https://orcid.org/0000-0003-1085-1354
Figures
Abstract
Background
Early childhood enteric infection with Shigella/EIEC, enteroaggregative E. coli (EAEC), Campylobacter, and Giardia has been associated with reduced child growth, yet a recent randomized trial of antimicrobial therapy to reduce these infections did not improve growth outcomes. To interrogate this discrepancy, we measured the enteric infections from this study.
Methods
We leveraged the Early Life Interventions for Childhood Growth and Development in Tanzania (ELICIT) trial, a randomized double-blind placebo-controlled trial of antimicrobial therapy with azithromycin and nitazoxanide provided quarterly to infants from 6 to 15 months of age. We tested 5,479 stool samples at time points across the study for 34 enteropathogens using quantitative PCR.
Results
There was substantial carriage of enteropathogens in stool. Azithromycin administration led to reductions in Campylobacter jejuni/coli, enteroaggregative E. coli, and Shigella/EIEC (absolute risk difference ranged from -0.06 to 0.24) 2 weeks after treatment however there was no effect after 3 months. There was no difference in Giardia after nitazoxanide administration (ARR 0.03 at the 12 month administration). When examining the effect of azithromycin versus placebo on the subset of children infected with specific pathogens at the time of treatment, a small increase in weight-for-age Z score was seen only in those infected with Campylobacter jejuni/coli (0.10 Z score, 95% CI -0.01–0.20; length-for-age Z score 0.07, 95% CI -0.06–0.20).
Conclusion
The antimicrobial intervention of quarterly azithromycin plus or minus nitazoxanide led to only transient decreases in enteric infections with Shigella/EIEC, enteroaggregative E. coli (EAEC), Campylobacter, and Giardia. There was a trend towards improved growth in children infected with Campylobacter that received quarterly azithromycin.
Citation: Guga G, Houpt ER, Elwood S, Liu J, Kimathi C, Mosha R, et al. (2023) Impact of azithromycin and nitazoxanide on the enteric infections and child growth: Findings from the Early Life Interventions for Childhood Growth and Development in Tanzania (ELICIT) trial. PLoS ONE 18(12): e0294110. https://doi.org/10.1371/journal.pone.0294110
Editor: David Joseph Diemert, George Washington University School of Medicine and Health Sciences, UNITED STATES
Received: July 11, 2023; Accepted: October 15, 2023; Published: December 21, 2023
Copyright: © 2023 Guga et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: Bill & Melinda Gates Foundation OPP1141342. Under the grant conditions of the Foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript” to the Funding Statement.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Early enteric infections in children in low-resource settings are an established risk factor for poor growth [1]. A detailed analysis of 35,622 stool samples from infants in 8 low resource settings found that substantial decrements in length-for-age Z score at 2 years were associated with intestinal carriage of specifically Shigella/EIEC, enteroaggregative E. coli (EAEC), Campylobacter, and Giardia, with smaller decrements associated with Cryptosporidium, norovirus, typical enteropathogenic E. coli, and Enterocytozoon bieneusi [2]. Whether any of these associations are causal is unclear. Strategies to reduce enteric infections, including through water, sanitation and hygiene interventions, have struggled to improve child growth [3] yet have also failed to reduce bacterial and protozoal pathogen burden [4, 5].
This background formed the rationale for the Early Life Interventions for Childhood Growth and Development in Tanzania (ELICIT) trial, a randomized double-blind placebo-controlled trial of antimicrobial therapy provided to infants [6]. The intervention included azithromycin because it has activity against most Shigella, Campylobacter, and enteroaggregative E. coli. For instance, a 3 day course of azithromycin administered to Indian infants more than halved the prevalence of these 3 bacterial pathogens in stool at a 2 week time point [7]. The intervention also included nitazoxanide to provide anti-protozoal activity. For instance, nitazoxanide reduced Giardia cysts detected by stool microscopy in a majority of Peruvian children at a 7–10 day time point [8]. Azithromycin was provided single dose at 6, 9, 12, and 15 months of age while nitazoxanide was given as a 3-day course at 12 and 15 months (nitazoxanide is not approved until 1 year of age). Importantly, the primary outcome of ELICIT has been reported and showed no effect of the antimicrobial intervention on child length at 18 months of age [9] nor any effect on cognitive scores [10]. ELICIT was a 2 x 2 factorial design and included a nutritional intervention of nicotinamide, which could also affect enteropathogens through regulation of gut dysbiosis and secretion of antimicrobial peptides [11, 12].
Here, we examine the results of qPCR testing for 34 enteric pathogens on stool samples collected at five timepoints. Specifically, we tested stool collected at 6 months (prior to the first dose of azithromycin or placebo), 6.5 months (two weeks after receiving azithromycin or placebo), 12 months (3 months after receiving azithromycin or placebo), 12.5 months (2 weeks after receiving azithromycin and nitazoxanide or placebo) and 18 months (3 months after receiving azithromycin and nitazoxanide or placebo). This sampling time frame allowed us to assess the extent to which the antimicrobial intervention had an effect on enteric infections and to interpret the lack of effect of this antimicrobial regimen on childhood growth.
Methods
Study enrollment and interventions
The ELICIT study methods have been previously reported in detail [6, 9]. Enrollment occurred between September 5, 2017 and August 31, 2018. Briefly, pregnant women and mothers of newborns were approached at their homes by field team members to inform them about the study and assess interest. Inclusion criteria were maternal age ≥18 years, child age ≤14 days, and the family’s stated intent to reside in the study area for the duration of the study. Exclusion criteria were multiple gestation, significant birth defect or neonatal illness, infant enrollment weight <1,500 g, and lack of intent to breastfeed. Overall 1,188 children were enrolled and allocated to one of 4 treatment groups in a 2 × 2 factorial manner in permuted blocks of eight, such that participants received either antimicrobials plus placebo, placebo plus nicotinamide, both active medications, or both placebos. Azithromycin was provided single dose (20 mg/kg) at 6, 9, 12, and 15 months of age while a 3-day course of nitazoxanide (100 mg twice daily) was provided at 12 and 15 months. The study protocol was approved by the National Institute for Medical Research (NIMR) of Tanzania (NIMR/HQ/R.8a/Vol.IX/2424 on 3 Mar 2017) and the Tanzanian Food and Drug Administration (TFDA; TFDA 0017/CTR/0005/02 on 5 July 2017)) and the Institutional Review Board at the University of Virginia (IRB-HSR #19465 on 3 May 2017). Written consent was obtained from all parents/guardians.
Data and sample collection
Use of non-study antibiotics by study children, defined as any antibiotic that was not part of the study intervention, was identified using a standardized questionnaire at monthly visits. Child lengths using measuring boards and weights using digital scales were measured every three months. Stool samples were collected at 6, 6.5, 12, 12.5, and 18 months of age, and 1,141 children contributed at least 1 sample. The 6 and 12 month samples were required to be collected prior to administration of study antimicrobials. The study protocol was approved by the National Institute for Medical Research (NIMR) of Tanzania and the Tanzanian Food and Drug Administration (TFDA) and the Institutional Review Board at the University of Virginia. Mothers gave written informed consent to participate either during pregnancy or at the time of enrollment. The study was registered at ClinicalTrials.gov: NCT03268902.
Stool testing
Procedures for sample extraction and testing have been previously detailed [13, 14] and the protocols are available at dx.doi.org/10.17504/protocols.io.5qpvo3k8xv4o/v1. We used custom-designed TaqMan Array Cards (ThermoFisher, Carlsbad, CA, USA) that compartmentalized probe-based quantitative PCR assays for 34 enteropathogens (S1 Table). A cycle threshold of 35 was used as the limit of detection and any detections above that were considered positive. Samples were stored at –80°C before extraction. Bacteriophage MS2 and phocine herpesvirus were used as external controls to monitor efficiency of nucleic acid extraction and amplification. We included one extraction blank per batch and one no-template amplification control per ten cards to exclude laboratory contamination.
Statistical analysis
Pathogens with an overall prevalence of at least 5% were included in the analyses of pathogen carriage. At a prevalence of 5% in the placebo arm, we had 80% power to detect a prevalence in the antimicrobial arm of 1.8%, a 3.2% difference; at a prevalence of 20%, we had 80% power to detect a prevalence of 13.6% in the antimicrobial arm, a 6.4% difference. To assess whether missing pathogen data were associated with treatment arm, we fit a logistic regression model with treatment arm as the predictor, adjusted for collection time. To estimate risk differences, we used g-computation via the R package riskCommunicator with Poisson regression, with clustering by individual to account for repeated measures, to estimate the effect of treatment arm on the absolute difference in pathogen detection at each time point [15]. Poisson regression was used as an approximation of log-binomial regression [16]. Confidence intervals were estimated by bootstrap with 1000 replicates. The exposure was coded as an interaction between the antimicrobial treatment arm and the time of sample collection, and we adjusted for the treatment arm assignment for the second study intervention (nicotinamide vs. placebo). To estimate the association between exposure to non-study antibiotics in the prior month and pathogen detection, we used the same modeling approach and outcome but with an interaction between exposure to any non-study antibiotics in the prior month and sample collection timepoint, adjusting for treatment arm assignment for the antimicrobial study intervention.
To assess whether the benefit of treatment was restricted to children who had an enteric infection, we evaluated whether the effect of treatment arm on short-term changes in growth (LAZ and WAZ) differed depending on baseline detection of each pathogen in the 6- or 12-month (pretreatment) stool samples. For this analysis, we restricted to children who had complete follow-up including growth outcomes at 18 months of age. We fit a generalized linear model with a Gaussian distribution to estimate the difference in length at 9 and 15 months (three months after antibiotic treatment) between treatment arms in subgroups defined by baseline detection of each pathogen. Adjustment covariates included pretreatment length and calendar month of birth based on previously observed differences in growth by birth month [9]. Pathogens were selected for this analysis based on a previous analysis of pathogens associated with growth deficits in a multisite study that included this study site [2]. To estimate the power for this analysis, we estimated an interclass correlation coefficient of 0.05 based on our observed sample, a standard deviation of 1 and an N of 1140 in each group (given each child was included at two time points), which would have 80% power to detect a 0.269 change in LAZ or WAZ for a pathogen with a prevalence of 20% and a 0.540 change in LAZ or WAZ for a pathogen with a prevalence of 5%. As a sensitivity analysis, we also restricted to children who had stool samples available at both time points (6 and 12 months).
We then estimated the association between antimicrobial treatment arm and detection of the macrolide resistance conferring genes mphA and ermB at each timepoint using Poisson regression with generalized estimating equations and adjusting for treatment arm assignment for the nicotinamide study intervention. We used geepack [https://www.jstatsoft.org/article/view/v015i02] for these regression models with generalized estimating equations. All analyses were conducted using R version 4.2.1.
Results
The baseline characteristics of the children in this study are summarized in Table 1. Of the 1,188 children in the ELICIT study, 1,141 contributed at least 1 stool sample for testing, of which there were 570 children in the antimicrobial arm and 571 that received placebo. Quantitative PCR testing for 34 pathogens was performed on 5,479 stool samples from collected at 6, 6.5, 12, 12.5, and 18-months. Receipt of external non-study antibiotic was common. For instance, 54–86% of children received some non-study antibiotic during the month prior to the 6, 12, and 18 month stool collections. There was no association between either treatment arm or receipt of non-study antibiotics and missing pathogen data, thus we performed complete case analyses. Length-for-age and weight-for age Z scores at 6 and 12 months were low but similar between the arms.
There was substantial detection of enteric pathogens as we have noted previously [2], including of enteroaggregative E. coli (EAEC), typical enteropathogenic E. coli (EPEC), Campylobacter, Shigella, Giardia, Cryptosporidium, and Enterocytozoon bieneusi (Fig 1). Shigella and Giardia carriage increased substantially in the second year of life, as expected. Pathogens with an overall prevalence of at least 5% were included in further analysis.
Y axis shows prevalence or proportion positive. EAEC = enteroaggregative E. coli; EIEC = enteroinvasive E. coli; LT-ETEC = heat labile toxin producing E. coli; ST-ETEC = heat stabile toxin producing E. coli; tEPEC = typical enteropathogenic E. coli.
We examined the effect of receiving study antimicrobial versus placebo on enteropathogen carriage. Fig 2 shows that receiving study antimicrobial substantially reduced detection of several pathogens in stool samples collected two weeks following the 6 or 12 month administrations. This included Campylobacter jejuni/coli at 6.5 and 12.5 months (absolute risk difference (ARD) -0.24; 95% CI: -0.26, -0.21 at 6.5 months and -0.28; -0.31, -0.22 at 12.5 months), tEPEC at 6.5 months (ARD -0.11; 95% CI: -0.15, -0.07), EAEC at 6.5 months (ARD -0.06; 95% CI: -0.11, - 0.02), and Shigella/EIEC at 12.5 months (ARD -0.16; 95% CI -0.20, -0.13). No significant difference was observed in carriage of LT-ETEC, ST-ETEC, or STEC 2 weeks after azithromycin (i.e., at the 6.5 or 12.5 month time points) nor of Cryptosporidium or Giardia 2 weeks after nitazoxanide (i.e., the 12.5 month time point). Moreover, at 3 months after antimicrobial administration (i.e., at the 12 and 18 month time points) none of the differences in Campylobacter, tEPEC, EAEC, or Shigella/EIEC carriage persisted. Stools were also tested in the children in the nicotinamide-receiving arms however no significant change in pathogens were observed in those versus placebo at any time point (S1 Fig).
X axis is the absolute difference in pathogen prevalence between the antibiotic and placebo arms. Negative values to the left of 0.0 represent lower carriage in the active arm.
Because children frequently received non-study antibiotics, we examined the effect of receiving such antibiotics on pathogens as well. Penicillins were the most frequently used class, followed by metronidazole, sulfonamides, and macrolides (S2 Fig). Use of non-study antibiotics in the month prior to stool sample collection was associated with a reduced risk of Shigella/EIEC (ARR −0.03; 95% CI: −0.04, −0.01) and STEC (ARR −0.06; 95% CI: −0.09, −0.03) at the 6 month time point, Giardia (ARR -0.10; 95% CI: -0.17, -0.02) at 12 months, and E. bieneusi (ARR −0.06; 95% CI: −0.11, −0.01) at 18 months (Fig 3).
X axis is the absolute difference in those that received non-study antibiotics during the month prior to 6, 12, and 18 months versus those that did not receive non-study antibiotics. Negative values to the left of 0.0 represent lower carriage with receipt of non-study antibiotics.
We examined whether a growth benefit was observed with study antimicrobial in the children who carried the enteric pathogens that have most consistently been associated with growth deficits, namely Shigella/EIEC, enteroaggregative E. coli (EAEC), Campylobacter jejuni/coli, Giardia, and Cryptosporidium [2]. There were no statistically significant differences in length-for-age or weight-for-age Z scores three months after receiving study antimicrobials in infected children compared to the children not infected with these pathogens (Fig 4), but there was a marginal 0.10 (95% CI -0.01–0.20) increase in weight-for-age Z score in children infected with Campylobacter jejuni/coli that received antimicrobials. There was also an increase in length-for-age Z score in children with Campylobacter jejuni/coli that received antimicrobials however this was less statistically significant (0.07, 95% CI -0.06–0.20). The results were almost identical in a sensitivity analysis that only included children with stool PCR results at both time points.
Shown are the subset of children at 6 and 12 months with Shigella/EIEC, enteroaggregative E. coli (EAEC), Campylobacter, Giardia, or Cryptosporidium detected compared with those without the pathogen detected. The X axis is the difference in the length (left) or weight (right) for age Z score in the antimicrobial arm–the Z score in the placebo arm, such that a difference in Z score above 0.0 represents an increase in child length or weight with antimicrobials.
We also tested for carriage of antimicrobial resistance genes by qPCR within this customized TAC card. Receiving antimicrobials rather than placebo slightly increased the risk of detecting the macrolide resistance genes mphA and ermB two weeks after azithromycin administration at 6 months (relative risk (RR) for mphA = 1.11, 95% CI 1.05, 1.18; for ermB = 1.02, 95% CI 1.00–1.04) and 12 months (RR for mphA = 1.19, 95% CI 1.12, 1.26; for ermB = 1.03, 95% CI 1.00–1.06). These differences disappeared at 12 months (i.e., 3 months after administration).
Discussion
The ELICIT trial administered single-dose azithromycin plus or minus nitazoxanide quarterly to infants under the hypothesis that antibiotic treatment of the intestinal infections associated with poor growth could reduce child stunting. In the primary analysis no effect on childhood growth or stunting was seen at 18 months of age [9]. The present work analyzed 5,479 stool samples in order to better interpret the findings and assess the hypothesis.
First, we did observe a clear and large effect of azithromycin on Campylobacter jejuni/coli and Shigella/EIEC, with an absolute difference in Shigella/EIEC carriage of about ~16% and Campylobacter jejuni/coli of about ~24–28% of children two weeks after administration. There was also some clearance seen of EAEC and tEPEC albeit of smaller magnitude. In an earlier study of Indian infants treated for 3 days with azithromycin we found greater reductions in EAEC, EPEC, Campylobacter, and Shigella/EIEC of about 50% [7] Prior work from Niger has also shown that Shigella was reduced by azithromycin mass drug administration to a greater extent than seen here [17]. Therefore, while reductions in pathogens were observed with azithromycin, it was lower than expected, particularly for diarrheagenic E. coli.
It was surprising to not see an effect of azithromycin against ETEC, as this antibiotic clearly has shown an effect on ETEC diarrhea [18]. It was also surprising that that there was no effect ascribable to nitazoxanide against Giardia or Cryptosporidium. It is possible that the high use of external non-study antimicrobials in the community obscured the study’s antimicrobial effects. For instance, the two most common non-study antibiotics were penicillins and metronidazole, and these could have “pre-reduced” Shigella/EIEC and Giardia, respectively, such that additional study azithromycin or nitazoxanide had a limited ability to show impact. It is also possible than antimicrobial resistance to azithromycin played some role, as rates of resistance to fecal E. coli in this study were previously noted to be 23–32% [9]. Perhaps the most important explanation for the low efficacy of study antimicrobials, however, was the high force of re-infection with these enteric infections from the environment. Notably, at 3 months after antimicrobials all of the earlier reductions were lost.
Therefore the underlying hypothesis of the ELICIT study was not able to be completely tested, as it remains possible that a better antimicrobial regimen that better and more durably reduces the pathogens associated with poor growth (Shigella/EIEC, EAEC, Giardia, and Campylobacter) could reduce stunting. In this setting however, it appears that this would require frequent dosing of azithromycin and is probably impractical.
There was a suggestion of improved weight-for-age Z score and length-for-age Z score in the subset of children infected with Campylobacter jejuni/coli that received azithromycin. It is certainly plausible that certain pathogens associated with poor growth are more important or causal (such as Campylobacter) than others, and if Campylobacter could be selectively targeted this would be important for further study. However the observed effect sizes were small (< = 0.10 Z score) and thus even if true may not translate into significant improvements in child health. Of course it is possible that more complete removal of this or other pathogens from the gut could have greater effects.
Limitations of this work include that our sampling frame only tested stools at 6 monthly time points so we do not know the kinetics of how long the azithromycin effect lasted or how quickly reinfections occurred. Additionally, the PCR approach could detect prolonged shedding of pathogens killed after antibiotics which could also obscure an antimicrobial effect.
In sum, the antimicrobial regimen tested in this study was not as effective as expected in treating the candidate enteropathogens. We do not believe that periodic azithromycin mass drug administration is the proper tool to improve childhood growth, at least in this setting, since the aggressive regimens of antimicrobial required to reduce enteric infections would carry complex risks and antimicrobial resistance. The suggestion that reducing Campylobacter could improve childhood growth should be pursued in future studies. If true, since Campylobacter is such a highly prevalent enteric pathogen, an effective and narrow-spectrum method to reduce this infection could be beneficial and should be studied given the intractable problem of childhood stunting.
Supporting information
S1 Table. Pathogens and gene targets tested by PCR in this study.
https://doi.org/10.1371/journal.pone.0294110.s001
(DOCX)
S1 Fig. Difference in pathogen detection in the nicotinamide versus placebo arms at 6, 6.5, 12, 12.5, and 18 months.
X axis is the absolute reduction in pathogen prevalence between the nicotinamide–placebo. Negative values to the left of 0.0 represent reduction with nicotinamide. 95% confidence intervals are included.
https://doi.org/10.1371/journal.pone.0294110.s004
(DOCX)
S2 Fig. Prevalence of non-study antibiotic use by drug class and month.
https://doi.org/10.1371/journal.pone.0294110.s005
(DOCX)
References
- 1. MAL-ED_Investigators. Relationship between growth and illness, enteropathogens and dietary intakes in the first 2 years of life: findings from the MAL-ED birth cohort study. BMJ Glob Health. 2017;2(4):e000370. pmid:29333282
- 2. Rogawski ET, Liu J, Platts-Mills JA, Kabir F, Lertsethtakarn P, Siguas M, et al. Use of quantitative molecular diagnostic methods to investigate the effect of enteropathogen infections on linear growth in children in low-resource settings: longitudinal analysis of results from the MAL-ED cohort study. Lancet Glob Health. 2018;6(12):e1319–e28. pmid:30287125
- 3. Pickering AJ, Null C, Winch PJ, Mangwadu G, Arnold BF, Prendergast AJ, et al. The WASH Benefits and SHINE trials: interpretation of WASH intervention effects on linear growth and diarrhoea. Lancet Glob Health. 2019;7(8):e1139–e46. pmid:31303300
- 4. Rogawski McQuade ET, Platts-Mills JA, Gratz J, Zhang J, Moulton LH, Mutasa K, et al. Impact of Water Quality, Sanitation, Handwashing, and Nutritional Interventions on Enteric Infections in Rural Zimbabwe: The Sanitation Hygiene Infant Nutrition Efficacy (SHINE) Trial. J Infect Dis. 2020;221(8):1379–86. pmid:31004129
- 5. Grembi JA, Lin A, Karim MA, Islam MO, Miah R, Arnold BF, et al. Effect of water, sanitation, handwashing and nutrition interventions on enteropathogens in children 14 months old: a cluster-randomized controlled trial in rural Bangladesh. J Infect Dis. 2020;227(3):434–47. pmid:32861214
- 6. DeBoer MD, Platts-Mills JA, Scharf RJ, McDermid JM, Wanjuhi AW, Gratz J, et al. Early Life Interventions for Childhood Growth and Development in Tanzania (ELICIT): a protocol for a randomised factorial, double-blind, placebo-controlled trial of azithromycin, nitazoxanide and nicotinamide. BMJ Open. 2018;8(7):e021817. pmid:29982218
- 7. Grassly NC, Praharaj I, Babji S, Kaliappan SP, Giri S, Venugopal S, et al. The effect of azithromycin on the immunogenicity of oral poliovirus vaccine: a double-blind randomised placebo-controlled trial in seronegative Indian infants. Lancet Infect Dis. 2016;16(8):905–14. pmid:27156189
- 8. Ortiz JJ, Ayoub A, Gargala G, Chegne NL, Favennec L. Randomized clinical study of nitazoxanide compared to metronidazole in the treatment of symptomatic giardiasis in children from Northern Peru. Aliment Pharmacol Ther. 2001;15(9):1409–15. pmid:11552913
- 9. DeBoer MD, Platts-Mills JA, Elwood SE, Scharf RJ, McDermid JM, Wanjuhi AW, et al. Effect of scheduled antimicrobial and nicotinamide treatment on linear growth in children in rural Tanzania: A factorial randomized, double-blind, placebo-controlled trial. PLoS Med. 2021;18(9):e1003617. pmid:34582462
- 10. Parpia T, Svensen E, Elwood S, Wanjuhi A, Blacy L, Bayo E, et al. Cognitive Outcomes at 18 Months: Findings from the Early Life Interventions for Childhood Growth and Development in Tanzania (ELICIT) Trial. Am J Trop Med Hyg. 2021;106(2):441–5. pmid:34872064
- 11. Hashimoto T, Perlot T, Rehman A, Trichereau J, Ishiguro H, Paolino M, et al. ACE2 links amino acid malnutrition to microbial ecology and intestinal inflammation. Nature. 2012;487(7408):477–81. pmid:22837003
- 12. Liang H, Dai Z, Kou J, Sun K, Chen J, Yang Y, et al. Dietary l-Tryptophan Supplementation Enhances the Intestinal Mucosal Barrier Function in Weaned Piglets: Implication of Tryptophan-Metabolizing Microbiota. Int J Mol Sci. 2018;20(1). pmid:30577574
- 13. Liu J, Gratz J, Amour C, Nshama R, Walongo T, Maro A, et al. Optimization of Quantitative PCR Methods for Enteropathogen Detection. PLoS One. 2016;11(6):e0158199. pmid:27336160
- 14. Liu J, Platts-Mills JA, Juma J, Kabir F, Nkeze J, Okoi C, et al. Use of quantitative molecular diagnostic methods to identify causes of diarrhoea in children: a reanalysis of the GEMS case-control study. Lancet. 2016;388(10051):1291–301. pmid:27673470
- 15. Grembi JA, Rogawski McQuade ET. Introducing riskCommunicator: An R package to obtain interpretable effect estimates for public health. PLoS One. 2022;17(7):e0265368. pmid:35849588
- 16. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–6. pmid:15033648
- 17. Platts-Mills JA, Ayoub EG, Zhang J, McQuade ETR, Arzika AM, Maliki R, et al. Impact of Biannual Mass Azithromycin Treatment on Enteropathogen Carriage in Children <5 Years Old in Niger. Clin Infect Dis. 2022;75(3):515–8.
- 18. Sanders JW, Frenck RW, Putnam SD, Riddle MS, Johnston JR, Ulukan S, et al. Azithromycin and loperamide are comparable to levofloxacin and loperamide for the treatment of traveler’s diarrhea in United States military personnel in Turkey. Clin Infect Dis. 2007;45(3):294–301. pmid:18688944