Should we prescribe oral metronidazole or probiotics for acute gastroenteritis in dogs? A

PICO question 
In dogs with acute gastroenteritis, is metronidazole faster, slower, or comparable in resolving clinical signs when compared to probiotic administration? 
  
Clinical bottom line 
Category of research question 
Treatment 
The number and type of study designs reviewed 
Five studies total, all were blinded, randomised controlled trials 
Strength of evidence 
Moderate 
Outcomes reported 
The use of probiotics as a treatment for acute, uncomplicated diarrhoea in dogs may improve clinical signs faster when compared to a placebo, but showed no difference when compared directly to metronidazole. Metronidazole, when compared to a placebo, produced mixed results with one study finding that treatment with metronidazole did significantly reduce the time to resolution of diarrhoea, while another study found the difference with placebo was not significant 
Conclusion 
Based on the evidence evaluated, the use of oral metronidazole will not decrease time to resolution of clinical signs in cases of acute, uncomplicated diarrhoea in dogs when compared to probiotic administration and thus should not be a first-line treatment in such cases 
  
How to apply this evidence in practice 
The application of evidence into practice should take into account multiple factors, not limited to: individual clinical expertise, patient’s circumstances and owners’ values, country, location or clinic where you work, the individual case in front of you, the availability of therapies and resources. 
Knowledge Summaries are a resource to help reinforce or inform decision making. They do not override the responsibility or judgement of the practitioner to do what is best for the animal in their care. 
  



Summary of the evidence 1. Shmalberg et al. (2019)
Population: Client-owned dogs that presented for acute diarrhoea (≤7 days) with or without vomiting, between 4-45 kg, and lacking any clinically relevant comorbidities (endocrinopathies, organ dysfunction, immune-mediated disease, suspected pancreatitis). Excluded were dogs with signs consistent with severe acute haemorrhagic diarrhoea syndrome (significant dehydration, hypovolemia, large volumes of haemtochezia). c) Dogs 10.1-20 kg received 250 mg; 12.5-25 mg/kg dose range. d) Dogs 20.1-45 kg received 400 mg; 8.9-20 mg/kg dose range. e) Typical dose range for metronidazole in cases of Giardiasis is 25 mg/kg and clostridial infections is 10-15 mg/kg when given orally. 3. Placebo group: capsules of sucrose equal in volume to probiotic and metronidazole capsules: a) Oral capsules given twice daily for 10 days. • Rescue treatment (treatment given to dogs that still had diarrhoea after 10 days or experienced worsening frequency of diarrhoea, worsening faecal score, increased haematochezia, or worsening straining) for any dog was tylosin, 30 mg/kg by mouth twice daily for 10 days. • Three dogs were eliminated from the study, one due to significant parasite burden (metronidazole group) and two due to failure of owners to give the treatment (one in the probiotic group and one in the placebo group). • A predetermined sample size of n = 20 per group was determined by power analysis prior to study. • Randomisation was performed using a schedule obtained from a random sorting feature of a commercial software program (Excel for Mac).
Study design: Double blinded, placebo controlled, randomised clinical trial.
Outcome studied: Outcome = number of days until first formed stool, as measured by a faecal score ≤3.

Main findings: (relevant to PICO question):
• Treatment group did not significantly affect the duration of diarrhoea when all dogs were included. Neither metronidazole treatment nor probiotic treatment significantly reduced the number of days until the first formed stool was appreciated in the dogs. 1. Probiotic group: Acceptable faecal score after 3.5 ±2 days (mean value with standard deviation). 2. Metronidazole group: Acceptable faecal score after 4.6 ±4 days (mean value with standard deviation). 3. Placebo group: Acceptable faecal score after 4.8 ±9 days (mean value with standard deviation). • P-value measured between the placebo group and the two treatment groups was p = 0.17. • After dogs positive for parasite ova on faecal float were removed, there was no significant difference between this finding and the findings when parasite ova positive dogs were included in the statistics (p = 0. Limitations: • The outcome was only one formed stool, which could be a sign of resolution, but could also represent an anomaly during treatment. • While complete blood count (CBC), chemistry panel, faecal floatation, and polymerase chain reaction (PCR) panels were done on most dogs, some dogs received empirical treatment (fenbendazole) or supportive care (maropitant, IV or SQ fluids) prior to study enrolment. • Owners received financial compensation for participating. • Dogs with a low parasite burden were included in the study (dogs with a high parasite burden were excluded). • The probiotic also contained prebiotics that were not found in the placebo and could have had an effect on the microbiome. • Metronidazole is typically given in the form of a tablet, capsule, or oral solution when used in a non-hospital setting, so the bioavailability of the powder form is unknown in this study. o Rescue treatment (metronidazole) was given if diarrhoea persisted after 7 days. • Any dog that presented with vomiting could receive one dose of maropitant 1 mg/kg SC (clinician's discretion). • Hematochezia (p = 0.48), fluid therapy treatment (p = 0.48), maropitant citrate treatment (p = 0.72) were not different between the two populations of dogs. • A predetermined sample size of n = 15 per group was determined by power analysis prior to study. • Dogs were randomised using a computer-generated log created by the pharmacist. The dog owner and clinician were blinded to treatment. • 14 dogs were excluded from the study. 10 dogs were excluded for gastrointestinal parasitism, two dogs for inability to obtain sufficient faeces for faecal testing, one dog for normal faecal score prior to therapeutic initiation, and one for ultrasonographic evidence of acute pancreatitis. • Three dogs were removed from the study. One dog for severe vomiting (test population), one dog for failure to provide faecal scores and return at day 7 for diagnostics, and one for failure of faecal scoring and adherence of dietary guidelines.

Langlois et al. (2020)
Study design: Double blinded placebo controlled randomised clinical trial.
Outcome studied: Outcome = time in days until resolution of diarrhoea, measured as two consecutive faecal scores ≤4.

Main findings: (relevant to PICO question):
• Metronidazole significantly reduced the number of days until resolution of diarrhoea (p = 0.04) in dogs when compared to the placebo group. o 12 in placebo group; 10 in probiotic group (no statistical analysis provided). • Doses for both placebo and probiotic paste was based on weight: o 1 mL for 1-10 kg; 2 mL for 10-25 kg; 3 mL for 25-50 kg. o Given by the owner by mouth three times daily (every 8 hours) until normalisation of stools. o First dose was a double dose in both placebo and probiotic group. • Randomisation protocol was not described, only stated that there was block randomisation. • Calculation of a predetermined group size (n) was not performed prior to study enrolment. • No adverse effects in either group were reported. Study design: Double-blind placebo control randomised clinical trial.
Outcome studied: Outcome = day (whole or half) until last abnormal stool and days until normalisation of stool.

Main findings: (relevant to PICO question):
• Probiotic treatment was a significant factor for time to diarrhoea resolution (p = 0.045) but was not significant for time until resolution of signs if diarrhoea and vomiting were both present (p = 0.55). The time to first normal stool after starting treatment was not significant between the two groups (p = 0.14).

Limitations:
• Faecal samples for bacterial and parasitic analysis were only collected from some patients, not all, and only at the initial enrolment in the clinical trial. • There was no standardised way to assess normalisation of stool consistency, only done by owner opinion and relied upon owner's recollection during interview on day 4 or 8. • No clearly stated definition of 'normalised stool', normalisation of stool was defined by owner's interpretation and was only subjective. • Small sample size. • All of the products used (probiotics and placebo paste) were supplied by the manufacturing company directly.

Nixon et al. (2019)
Population: Client owned dogs with acute diarrhoea ≥1 occurrences within 24 hours of presentation to primary care veterinarian. Dogs were excluded if their signs were deemed unsuitable for conservative management, if they received antimicrobial or probiotic treatment within 4 weeks before day 0 of enrolment. Dogs with comorbidities that did not affect diarrhoea were included in the study. Outcome studied: Outcome definition: time until passage of three consecutive faeces of normal consistency (faecal score ≤3) measured from initial presentation to the veterinary clinic within 10 days of initial study enrolment.

Main findings: (relevant to PICO question):
• Probiotic treatment was significant (p = 0.008) in decreasing time to resolution of diarrhoea as measured as the passage of three consecutive normal stools (faecal score ≤3). • No diagnostic testing to rule out infectious causes of diarrhoea or unknown comorbidities, which could impact the responsiveness of the diarrhoea to a probiotic alone. • While the Mann-Whitney test was used to determine differences in median age between the group was not significantly different, the value was very close to P and may confer a clinical difference. The median age of the treatment group was 45 months, while the median age of the placebo group was 24 months. • Dogs were allowed to be removed from the trial if they needed additional medical intervention; more were removed from the placebo group and could represent attrition bias. • Dogs that missed one dose in a day in either the placebo or treatment group were not excluded, which may impact results given the timing of resolution of diarrhoea was measured in hours and was around four doses for the treatment group (specific numbers were not given). This lack of strict compliance may mimic the realities of daily life (in respect to client compliance) therefore, it is not unreasonable to still include these dogs. • The placebo had a different composition than the probiotic; it would be a stronger study if the placebo was the same composition as the probiotic without the probiotics and prebiotics. • Two authors were employees of the company that produced the probiotic. Outcome studied:
• Resolution of diarrhoea was defined as stool scores that improved from 4 to ≤2 and remained ≤2 for at least 5 consecutive days.

Main findings: (relevant to PICO question):
• Treatment with probiotic significantly reduced days to resolution of diarrhoea when compared to the placebo group in all dogs, and in both German Shepherds and Labrador Retrievers, while the Labrador-Golden Retrievers or Golden Retrievers did not show a significant difference due to small sample size. • 5/13 dogs in the treatment group received metronidazole: o Mean days of resolution of diarrhoea in probiotic dogs that did not receive metronidazole: 3.33 ±26 days (based on reviewer's calculation). • 9/18 dogs in the control group received metronidazole: o Mean days to resolution of diarrhoea in placebo dogs that did not receive metronidazole: 6.11 ±88 days (based on reviewer's calculation). • When dogs receiving metronidazole were excluded from the study, resolution of diarrhoea by day 4 occurred in 3/9 dogs in the control group and 7/9 dogs in the probiotic groupstatistical analysis was not presented.

Limitations:
• Funded by Procter and Gamble Pet Care, which produced the probiotic used in the treatment group. • All dogs were from the same facility, which could be a confounder (a bias in management or population). • Dogs receiving metronidazole were included in the statistics and were more commonly in the control group. • Giardia positive dogs were included in the statistics and received metronidazole (standard of care for giardiasis); these dogs should have been eliminated from the study as the cause is infectious and not idiopathic. • Small samples sizes when considering the data with and without the metronidazole treated dogs. • Very small sample size in some subgroups. . The continued prescribing of metronidazole may be related to a perception that metronidazole administration resolves a patient's diarrhoea in a more timely manner and satisfies an owner's expectation of medication administration. However, there is also a rising concern about veterinary use of antimicrobials and the effects on human health (Herstad et al., 2010;and Prescott, 2019).

Appraisal, application and reflection
The use of antimicrobials in all fields of veterinary medicine can have effects on the microbial population, which in turn can impact human health. Research in companion animals shows the close relationship between humans and their pets allows for the transfer of microbes between the species, thus increasing the potential for transfer of methicillin-resistant bacteria between the human and pet. With increased antimicrobial resistance, the treatment of infections may become more challenging in both human and veterinary medical fields (Lloyd & Page, 2018). Thus an alternative treatment regimen needs to be established for uncomplicated acute diarrhoea in the veterinary population that both alleviates the patient's condition and placates the client. The practice of prescribing probiotics for treatment of acute diarrhoea in veterinary patients may offer an alternative solution.
The direct study of the time to resolution of clinical signs (diarrhoea) using metronidazole compared to probiotics has not been studied adequately in dogs that present for acute, uncomplicated diarrhoea. Only one study directly compares time to resolution of diarrhoea in dogs treated with metronidazole or probiotics, while the remaining data must be extrapolated from studies comparing either oral metronidazole use to a placebo, or probiotic use to a placebo.
Shmalberg et al.'s (2019) study directly comparing treatment with metronidazole, a probiotic product, and a placebo, found there was no significant reduction in time to resolution of acute diarrhoea when the dog received a probiotic or metronidazole. The metronidazole-treated group had a very similar average to the placebotreated group, but probiotic average days until resolution of diarrhoea was 1 day shorter than both, which may be clinically important for a client.
Langlois et al. (2020) found that metronidazole treatment for cases of acute, uncomplicated diarrhoea did have a significant impact on time to resolution of diarrhoea when compared to a placebo. Metronidazole treatment was found to resolve clinical diarrhoea 1.5 days sooner than the placebo. A limitation of the study is that the participants' diet was not controlled, which is also a mainstay of common treatment (Singleton et al., 2019). The sample sizes were small for this study as well.  2014) found that oral metronidazole treatment for cases of acute diarrhoea led to changes in the gastrointestinal microbiome composition, particularly bacteria responsible for bile acid metabolism, that persisted for up to 1 month post antibacterial treatment in some dogs. During oral metronidazole treatment, the faecal dysbiosis index was increased, but returned to normal after 2 weeks of metronidazole treatment (Pilla et al., 2020;and Nogueira et al., 2019). In contrast, probiotic use generally appears to be safe in veterinary species. The potential risk of probiotic use must be extrapolated from human research. The greatest apparent risk is the potential of a bacterium to become pathogenic, particularly if the intestinal barrier is not fully developed, but appeared safe when used in patients with a history of immunosuppression (either iatrogenic or infectious in origins) (Butel, 2014). Additionally, natural bacterial behaviours, such as genetic material transfer, can occur and could be detrimental if the probiotic bacterium conveyed antimicrobial resistant genetic material to the normal flora, though evidence of this occurring is lacking (Butel, 2014).
Further research needs to be performed in order to clinically answer this PICO question. Shmalberg et al. (2019) is the only study that directly compares the use of metronidazole and a probiotic product and did not show a significant advantage of either treatment compared to each other or a placebo. There is insufficient evidence to answer whether or not metronidazole affects the duration (shorter, or longer, or no difference) of clinical signs when compared to probiotic administration in cases of acute, uncomplicated diarrhoea.
Additionally, when evaluating the use of probiotics / synbiotic products as treatment to reduce the duration of acute, uncomplicated diarrhoea, the evidence remains weak.  diarrhoea, but only in certain dog breeds. Given the clinical scenario of an otherwise healthy dog presenting for acute and uncomplicated diarrhoea, prescribing probiotic / symbiotic products to satiate owner expectations (in insistence) would be a reasonable choice as the risk associated with these products is low and no long-term effects have been reported or studied.

Methodology Search strategy
Databases searched and dates covered: PubMed on NCBI Platform, 1999-2021. CAB Abstracts on OVID Platform, 1999-2021.
Search terms: PubMed: Acute AND (diarrhea OR diarrhoea OR gastroenteritis) AND (canine OR dog OR dogs) AND (metronidazole OR probiotic OR probiotics OR nutraceutical) CAB Abstracts: (Acute and (diarrhea or diarrhoea or gastroenteritis) and (dog or dogs or canine or canines) and (metronidazole or probiotic or probiotics or nutraceutical)) Dates searches performed: 19 Dec 2021

Exclusion / Inclusion criteria
Exclusion: Articles not in English, review articles, articles that did not measure the desired outcome, articles that studied diarrhoea that was not acute, or articles that studied diarrhoea of a known cause.
Inclusion: Articles available in English which were relevant to the PICO. Articles had to involve more than one dog and compare the PICO pharmacologics to a placebo or each other. Total relevant papers when duplicates removed