INTRODUCTION
Marcia Ryder
1, Elinor deLancey Pulcini
2, Albert Parker
2, and Garth James
2(1)
Ryder Science, Nashville, TN,
(2)Center for Biofilm Engineering, Montana State University-‐Bozeman, MT
…..medical biofilm research
Four PIVCs were tested: ViaValve® Safety I.V. Catheter (Smiths), the Autoguard™ BC Catheter (BD Medical), the Introcan Safety® 3 Catheter (B. Braun), and the Surflash® Plus Safety I.V. Catheter (Terumo/
CareFusion).
The inserJon of peripheral intravenous catheters (PIVC) is the most
common invasive procedure performed in hospitals. PrevenJng needlesJck injury and bloodborne pathogen exposure to healthcare workers during PIVC inserJon has been an ongoing challenge and places healthcare personnel at risk for HBV, HCV, or HIV infecJon.
The vascular catheter industry has responded with a new generaJon of advanced technologies that eliminates blood exposure during the inserJon procedure.
Valved blood control PIVC (BC-‐PIVC) technology requires the addiJon of internal components within the hub to prevent blood exposure to clinicians. The added components increase the internal surface area, dead space and volume within the catheter hub that is thought to increase biofilm formaJon and subsequent transfer of bacteria into the bloodstream increasing the risk of catheter-‐related bloodstream infecJon (CRBSI).
The 2011 Infusion Nursing Standards of PracJce indicates that the nurse should consider replacement of the short peripheral catheter when clinically indicated rather than the previous 72-‐96 hour resite standard. The increased potenJal for biofilm formaJon in the valved blood control catheters may increase the risk of PIV-‐CRBSI when le\
in place for extended periods of Jme.
Nine experiments were run with three Jme points measured within each run: 0, 72 and 96 hours. Catheters were precondiJoned with Bovine Serum Albumin (BSA) using a pressurized serum bo^le to mimic venous inserJon of the catheters.
A\er precondiJoning, a needleless connector (SmartSite®) was a^ached to each catheter, inoculated by flushing with 0.5 ml of a 104 colony forming units per ml (CFU/ml) of Staphylococcus aureus, and incubated at room temperature for 2 hours.
The connectors were then replaced with new sterile connectors and una^ached bacteria were rinsed from the fluid path using sterile Phosphate Buffered Saline (PBS).
Catheters were then either sampled or subjected to simulated clinical use by flushing 17 Jmes daily with 0.5 ml sterile nutrient and 1 flush at the end of the day with normal saline for 72 and 96 hours. The catheters were then either sampled or subjected to simulated clinical use by flushing 17 Jmes daily with 0.5 ml sterile nutrient and 1 flush at the end of the day with normal saline for 72 and 96 hours.
PURPOSE
The purpose of the this study was 1) to compare the bacterial transfer rate and intraluminal biofilm formaJon between three valved blood control PIVCs in a clinically simulated in vitro model and 2) to determine if the added components within the catheter hub with larger surface area and volume would also have larger bacterial transfer when flushed.
METHODS METHODS
ViaValve™ Safety I.V. Catheter
BD Insyte™ Autoguard™ BC Shielded I.V. Catheter
B.Braun Introcan Safety® 3 catheter
Surflash® Plus Safety I.V. catheter
RESULTS
Bacterial transfer, biofilm formation and effect of device
design in four peripheral IV blood-control catheters
Marcia Ryder1, Elinor deLancey Pulcini2, Albert Parker2, and Garth James2
There were staJsJcally significantly smaller bacterial mean log densiJes (LD) in Flush 1 and Flush 2 for the ViaValve™ compared to either the Autoguard™ BC (p = 0.003 and 0.001 respecJvely), Introcan® Safety 3 (p = 0.014 and 0.010 respecJvely) or Surflash®Plus Safety IV catheters (p = 0.014 and 0.005 respecJvely). There were no differences in the bacterial mean log densiJes between the Autoguard™ BC, Introcan® Safety 3 or Surflash® Plus Safety I.V. catheters in either Flush 1 or Flush 2 (p-‐value = 0.273 and 0.687 respecJvely).
For the ViaValve™, the rate of increase of the bacterial mean LD over Jme was small but not staJsJcally significant in either Flush 1 (0.16/day) or Flush 2 (0.23/day). For the Autoguard™ BC, there was a staJsJcally significant rate of decrease over Jme in the Flush 1 (-‐0.21/day) but not Flush 2 (-‐0.05/day). For the Introcan® Safety 3, a staJsJcally significant increase was found in both Flush 1 (0.71/day) and Flush 2 (0.91/day). For the Surflash®, a staJsJcally significant increase was found in both Flush 1 (0.43/day) and Flush 2 (0.61/day). The rates of change in bacterial mean LD over Jme amongst the three catheters were not significantly different in either Flush 1 (p-‐value ≥ 0.121) or Flush 2 (p-‐value ≥ 0.063).
Table 1. Yellow indicates that the mean LD of bacteria in Flush 1 through the ViaValve™ catheter was staWsWcally significantly smaller than either the mean LD for the Autoguard™ BC (p-‐value = 0.0030) or the mean LD for the Introcan® Safety 3 (p-‐
value = 0.014) and Surflash® Plus Safety I.V. catheter (p-‐value = 0.014). Green indicates that the mean LD the Autoguard™ BC, Introcan® Safety 3 and Surflash® Plus Safety I.V. catheter were not staWsWcally significantly different (p-‐value = 0.197).
Table 2. Yellow indicates that the mean LD of bacteria in Flush 2 through the ViaValve™ Safety catheter was staWsWcally significantly smaller than either the mean LD for the Autoguard™ BC (p-‐value < 0.0005), the mean LD for the Introcan® Safety 3 (p-‐value = 0.010) or the Surflash® Plus Safety I.V.
catheter (p-‐value = 0.005). Green indicates that the mean LD the Autoguard™ BC, Introcan® Safety 3 and Surflash® Plus Safety I.V. catheter were not staWsWcally significantly different (p-‐value = 0.687).
Catheters were sampled with a two-‐step procedure. First, each catheter was flushed to recover planktonic bacteria and plated to determine CFU/ml (Flush 1).
The connector surface was disinfected, sonicated in PBS to remove firmly a^ached (biofilm) bacteria, and flushed a second Jme and plated (Flush 2).
METHODS
In over nine experimental runs, the bacterial log densiJes (LD) in Flush 1 and Flush 2 were compared among four catheters types: the ViaValve™, Autoguard™ BC, Introcan® Safety 3, and Surflash®. Three experiments were performed with the Autoguard™ BC and the ViaValve™. Three other experiments with the Introcan® Safety 3 and the ViaValve™.
The last three experiments with the Surflash® and the ViaValve™. In each experiment, the bacterial log densiJes in Flush 1 and Flush 2 measured at three different Jme points:
0, 72 and 96 hours.
RESULTS
Figure 1 Figure 2
Figure 1 and 3: Each point in the graphs in Figure 1 and 2 is a LD for a single catheter in a single experiment. The lines indicate how the mean LDs for each catheter type change over Wme in each experiment.
Table 2 Table 1
RESULTS
Introcan 3 ViaValve
BD BC
B.Braun Introcan Safety® 3 Catheter BD Insyte™ Autoguard™ BC Shielded I.V. Catheter
A A
The log sum CFU for biofilm formed on all internal fluid pathway surfaces within Smiths Medical ViaValve™ Safety I.V. catheter was 3.84, BD Insyte™ Autoguard™ BC Shielded I.V. Catheter was 4.02, B.Braun Introcan Safety® 3 catheter was 4.90 and the Surflash®Plus Safety IV catheter was 5.43. Biofilm was observed on SEM on internal component surfaces with higher CFU counts
ViaValve™ Safety I.V. Catheter
Figure 3. Cross-‐secWonal schemaWc of the catheter hub accessed with syringe
E C B A
D
E D C B A Surflash® Plus Safety I.V. Catheter
Figure 4. DestrucWve sampling of catheter at 96 hours. Each part was placed in a
H G F E D C B A
Table 3. DestrucWve sampling CFU counts for each design component. The log sum CFU increased as the internal surface area increased.
B.Braun Introcan Safety® 3 Catheter BD Insyte™ Autoguard™ BC Shielded I.V. Catheter
ViaValve™ Safety I.V. Catheter
Surflash® Plus Safety I.V. Catheter
METHODS RESULTS
The hypothesis was that catheters with larger SAs and volumes would also have larger bacterial densiJes in Flush 1 and Flush 2. On the average, the staJsJcs support this hypothesis.
At each Jme point separately, there were significant increases in the Flush 1 and Flush 2 LDs as either SA or Volume increased (p ≤ 0.0001).
Across the Jme points, the effect of SA differed significantly (p ≤ 0.0150); and the effect of volume also differed significantly (p ≤ 0.0048). Based on R2 and informaJon criteria, neither SA nor Volume emerged as being be^er than the other when predicJng Flush 1 and Flush 2 results. In other words, both SA and volume are important.
InteresJngly, for both Flush 1 and Flush 2, the effect of SA was significantly different for different volumes (or, equivalently, the effect of volume was significantly different for different SAs (p ≤ 0.040).
The predicJve power of the models fit to the Flush 1 and Flush 2 LDs as a funcJon of both SA and volume and Jme can be roughly assessed by observing that the R2 ranged from 60 -‐ 63% for the Flush 1 and Flush 2 data respecJvely. In other words, 60 -‐ 63% of the variability in the Flush 1 and Flush 2 LDs are explained by the model, with 37-‐40% of the variability is due to unknown sources.(p-‐
value=0.014 and 0.010 respecJvely).
Figure 5. Comparison for surface area (in2) and biofilm count (CFU)
Bacterial transfer, biofilm formation and effect of device design in four peripheral IV blood-control catheters
Marcia Ryder1, Elinor deLancey Pulcini2, Albert Parker2, and Garth James2
in plunger
in catheter in hub
in seal in sleeve
between seal and hub in back of seal in actuator
BD BC in & around
plunger in seal
in hub in eyelet
in tube (enJre length not shown) ViaValve
Figure 6. Internal volume and surface area in contact with blood (in red). Blood locaWons used to calculate surface area and volume in Table 4. The cross-‐secWons, surface area (SA) and volume for each catheter was provided by Smith Medical engineers. .
Three different linear mixed effect models were fit to the Flush 1 LDs, and separately to the Flush 2 LDs. All of these models included a random effect for experiment, and a fixed effect for Jme (with 3 levels: 0, 72 and 96 hours). The first model included a covariate for SA and the 2-‐way interacJon between SA and Jme. The second model included a covariate for volume and the 2-‐way interacJon between volume and Jme. The third model included covariates for both SA and volume, their 2-‐way interacJon, and the 3-‐way interacJon between SA, volume and Jme. Random slopes for SA and Volume separately were also considered, but since it was not possible to fit random slopes to
“model 3” described above (that contains both SA and Volume), then this report does not consider any model with random slopes.
Residual and normal probability plots were used to check for outliers, and to check the normality and homogeneity of variance assumpJons of the models used.
RESULTS
Given the observaJon of increased biofilm formaJon in relaJon to surface area and volume, an analysis was done to determine if the added components within the catheter hub with larger surface area and volume would also have larger bacterial transfer when flushed.
Table 4. Comparison of the internal volume and surface area of the Autoguard™ BC, Introcan Safety®3, and Surflash Safety catheter hubs to the ViaValve™
Internal Volume and Surface Area in Contact with Blood
Internal surface area (in2) 0.34 0.96 2.8 x more SA
ViaValve™ Introcan®3 Difference
Internal volume (in3) 0.0038 0.017 4.5 x more vol Internal surface area (in2) 0.34 0.73 2.1 x more SA Internal volume (in3) 0.0038 0.0082 2.2 x more vol ViaValve™ Autoguard BC™ Difference
ViaValve™ Sureflash Plus® Difference
Internal volume (in3) 0.0038 0.0063 1.7 x more vol Internal surface area (in2) 0.34 0.66 1.9 x more SA
in hub in eyelet
Introcan 3
Table 4 . Surface area (in2) and biofilm count (CFU)
ViaValve™ Safety I.V. Catheter
BD Insyte™ Autoguard™ BC Shielded I.V. Catheter
B.Braun Introcan Safety® 3 catheter
Surflash® Plus Safety I.V. catheter
Surflash® Plus Safety I.V. catheter
Figure 3. Scanning electron microscope image of the white arm within the flow path. The large aggregates of spherical objects Indicate biofilm formaWon by Staph aureus.
CONCLUSIONS DISCUSSION
This project was funded by Smiths Medical Inc.
This informaJon was provided under a Montana State University TesJng Services Agreement and Ryder Science, Inc. and is not intended to endorse or recommend any product or service February 2015
As shown in the illustraJons and Table 1, the, BD Insyte™ Autoguard™ BC Shielded I.V. Catheter and the B.Braun Introcan Safety® 3 catheters have more complex flow paths than the ViaValve™
Safety I.V. catheter, as well as higher internal surface areas. Flow path irregulariJes, in parJcular areas that receive minimal fluid flow, are areas that promote bacterial a^achment and biofilm formaJon. This is analogous to dead-‐legs and surface irregulariJes in high purity water systems, which serve as reservoirs for biofilm accumulaJon and contaminaJon of the system.
Figure 5. This three-‐dimensional reconstrucWon of confocal scanning laser microscope images shows biofilm growth on a catheter septum. The biofilm was stained with the LIVE/
DEAD® BacLight™ Bacterial Viability Kit (Life Technologies CorporaWon, Carlsbad, CA). Live bacteria appear green.
Figure 4. Scanning electron microscopic image of the intraluminal surface of a sleeve within the catheter hub.
The large aggregates of spherical objects Indicate biofilm formaWon by Staph aureus.
RESULTS The log sum CFU for biofilm formed on all internal fluid
pathway surfaces within Smiths Medical ViaValve™ Safety I.V.
catheter was 3.84, BD Insyte™ Autoguard™ BC Shielded I.V.
Catheter was 4.02, B.Braun Introcan Safety® 3 catheter was 4.90 Surflash®Plus Safety IV catheter was 5.43.
Biofilm was observed on SEM on internal component surfaces with higher CFU counts
Figure 4. Scanning electron microscopic image of the intraluminal surface of a catheter tubing. The small groups of spherical objects are S. aureus cells.
Figure 4. Scanning electron microscopic image showing Staph aureus biofilm on the inner surface of a sleeve within a catheter hub.
Figure 2. Scanning electron microscopic image of the inside of a catheter hub. The small groups of spherical objects are S. aureus cells.
There are differences in biofilm formaJon among the devices with higher internal surface areas and volume. These differences increase the potenJal risk for transfer of bacteria into the bloodstream among the different blood control valved PIVC designs.
StaJsJcal analyses that considered the surface area and volume of each catheter indicated that there were significantly greater mean log bacterial densiJes in Flush 1 and Flush 2 as either surface area or volume was increased.
ViaValve™ Safety I.V. catheters had staJsJcally significantly fewer bacteria in flush counts compared to both Autoguard™ BC, Introcan®
Safety 3, and the Surflash Plus® catheters . It also had fewer bacteria on internal surfaces as well as a smaller internal surface area and less complex fluid path. These differences may reduce the potenJal risk for bacterial transfer and bloodstream infecJon.