ENCS
Tuesday, 11 August 2026
Your paper, your pace.

Antibiotics: what has actually slowed resistance, and what only looked like it

The EU promised to cut human antibiotic use by 20% by 2030. In 2024 it was using 2% more than in 2019, and not one member state had reached its own target.

Checked on

All ten problems

In 2024 the European Union used 20.3 defined daily doses of antibiotics per 1,000 people per day. A defined daily dose, or DDD, is the assumed average daily amount of a drug for an adult, so that figure means about two people in every hundred were taking an antibiotic on any given day. In 2023 the EU set itself a target: 20% less by 2030, which comes to 15.9. Consumption in 2024 was 2% above the 2019 baseline, not below it, and no member state had reached its own target. ECDC publishes a one-page factsheet per country, so you can look up your own.

Two words on this page do different work, and the difference is routinely lost. The Lancet’s GRAM study estimated that in 2021, 1.14 million deaths were attributable to bacterial resistance and 4.71 million were associated with it. Attributable asks what would have happened if the same infection had been treatable. Associated asks how many people died with a resistant infection present, whatever else was wrong. The second number is about four times the first, and it is the one usually quoted with the word dropped. ECDC’s separate estimate for the EU and EEA is more than 35,000 attributable deaths a year, on 2016 to 2020 data.

Nobody wins this one outright. Bacteria that survive a drug pass the trick on, and every dose does the sorting. The levers are how many doses are given, which drugs they are, and how many infections are stopped before a drug is needed at all. Below are five countries that pulled one of those levers and then measured what happened — including one that spent $770 million to develop a drug and sold $0.8 million of it.

How this is measured

One number, named and sourced. Everything on this page is argued against it, so you can check us.

Total consumption of antibacterials for systemic use (ATC group J01) in humans, community and hospital sectors combined, in defined daily doses per 1 000 inhabitants per day DDD per 1 000 inhabitants per day

ECDC, ESAC-Net Annual Epidemiological Report for 2024, published November 2025; country values from ECDC country factsheets 'Progress towards 2030 targets', 2025 update, 2024 data

Who is actually ahead

Ranked on the number above — not on reputation.

Total consumption of antibacterials for systemic use (ATC group J01) in humans, community and hospital sectors combined, in defined daily doses per 1 000 inhabitants per day
Country Result Why
Netherlands 9.8 DDD The lowest in the EU and EEA. Even so it is above its own 2030 target of 9.2, because it was already at 9.5 in 2019 and went up.
Austria 11.8 DDD Up from 11.6 in 2019 against a target of 11.2. Austria has one of the smallest cuts to make in Europe and has not made it.
Estonia 13.0 DDD Up 10% from 11.8 in 2019. Its target is 11.4, so the gap widened rather than closed.
Germany 13.8 DDD Up from 12.6 in 2019 against a target of 11.5. ECDC flags the German hospital share as estimated, so treat the level as approximate.
Czechia 19.0 DDD Up 12% from 16.9 in 2019, target 15.4. ECDC also notes Czechia changed how it collects the data during the period.
Cyprus 23.5 DDD The steepest fall in the EU: down 22% from 30.1 in 2019. Its Access-group share fell too, from 48.9% to 43.2%.
France 26.5 DDD Up 6% since 2019 against the toughest target in Europe, 18.3. France would have to cut more than a quarter in five years.
Greece 29.9 DDD The highest in the EU and EEA, and down 12% from 34.1 in 2019. Still 5.0 above its 2030 target of 24.9.

What has actually been tried

Real countries, real policies, and what the numbers did afterwards — including where it went wrong.

Netherlands

The Netherlands does two things at once and has done for thirty years. Doctors prescribe restrictively by professional agreement rather than by law, and hospitals run a 'search and destroy' policy: patients who have been treated in a hospital abroad, or who work with live pigs or veal calves, are screened for MRSA and isolated until cleared.

What happened. Total human consumption was 9.8 DDD per 1 000 inhabitants per day in 2024, the lowest of the 29 EU and EEA countries, against an EU mean of 20.3. NethMap 2025 found MRSA in 2% of Staphylococcus aureus blood isolates, stable over five years. Carbapenem resistance in Klebsiella pneumoniae was 0.2% in 2024.

The catch. Low is not the same as falling. Dutch consumption rose from 9.5 in 2019 to 9.8 in 2024, so the country is moving away from its own modest 2030 target of 9.2. Carbapenem-resistant Klebsiella climbed from 0.04% in 2020 to 0.2% in 2024 — five times as much, from almost nothing. Isolation rooms also cost beds and privacy, which no indicator records.

ECDC country factsheet, Netherlands (2024 data); RIVM, NethMap/MARAN One Health 2025, report 2025-0122

Sweden

Sweden built Strama in 1995: a voluntary network of doctors, pharmacists and county authorities that feeds every prescriber their own prescribing data, compares it with neighbours, and revises treatment guidelines locally. It has no power to refuse a prescription. In 2009 the country set itself a target of no more than 250 outpatient prescriptions per 1 000 inhabitants a year.

What happened. Outpatient prescribing fell from a peak of about 565 per 1 000 inhabitants a year in 1992 to 271 in 2024. Among children aged 0 to 4 it fell from 1,328 to 259, a drop of 81%. Sweden has still not reached the 250 target, seventeen years after setting it.

The catch. Prescribing less does not stop resistance arriving. MRSA in Swedish blood cultures rose from 2.1% in 2023 to 2.6% in 2024, and 3,937 MRSA cases were notified in 2024, 11% more than the year before. Sweden also reported no total-consumption figure to ECDC for 2024, so its line in the European table reads 'No data'.

Public Health Agency of Sweden and SVA, Swedres-Svarm 2024, published 2025; ReAct, Strama case study

Denmark

Denmark stopped feeding antibiotics to healthy animals to make them grow. Growth promoters were banned in broiler chickens and finisher pigs in 1998 and in weaner pigs in 1999, before the EU-wide ban of 2006. In 2010 Denmark added the Yellow Card scheme, which sets a consumption threshold per herd and puts farms above it under supervision.

What happened. Antibiotic use in Danish swine production fell by more than 50% between the 1992 peak and 2008, while pig production rose by nearly 50% over the same period. Diarrhoea in young pigs rose immediately after the ban, then fell back to pre-ban treatment levels within about a year. Pigs took roughly 1.6 days longer to reach slaughter weight.

The catch. This is the clearest measured reduction on the page, and it does nothing for human prescribing: Danish human consumption was 16.1 DDD in 2024, above its 13.9 target and up from 15.3 in 2019. The most detailed English-language account of the ban is a fact sheet by the Pew Charitable Trusts, an American foundation that campaigned on the issue; it compiles Danish DANMAP data rather than generating its own.

Pew Charitable Trusts, 'Denmark's Ban on Growth Promoting Antibiotics in Food Animals', comprehensive fact sheet, 2010, compiling DANMAP data; ECDC country factsheet, Denmark (2024 data)

United States

The United States tried to pay for new antibiotics with market exclusivity. The GAIN Act of 2012 gave any 'qualified infectious disease product' five extra years of protection from generic competition plus priority review. Achaogen, a small company backed partly by the US biodefence agency BARDA, spent fifteen years developing plazomicin for drug-resistant urinary infections.

What happened. Achaogen raised $770 million between 2004 and 2019, $200 million of it public money. The FDA approved plazomicin in June 2018. It sold $0.8 million of the drug that year. The company filed for bankruptcy on 15 April 2019; the assets fetched $16 million at auction, with Cipla paying $4.65 million up front. By 2018 the FDA had granted 147 GAIN designations and approved 12 products, none with a new mechanism of action.

The catch. Exclusivity is worth what the drug earns, and a good new antibiotic is deliberately kept on the shelf — the better the stewardship, the worse the sales. That is the mechanism, not bad luck. The financial reconstruction was published by university researchers in 2024, not by the FDA, which has published no post-mortem of its own.

'Novel insights from financial analysis of the failure to commercialise plazomicin', Humanities and Social Sciences Communications, 20 July 2024; 'Incentivizing antibiotic development: why isn't the GAIN Act working?', Open Forum Infectious Diseases, 28 January 2020

Greece

Greece has the highest antibiotic consumption in Europe and was given the toughest cut in the EU's 2023 Council Recommendation: 27% below its 2019 level by 2030. It has been running national stewardship and prescribing-control programmes throughout, and it is one of only two EU countries whose consumption has fallen substantially since 2019.

What happened. Consumption fell from 34.1 DDD per 1 000 inhabitants per day in 2019 to 29.9 in 2024, a cut of 12%. Over the same five years bloodstream infections with MRSA rose from 4.59 to 5.54 per 100 000 people (+20.7%), with third-generation cephalosporin-resistant E. coli from 2.58 to 5.4 (+109%), and with carbapenem-resistant Klebsiella pneumoniae from 13.05 to 14.89 (+14.1%).

The catch. Two honest readings. Five years is short for resistance to answer a change in prescribing, and Greek hospitals were absorbing a pandemic in the middle of it. Against that, the share of Greek consumption in the WHO Access group — the narrow, first-line drugs — fell from 46.8% to 43.0%, so fewer doses of worse drugs. ECDC also notes Greece changed its data collection process during the period, which affects comparability.

ECDC country factsheet, Greece, 'Progress towards 2030 targets', 2025 update (2024 data); ECDC, Antimicrobial resistance in the EU/EEA, annual epidemiological report for 2024

What a machine would optimise for

This is arithmetic, not advice and not a prediction. We name the single number being maximised and follow it wherever it goes. The point is to see the shape of the answer a calculator gives.

The number being maximised

The single number is total human antibiotic consumption in defined daily doses per 1 000 inhabitants per day. Drive the EU mean from 20.3 down to the 2030 target of 15.9, and nothing else enters the calculation.

  1. Buy doses where the doses are. Greece at 29.9 and France at 26.5 hold Europe's excess; the Netherlands at 9.8 has almost nothing left to remove. A euro spent on prescribing in Lyon moves the European mean many times further than the same euro spent in Utrecht, because the mean is weighted by population and France holds about 15% of it.
  2. Shorten courses before refusing patients. A defined daily dose is one day of treatment, so cutting a standard course from ten days to five halves that patient's contribution while treating exactly the same number of people. The indicator cannot tell the two apart.
  3. Remove the cheap doses first. The count treats one day of amoxicillin and one day of a last-line carbapenem as one each, so the fastest route to the target runs through the common, narrow, mostly harmless prescriptions — and leaves the broad-spectrum drugs that drive resistance almost untouched.
  4. Read a blank as a zero. Sweden's 2024 line in the European table says 'No data', because no total-consumption figure was reported to ECDC that year. A calculator working only from this table cannot distinguish a country that improved from a country that did not answer.

The arithmetic

EU/EEA, 2024: the population-weighted mean is 20.3 DDD per 1 000 inhabitants per day, against a 2030 target of 15.9. The gap is 4.4. France sits at 26.5 with about 68 million people, roughly 15% of the EU's 449 million. Bringing France alone to the EU mean removes 6.2 doses across 15% of the population — 6.2 × 0.15, or about 0.9 of the 4.4 gap. One country, a fifth of the whole European shortfall. Emptying the Netherlands to zero would remove about 0.4.

What the machine would miss

The most important part of this page. A number that goes up can still be paid for by somebody, and some things never make it into the number at all.

Who pays

The patient who never had the drug in the first place. In 2019, across Bangladesh, Brazil, Egypt, India, Kenya, Mexico, Pakistan and South Africa, there were close to 1.5 million infections with carbapenem-resistant Gram-negative bacteria, and in only 6.9% of cases was an appropriate antibiotic even potentially available — from 0.2% in the worst country to 14.9% in the best. A target that counts doses downwards scores every one of those missing courses as progress.

What the number flattens

A dose is not a decision. The count cannot see whether a prescription was right, wrong or life-saving, and it cannot see which drug it was. Greece removed 12% of its doses between 2019 and 2024 and its MRSA bloodstream infections still rose 20.7%, its resistant E. coli 109%. Across the EU the share of consumption in the WHO Access group — the narrow first-line drugs — fell from 61.3% in 2019 to 60.3% in 2024. Slightly more antibiotics, slightly worse ones.

What a spreadsheet cannot see

The operations that quietly stop being offered. Hip replacements, caesarean sections and chemotherapy all rely on antibiotics holding for a few days around the procedure; when they stop holding, no one records a resistant infection, they record a patient managed conservatively. The GRAM study counted 1.14 million deaths attributable to bacterial resistance in 2021 and 4.71 million associated with it. Neither figure counts an operation that was never booked.

Where does your country stand?

Open my country

Where the numbers come from

We are a newspaper, not a government and not an adviser. Nothing on this page is a recommendation to you or to anyone in office. It is what was tried, what it measured, and what a calculator would say if you let it loose on the same problem.

One good piece of thinking a day

The day's most worthwhile story, and the question underneath it. No spam, one click to leave.

One click to leave. We never sell or share your address.