Volume 22-04
15 February 1996

[Table of Contents]

Infection Control

PRELIMINARY REPORT: BIOSAFETY ANALYSIS OF ONE-WAY BACKFLOW VALVES FOR MULTIPLE PATIENT USE OF LOW OSMOLAR INTRAVENOUS CONTRAST SOLUTION

Low osmolar intravenous contrast solutions are widely used in hospitals for radiologic diagnostic studies, such as computed tomography scanning of the head, neck, thorax, and abdomen as well as angiography and venography. Owing to the relatively large volume and high cost of the intravenous solutions, multiple patient infusion of contrast from a single infusion bag has been proposed. Faced with budgetary restraints and increasing demand for radiologic and angiographic studies, many institutions have implemented this practice in their radiology departments.

A potential problem associated with multiple patient dosing from a single infusion bag is the risk of transmission of blood and bloodborne pathogens, such as hepatitis B, hepatitis C, human immunodeficiency virus, and other viral agents from one patient to the next during the diagnostic studies (1,2) . To prevent backflow of blood and avoid the potential transmission of bloodborne pathogens, some institutions have implemented the use of one-way backflow valves for multiple patient dosing. However, there are few published studies to support the integrity of a one-way backflow valve in preventing the contamination of the contrast solution. An advisory notice issued in August, 1994, by the National Steering Committee on Infection Control Guidelines Development, Laboratory Centre for Disease Control, Health Canada, advised against the use of intravenous contrast solution with more than one patient (3) .

Given this background, we designed a three-pronged, multidisciplinary study to assess the integrity of one-way valves in the prevention of backflow of potentially blood-contaminated or body fluid-contaminated solution into sterile multi-use intravenous contrast solution used during radiologic and angiographic studies.

Methods

Several one-way valves designed to protect the integrity of an infusion delivery system are available in Canada. In general, these one-way valves are check valves, which provide flow in a forward direction with no backflow leakage. Backflow valves from Medex Inc. (Hilliard, Ohio, USA), Merit Medical System (Salt Lake City, Utah, USA), and Namic (Namic Contrast Saving Delivery System, Glenn Falls, New York, USA) were obtained for the testing procedures. Valves from different lots were used in an effort to assess lot-to-lot production variability.

A series of test procedures were performed to assess the structural, functional, and biologic characteristics of the check valves. A three-part test protocol was developed by the Department of Medical Engineering at The Toronto Hospital and included a valve characteristics test, a static back pressure test, and a clinical simulation test (see Figure 1). Briefly, the valve characteristics test consisted of the determination of the forward operation of the valve. Pressure data for this and subsequent procedures were recorded using Biotek (Winooski, Vermont, USA) DPM III gauges with analog output sent to a personal computer for storage and subsequent analysis. The static back pressure test assessed the performance characteristics of the valve with various amounts of pressure applied against it. A syringe pump was used to supply a short-term pressure for 15 seconds simulating actual duration in clinical settings and a long-term test of 60 minutes to assess durability. A supraphysiologic pressure of 60 pounds per square inch (psi), equivalent to approximately 3,000 mm Hg, was chosen for this test. Each side of the valve was monitored for pressure profile changes during the period of application of the pressure. The clinical simulation test was conducted using the standard injection delivery system (Liebel-Flarsheim Company, Cincinnati, Ohio, USA) used in our Department of Radiology pressure tests to validate the presence of an active radionuclide during the entire period of the experiments.

Biologic testing of the check valves was performed using a staphylococcal bacteriophage (Group II, phage 55) as a human viral surrogate marker. This bacteriophage and the strain of Staphylococcus aureus (phage-propagating strain 3C) were obtained from the Reference Bacteriology Laboratory, Laboratory Services Branch, Ontario Ministry of Health, Toronto, Ontario (courtesy of A. Borczyk). This bacteriophage represents a medium-size virus particle that serves as a surrogate to such human pathogenic viruses as members of the herpes virus family or the human immunodeficiency virus type 1 (HIV-1), which might potentially be transmitted through blood or body fluids in a multiple-patient intravenous delivery system. The bacteriophage was prepared using a broth method according to a procedure obtained from the Laboratory Services Branch of the Ontario Ministry of Health (A. Borczyk: personal communication, 1995). A final working titre of the bacteriophage of 8.0 ´ 10 10 plaque- forming units/mL was subsequently injected into a one-litre bag of normal saline to be used in the one-way valve experiments.

The same experimental design as described for the radio-nuclide experiments was employed for the bacteriophage. Aliquots of 1 mL were obtained from the three-way stopcock just proximal to the one-way valve after each of three 1-minute pressure tests, as described previously. Positive control samples were obtained before and after all the pressure challenge experiments to ensure viability of the bacteriophage throughout the entire period of the testing. Twenty mL aliquots of the 1 mL fluid samples were dropped onto a propagating lawn of the fresh S. aureus strain (tryptic soy agar), and the plates were then read for lysis after an overnight incubation at 36o C.

Results

Three Medex valves from each of 10 separate lot numbers and three Merit valves from 2 separate lot numbers were used for the complete series of structural, functional, and biologic tests. One Namic valve was obtained and was used only for the structural test.

Structural Testing

The valve characteristics test indicated that the Medex valves required a significant (3.4 +/- 0.9 psi; mean +/- standard deviation) pressure to open them in the positive direction of flow and were thus considered a "sprung" type of valve. The Merit valves and the Namic valve were considered "unsprung", since the pressure needed to open them was less than 0.1 psi. The sprung nature of the valve represents a resting static pressure that must be overcome to open the valve membrane in the direction of flow. One of the 10 Medex valves exhibited pressure profile changes during the short-term back pressure valve test at a pressure of 15 psi (750 mm Hg). None of the Medex valves exhibited any pressure profile changes during the long-term back pressure valve test (60 psi for 1 h). In contrast, the Merit valves exhibited pressure profile changes of a significant nature in both the short-term and long-term tests. One of the two valves tested had a catastrophic failure during short-term testing. The other valve that passed the short-term back pressure testing failed the long-term testing. The Namic valve exhibited no pressure changes during short-term testing but exhibited pressure changes and leakage when subjected to the long-term back pressure testing. The clinical simulation valve tests demonstrated that the valves performed as expected with no flow across them when there was back pressure and no flow when the positive pressure exceeded the resting static pressure of the valve.

Functional Testing

Fifty samples were obtained, including pre-study and post-study samples. A 5.0 mL aliquot of each sample was counted for 1 minute (cpm) immediately following the procedure. Counts obtained were corrected for background radioactivity. There was no significant difference in the cpm of the pre-study (1.86 ´ 10 6 ) and post-study (1.88 ´ 10 6 ) samples, indicating no decay in the activity of the radionuclide during the time the experiments were per- formed. The first sample was collected with no pressure on the valve. Radioactivity was detected in none of the samples when the Medex valves were used. In one of two Merit valves tested, all three samples contained a mean of 1.8 ´ 10 6 cpm, indicating catastrophic failure of the valve from the pressure challenges. The methylene blue provided a visual aid to the failure of this valve.

Biologic Testing

Lysis was detected on control samples obtained before and after the experimental studies, indicating preservation of bacte- riophage activity throughout the time course of the experiments. Over 50 samples were obtained with one sample taken with no pressure and three others collected after a 1-minute pressure of 300 mm Hg (6 psi) against the valve. Lysis was detected in none of the samples obtained when the Medex valves were used. In one of the two Merit valves tested, all three samples contained significant amounts of bacteriophage with significant lysis of the S. aureus strain used as a marker for the bacteriophage. The valve that failed was from the same lot number as the valve that had failed the functional test procedure above.

Discussion

Our experiments tested two different types of valves: sprung and unsprung. None of the sprung valves exhibited failure at pressures that might normally be encountered in actual clinical settings, using multiple test procedures with very sensitive surrogate markers, including pressure measurements, a radionuclide molecule, and a bacteriophage. To further enhance the sensitivity of the experiment design, high concentrations of the radionuclide and the bacteriophage were used. In over 50 experiments performed with these sprung valves, only one gave suboptimal results. This valve exhibited a significant pressure-time profile change 3 seconds after beginning the short-term high back pressure test (15 seconds at 60 psi). It failed at a pressure of approximately 15 psi (equivalent to approximately 750 mm Hg), which greatly exceeds any back pressures that might be encountered in a clinical setting. All of the other sprung valves tested survived pressures of 60 psi without exhibiting failure,suggesting that intralot variability in the manufacturing process may account for this observation.

Although fewer unsprung valves were tested (two separate lots of the Merit valves for each of the test procedures and one Namic valve for the structural test procedure), the consistent failure of one of these unsprung valves (Merit) in the test procedures conducted at pressures that could easily be encountered in the clinical setting suggests that their utility as a safeguard for the prevention of backflow of blood or body fluids in a multi-dosing setting may be compromised. The limited number of experiments conducted on the other type of unsprung valve (Namic) limits the comments that may be made. The connection point on this valve leaked following the short-term high back pressure test, thus limiting further testing. From basic physical principles, plotting the flow rate against the opening area for different spring constant values demonstrates that with a sprung valve there is flow immediately upon opening it, which effectively prevents backflow. This principle in the design of backflow valves would make them desirable for the intended application of the prevention of any cross-contamination of bloodborne pathogens.

The results of these specimens, incorporating the basic physical principles discussed above, suggest that sprung one-way check valves may be safely used to prevent the backflow of potentially blood-contaminated or body fluid-contaminated fluids in a multi-dosing setting of intravenous contrast agents. It may be desirable to use a second valve as a backup to the first valve in the event of an unexpected catastrophic failure of the initial primary valve. Further studies are warranted by other investigators, but our findings suggest that the incorporation of sprung backflow valves should allow a high margin of safety in the use of multiple patient dosing of expensive low osmolar dyes, thus achieving significant cost reductions by reducing wastage.

Figure 1 Study Apparatus to Assess the Integrity of One-Way Valves

References

1. Heller RM, Horev G, Kirchner SG et al. AIDS awareness in the conduct of radiologic procedures: guidelines to safe practice. Radiology 1988;166:563-67.

2. CDC. Patient exposures to HIV during nuclear medicine procedures. MMWR 1992;41:575-79.

3. Paton S, Nicolle L. Advisory notice on low osmolar dye intravenous delivery systems. CCDR 1994;20:144.

Source: M Garcia, BScN, RN, S Woodward, MRT(N), R Reilly, MSc, PhM, K Anderson, RN, D Gretzinger, BEng, J Cafazzo, MHSc, PEng, J Ratner, PhD, A Easty, PhD, PEng, H Dedier, RT, J Yao, MD, W Wobeser, MD, J Conly, MD, Departments of Microbiology, Nuclear Medicine, Radiology and Medical Engineering, The Toronto Hospital, University of Toronto, Toronto, Ontario.

[Previous] [Table of Contents] [Next]