A rose-thorn scratch. A caesarean. A round of chemo. A new hip. None of these should be life-threatening, and for about eighty years they mostly haven’t been, because if an infection took hold, antibiotics cleared it. That quiet guarantee sits underneath almost all of modern medicine. We barely notice it, the way you don’t notice a floor until it gives.
It is starting to give. In 2019, drug-resistant bacterial infections killed an estimated 1.27 million people directly and had a hand in nearly 5 million deaths, more than HIV/AIDS and malaria combined (Murray et al., 2022). The bacteria are learning to shrug off the drugs faster than we are inventing new ones, and the thing feeding that faster than anything else is blunt: how much antibiotic we use. Not only in hospitals. On farms, too.
This is a story about that arithmetic. We are spending these drugs faster than we can replace them. The data shows where, it shows that humans aren’t the only ones doing the spending, and it shows the line still moving the wrong way. That last point is the one to sit with.
Average human antibiotic consumption, 2016 to 2022. Most heavy users sit far above the level seen in the world’s most careful prescribers.
Source: WHO GLASS (2024), via Our World in Data. DDD = Defined Daily Doses.
Iran sits at the top, getting through roughly four times the daily doses of the world’s most careful prescribers. Tanzania and Nepal aren’t far behind. But the eye-catching thing isn’t any one country. It’s the spread. The heavy users turn up in the Middle East, sub-Saharan Africa, South Asia and southern Europe all at once. There’s no tidy rich-versus-poor line to draw here.
The dashed line marks the Nordic average: Denmark, Sweden, Norway, Finland and Iceland, the countries that have spent decades treating antibiotics as something to ration rather than reach for by default. Almost everyone on this chart sits well to the right of it. That gap is the whole point. Prudent use is clearly possible, five countries are doing it, and most of the world isn’t. Which raises the obvious next question: is it only people swallowing these pills?
Antibiotic use in livestock, 2020. Hover any country for its exact intensity. The problem is global, and it lives on farms as much as in clinics.
Source: Mulchandani et al. (2023), processed by Our World in Data. Figures are modelled estimates; values for countries without reported data carry wider uncertainty. Colour capped at the 95th percentile; hover for exact values.
It isn’t. Most of the antibiotics produced each year never go near a person. By some estimates around 70% are given to farm animals, often to healthy ones, to head off disease in crowded sheds or to push a little extra growth (Mulchandani et al., 2023). The map shows where that use clusters: the big intensive-meat regions of East Asia, the Americas and parts of Europe.
This matters to human health because bacteria don’t respect the fence line. Resistant microbes, and the genes that hand resistance from one bug to the next, move between animals and people through meat, water, soil and the hands of the people doing the farming. A resistance gene that first shows up in a pig barn can end up in a hospital ward. So if you want to understand the pressure driving resistance, you have to count both sides of the ledger. Which sets up a question the next chart can actually test: do the two problems rise and fall together?
Human vs livestock antibiotic use, by country. If the two moved together, the dots would line up. They barely do.
Source: Our World in Data (2024). Correlation r = 0.26. Dashed line is a linear fit.
Here’s the surprise. If human and animal overuse were really one problem, the dots would climb together along the dashed line. They don’t. The correlation is weak. A country can be careful in its clinics and reckless on its farms, or the other way round.
That sounds like a technicality. It isn’t. It means there’s no single lever to pull. A health minister who clamps down on over-prescribing to GPs has barely touched the farm side; an agriculture department that cleans up livestock use hasn’t fixed the hospitals. The two problems were largely built separately, and they’ll have to be taken apart separately. Most public talk about antibiotic resistance pictures one tap to turn off. The data says there are at least two, in different rooms, and turning off one leaves the other running.
Share of invasive E. coli resistant to three major antibiotic classes at once (third-generation cephalosporins, fluoroquinolones and aminoglycosides), 2000 to 2024. The north holds the line; the south and east climb. Click legend items to isolate a country.
Source: European Centre for Disease Prevention and Control (2024). Measure: invasive E. coli with combined resistance to third-generation cephalosporins, fluoroquinolones and aminoglycosides. Grey lines show all other reporting European countries.
Zoom in on Europe, where surveillance is good enough to watch resistance move year by year. The measure here is a strict one: the share of serious E. coli infections that shrug off three major antibiotic classes at once (cephalosporins, fluoroquinolones and aminoglycosides), the kind of multidrug resistance that leaves doctors with few good options. The pattern is a continent quietly splitting in two. In the Nordic and Benelux countries that share has stayed low and roughly flat. Across much of southern and eastern Europe it has climbed, in some years steeply.
This is the same map as Chart 1, just one step downstream. The countries holding resistance down are mostly the careful prescribers; the ones where it’s rising tend to be the heavier users. Stewardship leaves a fingerprint. Two honest caveats, though. The lines are jumpy because some countries test only small numbers of samples each year, so a single bad batch can spike a curve. And even the “good” north is drifting gently upward. Across the EU, resistant E. coli bloodstream infections have kept rising rather than falling since 2019 (ECDC, 2025). Nobody here is actually winning. Some are just losing more slowly.
Consumption vs resistance across Europe. Each bubble is a country; bigger bubbles use more antibiotics on farms. Where use is high, resistance tends to follow.
Source: Our World in Data (2024); ECDC (2024). Correlation of consumption with resistance r = 0.38. Bubble size = livestock use.
Put it all together and a moderate signal comes through: countries that use more tend to have more resistance. It’s a tendency, not a law. This is an association measured across countries, not proof that one extra prescription causes one resistant infection, and plenty of other things shape the picture too: sanitation, hospital hygiene, how strains travel. But the direction is steady, and look at the biggest bubbles. Several of the high-use, high-resistance countries are also heavy users on their farms. The pressures pile up in the same places.
That’s the uncomfortable finding and, oddly, the hopeful one. Resistance isn’t weather. It’s a response to something we control. Every prescription that wasn’t needed, every healthy animal dosed out of habit, is a small vote for the bacteria. The flip side is that the careful prescribers have already shown the curve can be bent. Chart 4 is the proof. We don’t need a miracle drug to start; we need to use the ones we still have as though we’d like them to keep working. The cost of waiting is already being paid, in the 1.27 million lives a year we can’t save. And the bill is still going up.
European Centre for Disease Prevention and Control. (2024). Antimicrobial resistance (Escherichia coli: combined resistance to third-generation cephalosporins, fluoroquinolones and aminoglycosides) [Data set]. Surveillance Atlas of Infectious Diseases. https://atlas.ecdc.europa.eu/public/index.aspx
European Centre for Disease Prevention and Control. (2025). Antimicrobial resistance in the EU/EEA (EARS-Net): Annual epidemiological report for 2024. ECDC. https://www.ecdc.europa.eu/en/publications-data/antimicrobial-resistance-eueea-ears-net-annual-epidemiological-report-2024
Klein, E. Y., Van Boeckel, T. P., Martinez, E. M., Pant, S., Gandra, S., Levin, S. A., Goossens, H., & Laxminarayan, R. (2018). Global increase and geographic convergence in antibiotic consumption between 2000 and 2015. Proceedings of the National Academy of Sciences, 115(15), E3463–E3470. https://doi.org/10.1073/pnas.1717295115
Mulchandani, R., Wang, Y., Gilbert, M., & Van Boeckel, T. P. (2023). Global trends in antimicrobial use in food-producing animals: 2020 to 2030. PLOS Global Public Health, 3(2), e0001305. https://doi.org/10.1371/journal.pgph.0001305
Murray, C. J. L., Ikuta, K. S., Sharara, F., Swetschinski, L., Robles Aguilar, G., Gray, A., … Naghavi, M. (2022). Global burden of bacterial antimicrobial resistance in 2019: A systematic analysis. The Lancet, 399(10325), 629–655. https://doi.org/10.1016/S0140-6736(21)02724-0
Ortiz-Ospina, E., & Roser, M. (2016). Global health [Antibiotic consumption rate data, adapted from WHO GLASS]. Our World in Data. https://ourworldindata.org/grapher/antibiotic-consumption-rate
World Health Organization. (2024). Global Antimicrobial Resistance and Use Surveillance System (GLASS): Antimicrobial use data, contextual information and antimicrobial use estimates by ATC4 subgroup and AWaRe, 2016–2022. World Health Organization.
Anthropic. (2026). Claude [Large language model]. https://claude.ai
I used Claude (Anthropic, 2026) to help improve grammar and provide some brainstorming ideas. However, all decisions about the visualisations, improvements, data selection, interpretation of results, reconstruction, coding, and analysis were made by me. All data values were checked and verified using the original Our World in Data, WHO GLASS and ECDC sources before being included.