Introduction
Infection is a potential (severe) complication of endoscopic retrograde cholangiopancreatography
(ERCP), occurring in 1.4%–7.7% of patients, with a mortality rate of 7.8% [1 ]
[2 ]. Infectious complications post-ERCP can result from the translocation of endogenous
intestinal flora during the procedure or the introduction of exogenous microorganisms
via contaminated equipment. Contaminated duodenoscopes have caused multiple nosocomial
outbreaks, mainly involving multidrug-resistant organisms, resulting in cases of illness
and death [3 ]. Studies on duodenoscope contamination rates show significant variation. A recent
meta-analysis reported a contamination rate of 21.5% (95%CI 15.4%–27.6%) in nonoutbreak-initiated
studies [4 ].
A major factor responsible for duodenoscope contamination is biofilm formation. Risk
factors for biofilm formation include reprocessing lapses, delays before reprocessing,
endoscope damage, and insufficient drying [5 ]. Biofilms can reduce the efficacy of high level disinfection (HLD) and may cause
false-negative culture results [5 ]
[6 ]
[7 ]. Once a biofilm has formed in the endoscope channels, it is difficult to remove
and may require channel replacement [8 ].
Manual cleaning of duodenoscopes is considered a critical step in achieving adequate
reprocessing and involves flushing and brushing of endoscope channels [9 ]. Currently, the duodenoscope channel cleaning brushes advised by the duodenoscope
manufacturers consist of a wire with a single cleaning brush. However, an in vitro
study demonstrated that the Endoss “Push and Pull” brush (EPP; JPP50, Endoss BV),
a cleaning brush with a sweeper design, might be more efficient in cleaning duodenoscope
channels [10 ]. In this study, we aimed to evaluate the effect of EPP introduction on the contamination
rate of Pentax ED34-i10T2 duodenoscopes.
Methods
Setting
This retrospective before-and-after intervention study was performed in a large tertiary
care center, the Erasmus MC University Medical Center (Erasmus MC), Rotterdam, The
Netherlands, where approximately 750 ERCP procedures are performed on adult patients
annually. We included culture sets collected from eight Pentax ED34-i10T2 duodenoscopes
(with disposable caps) from March 2018 until June 2022. Reprocessing was performed
by dedicated reprocessing staff according to the manufacturer’s instructions.
Intervention
On December 15, 2020, the EPP was introduced for manual cleaning of the Pentax ED34-i10T2
duodenoscopes and replaced the Pentax single-use brush (CS5522A) (Fig. 1s , see online-only supplementary material).
Sampling
The duodenoscope culture sets consisted of five sample sites. First, the distal tip
of the duodenoscope was swabbed using a Copan Liquid Amies Elution Swab (eSwab; Copan).
Then 20 mL of sterile saline (0.9%) was flushed through each of the suction channel,
biopsy channel, and air water channel and collected separately in sterile containers.
Subsequently, a single-use endoscope cleaning brush (Pentax CS5522A) was pulled through
the suction and biopsy channels. The distal tip of the brush was cut using disinfected
pliers and placed in an eSwab container.
Starting in April 2021, sterile water was used as the flushing fluid instead of saline.
Routine surveillance cultures were taken approximately monthly. Data on the exact
timing of sampling and errors in the sampling process were not available.
Microbiological methods and interpretation
The eSwab containers were vortexed and poured over a sheep blood agar plate (Becton
Dickinson). The flushing fluid was filtered through a 0.22-µm filter (Milliflex Plus
Test System), after which the filter was placed on Reasoners2A agar (Becton Dickinson).
Plates were incubated for 3 days at 35°C. All morphologically distinct microorganisms
were identified and colony-forming units (CFUs) were counted. Identification was performed
using the matrix-assisted laser desorption/ionization time-of-flight analyzer (Bruker).
Contamination was defined in one of two ways: ≥1 CFU of a microorganism of gut or
oral origin (MGO); or ≥20 CFU/20 mL of any microorganisms, including those of waterborne
and skin origin (AM20) [11 ]
[12 ]
[13 ]. Once a duodenoscope tested positive for an MGO, it was quarantined and repeatedly
sampled until it tested negative. If the duodenoscope still tested positive after
three attempts, it was sent to the manufacturer for inspection and possible channel
replacement. From November 2020, MGO-positive duodenoscopes underwent routine borescope
inspections for channel damage and, if necessary, were sent to the manufacturer for
repair.
Subgroup analysis distinguished primary contamination from persistent contamination.
Primary contamination included cases with preceding negative culture sets or the emergence
of different microorganisms. Persistent contamination involved the same microorganisms
at species level across consecutive culture sets. Subgroup analysis excluded culture
sets from duodenoscopes with no patient exposure between sets.
Data collection
A sample size was not calculated as this study involved retrospectively retrieved
data and was not designed to detect a predefined difference. Duodenoscope usage data
were extracted from the endoscopy documentation system Endobase (Olympus) and the
electronic patient records. All available culture set data of the Pentax ED34-i10T2
duodenoscopes were extracted from the electronic laboratory information system of
the Department of Medical Microbiology and Infectious Diseases. The culture set result
was determined by combining the five duodenoscope sample site results. Additionally,
the duodenoscopes’ repair history and maintenance records were obtained from the manufacturer.
Statistical analyses
All analyses were performed using R version 4.1.3 [14 ]. Categorical variables are presented as absolute or relative frequencies (%), while
continuous variables are expressed as the median with the first and third quartile
(Q1, Q3), or as the mean (SD). Point estimates of contamination are accompanied by
Wilson score confidence intervals (95%CIs).
To analyze the effect of the EPP on contamination with MGOs or by the AM20 definition,
logistic mixed-effects regression models were employed, with endoscope-specific random
intercepts incorporated to account for potential correlation between observations
of the same duodenoscope [15 ]. The following covariates, were included: duodenoscope usage since the preceding
culture set, preceding culture set positive for an MGO, preceding culture set positive
by AM20, and duodenoscope usage since the last biopsy channel replacement. The covariates
were selected based on the existing literature and clinical expertise. To facilitate
model estimation, duodenoscope usage since the preceding culture was divided by 10,
and duodenoscope usage since the last biopsy channel replacement was divided by 30.
A subgroup analysis was conducted to assess the impact of the EPP specifically on
primary contamination. Additionally, we used mixed-model analyses to compare the odds
of contamination per sample site. To adjust for the increased risk of type-I errors
due to multiple testing, we applied the Bonferroni correction and set the significance
threshold to P <0.004.
Results
Culture characteristics
A total of 257 culture sets were collected from eight Pentax ED34-i10T2 duodenoscopes.
There were 176 culture sets (68.5%) collected pre-intervention (March, 2018 to December
15, 2020) and 81 culture sets (31.5%) collected during the intervention (December
15, 2020 to June 2022). [Table 1 ] presents an overview of the culture characteristics. The cultured MGOs are listed
in Tables 1s and 2s , and the microorganisms cultured by the AM20 definition in Tables 3s and 4s .
Table 1 Contamination of duodenoscopes before and after the introduction of the Endoss Push
and Pull brush.
Pentax single-use brush (CS5522A)
Endoss Push and Pull brush (JPP50)
AM20, any microbial growth with ≥20 colony forming units (CFUs)/20 mL, including waterborne
or skin-type microorganisms; MGO, ≥1 CFU of a microorganism of gut or oral microorganism;
Q1, quartile 1; Q3, quartile 3.
* Pentax ED34-i10T2.
Contamination
MGO
AM20
MGO
AM20
No
Yes
No
Yes
No
Yes
No
Yes
Duodenoscope* culture sets (n=257), n (%) [95%CI]
96 (54.5) [47.2–61.7]
80 (45.5) [38.3–52.8]
59 (33.5) [27.0–40.8]
117 (66.5) [59.2–73.0]
67 (82.7) [73.1–89.4]
14 (17.3) [10.6–26.9]
6 (7.4) [3.4–15.2]
75 (92.6) [84.8–96.6]
Sample sites (n=1285), n (%) [95%CI]
710 (80.7) [77.9–83.2]
170 (19.3) [16.8–22.1]
628 (71.4) [68.3–74.3]
252 (28.6) [25.7–31.7]
390 (96.4) [94.0–97.7]
15 (3.6) [2.3–6.0]
218 (53.8) [49.0–58.6]
187 (46.2) [41.4–51.0]
Air/water channel (n=257), n (%) [95%CI]
169 (96.0) [92.0– 98.1]
7 (4.0) [1.9–8.0]
161 (91.5) [86.4–94.8]
15 (8.5) [5.2–13.6]
79 (97.5) [91.4–99.3]
2 (2.5) [0.7–8.6]
79 (97.5) [91.4–99.3]
2 (2.5)
Biopsy channel (n=257), n (%) [95%CI]
125 (71.0) [63.9–77.2]
51 (29.0) [22.8–36.1]
90 (51.2) [43.8–58.4]
86 (48.8) [41.2–56.2]
78 (96.3) [89.7–98.7]
3 (3.7) [1.3–10.3]
8 (9.9) [5.1–18.3]
73 (90.1) [81.7–94.9]
Brush (n=257), n (%) [95%CI]
122 (69.3) [62.2–75.7]
54 (30.7) [24.3–37.8]
113 (64.2) [56.9–70.9]
63 (35.8) [29.1–43.1]
72 (89.9) [80.2–94.0]
9 (11.1) [6.0–19.8]
38 (46.9) [36.4–57.7]
43 (53.1) [42.3–63.6]
Forceps elevator (n=257), n (%) [95%CI]
171 (97.1) [93.5–98.8]
5 (2.9) [1.2–6.5]
169 (96.5) [92.0–98.1]
7 (3.5) [1.9–8.0]
80 (98.8) [93.3–99.9]
1 (1.2) [0.1–6.7]
79 (97.5) [91.4–99.3]
2 (2.5)
Suction channel (n=257), n (%) [95%CI]
123 (69.9) [62.7–76.2]
53 (30.1) [23.8–37.3]
95 (54.0) [46.6–61.2]
81 (46.0) [38.8–53.4]
79 (97.5) [91.4–99.3]
2 (2.5) [0.7–8.6]
14 (17.3) [10.1–26.9]
67 (82.7) [73.1–89.4]
Number of MGOs identified per culture set, median [Q1, Q3]
–
1.0
1.0
1.0
–
1.0
1.0
1.0
Number of AM20 identified per culture set, median [Q1, Q3]
2.0
2.0
–
2.0
3.0
4.0
–
3.0
Preceding culture set positive with an MGO (n=91), n (%)
36 (45.6%)
43 (54.4%)
31 (39.2%)
48 (60.8%)
12 (85.7%)
2 (14.3%)
2 (14.3%)
12 (85.7%)
Preceding culture set positive by AM20 (n=192), n (%)
59 (50.9%)
57 (49.1%)
32 (27.6%)
84 (72.4%)
63 (82.9%)
13 (17.1%)
6 (7.9%)
70 (92.1%)
Days since last culture set, median [Q1, Q3)
22.0
15.0
21.0
20.0
29.0
71.0
69.0
29.0
Number of uses since preceding culture set, median [Q1, Q3]
3.0
6.0
1.0
6.0
7.0
11.0
16.5
7.0
Number of uses since biopsy channel replacement, median [Q1, Q3]
48.0
65.0
48.0
56.0
78.0
31.5
55.5
76.0
Days since last biopsy channel replacement, median [Q1, Q3]
185.5
178.0
173.0
202.0
279.0
231.0
224.5
270.0
Contamination with MGOs
The introduction of the EPP statistically significantly reduced the odds of contamination
with an MGO (adjusted odds ratio [aOR] 0.25, 95%CI 0.11–0.58; P =0.001) ([Fig. 1 ]). We did not find a statistically significant association between the odds of contamination
with an MGO and duodenoscope usage since the preceding culture set (aOR 1.10, 95%CI
0.91–1.32; P =0.33) or biopsy channel replacement (aOR 1.01, 95%CI 0.89–1.16; P =0.84). Although not statistically significant, a preceding culture set positive with
an MGO seemed to increase the odds of contamination with an MGO in the subsequent
culture set ([Fig. 1 ]). This effect was similar in our subgroup analysis studying only primary contamination
(Fig. 2s ).
Fig. 1 Forrest plot with results of mixed-model analysis of duodenoscope culture sets by
contamination definition.
During the period that the Pentax single-use brush was used, the distal tip (aOR 0.08,
95%CI 0.03–0.20; P <0.001) and air/water channel (aOR 0.11,
95%CI 0.05–0.24; P <0.001) were associated with lower odds of
contamination with an MGO compared with the biopsy channel ([Fig. 2 ]). In the EPP period, the brush pulled through the biopsy and suction channels had
higher odds of being contaminated, although this effect was not statistically significant
(aOR 3.25, 95%CI 0.81–13.01; P =0.10).
Fig. 2 Forrest plot with results of mixed-model analysis of duodenoscope sample site contamination
with MGOs by type of brush used during manual cleaning.
Contamination according to the AM20 definition
The use of the EPP increased the odds of a positive culture set by AM20 (aOR 4.43,
95%CI 1.57–12.48; P =0.005), but did not reach statistical significance after correction for multiple
testing ([Fig. 1 ]). This effect was also slightly reduced in the subgroup analysis (aOR 3.05, 95%CI
1.03–9.04; P =0.04) (Fig. 2s ).
Duodenoscope usage was not statistically significantly associated with increased odds
of contamination by AM20 ([Fig. 1 ]). Although not statistically significant, a preceding culture set positive by AM20
was associated with higher odds of contamination in the subsequent culture set (aOR
1.86, 95%CI 0.94–3.69; P =0.08). Irrespective of the cleaning brush, the distal tip, air/water channel, and
culture of the brush were associated with lower odds of contamination according to
the AM20 definition compared with the biopsy channel ([Fig. 3 ]).
Fig. 3 Forrest plot with results of mixed-model analysis of duodenoscope sample site contamination
according to the AM20 definition by type of brush used during manual cleaning.
Discussion
After the introduction of the EPP for manual cleaning, we observed a 28.2 percentage
point reduction in contamination with MGOs in Pentax ED34-i10T2 duodenoscopes. This
is a remarkable finding, which has important clinical relevance. Literature reports
on outbreaks highlight the risks associated with contaminated duodenoscopes. Balan
et al. documented 24 outbreaks, involving 490 patients and resulting in over 30 deaths
[3 ]. The minimum base risk of exogenous duodenoscope infections per ERCP procedure has
been estimated to be 0.01% [16 ]. Contamination with an MGO indicates inadequate reprocessing and can occur even
in the absence of identified reprocessing breaches [17 ]. These findings highlight the importance of innovative approaches to improve reprocessing
outcomes.
In our study, Pseudomonas aeruginosa was the most commonly identified MGO, accounting for 14.4% of culture sets (37/257).
P. aeruginosa is notorious for its ability to form biofilms in challenging environments, which
demonstrate a certain level of tolerance to commonly used disinfectants in HLD.
Before the intervention, the duodenoscope contamination rate was 45.4%, significantly
higher than the 22.5% reported in a recent meta-analysis [4 ]. We hypothesize that multiple duodenoscopes harbored a robust P. aeruginosa biofilm, contributing to the elevated contamination rate. The introduction of the
EPP may have eliminated the biofilm, as only one culture set tested positive for P. aeruginosa after its implementation. The EPP’s design, incorporating an additional sweeper,
likely improves circumferential sealing of the duodenoscope channels. This could disrupt
biofilm formation and allow the disinfecting agents used during HLD to reach and eliminate
the embedded bacteria.
Although not statistically significant after correcting for multiple testing, the
introduction of the EPP led to an increase of culture sets contaminated by AM20, up
to 90%. Even though the clinical significance of AM20 contamination is likely low,
the biomatrix of environmental flora may protect MGOs during HLD [5 ]. The increase in AM20 contamination was observed specifically in sample sites treated
with the EPP, namely the biopsy and suction channels. We suggest that the sweeper
of the EPP becomes contaminated according to the AM20 definition during the manual
cleaning process and subsequently contaminates the duodenoscope channels.
Duodenoscope usage or biopsy channel replacement did not seem to influence the odds
of contamination with an MGO or by the AM20 definition. This is in line with the findings
of Rauwers et al. [13 ]. Borescope studies have shown that endoscope biopsy channels are often damaged,
which increases with use and has been associated with higher bacterial attachment
[18 ]
[19 ]. However, the risk of channel damage may depend less on the frequency of use and
more on ERCP characteristics, such as the instruments used.
This study has limitations associated with its before-and-after design [20 ]. Firstly, the order in which the brushes were used was not randomized, and no control
group was available. Therefore, we cannot establish a causal relationship between
the reduction in MGO contamination and the implementation of the EPP. Additionally,
as this study was retrospective, important information, such as the drying time after
reprocessing, the surveillance methods employed, and adherence to reprocessing and
sampling protocols, was not recorded. This may have led to biased estimates of the
impact of using the EPP. Furthermore, it is a single-site study and the EPP was used
with only one type of duodenoscope, limiting the generalizability of our findings
to other settings, types or brands of scope.
In conclusion, in this study, the introduction of the EPP was associated with significantly
lower odds of contamination with an MGO in Pentax ED34-i10T2 duodenoscopes. Therefore,
this seems a promising intervention to reduce contamination rates of ready-to-use
duodenoscopes and improve the prevention of duodenoscope-associated infections. Future
prospective multicenter studies in multiple duodenoscope brands should be performed
to confirm these observations.
