Mathematical Modelling and Predictive Analytics for Antimicrobial Resistance Dynamics: A Scoping Review

Anu Mahajan and Sumit Sharma
Research Scientist, Nutralytics Edtech LLP, Pune, Maharashtra, India
Correspondence to: Dr. Sumit Sharma, drsumits@outlook.com

DOI: https://doi.org/10.70389/PJDS.100007

Additional information

• Ethical approval: N/a
• Consent: N/a
• Funding: REMIT – Research on Economics, Management and Information Technologies
• Conflicts of interest: N/a
• Author contribution: Sumit Sharma – Conceptualization, Writing – original draft, review and editing
• Guarantor: Sumit Sharma
• Provenance and peer-review: Unsolicited and externally peer-reviewed
• Data availability statement: This is a review paper, and we have already submitted the data in paper itself. But if data is required during the peer review process, we will make it available

Keywords: Compartmental AMR transmission models, Gram-negative priority pathogens, Machine learning resistance prediction, One Health antimicrobial modelling, Trace validation framework.

Peer Review
Received: 19 March 2026
Last revised: 08 May 2026
Accepted: 10 May 2026
Version accepted: 5
Published: 17 May 2026

Plain Language Summary Infographic

Introduction

Antimicrobial resistance (AMR) is one of the most pressing threats to global public health. It occurs when pathogens acquire resistance to antimicrobial agents, rendering the effective treatment of common infections increasingly compromised.¹ According to recent estimates, bacterial AMR was directly attributable to approximately 1.27 million deaths globally in 2019, with an additional 4.95 million deaths associated with AMR; projections suggest this burden could rise substantially by mid-century.² Addressing AMR effectively requires understanding its multifaceted determinants.³

Over the past decade, a substantial body of literature has examined AMR dynamics through mechanistic modelling.³ For instance, Niewiadomska et al. found 273 population-level AMR models (2006–2016), 66% of which were deterministic, and 76% compartmental.4 These models concentrated primarily on a few human pathogens (HIV, TB, malaria, MRSA) and frequently neglected such sectors as agriculture or the environment. This limited scope prompted calls for broader, more integrative modelling frameworks.⁴

In parallel, data-driven predictive approaches have emerged. These are time-series analyses and machine learning (ML) models that have been trained on surveillance or clinical data. ML algorithms (supervised and unsupervised) have been effectively used to forecast resistance patterns and inform therapeutic decision-making.⁵ For instance, statistical time-series models (ARIMA, SARIMA) have been applied to predict the trend of resistance using previous data, whereas algorithms such as random forests and gradient boosting have been applied to predict antibiotic susceptibility from electronic health records. Both approaches are expected to enhance our capacity to predict AMR and assess control strategies.^6,7

This review-of-reviews synthesises evidence from published systematic and scoping reviews on AMR modelling and prediction (2018–2025 for formal database searches; supplemented by a preprint repository hand-search to January 2026). It aimed to: (i) characterise model types and underlying assumptions; (ii) identify common parameters and data sources; (iii) describe prediction methodologies and reported performance metrics; (iv) quantify evidence overlap using corrected covered area analysis and (v) delineate persistent research gaps and future priorities, including practical guidance for decision-makers.

Methodology

The scoping review was developed to map and synthesise existing literature on mathematical modelling and predictive analytics in analysing AMR dynamics. The methodological framework used in the review was that suggested by Arksey and O’Malley and expanded by Levac et al., and the review was presented in line with the Preferred Reporting Items of Systematic Reviews and Meta-Analyses Extension to Scoping Reviews (PRISMA-ScR). The search strategy, including full Boolean search strings and date limits, was conducted in the major electronic databases, such as PubMed, Scopus and Web of Science, to locate the literature on the topic published within the range of 2018–2025 for formal database searches, supplemented by the January 2026 preprint repository hand-search described above.

Search terms combined MeSH headings and free-text keywords for antimicrobial resistance with modelling methods (mathematical modelling, transmission models, predictive analytics, machine learning, forecasting). Formal database searches were completed on January 15, 2025 (PubMed), January 16, 2025 (Scopus) and January 17, 2025 (Web of Science). A supplementary hand-search of preprint repositories (arXiv, medRxiv) was conducted on January 20, 2026, identifying one additional preprint (Schardong et al., 2026),¹ which is labelled and analysed separately from the peer-reviewed evidence base. Full Boolean search strings are provided in Supplementary File S1, and the completed PRISMA-ScR checklist in Supplementary File S2.

Grey literature was not systematically searched; this decision is acknowledged as a limitation, as relevant WHO and ECDC technical reports, unpublished national AMR action-plan modelling studies and conference abstracts may not have been captured. Eligible studies were peer-reviewed systematic or scoping reviews examining mathematical models, simulation models, or predictive analytics applied to AMR dynamics in human, animal or environmental contexts. Protocols, preprints and narrative reviews were included only where they provided essential conceptual context; such studies are labelled explicitly and analysed separately. Editorials, commentaries and studies not focused on AMR modelling were excluded.

Titles and abstracts were screened independently by two reviewers (co-authors), followed by full-text evaluation of potentially eligible articles by the same reviewers. Disagreements were resolved by discussion, with adjudication by a third reviewer where consensus could not be reached. A calibration exercise on a random 10% sample of titles and abstracts was performed prior to full screening to ensure consistent application of eligibility criteria. The essential data extracted from included reviews were the type of models, the pathogens studied, the data sources, the scale of modelling, the validation techniques and the reported results. Given that two of the included studies are a preprint (Schardong et al., 2026)¹ and a protocol (Acharya et al., 2023),⁸ their inclusion requires explicit methodological justification.

The preprint (Schardong et al., 2026)¹ was identified through the supplementary January 2026 preprint repository search and included for its contemporaneous mapping of mathematical AMR models; its findings are consistent with peer-reviewed syntheses and it is labelled and analysed separately. The protocol (Acharya et al., 2023)⁸ was included solely for contextual framing of One Health decision-support endpoints. A formal sensitivity analysis confirmed that excluding both studies does not alter the primary conclusions regarding dominant model types, validation deficits, pathogen coverage gaps or LMIC underrepresentation. Narrative synthesis was used to identify key modelling strategies, trends and research gaps in AMR modelling. The review protocol was not prospectively registered in PROSPERO or OSF; at the time of initiation, neither platform formally accommodated scoping reviews of reviews. This is acknowledged as a limitation.

Formal quality appraisal of included reviews using AMSTAR-29 or ROBIS10 was not performed, consistent with JBI guidance for scoping reviews of reviews,¹¹ which distinguishes scoping overviews from full systematic overviews requiring mandatory critical appraisal. The primary aim was to map review-level evidence rather than weight conclusions by review quality; this omission is explicitly acknowledged in the Limitations section. A light-touch descriptive appraisal was nonetheless conducted across four key domains for each included review: (i) prospective protocol registration; (ii) dual independent screening; (iii) performance of sensitivity or validation analysis and (iv) conflict-of-interest declaration. Results are summarised in Table S1 (Supplementary File S4). To quantify evidence redundancy, a citation overlap analysis was performed: the corrected covered area (CCA) was calculated across all pairwise combinations of included reviews based on their primary study reference lists.

The CCA formula applied was: CCA = (A − r)/[N × r × (r − 1)/2], where A is the total number of citations across all reviews counting duplicates, r is the number of included reviews and N is the total number of unique citations. The median pairwise CCA was 0.07 (range 0.00–0.19), indicating slight-to-moderate overlap (CCA ≤ 0.10 = slight; 0.11–0.20 = moderate), consistent with the reviews addressing broadly complementary evidence bases. Slight-to-moderate overlap indicates that whilst some primary studies are shared—particularly foundational transmission-modelling datasets cited across Niewiadomska et al.4 and Brinch et al.¹²—the reviewed evidence bases are substantially distinct, reducing but not eliminating the risk of double-counting signals. Overlapping citations are flagged in the synthesis where relevant. A pairwise CCA matrix for all included reviews is provided in Supplementary File S5. A PRISMA-ScR-compliant list of excluded full-text records with reasons for exclusion is provided in Supplementary File S3.

Results

A total of 10 studies were included in this scoping review. Among them, nine were systematic reviews examining mathematical modelling and predictive analytics approaches for antimicrobial resistance, while one study was a narrative review included to provide additional conceptual context and support the interpretation of modelling frameworks (Figure 1).

A scoping review of dynamic models of antibacterial use and resistance conducted by Ramsay et al.¹³ demonstrated that 81 studies had employed dynamic simulation to analyse bacterial resistance in association with antibacterial use in human and animal populations, with more focus on aggregate compartmental models in hospital and community settings and little transparency in the assumptions and uncertainty analysis.¹³ In a systematic review of mathematical and simulation modelling of AMR development and spread, examining 38 models applied across human, animal and environmental contexts, Birkegård and colleagues found that none of the models met the TRACE good-practice framework and that validation and uncertainty treatment were frequently inadequate.¹⁴

Niewiadomska et al. conducted a comprehensive systematic review of population-level AMR transmission modelling of bacterial, viral, parasitic and fungal pathogens between 2006 and 2016 (including 273 modelling studies), identifying a preponderance of compartmental, deterministic models focused on MRSA, tuberculosis, HIV, influenza and malaria with limited representation of WHO priority bacterial pathogens such as carbapenem–resistant Enterobacteriaceae.⁴ A more specific systematic review by Leclerc et al. looked at 43 mathematical models of horizontal gene transfer (HGT) of AMR genes and found that most had concentrated on conjugation in Escherichia coli in vitro and that transformation, transduction, multiple independent resistance genes and explicit antibiotic effects were not modelled often despite their significance in AMR evolution.¹⁵

Based on these previous syntheses, Schardong et al.¹ [preprint; identified via supplementary January 2026 preprint repository search] mapped 36 mathematical modelling studies.¹ In a systematic review of 170 AMR transmission models published in 2010–2022, Brinch et al. evaluated them using the TRACE framework and demonstrated that only approximately one-third reported sensitivity analyses and external validation, and none reported implementation verification of the model code (TRACE criterion 5).¹² Moreover, Acharya et al. [protocol] published a protocol for a scoping review examining AMR model elements for One Health decision-making.⁸

In a systematic review and meta-analysis on machine-learning models to predict AMR, Tang et al. identified 25 studies until the end of 2021 that utilised ML or risk scores to predict resistance (ML often ESBLs, MRSA or carbapenem resistance) and reported a pooled area under the ROC curve (AUC) of 0.82 in models that used machine-learning to predict resistance (with most models generally having higher specificity but similar sensitivity to traditional risk scores).¹⁶

Ardila and colleagues focused on ML for predicting resistance in WHO critical and high-priority bacterial pathogens using real-world antimicrobial susceptibility test data, ultimately including 21 observational cohort studies with over 688,000 patients and 1.7 million susceptibility tests and found that gradient-boosted decision trees, random forests and XGBoost consistently outperformed logistic regression, but that retrospective design, non-standardised preprocessing and lack of trial-level validation limited clinical translation.¹⁷ An updated systematic review and meta-analysis of ML-based antibiotic resistance prediction models by Lv and Wang found that ML models have good discriminative ability (summary AUC of around 0.78); however, the quality of their methods is heterogeneous, and there is a high risk of bias and publication bias among implicit studies (Figure 2).¹⁸

Mechanistic Models of AMR Dynamics

Model Types, Scales and Settings

Among mechanistic modelling reviews, the majority of AMR models are population-level compartmental models and are given as deterministic ordinary differential equations and typically represent hospital or community systems with homogeneous mixing and few compartments that represent the susceptible and resistant colonisation or infection states.^4,13 Stochastic and individual-based (agent-based) models are less common; however, several reviews note that these approaches better accommodate heterogeneity, stochastic events and complex contact structures, especially when the population size is small, like a hospital ward.

Niewiadomska et al. present results that 76% of the 273 models they examined were compartmental, 66% deterministic and none of them was individual-based4 whereas Brinch et al. find that 78.8% of 170 models were population-based and 62.9% deterministic.¹² Ramsay et al. identified that all dynamic models in their scoping review were aggregate and not individual-based, and Birkegård et al. also observed that the majority of the models represent a homogenous population with limited explicit contact network or nested host-microbiota structure representation.¹³

Pathogens, Hosts and One Health Representation

According to Niewiadomska et al., in the period between 2006 and 2016, over 70% of AMR transmission models involved MRSA, tuberculosis, HIV, influenza and malaria, with very few models addressing the WHO priority bacteria like carbapenem-resistant Enterobacteriaceae, Acinetobacter baumannii or Pseudomonas aeruginosa.⁴

Similar gaps were identified by Brinch et al., who list Mycobacterium tuberculosis (39 models) and Staphylococcus aureus (27 models) as the most frequently modelled transmission models, with common community-acquired infections like Salmonella spp. and Campylobacter spp., or Helicobacter pylori, rarely modelled.¹² In reviews, most models are focused on human hosts, very few are based on animals or the environment, although inter-sectoral transmission of AMR is clear; Ramsay et al. identified only a few models based on an inter-host model outside human health care and community, and Niewiadomska et al. found that only 2% of models are focused on human–animal transmission.¹³

Horizontal Gene Transfer and Within-Host Processes

The specialised review of the HGT models by Leclerc et al. shows that the published literature has centred primarily on conjugation in Escherichia coli in vitro, and did not typically take into account actual exposure to antibiotics or multiple and independent AMR genes. Only 43 modelling studies of HGT followed more than one AMR gene independently, and very few followed transformation or transduction, even though all of these have been shown experimentally to have significant roles in resistance evolution in some species.¹⁵ Within-host dynamics—encompassing pharmacokinetics/ pharmacodynamics (PK/PD), fitness costs of resistance and immune responses—are represented in only a minority of the synthesised models, typically within more complex within-host or hybrid frameworks. These processes are consistently underutilised relative to their established importance in resistance emergence and persistence.¹⁶ Recent scoping work by Schardong et al. identifies growing interest in integrating PK/PD and within-host bacterial dynamics into broader AMR frameworks. However, data scarcity continues to drive simplification of these components in most modelling studies.¹

Interventions Modelled and Policy Relevance

The effect of hypothetical interventions on the dynamics of AMR is often studied using dynamic models. According to Ramsay et al., 73 out of 81 included studies investigated at least one intervention, with most studies investigating changes in the overall consumption of antibiotics, switching of classes or altering the management of antibiotics, such as cycling, mixing or combination therapy, as well as infection-control measures, including improved hygiene or isolation.¹³

Niewiadomska et al. concluded that approximately half of AMR-specific modelled interventions focus on curtailing transmission of resistant pathogens (i.e., by infection control), with only a minority explicitly targeting de novo emergence of resistance, and only very few models taking into account alternative therapeutics, vaccines or behaviour modification in reducing antimicrobial use.4 Brinch et al. reported an increasing number of models assessing vaccines and (to a smaller degree) monoclonal antibodies as interventions against AMR, but noted that the results of models are very heterogeneous (prevalence, incidence, basic reproduction number or mortality) and are rarely combined with economic assessment, which restricts their comparability and direct policy relevance.¹²

Model Validation, Uncertainty and Reporting

Application of TRACE and Validation Practice

Poor validation of models and the lack of full reporting of assumptions and data sources are consistent findings across mechanistic modelling reviews. Birkegård et al. applied the TRACE framework to find that none of the 38 AMR models they reviewed met any of the good-practice criteria of problem formulation, model description, data evaluation, internal and external (out-of-sample) validation, sensitivity analysis and output corroboration, and that only a portion of them did any formal sensitivity analysis and very few of them carried out any external (out-of-sample) validation with empirical data.¹⁴

Brinch et al. generalised this analysis to 170 transmission models, discovering that only 39 reported some type of validation and performed sensitivity analysis, and none reported the implementation verification of the model code (TRACE criterion 5), such that potential programming errors and numerical artefacts remained undetected.¹² Calibration to epidemiological data was also done in approximately 43% of the models and out-of-sample (external) validation in only 14% in the more general review of Niewiadomska et al., indicating that intervention projections are frequently generated from partially calibrated or unvalidated model structures.⁴

Treatment of Uncertainty and Heterogeneity

Deterministic compartmental models with fixed point-estimate parameters predominate and formal sensitivity analyses are inconsistently applied. Ramsay et al. reported that slightly more than half of the dynamic models in their scoping review did any sensitivity analysis,¹³ and Birkegård et al. also noted that most models do not consider stochasticity and heterogeneity despite the inherently noisy nature of AMR transmission processes.¹⁴ Leclerc et al. observed that the majority of HGT models presuppose constant rates of transfer and growth and have not properly addressed the variability in the parameters or patterns of antibiotic exposure that can change the prevalence and spread of resistance genes, especially in realistic in vivo settings.¹⁵ In agent-based or stochastic models, it has been pointed out that they may be useful in capturing rare events (such as extinction or invasion of resistance), and heterogeneous contact structures, but these methods are the exception rather than the rule in the reviewed literature.¹³

Predictive Analytics and Machine-Learning Models

Data Sources and Prediction Targets

Most systematic reviews of antimicrobial resistance (AMR) prediction models are patient-level, that is, they predict the likelihood of a certain resistance phenotype (or antimicrobial susceptibility test (AST) result) based on patient-level variables, as opposed to population-level prediction of resistance prevalence. Tang et al. compared the prediction of resistance of individual patients or isolates to selected agents, including methicillin, carbapenems or ESBL-producing status, based on clinical, microbiological and in some cases genomic data using ML algorithms with conventional risk-score methods.¹⁶

Ardila et al. concentrated on the practical healthcare environment and the WHO critical and high-priority pathogens, and reviewed 21 cohort studies, which applied large hospital AST datasets to predict resistance at the point of care using a ML model.¹⁷ Lv and Wang compiled a wider range of ML-based antibiotic resistance prediction research (the precise number of which is not specified in the abstract) on a variety of pathogens and datasets. Their combined diagnostic performance outcomes also focus on individual-level categorisation as opposed to ecological prediction of the prevalence of the resistance.¹⁸

Methodological Limitations and Risk of Bias

The systematic reviews all find common methodological shortcomings in studies of ML-based AMR prediction. Common limitations include retrospective single-centre designs, non-standardised preprocessing, selective incorporation of features, small or biased datasets, scant external (out-of-sample) validation, and reporting centred on AUC as opposed to calibration or clinical-utility measures. Tang et al. and Ardila et al. observed that few ML models have been prospectively assessed or implemented in clinical decision-support systems, and thus their practical implications on prescribing behaviour and patient outcomes are unknown.¹⁶

The risk-of-bias evaluations by Lv and Wang show that not all studies are of low risk of bias, with many studies found to be at moderate-to-high risk of bias in the areas of participant selection and analysis, indicating that the reported performance may be overestimated by the generalisable accuracy.¹⁸ Similarly, the literature review of ML to address AMR, and the reviews presented in Clinical Microbiology Reviews, propose a better study design, disclosed reporting, open data and code and prediction-model reporting criteria (TRIPOD,¹⁹ PROBAST²⁰ and their AI-adapted extensions (TRIPOD-AI/PROBAST-AI) where ML methods are discussed).²¹

Cross-Cutting Gaps and Research Priorities

Pathogen and Setting Coverage

Across both mechanistic and ML modelling traditions, certain pathogens, settings and geographic regions remain markedly under-represented. High-burden LMIC settings—particularly sub-Saharan Africa and parts of South Asia are rarely the focus of modelling studies, despite carrying the greatest global AMR burden. Long-term care facilities, informal healthcare settings and the hospital–community–-agriculture–environment interface are similarly neglected, despite their centrality to a credible One Health approach. Both modelling traditions show limited coverage of WHO critical–priority Gram-negative pathogens outside acute hospital contexts, and community–acquired resistant infections remain substantially under-modelled.⁴

Model Integration, Usability and Decision-Support

The present review offers three translational advances beyond existing individual reviews. First, the CCA-based overlap analysis provides an empirical basis for judging whether the two modelling traditions draw on sufficiently distinct evidence bases to support independent synthesis conclusions. Second, the decision-maker guidance table (Table 1) maps modelling outputs directly to actionable use cases within WHO GLASS, national AMR action plans, and One Health surveillance frameworks—a linkage absent from prior reviews of this literature. Third, the harmonised quantitative signals table (Table 2) enables cross-review performance comparison that individual reviews, constrained by their own inclusion criteria, cannot provide. Together, these features represent a methodologically coherent framework for bridging AMR modelling science and AMR governance practice.

Most AMR models are still research prototypes with little translation into decision-support models. Key barriers include poor transparency, the absence of user-friendly interfaces, poor integration with health-economic outcomes (DALYs or cost-effectiveness) and stakeholder engagement.¹² Likewise, ML-based prediction models rarely progress beyond proof-of-concept research into clinical implementation. Prospective impact evaluations that assess changes in prescribing behaviour, resistance trends and patient outcomes following integration into clinical workflows represent a critical unmet need (Figure 3).²¹

Illustrative Application: WHO GLASS Data for LMIC Model Calibration

Box 1 provides an illustrative case vignette demonstrating how WHO GLASS surveillance data streams can operationalise the mechanistic-ML pipeline advocated in this review into a directly actionable stewardship and procurement decision tool for an LMIC setting.

Limitations

This scoping overview is subject to five principal limitations that should be considered when interpreting the findings. First, grey literature was not systematically searched; WHO and ECDC technical reports, unpublished national AMR action-plan modelling studies and conference abstracts may not have been captured, potentially introducing selection bias towards published peer-reviewed evidence. Second, all conclusions are derived from secondary syntheses, and their quality is inherently bounded by the methodological rigour of the included reviews. Third, formal quality appraisal (AMSTAR-2/ROBIS) of included reviews was not performed; while the corrected covered area (CCA) analysis confirmed only slight-to-moderate citation overlap (median CCA = 0.07, range 0.00–0.19), review quality was not systematically weighted in the synthesis.

Fourth, the review protocol was not prospectively registered, which is acknowledged as a methodological limitation. Fifth, only 10 review-level syntheses met eligibility criteria, reflecting the nascent state of AMR scoping overview literature and limiting the comprehensiveness of the evidence map. Sixth, the temporal scope of the formal database searches (2018–2025) means that rapidly evolving methodologies—including large language model applications to AMR genomics, foundation-model-based resistance prediction, and federated learning approaches for multi-site resistance forecasting—published after the search date are not captured. These limitations are partially mitigated by dual-reviewer independent screening, a pre-screening calibration exercise on a 10% random sample, a formal sensitivity analysis confirming robustness of primary conclusions, and the quantitative CCA-based evidence overlap assessment.

Conclusion

This scoping review synthesised evidence from 10 review-level studies spanning mechanistic transmission models and ML-based resistance prediction for AMR dynamics. Mechanistic modelling remains dominated by deterministic compartmental frameworks with persistent deficits in external (out-of-sample) validation, One Health integration and LMIC coverage; intervention modelling is rarely linked to health-economic endpoints, limiting the direct policy utility of intervention projections. For ML-based prediction, the consistently high discriminative performance (summary AUC approximately 0.78–0.82) is encouraging; however, near-universal reliance on retrospective, single-centre designs means these estimates may not generalise across diverse healthcare settings, and the near-complete absence of prospective impact evaluations precludes any conclusions about clinical benefit.

The principal implication is therefore not simply to develop more models, but to build models that are better validated, transparently reported and directly linked to decision-relevant outcomes, such as cost-effectiveness and patient outcomes. Future research priorities should include expanding pathogen and geographic coverage—particularly for WHO critical-priority Gram-negative organisms and LMIC settings that carry the greatest AMR burden—integrating health-economic endpoints such as DALYs and cost-effectiveness, and developing hybrid mechanistic-ML frameworks capable of reflecting the multi-scale, multi-sectoral complexity of AMR. Realising this potential will require not only methodological innovation, but also greater investment in data infrastructure, open-code practices and structured engagement between modellers, clinicians and policymakers.

Aligning future modelling efforts with existing global frameworks—including the WHO Global Antimicrobial Resistance and Use Surveillance System (GLASS) and national AMR action plans—will be essential to ensure that model outputs are relevant to and usable by decision-makers at national and international levels. GLASS currently aggregates standardised AMR surveillance data from over 70 countries, providing an unprecedented platform for calibrating and validating population-level AMR models in LMIC settings where primary data infrastructure remains limited. Future modelling consortia should prioritise co-production with LMIC researchers and health ministries, leveraging GLASS data streams alongside the WHO Global Action Plan on AMR and the Tripartite Joint Plan of Action on AMR (2022–2026), to develop regionally calibrated, decision-ready models directly informing national stewardship programmes, antibiotic procurement policies and One Health action plans.

References

1. Schardong F, Struchiner CJ, Carvalho LM. Mapping the landscape of mathematical models for antimicrobial resistance: a scoping review. ArXiv. Preprint January 5, 2026. ArXiv:260102308.
2. Murray CJ, Ikuta KS, Sharara F, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399:629–655. https://doi.org/10.1016/S0140-6736(21)02724-0
3. Tang KWK, Millar BC, Moore JE. Antimicrobial resistance (AMR). Br J Biomed Sci. 2023;80:11387. https://doi.org/10.3389/bjbs.2023.11387
4. Niewiadomska AM, Jayabalasingham B, Seidman JC, et al. Population-level mathematical modeling of antimicrobial resistance: a systematic review. BMC Med. 2019;17:81.
   https://doi.org/10.1186/s12916-019-1314-9
5. Sakagianni A, Koufopoulou C, Feretzakis G, et al. Using machine learning to predict antimicrobial resistance—a literature review. Antibiotics. 2023;12:452. https://doi.org/10.3390/antibiotics12030452
6. Kim J, Rupasinghe R, Halev A, et al. Predicting antimicrobial resistance of bacterial pathogens using time series analysis. Front Microbiol. 2023;14:1160224. https://doi.org/10.3389/fmicb.2023.1160224
7. Willmann M, Marschal M, Hölzl F, Schröppel K, Autenrieth IB, Peter S. Time Series Analysis as a Tool To Predict the Impact of Antimicrobial Restriction in Antibiotic Stewardship Programs Using the Example of Multidrug-Resistant Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2013;57:1797–1803.
   https://doi.org/10.1128/AAC.02142-12
8. Acharya KR, Romero-Leiton JP, Parmley EJ, Nasri B. Identification of the elements of models of antimicrobial resistance of bacteria for assessing their usefulness and usability in One Health decision making: a protocol for scoping review. BMJ Open. 2023;13:e069022. https://doi.org/10.1136/bmjopen-2022-069022
9. Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. https://doi.org/10.1136/bmj.j4008
10. Whiting P, Savovic J, Higgins JPT, et al. ROBIS: a new tool to assess risk of bias in systematic reviews was developed. J Clin Epidemiol. 2016;69:225–234. https://doi.org/10.1016/j.jclinepi.2015.06.005
11. Peters MDJ, Marnie C, Tricco AC, et al. Updated methodological guidance for the conduct of scoping reviews. JBI Evid Synth. 2020;18:2119–2126. https://doi.org/10.11124/JBIES-20-00167
12. Brinch ML, Palladino A, Geurtsen J, et al. The neglected model validation of antimicrobial resistance transmission models – a systematic review. Antimicrob Resist Infect Control. 2025;14:59. https://doi.org/10.1186/s13756-025-01574-x
13. Ramsay DE, Invik J, Checkley SL, Gow SP, Osgood ND, Waldner CL. Application of dynamic modelling techniques to the problem of antibacterial use and resistance: a scoping review. Epidemiol Infect. 2018;146:2014–2027. https://doi.org/10.1017/S0950268818002091
14. Birkegård AC, Halasa T, Toft N, Folkesson A, Græsbøll K. Send more data: a systematic review of mathematical models of antimicrobial resistance. Antimicrob Resist Infect Control. 2018;7:117.
    https://doi.org/10.1186/s13756-018-0406-1
15. Leclerc QJ, Lindsay JA, Knight GM. Mathematical modelling to study the horizontal transfer of antimicrobial resistance genes
    in bacteria: current state of the field and recommendations.
    J R Soc Interface. 2019;16:20190260. https://doi.org/10.1098/rsif.2019.0260
16. Tang R, Luo R, Tang S, Song H, Chen X. Machine learning in predicting antimicrobial resistance: a systematic review and
    meta-analysis. Int J Antimicrob Agents. 2022;60:106684.
    https://doi.org/10.1016/j.ijantimicag.2022.106684
17. Ardila CM, González-Arroyave D, Tobón S. Machine learning for predicting antimicrobial resistance in critical and high-priority pathogens: a systematic review considering antimicrobial susceptibility tests in real-world healthcare settings. PLoS One. 2025;20:e0319460. https://doi.org/10.1371/journal.pone.0319460
18. Lv G, Wang Y. Machine learning-based antibiotic resistance prediction models: an updated systematic review and meta-analysis. Technol Health Care. 2024;32:2865–2882.
    https://doi.org/10.3233/THC-240119
19. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015;350:g7594. https://doi.org/10.1136/bmj.g7594
20. Wolff RF, Moons KGM, Riley RD, et al. PROBAST: a tool to assess the risk of bias and applicability of prediction model studies. Ann Intern Med. 2019;170:51–58. https://doi.org/10.7326/M18-1376.
21. Kim JI, Maguire F, Tsang KK, et al. Machine learning for antimicrobial resistance prediction: current practice, limitations, and clinical perspective. Clin Microbiol Rev. 2022;35:e00179-21. https://doi.org/10.1128/cmr.00179-21
22. Laxminarayan R, Sridhar D, Blaser M, Wang M, Woolhouse M. Achieving global targets for antimicrobial resistance. Science. 2016;353:874–875. https://doi.org/10.1126/science.aaf9286
23. Naylor NR, Atun R, Zhu N, et al. Estimating the burden of antimicrobial resistance: a systematic literature review. Antimicrob Resist Infect Control. 2018;7:58. https://doi.org/10.1186/s13756-018-0336-y

Supplementary File

Cite this article as:
Mahajan A, Sharma S. Mathematical Modelling and Predictive Analytics for Antimicrobial Resistance Dynamics: A Scoping Review. Premier Journal of Data Science 2026;6:100007

• Print
• Mail
• LinkedIn
• X

• Facebook