Passive sampling of hospital wastewater as a promising warning signal for Candidozyma auris in a low-prevalence setting: a proof-of-concept study

Abstract

Background

Candidozyma (Candida) auris is an opportunistic fungal pathogen with increasing worldwide prevalence. As it can spread quickly in healthcare settings, monitoring C. auris is essential for effective infection control. Current hospital surveillance of C. auris often focuses on screening high-risk patients, as prevalence in the Netherlands is low. Wastewater-based surveillance may provide a complementary approach by capturing signals from all hospital inpatients, but this commonly relies on expensive automated water samplers. This study examined the feasibility of using lower-cost, easily deployable passive samplers to detect C. auris in hospital wastewater of a low-prevalence setting.

Methods

Ten passive samplers containing one cotton bud, one piece of medical gauze, and two electronegative membrane filters were deployed for 24 h during a 16-week period in whole-hospital wastewater. Samplers were installed before, during, and after admission of a patient colonized with C. auris. DNA was extracted using a combination of phenol-chloroform and bead-beating, and C. auris presence was assessed by real-time polymerase chain reaction (RT-PCR).

Results

One patient was sampled during two separate admission periods and was identified as C. auris-colonized based on the detection of C. auris DNA in multiple specimens at both time points. The load of the clinical specimens decreased from a cycle threshold (Ct) value of 27.9 to 34.4 at second sampling. Hospital wastewater was C. auris negative before patient admission, low positive throughout patient stay, with average Ct-values of passive materials ranging from 33.2 to 38.0, and returned to negative after patient discharge.

Conclusion

Our findings highlight the possibility to detect C. auris in a passive sampler wastewater-based RT-PCR screening in a low-prevalence country during admission of a C. auris-positive patient. Although further studies are required, this proof-of-concept study identified passive wastewater sampling as potential C. auris-surveillance tool. This low-cost method may eventually complement existing risk-based strategies and provide warning signals to prevent C. auris transmission within the hospital.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13756-026-01773-0.

Introduction

Candidozyma (Candida) auris is an opportunistic fungal pathogen with increasing worldwide prevalence [1, 2]. It has been isolated from non-invasive body sites, including skin, as well as invasive sites, such as blood [3, 4]. Invasive infections are more frequent in critically ill patients, where they are associated with global mortality rates of up to 60% [3]. Furthermore, most clinical strains are resistant to one or more antifungal drug classes [3] and can rapidly spread in healthcare settings [3, 4]. Therefore, the WHO has designated C. auris as a fungal priority pathogen [5].

Monitoring C. auris in healthcare settings is critical. Colonized and infected patients shed C. auris in their environment [6, 7], potentially enabling transmission and colonization of other patients, which could lead to serious and prolonged outbreaks [4]. Therefore, a 2021 review recommended that detection of a single C. auris case should trigger an epidemiological investigation to prevent nosocomial transmission [6]. Hospital surveillance often focuses on screening high-risk patients (e.g. prior hospitalized abroad) at admission, but C. auris carriage is not limited to this group. For timely detection of C. auris in hospitals, wastewater-based surveillance (WBS) may provide a complementary approach as C. auris can enter hospital wastewater via human excreta and shower water. In this way, C. auris was detected in hospital wastewater [8] and in influent of a wastewater treatment plant despite low prevalence [9].

WBS can use water samples, which are typically obtained using resource-intensive automated samplers that collect wastewater time- or flow-proportionally. These samplers require a power source, maintenance, and experts for installation, removal, and programming [10], resulting in one-time mobilisation costs of up to €3000 and additional rental costs of €500 to €1000 per sampling week. Alternatively, water samples can be obtained via grab samples, where a volume of water is taken at a specific time. Although C. auris presence in ward wastewater was recently reported with grab samples [11], detection rates varied in whole-hospital wastewater [8], possibly because these grab samples only captured a limited timeframe. All water samples require pre-processing to concentrate the target microorganism [10, 12], either by vacuum filtration through a 0.45 μm pore-size filter [8] or by centrifugation-based partitioning [11, 13].

As an alternative to water samples, so-called passive samplers (‘passives’) made of adsorbent materials enclosed in a protective case [10, 12] may be used in WBS. These passives can be deployed for 4-48 h [14] and continuously capture microorganisms by filtration, adsorption or a combination of both mechanisms. The materials used can affect microbial attachment, as seen in viruses and bacteria [10, 15]. The estimated cost price is around €25 per passive, no expert or power source is needed for deployment, and elution can occur directly on the passive materials for downstream analysis. However, most studies to date examine passive sampling for viral and bacterial detection [10], with only one recent study reporting the possibility to detect C. auris in cotton Moore swabs [16]. While this singular study proposed passive sampling for wastewater surveillance, it remains unclear whether this approach is feasible in non-endemic settings and whether C. auris detection in wastewater correlates with its detection among hospital inpatients.

In this proof-of-concept study, we explore the detection of C. auris in hospital wastewater after admission of a colonized patient using passive sampling and real-time polymerase chain reaction (RT-PCR) in the Netherlands, a low-prevalence country [17].

Methods

Study site

The study site was an external sewer manhole directly downstream of a tertiary-care facility with 546 beds in Maastricht, the Netherlands (50°50’03.8"N 5°42’47.8"E). This manhole collects combined wastewater from the complete hospital, surrounding university buildings, and rainfall runoff. The hospital produces an estimated 394m wastewater per day, of which the untreated water travels directly to the study site in 5 to 30 min, depending on the originating ward, before continuing to a regional WWTP for treatment.

Passive deployment

Torpedo passive samplers [12] adapted for application in small water depths (type ‘pensives’) consisted of a 3D-printed case that contained a weight to keep the passive submerged and three different sorbent materials; one cotton bud, one piece of medical gauze, and two electronegative membrane filters (Fig. 1, Supplementary Fig. 1). This single-use sampler has been designed to sit in a small sewer line and allow wastewater to flow through the different materials (Fig. 1). These passives were deployed for 24 h-periods once every two weeks as part of a yearlong, ongoing investigation. In this period, there were two one-week shifts at Day 35 and Day 56 due to sampling possibilities, and an additional sampling round at Day 77 to investigate a possible delay in returning to C. auris negativity. After retrieval, the passive samplers were immediately deconstructed, and the cotton bud, the medical gauze, and the two electronegative membrane filters stored separately at −20 °C until DNA extraction. All other materials were disposed of according to the hospital guidelines.

Fig. 1.

(A) Schematic situation during passive deployment with (B) the individual components and layout of a torpedo passive sampler. The weight and top are placed on the 3D-printed casing containing the other components. Figures were created by converting original images to line drawings in chatGPT-5

DNA extraction of passive materials

DNA was extracted of the passive materials via the phenol-chloroform extraction method [18] combined with bead-beating, to maximize DNA yield. Passive materials were mixed with 1 mL 10x Tris-EDTA buffer (Pharmacia Biotech), 200 µg lysozyme (MP Biomedicals), and 10 µg MetaPolyzyme solutions (Sigma-Aldrich) before incubation at 37 °C for 1 h. Subsequently, the sample was incubated for 10 min with 30 µL 10% Sodium Dodecyl Sulfate solution (Invitrogen) followed by a 1 h incubation with 50 µg proteinase K (Sigma-Aldrich) at 56 °C.

DNA extraction was performed by adding phenol (Sigma-Aldrich) and chloroform (Sigma-Aldrich) with incubation for 5 min at room temperature. After centrifugation (5 min, 21000xg), the aqueous phase was collected. The remaining phenol phase, was subjected to bead beating using a FastPrep-24™ 5G Instrument (MP Biomedicals) with three 3 mm glass beads (Sigma-Aldrich), 100 µL 0.5 mm zirconia/silica beads (BioSpec Products, Inc.), and 200 µL nuclease-free water (Thermofisher Scientific). Following brief centrifugation, the supernatant was pooled with the initial aqueous phase.

The aqueous phases were further purified by a second phenol-chloroform extraction followed by a chloroform-only step. DNA was precipitated with 3 M sodium acetate (Sigma-Aldrich) and isopropanol (Sigma-Aldrich), incubated on ice for 10 min, and centrifuged (10 min, 500xg). Pellets were washed twice with 80% ethanol, air-dried, and resuspended in 50 µL nuclease-free water containing 1 µg RNase A (Life Technologies), followed by incubation at 37 °C for 30 min. The DNA concentration was determined using a Qubit fluorometer (Invitrogen).

RT-PCR analysis

Presence of C. auris DNA was detected by RT-PCR, as this could consistently detect C. auris in wastewater samples [9]. A RT-PCR assay validated on clinical specimens was performed, targeting the internal transcribed spacer 4 region of the rDNA [19], with sequences and concentrations as described in Supplementary Table 1. Wastewater samples were tested in replicate and each reaction consisted of 8 µL DNA extraction eluate, 10 µL TaqPath BactoPure Master Mix (Thermofisher Scientific), and 2 µL primer/probe mix (Supplementary Table 1). Samples were spiked with a plasmid construct (pGEM-3Z vector) containing the murine cytomegalovirus glycoprotein B (mCMV-gb) target and were diluted 2x and 5x to identify and resolve potential inhibition. The RT-PCR was performed on the Applied Biosystems Quantstudio5 (Thermofisher Scientific) with the following cycling conditions: 30 s on 60 °C, 2 min on 95 °C, 42 repeats of 10 s on 95 °C and 30 s on 60 °C, with a final step of 30 s on 60 °C.

Current C. auris screening

Current screening efforts for C. auris target patients with a history of hospitalization abroad, prior residence in asylum seeker accommodation facilities or those selected for screening based on expert opinion. These high-risk patients are immediately placed in strict isolation and screened according to the routine infection prevention and control guidelines, including culture and RT-PCR on throat, nose, axilla-groin, and rectal eSwabs (Copan). This combination of body sites can identify C. auris in 98% of cases [20]. In case of colonization, corresponding infection control measures are maintained until two independent clinical screening results are negative, at least 12 months after the last positive result [21]. In absence of colonization or infection, the corresponding precautionary measures are lifted.

Results

Since the implementation of the risk-based RT-PCR screening for C. auris, 136 patients were screened over a 26-week period, averaging 5.2 patients per week. In week 27, the first patient considered colonized was identified based on positive RT-PCR results from multiple specimens taken 40.5 h after patient’s admission, suggesting the patient to be a carrier on admission. This patient remained the only identified C. auris colonization during the admission period, which consisted of two admissions: the first lasting for 16 days and the second for 39 days. Table 1 displays the Ct-values of the two clinical specimen sets, one during each admission, and all positive wastewater sample materials. Notably, the axilla-groin and rectal swab tested during the first admission had a lower Ct-value than the swabs of the same body sites during the second admission (Table 1).

Table 1.

RT-PCR Ct-values of the two clinical specimen sets of the colonized patient and all positive wastewater sample materials per study day, material, dilution tested, and average Ct-value of technical replicates

Study day	Sample type	Material	Ct-value	Dilution	Replicate	Ct average
2	Clinical°	Nose swab	ND
2	Clinical°	Throat swab	ND
2	Clinical°	Axilla-groin swab	27.7
2	Clinical°	Rectal swab	27.9
15	Wastewater	Filter 2	33	2	1	34.2
15	Wastewater	Filter 2	34.5	2	2
15	Wastewater	Filter 2	35.0	2	3
15	Wastewater	Gauze	31.9	2	1	33.2
15	Wastewater	Gauze	34.6	2	2
29	Wastewater	Cotton bud	35.6	2	1	35.6
29	Wastewater	Cotton bud	36.1	2	2
29	Wastewater	Cotton bud	35.1	2	3
29	Wastewater	Gauze	36.6	5	1	38
29	Wastewater	Gauze	39.3	5	2
35	Wastewater	Filter 2	37.0	2	1	36.5
35	Wastewater	Filter 2	36.1	2	2
35	Wastewater	Gauze	32.2	2	1	35.3
35	Wastewater	Gauze	38.4	2	2
35	Wastewater	Gauze	34.7	5	1	34.5
35	Wastewater	Gauze	34.3	5	2
54	Clinical°	Nose swab	38.3
54	Clinical°	Axilla-groin swab	32.4
54	Clinical°	Rectal swab	34.6
54	Clinical°	Rectal swab	34.4
56	Wastewater	Cotton bud	36.2	2	1	36.2
56	Wastewater	Cotton bud	36.1	2	2
70	Wastewater	Gauze	36.0	2	1	37.6*
70	Wastewater	Gauze	39.2*	2	2

ND indicates no C. auris detection

* Inhibition, observed as the reduced detection of spike-in; observed Ct values might be an underestimation.

° Clinical samples consisted of 180 µL eSwab medium from which DNA was directly extracted and used undiluted in the RT-PCR.

Over a 16-week period, ten passives were deployed in hospital wastewater during patient presence or absence, and tested for C. auris presence by RT-PCR. C. auris DNA was detected most often in the medical gauze and the DNA load was consistently low in all materials, with average Ct-values ranging from 33.2 to 38 (Table 1). Figure 2 shows C. auris detection of the two clinical specimen sets and the passive materials before, during, and after patient admission. No C. auris DNA was detected on passive materials before patient admission, while it was detected in various materials of multiple passives after patient admission (Fig. 2). Interestingly, C. auris was not detected in the passive materials at Day 1 and 50, even though the colonized patient was admitted on those days (Fig. 2). Although C. auris was still detected five days after patient discharge, it was no longer detected after twelve days (Fig. 2).

Fig. 2.

RT-PCR results of C. auris detection in the two clinical specimen sets of the colonized patient and per passive material collected before, during, and after patient admission. Study day 0 represents the day of patient admission. On day −35 and −14 filter 2 was not available for testing

Discussion

Early detection of C. auris is of particular importance for hospital infection prevention and control, given its ability to persist in the hospital environment and its frequent multidrug resistance. Timely identification enables rapid containment measures to prevent hospital-wide outbreaks. In this study, we applied RT-PCR to passive samplers and demonstrated that C. auris could be detected on a variety of materials from multiple passive samplers during the admission of just a single patient colonized with C. auris. Notably, these detections were made in a hospital setting with low prevalence of C. auris, highlighting the sensitivity of RT-PCR-based environmental surveillance using passive sampling.

While universal screening of all hospital patients was not performed and potential additional cases could have been missed, the temporal emergence of positive wastewater samples after admission of the first colonized patient in the hospital suggests that this admitted patient was the most likely source. Moreover, no C. auris cases had been detected in this hospital before; even with implemented risk-based PCR screening up to six months prior to this admittance, and C. auris has rarely been reported in the Netherlands to date [17]. Unfortunately, low DNA loads in wastewater prevented whole-genome sequencing, so a direct C. auris patient–wastewater link could not be proven.

Fungal detection in wastewater by passives might depend on C. auris shedding, washroom facility usage by the patient, and the C. auris load in wastewater. The wastewater sampler deployed one day after patient admission was negative, suggesting C. auris concentrations below the limit of detection or lack of washroom use by the patient. Moreover, C. auris can persist within the clinical environment such as sink drains [22], potentially prolonging its presence in hospital wastewater. This persistence, together with the wastewater system layout, may explain the minimum five-day delay in returning to C. auris negativity after patient discharge. Follow-up beyond Day 77 was not possible because another colonized patient was admitted on Day 81. As this proof-of-concept study observed C. auris detection before, during, and after patient admission only once, additional studies are required to evaluate the detection timing, particularly the timing of first detection after admission and return to C. auris-negativity after patient discharge. Furthermore, as only one known colonized patient was present during this study, signal dynamics in the presence of multiple patients also remain to be investigated.

Quantifying C. auris using passives presents several challenges. First, no standardized water volume passes through the passive materials, requiring further investigations into the quantification of microorganisms when using passives. Second, microbial accumulation within passives may vary with water flow [23], potentially influencing detection on identical materials, as reflected by the different results between the two filter membranes in this study. Third, the RT-PCR might lead to an underestimation of C. auris if inhibitory substances are present [10] or if the low C. auris concentrations in wastewater approach the assay’s sensitivity threshold. Finally, abiotic factors such as rainwater co-collection and flow rate might affect microbial detection [24]. This dilution of wastewater might explain the negative signal on Day 50, when melting snow entered the hospital wastewater as ambient temperatures rose above freezing [25]. Future research could investigate whether passive sampling can be used to quantify C. auris in hospital wastewater and whether the detected wastewater load correlates with the number of C. auris-positive patients.

WBS using passive samplers could provide valuable information in a low-prevalence setting. Although no assumptions can be made for higher-prevalence settings, we detected one single C. auris-positive patient in our very low-prevalence setting. When performed frequently, WBS could give information about all hospital inpatients, notifying infection control teams of potential C. auris colonization or infection that might remain undetected with risk-based screening efforts only. Additionally, it could allow for early C. auris detection, which is, in combination with infection control measures, important in preventing C. auris transmission [26]. However, it must be noted that the resolution of complete hospital wastewater is low and more focused screening would be necessary to find the C. auris source, possibly via ward wastewater or broader clinical screenings.

Other important considerations could be the limited sample size, unknown C. auris viability, and the effect of wastewater characteristics. While C. auris was detected in the passive materials, the study was only observational and the small sample size prevented any indication of which material would be best for C. auris detection. Furthermore, C. auris presence was only detected by RT-PCR and no culture was performed. Therefore, viability of C. auris in hospital wastewater is unknown and no further analysis could be performed. Although it remains unexplored whether cell-free nucleic acids or intact fungal cells adhere to the passive materials, both would indicate the presence of C. auris in the hospital and could trigger infection control investigations and measures. Lastly, the wastewater flow at the study site was not examined during the deployment and study period. Water flow may affect microbial accumulation in passives [23], and wastewater characteristics, such as pH, temperature, and chemical content, may affect the breakdown rate of C. auris [8]. As these were not characterised, their effect on the results remain uncertain.

Conclusion

Our findings highlight the possibility to detect C. auris in hospital wastewater using passive samplers and RT-PCR. Although additional studies are needed to explore detection timing and signal dynamics further, we identified passive wastewater sampling as potential surveillance tool for the detection of C. auris in a low-prevalence hospital setting. This practical and low-cost method may eventually complement existing risk-based strategies and could provide warning signals to prevent C. auris transmission within the hospital.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

C. auris

Candidozyma auris

Ct

Cycle threshold

RT-PCR

Real-time polymerase chain reaction

Author contributions

AHV, MM, and LBA created the study concept. AHV, EVHB, and CJHW performed experiments to acquire the data. AHV wrote the main manuscript text, prepared the figures, and wrote the revised text. PHMS and LBA supervised the process and all authors reviewed the manuscript critically.

Funding

No external funding was received for this study.

Data availability

All data generated or analysed during this study are included in this published article (and its supplementary files).

Declarations

Ethics approval

The infection prevention and control team provided anonymized, retrospective data of patient samples (throat, nose, axilla-groin, and rectal swabs) collected during routine clinical care and did not alter usual clinical practice. Consequently, formal ethical review and approval were not required.

Generative AI in the manuscript preparation process

During the preparation of this work, the authors used chatGPT in order to optimize conciseness and clarity of textual pieces. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. Additionally, original photographs of sampling materials were converted to line drawings by use of chatGPT.

Competing interests

RPSS provided passive samplers at cost price. All other authors declare that they have no competing interests.