Two clusters of carbapenemase-producing bacteria linked to hospital utility room wastewater drains – an outbreak report

Abstract

Background

Outbreak investigations of two carbapenemase-producing bacteria (CPB)-clusters identified a link to utility room wastewater drains (URWD), prompting an investigation into CPB-colonization of the hospital’s URWD and into utility room management to identify possible breaches in infection prevention and control (IPC) measures.

Methods

At the University Hospital Basel, Switzerland, a cluster of four patients with VIM-1-producing Enterobacter hormaechei and two patients with KPC-3-producing Citrobacter freundii, occurred between 07/2022 – 07/2024.

A multidisciplinary team (IPC specialists, nurses, cleaning experts) evaluated utility room management. Wastewater drains in affected patient rooms and 25 utility rooms of the hospital building (including two wards in which the two transmission-cluster occurred) affected by the clusters were screened for CPB-colonization and whole genome sequencing was performed on clinical and environmental CPB isolates.

Results

The bla_VIM-1 and the bla_KPC-3 gene corresponding to both clusters were identified in sequences of Citrobacter freundii and Klebsiella pneumoniae isolates colonizing the ward’s URWD. Colonized drains were in proximity to affected patient’s rooms, while none of the wastewater drains in the respective patient rooms was colonized with CPB. Overall, CPB were identified in 52% of all URWD. Several breaches in utility room management and design facilitating colonization with CPB and potentially leading to contamination of shared medical equipment by dispersal were observed, including lacking separation between clean and contaminated zones in utility rooms (34%), and improper use of contaminated zones for storage of clean shared medical equipment (88.7%).

Conclusions

URWD serve as reservoirs for CPB. Breaches in utility room management and design, potentially leading to contamination of shared medical equipment by dispersal from contaminated wastewater drains, represent an intervention target for prevention of CPB-transmission.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13756-026-01772-1.

Background

Carbapenemase-producing bacteria (CPB) are an emerging global concern, with carbapenem-resistant Enterobacterales (CRE) and Acinetobacter baumannii (CRAB) classified as critical pathogens by the World Health Organization (WHO) Bacterial Priority Pathogen List 2024 [1].

In Switzerland, CPB are systematically and mandatorily reported to the Federal Office of Public Health, and prevalence of CPB has steadily increased within the past decade from 66 cases in 2013 to 646 cases in 2023 [2]. While Oxacillinase (OXA) type, New-Delhi-Metallo-beta-lactamase (NDM) type and Klebsiella pneumoniae-Carbapenemase (KPC) type are the predominant carbapenemases in Switzerland (N = 329/50.2%, N = 225/34.8% and N = 59/9.1%, respectively, in 2023), Verona-integrated Metallo-beta-lactamase (VIM) accounted for only 4.6% of all CPB isolates in 2023 [2]. Although CPB are still associated with importation from high-endemic areas, our recent data revealed that a considerable proportion of CPB-colonized patients lacked a history of travel outside of Switzerland, suggesting local or even within-institution acquisition [3]. Growing evidence indicates that sanitary installations and hospital wastewater reservoirs may be playing a major role in CPB-transmission [4, 5]. Biofilm formation of CPB in the moist environment of hospital water and wastewater draining systems facilitates survival of bacterial colonies and enables horizontal gene transfer, facilitating further spread of antimicrobial resistance mechanisms between different bacterial populations [6, 7]. The disposal of excretions from CPB-colonized patients into drains is likely a key source of contamination [8]. In addition, improper disposal of nutritional infusions and antibiotics into drainage systems may further promote bacterial growth and exert selection pressure within the drain microbiome [9, 10]. Dispersion of infectious particles through the air from contaminated drains facilitates dissemination of pathogens via contaminated equipment [11, 12] facilitating transmission of CPB within institutions and representing a potential intervention target for infection prevention and control (IPC) measures [5, 13].

In this study we describe two clusters of CPB in an academic Swiss tertiary care hospital linked to CPB colonization of the ward’s utility room wastewater drains (URWDs). We focus on the outbreak investigation, environmental sampling aiming to establish an inventory of the reservoir by assessing the CPB colonization of URWD and the analysis of potential IPC intervention targets in the management of utility rooms.

Methods

Setting

This is a retrospective outbreak report performed at the University Hospital Basel (UHB) in Switzerland. The UHB is a tertiary care centre admitting > 40′000 patients annually. The two clusters affected three medical wards of the same hospital wing, in which patients with haemato-oncological diseases are hospitalized. Each ward comprises up to 45 beds. Single rooms are equipped with en-suite bathrooms, multi-bedrooms usually share a bath- and a shower room on the hallway with one of the neighbouring rooms.

Each ward comprises two to three utility rooms, which are used for disposal of waste, patient’s excretions and for cleaning and storage of medical equipment. A utility room consist of two zones; the contaminated zone includes a wastewater drain (URWD, disposal unit made of a stainless steel hopper with integrated sink and flushing mechanism, supplementary Fig. 1A), used for the disposal of liquid and semi-solid clinical waste (emptying bedpans, urine bottles, disposing of vomit bowls etc.), and an automated preprocessor used for cleaning of bedpans and urinal bottles, the clean zone comprises a sink for hand hygiene as well as storage areas for cleaned medical equipment. Often further medical equipment such as infusion stands or toilet chairs are cleaned and stored inside utility rooms, however, usage and management differ among wards and wings due to different designs and user requirements.

Outbreak definition, case definition and active case finding

At the UHB, a CPB outbreak is defined as the occurrence of two or more patients with detection of the same CPB species harbouring the same CPB gene in any screening or clinical sample with an epidemiological link such as spatial overlap on room, ward or hospital level. [14]. A confirmed case of transmission is defined as any patient colonized or infected with the same CPB-strain as the one recovered from the index patient as confirmed by whole genome sequencing (WGS). Confirmed transmission was defined without a fixed temporal threshold, as all CPB isolates undergo routine, timely sequencing within the UHB surveillance program, allowing genetic links to be identified irrespective of the time interval between cases. These definitions were applied for the two reported clusters.

At the UHB, active surveillance screenings for carriage of multidrug-resistant organisms (MDROs) is performed in patients at risk [15], such as patients returning from other countries (rectal swab, urine culture, groin swab, respiratory specimen, and if applicable swabs of wounds or drainages) or patients with a history of colonization with MDRO (for carbapenemase-producing Enterobacterales the following specimens: rectal swab, urine culture, if applicable swabs from wounds or drainages, for carbapenemase-producing non-fermenting bacteria: rectal swab, groin swab, respiratory specimen, urine if urinary catheter in place, if applicable swabs of wounds and drainages). Furthermore, patients hospitalized on high-risk wards such as the intensive care units (ICU, for more than 24 h) or patients with hematologic malignancies are screened by rectal swab on admission and monthly thereafter.

Interventions

Patient-specific interventions

The following interventions are routinely applied following detection of a CPB-colonized patient and were therefore also applied for the patients within the clusters: (1) Droplet- and contact isolation precautions (surgical mask, gloves and gown) in single-bed rooms [15]. (2) Daily cleaning and disinfection of the patient room and the assigned bathroom / shower room, after patient discharge additionally including ultraviolet C disinfection (UVDI-360, Almedica, Galmiz, Switzerland). (3) Active case finding of direct contact patients using rectal swab screening. A direct contact patient is defined as any patient sharing the same room or bathroom as the index patient at any timepoint within the past 30 days. (4) Three point prevalence screenings using rectal swabs were performed by discretion of the IPC team on affected wards, starting as early as possible after detection of a patient harbouring CPB and then weekly. If nursing staff is clearly dedicated to ward sectors as defined by each individual ward, initial point prevalence screening might be limited to the corresponding sector, since staff cohorting is assumed to limit the risk of transmission, but will be extended to the entire ward as soon as an outbreak is suspected.

Environmental interventions and investigation

The following environmental interventions were applied to control the reported clusters: (1) Environmental CPB screening of in-room water drains and of sink and shower drains of shared bathrooms. (2) Environmental CPB screening of URWDs p-traps of the affected wards as well as the p-trap of the sinks localized in the ward’s pharmacy, offices and locker rooms. Sampling depth was approximately 15 cm into the p-trap. (3) Disinfection of CPB-colonized water drains with 6% hydrogen peroxide (Klercide™ Sporicidal Low Residue Peroxide, Ecolab, Switzerland) over three consecutive nights. (4) Ultraviolet C disinfection of utility rooms with detection of CPB in wastewater drains on affected wards. (5) Installation of covers of the disposal unit on CPB-colonized URWD (supplementary Fig. 1B) and partition walls next to the colonized drain in the ward’s pharmacy. (6) Replacement of the u-bend of the p-trap of CPB-colonized URWD.

Patient-specific and environmental interventions, as well as their timing, are illustrated in Figs. 2 and 3 for both clusters respectively.

Fig. 2.

Timeline VIM-1-producing Enterobacter hormaechei cluster A patient hospitalization and epidemiologic links based on hospital wards B environmental sampling and infection prevention and control measures

Fig. 3.

Timeline of the KPC-3producing Citrobacter freundii cluster A patient hospitalizations and epidemiologic links based on hospital wards B environmental sampling and infection prevention and control measures

Systematic environmental screening of URWDs of all wards of the hospital building, including the two wards affected by the clusters

Screening of 25 URWDs of ten medical wards (each comprising 2–3 utility rooms) localized in the building affected by the two CPB-clusters was performed in July 2024 by swabbing the inner walls of URWDs using eSwabs (length 15 cm, Copan, Brescia, Italy).

To assess potential airborne dispersion of CPB, microbiological sampling was performed before and after flushing of all five URWDs, known to be colonized with any CPB. These URWDs were known to be colonized due to the environmental sampling performed within the context of the outbreak investigation before the systematic screening of URWDs to assess the hospital reservoir was performed and hence were selected for air sampling. One of these five URWDs was already equipped with a drain cover as part of the outbreak control measures, therefore air sampling was performed both with the cover open and closed during flushing to confirm the effectiveness of preventing aerosolization. Air sampling was conducted at approximately 20 cm next to and above the URWD, using MAS 100 Microbial Air-Sampler (MBV, Stäfa, Switzerland), collecting 400L of air per sample on selective chromID® CARBA/OXA chromogenic agar plates (bioMérieux, Marcy-l’Étoile, France).

Hospital wide evaluation of management of utility rooms

Management of utility rooms was evaluated once between May and July 2024 for each utility room (N = 53) across the entire hospital—including, but not limited to, the building affected by the outbreak. Evaluation was performed by a multidisciplinary team including IPC specialists, the ward’s head nurse, nurse assistants and cleaning experts, assessing structural and work-flow related factors possibly promoting transmission of pathogens such as: (1) Missing indication of contaminated and clean zones, (2) storage of clean equipment or shared medical equipment in contaminated zones (2 m radius around URWDs), (3) damaged furniture possibly limiting effectiveness of disinfection, (4) disposal of liquids via the wastewater drain, (5) processes and responsibilities of cleaning and storage of shared medical equipment.

Microbiological analyses

Screening

Screening for CPB was performed using ChromID® Carba Smart agar (bioMérieux, Marcy-l’Etoile, France). Carbapenemase genes for Enterobacterales were tested using eazyplex® Superbug CRE panel detecting OXA-48, OXA-181, KPC, NDM, VIM (Amplex, Giessen, Germany). Pseudomonas spp. were analysed using GeneXpert Carba-R (Cepheid, Thalwil, Switzerland) detecting OXA-48, KPC, NDM, VIM, IMP-1 and Acinetobacter baumannii complex using eazyplex® SuperBug complete, a test system including KPC, VIM, OXA-48, OXA-23, OXA-40 and OXA-58 (Amplex).

Sequencing

Whole genome sequencing (WGS) was performed on a NextSeq500 (Illumina) [16] for short-read data, and on a GridIon (Oxford Nanopore Technologies) with flowcell R10.4.1, the rapid barcoding kit SQK-RBK004 and the basecalling model dna_r10.4.1_e8.2_400bps_hac@v4.2.0 for long-read data.

Assembly

Genome assembly from short reads and allele-based comparison were conducted using Ridom SeqSphere + software (Ridom GmbH, Münster, Germany). Strain relatedness was assessed based on core genome MLST (cgMLST) ad hoc schemes specific to the respective species respectively. Genome assembly from long reads was performed with Flye (v2.9.3) [17], polished once with medaka (v1.12.1, Oxford Nanopore), then polished up to five times with the previously obtained short reads with Pilon (v1.24) [18].

Phylogeny

Reads were first cleaned by removing adaptor sequences, low-quality-end trimming and removal of low-quality reads using BBTools (v39.19; https://github.com/bbushnell/BBTools). Single nucleotide polymorphisms were called using Snippy (v4.6.0; https://github.com/tseemann/snippy) with downsampling to 100 × coverage where needed, a required minimum coverage of 10, a required minimum fraction of 0.7, and with the reference genome for Enterobacter hormaechei subsp. xiangfangensis strain ECC-124 (NCBI accession NZ_CP143716.1). Snippy-core was then used to construct a whole genome alignment and from this the phylogenetic tree constructed by FastTree (v2.1.11; https://github.com/morgannprice/fasttree) with default parameters. The tree was visualised with a custom script in R (v4.3.2; https://www.R-project.org/) using the ape package (v5.8) [19].

Antimicrobial resistance genes

Antimicrobial resistance genes were identified using the AMRFinderPlus tool (NCBI), which provides gene calls based on a curated database of resistance determinants, including beta-lactamases (bla), aminoglycoside-modifying enzymes, and others. Identified genes were categorized into carbapenemases, extended-spectrum beta-lactamases (ESBL), and other beta-lactamases as displayed in Supplementary Tables 1 and 2. Gene sequences were compared for identity by multiple alignment using MAFFT (v7.520) [20]. The accession data of the publicly available sequences of cluster isolates in the European Nucleotide Archive (ENA) are available in Supplementary Table 3.

Plasmid replicon types were determined using PlasmidFinder and integron structures were screened using IntegronFinder for selected isolates.

Statistical analyses

Descriptive analyses were performed using STATA 16.1 software (StataCorp, College Station, TX).

Ethics

This study was assessed by the local ethics committee (Ethikkommission Nordwest- und Zentralschweiz [EKNZ]) as a quality assurance project, thus the requirement for approval does not apply (project no. Req-2023–01529). We adhered to the ORION guidelines for reporting of outbreak investigations [21].

Results

Description of clusters

bla_VIM-1 cluster ward A

In January 2024, a VIM-1 carbapenemase-producing Enterobacter hormaechei sequence type 90 (ST90) was identified in a clinical urine isolate of a patient (aged between 80–90 years, corresponding to patient 3 in Figs. 1 and 2A). Two weeks later, a VIM-1 producing E. hormaechei ST90 was recovered from a rectal swab of another patient (aged between 60–70 years, corresponding to patient 4 in Figs. 1 and 2A), hospitalized on the same ward. This rectal swab was performed as part of monthly routine screening of high-risk patients with hematologic malignancies. Sequencing confirmed clonal relatedness of the two isolates. Comparison of sequencing data with previous CPB isolates identified at the UHB revealed a close genetic link with the isolate of a VIM-1 producing E. hormaechei of two additional patients aged between 50–60 years hospitalized on the same ward in June 2022, and aged between 70–80 years, hospitalized on a different medical ward of the same hospital wing in August 2023 (corresponding to patients 1 and 2 in Figs. 1 and 2A).

Fig. 1.

Phylogenetic tree of the VIM1-producing Enterobacter hormaechei cluster genomes. Edge lengths are in units of substitutions per million sites, node labels show the approximate Bayes branch support

Baseline characteristics of these four patients are provided in Table 1. Of the three patients hospitalized on the same ward, only two had a temporal overlap, without sharing rooms or bathroom facilities (Fig. 2A).

Table 1.

Baseline characteristics of cluster patients

Patient	Date of CPB detection	Age	Sex	Ward	Diagnostic specimen	Previous negative CPB screening	CPB Infection	Underlying hematologic malignancy
VIM-1-producing Enterobacter hormaechei cluster
1	13.07.2022	59	Male	Outpatient hematology	Rectal swab	none	no	DLBCL
2	28.08.2023	77	Male	Medical	Sputum	none	no	DLBCL
3	23.01.2024	86	Male	External hospital, admitted to medical ward	Urine	none	no	Cerebral lymphoma
4	05.02.2024	65	Male	Medical	Rectal swab	12/2023	no	AML
KPC-3-producing Citrobacter freundii cluster
1	06.06.2024	63	Female	Hematology non isolation ward	Rectal swab	03/2024	no	MDS
2	09.07.2024	49	Male	Hematology non isolation ward	Rectal swab	11/2023	no	NK/T-cell Lymphoma

VIM = Verona integron-encoded metallo-β-lactamase, KPC = Klebsiella pneumoniae carbapenemase, DLBCL = diffuse large B-cell lymphoma, AML = acute myeloid leukemia, MDS = myelodysplastic syndrome

Environmental CPB screening detected VIM-producing Citrobacter freundii and Klebsiella pneumoniae in one out of two URWDs located on the affected ward, as well as a VIM-producing C. freundii in the water drain of the ward’s pharmacy (Fig. 2B). Sequencing confirmed presence of the same bla_VIM-1 cluster-gene located on the plasmid of the environmental CPB isolates.

As a result of the detection of these potential environmental reservoirs, the following infection prevention and control measures were implemented (Fig. 2B): All patient-specific interventions were applied. Covers on and partition walls next to colonized drains were installed to limit dispersion of CPB via droplets, colonized drains were disinfected with 6% hydrogen peroxide over three consecutive nights and the colonized u-bend of the URWD’s p-trap was replaced. Sampling one week after disinfection revealed persistent growth of VIM-producing CPB in the URWD, while no CPB were detected in the pharmacy ‘s drain. However, the disinfecting measures were unsuccessful in eliminating CPB colonization, as subsequent sampling three weeks after disinfection detected re-emergence of VIM-producing after three weeks. Eventually, the need for the sink and wastewater drain in the wards pharmacy was evaluated and deemed unnecessary and hence was removed later on. Enhanced staff education on standard hygiene was implemented twice weekly on affected wards by IPC specialists. The outbreak was declared over, as defined by three negative weekly screenings of all patients hospitalized on the affected ward according to national guidelines [14], almost two months after outbreak detection. The dates of point prevalence screenings are shown in Fig. 2.

bla_KPC-3 cluster ward B

In June 2024 a KPC-3 carbapenemase-producing C. freundii was detected in a 63-year-old patient by routine screening of patients with hematologic malignancies. She had previously been tested negative in March and April 2024. Three consecutive point prevalence screenings on the ward did not reveal any further cases. Based on the environmental findings of the previous VIM-1-outbreak, environmental investigation was performed as detailed in Fig. 3B. Colonization with KPC-3-producing C. freundii and K. pneumoniae was detected in one out of three of the ward’s URWD located opposite of the patient’s room, the bla_KPC-3 gene being related to the clinical isolate. One month later, a KPC-3-producing C. freundii was detected by routine screening of a 46-year-old patient, hospitalized in a neighbouring room of the index patient on the same ward (Fig. 3A). The two clinical isolates had a single nucleotide polymorphism (SNP) distance of 43 SNP.

All patient-specific interventions were applied. Notably, screening for CPB of patient room water drains (sinks and showers) was negative. The colonized URWD was disinfected, similarly as detailed for the VIM-1 cluster, the u-bend of the p-trap was replaced, and a cover was installed on the colonized disposal unit. Staff education on standard hygiene was performed by IPC specialists on the affected ward on a regular basis. Three consecutive ward screenings as well as screening of direct contact patients of both cluster patients did not reveal any further cases, therefore the outbreak was declared over almost two months after outbreak detection, according to national guidelines [14]. The dates of point prevalence screenings are shown in Fig. 3.

Environmental investigations

Following the detection of the two CPB-clusters, 25 URWD located in the same hospital wing were screened for colonization with CPB. CPB colonization was detected in 52% of URWD. (Table 2). The predominant CPB colonizing URWD were C. freundii (N = 10, 76.9%) and K. pneumoniae (N = 5, 38.5%) followed by Pseudomonas aeruginosa (N = 2, 15.4%) and Serratia marcescens (N = 1, 7.7%) (Table 2, supplementary Table 1) The most frequent carbapenemase gene was bla_KPC-3 (41.2%) followed by bla_OXA-48 (35.3%) (Table 3). Multiple species of CPB within the same URWD were detected in 31% of CPB-colonized URWD and multiple carbapenemase-genes within one URWD were found in 23%. Supplementary Table 1 provides a detailed overview on URWD screening results. CPB co-expressed extended-spectrum beta-lactamase (ESBL) genes in 38.8%.

Table 2.

Screening of URWD for colonization with carbapenem-resistant bacteria (n = 25)—Overall colonization rates of URWDs by carbapenemase-, beta-lactamase-groups and species

	N	%a
Carbapenemase	13	52.0
Citrobacter freundii	10	76.9
Klebsiella pneumoniae	5	38.5
Pseudomonas aeruginosa	2	15.4
Serratia marcescens	1	7.7
Extended-spectrum beta-lactamase	9	36.0
Citrobacter freundii	2	22.2
Klebsiella pneumoniae	4	44.4
Ochrobactrum intermedium	1	11.1
Pseudomonas aeruginosa	1	11.1
Pseudomonas alcaligenes	1	11.1
Broad/narrow-spectrum beta-lactamase	9	36.0
Aeromonas species	1	11.1
Citrobacter freundii	5	55.6
Klebsiella pneumoniae	3	33.3
Pseudomonas aeruginosa	2	22.2
Pseudomonas alcaligenes	1	11.1
Chromosomal/natural beta-lactamase	13	52.0
Aeromonas species	3	23.1
Citrobacter freundii	5	38.5
Elizabethkingia miricola	1	7.7
Klebsiella pneumoniae	1	7.7
Pseudomonas aeruginosa	4	30.8
Pseudomonas species	1	7.7
Stenotrophomonas maltophilia	4	30.8

URWD = utility room wastewater drain

^aPercentage refers to the proportion of URWDs positive for the corresponding gene or species

Table 3.

Distribution of carbapenemase-genes (A) Among all detected carbapenemase-genes (B) Among all screened URWDs (C) Among all URWD-colonizing species

(A) Proportions of carbapenemase-genes overall (N = 20), N (%)
KPC-2	1 (5.0)
KPC-3	7 (35.0)
GES-5	1 (5.0)
VIM-1	2 (10.0)
VIM-2	1 (5.0)
NDM-1	1 (5.0)
IMP-22	1 (5.0)
OXA-48	6 (30.0)

(B) Proportions of URWDs with detection of the respective carbapenemase-gene (N = 25), N (%)
KPC-2	1 (4.0)
KPC-3	6 (24.0)
GES-5	1 (4.0)
VIM-1	1 (4.0)
VIM-2	1 (4.0)
NDM-1	1 (4.0)
IMP-22	1 (4.0)
OXA-48	5 (20.8)

(C) Proportions of carbapenemase-genes per CPB species (N = 18), N(%)
	Citrobacter freundii	Klebsiella pneumoniae	Pseudomonas aeruginosa	Serratia marcescens
KPC-2	–	1 (5.6)	–	–
KPC-3	6 (33.3)	1 (5.6)	–	–
GES-5	–	–	–	1 (5.6)
VIM-1	1 (5.6)	1 (5.6)	–	–
VIM-2	–	–	1 (5.6)	–
NDM-1	–	1 (5.6)	–	–
IMP-22	–	–	1 (5.6)	–
OXA-48	5 (27.8)	1 (5.6)	–	–

URWDs = utility room wastewater drains, CPB = carbapenemase-producing bacteria, KPC = Klebsiella pneumoniae carbapenemase, GES = Guiana extended-spectrum, VIM = Verona integron-encoded metallo-β-lactamase, NDM = New Delhi metallo- β-lactamase, IMP = imipenemase, OXA = oxacillinase

Screening revealed growth of further carbapenem-resistant bacteria, harbouring ESBL genes or narrow spectrum beta-lactamase genes, with either potentially transmissible or natural chromosomal beta-lactamase genes (Table 2, supplementary Table 1).

Dispersion of CPB through the air was measured before and after flushing the disposal unit in all five utility rooms with CPB-colonized URWDs. One of the five disposal units was already equipped with a cover due to the environmental outbreak control interventions. No growth of CPB was detected in the air sampling cultures before flushing, nor in the one measurement during flushing with closed drain cover. Sampling after flushing and opened drain cover revealed growth of gram-negative bacteria in all five samples, with KPC-3-producing Enterobacter roggenkampii detected in one of the samples of an URWD with known colonization with KPC-producing C. freundii and K. pneumoniae (supplementary Table 2). Species detected in air sampling did correspond to species detected in URWD screening in only one out of five URWDs.

Analysis of management of utility rooms

Fifty-three utility rooms were inspected and audited by a multidisciplinary team. Results are displayed in Table 4. Contaminated and clean zones were clearly indicated in 66% of utility rooms. Fifty-two (98.1%) utility rooms were used as a general storage room for medical equipment and in 52.8% at least one piece of furniture was damaged. In 47 (88%) utility rooms, contaminated zones (2 m radius around URWDs) were used for storage of medical or non-medical equipment such as clean bedpans (41.5%) and clean gloves (24.5%). 90.6% of URWD were regularly used for disposal of patient’s excretions instead of the cleaning machine intended for this purpose. Parenteral nutrients were disposed in URWD in 28.3%. Feedback of analyses of the management of utility rooms was provided to staff head nurses immediately after analysis.

Table 4.

Evaluation of management of utility rooms N = 53

	N	%
Clear indication of contaminated and clean zones	35	66.0
Installation of faucet directly above drain	45	84.9
Storage of medical equipment in utility room	52	98.1
Storage of medical equipment in contaminated zone (2 m radius around URWDs):	47	88.7
Clean bedpans	22	41.5
Clean urinal bottles	13	24.5
Clean washing basins	2	3.8
Hand sanitizer	17	32.1
Disinfection wipes	12	22.6
Gloves	13	24.5
Wall bracket with cleaning brushes	20	37.7
Patient-related equipment	2	3.7
Othera	35	66.0
Damage of furniture:	28	52.8
Cabinets	16	30.2
Stainless steel surfaces	17	32.1
Shelf	5	9.4
Cleaning machine	3	5.7
Waste containers	3	5.7
Disposal of fluids through URWDs:
Parenteral nutrients	15	28.3
Food	2	3.8
Other medication	12	22.6
Dialysate	4	7.6
Other drained fluids (ascites, pleural or pericardial effusions)	7	13.2
Patients’ excretions	48	90.6
Otherb	17	32.1
Process of cleaning, disinfection and storage defined for:
Infusion poles and pumps	49	92.5
Mobile toilet chairs	50	94.4
Blood pressure monitors	46	86.8
Bedpans and urinal bottles	51	96.2
Flower vases	41	77.4

URWDs = utility room wastewater drains

^a e.g. cleaning products, magazines, scented spray, broom and shovel, measuring cups, material used for sampling urine or blood

^b e.g. cleaning products, disinfection products, water from flower vases

Discussion

This outbreak investigation identified two clusters of CPB transmission linked to colonized URWDs. Colonization with CPB was detected in more than half of all screened URWDs, highlighting their role as a significant hospital environmental CPB reservoir. Multiple breaches in utility room management and design and airborne CPB after flushing URWD were identified, potentially facilitating CPB transmission within our institution.

None of our six cluster patients suffered from CPB infection. The risk of progression from CPB carriage CPB bloodstream infection has recently been described as lower than previously reported (cumulative incidence of 2.4%), differing significantly by bacterial species [22].

Contamination rates of sinks and shower drains with CPB vary across institutions, recent reports ranging from 4% of hospital sinks and shower drains in Canada [23], to 24% of patient-room sinks in Israel [24], and 41.7% of ICU sinks in France [25]. The 52% contamination rate observed in our URWDs appears high in comparison but is likely attributable to their frequent use for disposal of contaminated bodily fluids, nutrients, and potentially antibiotics, promoting biofilm formation and selection of multidrug-resistant organisms [26].

Interestingly, the only ward without detection of URWD CPB colonization was a short-stay unit, which typically hosts healthier patients with fewer comorbidities and most likely lower prior healthcare exposure, as compared to the other wards screened which include the medical, hemato-oncological and neurology department, which typically admit patients with higher morbiditiy rates and longer hospital stays at the UHB. This suggests that patient population characteristics and antimicrobial pressure may be key drivers of environmental CPB contamination.

C. freundii was the predominant CPB detected in URWD, followed by K. pneumoniae. The predominance of C. freundii in sink and drain environments may be explained by its robust biofilm-forming capacity, metabolic adaptability to nutrient-poor wastewater, and frequent acquisition of resistance genes, enabling persistence in hospital plumbing systems even under selective pressure from disinfectants and antimicrobials [27]. While we did not detect E. hormaechaei, the species responsible for our VIM-1-cluster, the corresponding plasmid containing bla_VIM-1 was detected in environmental C. freundii and K. pneumoniae in URWD of the corresponding ward, suggesting the possibility of horizontal gene transfer between clinical and environmental Enterobacterales, a finding aligning with investigations of a study tracking plasmid transmission in clinical and environmental CRE isolates [28].

The predominance of bla_OXA and bla_KPC type carbapenemases is in line with the Swiss overall prevalence [2]. The presence of one IMP-carbapenemase gene in one of the URWD localized on the hematologic isolation ward was surprising, as IMP in Switzerland is rare with only three isolates detected in 2023 and none in 2024 [2], suggesting undetected CPB colonization. The frequent co-occurrence of bla_KPC-3 and bla_CTX-M-15 in C. freundii and K. pneumoniae isolates complicates phenotypic detection and limits treatment options, underlining the importance of molecular characterization in outbreak investigations.

Detection of CPB in airborne sampling after flushing of URWD underscores the importance of dispersal of CPB from contaminated drains as a driver of CPB transmission and the protective effect of covers limiting dispersion. While we did not identify the direct mode of transmission in our clusters, contamination of shared medical equipment (SME) is a probable path of pathogen transmission. Enhanced cleaning and disinfection of SME resulted in a significant reduction of hospital-acquired infections (HAI) [29], a target for IPC measures. Utility rooms at the UHB are regularly used for cleaning and storage of SME in the clean zone. However, our investigation revealed inconsistent separation of clean and contaminated zones in utility rooms with inadequate storage of SME in the contaminated zone in 88.7%, facilitating contamination of SME and transmission of pathogens. The IDSA practice recommendations of strategies to prevent HAI through hand hygiene recommend not to keep patient care supplies within one meter of sinks or if not feasible to install splash guards [30]. Garvey et al. demonstrated a sink splash zone of two meters, recommending avoidance of storing patient equipment within this zone [12], yet implementation may prove difficult in a setting of limited space.

While sensitivity for CPB screening of SME might be limited due to transient CPB colonization, our outbreak investigation focused on the detection and minimization of the reservoir, the promotion of standard IPC measures and consequently the limitation of dispersal from contaminated sinks by removing SME out of contaminated zones and installation of drain covers and splash guards. Elimination of the disposal unit may further represent a useful IPC measure as disposal of patient’s excretions and cleaning of urinal bottles and bedpans can be provided by the automatic cleaning machine, a closed cleaning system relevantly limiting dispersal of droplets and aerosols. Other fluids such as residues from parenteral nutrition or medications and antibiotics should be disposed of in accordance with the waste disposal policy of the respective institution.

Repetitive disinfection of colonized drains with 6% hydrogen peroxide did not result in eradication of CPB in the colonized URWD and only a short-termed eradication in the colonized pharmacy drain, hence p-traps were replaced primarily aiming to reduce bacterial load and therefore transmission risk, rather than eradication of CPB colonization [31]. Eradication of CPB from colonized drainage systems proves difficult, as both chemical disinfection and replacement of drains or drainage systems have been described as ineffective for permanent eradication of CPB from the hospital water system [5]. Migration of persisting biofilm in distal pipes connected to the p-trap system can lead to early recolonization of the drains [6, 32]. Based on this hypothesis, we additionally sampled rooms located on the ward below that were connected to the same wastewater stack; however, no VIM-1–producing bacteria were detected in these samples (Fig. 2). While a combination of different IPC measures in addition to reinforcing standard hygiene measures may be most promising, guidelines on IPC management of sinks or drains colonized with CPB were scarce although sink hygiene has come back into the spotlight in recent years [30, 33–35]. Further research on this emerging topic is needed and should in particular include the optimal structural design of a hospital’s water and wastewater facilities, possibly including water-free patient care to prevent the dispersion of pathogens as a primary preventive measure [36].

This study has several limitations. The single-centre retrospective study design limits its generalizability. Utility rooms are a concept not being universally used in other institutions and management of waste disposal and reprocessing of SME is handled differently. Nevertheless, similar wet environment CPB reservoirs seem to exist in various institutions, as growing evidence over recent years indicates, and IPC solution approaches might overlap.

As we performed a point prevalence screening of URWD, a statement cannot be made on long-term dynamics of URWD colonization. Contamination of sinks with the same CPB strain for up to 80 weeks has been reported [24], as well as that of other sanitary appliances for several years [26]. While the corresponding bacterial species were not detected in the wards URWD during the point-prevalence screening in cluster 1, this may reflect their transient occurrence—species such as E. hormaechei may only be intermittently present and therefore easily missed in a single screening. In contrast, other species, such as C. freundii, may be better adapted to persist in aquatic environments and could therefore act as reservoirs for antimicrobial resistance genes [27]. With exception to the VIM-1 cluster, we did not screen SME for CPB colonization, nor did we perform screening of healthcare staff, hence, no conclusion can be made about the direct transmission pathway.

In conclusion, URWD served as a CPB reservoir, possibly promoting CPB transmission. Targeted infection prevention and control measures in utility rooms may offer an effective strategy for reducing in-hospital CPB transmission.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CPB

Carbapenemase-producing bacteria

EKNZ

Ethikkommission Nordwest- und Zentralschweiz

ESBL

Extended-Spectrum Beta-Lactamase

ICU

Intensive Care Unit

IPC

Infection Prevention and Control

KPC

Klebsiella pneumoniae-Carbapenemase

NDM

New-Delhi-Metallo-beta-lactamase

OXA

Oxacillinase

SME

Shared Medical Equipment

UHB

University Hospital Basel

URWDs

Utility Room Wastewater Drains

VIM

Verona-integrated Metallo-beta-lactamase

WHO

World Health Organization

Author contributions

IV performed data acquisition, data analyses, interpretation of results and wrote the manuscript. GC, SK, PU, PD, DR, NG performed data acquisition and critically revised the manuscript. CN, CF, BI and PMK performed and described microbiological analyses, created microbiological figures und critically revised the manuscript. NK, SB, JP, JH critically revised the manuscript. STS conceived and supervised the study, analysed data, interpreted results and critically revised the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by the University Hospital Basel and the University of Basel.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was assessed by the local ethics committee (Ethikkommission Nordwest- und Zentralschweiz) as a quality assurance project, thus the requirement for approval does not apply (project no. Req-2023–01529).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.