Know your Microbe Foes: The Role of Surveillance in Combatting Antimicrobial Resistance

Abstract

Antibiotic-resistant organisms (AROs) are difficult and costly to treat, associated with high mortality rates, and are on the rise. In the United States, there is limited tracking of AROs, which can contribute to transmission and inhibit infection prevention interventions. Surveillance is limited by a lack of standardized methods for colonization screening and limited communication regarding patient ARO-status between healthcare settings. Some regional surveillance and reporting efforts are in place for extensively-resistant AROs such as carbapenem-resistant Enterobacterales (CRE), but need to be further expanded nationwide and to include other AROs such as extended-spectrum β-lactamase (ESBL) producing organisms. Increased surveillance of ARO infections and colonization will inform future targeted intervention and infection prevention strategies.

Introduction

Infections due to antibiotic-resistant organisms (AROs) are costly, have fewer treatment options, and have high rates of morbidity and mortality. The Centers for Disease Control and Prevention (CDC) estimates more than 2.8 million infections with AROs occur in the United States each year with more than 35,000 deaths [1].

Carbapenem-resistant Enterobacterales (CRE) and extended-spectrum β-lactamase producing Enterobacterales (ESBL-E) are AROs which pose urgent and serious threats, according to the CDC’s 2019 Antibiotic Resistance Threats Report [1]. CRE are resistant to at least one carbapenem antibiotic and may be resistant to additional classes of antibiotics, leaving few treatment options. ESBL-E harbor genes which encode extended-spectrum β-lactamases. For surveillance purposes, the CDC phenotypically defines ESBL-E as resistant to at least one extended-spectrum cephalosporin (ceftazidime, cefotaxime, or ceftriaxone) [2]. As with CRE, these bacteria can be resistant to other antibiotic classes, which can limit treatment. Bacteria in the Enterobacterales order, examples of which include Escherichia coli and Klebsiella pneumoniae, can cause a variety of life-threatening infections such as urinary tract infections (UTIs), blood stream infections (BSIs), and pneumonia. In the US alone, CRE and ESBL-E infections resulted in an estimated 1,100 and 9,100 deaths, respectively, in 2017 [1]. People can also be asymptomatically colonized with CRE, ESBL-E, and other AROs.

Microorganisms, including AROs, can transfer from person-to-person as well as between people and the environment [3,4]; therefore, surveillance is critical to stop transmission and create appropriate interventions. Surveillance for many AROs is limited, which may lead to further dissemination within hospitals, between hospitals, and throughout the community. This perspective will focus on ARO colonization and infection surveillance, with a particular emphasis on CRE and ESBL-E, specifically ESBL-E. coli.

Topics

ARO Infection Surveillance and Reporting

Infections due to AROs are the result of a clinical syndrome due to the presence of an ARO. While CRE infections have remained relatively stable since 2013, ESBL-E infections have increased, both in the hospital setting and in the community [1]. More public health measures exist for CRE surveillance and reporting, yet there is a marked gap in response for ESBL-E. Standardized methods and support for detecting ESBL-E colonization, tracking ESBL-E infections, and preventing ESBL-E transmission are lacking. For instance, there are no current recommendations for standardized universal screening for ESBL-E in the acute care setting. This hinders the ability to track and prevent ESBL-E transmission. More effort is needed to better understand and prevent ARO infections in healthcare and community settings.

The number of infections in both hospital and outpatient settings due to AROs is difficult to assess due to a lack of mechanism for reporting and tracking. In order to bill for hospitalizations, healthcare systems use national medical coding systems, such as the International Classification of Diseases, Tenth Revision (ICD-10). Infections as a diagnosis are generally well-coded, and when a patient has an infection, a coder assigns an appropriate ICD-10 code, sometimes with additional codes or modifiers to report the organism and antimicrobial resistance. However, both organism and drug resistance coding are frequently limited or inaccurate due to poor coding practices [5]. In one study, only 65.4% of ARO Enterobacteriaceae had the properly coded organism, and 3.3% had an ARO code [5]. Therefore, the use of hospital administrative data alone cannot reliably assess the frequency of AROs in hospitals. While organism and susceptibility pattern data can be attained through hospital microbiology laboratories, these data are not readily reported outside of the local level, and thus are of limited utility for nationwide tracking.

Limited surveillance and communication of a patient’s ARO status occurs across healthcare networks. Patients often visit multiple healthcare facilities; however, information regarding a patient’s ARO status may not be communicated across healthcare facilities as this is generally not required or regulated. For instance, during a CRE outbreak event in Chicago, patients of a particular hospital later visited other hospitals, including several where CRE was then later reported [6]. The outbreak was trackable due to the presence of New Delhi metallo-β-lactamase (NDM), a rare carbapenemase in Chicago at the time of the study. The hospital which shared the most patients from the original outbreak location had the highest incidence of outbreak-associated CRE [6]. This highlights the need to set up more regional surveillance of AROs to prevent dissemination among healthcare networks. Illinois has mandated CRE reporting through its extensively drug resistant organism (XDRO) registry, thus improving CRE surveillance [7,8]. Additionally, the registry works to improve inter-facility communication as healthcare facilities can use the registry to determine if a patient has previously had CRE. Systems, such as the Illinois XDRO registry described above, which promote communication between healthcare facilities regarding ARO-status are not routine. Enhanced ARO communication would allow for potential interventions such as contact isolation, enhanced environmental hygiene, or additional personal protective equipment requirements to prevent ARO transmission to others.

On a broader level, there is a lack of national reporting and surveillance of AROs. Carbapenemase-producing CRE has been a nationally notifiable organism to the CDC’s National Notifiable Diseases Surveillance System (NNDSS) since 2018 [9]. As a notifiable organism, states voluntarily provide case surveillance data [10]. The CDC’s Antibiotic Resistance Laboratory Network (ARLN) tracks carbapenemases and performs CRE colonization screening, if requested, following reported CRE-infections [11], but does not conduct CRE surveillance [12]. Other AROs, such as ESBL-E, are not currently notifiable at the national level.

Due to the lack of ARO infection tracking, incidences of ARO infection can be difficult to determine. As of July 2019, the Multi-site Gram-negative Surveillance Initiative (MuGSI) [2], an activity of the Healthcare-associated Infection Community Interface component of the CDC’s Emerging Infections Program (EIP), began conducting ESBL-E population- and laboratory-based surveillance. In an EIP pilot study, which occurred over a 3-month period in 2017 in select counties in New Mexico, New York, and Tennessee, Duffy et al. reported an overall ESBL-E infection incidence of 199.7 per 100,000 population [13]. Forty-seven percent of the ESBL-E infection cases in the study were considered community-associated [13], pointing to a need to strengthen community and environmental surveillance strategies. Importantly, MuGSI ESBL-E infection surveillance occurs across geographic regions, but as of 2022, is still limited in scope, occurring in one county each in Colorado, Maryland, New Mexico, and New York, two counties in Georgia, and four counties in Tennessee [2]. MuGSI also tracks CRE and carbapenem-resistant Acinetobacter baumanii.

The CDC’s National Healthcare Safety Network (NHSN) tracks the following healthcare-associated infections (HAIs) in inpatients: central line-associated blood stream infections, catheter-associated UTIs, surgical site infections, some ventilator-associated events, methicillin-resistant Staphylococcus aureus (MRSA) bacteremia, and Clostridioides difficile events [14]. For 2020, the NHSN reported 24.7% of E. coli isolates were resistant to extended-spectrum cephalosporins (ESCs) with an estimated 24.4% of isolates resistant in pediatric populations and 24.7% resistant in adults [15,16]. The NHSN data does not include all types of HAIs, nor does it encompass community-level data. Further, the data reported to NHSN may vary depending on state reporting requirements [17]. However, what these proportions show is that some AROs, including ESC-resistant E. coli (ie, potential ESBL-E. coli), are readily found in healthcare settings. Therefore surveillance and interventions are needed to prevent the spread of AROs in hospitals and beyond.

The SENTRY Antimicrobial Surveillance Program was established in 1997 and tracks pathogens and antimicrobial resistance trends from healthcare-associated and community-onset infections at participating centers worldwide [18]. In 2016, the prevalence of β-lactamase resistance genes among E. coli isolates was 11.6% for E. coli-causing UTIs and 16.1% for E. coli-causing BSIs among hospitalized patients from 36 states [19]. Of note, the SENTRY program has publicly available data, which is stratified by US census regions [20]; however, it is unclear which states or cities have hospitals participating as the SENTRY program is a service offered through JMI laboratories.

ARO Colonization Surveillance

In contrast to ARO infection, ARO colonization is asymptomatic. Depending on the ARO, colonization can occur at various body sites (eg, skin, gut, nares), and often precedes infection [21-23]. In particular, the gut can serve as a reservoir for AROs [22]. Gut colonization with ESBL-E. coli has been linked to increased risk for ESBL- E. coli UTIs [24]. The CDC estimates that 47% of ESBL-E infections are community-associated, making the spread of ESBL-E challenging to detect and contain [1]. Colonization surveillance across regions and demographics could provide insight into both community and healthcare-associated spread of AROs.

Globally, ESBL-E intestinal colonization prevalence is increasing worldwide, although the prevalence varies by region [25,26], with Southeast Asia noting the highest intestinal carriage of ESBL-E. coli among healthy individuals (27% prevalence) [25]. Fecal colonization surveillance in the US is limited, particularly among healthy individuals. In a 2004 study, which took place in Minnesota and Wisconsin, authors reported 0% and 2% prevalence of cephalosporin-resistant E. coli in 100 healthy vegetarians and 567 hospitalized patients, respectively [27]. In a 2011 study, 2% of 101 US military personnel based in Texas were colonized with multi-drug resistant E. coli, compared to 11% of 100 Afghanistan-based US military personnel [28], suggesting environmental risk factors for ARO colonization. Further, in a small 2012 study comprised of 60 travelers attending a New York travel medicine clinic, 1.7% of participants were colonized with ESBL-E. coli prior to traveling, and 25% of participants were colonized post-travel [29]. Given the fast-paced changing landscape of antimicrobial resistance, with estimated ESBL-E. coli or ESBL-E prevalence increases of 1.5-5.4% per year among healthy individuals [25,26], more current studies are needed to determine present-day ESBL-E. coli carriage throughout the US.

Data on ESBL-E colonization in the pediatric setting are even more limited. In a large pediatric multi-state study involving California, Texas, and Tennessee from 2013-2015, authors reported a 3.5% prevalence of ESBL-E. coli among 519 children, ages 14 days to 14 years [30]. It is unclear whether children in the US have higher or lower ESBL-E colonization than healthy adults.

Comparatively more studies are available, some highlighted here, regarding ARO colonization in long-term care facilities, nursing homes, and long-term acute care hospitals. In a 2016-2017 SHIELD Orange County point prevalence study from 18 nursing homes in Southern California, ESBL-E carriage was 34% (range 0-66%) [31]. The study also reported colonization prevalence of other AROs, specifically CRE, vancomycin-resistant Enterococci (VRE) and MRSA [31]. Like the SHIELD study, other groups have also taken a more regional approach to examine ARO colonization across healthcare facilities. In the Washington D.C. area, CRE colonization prevalence was 7% among seven long-term care facilities, with 5% prevalence among the area’s acute-care hospitals. ESBL-E prevalence was not assessed at any of the sites [32]. Therefore, some regions are beginning to survey ARO colonization across healthcare settings, including nursing homes; however, surveillance timing and methods are not standardized. Non-standardized surveillance approaches may be limited to isolated, single time point research studies and/or restricted to tackling surveillance of only the most extensively-drug resistant organisms such as CRE.

Domestic pets and farm animals can also be colonized with AROs. Cattle, pigs, sheep, poultry, and dogs have all been colonized with ESBL-E. coli [33,34]. The significance of animal reservoirs and routes of transmission (people to animals or the environment and vice versa) are still being determined. Surveillance studies incorporating a One Health approach (people, animals, and the environment) can aid in this assessment. In a study on Reunion Island (La Réunion, Département et Région d’Outre-Mer, France), located in the Southwest Indian Ocean, Miltgen et al. found a marked difference in ESBL-E. coli sequence types and resistance genes from animal sources (ST57, ST156; bla_CTX-M-1) compared to human sources (ST131, ST38, and ST10; bla_CTXM-15, bla_CTX-M-27, and bla_CTX-M-14), suggesting that in this setting animal-to-human transmission is not driving ESBL-E. coli colonization or infections in people [34]. Of note, this study did not include isolates from domestic animals living in shared spaces with people, but instead focused on cattle, pigs, poultry, and small ruminants [34].

ESBL-E Colonization and Infection Risk Factors

Both global and US-based studies have revealed potential risk factors for ESBL-E colonization in people. Risk factors for ESBL-E colonization in healthy individuals globally include prior antibiotic use (past 4 or 12 months) and international travel [26]. In the SHIELD study, ESBL-E colonization was associated with GI devices, a history of ESBL-E, and history of VRE [31]. In a systematic review and meta-analysis of >29,000 patients, gastric acid suppression was associated with CRE, ESBL, and VRE colonization [35], although infection data from UTIs were also included in the analysis. Household contacts are also an important consideration as patients can remain colonized, and strains can be shared among household contacts. In a study involving patients with community-associated ESBL-E infections in Spain, 16.7% of patients’ household members were also colonized [36].

Colonization is a risk factor for ESBL-E infection [23,37,38]. In one study, wherein patients were screened for ESBL-E on admission to the ICU, 25% of the ESBL-E-colonized patients had a positive ESBL-E clinical culture during their hospital stay compared to 0.6% of those not colonized with ESBL-E [38]. Other risk factors for ESBL-E infection include prior antibiotic use [39,40], UTI history [39], chronic indwelling vascular hardware [40], age ≥43 years [40], recent hospitalization [40,41], and residence in a long-term care facility [41].

While many healthcare and environmental risk factors are now known, key epidemiological data are missing. It is unclear whether certain regions of the US have more ESBL-E colonization and infection or if certain populations (eg, healthcare personnel, agricultural workers, those living in shared housing) have increased risk.

ESBL-E Strain Characteristics

Ideally, increased ARO surveillance and reporting would be coupled with genome data to allow rapid early detection of new or emerging antimicrobial resistance genes and strains. Strain data is important for outbreak tracing and for potential differences in clinical outcomes. Worldwide among ESBL-E, CTX-M-type β-lactamases predominate, particularly those encoded by bla_CTX-M-15 and bla_CTX-M-14, with bla_CTM-M-27 emerging [42,43]. These ESBL-genes are often plasmid associated, which is concerning, as plasmids and their associated antibiotic-resistance genes are transferable between strains or even across larger taxonomical families (eg, all Enterobacteriaceae). ST131 is the E. coli strain most associated with CTX-M carriage. Among a subset of E. coli isolates that were sequenced in the 2017 EIP surveillance study, most isolates harbored CTX-M-type β-lactamases (98%), and ST131 was the most common strain type (53.6%) [13]. Similarly, the SENTRY program determined ST131 accounted for 53.6% of E. coli BSI isolates and 58.2% of UTI isolates from hospitalized patients throughout the US, with nearly all ESBL-E. coli isolates carrying CTX-M-type β-lactamases (99.3%) [19]. However, given the lack of community ESBL- E. coli colonization data, it is unclear if ST131 is the predominant ESBL-E. coli colonizing strain in communities in the US and which colonizing strains and plasmids are more associated with infection or transmission. Knowing more about the strains of ESBL-E. coli colonizing the gut and their associations with infection may help develop strain-specific strategies for prevention. For other AROs, the microbiome and metabolome play a role in colonization resistance, but a recent report from the Netherlands suggests this may not be the case for ESBL-E. coli [44].

Future Research and Potential Interventions

The CDC has outlined multiple approaches for managing AROs, broadly grouped into categories of administrative support, judicious use of antimicrobials, surveillance, standard and contact precautions, environmental measures, education, and decolonization [45] Successful ARO control may require a multi-category approach, and efforts specifically targeting one ARO may also affect others. In one study targeting CRE-colonization in long-term acute care hospitals, not only was carbapenemase-producing Klebsiella pneumoniae colonization and infection decreased during the multi-faceted intervention, but so was all-cause bacteremia [46].

For ESBL-E infections, surveillance programs such as the SENTRY program and MuGSI are making strides in assessing ESBL-E infection surveillance nationwide. Future research is needed to determine the most effective interventions that prevent ARO dissemination in hospitals and the community. Additionally, collaborative approaches should be standardized and communication regarding the ARO-status among transferred patients amongst facilities should be enhanced, allowing facilities the opportunity to take infection prevention precautions.

Colonization surveillance is key to identifying and containing ESBL-E spread in the US. Active surveillance cultures (ASC) can identify people who are asymptomatically colonized with AROs in order to implement preventative measures. This may be limited due to cost and personnel resources, and intervention and surveillance strategies may differ depending on whether the community is a high- or low-endemic setting. In hospital settings, ASC colonization screening could take place upon hospital admission followed by increased infection prevention measures (eg, isolation precautions, single-use equipment) to prevent transfer between patients and healthcare personnel. For those who are colonized and most at-risk, microbiota-based strategies may be considered such as fecal microbiota transplant or engineered probiotics to prevent or reverse colonization.

More complete epidemiological studies will foster targeted interventions. Given some of the known risk factors for ARO and ESBL-E colonization and spread, specific interventions and practices could be established for certain at risk groups such as international travelers, residents of long-term care facilities, or household contacts. Antibiotic use is a risk factor, therefore antimicrobial stewardship is necessary in inpatient and outpatient settings. Future research is needed to find factors which protect against ARO colonization among those who are not yet colonized, and how to prevent community spread and progression to infection in those who are. Strain-specific data may be able to provide some insight into the emergence of new strains or strains of highest concern, which will allow a faster public health response. Some AROs such as ESBL-E also have potential reservoirs in the environment and animals Therefore, a multisector approach, which incorporates surveillance of people, animals, and the environment would assist in targeting prevention of transmission of AROs from different sources. For ESBL-E, present studies suggest that interventions with people, not food or agriculture would be most effective [47].

Conclusions and Outlook

ESBL-E colonization and infection incidence are increasing [25,48]; however, gaps remain in ESBL-E infection surveillance, and ESBL-E colonization surveillance is less standardized. Limited surveillance and reporting coupled with lack of interventions in the United States and worldwide will only lead to further ESBL-E transmission. While this perspectives piece has primarily focused on ESBL-E, the need for surveillance and reporting of both infections and colonization could be broadly applied to other AROs. It is time to accelerate testing of ARO colonization and infection prevention strategies. As many people are asymptomatically colonized and colonization does not pose an immediate risk, the challenge will be to find a strategy that is cost-effective, scalable, and where the benefits outweigh potential risks. To combat antimicrobial resistance, we first need to better understand the scope of the problem, from colonization prevalence to strain data, to best target infection prevention strategies for those most at-risk.

Glossary

ARLN

Antibiotic Resistance Laboratory Network

ARO

antibiotic-resistant organism

ASC

active surveillance cultures

CDC

Centers for Disease Control and Prevention

CRAB

carbapenem-resistant Acinetobacter baumanii

CRE

carbapenem-resistant Enterobacterales

EIP

Emerging Infections Program

ESBL

extended-spectrum β-lactamase

ESBL-E

extended-spectrum β-lactamase producing Enterobacterales

ESC

extended-spectrum cephalosporin

HAI

healthcare-associated infection

ICD-10

International Classification of Diseases, Tenth Revision

MRSA

methicillin-resistant Staphylococcus aureus

MuGSI

Multi-site Gram-negative Surveillance Initiative

NHSN

National Healthcare Safety Network

UTI

urinary tract infection

VRE

vancomycin-resistant Enterococci

XDRO

extensively drug resistant organism

Author Contributions

REB (ORCID: https://orcid.org/0000-0002-4844-4212) and JHK wrote the manuscript.