Abstract
Background Candida auris is an emerging fungal pathogen that is often multidrug-resistant. It can persist on skin and in hospital environments, leading to outbreaks and severe infections for patients at risk. Several countries and institutions are working on establishing guidelines and recommendations for prevention. This review aims to assess the evidence on factors associated with C. auris colonisation or infection, the duration of such colonisation, possible colonisation sites, and the risk of secondary cases to inform screening recommendations.
Methods We systematically searched five databases for primary studies and systematic reviews of our four outcomes. We excluded studies on treatment, management, laboratory methods, drug resistance, and environmental screening. From each paper, we extracted relevant data and summarised them in tables. Main findings were described narratively.
Findings We selected 117 studies for inclusion. Most of the studies were observational studies. Without taking the method of testing into account, the duration of C. auris colonisation varied, with up to and beyond a year being common. The predominant sites of colonisation were the axillae and groin, with the nares and rectum being less common sites. The risk of secondary cases saw considerable variation across the studies, and the secondary cases primarily involved patients and not health care workers. Critical care settings, invasive medical devices, recent antimicrobial use, and comorbidities were often associated with C. auris colonisation and infection.
Conclusion Our review highlights that, despite relevant findings on factors influencing C. auris colonisation and infection, substantial gaps remain in the evidence supporting screening practices. Most studies were conducted reactively, in outbreak settings, and lack systematic protocols. Given these limitations, screening guidelines are likely to be more successful if grounded in medical theory and yeast microbiology rather than relying solely on current studies. Rigorous, well- designed research is urgently needed to inform future C. auris screening and control efforts.
Background
Candida auris is a fungal pathogen that rapidly emerged globally. While initially isolated from the external ear canal of a Japanese patient in 2009 and retrospectively identified in South Korean blood samples dating back to 1996, the first large hospital outbreak triggering worldwide concern occurred in India and was described in 2013 (2–5). In the 2019 Antibiotic Resistance Threat Report by the U.S Centers for Disease Control and Prevention (CDC), C. auris was listed as an urgent threat and in 2022 the World Health Organization (WHO) classified it as a critical priority pathogen (6;7). The urgency communicated from these institutions stems from the rapid emergence of C. auris, its resistance to antifungal treatments, and its ability to cause outbreaks and severe, often fatal, infections in hospitalised patients, particularly those with underlying conditions(7).
C. auris can colonise and persist on the skin for longer periods and exhibits extended environmental survival on inanimate surfaces within healthcare settings. These characteristics contribute to transmission events and protracted outbreaks in healthcare settings, especially in intensive care units (ICU) (2;8). Furthermore, C. auris is capable of biofilm formation and can resist routine cleaning (9–11). Four major clades – South Asian (I), East Asian (II), South African (III) and South American (IV) – have accounted for the global spread of the pathogen (2). Clade I is the most prevalent. Genetic studies using whole genome sequencing (WGS) suggest a simultaneous emergence of these clades, which may have played a part in causing the rapid global spread. Recently one more clade; Iranian (V) has been identified, and a sixth clade is under investigation in Singapore (VI) (2;12). Variations in drug resistance and virulence, the ability to cause invasive disease, have been observed among different C. auris clades (2).
Although C. auris has been reported in over 60 countries across six continents since its discovery (9;13), case numbers likely underestimate the true burden as the fungus has proven notoriously difficult to culture (14). The first identified case of C. auris in Europe, belonging to clade I, was imported from India in 2007 and retrospectively identified (15). Since then, several European countries have witnessed a significant increase in cases, causing outbreaks, especially in ICU (2). Between 2013 and 2021, 15 EU/EEA countries identified a total of 1812 cases (11). The first European outbreak occurred in 2015-2016 in an ICU in the United Kingdom with a total of 70 cases (16). Between 2019 and 2021, five European countries (Denmark, France, Germany, Greece and Italy) reported a total of 14 C. auris outbreaks, encompassing 327 cases (17).
Several national and local institutions are moving to either recommend or mandate screening against C. auris in their prevention effort, e.g., in guideline documents. This highlights the need to assess the evidence on some key factors to consider when developing screening strategies.
These key factors are the associated factors of colonisation/infection, duration of colonisation, possible colonisation sites and risk of secondary cases. This systematic review aimed to comprehensively search for and narratively summarise the available literature on these key factors, including all relevant studies published to date.
Methods
Search and study selection
We systematically searched five databases on 2 February 2024: Embase, Ovid Medline, Cochrane Database of Systematic Reviews/ Cochrane Central Register of Controlled Trials, Web of Science, and Epistemonikos. The searches were complemented with hand-searches for grey literature and existing guidelines. The searches were performed by a specialist librarian, after internal peer review by another librarian. Search terms for C. auris combined with synonyms with appropriate truncations and abbreviations were used for searching title, abstract, author keywords, and controlled vocabulary. The search strategy was tailored to each database’s search interface. The search strategy can be found in Appendix 1. Deduplication was performed in EndNote 20 (18).
Three researchers piloted the inclusion and exclusion criteria on an initial sample of 50 abstracts to identify relevant inclusion criteria (Table 1). Subsequently, two researchers screened the remaining studies. We used EPPI-Reviewer 6 for screening (19). Disagreements or uncertainties were addressed through discussion with a third researcher. Two researchers performed single full-text screening. Inclusion conflicts were resolved through discussion between the researchers.
Data extraction and analysis
Two researchers extracted relevant data on three of the main outcomes: duration of colonisation, risk of secondary cases, and colonisation sites. One researcher extracted data regarding associated factors. For all studies we extracted information on the country of study, study period, setting, participants, follow-up period, and reported outcomes. Additionally, depending on the specific outcome of interest, we also extracted relevant data and statistics (e.g., odds ratio).
For each outcome we created tables summarising the relevant studies. Each table lists the studies reporting on the specific outcomes, along with the variables relevant for the outcome. There was a limited number of studies featuring control groups that investigated associated factors or participant characteristics, and only these were included in the interpreting and assessing factors associated with infection or colonisation. We narratively summarised each of the systematic reviews separately, and then findings for each outcome from the primary studies, where similar studies or studies reporting on comparable aspects to each outcome were grouped together for easier comparison. To assist in the synthesis and discussion we employed a set of definitions (Table 2). Neither a meta-analysis or a critical appraisal/quality assessment of the studies was performed due to the qualitative heterogeneity in the primary studies – both in terms of design and context. However, we highlighted methodological shortcomings throughout the text to provide context for our findings.
Findings
Selected studies
We identified 2,371 unique references after deduplication (Figure 1). Title and abstract screening identified 291 relevant studies. On full-text screening we included 117 studies. Studies were conducted in Asia (n=40), North America (n=29), Europe (n=21), South America (n=12), Africa (n=8), and Oceania (n=1). Six studies encompassed data from multiple countries.
Flow diagram of search strategy and study inclusion. Adapted from (1). *Databases include Ovid Medline® and Epub Ahead of Print, Embase, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, Web of Science and Epistemonikos.
Publication dates ranged from 2016 to 2024. We included 114 observational studies and three systematic reviews. Reported outcomes of interest were duration of colonisation (n=8), risk of secondary cases (n=15), colonisation sites (n=29), and associated factors/participant characteristics (n=112). Only a limited number of studies that investigated associated factors or participant characteristics featured control groups (n=14). In general, the included studies exhibited considerable variation in study populations and sample sizes. The body of literature on C. auris appear to primarily consist of single-centre studies with small sample sizes and descriptive statistics. These are often outbreak reports, summaries of experiences with the clinical management of defined clusters, or studies summarising data from laboratory findings over a specific period. An overview of the included studies is provided in Table 3.
Systematic reviews
We identified three systematic reviews (8;20;21). Sekyere authored a systematic review of different aspects of C. auris; molecular epidemiology, virulence and pathogenicity, resistance, crude mortality rates and infection prevention and control in 2018 (8), but did not collate data or make interpretations regarding our outcomes of interest. Vaseghi et al. carried out a systematic overview of the literature of COVID-19 associated C. auris infections (20). The relevant results of the review were underlying medical conditions and medical device interventions as risk factors/predictors for C. auris, reflecting a highly selected group of patients.
The most frequent predisposing factors were found to be hypertension (15/996; 1.5%), followed by diabetes mellitus (12/996; 1.2%) and cardiovascular diseases (7/996; 0.7%). Central venous catheter (CVC) was found to be the most applied medical device (prevalence rate: 96%). The estimated odds ratios indicated that COVID-19 patients who received a CVC had 2.6 times the odds of catching C. auris co-infection. Vinayagamoorthy et al. conducted a systematic review examining the prevalence, risk factors, treatment, and outcome of C. auris infections in COVID-19 patients (21). They found that hypertension (17/35; 48.6%) and diabetes mellitus (12/35; 34.3%) were the most prevalent underlying conditions among COVID-19- patients with C. auris candidemia. Regarding risk factors, the review identified broad-spectrum antibiotic usage (35/35; 100%), ICU stay (33/35; 94.3%), mechanical ventilation (24/35; 86.6%) and CVC (28/35;80%) as the most frequently found. However, the authors found no significant differences in underlying disease and risk factors among C. auris infections excluding candidaemia or colonisation, and C. auris candidemia cases.
Primary studies
Duration of C. auris colonisation
We included eight studies reporting on duration of C. auris colonisation (Table 4). The studies were from USA (n=5), UK (n=1), India (n=1), and South Korea (n=1), and were published between 2016 and 2023. The combined study population consisted of adult patients with a median age ranging from 51-72 years. Included participants were hospitalised in different types of hospitals and often with several underlying medical conditions. The assessment on colonisation duration was predominantly conducted within the healthcare settings, with only one study tracking patients post-discharge (22). The number of positive C. auris patients followed up with one or several screening tests varied from three to 75 patients among the studies. In some studies patients were lost to follow-up so the final number of patients in the included studies were three to 60 patients. Two of the studies followed less than ten patients (23;24). Follow-up time and number of screening tests taken of each patient varied between studies and among patients. Most studies followed some of the patients for at least 6 months, although follow-up time varied from a few days to 20 months.
Three studies lacked a pre-defined, standardised interval for the follow-up screening (23–25). Although the other studies had defined screening intervals, the time frames varied among the patients involved. Colonisation clearance was not clearly defined in five of the studies (23;24;26-28). All the studies, except one, obtained follow-up samples from multiple sites, primarily the groin and axilla. The exception conducted screenings solely from external ear discharge (24).
The absence of predefined, standardised protocols for research into colonisation duration, including set time intervals and the number of tests to conduct, coupled with the lack of unified definitions for spontaneous decolonisation, complicates the interpretation of duration of C. auris colonisation.
Seven studies looked at the length of colonisation among patients in hospitals (16;23-28). The duration of colonisation varied between studies and patients, ranging from days to years, with long-term colonisation of a few months to over a year being commonly observed. Eyre et al. followed 60 patients for up to 6 months with weekly screening and at discharge (16). They found the median duration of colonisation to be 2-3 months. Clearance was defined as two or three consecutive negative cultures. Pacilli et al. followed 51 patients with screening every week and found the duration of colonisation to be up to 10 months in some patients, but were unable to ascertain their definition of colonisation clearance (27). Adams et al. analysed follow-up samples from 38 patients, collected at least one week apart, and observed a duration of colonisation ranging from days and up to >200 days (25). Bergeron et al. assessed colonisation among patients post discharge with a follow-up time ranging from 0 to 20 months, taking samples every three months. They found that 62% of the patients did not remain colonised in the follow-up cultures after discharge to the community. They defined ‘serial negative’ as two consecutive negative PCR or fungal culture results. The median time for patients to become ‘serial negative’ was 8.6 months (interquartile range: 5.7-10.8) (22). However, across studies, persistent colonisation for up to a year post-discharge was still common, suggesting that clearance is inconsistent and may be influenced by screening frequency or methodology.
We extracted data on the diagnostic methods used in each study, documenting the sequence in which they were applied. These methods included culture-based approaches (selective agar, broth) with or without Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) confirmation and molecular methods such as polymerase chain reaction (PCR). Three studies incorporated PCR at some stage in the diagnostic process (22;26;27), but given the substantial heterogeneity in reported colonisation durations, no apparent differences were observed compared to studies that relied solely on culture-based methods. Given the known challenges in culturing C. auris, studies incorporating molecular identification at the initial stage may have had a higher detection sensitivity than those relying exclusively on culture. However, no study directly compared the two approaches in a way that allowed for conclusions about differences in measured colonisation duration, and as this review focused on colonisation duration rather than diagnostic accuracy, studies solely concerning microbiological methods were not included.
A critical question regarding recommendations for screening, as well as other infection control measures, is whether individuals previously diagnosed with C. auris should be considered permanent carriers or if negative test results can lead to the cessation of infection control measures. Multiple studies have demonstrated that patients can experience one or more negative screening result, prior to a subsequent positive result (26–28). Pacilli et al. observed that 55% of the patients who had at least one negative screening result later tested positive in a subsequent sample (27). In two separate studies, Arenas et al. and Bergeron et al. found that 34% and 25% of colonised patients, respectively, received a positive result after at least one negative result, with the time interval between screenings varying between patients in both studies (22;26). Arenas et al. used rt-PCR, while Bergeron et al. combined fungal culture and rt- PCR, which may have contributed to the differences in reported rates of intermittent negative results (22;26). Hence, negative test results for a colonised patient should be interpreted with caution when considering discontinuation of appropriate infection prevention and control precautions in healthcare settings. Such a careful approach is supported by international guidelines. The CDC does not recommend routine reassessments for C. auris colonisation, as intermittent negative results have been observed and patients may remain colonised long-term despite negative screenings (29). An expert meeting organised by the International Society for Antimicrobial Chemotherapy in 2019 recommended maintenance of infection control measures until discharge and to flag the patient chart for at least 1 year after the first negative screening culture (30).
Risk of secondary cases
We included 15 studies that reported on the risk of secondary cases of C. auris (Table 5). These were conducted in Australia, Canada, Colombia, India, Italy, Saudi Arabia, Republic of Korea, Spain, UK and USA. Studies were published between 2017 and 2024. The studies reported on outbreak investigations and surveillance screening. The study settings were hospitals, mainly ICUs, with two studies conducted in nursing facilities with ventilator-capability (31;32). Index C. auris cases, here defined as the primary cases that were followed for potential onward transmission, ranged from one to 50 adults, depending on study design, number of hospitals included, and the study period. The total number of tested (investigated) persons ranged from 17 to 960. Tested persons were mainly patients, with seven studies including healthcare workers (HCWs) and/or family members and visitors.
While seven studies reported more than 50 secondary cases, three studies observed zero to three secondary cases, despite testing between 180 and 600 patients (33). Secondary cases were mainly reported in patients. Due to differences in test strategies, follow-up periods, and number of defined primary cases, we could not calculate a secondary attack rate. Some studies did not report the total number of tested individuals. However, as the proportion of positive cases out of tested patients were >40% in some studies, a high risk of secondary cases has been documented in outbreak settings (27;34). Several studies showed that a high number of patients were already colonised shortly after the outbreak was recognised, but in some situations transmission was limited to a small number of patients, and the proportion of positive cases out of those tested were <2% (33;35-37). The reason for this may be multifactorial – i.e., transmission depend on the setting, the population, when the outbreak was detected, the amount of environmental contamination, and the infection control measures implemented.
Four studies detected positive cases among HCWs, identifying between one and four positive individuals from a pool of six to 258 tested persons (34;38-40). Schelenz et al. found one positive HCW out of 258 tested in an outbreak where HCWs had been caring for a patient colonised with C. auris (40). Escandon et al. found two HCWs that had positive hand/groin samples out of six tested (34). The two last studies found one and four HCWs where samples from their hands were positive, respectively, when performing point-prevalence surveys (PPS) in the units where there had been a positive case (38;39). One out of two studies testing family members found one positive family-member (25). The selected studies discovered minimal positive instances among HCWs, primarily featuring yeast present on their hands, even though extensive testing was conducted during outbreaks. Several authors have postulated that HCWs could potentially contribute to transmission chains due to transient hand contamination. However, to affirm this as a prevalent route of transmission, additional research is required.
Six studies reported results from PPS or surveillance screening (16;27;31;33;38;41). The authors reported testing of 101 to 900 persons and found 0 to 140 positives. Four studies did a combination of PPS, surveillance, and contact tracing (25;39;40;42). They tested 258 to 718 persons and found 1 to 62 positive cases. We were not able to find a denominator in three of the studies, and several studies had defined up to 50 primary cases in different hospitals and did surveillance screening on the wards where the cases were admitted (25;34;35). Six out of 15 studies did WGS, confirming genetically distinct clusters.
Not all studies differentiated between direct and indirect exposure as sources of secondary cases. Some studies clearly defined direct exposure and contact tracing, while in several others, the results were presented alongside the number of detections of C. auris in broad surveillance surveys. In a comprehensive approach, four studies carried out contact tracing, testing patients and HCWs with known exposure to a confirmed case. This exposure included direct contact, overlapping stays in the ICU, or other connections within the same ward (34–37). These studies tested 17 to 960 exposed persons and found one to eleven positive cases. Lee et al. concluded that the proportion of secondary cases of C. auris from close contacts was relatively high (25%) (39). However, Rossow et al. did not find an association between C. auris colonisation and residing in a room with a positive case (31).
To determine the transmissibility of C. auris, several factors are crucial, including the microbe’s inherent virulence, host susceptibility, the type and duration of contact, and persistence in the environment. Schelenz et al. conducted a root cause analysis and discovered that the minimum duration of contact with a positive case or a contaminated environment for acquiring C. auris was ≥4 hours. However, they were unable to identify a single specific source of transmission (40). Existing expert recommendations have emphasised the importance of prompt screening and infection prevention measures in healthcare settings to prevent nosocomial transmission of C. auris (43), as is supported by international guidelines. ECDC recommends prompt and robust measures if C. auris is detected in healthcare facilities including testing of close contacts (44), while CDC recommends testing patients sharing the same room, unit, or other care areas with a patient with C. auris (45). While these guidelines or their adherence have proven insufficient to prevent both major outbreaks and increasing prevalences (46), experiences from both the United Kingdom and Germany demonstrate that successful containment of such outbreaks is possible (16;47).
The included studies used different definitions of screening, and several studies presented results from contact tracing alongside the number of detections of C. auris in broad surveillance surveys or screening. Caution should be exercised when directly incorporating these results into decisions regarding admission screening recommendations. Conversely, numerous studies have reported secondary cases from screening patients who either shared a room as a positive case or occupied the neighbouring bed. This could support that an individual with a high degree of exposure is likely to have a high risk of becoming infected and may warrant their inclusion in screening protocols. Moreover, occupying the same environment as a C. auris case can potentially lead to colonisation or infection.
Associated factors of C. auris infection or colonisation
A total of 112 studies reported on factors associated with colonisation or infection with C. auris (Table 3). These studies were conducted in 30 different countries from all inhabited continents, with two studies including more than one country and two not specifying the country of sampling. Most studies primarily focused on detailing patients’ medical profiles or conditions without a control group and described critically ill patient populations or those with severe underlying medical conditions. ICUs were the predominant setting for these studies. Fourteen of the included articles included a control group and investigated factors associated with either colonisation or infection C. auris (Table 6) (31;48-60). These studies were predominantly small- scale (n<100) and based on outbreak investigations in heterogeneous populations. Most used bivariable screening to identify potential explanatory covariates, which were then included in multivariable regression models using stepwise selection techniques. However, this methodological approach may limit the validity of the identified factors due to several well- documented issues such as the multiple comparisons problem, exacerbated by the small sample sizes, and a lack of theoretical reasoning underpinning the modelling choices (61;62). Therefore, when reviewing these studies, counting how many times a factor has been found statistically significant in the included studies would not lead to robust inferences about their relative importance.
Exposure to the microbe is necessary for colonisation to occur. However, there have been very few studies investigating such exposure specifically, even though existing guidelines may target persons assumed to be exposed for instance through contact with healthcare system in endemic regions (63;64). We have only found one case series and one study examining the risk associated with travel and contact with healthcare abroad. In the case-series by Hamprecht et al., C. auris was found in seven patients in Germany between 2015 and 2017, six of whom had previously been treated in healthcare centres in the Middle East, Asia, Africa, or the United States (65). In a study of healthy travellers by Turbett et al. (66), 94 individuals were screened pre- and post-travel using self-collected axillary-inguinal groin swabs for C. auris. Travel destinations included all major regions of the world, with eastern Africa (31/94, 33%) being the most common region visited, and leisure being the predominant reason for travel (71/94, 76%). No C. auris was isolated from the samples. Few travellers (5%) reported providing or receiving medical care while traveling, indicating a low incidence of contact with healthcare systems during travel. Although the basis for inferences is sparse, the other studies align with the findings of these two studies, highlighting high-risk healthcare settings, particularly ICUs, as common sites of acquisition. However, recent studies confirm that C. auris may also persist in non-healthcare environments, such as stored fruit, where exposure to agricultural fungicides could contribute to antifungal resistance (67).
Prior reviews have identified correlations between C. auris infection with certain general factors, such as male gender along with severe underlying conditions such as immunosuppression, diabetes, and chronic kidney disease (8;68). Additionally, the presence of invasive devices and procedures is consistently cited as a risk factor (8;68;69). These general factors may be explained by our finding that most studies were conducted in critical care unit settings, as these patients are often implicated in hospital outbreaks or contract severe (and resistant) healthcare- associated infections, regardless of the specific causative agent. Some factors, like prior antifungal exposure, might be more specific, but it remains uncertain whether this selects for C. auris specifically or de-selects for other yeasts (8;68). Broad-spectrum antibiotics may also play a role by eliminating competing microbiota, thereby facilitating C. auris colonisation (68;69).
Colonisation with other MDROs could indicate shared risk factors between these multidrug- resistant pathogens (68). Past hospitalisation may also capture the statistical association of this shared risk set, and its inclusion in models should be carefully evaluated to first determine whether it acts as a mediator or confounder in the exposure-outcome relationship. The inclusion of these and other covariates without such a consideration may result in an attenuation of effect sizes, or even lead to incorrect conclusions.
The majority of studies which included a control group compared outcomes between patients with C. auris- candidaemia and those with other types of candidaemia. The research by van Schalkwyk et al. stands out due to its large scale, encompassing a national survey conducted over two years in an endemic region (60). This study evaluated a predefined hypothesis, among others, demonstrating a substantial increase in C. auris candidaemia. Rossow et al. concentrated their research on individuals colonised by C. auris, comparing them to non-colonised controls from identical exposed cohorts. They identified factors associated with colonisation, including the use of specific invasive medical equipment, recent use of antimicrobials, recent hospitalisations, and colonisation with other MDROs (31). When evaluating factors associated with infection or colonisation that are relevant to screening, the crucial variables are those helping us differentiate between incoming hospital patients who are at risk of being colonised and those with minimal risk. While comparisons with other candidaemias can highlight distinct features of C. auris, they offer limited insights into screening strategies for colonisation.
Colonisation sites
Twenty-nine studies reported on colonisation sites for C. auris and their positivity rates based on predetermined sampling materials (Table 7). The studies were conducted in 11 different countries, most of them in the USA (N=12), in the period 2013 to 2023. Armstrong et al. and Escandon et al. reported on the same materials sampled from patients, but not from HCWs (34;70). The studies varied in designs and populations, and revealed diverse patterns, prevalence, and positive sites for the detection of C. auris. Differences in diagnostic methods could be a source of the variation observed across studies. Among the 29 studies included, the majority (N=23) relied exclusively on culture-based methods, using either selective agar or broth enrichment, often followed by MALDI-TOF confirmation. Five studies incorporated PCR at some stage in the diagnostic process (63;71-74), and one study exclusively used PCR (75). While we did not observe any systematic differences in reported positivity rates between studies using PCR and those relying solely on culture-based methods, variations in diagnostic approaches may contribute to the heterogeneity seen in colonisation site detection.
The axillae and groin regions were the most common sites of C. auris colonisation, but also the body sites from which samples were most frequently taken. Positivity rates for screening sites in C. auris positive patients ranged from 36% to 100% for combined sampling of axilla and groin, 11.3% to 50% for axillae alone and 17.4% to 75% for groin separately (28;38;76). Colonisation of the nares was also observed, with positivity rates ranging from 17% to 64% (25;28;74-77). Piatti et al. used swab specimens from the skin of axilla, groin, auricular area and inframammary sulcus as screening sites and found a 27.3% (105/384) prevalence of skin colonisation of admission screenings in Italy (78). Additionally, positive cases were detected in other body sites such as rectum, oral cavity, respiratory tract, ear, urine, and wounds (76;79). Three studies reported separate results for the colonisation of C. auris on the hands of HCWs (34;38;39). Biswal et al. found that out of 145 screened HCWs, four (2.8%) tested positive. Whether the samples were reported as being taken bilaterally, in a composite swab, or individually varied between studies. ECDC recommends the axilla and groin as screening sites, as well as other sites (urine, wounds, catheter exit sites, throat etc.) if clinically relevant or indicated (44) and US CDC recommends bilateral screening of axillae and groin (80). Pan American Health Organization (PAHO) recommends screening the axilla, oropharynx, nostrils, groin, urine and rectum, but if it is not feasible to collect samples from all sites, they recommend at least pooled samples from the groin and axilla (81).
A literature review by the United Kingdom Health Security Agency draws attention to the study by Adams et al. where the addition of a sample from the nostril increases the probability of detection (25). However, Rowlands et al. and Southwick et al. reported lower positivity rates when axillae, groin and nares were combined as screening site than axillae and groin alone (74;75). Studies employing screening methods where the pre-test probability of a positive screening test is higher (contact tracing, PPS) also consistently identified the axillae and groin as primary sites for C. auris colonisation. Southwick et al. found that among 1668 patients, 144 (7%) were positive for C. auris colonisation, and positive results were predominantly observed in composite samples from the axilla and groin (81%), followed by the nares (64%) (74). Zhu et al. conducted an outbreak investigation in 2020 and included 11,035 samples of which 931 (8.4%) were positive for C. auris (76). The findings revealed that 80% (178/222) of positive samples were positive for C. auris from the axilla-groin (tested with a composite swab) and 58% (125/215) in the nares. They also found that if the nares are colonised, they harbour a relatively higher amount of C. auris than does the axilla or groin. Notably, when considering the combination of axillae-groin-nares, the positivity rate approached 100%. In other words, sampling from the axilla and groin bilaterally of patients at risk, as well as the nares, may find justification. While the cost-effectiveness of sampling from the nares warrants evaluation, the potential for increased detection rates may justify its inclusion, given the significant concerns surrounding over the spread of C. auris.
A limited number of the included studies aimed to discern the body sites where C. auris resides by testing various areas on infected patients. In only three studies, multiple body sites in patients with confirmed C. auris infection were screened (70;76;78). Armstrong et al. included five positive C. auris patients and screened each patient 10-11 times at different sites (70). There was significant variability in the body sites where C. auris was detected in screening. Zhu et al. reported on 298 microbiological samples where individuals had an infection, but the results reporting positive body sites are presented in combination with results from a surveillance survey, precluding data extraction regarding positive patients alone (76). Piatti et al. conducted a study using skin and rectal swabs to screen for the presence of C. auris on all patients admitted to at-risk units (78). Upon analysing patients with confirmed C. auris infection, the prevalence of colonisation in both skin and rectal areas was found to be nearly identical. Additional evidence suggests that screening multiple body sites improves detection rates (82).
A few of the included studies also tested wounds, tracheal secretions, and catheter urine and detected C. auris. Invasive devices have been found to be a factor associated with colonisation, although this may only apply to critically ill patients. Furthermore, some studies contain description of screening indwelling devices when present, but whether it was tested from these locations is uncertain, and no results are reported from this sample material (40;63;79).
However, there are different definitions in the literature regarding whether urine as a specimen should be considered colonisation or infection depending on clinical presentations/symptoms. The studies incorporated, which include description of urine samples, do not provide detailed descriptions. The same may apply to wounds, although Zhu et al. consider wounds as a source of colonisation (76).
Strengths and limitations of this review
Our review has several strengths. First, we employed a systematic and reproducible search with a comprehensive strategy, and a thorough review of included studies. We chose to do a broad search for all articles on C. auris and have likely been able to identify all literature relevant to our outcomes of interest. Furthermore, findings were presented in the context of all existing studies, providing the most up to date overview yet. Finally, our systematic approach avoids some of the pitfalls and biases of purely narrative reviews without systematic searches. While a narrowing of the inclusion criteria could have addressed some of the heterogeneity issues we revealed, such an approach would perhaps exclude a significant proportion of the current literature. We opted for a narrative summary of our findings to provide the most comprehensive overview of this broad research field, as the heterogeneous nature of the included studies precluded a meta- analysis or other quantitative synthesis. A formal quality assessment was not performed, though this limitation is somewhat mitigated by the high variability in study designs and settings. The open nature of our research question, however, introduces a risk of creating a less focused review, which may not allow for a full assessment of the strength of specific evidence.
Additionally, without a quantitative synthesis, the findings lack the weighted perspective that a meta-analysis could offer in highlighting patterns across studies.
Conclusion
Our systematic review summarises several key findings regarding the associated factors of colonisation/infection, the duration of colonisation, the potential for secondary cases, and possible colonisation sites of C. auris. However, our primary finding was that in screening for C. auris, several major knowledge gaps remain. Most studies reviewed described circumstances or outbreaks that arose unexpectedly, and lacked pre-developed protocols and a clear study aim that could be effectively addressed with the selected methods. The publication bias introduced by such a circumstantial body of literature must be acknowledged. If targeted screening of high- risk populations is chosen as a strategy to prevent establishment of C. auris, it remains unclear who should be screened, and there is no robust evidence either for or against current practices. Identifying individuals likely to have been exposed to C. auris remains the most critical factor in such guidelines, as highlighted also by others (68). Screening should continue to be performed in the axillae and groin, but considering additional sites like the nares could increase sensitivity, although there is conflicting evidence. Furthermore, routine screening of HCWs appears unnecessary unless there is clear evidence of direct exposure to contaminated settings. Given the lack of evidence that spontaneous decolonisation is a common occurrence among patients, it might be prudent to view exposed patients as potentially persistently colonised as a precaution.
Moving forward, the knowledge gaps we have identified herein should be addressed. There is an urgent need for primary research that methodically outlines the efficiency of current screening programs, including their sensitivity and specificity. This research should include findings from different anatomical sites and comparisons with composite swabs. Notably, given the substantial heterogeneity in the reported outcomes, it was difficult to identify any systematic difference between culture-based and molecular methods in the included studies. However, as challenges in detecting C. auris using culture alone are well-documented and recent studies have highlighted discrepancies between these methods, care should be taken in future research to ensure methodological consistency and accuracy (63). A suggested screening protocol advocates for using the most sensitive and specific method, PCR, initially, followed by culturing, an approach that appears promising (63). To determine colonisation duration, it is crucial to register positive patients and systematically monitor them at pre-set intervals over a specified period to delineate the natural timespan of C. auris colonisation. Furthermore, it is crucial to map how the hospital environments act as a vector in the transmission of C. auris. This not only includes the surfaces of the hospital, its ventilation, water and sanitation systems, but also the surrounding nature and animals. The emergence of C. auris fundamentally pertains to One Health, and this multidisciplinary approach should remain the predominant for addressing C. auris establishment in healthcare settings. Finally, the study of factors associated with C. auris colonisation should focus on identifying factors able to discriminate high-risk patients that should be screened in larger cohorts that are representative of all patients in healthcare settings. In addition to factors that may be coded from the medical charts, history of travel and contact with healthcare in endemic regions should be included in such studies. It is crucial to acknowledge that the most significant “risk factor” for contracting C. auris is exposure to the microbe itself. Such research initiatives as mentioned here could greatly bolster our comprehension, thus aiding the formulation of more effective screening guidelines and infection control measures for C. auris. In the absence of these research initiatives, however, the success of C. auris screening is more likely if it is grounded on medical theory, current knowledge of yeast microbiology, and existing literature on screening against other similar microbes.
Data Availability
All data produced in the present work are contained in the manuscript.
Data Availability
All data produced in the present work are contained in the manuscript.
Author contributions
LEØ, RR, and MM conceptualized the study and designed the methodology with JPWH. Data curation was performed by RAT. LEØ, MM, and ASD conducted the investigation and formal analysis. Project administration was overseen by MM and HMEV, while JPWH provided supervision. The original draft was written by ASD, LEØ, and MM. All authors reviewed and edited the manuscript.
Conflict of interests
None to declare.
Acknowledgements
We extend our thanks to all colleagues at the Norwegian Institute of Public Health (NIPH) and Statens Serum Institut (SSI) who have engaged in discussions on this topic, particularly those involved in the project to develop updated recommendations to contain and prevent the spread of resistant microbes in our countries.
Appendix 1: search strategies
Footnotes
Several changes in this version, but the major are: Merging results and discussion into a new section called 'Findings', and adding data extraction and discussion on diagnostic methods.
References
- 1.↵
- 2.↵
- 3.
- 4.
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.
- 73.
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.
- 84.
- 85.
- 86.
- 87.
- 88.
- 89.
- 90.
- 91.
- 92.
- 93.
- 94.
- 95.
- 96.
- 97.
- 98.
- 99.
- 100.
- 101.
- 102.
- 103.
- 104.
- 105.
- 106.
- 107.
- 108.
- 109.
- 110.
- 111.
- 112.
- 113.
- 114.
- 115.
- 116.
- 117.
- 118.
- 119.
- 120.
- 121.
- 122.
- 123.
- 124.
- 125.
- 126.
- 127.
- 128.
- 129.
- 130.
- 131.
- 132.
- 133.
- 134.
- 135.
- 136.
- 137.
- 138.
- 139.
- 140.
- 141.
- 142.
- 143.
- 144.
- 145.
- 146.
- 147.
- 148.
- 149.
- 150.
- 151.
