Efficacy of bismuth subsalicylate on the prevention of travellers′ diarrhoea, 2018–2023

Abstract

Background:

Travellers’ diarrhoea (TD) is the most common travel-related illness. Bismuth subsalicylate (BSS) is indicated for the treatment of TD. BSS is also used off-label for the prevention of TD, based on studies from the 1980s indicating TD protection; however, these studies have limitations. The objective of this study was to determine the efficacy of BSS in the prevention of TD.

Methods:

This study was a prospective, double-blinded, placebo-controlled, randomized clinical trial (NCT03535272) with two arms: BSS 4 tablets twice daily (2.1 grams of BSS total) vs placebo. Travellers were included if they were ≥ 18 and < 70 years of age at the time of enrollment, were leaving for an international trip ≥7 days after their pretravel consultation, travelling in country for ≥7 days but ≤ 21 days, and travelling to either Southeast Asia, South Central Asia, North Africa, or Sub-Saharan Africa. Self-completed web-based questionnaires were administered before, during, and after travel.

Results:

270 participants were included. The median age was 32 years (Interquartile range [IQR]: 27–44); 63% were female. Travellers were most frequently White (144; 61%) and non-Hispanic (239; 94%). The most frequent country of travel was Kenya (n = 87; 32%). The median total trip duration was 10 days ([IQR]: 8–13). The most frequent reason for travel was leisure/tourism (230, 85%). There was no significant difference among the groups for symptoms of loose stool and/or diarrhoea, although target sample size was not reached.

Conclusions:

This study provides the first new data since the 1980s about the potential use of BSS in the prevention of TD. These data should not be viewed as evidence in isolation because of sample size constraints; further studies are needed to determine if there is a benefit in certain traveller groups or under certain circumstances.

Introduction

Travellers’ diarrhoea (TD) affects between 30%–70% of all travellers and is the most common travel-related illness;¹ incidence of TD among short term (<100 days) travellers from high income countries may be between 20% and 56%.² The incidence of TD varies by travel destination, duration of travel, season, travel activities, and individual traveller attributes (e.g. comorbidities). Some areas of the world are considered high-risk, including Asia and Africa. TD can be associated with long-term morbidity, including post-infectious irritable bowel syndrome (PI-IBS), reactive arthritis, or Guillain-Barré syndrome.¹

Bismuth subsalicylate (BSS) has been approved by the US Food and Drug Administration since 1939 and is indicated for the treatment of TD, heartburn, indigestion, nausea, and stomach upset.³ BSS is also used off-label for the prevention of diarrhoea, particularly among international travellers.⁴ The active component of BSS with antimicrobial effects is bismuth, which is poorly absorbed and mostly remains in the gastrointestinal tract, while salicylate is systemically absorbed in the stomach.⁵ The precise mechanisms of action include the binding of ingested bacteria and marking of these organisms to be killed by host defences, rendering the bacteria unable to attach to receptor sites for colonization and infection,⁶ degrading the bacterial cell wall, inhibiting ATP synthesis, and prohibiting the function of the plasma membrane.⁵ This inhibition may occur in as little as 0.5–24 hours.⁵ Previous studies, performed in the 1980s, investigating the prophylactic use of BSS among international travellers, demonstrated lower incidence of TD among those receiving the drug^7–10 and a meta-analysis demonstrated that travellers who received BSS during travel had 3.5-times greater odds of remaining free of TD in comparison to those who received placebo.¹¹ However, many studies indicating protection by BSS have limitations—they were performed in individuals travelling to Latin America, among specific cohorts (e.g. students), over 30 years ago.

The objective of this study was to determine the efficacy of BSS in the prevention of TD among individuals returning to the United States after travel to Southeast Asia, South Central Asia, or Africa.

Methods

Trial design

This study was a prospective, double-blinded, placebo-controlled, randomized clinical trial (NCT03535272) with two arms: BSS 4 tablets twice daily (taken morning and night for 2.1 grams of BSS total) vs placebo.¹¹

A screening questionnaire was used to determine eligibility for inclusion (Appendix A). Travellers were included if they were ≥ 18 and < 70 years of age at the time of enrollment. Participants had to leave for an international trip ≥7 days after their pretravel consultation, travel in country for ≥7 days but ≤ 21 days, and travel to either Southeast Asia, South Central Asia, North Africa, or Sub-Saharan Africa for at least 7 days of their itinerary (Appendix B). Travellers must have been able to sign an informed consent in person or by e-signature (Appendix C). Travellers must also have been willing to refrain from taking any pre-biotics, probiotics, synbiotics, and/or herbal supplements throughout their study period (enrollment to post-travel follow-up).

Once recruited and consented, eligible participants received a one-page instructional handout regarding stool specimen collection and questionnaire completion (Appendix D). Self-completed web-based questionnaires were administered before, during, and after travel. Pretravel data was collected by study staff (not web-based) and included the participant’s medical history and travel plans (Appendix E). Follow-up telephone calls by study personnel were performed on pre-departure Days 7, 5, 3, and 1 as needed to complete the questionnaire. During travel, participants completed daily questionnaires that included data on adherence in taking the study tablets, presence of symptoms of TD, medication reactions, any medical care received, and trip specifics (Appendix F). The during travel questionnaire was available either as a web-based electronic form or a paper form (returned to the clinic mail) if the participant did not have reliable internet access. The electronic version of the form was only available for 48 hours and then was not accessible for a given day to limit recall bias. Paper forms sent to clinic were entered into REDCap by clinic staff. Post-travel data included information on TD symptoms, medication use, medical care received, and adverse events (Appendix G). Follow-up telephone calls by study personnel were performed on Days 3, 5, 7, and 10 after return as needed to ensure the collection of the post-travel questionnaire and paper questionnaires from during travel.

As part of standard pre-travel consultation practices, if indicated, travellers were prescribed anti-diarrheal medications and/or antibiotics for self-treatment of TD; if these medications are taken while abroad, the travellers were asked to note this on the during travel and post-travel questionnaires and instructed to continue their study medication.

Participants

Travellers were excluded if they had a known or suspected contraindication to taking BSS (e.g. kidney disease, diabetes, gout, a clotting disorder, or an allergy to any component of BSS); were pregnant, were planning to become pregnant, could become pregnant during travel (not actively using contraception and were sexually active), or were breastfeeding; routinely took a medication known to interact with BSS (e.g. insulin, other salicylates); took an antibiotic in the 30 days before departure; took any medication that may lower one’s ability to fight infection (e.g. steroids, monoclonal antibodies); had previous diagnoses of immunocompromising conditions (e.g. HIV/AIDS, undergoing active chemotherapy) or chronic gastrointestinal disorders (e.g. inflammatory bowel disease, celiac disease); had diarrhoea anytime in the previous 30 days, had diarrhoea at the pre-travel consultation, or developed diarrhoea before departure; or were given doxycycline for malaria prophylaxis for the current trip¹² (Appendix A).

Study participants were recruited from two clinical sites in the United States (New York and Massachusetts) that offer pretravel specialty services. Each site consented and enrolled participants, with the lead site coordinating medication distribution and follow-up.

Interventions

The intervention medication was bismuth subsalicylate (marketed as Pepto Bismol).

The placebo was comprised of microcrystalline cellulose, calcium carbonate, D&C red No. 27 aluminium lake, flavour, magnesium stearate, mannitol, povidone, saccharin sodium, and talc. It was formulated to have similar appearance (size, shape, colour, and packaging), texture, and taste to BSS. Each bottle used in the study included instructions for taking the tablets, study ID#, batch number, start date, date dispensed, quantity dispensed, expiration date, and clinic contact information.

Outcome

The primary outcome was to assess the efficacy of BSS on the presence or absence of any TD among enrolled participants during and after travel. There were no changes to the trial outcomes during the study.

Sample size

Sample size was estimated using a 40% presumed incidence of TD with a 30% minimum intervention (BSS) protection; with an alpha of 0.05, and 80% power, 482 participants with full compliance to medications and questionnaires (241 participants in each arm) were needed to fulfil the primary objective. 2023 PASS Analysis and Sample Size Software was used to calculate minimum sample size (NCSS, LLC. Kaysville, Utah, USA, ncss.com/software/pass).

Randomization

Randomization of participants was performed by a statistician unaffiliated with the study. Participants were randomized to ensure groups of equal sample sizes (1:1) using a computer-generated distribution in blocks of 4. Study medications for both the BSS and placebo group were blinded by a third party using a unique identifier and were distributed to participants at enrollment or were shipped to participants by the study clinics.

Blinding

Care providers, participants, and all study personnel except the individual who labelled and distributed the randomized bottles to the clinical sites were blinded.

Definitions

Diarrhoea was defined as ≥3 unformed stools in 24 hours, with or without associated symptoms such as fever, cramping, or abdominal pain. Loose stool was < 3 unformed stools in 24 hours. TD was defined as diarrhoea that began while travelling (after arrival in a country) or within 7 days after return. Severe TD was defined as hospitalization to an inpatient ward either during travel or after travel, with a diagnosis of diarrhoea. Travel companions were defined as two or more study participants with a common relation (e.g. friends, relatives) with an overlapping itinerary, who anticipated performing the same activities during travel. Loss to follow-up was defined as the failure to complete the post-travel questionnaire after three attempts to contact the participant or if submission of the post-travel stool specimen occurred more than 10 days after return from travel or was not submitted. These participants were removed from the final analysis.

Statistical methods

Travellers with incomplete post travel questionnaires (n = 55), missing screening questionnaire (n = 1), missing ID (n = 1), and having < 7 travel days (n = 6) were excluded from all analyses. If travellers identified themselves with more than one race category, they were classified as ‘Other’ race (n = 1).

Demographic characteristics, travel factors, and health statuses at baseline were analysed between arms using Wilcoxon rank-sum test for continuous variables and using Chi-square test for categorical variables.

The median number of travel days were compared between the two arms using Wilcoxon rank-sum test. The association between the occurrence of symptoms, medical care seeking, new medications, and consumption of yogurt or fermented or pickled foods during travel (if symptoms were reported for ≥1 day), was assessed between the two arms using Chi-square test. The weekly mean (in days) (e.g. the number of days in a week that a traveller had a given symptom) was assessed between the two arms using a two-sample t-test. Post-travel symptoms, medications, and medical care seeking were compared using Chi-square test. If cell counts were less than 5, the Chi-square test was replaced with Fisher Exact test.

Two additional sensitivity analyses were performed. One analysis was performed to assess if the COVID-19 pandemic was associated with a difference in the incidence of TD among travel completed prior to December 31, 2019 versus on or after January 1, 2020. Also, an analysis of participants who reported black stool versus those who did not report black stool (as a marker of drug ingestion) was performed to assess if those with black stool (presumably taking BSS) had any characteristics that differed from those that did not. Analyses were performed per protocol. Statistical significance was defined as P < 0.05, no adjustment for multiple tests was considered as this was exploratory in nature. All analyses were performed in SAS 9.4 (SAS Institute, Cary NC).

Research determination

This study was reviewed and approved as non-exempt human subjects research by the Quorum Review IRB (QR#:33046) (2018–2020) and Advarra IRB (Pro00036386) (2021–2024).

Results

Participant flow

In total, 415 travellers were screened (Figure 1) from May 2018 through December 2023, and 270 participants were included in the analysis.

Figure 1.

Flow chart of participants and exclusions, 2018–2023. ∗10 (3%) did not have the minimum 7 days in the specified regions; 6 (2%) had < 7 days of travel data; 3 (1%) were duplicate records; 1 (0.4%) had missing screening data; 1 (0.4%) took a disqualifying medication. ∗∗81 participants completed at least one daily during travel questionnaire on paper

Recruitment

The trial was terminated before achieving target sample size because of prolonged timeframe from the COVID-19 pandemic’s international travel disruptions and funding constraints.

Baseline data with numbers analysed

The median age of travellers was 32 years (Interquartile range [IQR]: 27–44); 63% were female (Table 1). Travellers were most frequently White (144; 61%) and non-Hispanic (239; 94%). The five most frequent countries of travel were Kenya (n = 87; 32%), Ghana (n = 51; 19%), Tanzania (n = 36; 13%), and Thailand (n = 24; 9%) (Supplementary Figure 1); most travellers visited more than one country. The median total planned trip duration was 10 days ([IQR]: 8–13). The most frequent reason for travel was leisure/tourism (230; 85%). Fifteen (6%) travellers had a medical condition and 44 (16%) took a medication regularly (most frequently birth control [10; 23%]). Thirty (12%) of 254 participants with information available travelled internationally in the 90 days before enrollment. Travellers in the placebo group were more frequently non-Hispanic (P = 0.04) and had a shorter trip duration (10 vs 11 days, P = 0.036).

Table 1.

Demographic characteristics of participants based on the pretravel questionnaire, 2018–2023 (N = 270)

	All participants N = 270	BSS n = 136 (50%)	Placebo n = 134 (50%)	p-value a
Age in years, median (IQR)	32 (27–44)	32 (27–40)	32 (27–45)	0.623
	N (%)	n (%)	n (%)
Sex, n (%)
Female	169 (63)	84 (62)	85 (63)	0.837
Male	100 (37)	51 (38)	49 (37)
Missing	1	1	0
Raceb, n (%)
White	144 (61)	73 (62)	71 (60)	0.618
African American	56 (24)	24 (21)	32 (27)
Asian/Pacific Islander	23 (10)	13 (11)	10 (8)
American Indian/Alaska Native	1 (0)	1 (1)	0 (0)
Other	11 (5)	6 (5)	5 (4)
Don’t know/Declined	35	19	16
Ethnicity, n (%)
Non-Hispanic	239 (94)	116 (91)	123 (97)	0.040
Hispanic	16 (6)	12 (9)	4 (3)
Missing/Declined	15	8	7
Anticipated trip duration in days, median (IQR)	10 (8–13)	11 (9–14)	10 (8–13)	0.073
Main reason for travel, n (%)				0.339
Leisure/Tourism	230 (85)	116 (86)	114 (85)
Visit Friends/Relatives	14 (5)	4 (3)	10 (7)
Education/research	12 (4)	8 (6)	4 (3)
Business	8 (3)	5 (4)	3 (2)
Non-medical service or volunteer work	5 (2)	2 (1)	3 (2)
Chronic medical condition or illnessc n (%)	15 (6)	8 (6)	7 (5)	0.802
Take medication regularly, n (%)	44 (16)	20 (15)	24 (18)	0.509
Have taken oral/IV antibiotics ≤3 months, n(%)	1 (0)	1 (1)	0 (0)	0.999
Have taken a probiotic ≤ 30 days, n (%)	4 (2)	2 (1)	2 (1)	0.999
Travelled internationally in the 90 days before enrollmentd, n (%)	30 (12)	15 (12)	15 (12)	0.927

BSS = bismuth subsalicylate

^aWilcoxon rank-sum test was used to compare the median estimates; Chi-square test was used to compare the distribution of frequencies and Fisher Exact test was applied when the counts were < 5. The p-value does not test the ‘missing’ data categories.

^bIf a participants selected more than one race category, then their race was classified as ‘other’.

^cAsthma (n = 4), hypertension (n = 3), hypothyroidism (n = 2), anxiety/depression (n = 1), fibroids (n = 1), chronic migraines (n = 1), pacemaker (n = 1), vasovagal syncope (n = 1), unknown (n = 1).

^dOf 254 participants with information available: 129 intervention and 125 placebo.

Outcomes and estimation

168 (60%) of travellers were self-reported to be compliant with all their medications; over 70% of participants in the placebo group were over 90% compliant (Supplementary Figure 2). There was no significant difference in self-reported TD incidence among the group that received BSS and the group that received placebo (Table 2). During travel, travellers reported a variety of symptoms, most frequently flatulence (134; 50%). Reported symptoms more frequent in the treatment group were vomiting (P = 0.036), dark/black stool (P < 0.001), dark/black tongue (P < 0.001), dizziness (P = 0.042), and muscle pain/soreness (P = 0.016). Five (2%) travellers sought medical care for their symptoms during travel and 105 (39%) started a new medication, most frequently the antimalarial atovaquone/proguanil (n = 66; 63%).

Table 2.

Travel days, symptoms, medical care seeking, and exposures during travel, 2018–2023

	Number of Participants (%)				Weekly Mean (in days)
	All participants N = 270	BSS n = 136 (50%)	Placebo n = 134 (50%)	p-value	All days	BSS	Placebo	p-value c
Median number of travel days recorded per participant (IQR)a	10 (8–13)	11 (9–13)	9.5 (8–12)	0.053	—	—	—	—
Symptoms during travelb	N (%)	n (%)	n (%)		Mean (min-max)	Mean (min-max)	Mean (min-max)
Flatulence	134 (50)	67 (49)	67 (50)	0.904	1.21 (0–7.0)	1.16 (0–7.0)	1.26 (0–7.0)	0.673
Loose stools	120 (44)	62 (46)	58 (43)	0.703	0.79 (0–6.5)	0.83 (0–6.5)	0.75 (0–5.5)	0.562
Dark/black stool	106 (39)	92 (68)	14 (10)	< 0.001	1.68 (0–7.0)	3.05 (0–7.0)	0.29 (0–6.5)	< 0.001
Bloating	84 (31)	44 (32)	40 (30)	0.657	0.66 (0–7.0)	0.66 (0–7.0)	0.65 (0–7.0)	0.951
Constipation	79 (29)	35 (26)	44 (33)	0.200	0.62 (0–7.0)	0.54 (0–7.0)	0.71 (0–7.0)	0.285
Urgency to relieve oneself	65 (24)	33 (24)	32 (24)	0.941	0.34 (0–6.0)	0.37 (0–6.0)	0.31 (0–6.0)	0.615
Dark/black tongue	61 (23)	56 (41)	5 (4)	< 0.001	0. 76 (0–6.5)	1.43 (0–6.5)	0.08 (0–5.7)	< 0.001
Abdominal pain	63 (23)	32 (24)	31 (23)	0.939	0.26 (0–5.0)	0.21 (0–2.6)	0.31 (0–5.0)	0.188
Abdominal cramps	58 (21)	27 (20)	31 (23)	0.512	0.28 (0–6.5)	0.31 (0–6.5)	0.26 (0–2.8)	0.587
Diarrhoea	53 (20)	27 (20)	26 (19)	0.926	0.26 (0–3.9)	0.23 (0–2.8)	0.28 (0–3.9)	0.493
Combined: loose stool plus diarrhoea	137 (51)	70 (51)	67 (50)	0.809	1.01 (0–6.5)	1.02 (0–6.5)	0.99 (0–5.5)	0.845
Nausea	54 (20)	33 (24)	21 (16)	0.078	0.24 (0–6.1)	0.27 (0–6.1)	0.22 (0–4.0)	0.524
Congestion/runny nose	45 (17)	26 (19)	19 (14)	0.276	0.30 (0–5.5)	0.36 (0–5.5)	0.25 (0–4.9)	0.309
Headache	47 (17)	21 (15)	26 (19)	0.391	0.19 (0–3.5)	0.18 (0–3.5)	0.19 (0–3.1)	0.842
Sore throat	38 (14)	18 (13)	20 (15)	0.690	0.23 (0–4.5)	0.20 (0–4.5)	0.27 (0–4.5)	0.408
Cough	32 (12)	17 (12)	15 (11)	0.740	0.23 (0–5.3)	0.27 (0–5.0)	0.19 (0–5.3)	0.387
Muscle pain/soreness	18 (7)	14 (10)	4 (3)	0.016	0.09 (0–5.3)	0.16 (0–5.3)	0.02 (0–1.0)	0.015
Dizziness	16 (6)	12 (9)	4 (3)	0.042	0.06 (0–3.5)	0.08 (0–3.5)	0.04 (0–3.1)	0.318
Fever	10 (4)	8 (6)	2 (1)	0.103	0.03 (0–2.0)	0.05 (0–1.4)	0.02 (0–2.0)	0.227
Rash	11 (4)	4 (3)	7 (5)	0.343	0.05 (0–3.5)	0.03 (0–1.4)	0.07 (0–3.5)	0.264
Drowsiness (severe)	8 (3)	4 (3)	4 (3)	0.999	0.02 (0–1.0)	0.02 (0–1.0)	0.01 (0–0.8)	0.695
Ringing in your ears	7 (3)	6 (4)	1 (1)	0.120	0.04 (0–5.2)	0.08 (0–5.2)	0.01 (0–0.9)	0.095
Vomiting	9 (3)	8 (6)	1 (1)	0.036	0.03 (0–1.4)	0.04 (0–1.4)	0.01 (0–1.4)	0.099
Bloody diarrhoea	6 (2)	3 (2)	3 (2)	0.999	0.02 (0–1.4)	0.02 (0–1.4)	0.01 (0–0.9)	0.851
Confusion	3 (1)	3 (2)	0 (0)	0.247	0.01 (0–1.0)	0.02 (0–1.0)	0 (0–0)	0.103
Joint pain/swelling/redness	3 (1)	3 (2)	0 (0)	0.247	0.02 (0–2.5)	0.03 (0–2.5)	0 (0–0)	0.125
Depression	5 (2)	3 (2)	2 (1)	0.999	0.02 (0–1.4)	0.02 (0–1.4)	0.01 (0–1.0)	0.528
Sought medical care during travelb	5 (2)	3 (2)	2 (2)	0.999	0.02 (0–1.1)	0.02 (0–1.1)	0.01 (0–0.8)	0.640
Started a new medication during travelb	105 (39)	57 (42)	48 (36)	0.304	0.52 (0–7.0)	0.49 (0–7.0)	0.55 (0–7.0)	0.659
Ate any of the following during travelb
Fermented or pickled foods	33 (12)	16 (12)	17 (13)	0.817	0.12 (0–5.2)	0.09 (0–2.3)	0.16 (0–5.2)	0.270
Yogurt	58 (21)	30 (22)	28 (21)	0.816	0.28 (0–7.0)	0.30 (0–7.0)	0.26 (0–5.3)	0.761

BSS = bismuth subsalicylate

^aCompared using Wilcoxon rank-sum test.

^bCompared using Chi-square test (with Fisher Exact test applied when the counts were < 5).

^cCompared using two-sample t-test.

Ancillary analyses

There was also no statistical difference among groups for symptoms of loose stool and/or diarrhoea when assessing pre-COVID pandemic and post-COVID pandemic travellers (Supplementary Table 1) or those who reported black stool and those who did not (Supplementary Table 2). Additional analyses were performed assessing the number of participants with TD reported during travel stratified by reported full daily adherence (Supplementary Table 3) and the number of participants in the treatment group with TD reported during travel by the participants’ reported daily adherence (Supplementary Table 4); there were no significant differences.

Harms

No severe adverse events (e.g. salicylism, neurotoxicity) were reported. Within 7 days of return from travel, travellers reported a variety of symptoms, and few took medications or sought medical care (Table 3).

Table 3.

Symptoms, medications, and medical care seeking since return from travel, 2018–2023, (N = 270)

	All participants N = 270	BSS n = 136 (50%)	Placebo n = 134 (50%)	p-value a
Symptoms since return
Dark or black stool	45 (17)	39 (29)	6 (4)	< 0.001
Loose stools<24 h	42 (16)	23 (17)	19 (14)	0.536
Flatulence	40 (15)	19 (14)	21 (16)	0.694
Constipation	30 (11)	13 (10)	17 (13)	0.414
Bloating	24 (9)	13 (10)	11 (8)	0.697
Congestion/runny nose	23 (9)	10 (7)	13 (10)	0.489
Diarrhoea<24 h	24 (9)	14 (10)	10 (7)	0.414
Combined: loose stool plus diarrhoea	59 (22)	33 (24)	26 (19)	0.334
Headache	18 (7)	13 (10)	5 (4)	0.055
Abdominal cramps	17 (6)	13 (10)	4 (3)	0.026
Abdominal pain	10 (4)	4 (3)	6 (4)	0.539
Cough	17 (6)	11 (8)	6 (4)	0.222
Sore throat	17 (6)	11 (8)	6 (4)	0.222
Dark or black tongue	11 (4)	11 (8)	0 (0)	0.001
Faecal urgency	12 (4)	6 (4)	6 (4)	0.979
Nausea	7 (3)	3 (2)	4 (3)	0.721
Muscle pain/soreness	7 (3)	6 (4)	1 (1)	0.120
Drowsiness (severe)	5 (2)	3 (2)	2 (1)	0.999
Depression	4 (1)	2 (1)	2 (1)	0.999
Ringing in your ears	4 (1)	3 (2)	1 (1)	0.622
Vomiting	4 (1)	2 (1)	2 (1)	0.999
Confusion	2 (1)	0 (0)	2 (1)	0.245
Dizziness	2 (1)	1 (1)	1 (1)	0.999
Rash	2 (1)	2 (1)	0 (0)	0.9498
Fever	1 (0)	0 (0)	1 (1)	0.496
Joint pain/swelling/redness	1 (0)	1 (1)	0 (0)	0.999
Bloody diarrhoea	0 (0)	0 (0)	0 (0)	—
Medications taken since return
Oral/IV antibiotics	7 (3)	3 (2)	4 (3)	0.721
Antidiarrheals	4 (1)	1 (1)	3 (2)	0.368
Antimotility agents	3 (1)	3 (2)	0 (0)	0.247
H2 blocker	2 (1)	1 (1)	1 (1)	0.999
Proton pump inhibitor	5 (2)	3 (2)	2 (1)	0.999
Other antacids	3 (1)	3 (2)	0 (0)	0.247
Probiotics	13 (5)	7 (5)	6 (4)	0.797
Oral/intravenous steroids	0 (0)	0 (0)	0 (0)	—
Local/natural/herbal remedies	3 (1)	1 (1)	2 (1)	0.621
Other	25 (9)	17 (12)	8 (6)	0.064
Sought medical care since return	4 (1)	1 (1)	3 (2)	0.367

BSS = bismuth subsalicylate

^aChi-square test was used to compare arms (Fisher Exact test applied when the counts were < 5).

Discussion

This study provides the first new data since the 1980s about the potential use of BSS in the prevention of TD and an assessment of TD rates among international travellers in the modern era. Despite this study not reaching the targeted sample size, there are nonetheless important considerations for healthcare providers conducting pretravel consultations for international travellers.

There was no difference between the percentage of travellers in the BSS or placebo groups with regards to reported occurrence of TD (20% vs 19%, respectively). The study sizing was based on historical placebo TD rates ranging from 38% to 64%.¹¹ However, the placebo TD rate observed in this study was 19% which affected the study’s ability to discriminate a treatment effect for BSS. Although participant’s reported adherence was high, it is possible that travellers did not take every dose of the medication as reported (since not all participants in the BSS arm reported black stool [a frequent result of BSS use]); however, over two-thirds of the BSS arm participants did report common side effects of BSS (e.g. black stool), suggesting it was at least taken regularly.

Most enrolled travellers in this study were predominantly White and non-Hispanic, travelling for leisure, and were enrolled at a fee-for-service travel health center before travel from New York City. These demographics contrast with those from previous studies on BSS and TD prevention, some of which enrolled students after their arrival to Mexico.⁸ It is unclear why these findings are different from prior studies. The difference may be due to host factors (e.g. anatomy, stress); however, travellers were excluded if they had known medication use or conditions that are known to impact TD or affect BSS. Destination factors, including destination countries and which pathogens the travellers may have been exposed to (especially given the timeframe between the current and prior studies), may have played a role; however, BSS has been shown to be effective against growth of Escherichia coli, Salmonella, Shigella, and other TD-causing bacteria.^13,14 Prior studies were predominantly conducted in Latin America, while this study focused on Asia and Africa. There are no studies to data on BSS use in travellers to Asia and limited data on travellers to Africa.¹¹ Diarrheal rates may vary by population or reason for travel, as prior studies involved students, while this study included large numbers of tourists. Although our study did not demonstrate a difference in pre-versus post-COVID pandemic diarrhoea rates, further study is needed to determine if hygiene-related practices (e.g. increased hand washing) may impact TD acquisition. However, raised public awareness of TD risks (e.g. street food, non-potable water), increased availability of personal hygiene items (e.g. hand sanitizers), and improved food preparation standards in the last few decades may also explain the lower incidence of TD seen in this study.

The findings of this study, although conducted robustly, should not be viewed as evidence in isolation because of sample size constraints (it was not powered to detect a difference), and previous studies (that were adequately powered and reached target sample size) found that BSS was effective. Based on the results of this study in concert with historical findings of other studies, healthcare practitioners may still consider recommending prophylactic use of BSS to prevent TD among international travellers to countries with a high incidence of TD while further studies are conducted to determine if there is benefit in specific traveller groups or under certain circumstances. A larger study might show benefit, although the benefit may be small given the results of this study.

BSS is generally well tolerated and has minimal side effects.^10,11 The most common symptoms reported in the BSS group (dark/black stool, flatulence, loose stool, and dark/black tongue) were benign and transient. Reported compliance with BSS was high, suggesting that adherence may not be an issue if it is prescribed.

More research is needed in modalities that aim to prevent TD. This study may also generate hypotheses that could help guide future study of possible uses of BSS in reducing the development of and treating TD.

This study has a few limitations. The study did not achieve the target sample size, and the TD observed rate was about half that expected (40%). Taken in combination, these limited the study’s power to detect a treatment effect, and we cannot make definitive conclusions about the ability of BSS to prevent TD; findings suggest the impact, if any, is likely not large. Participants may have had difficulty recounting symptoms for the post-travel questionnaire, which was administered within 7 days of travel. Participants in the BSS arm of the study experienced side effects of BSS (i.e. dark stool) and the appearance of these effects may have caused participants to suspect they were receiving BSS and not placebo and may have influenced responses to the questionnaire. The results from this study may not be generalizable to all international travellers due to various host factors and the exclusion of travellers going to regions outside of Asia and Africa. Participants were recruited from clinics providing specialized fee-for-service pre-travel medical advice and not all travellers seek this care; travellers who seek pre-travel care may be more likely to comply with safe food/water recommendations or medication adherence.