A colleague recently asked me about where she could access international data on antimicrobial resistance prevalence. In describing the ones I’ve used to her, it was again starkly apparent that in many ways we are flying blind in terms of AMR prevalence globally, in particular outside of hospital settings and at anything but a national level.  

Yes, I could point to the amazingly large and international Antimicrobial Testing Leadership and Surveillance (ATLAS) database shared by Pfizer and one of the several datasets in the Vivli database. These brilliant resources support analyses of AMR drivers Rahbe et al 2023 or enable methodology development for empiric prescribing design (Leclerc 2020) or resistance change predictions (Catalan 2022). ATLAS itself holds more than 6M isolates with different minimal inhibitory concentrations (MICs) values or resistance calls for over 3,500 pathogen-antibiotic pairs isolated from more than 600k patients in over 70 countries since 2004.  

However, these large numbers hide a fundamental problem – they are not sampled in a systematic surveillance framework and the only spatial data given is “country”. Not only will each country have different protocols for sampling patients, each hospital will likely choose different pathogen samples to provide in response to the requests from Pfizer which are very basic, vary with time and are aimed at purely trending AMR in clinical isolates. Fundamentally, once data is removed from a collection setting, it is also often hard to know exactly the sampling framework. Moreover, like much of AMR surveillance, the data is only from hospital settings. For the community we have, as far as I am aware, no open global datasets at the individual patient or sample level of asymptomatic carriage. The large numbers also hide that for many countries (e.g. in sub-saharan Africa) we have little to no surveillance data.  

Other international efforts such as the Global Antimicrobial Resistance and Use Surveillance System (GLASS) from the WHO and the European Centre for Disease Control do have a dashboards for exploring the data (here and here). To access the underlying open data you can download it from the ECDC and WHO websites (or use resources other have made to aggregate the available open data e.g. Leclerc 2022). However, this is again only at the country-level and this now has either disaggregation by wide age bands (0-4, 5-18, 19-64, 65+) or by sex (not both together). It also contains only isolates from the most serious, often bloodstream infections. 

Another initiative is the One Health Trust ResistanceMap which produces global maps and charts with more individual country data included outside of the above international surveillance efforts but it seems that you cannot download the nationally presented data. There is also AMRCloud based in Russia which appears to hold a range of downloadable individual study data as well as tools to explore AMR data. Outside of this, we must turn to individual country efforts such as the great sub-national indicators in the Fingertips for the UK or ThaiAMRwatch for Thailand.  

 While aggregation is necessary to protect patient privacy, it is concerning that we still lack an open albeit uncertain global AMR prevalence estimate —even for WHO priority pathogens, stratified by narrow age bands, sex, or other key factors such as comorbidities, ethnicity, or previous antibiotic exposure.

Despite these challenges, I remain hopeful. The field of AMR surveillance is evolving rapidly as health data digitization and data linkage systems improve. The WHO country surveys will provide detailed cross-cutting AMR analysis, potentially providing openly available data from well-structured sampling frameworks.

To make real progress, future datasets must prioritize not just AMR prevalence but also transparency in sampling methods —using frameworks like MICRO (Turner et al., 2019) to ensure data consistency and comparability. We are making strides, but we still have a long way to go.

References

• AMRCloud: https://amrcloud.net/en/

• Catalán P, Wood E, Blair JMA, Gudelj I, Iredell JR, Beardmore RE. Seeking patterns of antibiotic resistance in ATLAS, an open, raw MIC database with patient metadata. Nat Commun. 2022 May 25;13(1):2917. doi: 10.1038/s41467-022-30635-7. PMID: 35614098; PMCID: PMC9133080.

• GLASS WHO: https://www.who.int/initiatives/glass

• Leclerc QJ, Naylor NR, Aiken AM, Coll F, Knight GM. Feasibility of informing syndrome-level empiric antibiotic recommendations using publicly available antibiotic resistance datasets. Wellcome Open Res. 2020 Jun 24;4:140. doi: 10.12688/wellcomeopenres.15477.2. PMID: 32656364; PMCID: PMC7327722.

• Leclerc, Q.J. (2022). Compiled data from WHO GLASS 2022 report (Version 2.0) [GitHub Repository]. https://doi.org/10.5281/zenodo.7486150

• OneHealthTrust. ResistanceMap: [Page Name]. 2025. [URL]. Date accessed: [Date]. 
  Example: 1. OneHealthTrust. ResistanceMap: Antibiotic resistance. 2025. https://resistancemap.onehealthtrust.org/AntibioticResistance.php. Date accessed: May 31, 2025.

• Rahbe E, Watier L, Guillemot D, Glaser P, Opatowski L. Determinants of worldwide antibiotic resistance dynamics across drug-bacterium pairs: a multivariable spatial-temporal analysis using ATLAS. Lancet Planet Health. 2023 Jul;7(7):e547-e557. doi: 10.1016/S2542-5196(23)00127-4. PMID: 37437996.

• ThaiAMRWatch: https://www.thaiamrwatch.net/

• Turner P, Fox-Lewis A, Shrestha P, Dance DAB, Wangrangsimakul T, Cusack TP, Ling CL, Hopkins J, Roberts T, Limmathurotsakul D, Cooper BS, Dunachie S, Moore CE, Dolecek C, van Doorn HR, Guerin PJ, Day NPJ, Ashley EA. Microbiology Investigation Criteria for Reporting Objectively (MICRO): a framework for the reporting and interpretation of clinical microbiology data. BMC Med. 2019 Mar 29;17(1):70. doi: 10.1186/s12916-019-1301-1. PMID: 30922309; PMCID: PMC6440102.

• Vivli global AMR surveillance platform: https://amr.vivli.org/

• World Health Organization. Methodological principles of nationally representative surveys as a platform for global surveillance of antimicrobial resistance in human bloodstream infections. Geneva: World health Organization; 2023 (https://www.who.int/publications/i/item/9789240067004, accessed 04 February 2025).