Determination of minimum inhibitory concentrations using machine-learning-assisted agar dilution.

Effective policy to address the global threat of antimicrobial resistance requires robust antimicrobial susceptibility data. Traditional methods for measuring minimum inhibitory concentration (MIC) are resource intensive, subject to human error, and require considerable infrastructure. AIgarMIC streamlines and standardizes MIC measurement and is especially valuable for large-scale surveillance activities. MICs were measured using agar dilution for n = 10 antibiotics against clinical Enterobacterales isolates (n = 1,086) obtained from a large tertiary hospital microbiology laboratory. Escherichia coli (n = 827, 76%) was the most common organism. Photographs of agar plates were divided into smaller images covering one inoculation site. A labeled data set of colony images was created and used to train a convolutional neural network to classify images based on whether a bacterial colony was present (first-step model). If growth was present, a second-step model determined whether colony morphology suggested antimicrobial growth inhibition. The ability of the AI to determine MIC was then compared with standard visual determination. The first-step model classified bacterial growth as present/absent with 94.3% accuracy. The second-step model classified colonies as "inhibited" or "good growth" with 88.6% accuracy. For the determination of MIC, the rate of essential agreement was 98.9% (644/651), with a bias of -7.8%, compared with manual annotation. AIgarMIC uses artificial intelligence to automate endpoint assessments for agar dilution and potentially increases throughput without bespoke equipment. AIgarMIC reduces laboratory barriers to generating high-quality MIC data that can be used for large-scale surveillance programs. IMPORTANCE: This research uses modern artificial intelligence and machine-learning approaches to standardize and automate the interpretation of agar dilution minimum inhibitory concentration testing. Artificial intelligence is currently of significant topical interest to researchers and clinicians. In our manuscript, we demonstrate a use-case in the microbiology laboratory and present validation data for the model's performance against manual interpretation.

Challenges for global antibiotic regimen planning and establishing antimicrobial resistance targets: implications for the WHO Essential Medicines List and AWaRe antibiotic book dosing.

SUMMARYThe World Health Organisation's 2022 AWaRe Book provides guidance for the use of 39 antibiotics to treat 35 infections in primary healthcare and hospital facilities. We review the evidence underpinning suggested dosing regimens. Few (n = 18) population pharmacokinetic studies exist for key oral AWaRe antibiotics, largely conducted in homogenous and unrepresentative populations hindering robust estimates of drug exposures. Databases of minimum inhibitory concentration distributions are limited, especially for community pathogen-antibiotic combinations. Minimum inhibitory concentration data sources are not routinely reported and lack regional diversity and community representation. Of studies defining a pharmacodynamic target for ß-lactams (n = 80), 42 (52.5%) differed from traditionally accepted 30%-50% time above minimum inhibitory concentration targets. Heterogeneity in model systems and pharmacodynamic endpoints is common, and models generally use intravenous ß-lactams. One-size-fits-all pharmacodynamic targets are used for regimen planning despite complexity in drug-pathogen-disease combinations. We present solutions to enable the development of global evidence-based antibiotic dosing guidance that provides adequate treatment in the context of the increasing prevalence of antimicrobial resistance and, moreover, minimizes the emergence of resistance.

Antimicrobial treatment imprecision: an outcome-based model to close the data-to-action loop.

Health-care systems, food supply chains, and society in general are threatened by the inexorable rise of antimicrobial resistance. This threat is driven by many factors, one of which is inappropriate antimicrobial treatment. The ability of policy makers and leaders in health care, public health, regulatory agencies, and research and development to deliver frameworks for appropriate, sustainable antimicrobial treatment is hampered by a scarcity of tangible outcome-based measures of the damage it causes. In this Personal View, a mathematically grounded, outcome-based measure of antimicrobial treatment appropriateness, called imprecision, is proposed. We outline a framework for policy makers and health-care leaders to use this metric to deliver more effective antimicrobial stewardship interventions to future patient pathways. This will be achieved using learning antimicrobial systems built on public and practitioner engagement; solid implementation science; advances in artificial intelligence; and changes to regulation, research, and development. The outcomes of this framework would be more ecologically and organisationally sustainable patterns of antimicrobial development, regulation, and prescribing. We discuss practical, ethical, and regulatory considerations involved in the delivery of novel antimicrobial drug development, and policy and patient pathways built on artificial intelligence-augmented measures of antimicrobial treatment imprecision.

Antimicrobial learning systems: an implementation blueprint for artificial intelligence to tackle antimicrobial resistance.

The proliferation of various forms of artificial intelligence (AI) brings many opportunities to improve health care. AI models can harness complex evolving data, inform and augment human actions, and learn from health outcomes such as morbidity and mortality. The global public health challenge of antimicrobial resistance (AMR) needs large-scale optimisation of antimicrobial use and wider infection care, which could be enabled by carefully constructed AI models. As AI models become increasingly useful and robust, health-care systems remain challenging places for their deployment. An implementation gap exists between the promise of AI models and their use in patient and population care. Here, we outline an adaptive implementation and maintenance framework for AI models to improve antimicrobial use and infection care as a learning system. The roles of AMR problem identification, law and regulation, organisational support, data processing, and AI development, assessment, maintenance, and scalability in the implementation of AMR-targeted AI models are considered.

Vaccine-induced, but not natural immunity, against the Streptococcal inhibitor of complement protects against invasive disease.

Highly pathogenic emm1 Streptococcus pyogenes strains secrete the multidomain Streptococcal inhibitor of complement (SIC) that binds and inactivates components of the innate immune response. We aimed to determine if naturally occurring or vaccine-induced antibodies to SIC are protective against invasive S. pyogenes infection. Immunisation with full-length SIC protected mice against systemic bacterial dissemination following intranasal or intramuscular infection with emm1 S. pyogenes. Vaccine-induced rabbit anti-SIC antibodies, but not naturally occurring human anti-SIC antibodies, enhanced bacterial clearance in an ex vivo whole-blood assay. SIC vaccination of both mice and rabbits resulted in antibody recognition of all domains of SIC, whereas naturally occurring human anti-SIC antibodies recognised the proline-rich region of SIC only. We, therefore, propose a model whereby natural infection with S. pyogenes generates non-protective antibodies against the proline-rich region of SIC, while vaccination with full-length SIC permits the development of protective antibodies against all SIC domains.