Research and analysis

UK-VARSS 2023 Highlights

Published 19 November 2024

1. Antibiotic use and resistance

 A decade of bringing together antibiotic use and resistance data

This year’s UK-VARSS report mark 10 years of bringing together data on antibiotic sales, usage and resistance. This allows us to show long-term trends and demonstrate how reducing antibiotic use has been followed by a reduction in antimicrobial resistance (AMR) at a national level.

Infographic to show sales of antibiotics in food-producing animals in mg/kg from 2014-2024:

• 2014 - 62.3
• 2015 - 56.5
• 2016 - 39.0
• 2017 - 32.1
• 2018 - 29.0
• 2019 - 30.4
• 2020 - 30.2
• 2021 - 28.3
• 2022 - 25.7
• 2023 - 25.7

Infographic to show sales of HP-CIAs in mg/kg, in food- producing animals 2014-2024:

• 2014 - 0.67
• 2015 - 0.65
• 2016 - 0.38
• 2017 - 0.26
• 2018 - 0.21
• 2019 - 0.17
• 2020 - 0.14
• 2021 - 0.12
• 2022 - 0.12
• 2023 - 0.11

Infographic to show percentage of fully susceptible E.coli isolates 2014-2024 - Full susceptibility means there was no resistance to the antibiotics tested:

• 2014-2015: 18%
• 2015-2016: 20%
• 2016-2017: 23%
• 2017-2018: 34%
• 2018-2019: 33%
• 2019-2020: 37%
• 2020-2021: 39%
• 2021-2022: 40%
• 2022-2023: 43%

Infographic to show percentage of multi-drug resistant E.coli isolates 2014-2024. Multi-drug resistance means resistance to three or more classes of antibiotics:

• 2014-2015: 57%
• 2015-2016: 49%
• 2016-2017: 45 %
• 2017-2018: 36%
• 2018-2019: 36%
• 2019-2020: 32%
• 2020-2021: 31%
• 2021-2022: 31%
• 2022-2023: 27%

2. Antibiotic sales

2.1  Sales for food-producing animals (mg/kg)

Sales of veterinary antibiotics for use in food-producing animals, adjusted for animal population, were 25.7 mg/kg; this represents no change since 2022 and an overall 59% (36.6 mg/kg) decrease since 2014.

Infographic showing the sales in mg/kg between 2014 and 2022:

• 2014 - 62.3
• 2015 - 56.5
• 2016 - 39.0
• 2017 - 32.1
• 2018 - 29.0
• 2019 - 30.4
• 2020 - 30.2
• 2021 - 28.3
• 2022 - 25.7
• 2023 - 25.7 which is not changed compared to 2022 and a 59 % decrease since 2014

Sales of Highest Priority Important Antibiotics (HP-CIAs) in food-producing animals remain at very low levels at 0.11 mg/kg in 2023 and account for less than 0.5% of total sales.

Infographic showing the classes of HP-CIAs and total sales for 2014, 2022 and 2023:

• Total HP-CIAs: In 2014 0.67 mg/kg were sold, in 2022 0.12 mg/kg were sold and in 2023 0.11 mg/kg were sold which represents a 84% decrease since 2014.
• Total Fluoroquinolones: In 2014 0.35 mg/kg were sold, in 2022 0.10 mg/kg were sold and in 2023 0.09 mg/kg were sold which represents a 74% decrease since 2014.
• Total 3rd and 4th generation cephalosporins: In 2014 0.19 mg/kg were sold, 0.02 mg/kg were sold and in 2022 0.02 mg/kg were sold which represents a 91% decrease since 2014.
• Colistin: In 2014 0.12 mg/kg was sold, in 2021 0.00 mg/kg was sold and in 2022 0.00 mg/kg was sold which represents a 100% decrease since 2014.

A new harmonised mg/kg metric for all animals has been developed for analysing sales data. This uses different animal categories and weights, meaning it is not comparable with the current food producing animal mg/kg metric. This metric is expected to be used when the EU publishes their 2023 sales in 2025. We have introduced this new metric as we recognise the value of harmonised regional surveillance. The 2023 UK sales using this metric is 16.5 mg/kg.

2.2 Sales of antibiotics for all animals in tonnes

In 2023 the total quantity of antibiotic active ingredient sold in the UK was 189 tonnes, the lowest sales to date.

Infographic showing the total sales for all animals in tonnes:

• 2014 - 447
• 2015 - 406
• 2016 - 293
• 2017 - 246
• 2018 - 223
• 2019 - 229
• 2020 - 227
• 2021 - 212
• 2022 - 193
• 2023 - 189 which is a reduction of 58% since 2014.

Over 60% of all antibiotics sold were either tetracyclines (23%) or penicillins (28%)

2.3 Sales of HP-CIAs for all animals in tonnes

Sales of HP-CIAs, not adjusted for animal population size, for 2023 was 0.82 tonnes, representing a decrease of 9% (0.09 tonnes) since 2022 and an 83% decrease (3.96 tonnes) since 2014. Sales of HP-CIAs continue to represent a small proportion (<0.5%) of total veterinary antibiotic sales in tonnes.

3. Antibiotic usage

Antibiotic usage refers to the amount of antibiotics prescribed and/or administered per sector. The data have been collected and provided to the VMD by the animal industry on a voluntary basis. Coverage is at least 85% for all sectors shown, Ruminant coverage of use data is low and may not be representative of the national    sectors and is not included in the report.

3.1 Antibiotic usage by food-producing animal species

Infographic showing the antibiotic usage, in tonnes, by food-producing animal species for 2023, the change since 2022 and the percentage change since the date of first publication:

• Pigs: 61.2 which is equivalent to 84.8 mg/kg. This is a 13.0 mg/kg increase since 2022 and represents an overall decrease of 69% since date of first publication in 2015
• Broilers: 13 which is equivalent to 13.5 mg/kg. This is a 0.5 mg/kg decrease since 2022 and represents an overall decrease of 72% since date of first publication in 2014.
• Gamebirds: 9.9 which is a 3.1 tonne increase since 2022 and represents an overall decrease of 51% since date of first publication in 2016.
• Salmon: 3.7 which is equivalent to 19.9 mg/kg. This is a 1.4 mg/kg increase since 2022 and represents an overall increase of 24% since date of first publication in 2017.
• Turkeys: 1.7 which is equivalent to 33.6 mg/kg. This is a 1.7 mg/kg decrease since 2022 and represents an overall decrease of 85% since date of first publication in 2014.
• Laying hens: 1.5 which is reported at 0.22% bird days. This is a 0.02% bird days decrease since 2022 and represents a 67% decrease since date of first publication in 2016.
• Trout: 0.09 which is equivalent to 6.9 mg/kg. This is a 37.2 mg/kg increase since 2022 and represents an overall decrease of 64% since date of first publication in 2017.
• Ducks: 0.01 which is equivalent to 0.4 mg/kg. This is a 0.1 mg/kg increase since 2022 and represents an overall decrease of 97% since date of first publication in 2014.

3.2 Highest Priority Critically Important Antibiotics by food-producing animal species

Infographic showing the use of HP-CIAs in kg and change since 2022 as well as change since when data was first published:

• Gamebirds: 27.7 kg. This is a 4.5 kg increase since 2022 and represents an overall decrease of 57% since date of first publication in 2016.
• Trout: 22.7 kg which is equivalent to 1.8 mg/kg. This is a 0.4 mg/kg decrease since 2022 and represents an overall decrease of 73% since date of first publication in 2017.
• Pigs: 5.2 kg which is equivalent to 0.007 mg/kg. This is a 0.005 mg/kg decrease since 2022 and represents an overall decrease of 99% since date of first publication in 2015.
• Meat poultry: 5.1 kg which is equivalent to 0.00441 mg/kg. This is a 0.0044 mg/kg increase since 2022 and represents an overall decrease of 99% since date of first publication in 2014.

4. Antibiotic resistance in zoonotic and commensal bacteria from healthy animals at slaughter

4.1 Key resistance outcome indicators: E. coli

The harmonised monitoring outcome indicators combine results from healthy pigs and poultry at slaughter to give an overall picture of antimicrobial resistance (AMR), and are internationally comparable. Results from 2023 are very positive, with considerable improvements since 2015. The percentage of E. coli isolates showing full susceptibility to the panel of antibiotics tested is at a new high of 43%, more than doubling since 2014/2015. The percentage of multi-drug resistant isolates, that is resistant to three or more antibiotic classes, is at a new low of 27%.

Infographic showing the proportion of E. coli isolates showing full susceptibility to the panel of antibiotics tested and the percentage of multi-drug resistant isolates:

• 2014/2015, the proportion of E. coli isolates fully susceptible was 18%. The percentage of multi-drug resistant isolates was 57%.
• 2015/2016, the proportion of E. coli isolates fully susceptible was 20%.  The percentage of multi-drug resistant isolates was 49%.
• 2016/2017, the proportion of E. coli isolates fully susceptible was 23%. The percentage of multi-drug resistant isolates was 45%.
• 2017/2018, the proportion of E. coli isolates fully susceptible was 34%. The percentage of multi-drug resistant isolates was 36.%.
• 2018/2019, the proportion of E. coli isolates fully susceptible was 33%. The percentage of multi-drug resistant isolates was 36.%.
• 2019/2020, the proportion of E. coli isolates fully susceptible was 37%.  The percentage of multi-drug resistant isolates was 32.%.
• 2020/2021, the proportion of E. coli isolates fully susceptible was 39%.  The percentage of multi-drug resistant isolates was 31.%.
• 2021/2022, the proportion of E. coli isolates fully susceptible was 40%. The percentage of multi-drug resistant isolates was 31.%.
• 2023/2023, the proportion of E. coli isolates fully susceptible was 43. The percentage of multi-drug resistant isolates was 27%.

4.2 New AMR Surveillance

This year’s harmonised monitoring includes testing for three new species of bacteria in pigs: Campylobacter coli, Enterococcus faecalis and Enterococcus faecium. The addition of enterococci allows for detection of vancomycin-resistant enterococci (VRE), which are of clinical importance in people. No VRE were detected in pigs in 2023.

4.3 Pathogen Surveillance in Agriculture, Food and Environment (PATH-SAFE)

This year’s report includes results from AMR surveillance pilots in milk from dairy cattle, beef cattle and sheep carried out under the PATH-SAFE programme. These surveys mirror the methodology used in our routine surveillance in pigs and poultry. However, the PATH-SAFE surveys were not conducted over an entire calendar year, and coverage of animal populations (27-28%) was less than that achieved in our routine harmonised monitoring (81% in 2023). Nonetheless these results provide an initial baseline for AMR in cattle and sheep. Results show that full susceptibility of E. coli from ruminants was extremely high (>86%), although there were some notable sector-specific findings, for example, carriage of ESBL/AmpC-producing E. coli in beef cattle.

4.4 E. coli in different animal species

Full susceptibility in E. coli.

Infographic showing the percentage of E. coli isolated from different animal species that are fully susceptible to the panel of antibiotics tested, that is, no resistance was found.

These E. coli were isolated from our routine surveillance, including harmonised monitoring, and PATH-SAFE pilot surveys, which took samples of gut (caecal) content from healthy animals at slaughter, or for dairy cattle, from bulk milk tanks:

• Pigs: Harmonised Monitoring 2023: 42 % of isolates fully susceptible to E. coli
• Chickens: Harmonised Monitoring 2022: 45% of isolates fully susceptible to E. coli
• Turkeys: Harmonised Monitoring 2022: 22% of isolates fully susceptible to E. coli
• Beef cattle: PATH-SAFE Pilot 2023: 90% of isolates fully susceptible to E. coli
• Sheep: PATH-SAFE Pilot 2023: 89% of isolates fully susceptible to E. coli
• Dairy cattle: PATH-SAFE Pilot 2023: 86% of isolates fully susceptible to E. coli

Using selective media to detect ESBL-and/or AmpC-producing E. coli:

• We also perform a more sensitive type of testing using selective media which inhibits the growth of susceptible bacteria but allows ESBL-/AmpC-producing E. coli to multiply, making them easier to detect.
• This tells us the percentage of individual animals carrying resistance to 3rd and 4th generation cephalosporins even at very low levels.

Infographic showing the percentage of individual animals carrying resistance to 3rd and 4th generation cephalosporins even at very low levels:

• Pigs - 23
• Chickens - 11
• Turkeys - 8.5
• Beef cattle - 27
• Sheep - 8.6

5. Antibiotic resistance in clinical surveillance

Clinical surveillance aims to provide veterinarians with relevant treatment information using results from bacteria isolated from diagnostic samples. As this kind of scanning surveillance is subject to biases and differences in the number of samples, the results are not representative of the UK’s wider animal populations.

5.1 Key findings

• 7,415 isolates were tested for AMR in England and Wales predominantly Salmonella and E. coli
• Resistance in E. coli across all animal species mostly shows decreasing resistance since 2014
• The percentage of isolates tested by animal species were:
  • pigs - 13%
  • poultry - 30%
  • cattle - 15%
  • sheep - 7%
  • dogs - 10%
  • trout - <1%

5.2 Resistance in Escherichia coli

• 1,168 E. coli isolates were tested from all species.
• 18% of isolates were resistant to four or more individual antibiotics, which could limit treatment options for veterinarians. This was most frequent in isolates from cattle (24%) and pigs (23%).
• Across species, resistance tended to be higher in younger animals, which may reflect more frequent antibiotic use in this age group.

Infographic showing the percentage of E. coli isolates with limited treatment options in 2023, that is, resistant to four or more individual antibiotic:

• pigs - 23%
• chickens - 10%
• turkeys - 0%
• cattle - 24%
• sheep - 5.7%

5.3 Highest priority critically important antibiotics (HP-CIAs)

Resistance was low, very low or not detected in E. coli: cefotaxime (8.7%), cefpodoxime (0.7%), ceftazidime (4.9 %) and enrofloxacin (1.7%).

5.4 Resistance in Salmonella spp. from animals and their environment:

• Of the 5,513 Salmonella isolates tested, 73% were fully susceptible. The lowest levels of susceptibility were in pigs (19%) and turkeys (30%).
• When looking at data since 2014, there are sector specific differences, with pigs, chickens and turkeys all showing trends of increasing full susceptibility. Full susceptibility in Salmonella from sheep and cattle remain high.
• 14% of Salmonella from all species indicated limited treatment options, that is, resistant to four or more individual antibiotics. This was most frequent in isolates from pigs (76%).

Infographic showing percentage of isolates fully susceptible to all antibiotics tested in 2023:

• pigs - 19%
• chickens - 81%
• turkeys - 30%
• cattle - 89%
• sheep - 97%
• dogs - 82%
• horses - 92%

5.5 Highest priority critically important antibiotics (HP-CIAs)

Resistance was low, very low or not detected in Salmonella spp.: cefotaxime and ceftazdime (>0.9%) and ciprofloxacin (<2.6%).

5.6 National Biosurveillance Network (NBN)

The NBN is a major cross-government initiative to pilot and improve surveillance for biological threats across the One Health spectrum. It will assess how to better share and bring together surveillance outputs, including facilitating sharing of privately-held data, to generate better insights into threats. Most AMR data from animals in the UK is generated and held by private veterinary laboratories (PVLs) and does not ordinarily feed into AMR surveillance. Under the NBN pilot, the VMD is expanding on previous work with PVLs to facilitate data sharing with government across species sectors.

6. Background

6.1 How sales data are collected

In the UK, from 2005 it has been a statutory requirement for pharmaceutical companies to report to the VMD the amount of antibiotic products sold for use in animals. The quantity of active ingredient is calculated from the amounts sold and the product characteristics. These sales data do not take into account wastage of veterinary antibiotics. However, this is the best currently available approximation of the quantity of antibiotics administered to animals in the UK.

6.2 How usage data are collected

Data have been voluntarily provided by producers in the pig, poultry and laying hen sectors, feed companies for gamebirds and veterinary practices for gamebirds and fish. Usage data collection systems have been put in place to collect data from:

• British Poultry Council (meat poultry)
• British Egg Industry Council (laying hen sector)
• Game Farmers Association (gamebirds)
• The electronic Medicines Book (pigs)
• British Trout Association (trout)
• Scottish Salmon Producers’ Association (salmon)

Usage data, the amount of antibiotics purchased, prescribed and/or administered, have the potential to provide much more precise estimates of use. The VMD has been working with the animal production sectors to develop sector-led data collection systems to monitor their antibiotic usage.

6.3 Population Correction Unit (PCU)

Trends in sales of antibiotics between years and different countries cannot be determined without taking into consideration variations in the number and size of animals that may require treatment. Therefore, sales data are analysed using the population correction unit (PCU). This is a standard technical unit of measurement developed by the European Medicines Agency and adopted by EU countries. This allows data to be presented as mg of antibiotic per kg of livestock biomass. For more details refer to, Understanding the mg/PCU calculation used for antibiotic monitoring in food producing animals.

6.4 Critically Important Antibiotics (CIAs)

Certain antibiotic classes are categorised by the World Health Organization (WHO) as critically important antibiotics for human use, of which several are designated as ‘highest priority critically important antibiotics’(HP-CIA). In January 2020, the European Medicines Agency published new scientific advice on the risk to humans from antibiotic resistance caused by the use of (HP-CIAs) in animals. The report was prepared by Antimicrobial Advice Ad Hoc Expert Group (AMEG). Quinolones, third and fourth generation cephalosporins and polymyxins were classified as category B, where the use of these antibiotics should be restricted, as a result of their critical importance in human medicine. For more details refer to the EMA document, AMEG 2018 - Categorisation of antibiotics.

6.5 How antibiotic resistance is interpreted

Antibiotic resistance in bacteria isolated from animals is monitored through two distinct surveillance programmes: harmonised monitoring and clinical surveillance. The harmonised monitoring scheme is a UK-wide programme in which we test bacteria from the gut of healthy pigs and poultry at slaughter, and the National Control Programme for Salmonella, giving us a representative picture of resistance in key livestock species entering the food chain. Clinical surveillance involves the testing of bacteria that have been isolated from clinical samples submitted by farmers and private veterinarians to government laboratories in England and Wales.

Susceptibility testing for harmonised monitoring is performed using broth microdilution to determine minimum inhibitory concentrations (MICs). Resistance is assessed using EUCAST (European Committee on Antimicrobial Susceptibility Testing) epidemiological cut-off values (ECOFFs) where possible. Where no EUCAST values were available, European Food Safety Authority (EFSA) recommended cut-off values were used. Where neither defined values are available, tentative EUCAST ECOFF values were applied. Results interpreted using ECOFFs are reported in full in Supplementary Material 2.

In the 2023 clinical surveillance programme, the majority of the results presented were assessed by disc diffusion techniques, and interpreted using BSAC (British Society for Antimicrobial Chemotherapy) human clinical breakpoints (CBPs) where available. Broth microdilution testing was introduced in UK-VARSS 2020 and continues to be developed and applied to an increasing number of organisms over the last four years. MIC results have been interpreted using veterinary clinical breakpoints from Clinical and Laboratory Standards Institute (CLSI) in the first instance, or Comité de l’Antibiogramme de la Société Française de Microbiologie (CA-SFM) when these are not available; if veterinary clinical breakpoints were not available, human clinical breakpoints were used. Full details of the methods used are available in Supplementary Material 2.