Antibiotics for dogs: benefits, risks and resistance

Reading time: about 10 min

When a veterinarian prescribes antibiotics for a dog, we naturally assume they are needed. But is the inflammation bacterial? Must treatment reach the whole body, or could a local treatment suffice? Antibiotics can save lives in serious infections. The aim is not to question necessary care, but to understand why a particular treatment is best for a particular dog.

These questions are not meant to challenge necessary veterinary care. They help us understand why a particular treatment is appropriate for this particular dog. In a serious infection an antibiotic may save a life; in a mild, localized problem avoiding unnecessary systemic exposure can matter just as much.

Antibiotics are valuable, but are they always needed?

Antibiotics act against bacteria, not viruses or yeast. Redness, itching and pain can also arise from allergy, injury or irritation. Serious or spreading bacterial infections may require prompt systemic treatment. The question is whether the medicine, its spectrum and its route are appropriate.

The word ‘inflammation’ does not itself establish a bacterial infection. Allergy, a foreign body or trauma may cause redness and pain. If an infection is severe, spreading into deeper tissues or affecting the dog’s general condition, delaying an indicated antibiotic is not sensible. The question is whether the drug is indicated here, whether the active ingredient is appropriate and whether the same result can be achieved with less avoidable risk.

Topical versus systemic treatment

Topical medication is applied directly to skin or ears. Oral or injected antibiotics circulate through the body. Local treatment may achieve high concentrations at the affected site with less systemic exposure, but suitability depends on the site, depth, agent and, for ears, eardrum integrity. The 2025 ISCAID canine pyoderma guidance favors topical antimicrobial treatment for many superficial bacterial skin infections and reserves systemic antibiotics for deep disease or when topical treatment is impractical or unsuccessful. In some cases a topical antiseptic, such as an appropriate chlorhexidine preparation, is preferable to a topical antibiotic. These recommendations are not permission to apply household antiseptics indiscriminately.

Local treatment is not automatically weaker treatment. It can deliver a high concentration at the affected site without exposing the whole body to the same extent. Conversely, topical products are not risk-free, and the state of the eardrum matters in ear disease. The 2025 guidance also distinguishes a topical antibiotic from a topical antiseptic: for many superficial bacterial skin infections, an appropriate antiseptic such as chlorhexidine may be preferred. This guidance concerns the skin conditions studied; it is not permission to put home antiseptic preparations into a dog’s ear or apply them indiscriminately to damaged skin.

Ossu’s ear infection

Ossu developed his first outer ear infection. Cytology showed bacteria and yeast. The veterinarian prescribed ear treatment and proposed adding an oral antibiotic immediately. I asked whether we could first assess the response to topical treatment over two or three days. The veterinarian agreed. This was not a request to leave infection untreated. Yeast does not respond to antibiotics, and a bacterial finding alone does not always establish a need for systemic medication. Oral treatment can be needed for deeper or middle-ear disease. Ossu’s experience does not prove the additional antibiotic would have been unnecessary; it shows why asking for the reason is worthwhile.

This was Ossu’s first episode, not a history of repeatedly recurring ear infections. Cytology found both bacteria and yeast. The veterinarian agreed that we could assess the response to the prescribed ear treatment for two or three days before deciding whether to add a systemic antibiotic. The important distinction is between questioning an additional medicine and refusing to treat an infection. Cytology helps guide treatment, but yeast does not respond to antibiotics, and the presence of bacteria does not automatically prove that oral treatment is required. The experience cannot establish that the extra drug would certainly have been unnecessary.

What happens to the gut microbiome?

Antibiotics affect susceptible bacteria, including members of the intestinal microbiome. Effects depend on the agent, dose, duration, diet and individual dog. Recovery has several meanings: microbial diversity, community composition, the abundance of particular groups and metabolic function may recover at different rates.

In a 2020 study of healthy dogs receiving metronidazole for 14 days, not all measured changes had resolved four weeks after treatment. The median relative abundance of Fusobacteria fell from 14.5% before treatment to 0.6% during treatment and remained at 1.8% four weeks afterward, although other groups approached baseline. Some metabolic measures also remained altered. The appropriate conclusion is that certain effects persisted at least four weeks, not that everything had recovered.

Microbiome ‘recovery’ is not one measurement. Diversity describes how many bacterial groups are present and how evenly they are distributed. Composition describes which groups are present; relative abundance describes the proportion represented by each group. Researchers can also measure metabolic activity. Normalization of one measure does not establish recovery of all the others. In the 2020 study, Fusobacteria’s median relative abundance was 14.5% before treatment, 0.6% during metronidazole and still only 1.8% four weeks after treatment. Several other major groups had moved closer to baseline, while some metabolic indicators remained altered. The supported conclusion is that some measured effects lasted at least four weeks, not that the whole microbiome had recovered by then.

How long does recovery take?

A 2024 study followed 22 dogs after metronidazole for up to six months. Overall bacterial diversity recovered relatively quickly, while community structure became similar to baseline after approximately four to six weeks. All dogs remaining in the study approached their measured baseline composition during follow-up. The 2020 and 2024 findings are not necessarily contradictory: they measured different outcomes and involved different dogs and treatment schedules. The 2020 study involved healthy dogs given metronidazole for 14 days; the 2024 study involved dogs with gastrointestinal signs and subsequent giardiasis, treated for seven days. In the latter study, Shannon diversity was reduced up to three days after treatment, while the broader bacterial community structure took roughly four to six weeks to resemble baseline. Recovery of overall diversity or relative composition does not prove that every individual bacterial group, its absolute abundance or every metabolic function has recovered. A cat study reported changes associated with clindamycin as late as 603 days after treatment, but that result cannot be directly applied to dogs. Nor does six-month follow-up rule out every possible longer-lasting change.

The 2024 study followed 22 dogs for as long as six months. Its dogs had gastrointestinal signs and were subsequently diagnosed with giardiasis; they received seven days of metronidazole, whereas the 2020 study used 14 days in healthy dogs. Shannon diversity was reduced up to three days after treatment, and community structure resembled baseline after roughly four to six weeks. All dogs remaining under observation eventually approached their measured baseline composition within six months. That does not rule out longer changes in individual strains, absolute bacterial numbers or functions that were not measured. In cats treated with clindamycin, Whittemore and colleagues reported changes in individual bacterial groups and metabolites more than 600 days after stopping treatment. The feline result cannot be directly generalized to dogs. Asking ‘which part of the microbiome, after which drug, measured how?’ is more useful than assuming a universal recovery time.

Do probiotics eliminate antibiotic risks?

No. Some probiotics or synbiotics may reduce digestive side effects in particular circumstances, but they do not automatically restore the entire pre-treatment microbial community. Effects depend on strain, product and context. Supportive treatment can be discussed with a veterinarian; it does not justify an unnecessary antibiotic.

Clinical relief and restoration of the original microbiome are not identical outcomes. A probiotic or synbiotic’s effects depend on the strain, formulation and circumstances. Nor should every supposedly natural microbiome-supporting product be assumed safe or effective. Supportive care can be discussed with the veterinarian when antibiotics are needed; it cannot turn an unnecessary course into an appropriate one.

Why do infections return?

Recurrent ear problems may be driven by allergy, anatomy, moisture or other persistent factors. Treating microbial overgrowth without addressing the cause can lead to recurrence. Both conventional and integrative approaches should be assessed by evidence, safety and the dog’s actual response, rather than by labels.

A first episode and a problem returning every few months are not the same situation. Recurrent ear inflammation may be associated with allergies, ear anatomy, moisture or another persistent factor. Conventional and integrative veterinary medicine meet on an important question: what caused the condition, and what keeps it going? Proposed explanations and interventions can differ, and their value should be judged by evidence, safety and the dog’s actual response, not by whether a method is called ‘natural’ or ‘conventional’.

Empirical treatment need not mean blind treatment

Sometimes treatment must begin before culture results are available, especially if disease is serious or rapidly worsening. There should still be a plan for reassessment. Ossu’s forepaw problem initially raised concern about bone cancer. Antibiotics were started empirically, and he improved within days, but he was also receiving pain relief, so the improvement cannot be attributed solely to antibiotics. I sent the veterinarian videos and updates; decisions were made in stages. Staphylococcus was identified later. This story does not support stopping antibiotics early without veterinary advice. Duration should be guided by diagnosis, response and new information.

Empirical treatment should include a plan for what to monitor, when to reassess and what new information would justify changing the initial decision. In Ossu’s forepaw case, the early improvement could not be attributed solely to the antibiotic because he was also receiving pain medication. I sent the veterinarian videos and updates, and we made continuation decisions step by step. Staphylococcus was identified later. This is not an argument for owners to stop a prescribed course early or to assume a long course is always wrong. Treatment duration should be linked to diagnosis, clinical response and new findings. Our earlier FendaF article on antibiotic choice and susceptibility testing discusses that process in more detail.

Bacteria become resistant, not the dog

Resistance occurs when bacteria possess or acquire traits allowing survival despite an antibiotic. Treatment can favor resistant organisms by removing susceptible competitors. Not every course causes a dangerous resistant infection, but unnecessary, repeated or excessively broad treatment increases selection pressure. A 2026 metagenomic study identified 217 distinct resistance genes across samples from 273 dogs, with a median of 24.5 per dog. This does not mean the dogs were ill or that every gene causes clinical resistance. It shows that resistance genes are part of microbial ecology and may sometimes spread between bacteria. The One Health perspective considers people, animals and shared environments. Appropriate treatment protects both individual dogs and broader public health.

Some bacteria already carry traits or genes that help them survive a particular antibiotic before treatment begins. By removing susceptible competitors, treatment can select for resistant organisms. This is selection pressure, not proof that every course creates a dangerous resistant infection. Unnecessary courses, unjustifiably broad-spectrum drugs and repeated treatment increase opportunities for selection. The 2026 study found 217 different resistance genes across samples from 273 dogs, with a median of 24.5 genes per dog. The presence of a gene does not mean the dog is sick or that every gene produces clinically important resistance. Some genes may be carried on mobile genetic elements, making transfer between bacteria possible. One Health recognizes that people, dogs and shared environments are connected; it does not mean a family dog should be treated as a danger.

When is culture especially useful?

Culture and susceptibility testing are particularly valuable in recurrent, deep, unusual or poorly responding infections, or after failed empirical treatment. They are not mandatory for every first superficial infection, but can prevent repeated blind trials of different drugs.

Culture is particularly worth considering for recurrent, deep, unusual or poorly responding infections and after unsuccessful empirical treatment. It is not necessary for every first superficial infection, but when treatment keeps failing, susceptibility testing can prevent one blind antibiotic trial after another.

What alternatives really exist?

For some superficial skin infections, topical antiseptics have good evidence. Some ear infections respond to suitable local treatment addressing both bacteria and yeast. Recurrent disease requires attention to underlying causes. Integrative supportive products should be evaluated individually for dog-specific evidence, concentration and safety. Do not put homemade oils, acids or antiseptics in ears, particularly when the eardrum may be damaged.

An alternative does not have to be a herb. Evidence-based options can include topical antiseptics for certain superficial skin infections, suitable local ear treatment that addresses bacteria and yeast, and finding and managing the underlying causes of recurrence. Local and systemic treatments can also be combined when clinically indicated. Integrative supportive products must each be assessed for dog-specific evidence, the condition being treated, concentration and safety. In particular, homemade oils, acids and antiseptics should never be assumed safe for ears, especially if the eardrum is damaged.

Questions to ask your veterinarian

What indicates bacterial infection? Is immediate antibiotic treatment necessary? Could the problem be treated locally? What additional benefit does a systemic antibiotic provide? Which side effects should be monitored, and when should we reassess? Do not delay care independently if your dog is lethargic, feverish, in severe pain or has a rapidly spreading or deep infection.

Ask what specifically indicates bacterial infection; whether treatment must begin immediately or further assessment is possible; whether the problem can be treated at the site; what additional benefit a systemic drug would provide; which adverse effects to watch for; and when to report back. If your dog is lethargic, has a high fever or severe pain, or the infection is rapidly spreading, deep or systemic, do not postpone treatment on your own. The aim is informed cooperation with the veterinarian, not delay.

Conclusion

Antibiotics are valuable and sometimes lifesaving. The best treatment is not necessarily the most extensive one, but the treatment that resolves the individual dog’s problem effectively, appropriately and with monitoring.

Sources and further reading

• ISCAID guidance (2025): canine superficial pyoderma
• ISCAID antimicrobial treatment review
• Pilla et al. (2020): Metronidazole and the canine microbiome
• Marshall-Jones et al. (2024): Canine microbiome recovery
• 2026 canine antimicrobial resistance study
• Review: antimicrobial resistance in companion animals
• Judy Morgan: Is Your Veterinarian Over-Prescribing Antibiotics?
• Judy Morgan: Itchy, Smelly Ears
• FendaF: Antibiootikumid ja mida neist teadma peaks?

• Merck Veterinary Manual: Otitis Externa in Animals
• Whittemore et al. (2018): Clindamycin and feline microbiome, 603-day follow-up (PMID 30038854)

---

---