. 3Division of Pediatric Surgery, Stanford Children's Health, Palo Alto, California, USA
1. Correspondence to Dr Carl L Dambkowski, Department of Emergency Medicine, Stanford Health Care, 300 Pasteur Drive, Palo Alto 94305, CA, USA; carld{at}stanford.edu

## Abstract

Introduction Central line-associated bloodstream infections (CLABSIs) in neonates with umbilical catheters occur at a rate that is 5 times higher than CLABSIs in adults with central catheters. No device currently exists tailored to the unique constraints umbilical catheters pose. The current study examined the natural progression of bacterial colonisation in goat umbilical cords and the relationship between embodiments of a novel neonatal umbilical catheter protection device and bacterial colonisation rates. The authors hypothesise that device venting is required to minimise bacterial colonisation in the unique umbilical stump environment.

Methods The natural progression of bacterial colonisation in goat umbilical cord segments was studied by examining bacterial colonisation rates each day over 7 days. To understand the relationship between protection and bacterial colonisation, umbilical catheters were placed in goat umbilical cord segments and secured with 1 of 3 embodiments of a novel umbilical catheter protection device, which offered varying degrees of venting. After a 7-day period of incubation, colony counts were compared.

Results Bacterial load was largest when umbilical cord segments were fresh and subsequently decreased over time. Further, bacterial colonisation rates showed a statistically significant difference between device embodiments (F(2,9)=4.26, p<0.05), with the non-vented embodiment showing the highest bacterial colonisation rate.

Conclusions A device to better stabilise and protect umbilical catheters in order to reduce bloodstream infections in neonates is greatly needed. The current experiments confirm the hypothesis that completely enclosed, or non-vented, protection device will have increased bacterial growth.

• Medical devices
• Neonatal
• Umbilical catheter
• Infection
• LIFEbubble™

## Introduction

Each year, over 450 000 infants are born prematurely in the USA1 and approximately 250 000 are admitted to the neonatal intensive care unit2 where many receive umbilical catheters (UCs) for an average of 6–8 days.2–4 Umbilical catheterisation is the preferred route of access for many of these neonates as the procedure offers relatively easy and reliable access to the venous and arterial systems with the necessary flow rates to deliver medication, fluids, parenteral nutrition, as well as dependable haemodynamic monitoring. However, the rate of central line-associated bloodstream infections (CLABSIs) in neonates with UCs is approximately five times greater than CLABSIs in the adult population with central catheters.2 ,5–8 Studies estimate that 5–10% of neonates with UCs develop CLABSIs, with the highest rates seen in infants weighing <1250 g.2 ,6–12 CLABSIs are the most common complication associated with UC placement and result in increased morbidity and a mortality rate of 7–11%.6 ,10 ,13 ,14 The current standard of care for securement and protection of UCs involves attaching the catheter to the umbilical stump with a suture, and securing the catheter with tape in a special configuration, known as a ‘goal post’ or ‘tape bridge’. 4

Research has demonstrated that two of the most important sources of infection are: bacterial entry from the external environment after placement of the catheter and lack of proper securement of the catheter.15–21 Infectious agents have the ability to pass along the external surface of the catheter from the insertion site to the distal tip of the catheter where bacterial colonisation most often occurs.15 The period after catheterisation is potentially the most vulnerable to infection because it is the only period of catheterisation that does not occur under sterile conditions and is the longest period.17 ,18 With regard to proper securement, one study in adults showed a reduction in central catheter-associated bloodstream infections from 9.4% to 1.1% when a properly manufactured external securement device was used instead of sutures.20

While multiple devices exist in the adult market related to central catheter protection and stabilisation, the unique constrains of the umbilical area lead to a different set of design constraints that do not allow for straightforward translation of adult-centric devices to neonatal applications. Specifically, UCs are the only catheters inserted through an area of tissue that will necrose during the subsequent weeks (ie, the umbilical cord stump). Further, the entry angle for UCs is different from other central catheters (ie, perpendicular instead of almost parallel to the skin at the point of entry). These design constraints necessitate a different approach than adult catheters, including fundamental research about the unique UC environment. To date, no device tailored to UCs and responsive to the aforementioned design constraints exists, leaving open an opportunity to decrease the high rate of UC-associated bloodstream infections.

The authors have designed a potential solution, called LIFEbubble, that addresses both catheter motion and bacterial migration in UCs (patent pending; US provisional patent number 62 156 120). The design of this device includes a protective dome or ‘bubble’ around the UC insertion site once catheterisation is complete. Adult catheters are completely enclosed with barrier devices; however, due to the fact that UCs are inserted through the umbilical stump, which will subsequently necrose while the catheter is still in place, the maximum amount of protection that can be used is unknown. In this study, the authors examined the natural progression of bacterial colonisation of the umbilical stump over a period of time as well as the effect of different degrees of protective enclosure on bacterial colonisation in ex vivo umbilical stumps.

To examine the natural progression of bacterial colonisation of the umbilical stump, goat umbilical cord segments were used to test the bacterial colonisation rate on a daily basis over a 1-week time period. The authors hypothesised that bacterial colonisation rate would decrease over the course of the 1-week time period.

In order to investigate the maximum amount of protection that can be provided without increasing bacterial growth, three embodiments of the device were designed with varying degrees of enclosure via a plastic dome structure. Neonatologists have long believed that ‘airing out’ of the umbilical stump is critical to decreased bacteria around the umbilical stump; however, no study to date as studied the effect of venting or ‘airing out’ on bacterial load as it relates to umbilical cords. One of the embodiments has an open vented system incorporated into the enclosure (figure 1A), a second has a vented system covered by a semipermeable membrane (see figure 1B), and a third has a complete enclosure (see figure 1C). Venting in the first two device embodiments is passive. These three embodiments of the novel neonatal UC protection device were compared in an in vitro model utilising goat umbilical cord segments and testing bacterial colonisation rates after 1 week, the current average length of time an UC is in place.3 ,4 The authors hypothesise that complete enclosure of the umbilical stump would lead to higher bacterial colonisation rates and that some amount of venting would yield reduced bacterial load in the unique umbilical stump environment.

Figure 1

(A) Domed enclosure with rectangular vents. (B) Domed enclosure with semipermeable membranes over rectangular vents. (C) Domed enclosure without vents.

## Methods

### Experiment 1: the natural progression of bacterial colonisation of goat umbilical cord stumps

Four fresh goat umbilical cord segments were collected from Harley Farms (Pescadero, California, USA) within 24 h of live goat birth. Two-centimeter segments (n=4) were placed on petri dishes. Using a Lazy-L spreader (Excel Scientific, Victorville, California, USA), umbilical cord segments were swiped multiple times and then the spreader was swiped on a tryptic soy agar (TSA) with 5% sheep blood plate. The agar plate was placed in an incubator (Thermo Scientific Steri-Cycle CO2 Incubator, Waltham, Massachusetts, USA) at 37°C for 24 h, at which time bacterial colonisation counts were performed. Segments of the four goat umbilical cords were placed in the incubator at 37°C, and the above procedure was repeated every 24 h for 168 h (7 days) for a longitudinal examination of bacterial colonisation.

### Experiment 2: the bacterial colonisation of UCs placed in one of three embodiments of a novel neonatal UC protection device after 1 week

All embodiments of the device were created on SolidWorks (Dassault Systèmes, S.A., Vélizy, France) and printed via additive manufacturing technology on a Dimension 1200es BTS (Stratasys, Edina, Minnesota, USA). The device embodiments were created of acrylonitrile butadiene styrene plastic from the additive manufacturing process. In the embodiment of the device with a vented system covered by a semipermeable membrane, the semipermeable membrane was a thin polyurethane membrane that is permeable to both water vapour and oxygen, but impermeable to microorganisms (Opsite Flexigrid, Smith & Nephew, Hull, UK). Devices were chemically sterilised with 70% ethanol prior to experimentation. After chemical sterilisation, a 3 cm segment of 5 French UC (Vygon, Lansdale, Pennsylvania, USA) was sterilely secured in one of the three device embodiments (see figures 1A–C). The distal end of the catheter was then inserted into a 1 cm segment of goat umbilical cord (n=4) collected from Harley Farms (Pescadero, California, USA) and used within 24 h of live goat delivery. The completed model was then placed on a sterile petri dish and placed in an incubator (Thermo Scientific Steri-Cycle CO2 Incubator, Waltham, Massachusetts, USA) at 37°C for 7 days. The device was taped to the bottom of the petri dish for securement. After 7 days, the distal 2 cm of the catheter was clipped and rolled on a TSA with 5% sheep blood plate. The agar plate was incubated for 24 h at 37°C after which time colony counts were performed.

The aforementioned procedure was repeated, with the exception of the inserting the setup into goat umbilical cord segments, as a control. In this experimental control setup, a bare catheter with one of the three embodiments of the device was set up with no goat umbilical cord segment.

## Results

### Experiment 1: the natural progression of bacterial colonisation of goat umbilical cord stumps

Bacterial colony counts were obtained daily after a 24 h incubation. Bacterial colony counts are reported in table 1 and graphically in figure 2. When compared with the previous day, bacterial colony counts at days 1 and 2 were statistically significantly lower (respectively: M=871.8, SD=230.5, p<0.01; M=122.3, SD=57.7, p<0.01); however, bacterial colony counts for days 3 through 7 did not reach statistical significance when compared with the previous day (respectively: M=182.3, SD=101.1, p=0.34; M=44.8, SD=31.4, p=0.06; M=17.2, SD=18.5, p=0.18; M=20.0, SD=18.5, p=0.84; M=34.5, SD=26.1, p=0.40).

Table 1

Bacterial colony count given days of incubation in goat umbilical cords

Figure 2

Bacterial colony count given days of incubation in goat umbilical cords segments without device embodiments. *Statistically significant when compared with previous day using Student t test.

### Experiment 2: the bacterial colonisation of UCs placed in one of three embodiments of a novel neonatal UC protection device after 1 week

Absolute bacterial colony counts are reported in table 2 and graphically in figure 3. A one-way analysis of variance was calculated on bacterial colonisation rates of the three embodiments of the novel neonatal umbilical protection device used in the current study. The analysis was significant (F(2,9)=4.26, p<0.05). Colony counts were highest when the non-vented device embodiment was used to secure the UC (M=1740, SD=664). Bacterial colony counts were lower when the UC was secured with the vented with semipermeable membrane device embodiment (M=1263, SD=372) and lowest when the vented device embodiment was used to secure the UC (M=708, SD=418). Post hoc analysis using Turkey's honest significant difference (HSD) test showed differences in bacterial colony counts between the non-vented and vented device embodiments that reached statistical significance (p=0.04). However, bacterial colony counts were not statistically significantly different between vented and vented with semipermeable membrane device embodiments (p=0.09) or between vented with semipermeable membrane and non-vented device embodiments (p=0.26). All control models had no bacterial growth.

Table 2

Bacterial colony count given device embodiment after 7-day incubation

Figure 3

Bacterial colony count on goat umbilical cords inside device embodiments with different venting, after 7-day incubation. *Statistically significant when compared with one-way analysis of variance: (p<0.05; (F(2,9)=4.26)) and post hoc analysis using Turkey's HSD test showed differences in bacterial colony counts between the non-vented and vented device embodiments (p=0.04).

## Discussion

This study is the first to describe the bacterial colonisation of mammalian (ie, goat) umbilical cord segments and measure the effects of a novel neonatal UC protection device on bacterial colonisation rates. These experiments provide insight that can guide the design for devices for neonatal UC protection with the potential to reduce CLABSIs in vulnerable neonates with UCs.

With regard to the first experiment, which examined the natural progression of bacterial colonisation in goat umbilical cord segments, the results demonstrate that bacterial colonisation decreases over time, reaching a nadir after 2 days and subsequently staying low thereafter. There was a slight rise from day 2 to 3; however, this was not statistically significant. This suggests that the necrosis of the umbilical cord segment in an open environment over time is critical to decreasing bacterial load. Many neonatologists have long believed that necrosis of the umbilical stump through a drying process that occurs while the umbilical stumps is aired out is important to decrease infection rates; however, no study had previously directly examined this in vitro or in vivo. The results of this experiment support the belief that the death of the stump can lead to lower bacterial colonisation rates. The possible mechanism that causes decreased bacterial colonisation over time in the umbilical cord segments is drying of the umbilical stump, which could lead to a less hospitable environment for bacterial growth. Subsequently, there is a theoretic concern that if the umbilical stump is unable to be aired out and dry, then the bacterial colonisation rates would be higher and could lead to higher CLABSI rates. The knowledge gained in this initial experiment is of critical importance to designing a neonatal UC protection device as it provides a unique constraint that does not exist for adult central catheter protection devices.

The results of the second experiment, examining the effect of one of three embodiments of the novel neonatal UC protection device on bacterial colonisation of the UC after a 7-day incubation period, show that the vented prototype leads to lower bacterial colonisation rates than the completely enclosed prototype, with the vented with semipermeable membrane prototype at a bacterial colonisation rate in between the other two embodiments (and not statistically significantly different from either). In conjunction with the results of the first experiment, these results suggest that venting of the device might allow for the necrosis of the umbilical stump, which seems to lead to lower bacterial colonisation rates. The non-vented prototype may act to increase bacterial growth by creating a hospitable environment for bacterial colonisation. More experimentation needs to be completed, with greater statistical power, before making definitive conclusions on the effect of the vented with semipermeable membrane device embodiment on bacterial colonisation of umbilical cord segments and to investigate the direct correlation between drying of the umbilical stump and bacterial colonisation rates. Overall, this experiment suggests that venting is a crucial design element that needs to be integrated into the final design of the novel neonatal UC protection device, although the exact amount of venting necessary is not yet known. The main aspect of the device's protective qualities that may help to reduce CLABSIs associated with UC use is to prevent inadvertent contact from caregivers, which was not tested in this in vitro experiment. Given that inadvertent contact may be leading to CLABSIs in neonates, the least amount of venting necessary while not creating a more hospitable environment for bacterial growth would likely lead to the greatest reduction in CLABSIs.

Before human use, similar experiments to those described in this manuscript need to be undertaken using human umbilical cord segments in order to confirm that the results between goat and human umbilical cord segments show similar trends. Furthermore, while these studies suggest that venting is critical in the final design of a neonatal UC protection and stabilisation device, the exact amount of venting necessary is still unknown and merits exploration. In future experiments, a greater statistical power should be used in order to be able to differentiate between various amounts of venting in a novel UC protection device, as limited statistical power was a limitation of the current experiment. The current experiments did also not examine the stabilisation qualities of the device, which is another potential aspect of a novel device that has the potential to reduce CLABSIs. Finally, further experiments will need to explore the material the device is made of, which in this study was acrylonitrile butadiene styrene plastic but would likely be a softer material prior to human clinical trials, and the method used to secure the device to the infant (tape in this experiment), as both are areas that could affect the bacterial colonisation rate.

Despite the limitations of the current study and the need for further experimentation, the current study is the first step to understanding how a novel neonatal UC protection device can reduce CLABSIs in neonates, save the lives of many infants and potentially reduce the cost burden related to CLABSIs on the healthcare system. Owing to recent changes in healthcare policy, CLABSIs are considered ‘never events’ and are no longer reimbursed by the government and insurance providers.22 Each CLABSI can now lead to US$30 000–US$50 000 that hospitals no longer receive in reimbursement.23–26 In addition to the life-saving potential of a novel neonatal UC protection and stabilisation device, hospitals have an incentive to realise the savings that could result from such a device.

## Conclusion

Necrosis of the umbilical stump is over time crucial to reduce the bacterial load of the umbilical cord; therefore a novel device to protect the umbilical stump, and secure UCs must have some amount of venting to best reduce UC CLABSIs.

## Acknowledgments

The authors would like to thank Dee Harley of Harley Farms in Pescadero, California, who helped the team collect goat umbilical catheters and Shawna Hannon, RN at Lucile Packard Children's Hospital for her invaluable insights into umbilical catheter placement and care.

View Abstract

## Footnotes

• Contributors CLD designed the in vitro trial, performed the in vitro trial, designed data collection tools, analysed the data, and drafted and revised the manuscript. EFC designed the in vitro trial and revised the manuscript. JDS designed the in vitro trial, performed the in vitro trial, analysed the data and revised the manuscript. RV designed the in vitro trial and revised the manuscript. JKW designed the in vitro trial, analysed the data, and revised the manuscript.

• Funding This project was generously supported by the Stanford Medical Scholars Fellowship Program and Stanford Children's Health, Division of Pediatric Surgery.

• Disclaimer Funders had no role in study design, collection, analysis and interpretation of the data, in writing the report, or in the decision of the paper for publication.

• Competing interests None declared.

• Provenance and peer review Not commissioned; externally peer reviewed.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.