Acute diarrhoea is a common complaint seen in first-opinion small animal veterinary practice. Whilst it is often self-limiting and likely to resolve with symptomatic treatment alone, antibiotics are frequently prescribed as part of the therapeutic management plan. In part one of this article (Veterinary Practice, June 2020), we looked at the common causes of acute diarrhoea and considered the use of faecal analysis. In part two, we will explore when and why antibiotics are used in the management of acute diarrhoea, whether they are truly indicated and the potential adverse effects of antibiotic usage.
When and why are antibiotics prescribed?
Ideally antibiotics should only be prescribed to treat a known bacterial infection or septicaemia. However, this is not always the case; it was found that 71 percent of canine cases presenting with diarrhoea in first-opinion practice were given antibiotics (German et al., 2010). This statistic has reduced according to a more recent study, in which 49.7 percent of dogs were prescribed antibiotics when initially presented at the vets with diarrhoea (Singleton et al., 2019). Despite the fact that almost half of the dogs in the latter study were dispensed antibiotics, faecal bacteriology/parasitology was performed in only 3.2 percent of these cases. This suggests a large discrepancy between the administration of antibiotics and prior demonstration of an underlying bacterial infection. Antibiotic prescription was associated with pyrexia in both studies, and with haemorrhagic diarrhoea in the more recent study. One could therefore argue that prescription, at least in some cases, may have been due to concern regarding the development of septicaemia caused by bacterial translocation across a damaged intestinal epithelium. However, it is also of note that 88 percent and 37.5 percent of normothermic animals were prescribed antibiotics in each study respectively, making concern over sepsis a less likely justification for antibiotic use in these cases.
In certain cases, antibiotics may be prescribed due to concern over zoonotic disease, particularly in the knowledge that children or immunocompromised people are in contact with the animal. In other cases, vets may use antibiotics in the hope that they will accelerate the resolution of the diarrhoea. A recent study of 31 dogs reported that dogs given metronidazole had a shorter time to resolution of diarrhoea compared to control dogs receiving a placebo (Langlois et al., 2019). However, another study of 60 dogs reported that whilst dogs given metronidazole recovered slightly faster than those given a placebo, the group given probiotics recovered fastest (Shmalberg et al., 2019). Several further studies have reported the use of various probiotic formulations to accelerate the resolution of acute diarrhoea compared to a placebo (Kelley et al., 2009; Herstad et al., 2010; Ziese et al., 2018; Nixon et al., 2019). Given the widespread concern over growing antimicrobial resistance, nutraceuticals may be a more appropriate choice for first-line management of acute diarrhoea compared to antibiotics (Langlois et al., 2019).
Finally, it has been shown that medications may be prescribed due to owner expectation or pressure. In human medicine, doctors have been reported to be under pressure to prescribe a material management plan (Little et al., 2004; Lewis and Tully, 2011). The same appears to be the case within veterinary medicine, when antibiotics may be selected based on owner compliance and willingness to pay, rather than suitability for the bacterial infection in question (Mateus et al., 2014).
Considerations when prescribing antibiotics
Are antibiotics required following detection of bacteria in the faeces?
Irrespective of the pathogen isolated, the empirical use of antibiotics is not recommended in cases of uncomplicated acute diarrhoea. Antibiotics have been shown to increase adverse effects and prolong Salmonella shedding in the stools of people with non-severe diarrhoea attributed to Salmonella (Sirinavin and Garner, 2000). In people with Campylobacter enteritis, antibiotics are reserved for those with a high fever or haemorrhagic diarrhoea; who are passing more than eight stools per day; who have had symptoms for more than one week or have worsening symptoms; or who are immunocompromised (Acheson and Allow, 2001; Blaser and Allos, 2005). Antibiotic use for the treatment of diarrhoea thought to be associated with E. coli infections is controversial due to the high levels of intrinsic antimicrobial resistance and the risk that antibiotic use may increase toxin release and worsen the disease process (Panos et al. 2006).
Antibiotic use should be considered only following detection of the bacteria, or its toxin, in patients with acute, severe disease (eg haemorrhagic gastroenteritis), or where there is concern regarding the development of sepsis. In these cases, antimicrobial selection should ideally follow in vitro sensitivity testing (Hall, 2009; Weese, 2011).
Are antibiotics actually indicated if there is concern over potential bacterial translocation?
Dysbiosis, immunosuppression or damage to the gastrointestinal epithelium may contribute to the translocation of viable bacteria from the lumen of the gut to extra-intestinal sites (Brenchley and Douek, 2012). These bacteria may originate from the commensal population or may have been introduced into the gut via ingestion. If the bacteria reach the blood, they could cause disseminated infections or even life-threatening septicaemia or sepsis. Therefore, antibiotics may be given to individuals that are at a higher risk of translocation of potentially pathogenic bacteria, such as animals with acute haemorrhagic diarrhoea where the gastrointestinal barrier may be compromised (Armstrong, 2013). Further recommendations to try to limit bacterial translocation include restoring intestinal perfusion, commencing early feeding, maintaining a normal gastrointestinal microbiota and supporting the gut barrier function (Krentz and Allen, 2017).
However, it has also been suggested that antibiotic use may not be the solution in the face of potential bacterial translocation and could even worsen the situation. Research in human medicine has reported that antibiotic use can induce translocation of commensal bacteria, increase inflammatory responses and increase the release of Shiga-like toxin by E. coli O157 (Panos et al., 2006; Knoop et al., 2016). It has been proposed that anaerobic bacteria may play a role in preventing bacterial translocation and that the use of antibiotics which selectively kill anaerobes may therefore increase risk of bacterial translocation; mice that received metronidazole experienced increased dissemination of intestinal bacteria to their mesenteric lymph nodes compared to control animals (Wells et al., 1987).
A prospective study looking at bacteraemia in canine patients with acute haemorrhagic diarrhoea syndrome (AHDS) found no significant difference in the incidence of bacteraemia between dogs with AHDS and control dogs. Nor were there significant differences in the clinical signs, laboratory parameters, duration of hospitalisation or mortality between blood culture positive and blood culture negative affected dogs (Unterer et al., 2015). The use of antibiotics (amoxicillin/clavulanic acid) in dogs with AHDS and a positive blood culture showed no benefit; however, dogs with potential signs of sepsis were excluded from this study (Unterer et al., 2011). Another study found that dogs with AHDS and a left shift neutrophilia recovered well in the absence of antimicrobial therapy (Mortier et al., 2015). By studying cultures of mesenteric lymph nodes, bacterial translocation was even demonstrated in 52 percent of healthy dogs undergoing routine ovariohysterectomy (Dahlinger et al., 1997). All of these studies demonstrate that bacteraemia is not necessarily associated with clinical disease, questioning the use of antibiotics even if bacteraemia may be present.
The term “bacteraemia” refers to the presence of bacteria within the bloodstream and is confirmed by repeat isolation of a pathogen from culture of the patient’s blood (Nostrandt, 1990). Whilst bacteraemia is not necessarily associated with clinical disease, the potential for development of septicaemia or sepsis is a valid concern. Septicaemia and sepsis are respectively defined as “systemic disease caused by the spread of microorganisms and their toxins via the circulating blood” and “life-threatening organ dysfunction caused by a dysregulated host response to infection” (Odeh, 1996; Singer et al., 2016). If septicaemia or sepsis is even suspected, due to deteriorating systemic clinical signs, haematological changes and/or positive blood culture, then appropriate parenteral antibiotics should be started immediately.
Despite the lack of studies demonstrating that haemorrhagic diarrhoea is associated with an increased risk of bacterial translocation or sepsis, some clinicians will still elect to use antibiotics in patients with acute haemorrhagic diarrhoea (Armstrong, 2013). This is likely due to the potential severity of septicaemia should it occur, and the difficulty involved in confirming bacterial infection as the cause of diarrhoea.
Are antibiotics indicated if zoonotic pathogens are detected in the stools?
This remains a difficult topic, with many factors to consider. As previously discussed, it is reasonable to use antibiotics, if indicated, to manage the clinical signs in an affected animal. However, once these signs have resolved there is a lack of evidence to support the continued use of antibiotics until multiple consecutive faecal samples are clear (Hall, 2009). There may be a reasonable justification to use antibiotics following detection of zoonotic pathogens if the owner is also affected, or if the animal shares its home with children or immunocompromised people. However, use of antibiotics in healthy individuals can promote the establishment of a carrier state and may favour development of antimicrobial resistance (Weese, 2011; Werner et al., 2020). In these cases, it is always vital to enhance infection control, for example by improving hygiene and barrier nursing protocols.
Antibiotic use can be detrimental at a patient and/or population level
Antibiotic use can disrupt the normal microbiota of an individual
Antibiotics used to treat a specific infection can simultaneously affect bacterial species within the normal gastrointestinal microbiota, resulting in a microbial imbalance or “dysbiosis”. The normal microbiota balance will usually recover once the antibiotics are stopped; however, a prolonged dysbiosis may occur in some individuals which could result in persistent signs of gastrointestinal upset (Canine Microbiota Dysbiosis Index, 2020). Administration of metronidazole to healthy dogs has been reported to decrease the numbers of specific groups of bacteria that are thought to be important for gut health; four weeks after cessation of a two-week metronidazole course, some of these changes were still present (Olson et al., 2015). Ciprofloxacin has also been shown to reduce taxonomic richness and diversity of gut bacterial populations in people (Dethlefsen et al., 2008). An individual’s microbiota appears to be intimately involved in many normal body functions, therefore its disruption through use of antibiotics should always be considered and requires reasonable justification.
Antibiotic use can cause adverse effects for the patient
Antibiotics can cause acute adverse effects in some patients, with gastrointestinal upset commonly reported in people, dogs and cats (Noli and Booth, 1999; McFarland, 2008; Pallo-Zimmerman et al., 2011; Schulz et al., 2011; Olson et al., 2015; Whittemore et al., 2019). Metronidazole, which is frequently used in the management of acute diarrhoea, may induce neurotoxicity even when given at doses that are less than the current licensed dose of 25mg/kg BID (Tauro et al., 2018). Furthermore, use of antibiotics early in life in people has been associated with development of various diseases in later life, including obesity, Crohn’s disease, asthma, atopy and even behavioural disorders (Wickens et al., 1999; Johnson et al., 2005; Hvid et al., 2011; Transade et al., 2013; Bailey et al., 2014; Ahmadizar et al., 2017; Slykerman et al., 2017). More studies are required to investigate whether similar links may occur in companion animals.
Antibiotic use can increase antimicrobial resistance
Levels of antimicrobial resistance are increasing and antibiotic use in animals has been linked to resistant bacterial strains in humans (Normand et al., 2000; Carattoli et al., 2005; Shea et al., 2011; O’Neill, 2016). Given that widespread use of antibiotics can promote the emergence of resistant bacterial populations (Shea et al., 2011), empirical or unnecessary use of antibiotics is difficult to justify and may result in future scenarios where we lack effective treatment options for currently curable infections. A recent study reported that a short course of amoxicillin-clavulanic acid predisposed dogs with non-complicated acute diarrhoea to the establishment of amoxicillin-resistant E. coli in their faeces, which persisted for as long as three weeks post treatment. Antibiotic use in these dogs did not provide a clinical benefit compared to those given placebo (Werner et al., 2020).
Acute diarrhoea is often self-limiting and likely to resolve without veterinary intervention. Diagnostic investigations are infrequently performed in these cases and even when faecal analysis is carried out, the results can be difficult to interpret given that many potentially pathogenic bacteria are also found in healthy individuals. However, antibiotics are still frequently prescribed to dogs or cats that present with acute diarrhoea. Detection of pathogenic bacteria in the faeces of animals with severe disease, or in those at risk of septicaemia, may warrant the use of antibiotics. However, the use of antibiotics following the detection of zoonotic bacteria in the absence of clinical disease remains questionable, and their use should never be influenced by owner demands alone. Antibiotic use can also contribute to the development of antibacterial resistance, disrupt the normal gastrointestinal microbiota and result in adverse effects for the patient. Therefore, antibiotics should be reserved for specific cases where bacterial infection is confirmed or highly suspected, and/or there is concern of bacterial translocation or sepsis.