5 Comparative analysis
5.1 Comparison of antibiotic sales in human and veterinary medicine
5.1.1 Data included and calculations
For human medicinal products with antibiotics, the numbers on the total amount of antibiotics consumed for systemic use to humans (ATC group A07AA oral glycopeptides and J01 excluding methenamine; sales on requisition and on prescriptions to individuals; ATC/DDD index version 2025) were retrieved as defined daily doses and calculated to kg active substance.
For veterinary medicinal products with antibiotics, data on sales of antibiotics for use in animals (QJ01 and QA07AA) are those presented in Sales of antibiotics for animals except products for intramammary and intrauterine use (QG01 and QJ51). Sales for aquaculture were not included, nor were sales of drugs authorised for human use but sold for use in animals.
To estimate the biomass of the human population, data on population numbers by age were multiplied with the corresponding average body weights from studies made by Statistics Sweden in 2016. For animal body mass, the data on population correction unit for 2024 was used as a proxy for 2025 (European Medicines Agency (EMA) 2025). This unit roughly corresponds to the total biomass of major animal populations, excluding dogs and cats.
5.1.2 Total sales
A total of 61.7 and 9.0 tonnes of antibiotics were consumed in human and veterinary medicine, respectively, in 2025 from the included ATC classes. Beta-lactam antibiotics remain the most prescribed antibiotics in both human and veterinary medicine and represent the largest volumes consumed, measured in kilograms. Other antibiotic products were consumed in smaller quantities than beta-lactams but considering their chemical and pharmacological properties, they could have a greater impact on the environment and the emergence of antibiotic resistance. The largest difference is noted for fluoroquinolones, where sales for humans are more than 150 times higher than for animals and constitute approximately 4% of total sales for humans included in this analysis. For animals, sales of fluoroquinolones constitute 0.2% of the total sales.
In total, 86.5 and 11.8 mg active antibiotic substance per kg estimated biomass were sold in 2025 in human and veterinary medicine, respectively. Total sales data do not take the heterogeneity of likelihood of exposure within the population into account. This is especially true for data on sales for use in animals, as certain substances may only or mainly be sold for use in one particular animal species. Consequently, the selective pressure in a particular subset of the population (i.e. a particular animal species) could be far larger than in the total population. Both in tonnes active substance and in mg per kg estimated biomass, antibiotic sales are higher for humans than for animals in Sweden.
5.2 Comparison of antibiotic resistance in human and veterinary medicine
5.2.1 ESBL-producing Enterobacterales
Enterobacterales with ESBLA or ESBLM, and their corresponding genes, can transfer between animals and humans (European Food Safety Authority (EFSA) 2011; Been et al. 2014). The main route would be via food, but the possibility for direct transfer when handling animals should be kept in mind.
The available data show that ESBL-producing bacteria are generally rare in animals and food in Sweden. Previously, the occurrence in intestinal samples from broilers was high but it has decreased considerably in recent years. Moreover, previous investigations when the occurrence was higher has shown that ESBLA- or ESBLM-producing E. coli constitute a small part of all the E. coli in the intestinal flora in a majority of the broiler samples. Finally, it has previously been shown that most isolates from humans in Sweden are not of the same types of ESBLA or ESBLM as in broilers. Hence, nothing indicates a need to revise the conclusion that food on the Swedish market is a limited source for ESBLs for humans (Börjesson et al. 2016). Nevertheless, continued vigilance is warrented to prevent the development of reservoirs of ESBL-producing Enterobacterales in animals.
5.2.2 MRSA
Zoonotic transmission of MRSA occurs by direct or indirect contact. MRSA is reported globally in farm animals, companion animals, horses and wildlife. During the year, MRSA was isolated from pigs in a study with selective culture methods and sporadically in clinical samples from dogs, horses and cats. From dogs and cats, different spa-types were isolated, most of them previously found in humans. The situation among humans is also favourable.
Livestock-associated MRSA
In the last two decades, the zoonotic aspects on MRSA in farm animals has widened in many countries, largely due to spread of livestock-associated MRSA, and concerns pigs, veal calves, broilers and dairy cows. For pigs, the main clonal complex (CC) is 398.
Based on surveillance of MRSA in livestock, with occasional findings in samples from cow, pig, goat and sheep, the situation has previously been considered favourable in Sweden. The prevalence in pigs is insufficiently studied, but in an EU baseline study with sampling of fattening pigs at slaughter and selective culture methods, MRSA was detected in 29% of investigated slaughter batches. In all positive batches, MRSA of CC398 was detected. Spa-type t034 was most common (n=50), followed by t011 (n=7) and a few t571 (n=3). How well this reflects the occurrence on farm is not known but sampling of farms with piglets is being performed in 2025/2026. Furthermore, MRSA CC398 occurs among horses and spa-type t011 (n=5) was the most common in 2025.
In Sweden, human cases of MRSA CC398 are rare. In 2025, among all MRSA cases with available typing data, 23 cases were identified with the following spa-types: t034 (n=14), t011 (n=3), t1451 (n=2), t4652 (n=1), t4885 (n=1), t1456 (n=1), and t899 (n=1). Of these, ten isolates tested negative for PVL, while the PVL status of the remaining 13 isolates was unknown. Additionally, eleven isolates belonging to CC398 were detected through Regional whole genome sequencing. The possibility of animal contacts as a source is often not pursued. Consequently, epidemiological information regarding this is scarce.
MRSA with mecC
Isolates of MRSA with mecC were first reported internationally from dairy cows and humans in 2011 (García-Álvarez et al. 2011; Shore et al. 2011; Ito et al. 2012). Throughout the years, MRSA with mecC have been isolated from several animal species (cat, cow, dog, hedgehog, goat, pig and sheep). The total number of cases are low even if there are a number of isolates from hedgehogs in research projects and from goats in an outbreak at a zoo.
Human cases of mecC-positive MRSA are also uncommon in Sweden. Five reported cases with the following spa-types were identified in 2025: t843 (n=2), t16761 (n=2) and t208 (n=1). In addition, another mecC case was typed to ST130 by whole genome sequencing. The epidemiological information concerning possible animal contacts is scarce but some of the spa-types in cases from humans have also been found in cases from animals. However, even if there would be zoonotic transfer, it is not currently considered a public health problem, as the number of cases of MRSA with mecC in humans in Sweden is low.
MRSA-types typically associated with humans
MRSA isolated from dogs and cats often belong to spa-types seen in MRSA from humans. This supports the view that humans often are the source of MRSA in companion animals (European Food Safety Authority (EFSA) 2009; Committee for medicinal products for veterinary use (CVMP) 2009). Spread can subsequently occur from animals to humans. However, the impact of companion animals as vectors for spread between humans is not known.
Conclusions
The MRSA situation in Sweden is in general still favourable both in humans and in animals. Biosecurity, with caution in trade of live animals and measures to prevent introduction by indirect routes, is important for preventing introduction and spread of MRSA in animal populations. However, with the detection of MRSA in the Swedish pig production, further emphasis need to be put not only on these matters to reduce further spread but also on hygiene measures to prevent spread from farms to the general public. Furthermore, antibiotic stewardship as well as hygiene and infection prevention and control measures are important to prevent healthcare-related spread between people, between animals or between people and animals.
For more information on MRSA in Sweden, see Chapter 3 - Antibiotic resistance in humans and Chapter 4 - Antbiotic resistance in animals.
5.2.3 MRSP
Staphylococcus pseudintermedius may act as an opportunistic pathogen in humans and there are several reports in the literature of infections in humans with a varying degree of severity. However, MRSP is not generally considered to be a zoonotic pathogen.
5.2.4 VRE
VRE have historically been isolated from a large proportion of broilers in Sweden using selective media. This occurrence has however decreased considerably. The occurrence in humans varies between years, mainly due to nosocomial outbreaks causing high occurrence in some years. However, based on genotypical investigations of isolates, there are no indications that the presence of VRE in broilers in Sweden has affected the situation in Swedish healthcare.
5.2.5 Salmonella
Occurrence of Salmonella among farm animals, as well as among other animals, in Sweden is low and few incidents involve multiresistant strains. In 2025, the majority of isolates (92 of 108; 85%) were susceptible to all antibiotics tested. Resistance to fluoroquinolones (e.g. ciprofloxacin) is rare and in 2019, a strain with ESBL was detected for the first time, in an environmental sample from a farm. Thus, the overall situation in the veterinary sector is favourable, largely due to the strategies in the Swedish salmonella control programme initiated in the 1950s. In 2025, half of the notifiable infections of Salmonella enterica were reported as domestic and 48% were acquired abroad. The origin of the isolates used in generating AST results from Svebar are not known. Considering the low occurrence of Salmonella in food-producing animals in Sweden, the majority of food-related infections presumably have a foreign source. The high occurrence of resistance to fluoroquinolones in isolates from humans (21%) in comparison to the very rare occurrence of such resistance in isolates from Swedish food-producing animals suggests that most of these isolates from human infections do not have a domestic origin.
5.2.6 Campylobacter
Resistance to fluoroquinolones, tetracycline and erythromycin among faecal isolates of Campylobacter jejuni from humans was 56%, 30% and 0.8% respectively. From animals, 176 isolates of C. coli from healthy pigs were tested. The resistance found in C. coli from pigs against fluoroquinolones was 36% and one isolate was resistant to erythromycin.
Resistance to erythromycin, the drug of choice for treatment of human campylobacteriosis, is rare among isolates from humans as well as animals in Sweden. In animals, macrolide resistance has only been found in five isolates, two from Swedish broiler meat in 2013 and one isolate each from pigs in 2017, 2023 and 2025. The isolates have been genome sequenced and no transferable macrolide resistance gene has been found.
5.2.7 Clinical resistance in Escherichia coli from humans and animals
Comparison of resistance in bacteria from humans and different animal categories may indicate the magnitude of possible transfer of resistance between sectors and give insight into the drivers for resistance in the specific populations. However, in Swedres-Svarm, direct comparison of resistance is problematic as different interpretative criteria are used for bacteria from humans and animals. Data for bacteria from humans are interpreted with clinical breakpoints and presented as the proportion of isolates with clinical resistance. In contrast, data for bacteria from animals are mainly interpreted with epidemiological cut-off values (ECOFF) and presented as the proportion of isolates of non-wild type. For further information on interpretive criteria, see sections Guidance for readers and Chapter 6 - Materials and methods.
For the purpose of the comparison in this section, some data sets for E. coli from animals presented in Swedres-Svarm have been interpreted using clinical breakpoints for humans (Table 5.1).
Resistance was generally more common in E. coli from humans than in isolates from animals (Table 5.1). Notably, clinical resistance to fluoroquinolones or 3rd generation cephalosporins was considerably more common in E. coli from humans than in isolates from animals, with the highest occurrence in blood stream isolates from humans (Table 5.1). This is in line with the very low use of these antibiotic classes in animals (see Chapter 2 - Sales of antibiotics for animals). However, although few isolates of E. coli from animals show clinical resistance to fluoroquinolones, reduced susceptibility (i.e. non-wild type) is more common in some categories of diseased and healthy animals (see Chapter 4 - Antibiotic resistance in animals and previous reports). Possibly, the selection pressure from use of fluoroquinolones in animal populations is not sufficient to select for further mutations to clinical resistance in isolates with reduced susceptibility.
For the antibiotics commonly used in both animals and humans, e.g. ampicillin and trimethoprim, resistance is more frequent. In particular, the occurrence of resistance is high among clinical isolates from calves, pigs and humans (Table 5.1, Figure 5.1 and Figure 5.2). When comparing resistance to trimethoprim, it should be considered that for some categories (i.e. clinical isolates from animals and blood isolates from humans), trimethoprim-sulphamethoxazole was tested. This could possibly result in a lower occurrence of resistance than if susceptibility to only trimethoprim had been tested.
Occurrence of resistance to ampicillin or trimethoprim could also be due to co-selection by use of other antibiotics or to other factors selecting for resistance. For example, although exact data are missing, use of ampicillin or amoxicillin in cattle is believed to be low in Sweden. Nevertheless, resistance to ampicillin is common in both isolates from diseased calves and dairy cows. However, it is well known that multiresistant E. coli is common in pre-weaned dairy calves but that resistant strains are cleared as calves mature.
Moreover, the high occurrence of resistance to ampicillin or trimethoprim may be influenced in some categories by a low number of isolates and possible sampling bias, where animals are sampled due to therapeutic failures, inferring a selection of problematic cases.
| Category | Sample type | Year | Number of isolates | Amp (>8) | Cip (>0.5) | Ctx (>2) | Gen (>2) | Mer (>8) | Nit (>64) | Tmp (>4) |
|---|---|---|---|---|---|---|---|---|---|---|
| Cat (UTI) | Urinary | 2025 | 428 | 19 | 1.2(a) | 0.5 | 0 | 0 | 0 | – |
| Dog (UTI) | Urinary | 2025 | 846 | 15 | 0.9(a) | 0.5 | 1 | 0 | 0 | – |
| Horse (e.g., endometritis) | Genital tract | 2025 | 210 | 3 | 0(a) | 0.5 | 2 | 0 | – | 10.0(b) |
| Calf (enteritis) | Faeces/Post-mortem | 2021-22 | 46 | 76 | 0(a) | 0 | 0 | 0 | – | 30(b) |
| Dairy cow (mastitis) | Milk | 2025 | 49 | 20 | 0(a) | 2 | 0 | 0 | – | 12(b) |
| Laying hens | Post-mortem | 2024-25 | 79 | 13 | 1.3(a) | 0 | 1 | 0 | – | 0(b) |
| Pig (enteritis) | Faeces/Post-mortem | 2025 | 55 | 36 | 0(a) | 0 | 2 | 0 | – | 27(b) |
| Broiler (healthy) | Intestinal content | 2024 | 173 | 11 | 0.6 | 0 | 1 | 0 | – | 7.5 |
| Cattle under 1 year (healthy) | Intestinal content | 2020-21 | 56 | 2 | 0 | 0 | 0 | 0 | – | 1.8 |
| Horses (healthy) | Intestinal content | 2025 | 160 | 5 | 0 | 0 | 0 | 0 | – | 5 |
| Laying hens (healthy) | Intestinal content | 2022-23 | 86 | 1 | 0 | 0 | 2 | 0 | – | 0 |
| Pig (healthy) | Intestinal content | 2025 | 174 | 21 | 0 | 0 | 1 | 0 | – | 17 |
| Turkey (healthy) | Intestinal content | 2024 | 29 | 3 | 0 | 0 | 3 | 0 | – | 3.4 |
| Humans (UTI) | Urinary | 2025 | 220589 | 29 | 12 | 5.5(c) | – | – | 1 | 20 |
| (a)Enrofloxacin tested, BP >1mg/L; (b)Trimethoprim-sulphamethoxazole tested, BP >4 mg/L; (c)Cefadroxil tested, BP >16 mg/L, NordicAST v. 15.0 | ||||||||||