Antibiotic treatment to reduce the duration and severity of travellers’ diarrhoea

Objectives

This is a protocol for a Cochrane Review (intervention). The objectives are as follows:

To assess the effects of antibiotic treatment for travellers’ diarrhoea in relation to:

• illness duration;

• adverse effects;

• acquisition of multidrug‐resistant organisms.

Background

Description of the condition

The passing of three or more watery or unformed stools in 24 hours by a person travelling from a high‐income country to a low‐ or middle‐income country is often described as travellers’ diarrhoea (TD) (CDC 2019). Loose bowel motions which are passed more frequently than usual are sometimes accompanied by nausea, vomiting, abdominal pain, or cramping (De Bruyn 2000). TD is typically acute and benign, and functional impact is used by the expert panel of the International Society of Travel Medicine to define its severity. The following definitions are used: mild (acute) diarrhoea is tolerable, not distressing, and does not interfere with planned activities; moderate (acute) diarrhoea is distressing or interferes with planned activities; severe (acute) diarrhoea is incapacitating or completely prevents planned activities (Riddle 2017).

Although usually self‐limiting, with those affected normally recovering in five days or less (De Bruyn 2000), TD can lead to dehydration, and in severe cases, significant complications (Angelo 2018; Bae 2018; Ericsson 2018). In about 10% of cases, symptoms will persist beyond seven days (Farthing 1992; Kass 2005). When TD lasts 14 days or more, it is sometimes termed chronic or persistent diarrhoea and its management differs from usual TD (De Bruyn 2000). In a study evaluating risk factors associated with TD, the use of TD self‐treatment, and the risk of irritable bowel syndrome (IBS) during travel, 3.7% of TD cases developed dysentery (Lalani 2015), which is defined as febrile diarrhoea mixed with visible blood, indicating a potentially serious condition that warrants separate clinical investigation and management (De Bruyn 2000).

TD causes vary by region, though the source is rarely identified in less severe cases. Viral and parasitic infections can cause symptoms, but the bacterial source enterotoxigenic Escherichia coli (ETEC) represents the most common pathogen, responsible for nearly 30% of cases. Other common bacterial species causing TD include Campylobacter jejuni, Shigella spp., and Salmonella spp. The most common parasitic source is Giardia intestinalis, while Entamoeba histolytica and Cryptosporidium spp. are other responsible pathogens (Connor 2018; Dunn 2020; Shirley 2018). Viral causes include norovirus, most common in adults, and rotavirus, which until the advent of global vaccination initiatives had been most common in children (Hall 2013; Kirk 2015; Phillips 2010). Around 20% to 30% of antibiotic‐associated diarrhoea is due to disruption of the usual gut flora which causes Clostridioides difficile infection (CDC 2019; NICE 2015).

One of the determinants of risk for TD is travel to low‐income countries, and causes depend on the season, destination, and setting (Connor 2012; Riddle 2006). A recent review on persistent or chronic diarrhoea in returning travellers found that TD is a leading syndromic diagnosis (condition characterized by a group of symptoms that collectively indicate the presence of the disorder) across all global regions. The review found 6% incidence (proportionate morbidity of 60) observed in over 300,000 global travellers, which is comparable to other previous estimates (Duplessis 2017).

Description of the intervention

Antibiotics treat and prevent bacterial infections by either killing or inhibiting the growth of bacteria and are classed as antimicrobial drugs (Waksman 1947). Antibiotic groups are based on the following criteria: (1) their class (e.g. quinolones, penicillins, macrolides, aminoglycosides, tetracyclines); (2) their mechanism of action (bacteriostatic or bactericidal); and (3) their bacterial spectrum (narrow or broad) (Bérdy 2005). Bacterial enteropathogens are the predominant risk, thought to account for ≥ 80% to 90% of TD cases (CDC 2019).

There is increasing resistance to trimethoprim‐sulfamethoxazole and ampicillin. Therefore, these are less suitable for blind therapy. Fluoroquinolone antibiotics were once the treatment of choice, with standard doses for three to five days reducing the severity and duration of illness by at least 50% (Ericsson 1987; Mattila 1993). However, there is a growing resistance to fluoroquinolones among Campylobacter spp. in some Southeast Asian countries, which poses a significant challenge to its usefulness for TD treatment (Ericsson 2017). A prospective, multisite cross‐sectional study evaluating antibiotic prescriptions for the self‐treatment of TD among adult international travellers between 2009 and 2018 found that fluoroquinolone prescribing has declined dramatically, especially among travellers to Southeast Asia, while azithromycin is now the most frequently prescribed antibiotic for TD (Gandhi 2020). Since the last Cochrane Review on antibiotics for treatment of TD in 2000, there has been a notable decrease of fluoroquinolone use. This is not just due to the emergence of resistance, but also to 'black box' warnings in the USA, which has limited their use not only in children, but in adults too (Kuula 2019).

Although fluoroquinolone antibiotics are generally well‐tolerated, concerns remain about fluoroquinolone‐induced articular toxicity in immature animals and the fact that the quality of evidence supporting clinical use in children is less robust. This has limited their use in children in some areas. However, clinicians still prescribe them, even as antimicrobial‐resistant pathogens continue to emerge (Patel 2016).

How the intervention might work

The efficacy of antibiotics has been shown in the treatment of TD in many randomized placebo‐controlled and comparative trials (De Bruyn 2000). While acknowledging the importance of reducing the use of unnecessary antibiotics, there are circumstances in which these drugs are needed and are potentially life‐saving, particularly in severely sick children, those who have chronic conditions or specific risk factors, or in particular settings (Bruzzese 2018). As shown by several trials, antibiotics taken as a single dose or for up to three days will improve the condition within 20 to 36 hours, shortening the duration of diarrhoea by one to two days, when compared with controls taking placebos (Hill 2008). A graded position statement from Canadian health authorities endorsed the use of antibiotics for treatment of some cases of TD in 2015 (CATMAT 2015).

In 2017, a guideline on the prevention and treatment of TD was published, in which an expert panel used adapted GRADE methodology to make recommendations based on the available evidence. This guideline suggests that there is high‐certainty evidence supporting the effectiveness of antimicrobial therapy in most cases of moderate to severe TD. For management of most cases of mild diarrhoea, either increasing fluid intake only or loperamide or bismuth subsalicylate may suffice (Riddle 2017). Where indicated, a single‐dose regimen of azithromycin, a broad‐spectrum macrolide antibiotic with a long half‐life and excellent tissue penetration, is listed as a treatment choice for enteric infections (McMullan 2015).

Travellers from low‐ and middle‐income countries apparently develop some partial protective immunity for a few months during subsequent travel, and those from countries with an intermediate or high risk of TD have a far lower incidence rate compared with those from low‐risk countries (Angelo 2017; Dupont 1977; Kuenzli 2017; Steffen 2003).

TD is potentially serious in the elderly and very young, those with an impaired immune response, reduced gastric acidity as well as those with underlying health issues such as heart disease (De Bruyn 2000). TD can also result in other comorbid conditions and complications such as cardiovascular symptoms (Martins 2016), and other triggers such as IBS, lactose intolerance, and reduced effectiveness of some medications (Diemert 2006).

Why it is important to do this review

Increased globalized trade and travel are risk factors for infectious disease emergence (IOM 2010), including TD. Although TD is unlikely to result in death, it can disrupt travel plans and lead to severe or incapacitating symptoms (De Bruyn 2000). TD remains a very relevant cause of death, mainly among infants and children (Steffen 2015; WHO 2022), and despite a significant decline from 2.5 million deaths in 2000, diarrhoeal infections are among the world’s top 10 causes of mortality, particularly in low‐income countries, where they rank fifth. They accounted for more than 1.5 million global deaths in 2019 (WHO 2020).

Days lost to illness can be unexpectedly disruptive to international travellers, who are often on restricted schedules, in addition to other factors that might limit their accessibility to TD treatment. These include unfamiliar health systems, language barriers, remoteness, and time constraints, thereby creating a demand for prompt relief. There is a risk of significant impact on travel plans due to TD, with up to 52% of travellers reporting partial or complete incapacitation for a day (Lalani 2015). In up to 30% of travellers, significant or complete disablement has been reported (Riddle 2017). Despite an abundance of information on food and drink safety prior to overseas travel, travellers may not always comply with guidance for various reasons, such as the desire to sample local foods (Kass 2005). A multicentre study of 67,000 people travelling to India, Jamaica, Kenya, and Brazil found that more than 96% of travellers did not follow recommended safe food and water practices (von Sonnenburg 2000).

Previous systematic reviews have demonstrated a reduced illness duration in treatment of TD with antibiotics (De Bruyn 2000). The available evidence on the use of antibiotics for acute diarrhoea treatment in children still requires careful consideration of clinical and epidemiological issues (Bruzzese 2018), and for this reason, we will consider children as a subgroup of interest in this review.

Antibiotics are effective in reduction of the duration of diarrhoea by about a day in cases of bacterial pathogenesis, if pathogens are susceptible to the antibiotic prescribed. The risk of potential harm increases in the immunosuppressed, or women who may be prone to urinary tract infection, with the added risk and possibility of introducing resistant bacteria into the community (CDC 2019).

There has been a change in patterns of antimicrobial resistance in enteropathogens and in the preferred antimicrobials since the last major graded recommendations (CATMAT 2015; Riddle 2017), hence the need to incorporate more recently published evidence on antibiotic effectiveness. The GRADE expert panel has acknowledged the need to address knowledge gaps regarding global health risks of multidrug resistance acquisition (Riddle 2017). Antibiotic use against non‐bacterial pathogens increases infection risk due to removal of protective microflora from the bowel (CDC 2019). There is new evidence relating to increased prevalence of colonization by antimicrobial resistant organisms as a cause of illness. There is also an increased risk of travellers becoming colonized with extended‐spectrum β‐lactamase (ESBL)‐producing bacteria, particularly when abroad.

Beyond the scope of this review are the costs of antibiotic treatment, including the treatment of acute infectious diarrhoea with probiotics, which might be better addressed in an economic evaluation of TD study. Acute infectious diarrhoea treatment with probiotics is addressed in a recent Cochrane Review (Collinson 2020), while a systematic review of non‐pharmacotherapeutic interventions in TD also investigates both pre‐ and probiotics (Evans 2018).

The authors of the Cochrane Review on antibiotic treatment for TD published in 2000 were unable to perform a quantitative analysis of data relating to duration of diarrhoea, as data from the included trials were not combinable in a meta‐analysis. Concerns around adverse consequences of antibiotic treatment of TD as well as new evidence and developments in the antibiotic treatment of TD warrant an updated review, and more recent trials may provide more information on key patient‐important outcomes.

Objectives

To assess the effects of antibiotic treatment for travellers’ diarrhoea in relation to:

• illness duration;

• adverse effects;

• acquisition of multidrug‐resistant organisms.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomized controlled trials (RCTs) in all languages, including any unpublished studies.

Types of participants

We will include adults and children with diarrhoea, defined as at least three or more watery or unformed stools in 24 hours, including dysentery, lasting for up to 14 days, who develop symptoms after travelling outside their normal country of residence, or who have been resident in the destination country for less than six months. We will accept and document similar definitions.

We will define the population subgroup ages as follows:

• < 1 year: infants;

• 1 to 4 years: toddlers;

• 5 to 17 years: children;

• 18 to 64 years: adults;

• ≥ 65 years: older adults.

Types of interventions

Intervention

Treatment with any single antibiotic or combination of antibiotics given as empirical treatment (defined as therapy directed against the anticipated most likely cause of diarrhoea, prescribed prior to the availability of results of any investigations on its aetiology). Interventions will be eligible regardless of co‐interventions, but co‐interventions must be the same in both treatment arms.

Control

• Placebo

• No treatment or symptomatic treatment

• Any single antibiotic or combination of antibiotics that is different to the intervention antibiotic

• The intervention antibiotic given at a different dosing regimen

Types of outcome measures

Primary outcomes

• Duration of diarrhoea (time to last unformed stool (TLUS)) measured in hours.

Secondary outcomes

• Number of participants requiring admission to hospital

• Number of participants requiring intravenous fluid therapy

• Colonization with antibiotic‐resistant bacteria within 30 days of treatment (we will accept and document any primary study definition of 'antibiotic‐resistant bacteria')

• Number of participants with ongoing diarrhoea at the end of the follow‐up period (up to 30 days)

• Number of participants experiencing any adverse effects (defined as any systemic event such as fever, rash, myalgia, headache, fatigue, or any untoward medical occurrence that at any dose results in persistent or significant disability/incapacity, requires in‐patient hospitalization or prolongation of existing hospitalization, is life‐threatening, or results in death)

Search methods for identification of studies

We will attempt to identify all relevant studies regardless of language, publication status, or publication date limit.

Electronic searches

We will search the following databases using the search terms and strategy described in Appendix 1: the Cochrane Central Register of Controlled Trials (CENTRAL, published in the Cochrane Library), MEDLINE (via Ovid), Embase (via Ovid), Scopus (Elsevier), and we will check Google Scholar. We will also search the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) and ClinicalTrials.gov for trials in progress.

Searching other resources

We will also check the reference lists of relevant studies. We will use the “Similar articles” feature in PubMed to identify additional references.

Data collection and analysis

Two review authors (RT and WN) will conduct each step of the study selection and data extraction process independently and in duplicate. This includes assessment for participants, intervention, control, outcomes, designs (PICOD) as well as allocation concealment, generation of allocation sequence, and whether all randomized participants are included. Any disagreements will be resolved through discussion and reviewed by a third review author.

Selection of studies

Two review authors (RT and WN) will independently and in duplicate screen the titles, abstracts, and keywords of each record identified in the searches using Covidence (Higgins 2020). We will retrieve the full texts of all potentially relevant records. Two review authors (RT and WN) will independently and in duplicate apply the inclusion criteria to each of the full‐text records obtained following screening.

Any differences will be resolved through discussion between review authors. We will attempt to contact trial authors for clarification of eligibility of a trial if this is unclear, or if further information is required. We will examine each trial report to ensure that we include multiple publications from the same trial only once. We will list the studies excluded after full‐text review and the reasons for their exclusion in a 'Characteristics of excluded studies' table.

Data extraction and management

Two review authors (RT and WN) will independently and in duplicate extract data on participant characteristics, diagnostic criteria, disease severity, comorbidity, antibiotic dose and administration, other treatments given, and outcome measures using a piloted data extraction form that we have adapted for this review (Appendix 2) (ADMMH 2016). Any disagreements will be resolved through discussion. We will contact the corresponding trial authors in the case of unclear or missing data.

Trials involving patient‐initiated treatment will be differentiated from those involving physician‐initiated treatment. We will gather destination or travel location on contracting TD due to differences in aetiology between locations that may influence whether one antibiotic is more efficacious over another.

For dichotomous outcomes, we will record the number of participants who experienced the event and the number of participants randomized to each treatment group. We will record the number of participants analyzed in each treatment arm and use the discrepancy between the figures to calculate the number of participants lost to follow‐up, which will allow us to perform sensitivity analyses to investigate the effect of missing data if necessary. For continuous outcomes, we plan to extract the standard deviation and the mean for the outcome in each group; we will also record medians for narrative comparisons where means are unavailable.

Assessment of risk of bias in included studies

We will estimate the effect of assignment to the interventions at baseline, regardless of whether the interventions are received as intended (the ‘intention‐to‐treat' effect).

Two review authors (RT and WN) will independently perform risk of bias assessments on this effect, and also assess the risk of bias for each outcome in all included trials according to Cochrane's RoB 2 tool (Higgins 2022). We will assess risk of bias according to the domains below that are currently understood to affect the results of randomized trials:

• bias arising from the randomization process;

• bias due to deviations from intended interventions;

• bias due to missing outcome data;

• bias in measurement of the outcome;

• bias in selection of the reported result.

We will use the signalling questions below to elicit information relevant to an assessment of risk of bias:

• yes;

• probably yes;

• probably no;

• no;

• no information.

We will classify each domain understood to affect the results of randomized trials as low risk of bias, some concerns, or high risk of bias. For each outcome, we will consider the implications of the judgements in each domain for the overall risk of bias judgement, asking whether the problems identified are likely to affect the ability to draw reliable conclusions from the study. As per Table 8.2.b in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022), we will consider the overall risk of bias to be low risk of bias if all domains are at low risk; some concerns if at least one domain is some concerns and no domain is at high risk; and high risk of bias if at least one domain is at high risk or several domains are some concerns. We will attempt to contact the trial authors if information is not specified or is unclear, resolving any disagreements by discussion between review authors. To maintain full, consistent evaluations of all included evidence, and in the interests of transparency and continuity, we plan to include all eligible trials including those with a high risk of bias and perform sensitivity analyses.

For cluster‐RCTs, we will use the version of the tool tailored to this trial design, with an additional domain to assess bias associated with timing, and recruitment of participants. We will take into account the risk of bias for all the domains including (1.b) bias arising from identification or recruitment of individual participants within clusters, or the cross‐over RCT domain (S) (domain addressing two issues that are specific to cross‐over trials: period effects and carryover). This is described in the RoB 2 tool (Higgins 2022). We will conduct a count of adverse events by participant associated with the antibiotic received.

We will use the purpose‐built Excel tool to manage the data generated during the risk of bias assessment, and these will be published online with a link to access them in the final review (Higgins 2022).

Measures of treatment effect

For continuous data, where outcomes are measured in the same way between trials, we will use mean differences (MDs). We will use standardized mean differences (SMDs) where trials measured the same outcome in different ways. Where trials report continuous data as median and interquartile ranges (IQR), we will use the median as the mean where data are distributed normally. Where data are skewed, we will impute the missing mean according to the formula by Wan 2014, based on the lower quartile, median and upper quartile summary statistics. We will test for skewness by calculating the observed mean minus the lowest possible value (or the highest possible value minus the observed mean), and dividing this by the standard deviation. A ratio less than 2 suggests that the data are skewed, while a ratio of less than 1 suggests strong evidence of a skewed distribution (Altman 1996). We will not include studies with skewed data that report on medians and IQRs in the meta‐analysis, but will instead report their results separately.

For dichotomous data, we will use risk ratio as the treatment effect for analysis as suggested in Chapter 10 of the Cochrane Handbook (Higgins 2022).

Unit of analysis issues

If we identify multi‐arm trials, we will select relevant arms for inclusion in our analyses, or if more than two arms are relevant to the review, we will combine intervention arms to allow a single comparison. Alternatively, we will split the control group between multiple comparisons so that participants are not double‐counted in the meta‐analysis.

Dealing with missing data

If data are incompletely reported, we will contact the study authors to request additional information.

Missing outcomes and summary data may be the result of selective reporting bias; missing individuals may be the result of attrition from the study or lack of intention‐to‐treat analysis. We plan to address these sources of missing data using the RoB 2 tool (Higgins 2022). Where missing data introduce serious risk of bias, we will explore the impact of including such studies in the overall assessment of results through a sensitivity analysis.

We aim to identify all research that meets our predefined eligibility criteria, including searching for completed non‐published trials in trial registers to avoid missing studies, which could introduce reporting bias to the analysis. Where data are missing, we will impute the missing data with replacement values, and treat these as if they were observed (e.g. last observation carried forward, imputing an assumed outcome such as an assumption that all were poor outcomes, imputing the mean, or imputing based on predicted values from a regression analysis).

Assessment of heterogeneity

We will assess heterogeneity using visual inspection of the forest plots, by comparing the heterogeneity statistic Q with the Chi² distribution, and determining the amount of heterogeneity using the I² statistic.

We will consider statistical heterogeneity to be present if the P value is less than 0.1 for the Chi² statistic, and will use the following ranges to interpret the I² statistic as outlined in the Cochrane Handbook (Higgins 2019):

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity;

• 50% to 90%: may represent substantial heterogeneity;

• 75% to 100%: considerable heterogeneity.

We will explore any substantial statistical heterogeneity through subgroup analyses.

Assessment of reporting biases

If information is not specified, unclear, or if any other uncertainty arises regarding an included trial, we will attempt to resolve the issue through contact with the trial authors, or by a team review of the evidence to reach a consensus. If 10 or more trials are included for a given outcome, we will create a funnel plot to investigate publication bias.

Data synthesis

We will analyse data in pairwise comparisons using Review Manager Web (Deeks 2022; RevMan Web 2023).

We plan to assess the following comparisons:

• any antibiotic treatment versus no antibiotic treatment or placebo;

• any antibiotic treatment versus any other antibiotic treatment;

• various dosing regimens of the same antibiotic.

If clinical and methodological characteristics of individual trials are sufficiently similar, we will pool the data in meta‐analyses using Review Manager Web (RevMan Web 2023). Where there are no concerns about clinical or statistical heterogeneity, we will use the fixed‐effect model in meta‐analyses. Where we detect clinical or statistical heterogeneity but still consider pooling of data to be appropriate, we will use the random‐effects model. Where a meta‐analysis is not appropriate, we will conduct a narrative synthesis of the data.

Subgroup analysis and investigation of heterogeneity

We will test for subgroup interactions and investigate heterogeneity in the following subgroup analyses:

• toddlers aged 1 to 4 years versus children aged 5 to 17 years versus adults ≥ 18 years;

• patient versus physician‐initiated treatment;

• destination or travel location on contracting TD;

• adults with bloody diarrhoea versus adults with non‐bloody diarrhoea;

• vomiting at recruitment;

• immunocompromised patients including people living with HIV;

• severity of diarrhoea (mild, moderate, or severe) as defined by study authors.

Sensitivity analysis

We will perform a sensitivity analysis on the primary outcome, TLUS, based on the risk of bias assessment and validate the findings of this outcome based on data extracted from studies at low risk of bias.

Summary of findings and assessment of the certainty of the evidence

Two review authors will independently and in duplicate assess the certainty of the evidence for all outcomes using the five GRADE framework considerations (risk of bias, inconsistency, imprecision, indirectness, and publication bias). Any differences in assessment will be resolved by discussion or by consultation with a third review author.

We will include a summary of findings table for the following outcomes (we will perform a risk of bias assessment for each) in all trials:

• duration of diarrhoea;

• number of participants requiring admission to hospital during the study period;

• number of participants requiring intravenous fluid therapy during the study period;

• colonization with antibiotic‐resistant bacteria;

• number of participants with ongoing diarrhoea at the end of the follow‐up period;

• number of participants experiencing any adverse effect.

We plan to include all six outcomes (primary and secondary) in the summary of findings table. This table will present key information regarding the certainty of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schünemann 2019). We will use version 3 of GRADEpro GDT to present the summary of findings table in Review Manager Web (GRADEpro GDT; RevMan Web 2023).

The overall RoB 2 assessment will be used to feed into the GRADE assessment, and we will report the certainty of evidence as high, moderate, low, or very low according to Chapter 14 of the Cochrane Handbook (Schünemann 2019). We will justify all decisions to downgrade the certainty of the evidence using footnotes, and will use GRADE guidelines for informative statements as suggested by Santesso 2020.

Appendices

Appendix 1. Search strategy

Ovid MEDLINE(R) and In‐Process, In‐Data‐Review & Other Non‐Indexed Citations <1946 to present >

1. exp Anti‐Bacterial Agents/

2. (antibiotic* or anti?biotic* or antimicrobial* or anti?microbial* or antibacterial* or anti?bacterial* or antimycobacterial* or anti?mycobacterial* or bacteriocid*).mp.

3. 1 or 2

4. traveler* diarrhea.mp.

5. Travel/

6. travel*.mp.

7. Diarrhea/

8. (diarrh* or gastroenteritis).mp.

9. 5 or 6

10. 7 or 8

11. (Travel/ or travel*.mp.) adj2 (Diarrhea/ or (diarrh* or gastroenteritis).mp.)

12. 4 or 11

13. . 3 and 12

14. (randomized controlled trial or controlled clinical trial).pt.

15. (randomized or placebo).ab.

16. drug therapy.fs.

17. (randomly or trial or groups).ab.

18. 14 or 15 or 16 or 17

19. animals/ not humans/

20. 18 not 19

21. 13 and 20

This search strategy will be applied to MEDLINE Ovid and adapted for other databases.

Appendix 2. Data extraction form

Population: less than 1 year old 1 to 4 years old 5 to 17 years old 18 to 64 years old ≥ 65 years old with acute diarrhoea for up to a maximum of 14 days (subgroup ‐ bloody diarrhoea of any description) travelling outside normal country of residence resident in destination country of study for less than 6 months	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Intervention: any antibiotic treatment versus no antibiotic treatment or placebo any antibiotic treatment versus any other antibiotic treatment various dosing regimens of the same antibiotic	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Comparison: placebo/no treatment symptomatic treatment non‐antimicrobial (e.g. oral hydration salts, pre‐travel vaccinations) antimotility agents (e.g. loperamide) preventative measures (e.g. handwashing regimen)	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Outcomes: Primary 1. duration of diarrhoea (time to last unformed stool (TLUS)) in hours Secondary 1. number of participants requiring admission to hospital during the study period 2. number of participants requiring intravenous fluid therapy during the study period 3. colonization with antibiotic resistant bacteria within 30 days of treatment (we will accept and document any primary study definition of 'antibiotic‐resistant bacteria') 4. number of participants with ongoing diarrhoea at the end of the follow‐up period (30 days) 5. number of participants experiencing any adverse effects (defined as any systemic event such as fever, rash, myalgia, headache, fatigue, or any untoward medical occurrence that at any dose results in persistent or significant disability/incapacity, requires inpatient hospitalization or prolongation of existing hospitalization, is life‐threatening, or results in death)	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Design: RCT	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Allocation concealment?	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Generation of allocation sequence?	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
All randomized participants included?	Yes √ ↓ Next question	Unclear ↓ Next question	No → EXCLUDE
Final decision	INCLUDE √	UNCLEAR (Re‐evaluate)	EXCLUDE (Reason)

(Adapted from Surgeon General's review strategy, 2016)

Contributions of authors

All authors participated in developing the protocol. All authors read and approved the final version of the protocol prior to publication.

Sources of support

Internal sources

• Liverpool School of Tropical Medicine, UK

• Defence Medical Services, UK

External sources

• Foreign, Commonwealth, and Development Office (FCDO), UK

  Project number 300342‐104

Declarations of interest

RT, a member of the British Army, is part of the Military Enteric Diseases Group (UK) and investigates travellers' diarrhoea, with a primary focus on military populations. He has no known conflicts of interest.

WN is an Infectious Diseases Registrar, Royal Air Force: Higher Specialist Trainee in Infectious Diseases and General Internal Medicine, part of the Defence Deanery and employed by the Royal Air Force, working in the NHS, and has no known conflicts of interest.

RK is a General Practitioner. She is a Cochrane Infectious Diseases Group Research Associate, and was not involved in the editorial process. She has no known conflicts of interest.

HR works as a medical doctor at Liverpool University Hospitals NHS Foundation Trust (LUHFT NHS), and has no known conflicts of interest.

NJB was a co‐author of previous graded guidelines on travellers’ diarrhoea (Riddle 2017), and has authored several publications and reviews in peer‐review literature, unfunded, or derived from non‐commercial grants received more than three years ago. Independent trial monitor for publication likely to be included. Supervisor of research on screening military personnel including for diarrhoea‐inducing pathogen. Military Enteric Diseases Group (advisory role) UK. International Society of Travel Medicine (ISTM) Fellow, and received ISTM travel and publication expenses. European Study Group on Infections in Travellers and Migrants. Guidelines on travellers' diarrhoea in organ transplant recipients. He has no known conflicts of interest.

TF is a Liverpool University Hospitals NHS Foundation Trust (LUHFT NHS) Infectious Diseases Consultant, and has no known conflicts of interest.

DB is a Military Infectious Diseases Consultant, treating patients with the condition considered by the review. As part of the Military Enteric Diseases Group (UK), he advises the UK military on travellers' diarrhoea. He has no known conflicts of interest.

New