- Email: [email protected]
Significant factors in the disinfection and sterilization of flexible endoscopes
Significant factors in the disinfection and sterilization of flexible e n d o s c o p e s Donald Vesley, PhD a Kathy G. Norlien' Barbara Nelson, LPN, C G C b Beverly Ott, C G C b Andrew J. Streifel, M P H c Minneapolis and Rochester, Minnesota
Background: Many nosocomial infection Outbreaks have been linked to improper
disinfection of the flexible endoscopes used in hospitals and clinics. The objective of this study was to evaluate the effÉcacyof scope disinfection with glutaraldehyde and hydrogen peroxide in manual and mechanical protocols. Methods: Bacillus subtilis and Pseudomonas cepacia were the test organisms. Each channel in two different endoscopes was seeded and evaluated separately. Residual chemical germicide levels in the channels and in .the .work environment We,re also :' measured. ,: ' " : " ~ ": Results: Parametric analyses Were carried out on log transformations of number of colony-forming units recovered. Repeated measures analysis demonstrated that both the type of disinfectant and the method of washing were significant factors for disinfection. Conclusions: Hydrogenperoxide proved to be more efficacious than glutaraldehyde for killing or rernoving B. Subtilis in a 10-minute contact period. Automatic disinfection was more efficacious than manual disinfection for killing or removing B. subtilis in a 10-minute contac~ Period. The ichannel being disinfected also proved to be a significant factor, with cba-bondioxide: and eieqator channels the most difficult to disinfect consistently. (~IIC AM'J INFECTCONTROL199.2;20:29!-300) :',
Cases of B~¢t6remia c a u s e d b y the use of c o n t a m i n a t e d ejr f l d o s c o p e s d u r i n g e n d o s c o p i c retr o g r a d e c h o l m a g i o p a n c r e a t o g r a p h y (ERCP) a n d sigmoidoscop~¢ p r o c e d u r e s a r e well d o c u m e n t e d in i n f e c t i o n 90ntrol l i t e r a t u r e . 1 S a l m o n e l l a a n d P s e u d o m o n a ( species h a v e b e e n i m p l i c a t e d in m a n y of thege n o s o c o m i a l infections. 2-5 The inc r e a s e of t u b e r c u l o s i s a n d H I V cases has f u r t h e r i n c r e a s e d the c o n c e r n a b o u t p r o p e r d i s i n f e c t i o n o r sterilization t e c h n i q u e s f o r these e n d o s c o p e s . 6 The use of flexible e n d o s c o p e p r o c e d u r e s in hospitals a n d clinics t h r o u g h o u t the U n i t e d States is r a p i d l y i n c r e a s i n g . B e c a u s e of the h i g h cost From the School of Public Health, University of Minnesota, Minneapolis, a the Mayo Clinic, Rochester, ~ and the Department of Environmental Health and Safety, University of Minnesota, Minneapolis. Supported by MediVators Inc., Cannon Falls, Minnesota. Reprint requests: Donald Vesley, PhD, Division of Environmental and Occupational Health, University of Minnesota, Box 197 Mayo, 420 Delaware St. S.E., Minneapolis, MN 55455.
17/46/39566
of these i n s t r u m e n t s , t h e r e is a v i t a l n e e d for c l e a n i n g a n d high-level d i s i n f e c t i o n o r sterilization p r o c e d u r e s w i t h m i n i m a l t u r n a r o u n d time. The p r o c e s s is c o m p l i c a t e d b y the a r r a y of channels in these i n s t r u m e n t s ; s o m e are e x t r e m e l y small in d i a m e t e r , w i t h l u m e n e n t r y p o r t s difficult to a c c e s s individually. I n addition, it is i m p o s s i b l e t o e n s u r e the p e r f u s i o n of all l u m e n s d u r i n g r o u t i n e ' r e p r o c e s s i n g , p a r t i c u l a r l y w h e n t h e r e ma); be slight v a r i a t i o n s in the t e c h n i q u e f r o m individual to i n d i v i d u a l a n d f r o m d a y to day. P r o c e d u r e s c u r r e n t l y u s e d i n c l u d e e t h y l e n e o x i d e (ETO) gas s t e r i l i z a t i o n a n d l i q u i d c h e m i c a l soaking. ETO sterilization has a h i g h t u r n a r o u n d t i m e of 24 hours, w h i c h m a k e s its use i m p r a c t i c a l in m o s t e n d o s c o p y units. G l u t a r a l d e h y d e a n d h y d r o g e n p e r o x i d e in form u l a t i o n s a n d c o n c e n t r a t i o n s c a p a b l e of sterilization a r e the u s u a l liquid c h e m i c a l s used. A s h o r t e r c o n t a c t t i m e is u s e d w i t h these c h e m i c a l s to a c h i e v e high-level'disinfection. R e p o r t s of w o r k e r sensitivity to g l u t a r a l d e h y d e , 7-9 the e s t a b l i s h m e n t of the O c c u p a t i o n a l S a f e t y a n d H e a l t h Adminis291
292
Vesley et al.
tration permissible ceiling exposure limit for glutaraldehyde of 0.2 ppm, 1° and a concern about chemical waste m a n a g e m e n t have further complicated this option. As a result, current attention has been focused on high level disinfection procedures with 6% hydrogen peroxide, both at room temperature as a high-level disinfectant and at elevated temperatures (50 ° C) as an Environmental Protection A g e n c y - a p p r o v e d sterilant. Additionally, septicemia after ERCP has been linked to the use of automatic endoscope disinfecting machines.n There is thus a need to evaluate critically mechanical endoscope disinfecting machines. The availability of the MediVator ED 105 mechanical disinfector (MediVators, Inc., Cannon Falls, Minn.), with high-pressure penetration of the disinfectant into the n a r r o w channels, provided an opportunity for comparative testing of the various treatment modes against appropriate microorganisms in a standard manner. METHODS
Pseudomonas cepacia (kingii) American Type Culture Collection (ATCC) 25609 and Bacillus subtilis ATCC 6633 were obtained from ATCC (Rockville, Md.). The freeze-dried culture of P. cepacia was aseptically rehydrated with trypticase soy broth (TSB); this. rehydrated mixture was added to 50 ml TSB, which was then incubated for 24 hours at 35 ° C. After incubation, P. cepacia was washed, vortexed, and spun d o w n at 4000 r p m two times before the pellet was placed in 1.5 L sterile buffered distilled water (SBDW). This mixture was placed in a dark cabinet for 48 hours to stabilize the population. For each day on which trials w e r e run, bacterial counts were obtained by serial dilution followed by enumeration of the bacteria on duplicate pour plates with standard methods agar (SMA). Between trials, the bacterial Solution was stored in a dark cabinet at r o o m temperature. Stock cultures were maintained on SMA at 4 ° C. B. subtilis ATCC 6633 freeze-dried culture was aseptically rehydrated with TSB; this rehydrated mixture was added to 50 ml TSB, which was then incubated for 24 hours at 35 ° C along with SMA-prepared Plates. Cultures were maintained on SMA at 4 ° C. Rapid disk assay method 12 was used to enumerate the B. subtilis and to obtain sufficient bacterial counts within the 106 to 108 colony-forming units (cfu)/ml range. A sample of this prepared mixture was stained by the Schaeffer-Fulton method, a modification of the Ziehl-Neelsen method, with
AJIC December 1992
malachite green a n d a safranine counterstain. Both spores and a limited nuinber of vegetative cells were visible with phase-contrast microscopy. The two endoscopes used in this study included an end-viewing sigmoidoscope (CF-P10S) and a side-viewing e n d o s c o p e used for ERCP (ERCP/JF-10 or duodenoscope), both manufactured by Olympus Optical Co., Ltd. (Tokyo, Japan). The 60 c m sigmoidosc0pe has air, water, suction, and carbon dioxide channels; the 120 cm ERCP scope has air, water, suction, and elevator channels. Reflux of body fluids from the patient m a y occur in any of the standard channels; it is therefore of utmost importance to clean and disinfect all channels of the endoscope thoroughly b e t w e e n patients. Both endoscopes were approximately the s a m e age and both were in excellent near-new condition, as demonstrated by visual inspection of the biopsy and suction channels. 13 The a m o u n t of inoculum needed to fill each channel was determined by trial and error with tap water. It w a s found that the suction channel could hold 15 ml, the air and water channels could hold 5 ml each, the c a r b o n dioxide channel could hold 3 ml, and the elevator channel could hold 1 ml. For consistency, these same volumes were used for inocula in each trial, as well as for S B D W rinse used for recovery after each individual treatment. Volumes of the inoculum and the S B D W rinse solutions were measured with a pipette. Delivery into the air, water, and special (carbon dioxide and elevator) channels-was accomplished with a syringe. The suction channel was also inoculated with a syringe, but recovery was accomplished by connecting the endoscope to a v a c u u m p u m p and drawing the sample through in the m a n n e r described in the Endoscope Channel Sampling Kit.* All trials were carried out in the same manner. The o r d e r in which the channels were inoculated was as follows: 1, air; 2, water; 3, suction; and 4, elevator or carbon dioxide channel. The order used for recovery was the same as that used for the inoculum. After the endoscope was seeded (inoculated) with the appropriate bacteria to be tested (B. subtilis or P. cepacia), it was manually washed and put through the disinfection or sterilization process t o be tested. Endoscopes were mechanically disinfected b e t w e e n each trial. *O-FI channel sampling kit for flexible endoscopes. MediVators, Inc. 1652 Greenview Dr. S.W., P.O. Box 6431, Rochester, MN 55903.
Volume 20
Number6 Inoculation of the endoscope Two individuals, an experimenter and an assistant, were required to carry out the inoculation for each endoscope trial. To inoculate the scope, the air inlet at the end of, the universal cord was blocked with the plug provided in the sampling kit. A 35 ml syringe was placed on the water port adapter and tightly attached to the water bottle port on the endoscope. For the air channels of the insertion tube and universal cord, 5 ml of inoculure was placed in the syringe while the assistant placed a finger over the air/water valve. The 5 ml was then injected into the scope. Another five ml of inoculum was placed in the same syringe and injected into the endoscope while the assistant depressed the air/water valve, which filled the water channels. Fifteen ml was then placed in the 35 ml syringe, which was connected to the suction port of the endoscope with a rubber tubing adaptor. The assistant then depressed the suction valve at the control head while the experimenter injected the inoculum into the scope. For the elevator channel of the duodenoscope, one ml of inoculum was placed in a 3 ml syringe that had been firmly attached with an adaptor to the elevator port of the endoscope. This inoculum was then injected into the endoscope. To inoculate the carbon dioxide channel, 3 ml was transferred by pipette into the 35 ml syringe, which had been connected to the carbon dioxide port in the universal cord. The carbon dioxide valve was depressed at the control head while the inoculum was slowly injected into the scope. Once inoculation of the scope was complete, either the inoculure was directly recovered for the control trials or the scope was put through the manual cleaning process followed by scope disinfection. Manual cleaning procedures After it was seeded with the test organism, the endoscope was placed in a laboratory sink. Liquid enzymatic cleaning solution (Endozyme; Ruhof Corp., Valley Stream, N.Y.) was then added to the wash water in accordance with manufacturer recommendations. Exterior surfaces of the endoscope were washed in this enzymatic solution with a soft-bristle toothbrush and clean cloth. The scope was then connected to the light source and the air/water valve was activated to ensure unblocked channels. The air/water valve was replaced with the air/water cleaning adaptor provided by the manufacturer. All accessible channels of the scope, in particular the suction channel of the universal cord, were thoroughly brushed. A
Disinfection of flexible endoscopes 293 small, hand-held cervical cytology brush was then used in the suction channel valve opening, biopsy port opening, and the suction tubing connection on the umbilical cord. During cleaning of the duodenoscope, the special cleaning h o s e was attached at the head of the scope for washing and rinsing the elevator channel. A 3 ml syringe was used to deliver 9 ml for each washing and rinsing of this elevator channel. The all-channe! irrigator was then attached and the scope was completely submerged in the cleaning solution and water. A 60 ml syringe was used to infuse cleaning solution through the scope. The wash and tap water rinses each used 250 ml of the solution. Disinfection/sterilization procedures After the manual washing was complete, the endoscope was put through a test disinfectant treatment. The test treatments were as follows: (1) manual disinfection with a 2% activated glutaraldehyde solution (MetriCide MX-1400, Metrex Research Corp., Parker, Colo.), (2) a mechanical disinfection (MediVator ED-105; MediVators, Inc.) with 2% Metricide at room temperature, (3) manual disinfection with a 6% hydrogen peroxide/0.85% phosphoric acid solution (EndoSpore; Globe Medical Instruments, Inc., Clearwater, Fla.), (4) mechanical disinfection with 6% hydrogen peroxide/0.85% phosphoric acid solution a t room temperature, (5) mechanical disinfection with 6% hydrogen peroxide/0.85% phosp h o r i c acid at 50 ° C, (6) ETO sterilization in a standard hospital cycle, and (7) manual wash followed by no disinfection. Each treatment was repeated three times for each of the two types of endoscopes. Between each trial, scopes were manually washed and automatically disinfected. B. subtilis and P. cepacia were always used in separate tests. Glutaraldehyde and hydrogen peroxide solutions were fresh at the beginning of the experiment; no solution was used for any purpose other than this experiment. Disinfectant solutions were replaced every week and glutaraldehyde was periodically checked with Sterilog colorimeter strips (Ihjmah Corp., Somerville, N.J.), which measure,whether the glutaraldehyde concentration is at least 1%. If the endoscope was to be manually disinfected after the manual wash described previously, the all-channel irrigator remained attached to the scope and tl~e scope was then submerged in the appropriate disinfectant. The 60 mA syringe was used to infuse 180 ml disinfectant through the
A31C 294
Vesley et al.
scope. If the duodenoscope was being disinfected, 9 ml disinfectant was infused separately into the elevator channel with the 3 ml syringe. The scope was then left to soak in the disinfectant for 10 minutes. Ten minutes was used for the liquid chemical contact time because it is the time m o s t often used in gastrointestinal endoscopy labs a n d because a longer contact time is believed to be impractical during m o s t endoscopy schedules. Although it was not measured, an additional 1 to 1.5 minutes of contact time was probable for all methods during the filling and emptying of channels throughout the various steps of reprocessing. A total of 250 ml tap w a t e r was used as a final rinse to flush out the disinfectant. Scopes were dried only at the end of the day, for 10 minutes with all proximal ports and valve openings blocked and for 10 minutes with ports open. Manufacturer directions were always followed for scope hook-up when mechanical disinfection was used. Before ETO sterilization the scope was p!aced'.in a foam nest for safe transport and handling, after all valves had been removed and the gas cap h a d been placed on the scope to allow for ga s flow to avoid damage to the distal end of the scope.; The foam nest was then w r a p p e d and ETO indicator tape w a s used for verification of ETO cycling. Er~doscopes were processed ~hrough:,the standard ET0,hospital cycling, which consisted of a 2-hour expostire at 540 ~C with a standard ! 2 : 8 8 Freon/ETO mix (Penngas; Pennsylvania Engineering Co., Philadelphia, Pa.) ]n a Castle ETO sterilization unit (Castle. MDT, ~Rochester, N.Y.). Sterilization was followed by a 30-minute aeration within the sterilizer at 54 ° C and a final 12-hour aeration at r o o m temperature outside the sterilization unit. Recovery procedures
Recovery from each channel paralleled the procedures used to inoculate the endoscopes. SBDW was substituted for inoculum and air was allowed through until no m o r e of this S B D W "carrier" could be collected in the sterile plastic collection bottles from the distal end of the scope. The exception was from the suction channel. F o r this channel, the suction trap was connected to the suction port and v a c u u m supply. The distal tip of the scope was placed in a sterile container with 15 ml of SBDW. The assistant depressed the suction valve at the control head as the v a c u u m source was slowly turned on, until the entire 15 ml rinse h a d passed through the endoscope and w a s collected in t h e sterile .~ucti0n t r a p . Additionally,-direct
December 1992
recovery trials were completed for each test bacterium and each endoscope. These trials consisted of a syringe of inoculum followed by a syringe full of air. Once effluent was collected, serial dilutions were made a n d the samples were plated in 100 × 15 m m petri plates with SMA. Plates were then incubated for 24 hours (13. subtilis) or 48 hours (P. cepacia)at 35 ° C. Plates with counts between 25 and 250 cfu were read and the average was calculated for duplicate plates. Counts of plates with less than 25 cfu were also recorded, but these were used only w h e n no dilution yielded plate counts within the 25 to 250 cfu range. Residual disinfectant/sterilant testing from endoscope
Samples of residual levels of chemical disinfect a m within each separate channel of the endoscope were collected after disinfection. To obtain r e s i d u a l chemical levels, S B D W in the same volumes used for recovery of bacterial samples w a s inoculated and directly recovered from the, endoscope to be~tested. For hydrogen peroxide residual levels, t h e s e samples were analyzed along with a blank by means of CHEM.et (0 to 1 and 1 to !0 ppm) and VACUette (0 to 25 and 25 to 250 ppm) colorimetric kits (FeSCN; CHEMetrics, Inc., Calverton, Va.). This test kit is based on hydroge n peroxide oxidation of ferrous iron to the ferric state, resuking in the formation of a red thiocyanate complex. Glutaraldehyde residual levels Were obtained in the same m a n n e r as hydrogen peroxide levels, except that analysis was completed with a glutaraldehyde colorimetric test (LaMotte Chemical Products Co., Chestertown, Md.). Scopes were tested for residual glutaraldehyde after disinfection followed by rinses with a series of different; rinse temperatures. Four trials were run after tepid rinses, two were run after hot rinses (44 ° C), and two were run after cold rinses (13 ° C). ETO residual levels were obtained by injecting air through the sterilized/ventilated endoscope while measuring the expelled air with a spectrophotometer (Miran 1A Portable Gas Analyzer; Wilks Scientific Corp., Norwalk, Conn.), which was connected to a plotter. Worker exposure to 2 % activated glutaraldehyde
The spectrophotometer was also used in a separate series of manual disinfection schemes w i t h h o t and cold water rinses, as well a s during
Volume 20
Number6 m a c h i n e disinfection, to m e a s u r e exposure levels of personnel working on the disinfection of these scopes. Throughout these trials, the r o o m air exchange rate remained at 15 air changes per hour. Disinfectant immersion baths always rem a i n e d covered, except w h e n uncovering was necessary for infusing disinfectant through endoscope channels o r during other necessary scope manipulation. RESULTS
B. subtilis was far m o r e resistant to all disinfection methods and disinfectants than was P. cepacia. P. cepacia was so readily killed, flushed, or both from the scopes by all cleaning and disinfection methods that no statistically significant differences between products or methods could be ascertained. Because of its hardiness, B. subtilis was chosen for comparison of various disinfection procedures. The different channels of the endoscope were treated as independent experimental units. Nonparametric tests were used because the within cell variances were quite different for the four channels even w h e n the data were subjected to loga, rithmic transformations. First a F r i e d m a n test was performed on the four combinations of the two factors, disinfectant (two levels) and device (two levels), u n d e r investigation. Plate counts in which no reduction was observed were replaced by total bacterial count from direct recovery. The Wilcoxon signed-rank test was applied to examine the effects of device given the disinfectant and of the disinfectant given the device; each comparison w a s interest for its own sake. Adjustment for multiple testing was therefore not warranted. For ease of presentation, the n u m b e r of times the results were in favor of hydrogen peroxide or in favor of the mechanical device are given, although Wilcoxon's signed rank test takes the a m o u n t of the difference between the pair of observations into account. Parametric analyses were carried out on the log transforms of the bacterial recovery rates. Each different channel of the endoscope was treated as an independent experimental unit. Tables 1 and 2 show the m e a n log reductions for the ERCP and sigmoidoscope trials, respectively. All bacterial recovery rates in the log reduction tables are recorded in cfu per milliliter of inoculum recovered. Log reductions denoted as "greater than" in these two tables represent an average recovery of at least 0.5 cfu/ml. It is evident from Tables 1 and 2 that ETO sterilization proved m o r e efficacious in
Disinfection of flexible endoscopes
295
killing B. subtilis than did high-level disinfection procedures, although a few B. subtilis organisms did survive in the elevator channel of the ERCP scope. Plates for which no reduction was observed, as shown in Table 2, occurred only after m a n u a l cleaning and m a n u a l disinfection procedures and only f r o m the carbon dioxide channel of the sigmoidoscope. Several of these events also occurred with m a n u a l disinfection trials with P. cepacia. The type of scope tested did not make a difference in the efficacy of the disinfection method (Table 3). Because the within-channel icariances were different, separate analyses were carried out for the air and carbon dioxide or elevator channels (1 and 4, respectively). The water and suction channels (2 and 3, respectively) had similar variances and w e r e therefore calculated together. Both the type of disinfectant and the channel were important in the efficacy of high-level disinfection achieved for the water and suction Channels (p < 0.001); the type of disinfectant was statistically significant (p = 0.001) for the air channel. Pair-wise comparisons with the Wilcoxon sign test (Table 4) show the MediVator automatic disinfection to be significantly more efficacious than manual disinfection. Hydrogen peroxide was m o r e efficacious than glutaraldehyde for highlevel disinfection, on the basis of killing or removal of B. subtilis in a 10-minute period. This test took the a m o u n t of the difference into account and was therefore m o r e sensitive than a test ignoring these numerical differences. Although the r~itios may be the same, the p values can differ l~ecause they reflect the n u m e r i c amount of difference between trials. ' When comparing manual disinfection with glu~: taraldehydle or hydrogen peroxide, and m a c h i n e disififection with glutaraldehyde or hydrogen peroxide with the control (manual wash with no disinfection), we found that all four treatments werestatistically different (p < 0.0001). Machine disinfection was shown to be more effective than m a n u a l d i s i n f e c t i o n ( / ) < 0.004) and hydrogen peroxide found to be m o r e effective than glutaraldehyde (p < 0.0001). No advantage was found in use of heated hydrogen peroxide for disinfection in conjunction with m a c h i n e disinfection (p = 0.4). During m a n u a l disinfection, the hydrogen p e r o x i d e p e r f o r m e d better than did the glutaraldehyde in 23 of 24 trials (Table 4). When the nonparametric Friedman test was employed to compare the four groups, machine disinfection with hydrogen peroxide proved most
AJIC December 1992
Vesley et al.
296
1 . Log reduction in B. subtilis, ERCP s c o p e
Table
Hydrogen peroxide Channel Air
Water
Suction
Elevator
Table
Glutaraldehyde
Wash only
Manual
Machine at 50 ° C
Machine at 21 ° C
Manual
Machine
ETO
4.85 4.38 4.37 4.37 4,27 4.35 5.23 4.96 5.04 3.88 3.37 2.84
6.53 6.06 6.31 5.71 5.91 4.51 > 7.64 7.47 6.49 5.71 6.11 5.27
8,05 8,27 6,96 8,48 7.90 6.22 > 8.48 8,05 >6.96 5.17 6.27 5.75
6.64 6.94 7.11 5.77 7.11 > 7.11 7.11 > 7.11 > 7.11 -4.39 4.77 6.81
4,23 6,03 5,10 4 95 5,03 5,13 5,46 5,63 5,95 4.10 4.35 4.95
5.06 5.41 5,89 4.99 5,41 5.70 5.65 6.14 6.48 4.87 4.93 5.02
> 6.96 > 6.95 > 7.92 > 6.96 > 6.95 > 7.92 > 6.96 > 6.95 > 7.92 5.70 > 6.95 7.02
2 . Log reduction in B. subtilis, s i g m o i d o s c o p e
Hydrogen peroxide Channel
Glutaraldehyde
. Wash only
Manual
Machine at 122 ° F
Machine at 70 ° F
Manual
Machine
ETO
4.46 4.38 3.81 4.46 4.25 4,15 4.81 4.56 4.93 4.18 4,52 NRO
5.96 6.17 5.88 6.03 6.06 5.25 > 7.64 5.41 > 7.92 6.39 5.69 5.95
5.88 5.71 6.56 6.60 6.58 > 6.96 > 8.48 8.48 > 6.96 6.97 6.8i 6.96
6.57 7.11 > 6.73 6.81 6.64 > 6.73 > 7.11 > 7.11 > 6.73 6.81 6.72 6.73
4.81 4.42 4.42 4.57 4.92 4.27 6.49 5.85 5.88 5.24 4.88 NRO
4.77 4.98 5.48 4.99 5.18 5,44 5.97 6,34 6.93 5.23 5.44 5.53
> 6,96 > 6,95~ > 7.92 > 6.96 > 6.95 > 7.92 > 6.96 >6.95 > 7.92 > 6,96 > 6.95 > 7.92
Air
Water
Suction
Carbon dioxide
NRO, No reduction observed.
effective, followed by manual disinfection with hydrogen peroxide, machine disinfection with glutaraldehyde, and finally manual disinfection with glutaraldehyde. Residual levels of glutaraldehyde in the endoscopes after manual disinfection and final rinses with hot (44 ° C), cold (13 ° C) and tepid (33 ° C) water were inconsistent and somewhat dependent on thechannel tested. Dependence on the channel is consistent with results of bacteriologic testing, which show that the channel tested was a significant factor in the disinfection procedure (Table 3). Residual levels in the air channel were consistently 5 ppm or less, regardless of rinse temperature. Water channel concentrations averaged 12 ppm, although one value after a cold rinse was 35 ppm. The suction channel appears to be most difficult to rinse, with an average residual glutaraldehyde concentration of 48 pprn.-:The l a r g e r lumen of this channel and greater surface area for absorption may contribute to difficulty in rinsing
out the residual chemicalsJ 4 The carbon dioxide and' elevator channels had average residual levels of 10 ppm, with a range from less than 5 ppm to 20 ppm. More sampling pertaining to the residual levels found in the endoscopes after various temperature rinses should be completed in the future. Residual levels of disinfectant or sterilant were also obtained for endoscopes after each type of disinfecting regimen. These average residual chemical disinfectant levels are shown in Fig. 1. Glutaraldehyde residuals after machine disinfection with a 4 minute, 8 second rinse cycle averaged 8.4 ppm, as compared with the 69 ppm residual after manual disinfection. Residual levels of hydrogen peroxide after machine disinfection averaged 77.4 ppm, as compared with 88.3 ppm after manual disinfection with hydrogen peroxide. In both cases, machine disinfection resuked in lower residual levels than did manual disinfection (Fig. 1).
Volume 20 Number 6
Table
Disinfection of flexible endoscopes
297
3 . Analysis of variance with B. subtilis as the test m i c r o o r g a n i s m
Air channel Disinfectant Within Device Within Disinfectant x device Within Water and suction channels Channel Within Disinfectant Within Device Within .Channel x disinfectant Channel x device Disinfectant x device Channel x disinfectant x device Elevator or co2 channel Disinfectant Within Device Within Disinfectant x device Within
SS
df
MS
F
p Value
71.96 6.35 12.55 6.3 1.17 3.38
1 5 1 5 1 5
71.96 1.27 12.55 1.26 1.17 0.68
56.69
0.001
9.95
0.025
1.74
0.25
2507.251 31.03 111.18 7.62 27.49 26.1 0.28 6.27 '3.13 4.61
1 10 1 10 1 10 1 1 1 1
2507.25 3.1 111.18 0.76 27.49 2.61 0.28 6.27 3.13 4.61
808.14
< 0.001
146
< 0.001
49.88 15.92 10.69 26.09 2.26 14.45
1 5 1 5 1 5
49.88 3.18 10.69 5.22 2.26 2.89
10.53
0.009
0.36 2.4 2.69 3.96
0.56 0.15 0.13 0.075
15.67
0.011
2.05
0.21
0.78
0.42
SS, Sum of squares; MS, mean square; df, degrees of freedom.
ETO residual levels averaging 66.2 p p m were found in the endoscopes, despite a standard 12.5 hour aeration period after sterilization. The suction channel proved t o h a v e the highest residuals, with levels of 190'ppm, 105 ppm, more t h a n 200 ppm, and 105 ppm. Fig. 2 shows that personnel glutaraldehyde exposure during a simulation of manual disinfection of an endoscope exceeded the 0.2 p p m Occupational Safety and Health Administration ceiling limit in five out of six trials with a 4.3 ° C cold water rinse (mean, 0.5 _+ 0.205). In all trials with a 43.3 ° C hot w a t e r rinse, the 0.2 p p m Ceiling limit was exceeded (mean, 3.1 _ 0.6 ppm). The average breathing zone glutaraldehyde exposure was approximately six times higher with hot water rinsing of the flexible endoscope than with cold water rinsing. The Miran 1A Portable Gas Analyzer was also used to test for glutaraldehyde levels directly over and around the MediVator ED105 during its operation. Glutaraldehyde levels measured were consistently less than 0.2 p p m for both the automatic disinfection and automatic rinse cycles. DISCUSSION
In this study-the MediVator Disinfector (ED 105) proved to be m o r e efficacious and consistent in reducing B. subtilis counts than did manual
Table
4 . Pair-wise comparisons with Wilcoxon
Sign Test of disinfectant regimens for B. subtilis
Comparisons Glutaraldehyde (mechanical better than manual) Hydrogen peroxide (mechanical better than manual Manual disinfection (Hydrogen peroxide better than glutaraldehyde) Mechanical disinfection (Hydrogen peroxide better than glutaraldehyde) Mechanical disinfection with hydrogen peroxide (21 ° C better than 50° C) Mechanical better than manual Hydrogen peroxide better than glutaraldehyde Interaction between product an~ method
Ratio of trials
p Value
20/24
< 0,001
19/24
0.004
23/24
<0.0001
22/24
< 0.0001
10/24
0.4
20/24
0.0004
24/24
< 0,0001
16/24
>0.3
Difference in p va ues despite like ratios reflects Wilcoxontest's accounting for the numeric differences betweentrials. ' q
disinfection. Additionally, the pressure in the channels created by the ED105 apparently prevented air locks from occurring i n the carbon dioxide channel; this failure was not detected after any of the mechanical disinfection protocols but
AdlC December 1992
2911 Vesley et al.
400- /1 E 350Q. -~ 300-
]
m¢O
250200- /1 n/1 -~ 150._t2 E 100~- 50- /1 C) J
f E'TO
........ v--,f MacWGlu
7 MaNGlu
Ma ch l HP
----~ Man/HP
Treatment Regimen
- - ~ Air
~
Water- ~ ]
Suction ~
Special ]
Fig, 1, Average residual chemical levels within channels after rinse.
20 18
E ~. 16. uJ 14a >-1- 12tu Q 10.-I
< rl-
8
~ _.1
42i (3-6"c)
r
i
i ('42.4.5 * C )
i
RINSE WATER TEMPERATURE ('C)
[--~- Breathing Zone
~
Sink Top
>K Water Level
]
Fig. 2. Glutaraldehyde personnel exposure with manual disinfection.
occurred in three of 18 manual trials with the sigmoidoscope. This 17% failure rate could cause contamination between patients, resulting in nosocomial infections. The endoscope used for ERCP was not associated with the air-lock problem, yet this particular scope has an elevator channel that needs to be cleaned and disinfected separately even when the rest of the instrument is disinfected in a machine. Unfortunately, the airlock problem or simple omission of cleaning and disinfecting of the elevator channel would be detectable only after infection had resulted from this cross-contamination. It is thus obvious that a clear understanding of the scopes' internal mechanisms is necessary to ensure both consis-
tency and improved cleaning and disinfection methodsF 5 Hydrogen peroxide was significantly more effective in the high-level disinfection of the flexible endoscopes than was the 2% glutaraldehyde solution used in these trials. This conclusion is based on destruction, removal, or both of B. subtilis and does not imply that glutaraldehyde would necessarily be less effective against vegetative bacteria, fungi, or viruses. No distinction was made between inactivation of the bacteria and removal. Hydrogen peroxide could be simply removing the B. subtilis spores more efficaciously than glutaraldehyde. However, hydrogen peroxide is usually not recommended by the manufacturers of flexible
Volume 20
Number6 endoscopes because there continues to be some doubt as to the compatibility of hydrogen peroxide with endoscopes and the possibility that its oxidizing properties may be harmful to some materials of the instruments. Many experts involved with endoscopy are calling for manufacturers to redesign flexible endoscopes with materials resistant enough to withstand chemical disinfection and with removable parts to allow for ease of reliable cleaning and high-level disinfection procedures. 16-1a The apparent advantages of hydrogen peroxide to glutaraldehyde as a high-level disinfectant, which we have demonstrated, give added credence to that recommendation. Heated hydrogen peroxide is particularly efficacious in destroying heat-sensitive microorganisms; however, in these trials there was no advantage to heating the hydrogen peroxide to 50 ° C for destruction of B. subtilis, a result that was unexpected and has no obvious explanation. ETO sterilant was also tested. ETO is not routinely used for endoscope sterilization because of the lengthy turnaround time; the high cost of endoscopes prohibits medical facilities from owning multiple scopes. Although ETO sterilization proved highly effective in our study, a mere 2% of the respondents in a recent questionnaire on disinfection practices for endoscopes and other semicritica] items reported that ETO was the method they used in the disinfection of gastrointestinal endoscopes, whereas 59% used a 2% glutaraldehyde solution. 6 The number of bacteria recovered after ETO sterilization in our trials was extremely low. Some viable spores were recovered from the elevator channel of the ERCP scope, however, suggesting that even this timeconsuming process cannot always guarantee sterility. The residual levels of ETO, averaging 66.2 ppm and ranging from 0 to greater than 200 ppm even after a standard 12.5 hour degassing time, are of greater concern. These test results provide evidence that scopes should be manually aired out by inoculating each channel with air to eliminate residual ETO before they are used on patients. High-level disinfection is defined as "a procedure that inactivates all fungi, viruses, and vegetative microorganisms, but not necessarily all bacterial spores, ''19 and glutaraldehyde thus may be appropriate for this purpose. Despite the fact that many institutions use a glutaraldehyde or glntaraldehyde-phenol-phenate mixture in the high-level disinfection of flexible endoscopes, debate continues as to which type of glutaraldehyde product is best. 2° With this wide use, glutaralde-
Disinfection of flexible endoscopes 299 hyde has been implicated frequently in chemical sensitization of workers. 7'8'21 The MediVator ED 105 reduces employee exposure to glutaraldehyde and efficiently supplies clinics and hospitals with a consistent, standardized method of disinfection, including a microbial filtered (0.05 ~m) water rinse. Additionally, flow of disinfectant was more reliable and no air locks were observed during these trials, whereas air locks were observed several times during manual disinfection. Variances in glutaraldehyde temperature between automatic and manual methods could have contributed to the improved results with automatic disinfection. The pump and other electronic components of the MediVator unit have been shown to elevate the temperature of the fluid in the disinfectant reservoir to 25 ° to 27 ° C after the first cycle of the day in earlier testing at Mayo Clinic. Because significant temperature differences during hydrogen peroxide disinfection did not significantly alter results, however, the most likely conclusion is that superior results with the automatic processing were due to the reliable and continuous pei-fusion of endoscope channels. Thus there is convincing evidence of the machine's advantages over manual processes. 6' 22 Drawbacks of automatic disinfection include the cost of the equipment and the fact that a manual cleaning of the endoscope must first be completed before the endoscope undergoes the disinfection process. This cleaning process may be subject to human error and failure because of instrument and cleaning tool design. Special channels, such as the elevator channel on the JF-10 ERCP, must still be disinfected manually. Record-keeping and other errors must be guarded against regardless of treatment method. Finally, some automatic disinfecting machines have had problems with bacterial growth within reservoirs in the machinery, exacerbating transmission of disease through endoscopes treated with this disinfection process. 23 The MediVator disinfector has a unique tap-water filter system for rinse cycles and is designed to guard against providing a reservoir for bacterial growth. This is highly preferable to rinsing endoscopes with unfiltered tap water, as reported by 54% of survey respondents. 6 PeriOdic sampling of the reservoir during these trials did not detect any contamination. The cleaning and disinfection responsibilities of endoscopic personnel should be taken seriously. It is of utmost importance to have specially trained personnel and a quality control system to maintain high-level disinfection of the scopes, regardless of
300
Vesley et al.
methods or products used. Walter Bond of Centers for Disease Control has pointed out, "In virtually every reported instance of disease transmission associated with endoscopes, there has been a major error in either the cleaning, disinfecting or sterilizing of the instrument.' ,24 Trained personnel are needed tO inspect scopes for damage, manual cleaning, disinfection, rinsing, drying, and proper storage to prevent damage to the endoscopes and disease transmission between patients. Our trials were limited by a lack of organic material buildup within the various channels. Such a buildup is common and could compromise disinfection of flexible endoscopes.13" 25 In hospital settings other factors besides the efficacy of highlevel disinfection procedures may increase the risk of nosocomial infections, including the degree of debilitation of the patient. All of the risks and benefits of various combinations of products and methods and procedures for cleaning and disinfection of flexible endoscopes must be considered in institutional decisions. A special thanks to Dorothy Aeppli, PhD, from the Division of Biostatistics in the School of Public Health at the University of Minnesota, who performed the statistical analyses on our data.
References
1. Botornan VA, Surawicz CM. Bacteremia with gastrointestinal endoscopic procedures. Gastrointest Endosc 1986; 32:342-6. 2. Earnshaw JJ, Clark AW, Thorn BT. Outbreak of Pseudomonas aeruginosa following endoscopic retrograde cholangiopancreatography. J Hosp Infect 1985;6:95-7. 3. Cryan EMJ, Falkiner FR, Mulvihill TE, Keane CT, Keeling PWN. Pseudomonas aeruginosa following endoscopic retrograde cholangiopancreatography. J Hosp Infect 1984;5: 371-6. 4. Beecham HJ III, Cohen ML, Parkin WE. Salmonella typhimurium-transmission by fiberoptic upper gastrointestinal endoscope. JAMA 1979;241:1013-5. 5. Dwyer DM, Klein EG, Istre GR, Robinson MG, Neumann DA, McCoy GA. Salmonella newport infections transmitted by fiberoptic colonoscopy. Gastrointest Endosc 1987;33: 84-7. 6. Rutala WA, Clontz EP, Weber DJ, Hoffmann KK. Disinfection practices for endoscopes and other semicritical items. Infect Control Hosp Epidemiol 1991;12:282-8. 7. Fowler JF Jr. Allergic contact dermatitis from glutaraldehyde exposure. J Occup Med 1989;31:852-3. 8. Wiggins P, McCurdy SA, Zeidenberg W. Epistaxis due to glutaraldehyde exposure. J Occup Med 1989;31:854-6.
AJIC December 1992
9. Corrado Off, Osman J, Davies RJ. Asthma and rhinitis after exposure to gkttaraldehyde in endoscopy units. Hum Toxicol 1986;5:325-7. 10. American Conference of Governmental Industrial Hygienists. Documentation of the threshold limit values and biological exposure indices. 8th ed. Cincinnati: American Conference of Governmental Industrial Hygienists, 1989. 11. Struelens MJ, Rost F, Loriers M, et al. Septicemia after ERCP: outbreak linked to an automatic endoscope disinfecting machine. Abstracts of the internationalsymposium on chemical germicides. Washington, DC: American Society for Microbiology, 1990: poster/abstract no 34, p 21. 12. Arret B, Kirshbaum A. A rapid disc assay method for detecting penicillin in milk. J Milk Food Technol 1959;22: 329-30. 13. Bond WW, Ott BJ, Franke KA, McCracken JE. Effective use of liquid chemical germicides on medical devices: instrument design problems. In: Block SS. Disinfection, sterilization, and preservation. Philadelphia: Lea and Febiger, 1991:1097-106. 14. Power EGM, Russell AD. Glutaraldehyde: its uptake by sporing and non-sporing bacteria, rubber, plastic, and an endoscope. J Appl Bacteriol 1989;67:329-42. 15. Ott BJ, Nelson B. Understanding the structure and function of endoscope channels: the inside story. Soc Gastrointest Assist J 1987;9:184-7. 16. Gorse GJ, Messner RL. Infection control practices in gastrointestinal endoscopy in the United States: a national survey. Infect Control Hosp Epidemiol 1991;12:289-96. 17. Babb JR, Bradley CR. Decontamination of fiberoptic endoscopes: an update. 1 Sterile Serv Mgmt 1988;5:9-11. 18. Bond WW, Favero MS, Mackel DC, Mallison GF. Sterilization or disinfection of flexible fiberoptic endoscopes [Letter]. AORN J 1979;30:350-2. 19. Spaulding EH. Chemical disinfection of medical and surgical materials. In: Lawrence CA, Block SS, eds. Disinfection, sterilization and preservation. Philadeplhia: Lea and Febiger, 1968:517-31. 20. Gurevich I, Yarmelli B, Ctmha BA. The disinfectant dilemma revisited. Infect Control Hosp Epidemiol 1990; 11:96-100. 21. Burge PS. Occupational risks of glutaraldehyde-may cause respiratory, nasal, and skin problems at low concentration. BMJ 1989;299:342. 22. Frank U, Daschner FD. Disinfection in gastrointestinal endoscopy: current status. Endoscopy 21 1989;21:276-9. 23. Alvarado CJ, Stolz SM, Maki DG. Nosocomial P. aeruginosa infections from contaminated endoscopes, Abstracts of the international symposium on chemical germicides. Washington, DC: American Society for Microbiology, 1990. 24. McGregor P, Connell P. Recommended guidelines for infection control in gastrointestinal endoscopy setting. Rochester, New York: Society of Gastroenterology Nurses and Associates, 1990:1-8. 25. Favero MS. Strategies for disinfection and sterilization of endoscopes: the gap between basic principles and actual practice. Infect Control Hosp Epidemiol 1991; 12:279-81.
Background: Many nosocomial infection Outbreaks have been linked to improper
disinfection of the flexible endoscopes used in hospitals and clinics. The objective of this study was to evaluate the effÉcacyof scope disinfection with glutaraldehyde and hydrogen peroxide in manual and mechanical protocols. Methods: Bacillus subtilis and Pseudomonas cepacia were the test organisms. Each channel in two different endoscopes was seeded and evaluated separately. Residual chemical germicide levels in the channels and in .the .work environment We,re also :' measured. ,: ' " : " ~ ": Results: Parametric analyses Were carried out on log transformations of number of colony-forming units recovered. Repeated measures analysis demonstrated that both the type of disinfectant and the method of washing were significant factors for disinfection. Conclusions: Hydrogenperoxide proved to be more efficacious than glutaraldehyde for killing or rernoving B. Subtilis in a 10-minute contact period. Automatic disinfection was more efficacious than manual disinfection for killing or removing B. subtilis in a 10-minute contac~ Period. The ichannel being disinfected also proved to be a significant factor, with cba-bondioxide: and eieqator channels the most difficult to disinfect consistently. (~IIC AM'J INFECTCONTROL199.2;20:29!-300) :',
Cases of B~¢t6remia c a u s e d b y the use of c o n t a m i n a t e d ejr f l d o s c o p e s d u r i n g e n d o s c o p i c retr o g r a d e c h o l m a g i o p a n c r e a t o g r a p h y (ERCP) a n d sigmoidoscop~¢ p r o c e d u r e s a r e well d o c u m e n t e d in i n f e c t i o n 90ntrol l i t e r a t u r e . 1 S a l m o n e l l a a n d P s e u d o m o n a ( species h a v e b e e n i m p l i c a t e d in m a n y of thege n o s o c o m i a l infections. 2-5 The inc r e a s e of t u b e r c u l o s i s a n d H I V cases has f u r t h e r i n c r e a s e d the c o n c e r n a b o u t p r o p e r d i s i n f e c t i o n o r sterilization t e c h n i q u e s f o r these e n d o s c o p e s . 6 The use of flexible e n d o s c o p e p r o c e d u r e s in hospitals a n d clinics t h r o u g h o u t the U n i t e d States is r a p i d l y i n c r e a s i n g . B e c a u s e of the h i g h cost From the School of Public Health, University of Minnesota, Minneapolis, a the Mayo Clinic, Rochester, ~ and the Department of Environmental Health and Safety, University of Minnesota, Minneapolis. Supported by MediVators Inc., Cannon Falls, Minnesota. Reprint requests: Donald Vesley, PhD, Division of Environmental and Occupational Health, University of Minnesota, Box 197 Mayo, 420 Delaware St. S.E., Minneapolis, MN 55455.
17/46/39566
of these i n s t r u m e n t s , t h e r e is a v i t a l n e e d for c l e a n i n g a n d high-level d i s i n f e c t i o n o r sterilization p r o c e d u r e s w i t h m i n i m a l t u r n a r o u n d time. The p r o c e s s is c o m p l i c a t e d b y the a r r a y of channels in these i n s t r u m e n t s ; s o m e are e x t r e m e l y small in d i a m e t e r , w i t h l u m e n e n t r y p o r t s difficult to a c c e s s individually. I n addition, it is i m p o s s i b l e t o e n s u r e the p e r f u s i o n of all l u m e n s d u r i n g r o u t i n e ' r e p r o c e s s i n g , p a r t i c u l a r l y w h e n t h e r e ma); be slight v a r i a t i o n s in the t e c h n i q u e f r o m individual to i n d i v i d u a l a n d f r o m d a y to day. P r o c e d u r e s c u r r e n t l y u s e d i n c l u d e e t h y l e n e o x i d e (ETO) gas s t e r i l i z a t i o n a n d l i q u i d c h e m i c a l soaking. ETO sterilization has a h i g h t u r n a r o u n d t i m e of 24 hours, w h i c h m a k e s its use i m p r a c t i c a l in m o s t e n d o s c o p y units. G l u t a r a l d e h y d e a n d h y d r o g e n p e r o x i d e in form u l a t i o n s a n d c o n c e n t r a t i o n s c a p a b l e of sterilization a r e the u s u a l liquid c h e m i c a l s used. A s h o r t e r c o n t a c t t i m e is u s e d w i t h these c h e m i c a l s to a c h i e v e high-level'disinfection. R e p o r t s of w o r k e r sensitivity to g l u t a r a l d e h y d e , 7-9 the e s t a b l i s h m e n t of the O c c u p a t i o n a l S a f e t y a n d H e a l t h Adminis291
292
Vesley et al.
tration permissible ceiling exposure limit for glutaraldehyde of 0.2 ppm, 1° and a concern about chemical waste m a n a g e m e n t have further complicated this option. As a result, current attention has been focused on high level disinfection procedures with 6% hydrogen peroxide, both at room temperature as a high-level disinfectant and at elevated temperatures (50 ° C) as an Environmental Protection A g e n c y - a p p r o v e d sterilant. Additionally, septicemia after ERCP has been linked to the use of automatic endoscope disinfecting machines.n There is thus a need to evaluate critically mechanical endoscope disinfecting machines. The availability of the MediVator ED 105 mechanical disinfector (MediVators, Inc., Cannon Falls, Minn.), with high-pressure penetration of the disinfectant into the n a r r o w channels, provided an opportunity for comparative testing of the various treatment modes against appropriate microorganisms in a standard manner. METHODS
Pseudomonas cepacia (kingii) American Type Culture Collection (ATCC) 25609 and Bacillus subtilis ATCC 6633 were obtained from ATCC (Rockville, Md.). The freeze-dried culture of P. cepacia was aseptically rehydrated with trypticase soy broth (TSB); this. rehydrated mixture was added to 50 ml TSB, which was then incubated for 24 hours at 35 ° C. After incubation, P. cepacia was washed, vortexed, and spun d o w n at 4000 r p m two times before the pellet was placed in 1.5 L sterile buffered distilled water (SBDW). This mixture was placed in a dark cabinet for 48 hours to stabilize the population. For each day on which trials w e r e run, bacterial counts were obtained by serial dilution followed by enumeration of the bacteria on duplicate pour plates with standard methods agar (SMA). Between trials, the bacterial Solution was stored in a dark cabinet at r o o m temperature. Stock cultures were maintained on SMA at 4 ° C. B. subtilis ATCC 6633 freeze-dried culture was aseptically rehydrated with TSB; this rehydrated mixture was added to 50 ml TSB, which was then incubated for 24 hours at 35 ° C along with SMA-prepared Plates. Cultures were maintained on SMA at 4 ° C. Rapid disk assay method 12 was used to enumerate the B. subtilis and to obtain sufficient bacterial counts within the 106 to 108 colony-forming units (cfu)/ml range. A sample of this prepared mixture was stained by the Schaeffer-Fulton method, a modification of the Ziehl-Neelsen method, with
AJIC December 1992
malachite green a n d a safranine counterstain. Both spores and a limited nuinber of vegetative cells were visible with phase-contrast microscopy. The two endoscopes used in this study included an end-viewing sigmoidoscope (CF-P10S) and a side-viewing e n d o s c o p e used for ERCP (ERCP/JF-10 or duodenoscope), both manufactured by Olympus Optical Co., Ltd. (Tokyo, Japan). The 60 c m sigmoidosc0pe has air, water, suction, and carbon dioxide channels; the 120 cm ERCP scope has air, water, suction, and elevator channels. Reflux of body fluids from the patient m a y occur in any of the standard channels; it is therefore of utmost importance to clean and disinfect all channels of the endoscope thoroughly b e t w e e n patients. Both endoscopes were approximately the s a m e age and both were in excellent near-new condition, as demonstrated by visual inspection of the biopsy and suction channels. 13 The a m o u n t of inoculum needed to fill each channel was determined by trial and error with tap water. It w a s found that the suction channel could hold 15 ml, the air and water channels could hold 5 ml each, the c a r b o n dioxide channel could hold 3 ml, and the elevator channel could hold 1 ml. For consistency, these same volumes were used for inocula in each trial, as well as for S B D W rinse used for recovery after each individual treatment. Volumes of the inoculum and the S B D W rinse solutions were measured with a pipette. Delivery into the air, water, and special (carbon dioxide and elevator) channels-was accomplished with a syringe. The suction channel was also inoculated with a syringe, but recovery was accomplished by connecting the endoscope to a v a c u u m p u m p and drawing the sample through in the m a n n e r described in the Endoscope Channel Sampling Kit.* All trials were carried out in the same manner. The o r d e r in which the channels were inoculated was as follows: 1, air; 2, water; 3, suction; and 4, elevator or carbon dioxide channel. The order used for recovery was the same as that used for the inoculum. After the endoscope was seeded (inoculated) with the appropriate bacteria to be tested (B. subtilis or P. cepacia), it was manually washed and put through the disinfection or sterilization process t o be tested. Endoscopes were mechanically disinfected b e t w e e n each trial. *O-FI channel sampling kit for flexible endoscopes. MediVators, Inc. 1652 Greenview Dr. S.W., P.O. Box 6431, Rochester, MN 55903.
Volume 20
Number6 Inoculation of the endoscope Two individuals, an experimenter and an assistant, were required to carry out the inoculation for each endoscope trial. To inoculate the scope, the air inlet at the end of, the universal cord was blocked with the plug provided in the sampling kit. A 35 ml syringe was placed on the water port adapter and tightly attached to the water bottle port on the endoscope. For the air channels of the insertion tube and universal cord, 5 ml of inoculure was placed in the syringe while the assistant placed a finger over the air/water valve. The 5 ml was then injected into the scope. Another five ml of inoculum was placed in the same syringe and injected into the endoscope while the assistant depressed the air/water valve, which filled the water channels. Fifteen ml was then placed in the 35 ml syringe, which was connected to the suction port of the endoscope with a rubber tubing adaptor. The assistant then depressed the suction valve at the control head while the experimenter injected the inoculum into the scope. For the elevator channel of the duodenoscope, one ml of inoculum was placed in a 3 ml syringe that had been firmly attached with an adaptor to the elevator port of the endoscope. This inoculum was then injected into the endoscope. To inoculate the carbon dioxide channel, 3 ml was transferred by pipette into the 35 ml syringe, which had been connected to the carbon dioxide port in the universal cord. The carbon dioxide valve was depressed at the control head while the inoculum was slowly injected into the scope. Once inoculation of the scope was complete, either the inoculure was directly recovered for the control trials or the scope was put through the manual cleaning process followed by scope disinfection. Manual cleaning procedures After it was seeded with the test organism, the endoscope was placed in a laboratory sink. Liquid enzymatic cleaning solution (Endozyme; Ruhof Corp., Valley Stream, N.Y.) was then added to the wash water in accordance with manufacturer recommendations. Exterior surfaces of the endoscope were washed in this enzymatic solution with a soft-bristle toothbrush and clean cloth. The scope was then connected to the light source and the air/water valve was activated to ensure unblocked channels. The air/water valve was replaced with the air/water cleaning adaptor provided by the manufacturer. All accessible channels of the scope, in particular the suction channel of the universal cord, were thoroughly brushed. A
Disinfection of flexible endoscopes 293 small, hand-held cervical cytology brush was then used in the suction channel valve opening, biopsy port opening, and the suction tubing connection on the umbilical cord. During cleaning of the duodenoscope, the special cleaning h o s e was attached at the head of the scope for washing and rinsing the elevator channel. A 3 ml syringe was used to deliver 9 ml for each washing and rinsing of this elevator channel. The all-channe! irrigator was then attached and the scope was completely submerged in the cleaning solution and water. A 60 ml syringe was used to infuse cleaning solution through the scope. The wash and tap water rinses each used 250 ml of the solution. Disinfection/sterilization procedures After the manual washing was complete, the endoscope was put through a test disinfectant treatment. The test treatments were as follows: (1) manual disinfection with a 2% activated glutaraldehyde solution (MetriCide MX-1400, Metrex Research Corp., Parker, Colo.), (2) a mechanical disinfection (MediVator ED-105; MediVators, Inc.) with 2% Metricide at room temperature, (3) manual disinfection with a 6% hydrogen peroxide/0.85% phosphoric acid solution (EndoSpore; Globe Medical Instruments, Inc., Clearwater, Fla.), (4) mechanical disinfection with 6% hydrogen peroxide/0.85% phosphoric acid solution a t room temperature, (5) mechanical disinfection with 6% hydrogen peroxide/0.85% phosp h o r i c acid at 50 ° C, (6) ETO sterilization in a standard hospital cycle, and (7) manual wash followed by no disinfection. Each treatment was repeated three times for each of the two types of endoscopes. Between each trial, scopes were manually washed and automatically disinfected. B. subtilis and P. cepacia were always used in separate tests. Glutaraldehyde and hydrogen peroxide solutions were fresh at the beginning of the experiment; no solution was used for any purpose other than this experiment. Disinfectant solutions were replaced every week and glutaraldehyde was periodically checked with Sterilog colorimeter strips (Ihjmah Corp., Somerville, N.J.), which measure,whether the glutaraldehyde concentration is at least 1%. If the endoscope was to be manually disinfected after the manual wash described previously, the all-channel irrigator remained attached to the scope and tl~e scope was then submerged in the appropriate disinfectant. The 60 mA syringe was used to infuse 180 ml disinfectant through the
A31C 294
Vesley et al.
scope. If the duodenoscope was being disinfected, 9 ml disinfectant was infused separately into the elevator channel with the 3 ml syringe. The scope was then left to soak in the disinfectant for 10 minutes. Ten minutes was used for the liquid chemical contact time because it is the time m o s t often used in gastrointestinal endoscopy labs a n d because a longer contact time is believed to be impractical during m o s t endoscopy schedules. Although it was not measured, an additional 1 to 1.5 minutes of contact time was probable for all methods during the filling and emptying of channels throughout the various steps of reprocessing. A total of 250 ml tap w a t e r was used as a final rinse to flush out the disinfectant. Scopes were dried only at the end of the day, for 10 minutes with all proximal ports and valve openings blocked and for 10 minutes with ports open. Manufacturer directions were always followed for scope hook-up when mechanical disinfection was used. Before ETO sterilization the scope was p!aced'.in a foam nest for safe transport and handling, after all valves had been removed and the gas cap h a d been placed on the scope to allow for ga s flow to avoid damage to the distal end of the scope.; The foam nest was then w r a p p e d and ETO indicator tape w a s used for verification of ETO cycling. Er~doscopes were processed ~hrough:,the standard ET0,hospital cycling, which consisted of a 2-hour expostire at 540 ~C with a standard ! 2 : 8 8 Freon/ETO mix (Penngas; Pennsylvania Engineering Co., Philadelphia, Pa.) ]n a Castle ETO sterilization unit (Castle. MDT, ~Rochester, N.Y.). Sterilization was followed by a 30-minute aeration within the sterilizer at 54 ° C and a final 12-hour aeration at r o o m temperature outside the sterilization unit. Recovery procedures
Recovery from each channel paralleled the procedures used to inoculate the endoscopes. SBDW was substituted for inoculum and air was allowed through until no m o r e of this S B D W "carrier" could be collected in the sterile plastic collection bottles from the distal end of the scope. The exception was from the suction channel. F o r this channel, the suction trap was connected to the suction port and v a c u u m supply. The distal tip of the scope was placed in a sterile container with 15 ml of SBDW. The assistant depressed the suction valve at the control head as the v a c u u m source was slowly turned on, until the entire 15 ml rinse h a d passed through the endoscope and w a s collected in t h e sterile .~ucti0n t r a p . Additionally,-direct
December 1992
recovery trials were completed for each test bacterium and each endoscope. These trials consisted of a syringe of inoculum followed by a syringe full of air. Once effluent was collected, serial dilutions were made a n d the samples were plated in 100 × 15 m m petri plates with SMA. Plates were then incubated for 24 hours (13. subtilis) or 48 hours (P. cepacia)at 35 ° C. Plates with counts between 25 and 250 cfu were read and the average was calculated for duplicate plates. Counts of plates with less than 25 cfu were also recorded, but these were used only w h e n no dilution yielded plate counts within the 25 to 250 cfu range. Residual disinfectant/sterilant testing from endoscope
Samples of residual levels of chemical disinfect a m within each separate channel of the endoscope were collected after disinfection. To obtain r e s i d u a l chemical levels, S B D W in the same volumes used for recovery of bacterial samples w a s inoculated and directly recovered from the, endoscope to be~tested. For hydrogen peroxide residual levels, t h e s e samples were analyzed along with a blank by means of CHEM.et (0 to 1 and 1 to !0 ppm) and VACUette (0 to 25 and 25 to 250 ppm) colorimetric kits (FeSCN; CHEMetrics, Inc., Calverton, Va.). This test kit is based on hydroge n peroxide oxidation of ferrous iron to the ferric state, resuking in the formation of a red thiocyanate complex. Glutaraldehyde residual levels Were obtained in the same m a n n e r as hydrogen peroxide levels, except that analysis was completed with a glutaraldehyde colorimetric test (LaMotte Chemical Products Co., Chestertown, Md.). Scopes were tested for residual glutaraldehyde after disinfection followed by rinses with a series of different; rinse temperatures. Four trials were run after tepid rinses, two were run after hot rinses (44 ° C), and two were run after cold rinses (13 ° C). ETO residual levels were obtained by injecting air through the sterilized/ventilated endoscope while measuring the expelled air with a spectrophotometer (Miran 1A Portable Gas Analyzer; Wilks Scientific Corp., Norwalk, Conn.), which was connected to a plotter. Worker exposure to 2 % activated glutaraldehyde
The spectrophotometer was also used in a separate series of manual disinfection schemes w i t h h o t and cold water rinses, as well a s during
Volume 20
Number6 m a c h i n e disinfection, to m e a s u r e exposure levels of personnel working on the disinfection of these scopes. Throughout these trials, the r o o m air exchange rate remained at 15 air changes per hour. Disinfectant immersion baths always rem a i n e d covered, except w h e n uncovering was necessary for infusing disinfectant through endoscope channels o r during other necessary scope manipulation. RESULTS
B. subtilis was far m o r e resistant to all disinfection methods and disinfectants than was P. cepacia. P. cepacia was so readily killed, flushed, or both from the scopes by all cleaning and disinfection methods that no statistically significant differences between products or methods could be ascertained. Because of its hardiness, B. subtilis was chosen for comparison of various disinfection procedures. The different channels of the endoscope were treated as independent experimental units. Nonparametric tests were used because the within cell variances were quite different for the four channels even w h e n the data were subjected to loga, rithmic transformations. First a F r i e d m a n test was performed on the four combinations of the two factors, disinfectant (two levels) and device (two levels), u n d e r investigation. Plate counts in which no reduction was observed were replaced by total bacterial count from direct recovery. The Wilcoxon signed-rank test was applied to examine the effects of device given the disinfectant and of the disinfectant given the device; each comparison w a s interest for its own sake. Adjustment for multiple testing was therefore not warranted. For ease of presentation, the n u m b e r of times the results were in favor of hydrogen peroxide or in favor of the mechanical device are given, although Wilcoxon's signed rank test takes the a m o u n t of the difference between the pair of observations into account. Parametric analyses were carried out on the log transforms of the bacterial recovery rates. Each different channel of the endoscope was treated as an independent experimental unit. Tables 1 and 2 show the m e a n log reductions for the ERCP and sigmoidoscope trials, respectively. All bacterial recovery rates in the log reduction tables are recorded in cfu per milliliter of inoculum recovered. Log reductions denoted as "greater than" in these two tables represent an average recovery of at least 0.5 cfu/ml. It is evident from Tables 1 and 2 that ETO sterilization proved m o r e efficacious in
Disinfection of flexible endoscopes
295
killing B. subtilis than did high-level disinfection procedures, although a few B. subtilis organisms did survive in the elevator channel of the ERCP scope. Plates for which no reduction was observed, as shown in Table 2, occurred only after m a n u a l cleaning and m a n u a l disinfection procedures and only f r o m the carbon dioxide channel of the sigmoidoscope. Several of these events also occurred with m a n u a l disinfection trials with P. cepacia. The type of scope tested did not make a difference in the efficacy of the disinfection method (Table 3). Because the within-channel icariances were different, separate analyses were carried out for the air and carbon dioxide or elevator channels (1 and 4, respectively). The water and suction channels (2 and 3, respectively) had similar variances and w e r e therefore calculated together. Both the type of disinfectant and the channel were important in the efficacy of high-level disinfection achieved for the water and suction Channels (p < 0.001); the type of disinfectant was statistically significant (p = 0.001) for the air channel. Pair-wise comparisons with the Wilcoxon sign test (Table 4) show the MediVator automatic disinfection to be significantly more efficacious than manual disinfection. Hydrogen peroxide was m o r e efficacious than glutaraldehyde for highlevel disinfection, on the basis of killing or removal of B. subtilis in a 10-minute period. This test took the a m o u n t of the difference into account and was therefore m o r e sensitive than a test ignoring these numerical differences. Although the r~itios may be the same, the p values can differ l~ecause they reflect the n u m e r i c amount of difference between trials. ' When comparing manual disinfection with glu~: taraldehydle or hydrogen peroxide, and m a c h i n e disififection with glutaraldehyde or hydrogen peroxide with the control (manual wash with no disinfection), we found that all four treatments werestatistically different (p < 0.0001). Machine disinfection was shown to be more effective than m a n u a l d i s i n f e c t i o n ( / ) < 0.004) and hydrogen peroxide found to be m o r e effective than glutaraldehyde (p < 0.0001). No advantage was found in use of heated hydrogen peroxide for disinfection in conjunction with m a c h i n e disinfection (p = 0.4). During m a n u a l disinfection, the hydrogen p e r o x i d e p e r f o r m e d better than did the glutaraldehyde in 23 of 24 trials (Table 4). When the nonparametric Friedman test was employed to compare the four groups, machine disinfection with hydrogen peroxide proved most
AJIC December 1992
Vesley et al.
296
1 . Log reduction in B. subtilis, ERCP s c o p e
Table
Hydrogen peroxide Channel Air
Water
Suction
Elevator
Table
Glutaraldehyde
Wash only
Manual
Machine at 50 ° C
Machine at 21 ° C
Manual
Machine
ETO
4.85 4.38 4.37 4.37 4,27 4.35 5.23 4.96 5.04 3.88 3.37 2.84
6.53 6.06 6.31 5.71 5.91 4.51 > 7.64 7.47 6.49 5.71 6.11 5.27
8,05 8,27 6,96 8,48 7.90 6.22 > 8.48 8,05 >6.96 5.17 6.27 5.75
6.64 6.94 7.11 5.77 7.11 > 7.11 7.11 > 7.11 > 7.11 -4.39 4.77 6.81
4,23 6,03 5,10 4 95 5,03 5,13 5,46 5,63 5,95 4.10 4.35 4.95
5.06 5.41 5,89 4.99 5,41 5.70 5.65 6.14 6.48 4.87 4.93 5.02
> 6.96 > 6.95 > 7.92 > 6.96 > 6.95 > 7.92 > 6.96 > 6.95 > 7.92 5.70 > 6.95 7.02
2 . Log reduction in B. subtilis, s i g m o i d o s c o p e
Hydrogen peroxide Channel
Glutaraldehyde
. Wash only
Manual
Machine at 122 ° F
Machine at 70 ° F
Manual
Machine
ETO
4.46 4.38 3.81 4.46 4.25 4,15 4.81 4.56 4.93 4.18 4,52 NRO
5.96 6.17 5.88 6.03 6.06 5.25 > 7.64 5.41 > 7.92 6.39 5.69 5.95
5.88 5.71 6.56 6.60 6.58 > 6.96 > 8.48 8.48 > 6.96 6.97 6.8i 6.96
6.57 7.11 > 6.73 6.81 6.64 > 6.73 > 7.11 > 7.11 > 6.73 6.81 6.72 6.73
4.81 4.42 4.42 4.57 4.92 4.27 6.49 5.85 5.88 5.24 4.88 NRO
4.77 4.98 5.48 4.99 5.18 5,44 5.97 6,34 6.93 5.23 5.44 5.53
> 6,96 > 6,95~ > 7.92 > 6.96 > 6.95 > 7.92 > 6.96 >6.95 > 7.92 > 6,96 > 6.95 > 7.92
Air
Water
Suction
Carbon dioxide
NRO, No reduction observed.
effective, followed by manual disinfection with hydrogen peroxide, machine disinfection with glutaraldehyde, and finally manual disinfection with glutaraldehyde. Residual levels of glutaraldehyde in the endoscopes after manual disinfection and final rinses with hot (44 ° C), cold (13 ° C) and tepid (33 ° C) water were inconsistent and somewhat dependent on thechannel tested. Dependence on the channel is consistent with results of bacteriologic testing, which show that the channel tested was a significant factor in the disinfection procedure (Table 3). Residual levels in the air channel were consistently 5 ppm or less, regardless of rinse temperature. Water channel concentrations averaged 12 ppm, although one value after a cold rinse was 35 ppm. The suction channel appears to be most difficult to rinse, with an average residual glutaraldehyde concentration of 48 pprn.-:The l a r g e r lumen of this channel and greater surface area for absorption may contribute to difficulty in rinsing
out the residual chemicalsJ 4 The carbon dioxide and' elevator channels had average residual levels of 10 ppm, with a range from less than 5 ppm to 20 ppm. More sampling pertaining to the residual levels found in the endoscopes after various temperature rinses should be completed in the future. Residual levels of disinfectant or sterilant were also obtained for endoscopes after each type of disinfecting regimen. These average residual chemical disinfectant levels are shown in Fig. 1. Glutaraldehyde residuals after machine disinfection with a 4 minute, 8 second rinse cycle averaged 8.4 ppm, as compared with the 69 ppm residual after manual disinfection. Residual levels of hydrogen peroxide after machine disinfection averaged 77.4 ppm, as compared with 88.3 ppm after manual disinfection with hydrogen peroxide. In both cases, machine disinfection resuked in lower residual levels than did manual disinfection (Fig. 1).
Volume 20 Number 6
Table
Disinfection of flexible endoscopes
297
3 . Analysis of variance with B. subtilis as the test m i c r o o r g a n i s m
Air channel Disinfectant Within Device Within Disinfectant x device Within Water and suction channels Channel Within Disinfectant Within Device Within .Channel x disinfectant Channel x device Disinfectant x device Channel x disinfectant x device Elevator or co2 channel Disinfectant Within Device Within Disinfectant x device Within
SS
df
MS
F
p Value
71.96 6.35 12.55 6.3 1.17 3.38
1 5 1 5 1 5
71.96 1.27 12.55 1.26 1.17 0.68
56.69
0.001
9.95
0.025
1.74
0.25
2507.251 31.03 111.18 7.62 27.49 26.1 0.28 6.27 '3.13 4.61
1 10 1 10 1 10 1 1 1 1
2507.25 3.1 111.18 0.76 27.49 2.61 0.28 6.27 3.13 4.61
808.14
< 0.001
146
< 0.001
49.88 15.92 10.69 26.09 2.26 14.45
1 5 1 5 1 5
49.88 3.18 10.69 5.22 2.26 2.89
10.53
0.009
0.36 2.4 2.69 3.96
0.56 0.15 0.13 0.075
15.67
0.011
2.05
0.21
0.78
0.42
SS, Sum of squares; MS, mean square; df, degrees of freedom.
ETO residual levels averaging 66.2 p p m were found in the endoscopes, despite a standard 12.5 hour aeration period after sterilization. The suction channel proved t o h a v e the highest residuals, with levels of 190'ppm, 105 ppm, more t h a n 200 ppm, and 105 ppm. Fig. 2 shows that personnel glutaraldehyde exposure during a simulation of manual disinfection of an endoscope exceeded the 0.2 p p m Occupational Safety and Health Administration ceiling limit in five out of six trials with a 4.3 ° C cold water rinse (mean, 0.5 _+ 0.205). In all trials with a 43.3 ° C hot w a t e r rinse, the 0.2 p p m Ceiling limit was exceeded (mean, 3.1 _ 0.6 ppm). The average breathing zone glutaraldehyde exposure was approximately six times higher with hot water rinsing of the flexible endoscope than with cold water rinsing. The Miran 1A Portable Gas Analyzer was also used to test for glutaraldehyde levels directly over and around the MediVator ED105 during its operation. Glutaraldehyde levels measured were consistently less than 0.2 p p m for both the automatic disinfection and automatic rinse cycles. DISCUSSION
In this study-the MediVator Disinfector (ED 105) proved to be m o r e efficacious and consistent in reducing B. subtilis counts than did manual
Table
4 . Pair-wise comparisons with Wilcoxon
Sign Test of disinfectant regimens for B. subtilis
Comparisons Glutaraldehyde (mechanical better than manual) Hydrogen peroxide (mechanical better than manual Manual disinfection (Hydrogen peroxide better than glutaraldehyde) Mechanical disinfection (Hydrogen peroxide better than glutaraldehyde) Mechanical disinfection with hydrogen peroxide (21 ° C better than 50° C) Mechanical better than manual Hydrogen peroxide better than glutaraldehyde Interaction between product an~ method
Ratio of trials
p Value
20/24
< 0,001
19/24
0.004
23/24
<0.0001
22/24
< 0.0001
10/24
0.4
20/24
0.0004
24/24
< 0,0001
16/24
>0.3
Difference in p va ues despite like ratios reflects Wilcoxontest's accounting for the numeric differences betweentrials. ' q
disinfection. Additionally, the pressure in the channels created by the ED105 apparently prevented air locks from occurring i n the carbon dioxide channel; this failure was not detected after any of the mechanical disinfection protocols but
AdlC December 1992
2911 Vesley et al.
400- /1 E 350Q. -~ 300-
]
m¢O
250200- /1 n/1 -~ 150._t2 E 100~- 50- /1 C) J
f E'TO
........ v--,f MacWGlu
7 MaNGlu
Ma ch l HP
----~ Man/HP
Treatment Regimen
- - ~ Air
~
Water- ~ ]
Suction ~
Special ]
Fig, 1, Average residual chemical levels within channels after rinse.
20 18
E ~. 16. uJ 14a >-1- 12tu Q 10.-I
< rl-
8
~ _.1
42i (3-6"c)
r
i
i ('42.4.5 * C )
i
RINSE WATER TEMPERATURE ('C)
[--~- Breathing Zone
~
Sink Top
>K Water Level
]
Fig. 2. Glutaraldehyde personnel exposure with manual disinfection.
occurred in three of 18 manual trials with the sigmoidoscope. This 17% failure rate could cause contamination between patients, resulting in nosocomial infections. The endoscope used for ERCP was not associated with the air-lock problem, yet this particular scope has an elevator channel that needs to be cleaned and disinfected separately even when the rest of the instrument is disinfected in a machine. Unfortunately, the airlock problem or simple omission of cleaning and disinfecting of the elevator channel would be detectable only after infection had resulted from this cross-contamination. It is thus obvious that a clear understanding of the scopes' internal mechanisms is necessary to ensure both consis-
tency and improved cleaning and disinfection methodsF 5 Hydrogen peroxide was significantly more effective in the high-level disinfection of the flexible endoscopes than was the 2% glutaraldehyde solution used in these trials. This conclusion is based on destruction, removal, or both of B. subtilis and does not imply that glutaraldehyde would necessarily be less effective against vegetative bacteria, fungi, or viruses. No distinction was made between inactivation of the bacteria and removal. Hydrogen peroxide could be simply removing the B. subtilis spores more efficaciously than glutaraldehyde. However, hydrogen peroxide is usually not recommended by the manufacturers of flexible
Volume 20
Number6 endoscopes because there continues to be some doubt as to the compatibility of hydrogen peroxide with endoscopes and the possibility that its oxidizing properties may be harmful to some materials of the instruments. Many experts involved with endoscopy are calling for manufacturers to redesign flexible endoscopes with materials resistant enough to withstand chemical disinfection and with removable parts to allow for ease of reliable cleaning and high-level disinfection procedures. 16-1a The apparent advantages of hydrogen peroxide to glutaraldehyde as a high-level disinfectant, which we have demonstrated, give added credence to that recommendation. Heated hydrogen peroxide is particularly efficacious in destroying heat-sensitive microorganisms; however, in these trials there was no advantage to heating the hydrogen peroxide to 50 ° C for destruction of B. subtilis, a result that was unexpected and has no obvious explanation. ETO sterilant was also tested. ETO is not routinely used for endoscope sterilization because of the lengthy turnaround time; the high cost of endoscopes prohibits medical facilities from owning multiple scopes. Although ETO sterilization proved highly effective in our study, a mere 2% of the respondents in a recent questionnaire on disinfection practices for endoscopes and other semicritica] items reported that ETO was the method they used in the disinfection of gastrointestinal endoscopes, whereas 59% used a 2% glutaraldehyde solution. 6 The number of bacteria recovered after ETO sterilization in our trials was extremely low. Some viable spores were recovered from the elevator channel of the ERCP scope, however, suggesting that even this timeconsuming process cannot always guarantee sterility. The residual levels of ETO, averaging 66.2 ppm and ranging from 0 to greater than 200 ppm even after a standard 12.5 hour degassing time, are of greater concern. These test results provide evidence that scopes should be manually aired out by inoculating each channel with air to eliminate residual ETO before they are used on patients. High-level disinfection is defined as "a procedure that inactivates all fungi, viruses, and vegetative microorganisms, but not necessarily all bacterial spores, ''19 and glutaraldehyde thus may be appropriate for this purpose. Despite the fact that many institutions use a glutaraldehyde or glntaraldehyde-phenol-phenate mixture in the high-level disinfection of flexible endoscopes, debate continues as to which type of glutaraldehyde product is best. 2° With this wide use, glutaralde-
Disinfection of flexible endoscopes 299 hyde has been implicated frequently in chemical sensitization of workers. 7'8'21 The MediVator ED 105 reduces employee exposure to glutaraldehyde and efficiently supplies clinics and hospitals with a consistent, standardized method of disinfection, including a microbial filtered (0.05 ~m) water rinse. Additionally, flow of disinfectant was more reliable and no air locks were observed during these trials, whereas air locks were observed several times during manual disinfection. Variances in glutaraldehyde temperature between automatic and manual methods could have contributed to the improved results with automatic disinfection. The pump and other electronic components of the MediVator unit have been shown to elevate the temperature of the fluid in the disinfectant reservoir to 25 ° to 27 ° C after the first cycle of the day in earlier testing at Mayo Clinic. Because significant temperature differences during hydrogen peroxide disinfection did not significantly alter results, however, the most likely conclusion is that superior results with the automatic processing were due to the reliable and continuous pei-fusion of endoscope channels. Thus there is convincing evidence of the machine's advantages over manual processes. 6' 22 Drawbacks of automatic disinfection include the cost of the equipment and the fact that a manual cleaning of the endoscope must first be completed before the endoscope undergoes the disinfection process. This cleaning process may be subject to human error and failure because of instrument and cleaning tool design. Special channels, such as the elevator channel on the JF-10 ERCP, must still be disinfected manually. Record-keeping and other errors must be guarded against regardless of treatment method. Finally, some automatic disinfecting machines have had problems with bacterial growth within reservoirs in the machinery, exacerbating transmission of disease through endoscopes treated with this disinfection process. 23 The MediVator disinfector has a unique tap-water filter system for rinse cycles and is designed to guard against providing a reservoir for bacterial growth. This is highly preferable to rinsing endoscopes with unfiltered tap water, as reported by 54% of survey respondents. 6 PeriOdic sampling of the reservoir during these trials did not detect any contamination. The cleaning and disinfection responsibilities of endoscopic personnel should be taken seriously. It is of utmost importance to have specially trained personnel and a quality control system to maintain high-level disinfection of the scopes, regardless of
300
Vesley et al.
methods or products used. Walter Bond of Centers for Disease Control has pointed out, "In virtually every reported instance of disease transmission associated with endoscopes, there has been a major error in either the cleaning, disinfecting or sterilizing of the instrument.' ,24 Trained personnel are needed tO inspect scopes for damage, manual cleaning, disinfection, rinsing, drying, and proper storage to prevent damage to the endoscopes and disease transmission between patients. Our trials were limited by a lack of organic material buildup within the various channels. Such a buildup is common and could compromise disinfection of flexible endoscopes.13" 25 In hospital settings other factors besides the efficacy of highlevel disinfection procedures may increase the risk of nosocomial infections, including the degree of debilitation of the patient. All of the risks and benefits of various combinations of products and methods and procedures for cleaning and disinfection of flexible endoscopes must be considered in institutional decisions. A special thanks to Dorothy Aeppli, PhD, from the Division of Biostatistics in the School of Public Health at the University of Minnesota, who performed the statistical analyses on our data.
References
1. Botornan VA, Surawicz CM. Bacteremia with gastrointestinal endoscopic procedures. Gastrointest Endosc 1986; 32:342-6. 2. Earnshaw JJ, Clark AW, Thorn BT. Outbreak of Pseudomonas aeruginosa following endoscopic retrograde cholangiopancreatography. J Hosp Infect 1985;6:95-7. 3. Cryan EMJ, Falkiner FR, Mulvihill TE, Keane CT, Keeling PWN. Pseudomonas aeruginosa following endoscopic retrograde cholangiopancreatography. J Hosp Infect 1984;5: 371-6. 4. Beecham HJ III, Cohen ML, Parkin WE. Salmonella typhimurium-transmission by fiberoptic upper gastrointestinal endoscope. JAMA 1979;241:1013-5. 5. Dwyer DM, Klein EG, Istre GR, Robinson MG, Neumann DA, McCoy GA. Salmonella newport infections transmitted by fiberoptic colonoscopy. Gastrointest Endosc 1987;33: 84-7. 6. Rutala WA, Clontz EP, Weber DJ, Hoffmann KK. Disinfection practices for endoscopes and other semicritical items. Infect Control Hosp Epidemiol 1991;12:282-8. 7. Fowler JF Jr. Allergic contact dermatitis from glutaraldehyde exposure. J Occup Med 1989;31:852-3. 8. Wiggins P, McCurdy SA, Zeidenberg W. Epistaxis due to glutaraldehyde exposure. J Occup Med 1989;31:854-6.
AJIC December 1992
9. Corrado Off, Osman J, Davies RJ. Asthma and rhinitis after exposure to gkttaraldehyde in endoscopy units. Hum Toxicol 1986;5:325-7. 10. American Conference of Governmental Industrial Hygienists. Documentation of the threshold limit values and biological exposure indices. 8th ed. Cincinnati: American Conference of Governmental Industrial Hygienists, 1989. 11. Struelens MJ, Rost F, Loriers M, et al. Septicemia after ERCP: outbreak linked to an automatic endoscope disinfecting machine. Abstracts of the internationalsymposium on chemical germicides. Washington, DC: American Society for Microbiology, 1990: poster/abstract no 34, p 21. 12. Arret B, Kirshbaum A. A rapid disc assay method for detecting penicillin in milk. J Milk Food Technol 1959;22: 329-30. 13. Bond WW, Ott BJ, Franke KA, McCracken JE. Effective use of liquid chemical germicides on medical devices: instrument design problems. In: Block SS. Disinfection, sterilization, and preservation. Philadelphia: Lea and Febiger, 1991:1097-106. 14. Power EGM, Russell AD. Glutaraldehyde: its uptake by sporing and non-sporing bacteria, rubber, plastic, and an endoscope. J Appl Bacteriol 1989;67:329-42. 15. Ott BJ, Nelson B. Understanding the structure and function of endoscope channels: the inside story. Soc Gastrointest Assist J 1987;9:184-7. 16. Gorse GJ, Messner RL. Infection control practices in gastrointestinal endoscopy in the United States: a national survey. Infect Control Hosp Epidemiol 1991;12:289-96. 17. Babb JR, Bradley CR. Decontamination of fiberoptic endoscopes: an update. 1 Sterile Serv Mgmt 1988;5:9-11. 18. Bond WW, Favero MS, Mackel DC, Mallison GF. Sterilization or disinfection of flexible fiberoptic endoscopes [Letter]. AORN J 1979;30:350-2. 19. Spaulding EH. Chemical disinfection of medical and surgical materials. In: Lawrence CA, Block SS, eds. Disinfection, sterilization and preservation. Philadeplhia: Lea and Febiger, 1968:517-31. 20. Gurevich I, Yarmelli B, Ctmha BA. The disinfectant dilemma revisited. Infect Control Hosp Epidemiol 1990; 11:96-100. 21. Burge PS. Occupational risks of glutaraldehyde-may cause respiratory, nasal, and skin problems at low concentration. BMJ 1989;299:342. 22. Frank U, Daschner FD. Disinfection in gastrointestinal endoscopy: current status. Endoscopy 21 1989;21:276-9. 23. Alvarado CJ, Stolz SM, Maki DG. Nosocomial P. aeruginosa infections from contaminated endoscopes, Abstracts of the international symposium on chemical germicides. Washington, DC: American Society for Microbiology, 1990. 24. McGregor P, Connell P. Recommended guidelines for infection control in gastrointestinal endoscopy setting. Rochester, New York: Society of Gastroenterology Nurses and Associates, 1990:1-8. 25. Favero MS. Strategies for disinfection and sterilization of endoscopes: the gap between basic principles and actual practice. Infect Control Hosp Epidemiol 1991; 12:279-81.