Comparison of Bacterial Transfer and Biofilm Formation on Intravenous Catheter Surfaces Among Twenty Valves in a Clinically Simulated In Vitro Model

Ryder M, Pul­ci­ni E, Park­er A, James G. Pre­sent­ed at the World Con­gress on Vas­cu­lar Access, June 2016.

Summary

The risk of blood­stream infec­tions asso­ci­at­ed with vas­cu­lar access devices and needle­less valves is a grow­ing con­cern. Many ques­tions remain regard­ing the poten­tial infec­tion risks among dif­fer­ent types of con­nec­tors. Dis­in­fec­tion before access is cru­cial in pre­vent­ing the entry of microor­gan­isms, but adher­ence to this prac­tice is often insuf­fi­cient. This study, along with its accom­pa­ny­ing sta­tis­ti­cal analy­sis, com­pares the rate of bac­te­r­i­al trans­fer through 20 needle­less valves and exam­ines biofilm for­ma­tion inside the valve, with­in the catheter con­nec­tor, and inside the catheter.

The in vit­ro mod­el was designed to sim­u­late four dai­ly antibi­ot­ic infu­sions using the SASH (saline, antibi­ot­ic, saline, and heparin) method. Dur­ing each five-day study, the con­nec­tor of each valve was con­t­a­m­i­nat­ed twice dai­ly with approx­i­mate­ly 10^6 colony-form­ing units (CFU) of Staphy­lo­coc­cus aureus ATCC #6538. After each inoc­u­la­tion, the valves were flushed six times, and cul­tures were tak­en from the first and last flush­es to count the bac­te­ria. The aver­age log­a­rith­mic den­si­ties (LD) of CFU inside the valve, with­in the catheter con­nec­tor, and inside the catheter were also mea­sured.

Results

Neu­tron® and Micro­Clave® valves exhib­it the low­est aver­age log­a­rith­mic den­si­ty of bac­te­ria in the flush flu­id com­pared to all oth­er valves (= 3.14 and 3.20 log(CFU/flush)). Micro­Clave and Neu­tron did not dif­fer sig­nif­i­cant­ly from each oth­er sta­tis­ti­cal­ly and are the only devices in the group that had a sta­tis­ti­cal­ly sig­nif­i­cant, low­est aver­age log­a­rith­mic den­si­ty of bac­te­ria in the flush flu­id.

Conclusions

Ryzyko prze­niesienia bak­terii przez zawór, w łącznik cewni­ka i światło cewni­ka do krwioo­biegu z zanieczyszc­zonej powierzch­ni zaworu zależy od rodza­ju zas­tosowanego zaworu. Anal­izy regresji sugeru­ją, że log (CFU/złącze) jest najlep­szym poje­dynczym predyk­torem śred­niej dobowej LD bak­terii w popłuczy­nach (spośród trzech predyk­torów: wewnątrz zaworu, w łączniku cewni­ka i wewnątrz cewni­ka.

Introduction

Pre­vi­ous stud­ies indi­cate that the design ele­ments of needle­less valves affect the poten­tial for bac­te­ria to trans­fer from the valve sur­face to the flow path, catheter con­nec­tor, and catheter lumen. Biofilm with­in the lumen is the pre­dom­i­nant source of catheter-relat­ed blood­stream infec­tions (CRBSI).

Objective

The objec­tive of the study was to com­pare the rate of bac­te­r­i­al trans­fer for 20 needle­less valves and to com­pare the for­ma­tion of biofilm inside the valve, with­in the catheter con­nec­tor, and inside the catheter.

Biofilm form­ing with­in the catheter lumen is the pre­dom­i­nant source of catheter-relat­ed blood­stream infec­tions (CRBSI) dur­ing catheter care.

Methods

A total of 20 needle­less valve designs were eval­u­at­ed. Four valves of each type were assessed in three inde­pen­dent stud­ies (n=9), with Micro­Clave serv­ing as the con­trol in each of the 33 series. The valve mem­brane was con­t­a­m­i­nat­ed twice dai­ly with approx­i­mate­ly 10^6 CFU of Staphy­lo­coc­cus aureus ATCC #6538. The inoc­u­lat­ed valve was left to dry for 30 min­utes, then con­nect­ed to a 50 cm polyurethane periph­er­al­ly insert­ed cen­tral catheter (PICC).

Each catheter-valve set was flushed with 3.0 ml of ster­ile saline solu­tion, which was col­lect­ed and cul­tured on a plate (first flush). The catheter-valve sets were then flushed with ster­ile saline solu­tion (NS) two more times and blocked with ster­ile Brain Heart Infu­sion Broth (BHI) for one hour. Sub­se­quent­ly, they were flushed with NS three more times. The final flush was also cul­tured on a plate (last flush).

The Neu­tron and Micro­Clave con­nec­tors had sta­tis­ti­cal­ly low­er mean log­a­rith­mic den­si­ties (LD) of bac­te­ria after flush­ing, when all flush­es, con­t­a­m­i­na­tions, days, and cycles were summed up, com­pared to oth­er con­nec­tor types (3.14 and 3.20 log(CFU/flush)).

The catheter-valve sets were con­t­a­m­i­nat­ed for a sec­ond time each day after the sixth ster­ile flush­ing with saline, fol­lowed by a sec­ond series of flush­ing, trans­fer to a plate, and block­ing for a total of 18 valve acti­va­tions per day, which is con­sid­ered a rou­tine num­ber of acti­va­tions in the Inten­sive Care Unit.

The entire pro­ce­dure was repeat­ed dai­ly for five days. Except for days 4 and 5, when two catheter-valve sets were sam­pled for micro­bi­o­log­i­cal and destruc­tive micro­scop­ic exam­i­na­tion.

Sta­tis­ti­cal analy­sis was con­duct­ed using ANOVA and Tukey tests to deter­mine sig­nif­i­cant mean dif­fer­ences in the log­a­rith­mic den­si­ty of bac­te­ria in the flush, with­in the valve, in the catheter con­nec­tor, and with­in the catheter among dif­fer­ent needle­less con­nec­tors. Mul­ti­ple lin­ear regres­sion was employed to deter­mine if any com­bi­na­tion of bac­te­r­i­al log­a­rith­mic den­si­ty with­in the valve, catheter con­nec­tor, and with­in the catheter could sig­nif­i­cant­ly pre­dict the log­a­rith­mic den­si­ty of bac­te­ria in the flush.

 

Results

The Neu­tron and Micro­Clave valves gen­er­al­ly exhib­it the low­est mean log­a­rith­mic den­si­ty of bac­te­ria in the flush among all com­po­nents of the sys­tem (=3.14 and 3.20 log­a­rithm (CFU/flush)). Micro­Clave and Neu­tron did not sig­nif­i­cant­ly dif­fer from each oth­er and are the only devices that are in the group with the small­est sig­nif­i­cant mean log­a­rith­mic den­si­ty of bac­te­ria after flush­ing.

In the Q‑Syte® and Ultr­a­Site® valves, sig­nif­i­cant­ly high­er mean LD of bac­te­ria was observed after flush­ing com­pared to oth­er types of valves (= 5.37 and 5.15 log (CFU/flush)). Q‑Syte and Ultr­a­Site did not dif­fer sig­nif­i­cant­ly (p = 0.9101).

TABLE 1: AVERAGE DAILY BACTERIAL DENSITY IN FLUSH SOLUTION.

TABLE 2: AVERAGE SQUARE OF FLUSHES ACROSS ALL DAYS AND ALL FLUSHES.

Discussion

The risk of bac­te­r­i­al trans­fer through the inte­ri­or of the valve, catheter con­nec­tor, and inside the catheter into the blood­stream from a con­t­a­m­i­nat­ed valve sur­face depends on its type. The study’s find­ings con­firm that the for­ma­tion of biofilm in the catheter con­nec­tor and with­in the catheter lumen may result from bac­te­r­i­al trans­fer through the needle­less valve. More­over, it has been demon­strat­ed that detached or plank­ton­ic bac­te­ria detach­ing from the biofilm are sub­se­quent­ly intro­duced into the blood­stream through infu­sion. Regres­sion analy­sis indi­cates that biofilm for­ma­tion in the valve was the best pre­dic­tor of the num­ber of bac­te­ria intro­duced into the blood­stream (R2 = 95%). Thus, the use of a valve with a low rate of microor­gan­ism trans­fer can min­i­mize the risk of blood­stream infec­tion. It also under­scores the impor­tance of con­sis­tent and effec­tive valve and catheter hub dis­in­fec­tion meth­ods before obtain­ing access as a key strat­e­gy for pre­vent­ing CRBSI. The data also sug­gest that the com­mon clas­si­fi­ca­tion regard­ing valve char­ac­ter­is­tics, such as split sep­tum and mechan­i­cal valve, is an unre­li­able approach to device selec­tion based on infec­tion risk.

The risk of bac­te­r­i­al trans­fer through the inte­ri­or of the valve, catheter con­nec­tor, and inside the catheter into the blood­stream from a con­t­a­m­i­nat­ed valve sur­face depends on its type.

Conclusions

The risk of bac­te­r­i­al trans­fer from the con­t­a­m­i­nat­ed valve sur­face through the catheter con­nec­tor and catheter lumen into the blood­stream depends on the type of valve used, and Micro­Clave and Neu­tron valves demon­strat­ed sig­nif­i­cant­ly low­er bac­te­r­i­al trans­fer rates than any oth­er valves test­ed. Addi­tion­al­ly, the fre­quen­cy of valve replace­ment may depend on the poten­tial for bac­te­r­i­al trans­fer in each device design. Data from this study also sug­gest that the com­mon clas­si­fi­ca­tion of split-sep­tum and mechan­i­cal valves is too sim­plis­tic and unre­li­able as an approach to device selec­tion based on infec­tion risk.

It has been demon­strat­ed that Micro­Clave and Neu­tron valves exhib­it sig­nif­i­cant­ly low­er bac­te­r­i­al trans­fer rates than any oth­er valves test­ed..

Bibliography:

  1. Brown JD, Moss HA, Elliott TSJ. The poten­tial for catheter micro­bial con­t­a­m­i­na­tion from a needle­less con­nec­tor. J Hosp Infect. 1997; 36:181–189
  2. Yebenes J, Del­ga­do M, Sauca G, Ser­ra-Prat M, Sol­sona M, Almi­rall J, et al. Effi­ca­cy of three dif­fer­ent valve sys­tems of needle­free closed con­nec­tors in avoid­ing access of microor­gan­isms to endovas­cu­lar catheters after incor­rect han­dling. Crit Care Med 2008; 36: 2558–2561
  3. Ryder M, Fish­er S, Hamil­ton G, Hamil­ton M, James G. Bac­te­r­i­al trans­fer through needle­free con­nec­tors: com­par­i­son of nine dif­fer­ent devices. Pre­sent­ed at the Soci­ety for Health­care Epi­demi­ol­o­gy of Amer­i­ca Annu­al Sci­en­tif­ic Meet­ing, April 2007