Understanding the landscape of carbapenemase-producing organisms (CPOs), and spotlighting opportunities for control in England

SUMMARY

Carbapenemase-producing organisms (CPOs) are an increasing global public health threat for which there are limited effective and affordable therapeutic options. The rise in the incidence of CPO infections and colonisations recorded across the UK and beyond over the last 20 years necessitates a need to review and update strategies for control. It is important to review other countries’ frameworks for CPO control and significant CPO outbreaks as these could inform the design of an efficient public health response. Despite many nations reacting to the rise by upscaling public health surveillance of CPOs (and the introduction of mandatory notification in England), improvements in laboratory detection methods, and the linkage of data across jurisdictions, significant challenges remain. For example, though predominantly acquired via cross-transmission between patients in healthcare, there are reports of human infections putatively acquired from CPOs harboured in the natural environment. Given the role of one-health in AMR, this is an important consideration. In this article, we outline some of the CPO control strategies adopted across different countries to give a sense of the global picture, and expand on why, in addition to controls in healthcare, there is a strong need to consider a One-Health approach. We describe the existing framework for CPO control in England and emphasise the importance of an integrated, multi-disciplinary and cross-sectoral strategy for dealing with the multifaceted problem posed by CPO in England.

Background

Carbapenemase-producing organisms (CPOs) are a serious and increasing public health threat [1] for which there are limited effective and affordable therapeutic options, particularly in oral formulation. In 2017, the World Health Organization (WHO) declared carbapenem-resistant Enterobacterales, Pseudomonas aeruginosa and Acinetobacter baumannii the top three critical pathogens for which new antibiotics are urgently required [2,3]. In 2024, critical status was maintained for all except carbapenem-resistant P. aeruginosa which was downgraded to the high priority group [4].

Although CPOs can colonize individuals without causing infection, CPO infections currently primarily affect immunocompromised and critically ill patients [13]. Infections with CPOs are associated with clinical outcomes, prolonged hospitalisation and higher mortality rates in comparison to those with carbapenem-susceptible strains [[14], [15], [16]]. CPO infections have been associated with mortality rates as high as 31–53% in studies conducted in China, Singapore, Saudi Arabia and Brazil [[17], [18], [19], [20]]. A study in New York on NDM-positive CPO bacteraemia (with Klebsiella pneumoniae being the predominant organism isolated (40/61, 66%)), had mortality rates of 42% in 2022 and 33% in 2023, while those with NDM-positive urine infections had a mortality rate of 6%. The last 20 years have witnessed a rise in the incidence of CPO infections across the UK and elsewhere [1,11]. In England, this has proven to be challenging and financially costly for NHS Integrated Care Boards and acute trusts [12].

Carbapenem resistance is mediated by several mechanisms: carbapenemase enzymes, efflux pumps, porin loss, alteration to penicillin-binding proteins and biofilm production [5]. Carbapenemase enzymes break down carbapenem antibiotics, rendering them ineffective. Efflux pumps work by actively transporting carbapenems out of the cell, preventing them from reaching their target. Penicillin-binding proteins are the targets of carbapenems. Therefore, alterations to penicillin-binding proteins can reduce the affinity of the antibiotic for its target. Several factors mediate the increased resistance to antibiotics which is characteristic of some bacteria in biofilms, and these include reduced penetration (physical barriers), altered metabolic activity and genetic adaptations.

Some bacteria are intrinsically resistant to carbapenems while others acquire resistance, often in the form of carbapenemases. Carbapenemases are classified into two types: serine carbapenemases (Ambler class A or D) and metallo-β-lactamases (Ambler class B), which hydrolyse most β-lactam antibiotics [6]. There are several families of carbapenemases reported globally of which the K. pneumoniae carbapenemase (KPC), oxacillinase-48-type carbapenemases (OXA-48-like), New Delhi metallo-β-lactamase (NDM), imipenemase metallo-β-lactamase (IMP) and Verona integron-mediated metallo-β-lactamase (VIM) families (often referred to as the ‘big five’) are the most commonly identified [7].

CPOs are frequently resistant to multiple antibiotic classes [8]. Indeed, the resistance genes are often encoded on plasmids and other mobile genetic elements (MGE) and can therefore easily be acquired by organisms already carrying other types of resistances. Resistance genes to different antibiotic classes, including carbapenemases, can be co-located in the same bacteria thanks to MGEs. The resulting multi-drug resistance significantly reduces the therapeutic options available to treat CPOs. As Enterobacterales commonly give rise to community-associated and healthcare-associated infections [6], the acquisition of MGEs harbouring resistance mechanisms gives greater cause for concern.

Though the most frequently reported CPOs in English healthcare settings are Enterobacterales, (known as carbapenemase-producing Enterobacterales (CPE)), carbapenemase-producing P. aeruginosa and A. baumannii are well-described nosocomial pathogens, particularly in intensive care units (ICUs) [9]. Carbapenemase-producing A. baumannii has been associated with an increased risk of mortality ranging from 8% to 40% from global data [10].

Both clonal expansion and horizontal transfer of carbapenemase genes on MGEs (e.g. transposons and plasmids) contribute to the spread of CPOs [21]. Person-to-person transmission can occur through patient transfers within and between healthcare facilities including across international borders [13]. Health tourism, including cosmetic tourism, and medical repatriation are recognised routes of introduction of resistant bacteria to receiving institutions [22]. Wound infection represented approximately 25% of cosmetic tourism complications presenting to the NHS in one study [23].

Community transmission in long-term care facilities as part of outbreaks has been described, though further studies on community transmission are needed to determine its role in CPO spread. Transmission of CPOs occurs through direct contact or indirectly, via shared equipment or common reservoirs [24], such as hospital sinks. Carriage of CPOs in the gut of humans can be prolonged and symptomless, contributing to their propagation. There are currently no standardised methods to eradicate CPO carriage. This is in direct contrast to methicillin-resistant Staphylococcus aureus (MRSA), that was successfully controlled in England, in part, by reducing carriage using decolonisation as part of a whole bundle approach [25].

The utility of the One-Health approach in addressing CPO transmission

It is thought that the predominant means of acquisition of CPOs is via human-to-human spread, for example in healthcare, but there have been reports of human infections putatively acquired from environmentally sourced CPOs [26]. Previous studies have reported the presence of carbapenemase genes in the natural environment on almost every continent. Hospital and municipal wastewater, drinking water, recreational waters, wildlife, companion animals, agricultural environments, and retail food products have been identified as current reservoirs of CPOs and carbapenemase genes [26]. Though difficult to prove human acquisition of CPOs from the environment, cases have been reported from China and France [27,28]. A study conducted in China describes whole genome sequencing (WGS) of bla_NDM-positive E. coli shared among farms, flies, dogs and farmers, providing direct evidence of carbapenem-resistant E. coli transmission and environmental contamination [27]. There was also a human case of bacteraemia caused by a carbapenemase-producing Enterobacter asburiae after accidental near-drowning in a French river^28. Identical antibiotypes were identified from E. asburiae samples collected from the river water and the patient's blood, sharing the same genomic fingerprint by pulsed-field gel electrophoresis (PFGE). No matching strain of E. asburiae was isolated from the hospital environment where the patient was treated during a 2-year period [28]. Thus, the authors conclude that there was transmission from river to human [28]. All of this underscores the importance of considering the environment to appreciate the full picture of CPO transmission.

The use of antibiotics as clinical agents has generated pressures favouring the evolution and dissemination of resistance genes. Subclinical concentrations of antibiotics in vivo and in vitro promote the mobilization and horizontal transfer of a large range of antibiotic resistance genes (ARGs) in many bacterial species. Furthermore, resistant bacteria from animals and humans have the potential to transmit in both directions, through human contact with animals or their environments, through ingestion of contaminated food and through contact with effluent waste from humans, animals and industry [29].

There are increasing examples of CPOs in the environment [30] including agricultural sectors [31]. In this respect, the soil plays a key role in the evolutionary dynamics of antimicrobial resistance development [32]. Anthropogenic activities have affected soil microbial diversity and facilitated the spread of clinically relevant ARGs in the environment [31]. Consequently, these bacteria might already be sheltered in soils, ground water or surface water which provide ideal conditions for long-term harbouring of these strains, amplification, and dissemination to multiple sources [31]. Furthermore, the presence of medically relevant CPOs in soils surrounding urban areas can accelerate their transfer to humans [31].

Studies conducted in Brazil and China have identified bla_NDM, bla_VIM, and bla_IMP in E. coli, Citrobacter freundii, Citrobacter sedlakii, Enterobacter cloacae complex, and P. aeruginosa strains from soils [33,34] Although data regarding the dissemination of CPO in soils is limited, there is evidence for interspecies and intraspecies transmission of carbapenemase-encoding genes in soil [31]. Although carbapenems are not used in animals in the UK, there is some potential evidence of human sewage infecting wildlife [35]. The third UK One-Health report contains a case-study of the likely origin of carbapenem-resistant (resistance mechanism being bla_OXA-48) bacteria in a seal, which was collected from a beach near a sewage outlet, the inference being that human waste was the vehicle for the resistance gene and its carrier plasmid.

Wastewater can be a reservoir for CPO. Suggested strategies for wastewater management to prevent CPO transmission include proper waste disposal, use of targeted cleaning protocols, minimising wastewater exposure and environmental cleaning and decontamination. With respect to proper waste disposal, the use of designated protocols for disposing wastewater and segregation of liquid waste from patient care areas, especially those with known CPO colonization or infection, have been suggested by some. Measures to minimise wastewater exposure include the reducing the number of handwashing basins and shower drains, particularly in patient rooms (where feasible).

As bacteria and genes can cross environments and species boundaries [36], it is critical to understand and acknowledge the connections between the human, animal and environmental microbiota (the One Health Concept) to manage this global health challenge.

Examples of CPO surveillance, significant outbreaks and control outside of England

In Europe, large surveillance networks exist for CPOs. Coordinated by the European Centre for Disease Prevention and Control (ECDC), the European Antimicrobial Resistance Genes Surveillance Network (EURGen-Net) is one such network for genomic-based surveillance of multidrug-resistant bacteria that involves the national reference laboratories or equivalent laboratories of 37 European countries [37]. There is a Joint Action Plan between the UK Health Security Agency (UKHSA) and ECDC which facilitates the participation of England in this surveillance network [38]^. In 2013, the ECDC launched a self-assessment tool called the “European survey of carbapenemase-producing Enterobacteriaceae (EuSCAPE)”. For the EuSCAPE project, a joint agreement on diagnostic and improvement of quality-assessed diagnostic capacity among national expert laboratories (NELs) was reached to facilitate a structured survey using a standard sampling protocol in all participating sites. Technical staff from all NELs were trained to use a set of standard phenotypic and genotypic tests in accordance with European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines (version 1.0, December 2013). Subsequently, the United Kingdom National External Quality Assessment Service (UK NEQAS) conducted an External Quality Assessment (EQA) exercise for all NELs and successful completion was a prerequisite for participation.

The EuSCAPE tool has successfully described the European epidemiology of CPOs via two surveys in 2013 and 2015 [39]. These demonstrated that CPOs are expanding, with 13 of 38 countries in 2015 reporting either inter-regional spread or endemicity as compared to six countries in 2013 [40]. It provided evidence of the occurrence of carbapenemases in carbapenem non-susceptible E. coli and K. pneumoniae clinical isolates, which were the subject of the study. In 2015, only 25 (66%) of 38 participating European countries had a surveillance system in place with mandatory notification of CPO cases to health authorities [40]. Three countries had not identified a single CPO case in 2015; by 2018, all participating countries (37) had reported CPO cases following further surveillance [41].

Since then, the carbapenem- and/or colistin-resistant Enterobacteriaceae (CCRE survey) has been launched in 37 European countries to obtain genomic data to complement the phenotypic antimicrobial resistance data available from the European Antimicrobial Resistance Surveillance Network (EARS-Net) [42].

With respect to large outbreaks, a nationwide intervention was launched in Israel in 2006 in response to a clonal outbreak of carbapenem-resistant K. pneumoniae, mediated by KPC, that affected 1,275 patients in 27 hospitals [43]. This involved issuing guidelines mandating physical separation of hospitalized carriers and appointing a professional task force that paid site visits at acute-care hospitals. Following this, the continuous increase in the incidence of CPO acquisition was halted, and a direct correlation between compliance with isolation guidelines and success in containment of transmission was demonstrated (P=0.02) [43]. The importance of strategic planning and national oversight was again demonstrated when the National Center for Infection Control (NCIC) in Israel initiated a coordinated, comprehensive intervention to control carbapenem-resistance organisms on the back of the national CPO outbreak, in long-term care facilities (LTCFs) encompassing approximately 25,000 beds in over 300 institutions [44]. These measures included implementation of population-tailored contact precautions and early detection of colonised patients. A real-time repository of all patients colonised or infected with CPO was created and there was supervised information exchange between healthcare facilities during local outbreaks [44]. This intense nationwide effort resulted in a decline in the incidence of CPOs (both infections and colonisations) in all facility types, to approximately 50% of the baseline.

The first documented outbreak of CPOs in Australia involved 10 cases which were identified in 2012 [45]. The measurement of CPO burden in Australia has been based on data from individual studies, state surveillance programmes [24], and the National Alert System for Critical Antimicrobial Resistances (CARAlert) [46]. The number of CPO episodes (screening specimens and clinical infection specimens) reported in Australia increased by 45.4% between 2022 (n=829) and 2023 (n=1,205) according to data from CARAlert. Of these isolates from 2023, 57.6% (n=694/1,205) were from clinical specimens (of which urine accounted for 51.4%; n=357/694) while the rest were from screening swabs. However, the numbers of cases of CPOs in Australia are far lower than those reported in Europe, the Middle East or North America. These relatively lower numbers of CPO cases have been linked to good IPC practices, antimicrobial stewardship (AMS) in ICUs and a limited number of medical transfers from high-risk continents where CPO is common. Suggestions for improving the control of CPO in Australia have included: 1) Establishing nationally consistent notification of CPOs; 2) Creating funding mechanisms to support laboratory detection of CPOs, pathogen genomics and public health surveillance; 3) Establishing structures to enable sharing and linkage of epidemiological, clinical, and genomic data across jurisdictions; 4) Integration of surveillance data from animals and the environment using a One-Health framework [46].

Significant progress has been made in the Republic of Ireland with CPE control since CPEs were made a notifiable infection there in 2011. With strong policy commitment and the establishment of management systems and structures, the coordinated and multi-agency approach to control has yielded positive results. The National Carbapenemase-Producing Enterobacterales Reference Laboratory (NCPERL) was established in 2012 to coordinate testing of CPE isolates across Ireland [47]. With the declaration of a national public health emergency (NPHE) in October 2017 to address CPE in acute hospitals came the launch of Ireland's first National Action Plan on Antimicrobial Resistance (iNAP). This was complimented by the establishment of a National Public Health Emergency Team (NPHET) and an expert advisory group (EAG). The remit of the NPHET was to improve resource provision, surveillance, enhance communication, increase testing and develop policies and guidance to limit the spread of CPE. The role of the EAG was to provide evidence-based guidance to the NPHET [47]. The EAG made several key findings that have implications for policy making on control: these included that damp environmental reservoirs in hospitals were inadequately controlled; that there was no current requirement to extend screening to detect CPO outside of acute hospitals and antibiotic stewardship remained vital to control efforts [47].

In Singapore, a retrospective cohort study analysed 779 patients who acquired CPEs (1,215 CPE isolates) over five years in all multi-disciplinary public hospitals [48]. It provided important insights into the dynamics of CPE introduction and dissemination in healthcare facilities. After implementation of various IPC measures consistent with US Centre for Disease Control and Prevention (CDC) [49] and WHO guidelines [6], there was an overall reduction in putative clonal CPE transmission and a continued increase in putative plasmid-mediated CPO transmission. The decline in putative clonal transmission appeared to be mainly due to a reduction in transmission events arising from direct ward contact. The independent association of indirect ward and indirect hospital contact with putative clonal CPE transmission suggests the presence of persistent reservoirs of CPE in the hospital environment [48]. The conclusion was that undetected CPE reservoirs continue to evade hospital IPC measures and new measures were needed to address plasmid-mediated CPE transmission, which accounted for 50% of CPO dissemination.

In Canada, CPO remains a growing threat with CPE rates (both infection and colonization) reportedly rising from 0.06 to 0.14 per 10,000 patient-days between 2018 and 2022. The rise in CPE cases has been attributed to both increased ascertainment (due to increased awareness) and alterations to screening practices) and true increases in colonized cases. Evidence suggests a significant proportion of CPO acquisition in Canada is occurring with no epidemiolocal link to inpatient healthcare exposures or hospital CPO outbreaks. This has fuelled a growing concern of CPO reservoirs that may exist outside of hospital settings in Canada.

In the United States (US), CPO infections have been reported in all 50 states with CRE reportedly causing about 13,100 infections in hospital patients and about 1,100 deaths in 2017. The US Centers for Disease Control and Prevention (CDC) track and report on Carbapenemase-producing Carbapenem-resistant Enterobacteriaceae (CP-CRE). These data are available through the National Notifiable Diseases Surveillance System (NNDSS). Following an outbreak of bla_KPC-positive (bla_KPC⁺) Klebsiella pneumoniae in the NIH Clinical Centre (NIHCC) in the USA in 2011, environmental sampling was conducted at several locations, including external manholes, wastewater from hospital internal pipes and housekeeping closets. All samples from the ICU pipe wastewater and external manholes contained CPOs, suggesting a vast, resilient reservoir. Although there were species and susceptibility profile differences between environmental and patient populations of CPOs, some plasmid backbones were common to both populations, highlighting an environmental reservoir of mobile elements with potential for spread.

In Belgium, active case finding, environmental sampling, WGS analysis of patient and environmental strains, and implemented control strategies were effectively used to deal with an ICU long-term CPO outbreak involving multiple species and a persistent environmental reservoir. WGS confirmed the epidemiological link between clinical and environmental strains collected from the sink drains with clonal strain dissemination and horizontal gene transfer mediated by plasmid conjugation and/or transposon jumps. Quaternary ammonium-based disinfectant and replacement of contaminated equipment failed to eradicate environmental sources. As removing sinks was not feasible, targeted cleaning protocols involving the use of enzyme-based cleaning agents to degrade biofilms and disinfectants (containing peracetic acid and hydrogen peroxide) were used and found to be effective in preventing the recolonization of the proximal sink drain by CPO.

Control of CPOs in England

Minimising the transmission of CPOs within health and social care is a priority in England [50]. However, despite a constantly evolving global evidence base, several knowledge gaps remain. These include detailed aspects of CPO transmission and understanding the most effective control measures and how best to implement them. The answers may differ between Enterobacterales, Pseudomonas spp and Acinetobacter spp which are carbapenemase-producing. There may also be differences based on other factors such as affected population and care setting.

In 2013, Public Health England (now UKHSA) published recommendations on the detection and control of CPEs within acute settings [51]. All carbapenemase-producing organisms (CPOs) were added to the list of causative agents under Schedule 2 of the Health Protection (Notifications) Regulations in 2020 in England and thus became notifiable in England from 1 October 2020 [52]. All diagnostic laboratories that identify CPOs in human samples electronically report these results via UKHSA's laboratory reporting surveillance system [7]. To complement this, UKHSA published a framework of actions, which focuses on the control of CPE [50] (although it states that some interventions may be common to carbapenemase-producing Pseudomonas spp. and Acinetobacter spp., these latter organisms are not referred to within the document given the differences in epidemiology, microbiology, transmission, and environmental persistence). The guidance document sets out a range of measures aimed at minimising the impact of CPO in the healthcare setting.

Critical measures have been utilised over time with the aim to limit the transmission of CPOs in healthcare facilities [6,50]. The CPE framework recommends targeted screening for CPE on admission, isolation and contact precautions for identified colonised or infected patients with use of dedicated nursing teams, screening of contacts and environmental cleaning/decontamination to minimise CPO reservoirs [50]. There appears to have been widespread adoption of AMS programmes to minimise inappropriate use of broad-spectrum antibiotics and robust laboratory methods for rapid and accurate detection of carbapenemases at local levels. The UKHSA has advised NHS trusts to produce and use locally developed risk assessments to accurately identify high-risk individuals for screening and to implement the most appropriate method of CPE testing [50]. These risk assessments should be based on regional prevalence, patient mix, and transmission risk from interactions and patient transfers from other healthcare provider.

In England, a total of 22,237 acquired CPO episodes were reported between October 2020 and March 2025, of which 71.3% were identified in screening samples and 1,102 (4.6%) were from sterile site specimens (Figure 1) [7]. There has been an overall rise in CPO episodes reported since the introduction of mandatory reporting in quarter 4 of 2020, with the latest overall annual rate of CPO episodes in England (Q2 2024 to Q1 2025) being 13.1 per 100,000 population [7]. The thirty-day all-cause mortality due to a CPO infection from a sterile site was 22.9% in 2023 [11].

Figure 1.

Quarterly rate of acquired CPO episodes by specimen type and quarter (England): October 2020 to March 2025. Note 2. Samples that did not fall into either ‘sterile site’ or ‘screening’ samples, for example, urine and lower genital tract specimens.

Source: UKHSA. Carbapenemase-producing Gram-negative organisms in England since October 2020: quarterly update, Q1 2025.

The Antimicrobial Resistance and Healthcare Associated Infections (AMRHAI) Reference Unit has also documented year-on-year rises in referrals of CPEs from 2020. Local diagnostic laboratory carbapenemase-enzyme detection capabilities have evolved, and many can detect the ‘big five’ carbapenemase mechanisms which account for >97% of all CPOs referred to AMRHAI [11]. Diagnostic laboratories are only expected to send CPO isolates if obtained from either an invasive site or other clinically relevant sites, or for colonised patients where the laboratory is unable to perform CPO testing locally or require confirmation of their results. Other criteria for referral include isolates negative for the ‘big five’ carbapenemases to rule out presence of rarer carbapenemase families and CPOs exhibiting unusual resistance such as ceftazidime/avibactam, meropenem/vaborbactam or imipenem/relebactam resistance where isolates are confirmed as negative for class B (NDM, VIM or IMP) carbapenemases [11].

In England, there is considerable regional variation in both the frequency and type of carbapenemases being recorded (Figure 2) with the North West and London continuing (since October 2020) to report the highest number of CPOs.

Figure 2.

Regional notifications per 100,000 population of acquired carbapenemase-producing organisms by carbapenemase family in England, 2023.

Source: UK Health Security Agency (UKHSA). English surveillance programme for antimicrobial utilisation and resistance (ESPAUR) Report 2022 to 2023 UKHSA 2023.

Despite the significant gains in detection, surveillance and guidance in England, further work is required in certain areas. These include, but are not limited to, a deeper understanding of the epidemiology and transmission of CPOs in hospitals and the community and the impact of travel including health tourism; increased use of genomic data (particularly in conjunction with epidemiological insights); a better understanding of the role of plasmids and other MGEs in the spread of carbapenemase genes; the impact of antibiotic prescribing on CPO epidemiology; estimating baseline CPO colonization from wastewater to understand if wastewater from hospitals include more CPOs than from non-hospitals; additional evidence around the cost-effectiveness of interventions, and additional engagement across public health, medical and estates disciplines and sectors.

Priorities and next steps

Acquired CPOs are now notifiable in England. UKHSA has increased engagement with key stakeholders on CPOs in England, including a CPO Forum in March 2024, organised by UKHSA and NHS England. The main objective has been to facilitate information delivery, increase awareness of the current CPO work in England, and enable increased cross-sector, multi-disciplinary working. There is an urgent need for a national, multi-disciplinary approach, with consideration of the role of One Health, and collaborative working by healthcare stakeholders to address the nation's growing burden of CPOs. The spread and impact of AMR in the environment between and among humans and animals remains poorly understood, including carbapenem resistance. There is a requirement to research and understand this dynamic and where necessary, control it, to reduce the burden of CPO [35]. It will be vital that new evidence and lessons learned are considered both from a national perspective as well as used to guide targeted responses at local trust level to ensure best practices are being continually implemented and reviewed. Whilst national CPO reports are published on a quarterly basis and monthly regional reports are disseminated to respective UKHSA regional teams, further timely and accessible data are being developed to enhance access to national data.

The issues around controlling CPO are complex; this reflects the need for greater stakeholder engagement to address the various aspects of this multifaceted problem. Only by adopting an integrated, multi-disciplinary and cross-sectoral approach to CPO surveillance and action can lasting gains be made in managing this problem in England and beyond.

CRediT authorship contribution statement

Akaninyene Otu: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Jacquelyn McCormick: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Katherine L. Henderson: Data curation, Methodology, Writing – original draft, Writing – review & editing. Alice Ledda: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Danièle Meunier: Formal analysis, Methodology, Writing – original draft, Writing – review & editing. Bharat Patel: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Colin S. Brown: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Susie Singleton: Data curation, Formal analysis, Writing – original draft. Emily L. Mason: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Jasmin Islam: Conceptualization, Data curation, Writing – original draft. Ginny Moore: Conceptualization, Data curation, Writing – original draft. Katie L. Hopkins: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Dakshika Jeyaratnam: Conceptualization, Methodology, Writing – original draft, Writing – review & editing.

Ethics

Not applicable.

Funding

None.

Conflict of interest

None.