Antimicrobial resistance surveillance systems: Are potential biases taken into account?

Bayesian estimation of the prevalence of antimicrobial resistance: a mathematical modelling study

BACKGROUND

Estimates of the prevalence of antimicrobial resistance (AMR) underpin effective antimicrobial stewardship, infection prevention and control, and optimal deployment of antimicrobial agents. Typically, the prevalence of AMR is determined from real-world antimicrobial susceptibility data that are time delimited, sparse, and often biased, potentially resulting in harmful and wasteful decision-making. Frequentist methods are resource intensive because they rely on large datasets.

OBJECTIVES

To determine whether a Bayesian approach could present a more reliable and more resource-efficient way to estimate population prevalence of AMR than traditional frequentist methods.

METHODS

Retrospectively collected, open-source, real-world pseudonymized healthcare data were used to develop a Bayesian approach for estimating the prevalence of AMR by combination with prior AMR information from a contextualized review of literature. Iterative random sampling and cross-validation were used to assess the predictive accuracy and potential resource efficiency of the Bayesian approach compared with a standard frequentist approach.

RESULTS

Bayesian estimation of AMR prevalence made fewer extreme estimation errors than a frequentist estimation approach [n = 74 (6.4%) versus n = 136 (11.8%)] and required fewer observed antimicrobial susceptibility results per pathogen on average [mean = 28.8 (SD = 22.1) versus mean = 34.4 (SD = 30.1)] to avoid any extreme estimation errors in 50 iterations of the cross-validation. The Bayesian approach was maximally effective and efficient for drug-pathogen combinations where the actual prevalence of resistance was not close to 0% or 100%.

CONCLUSIONS

Bayesian estimation of the prevalence of AMR could provide a simple, resource-efficient approach to better inform population infection management where uncertainty about AMR prevalence is high.

Antimicrobial susceptibility of mastitis pathogens isolated from North American dairy cattle, 2011-2022

A total of 10,890 bacterial isolates of Streptococcus dysgalactiae, Streptococcus uberis, Staphylococcus aureus and Escherichia coli isolated as etiological agents from dairy cows with mastitis by 29 veterinary laboratories across North America between 2011 and 2022 were tested for in vitro antimicrobial susceptibility by broth microdilution to ampicillin, cefoperazone, ceftiofur, cephalothin, erythromycin, oxacillin, penicillin-novobiocin and pirlimycin according to CLSI standards. Using available clinical breakpoints, antimicrobial resistance among S. dysgalactiae (n = 2406) was low for penicillin-novobiocin (0% resistance), ceftiofur (0.1%), erythromycin (3.2%) and pirlimycin (4.6%). Among S. uberis (n = 2398), resistance was low for ampicillin (0%) and ceftiofur (0.2%) and moderate for erythromycin (11.9%) and pirlimycin (18.4%). For S. aureus (n = 3194), resistance was low for penicillin-novobiocin (0%), ceftiofur (0.1%), oxacillin (0.2%), erythromycin (0.7%), cefoperazone (1.2%) and pirlimycin (2.8%). For E. coli (n = 2892), resistance was low for ceftiofur (2.8%) and cefoperazone (3.4%) and moderate for ampicillin (9.2%). Overall, the results indicate that mastitis pathogens in the United States and Canada have not shown any substantial changes in the in vitro susceptibility to antimicrobial drugs over the 12 years of the study, or among that of the proceeding survey from 2002-2010. The data support the conclusion that resistance to common antimicrobial drugs among mastitis pathogens, even to drugs that have been used in dairies for mastitis management for many years, continues to remain low.

Strengthening antimicrobial resistance surveillance systems: a scoping review

BACKGROUND

Antimicrobial resistance (AMR) is an emerging global public health crisis. Surveillance is a fundamental component in the monitoring and evaluation of AMR mitigation endeavours. The primary aim of the scoping review is to identify successes, barriers, and gaps in implementing AMR surveillance systems and utilising data from them.

METHODS

PubMed, Web of Science, SCOPUS, and EMBASE databases were searched systematically to identify literature pertaining to implementation, monitoring, and evaluation of AMR surveillance systems. A thematic analysis was conducted where themes within the literature were inductively grouped based on the described content.

RESULTS

The systematic search yielded 639 journal articles for screening. Following deduplication and screening, 46 articles were determined to be appropriate for inclusion. Generally, most studies focused on human AMR surveillance (n = 38, 82.6%). Regionally, there was equal focus on low- and middle-income countries (n = 7, 15.2%) and trans-national contexts (n = 7, 14.5%). All included articles (n = 46, 100.0%) discussed barriers to either implementing or utilising AMR surveillance systems. From the scoping review, 6 themes emerged: capacity for surveillance, data infrastructure, policy, representativeness, stakeholder engagement, and sustainability. Data infrastructure was most frequently discussed as problematic in evaluation of surveillance systems (n = 36, 75.0%). The most frequent success to surveillance system implementation was stakeholder engagement (n = 30, 65.2%).

CONCLUSIONS

Experiences of AMR surveillance systems are diverse across contexts. There is a distinct separation of experiences between systems with emerging surveillance systems and those with established systems. Surveillance systems require extensive refinement to become representative and meet surveillance objectives.

Population-based study of Morganella-Proteus-Providencia group bloodstream infections

Background

There is a paucity of studies investigating the population-based epidemiology of Morganella-Proteus-Providencia (MPP) group infections. Our objective was to determine the incidence, risk factors, and outcome of MPP group bloodstream infections (BSI), and explore species-specific differences.

Methods

Population-based surveillance was conducted in the western interior of British Columbia, Canada, between April 1, 2010 and March 30, 2020.

Results

Sixty-two incident MPP group BSI occurred for an annual incidence of 3.4 per 100,000 residents; rates for Morganella morganii, Proteus mirabilis, and Providencia species were 0.5, 2.6, and 0.3 per 100,000 population, respectively. The median year of age was 72.5 and was different (p = 0.03) among the groups. Most (92%) MPP group BSIs were of community-onset. Significant differences were observed in the distribution of clinical focus of infection, with most notably 81% of P. mirabilis BSI due to genitourinary focus as compared to 60% and 22% for Providencia species and M. morganii, respectively. Comorbid illnesses that increased the risk for development of MPP group BSI (incidence rate ratio; 95% CI) were HIV infection (37.0; 4.4-139.6), dementia (11.5; 6.1-20.7), cancer (6.4; 3.2-11.9), stroke 6.5 (2.8-13.3), and diabetes 2.7 (1.3-5.0). Thirteen, one, and none of the cases with P. mirabilis, M. morganii, and Providencia species BSI died within 30 days of index culture for respective all cause case-fatalities of 27%, 11%, and 0% (p = 0.1).

Conclusions

Although collectively responsible for a substantial burden of illness, the epidemiology of MPP group BSI varies significantly by species.

Sphingomonas paucimobilis bloodstream infection is a predominantly community-onset disease with significant lethality

BACKGROUND

Small case series and reports suggest that Sphingomonas paucimobilis is predominantly a cause of nosocomial bloodstream infections (BSI) with very low associated mortality. Our objective was to describe the epidemiology and outcome of Sphingomonas species BSI in a large Australian population.

METHODS

All residents of Queensland Australia with BSI due to Sphingomonas species identified within the publicly funded system during 2000-2019 were included.

RESULTS

A total of 282 incident episodes of Sphingomonas species BSI were identified for an age- and sex-adjusted incidence of 3.2 per million population annually. Incidence rates were highest in the tropical regions of the state. Most (94%) of the isolates were confirmed as Sphingomonas paucimobilis. Seventy-seven percent of the infections were of community-onset of which 48% were community-associated and 30% were healthcare-associated. The very young and old and males were at highest risk. Patients with community-associated disease were on average younger, had fewer co-morbidities, and were less likely to have polymicrobial infections. At least one comorbidity was identified in 62% with malignancy, diabetes, and lung disease most prevalent. The overall all cause 30-day case-fatality rate was 6%.

CONCLUSION

Sphingomonas paucimobilis BSI is a predominantly community-onset disease associated with a significant risk for death.

Surveillance of antimicrobial resistance in low- and middle-income countries: a scattered picture

Data on comprehensive population-based surveillance of antimicrobial resistance is lacking. In low- and middle-income countries, the challenges are high due to weak laboratory capacity, poor health systems governance, lack of health information systems, and limited resources. Developing countries struggle with political and social dilemma, and bear a high health and economic burden of communicable diseases. Available data are fragmented and lack representativeness which limits their use to advice health policy makers and orientate the efficient allocation of funding and financial resources on programs to mitigate resistance. Low-quality data means soaring rates of antimicrobial resistance and the inability to track and map the spread of resistance, detect early outbreaks, and set national health policy to tackle resistance. Here, we review the barriers and limitations of conducting effective antimicrobial resistance surveillance, and we highlight multiple incremental approaches that may offer opportunities to strengthen population-based surveillance if tailored to the context of each country.

Antimicrobial resistance among uropathogens in the Asia-Pacific region: a systematic review

BACKGROUND

Antimicrobial resistance (AMR) in urinary tract infections (UTI) is a global public health problem. However, estimates of the prevalence of AMR, required for empirical treatment guidelines, are lacking for many regions.

OBJECTIVES

To perform a systematic review and summarize the available information about AMR prevalence among urinary Escherichia coli and Klebsiella pneumoniae, the two priority uropathogens, in the Asia-Pacific region (APAC).

METHODS

PubMed, EBSCO and Web of Science databases were searched for articles (2008-20), following PRISMA guidelines. The prevalence of resistance was calculated and reported as point estimate with 95% CI for antimicrobial drugs recommended in WHO treatment guidelines. Data were stratified by country and surveillance approach (laboratory- or population-based surveillance). The quality of included articles was assessed using a modified Newcastle-Ottawa Quality Assessment Scale.

RESULTS

Out of 2400 identified articles, 24 studies, reporting on 11 (26.8%) of the 41 APAC countries, met the inclusion criteria. Prevalence of resistance against trimethoprim/sulfamethoxazole, ciprofloxacin, and ceftriaxone ranged between 33% and 90%, with highest prevalence reported from Bangladesh, India, Sri Lanka and Indonesia. Resistance against nitrofurantoin ranged between 2.7% and 31.4%. Two studies reported data on fosfomycin resistance (1.8% and 1.7%). Quality of reporting was moderate.

CONCLUSIONS

We show very high prevalence estimates of AMR against antibiotics commonly used for the empirical treatment of UTI, in the limited number of countries in the APAC for which data are available. Novel feasible and affordable approaches that facilitate population-based AMR surveillance are needed to increase knowledge on AMR prevalence across the region.

Laboratory-based versus population-based surveillance of antimicrobial resistance to inform empirical treatment for suspected urinary tract infection in Indonesia

Surveillance of antimicrobial resistance (AMR) enables monitoring of trends in AMR prevalence. WHO recommends laboratory-based surveillance to obtain actionable AMR data at local or national level. However, laboratory-based surveillance may lead to overestimation of the prevalence of AMR due to bias. The objective of this study is to assess the difference in resistance prevalence between laboratory-based and population-based surveillance (PBS) among uropathogens in Indonesia. We included all urine samples submitted to the laboratory growing Escherichia coli and Klebsiella pneumoniae in the laboratory-based surveillance. Population-based surveillance data were collected in a cross-sectional survey of AMR in E. coli and K. pneumoniae isolated from urine samples among consecutive patients with symptoms of UTI, attending outpatient clinics and hospital wards. Data were collected between 1 April 2014 until 31 May 2015. The difference in percentage resistance (95% confidence intervals) between laboratory- and population-based surveillance was calculated for relevant antibiotics. A difference larger than +/- 5 percent points was defined as a biased result, precluding laboratory-based surveillance for guiding empirical treatment. We observed high prevalence of AMR ranging between 63.1% (piperacillin-tazobactam) and 85% (ceftriaxone) in laboratory-based surveillance and 41.3% (piperacillin-tazobactam) and 74.2% (ceftriaxone) in population-based surveillance, except for amikacin and meropenem (5.7%/9.8%; 10.8%/5.9%; [laboratory-/population-based surveillance], respectively). Laboratory-based surveillance yielded significantly higher AMR prevalence estimates than population-based surveillance. This difference was much larger when comparing surveillance data from outpatients than from inpatients. All point estimates of the difference between the two surveillance systems were larger than 5 percent points, except for amikacin and meropenem. Laboratory-based AMR surveillance of uropathogens, is not adequate to guide empirical treatment for community-based settings in Indonesia.

High Burden of Extended-Spectrum β-Lactamase-Producing Escherichia coli and Klebsiella pneumoniae Bacteremia in Older Adults: A Seven-Year Study in Two Rural Thai Provinces

Bloodstream infection surveillance conducted from 2008 to 2014 in all 20 hospitals in Sa Kaeo and Nakhon Phanom provinces, Thailand, allowed us to look at disease burden, antibiotic susceptibilities, and recurrent infections caused by extended-spectrum β-lactamase (ESBL)-producing Escherichia coli and Klebsiella pneumoniae. Of 97,832 blood specimens, 3,338 were positive for E. coli and 1,086 for K. pneumoniae. The proportion of E. coli isolates producing ESBL significantly increased from 19% to 22% in 2008-2010 to approximately 30% from 2011 to 2014 (P-value for trend = 0.02), whereas ESBL production among K. pneumoniae cases was 27.4% with no significant trend over time. Incidence of community-onset ESBL-producing E. coli increased from 5.4 per 100,000 population in 2008 to 12.8 in 2014, with the highest rates among persons aged ≥ 70 years at 79 cases per 100,000 persons in 2014. From 2008 to 2014, community-onset ESBL-producing K. pneumoniae incidence was 2.7 per 100,000, with a rate of 12.9 among those aged ≥ 70 years. Although most (93.6% of E. coli and 87.6% of K. pneumoniae) infections were community-onset, hospital-onset infections were twice as likely to be ESBL. Population-based surveillance, as described, is vital to accurately monitor emergence and trends in antimicrobial resistance, and in guiding the development of rational antimicrobial therapy recommendations.

Harnessing alternative sources of antimicrobial resistance data to support surveillance in low-resource settings

One of the most pressing challenges facing the global surveillance of antimicrobial resistance (AMR) is the generation, sharing, systematic analysis and dissemination of data in low-resource settings. Numerous agencies and initiatives are working to support the development of globally distributed microbiology capacity, but the routine generation of a sustainable flow of reliable data will take time to establish before it can deliver a clinical and public health impact. By contrast, there are a large number of pharma- and academia-led initiatives that have generated a wealth of data on AMR and drug-resistant infections in low-resource settings, together with high-volume data generation by private laboratories. Here, we explore how untapped sources of data could provide a short-term solution that bridges the gap between now and the time when routine surveillance capacity will have been established and how this could continue to support surveillance efforts in the future. We discuss the benefits and limitations of data generated by these sources, the mechanisms and barriers to making this accessible and how academia and pharma might support the development of laboratory and analytical capacity. We provide key actions that will be required to harness these data, including: a mapping exercise; creating mechanisms for data sharing; use of data to support national action plans; facilitating access to and use of data by the WHO Global Antimicrobial Resistance Surveillance System; and innovation in data capture, analysis and sharing.

Azole resistance surveillance in Aspergillus fumigatus: beneficial or biased?

Azole resistance is a growing concern with Aspergillus fumigatus, and may cause increased mortality in patients with azole-resistant invasive aspergillosis (IA). Microbial surveillance has been recognized as a fundamental component of resistance management. Surveillance information may be used to inform decisions regarding health services and research funding allocation, to guide local infection control in hospitals and communities, and to direct local and national drug policies and guidelines. Azole resistance frequencies have been based on screening of unselected A. fumigatus isolates, on the number of azole-resistant cases within a cohort of patients with a specific Aspergillus disease, or on analysis of patients within a specific risk group. The various surveillance approaches differ in their aims, as well as in their associated advantages and drawbacks. Nevertheless, a wide range of azole resistance frequencies has been reported, partly due to the denominator used. As most azole resistance is believed to develop in the environment and, as a consequence, azole-naive patients may present with azole-resistant aspergillosis, experts recommended a 10% resistance frequency threshold above which the standard treatment choice, i.e. voriconazole, should be reconsidered. We believe that local resistance rates based on Aspergillus disease and/or risk group should be leading for decisions regarding empirical antifungal therapy in specific units. In addition, patient factors should be considered, such as admission to the ICU. Collecting valid surveillance data may be challenging in azole resistance due to numerous factors that present potential biases. Surveillance research may benefit from further standardization, which may be facilitated through the recently instituted International Society for Human and Animal Mycology (ISHAM) Aspergillus Resistance Surveillance Working Group.

Burden of community-onset bloodstream infections, Western Interior, British Columbia, Canada

Although community-onset bloodstream infection (BSI) is recognized as a major cause of morbidity and mortality, its epidemiology has not been well defined in non-selected populations. We conducted population-based surveillance in the Interior Health West region of British Columbia, Canada in order to determine the burden associated with community-onset BSI. A total of 1088 episodes were identified for an overall annual incidence of 117·8/100 000 of which 639 (58·7%) were healthcare-associated (HA) and 449 (41·3%) were community-associated (CA) BSIs for incidences of 69·2 and 48·6/100 000, respectively. The incidence of community-onset BSI varied by age and gender and elderly males were at the highest risk. Overall 964 (88·6%) episodes resulted in hospital admission for a median length of stay of 8 days; the total days of acute hospitalization associated with community-onset BSI was 13 530 days or 1465 days/100 000 population per year. The in-hospital mortality rate was 10·6% (102/964) and this was higher for HA-BSI (72/569, 12·7%) compared to CA-BSI (30/395, 7·6%, P = 0·014) episodes. Community-onset BSI, especially HA-BSI, is associated with a major burden of illness.

Laboratory-based nationwide surveillance of antimicrobial resistance in Ghana

Global efforts are underway to combat antimicrobial resistance (AMR). A key target in this intervention is surveillance for local and national action. Data on AMR in Ghana are limited, and monitoring of AMR is nonexistent. We sought to generate baseline data on AMR, and to assess the readiness of Ghana in laboratory-based surveillance. Biomedical scientists in laboratories across Ghana with capacity to perform bacteriological culture were selected and trained. In-house standard operating protocols were used to perform microbiological investigations on clinical specimens. Additional microbiological tests and data analyses were performed at a centralized laboratory. Surveillance data were stored and analyzed using WHONET program files. A total of 24 laboratories participated in the training, and 1,598 data sets were included in the final analysis. A majority of the bacterial species were isolated from outpatients (963 isolates; 60.3%). Urine (617 isolates; 38.6%) was the most common clinical specimen cultured, compared to blood (100 isolates; 6.3%). Ten of 18 laboratories performed blood culture. Bacteria isolated included Escherichia coli (27.5%), Pseudomonas spp. (14.0%), Staphylococcus aureus (11.5%), Streptococcus spp. (2.3%), and Salmonella enterica serovar Typhi (0.6%). Most of the isolates were multidrug-resistant, and over 80% of them were extended-spectrum beta-lactamases-producing. Minimum inhibitory concentration levels at 50% and at 90% for ciprofloxacin, ceftriaxone, and amikacin on selected multidrug-resistant bacteria species ranged between 2 µg/mL and >256 µg/mL. A range of clinical bacterial isolates were resistant to important commonly used antimicrobials in the country, necessitating an effective surveillance to continuously monitor AMR in Ghana. With local and international support, Ghana can participate in global AMR surveillance.

Susceptibility to antimicrobial agents among bovine mastitis pathogens isolated from North American dairy cattle, 2002–2010

Approximately 8,000 isolates of Streptococcus agalactiae, Streptococcus dysgalactiae, Streptococcus uberis, Staphylococcus aureus, and Escherichia coli, isolated by 25 veterinary laboratories across North America between 2002 and 2010, were tested for in vitro susceptibility to beta-lactam, macrolide, and lincosamide drugs. The minimal inhibitory concentrations (MICs) of the beta-lactam drugs remained low against most of the Gram-positive strains tested, and no substantial changes in the MIC distributions were seen over time. Of the beta-lactam antimicrobial agents tested, only ceftiofur showed good in vitro activity against E. coli. The MICs of the macrolides and lincosamides also remained low against Gram-positive mastitis pathogens. While the MIC values given by 50% of isolates (MIC50) for erythromycin and pirlimycin and the streptococci were all low (≤0.5 µg/ml), the MIC values given by 90% of isolates (MIC90) were higher and more variable, but with no apparent increase over time. Staphylococcus aureus showed little change in erythromycin susceptibility over time, but there may be a small, numerical increase in pirlimycin MIC50 and MIC90 values. Overall, the results suggest that mastitis pathogens in the United States and Canada have not shown any substantial changes in the in vitro susceptibility to beta-lactam, macrolide, and lincosamide drugs tested over the 9 years of the study.