Effects of an urban sanitation intervention on childhood enteric infection and diarrhoea in Mozambique

Background. Onsite sanitation serves more than 740 million people in urban areas, primarily in low-income countries. Although this critical infrastructure may play an important role in controlling enteric infections in high-burden settings, its health impacts have never been evaluated in a controlled trial. Methods. We conducted a controlled before and after trial to evaluate the impact an onsite urban sanitation intervention on the prevalence of bacterial and protozoan infection (primary outcome), soil transmitted helminth (STH) re-infection, and seven-day period prevalence of diarrhoea among children living in informal neighborhoods of Maputo, Mozambique. A non-governmental organization replaced existing shared latrines in poor condition with engineered pour-flush toilets with septic tanks serving household clusters. We enrolled children aged 1-48 months at baseline and measured outcomes before the intervention and at 12 and 24 months following intervention. We measured outcomes concurrently among children served by the sanitation improvements and those in a comparable control arm served by existing poor sanitation. The trial was registered at ClinicalTrials.gov, number NCT02362932. Findings. At baseline, we enrolled 454 children from 208 intervention clusters and 533 children from 287 control clusters. We enrolled or re-visited 462 intervention and 477 control children 12 months 60 after intervention and 502 intervention and 499 control children 24 months after intervention. Despite nearly exclusive use of the intervention, we found no evidence that engineered onsite sanitation affected the overall prevalence of any measured bacterial or protozoan infection (12-month adjusted prevalence ratio 1.05, 95% CI [0.95-1.16]; 24-month adjusted prevalence ratio 0.99, 95% CI [0.91-1.09]), any STH re-infection (1.11 [0.89-1.38]; 0.95 [0.77-1.17]), or diarrhoea (1.69 [0.89-3.21]; 0.84 [0.47-1.51]) after 12 or 24 months of exposure. Among children born into study sites after the intervention and measured at the 24-month visit, we observed a reduced prevalence of any STH re-infection of 49% (adjusted prevalence ratio 0.51 [95% confidence interval 0.27 - 0.95]), Trichuris of 76% (0.24 [0.10 - 0.60]), and Shigella infection by 51% (0.49 [0.28-0.85]) relative to the same age group at baseline. Interpretation. The intervention did not reduce the overall prevalence of enteric infection and diarrhoea among all enrolled children but may have substantially reduced the prevalence of STHs and Shigella among children born into clusters with sanitary improvements.


Introduction
Participants 160 We enrolled eligible children at three time points: baseline (0 months), 12 months post-161 intervention, and 24 months post-intervention. Children aged 1-48 months old were eligible for 162 baseline enrolment if we received written informed consent from a parent or guardian and if the 163 head of the compound provided verbal assent for the compound to be included in the study. 164 Children were eligible for enrolment at 12-and 24-month visits if they were aged 1-48 months or 165 if they were eligible for enrolment at baseline but absent during that study visit. Children who 166 moved into the compound fewer than six months before the 12-month or 24-month visit were not 167 eligible for enrolment during that phase given their limited exposure to their new compound. 168 Procedures 169 Trained field enumerators completed consent procedures and surveys in the participant's preferred 170 language (Portuguese or Changana) and collected biological specimens from enrolled children 171 (supplemental information p 9). At baseline, we aimed to visit intervention compounds two weeks 172 prior to the opening of the new latrines. We scheduled follow-up visits to be 12 months (±2 weeks) 173 and 24 months (±2 weeks) from the date compound members began using their new latrines, with 174 visits to control compounds made concurrently (±2 weeks). 175 We collected stool specimens independently of reported symptomology. If we were unable to 176 collect a stool sample after multiple attempts, a registered nurse collected a rectal swab after 177 obtaining written consent for the procedure from a parent or guardian. Stool samples were kept 178 cold and delivered to the Laboratory of Molecular Parasitology at the Instituto Nacional de Saúde 179 (INS) within six hours of collection for analysis and storage at -80°C. 180 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020 Samples were shipped frozen with temperatures monitors to the Georgia Institute of Technology 181 (Atlanta, USA) where we used the xTAG Gastrointestinal Pathogen Panel (Luminex Corp,Austin,182 USA), a qualitative multiplex molecular assay, to detect 15 enteric pathogens in stool samples: 183 Campylobacter jejuni/coli/lari; Clostridium difficile, toxin A/B; enterotoxigenic Escherichia coli For the 12-month analysis, we pre-specified the primary outcome as infection with one or more of 201 the 12 bacterial or protozoan enteric pathogens detected by the GPP and secondary outcomes as 202 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020 re-infection with one or more STH as detected by Kato-Katz (following albendazole treatment at 203 baseline), and seven-day period prevalence of caregiver-reported diarrhoea. All three outcomes 204 were considered secondary outcomes in the 24-month analysis. We defined diarrhoea as the 205 passage of three or more loose or liquid stools in a 24-hour period or any stool with blood. (Arnold 206 et al., 2013) We excluded viral enteric pathogens from the primary outcome definition. The 207 intervention may not have interrupted virus transmission due to their low infectious doses, high 208 concentration shed in faeces and extended period of shedding, environmental persistence, and 209 capability for direct person-to-person transmission. (Julian, 2016) Following reported specificity 210 issues with the Salmonella target of the GPP, we removed it from our GPP-based outcome 211 definitions. (Duong et al., 2016;Kellner et al., 2019) In addition to the pre-specified outcomes, we 212 evaluated the effect of the intervention on specific pathogen types (bacterial, protozoan, viral) and 213 on individual pathogens. 214

Statistical analysis 215
Our sample size calculation has been described previously. (Brown et al., 2015) We used an 216 intention-to-treat approach for our analysis. We included all enrolled children at each visit and 217 analysed data as repeated cross-sectional observations. We examined the effect of the intervention 218 at the 12-month and 24-month phases separately. We conducted two sets of exploratory sub-group 219 analyses. The first assessed the effect of the intervention on children with repeat observations at 220 baseline and 12-months and at baseline and 24-months visits. This longitudinal analysis also 221 served as a sensitivity analysis of the impact of participant migration on effect estimates. The 222 second sub-group analysis compared children who were born into study sites after the intervention 223 (or after baseline in controls) but before the 12-month or 24-month visit with children of a similar 224 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020 age distribution at baseline. This allowed us to explore whether exposure to the intervention from 225 birth would reduce enteric pathogen infection during the first 1-2 years of life. 226 We used a difference-in-difference (DID) approach to assess the impact of the intervention on all 227 outcomes at the 12-and 24-month visits. We used generalized estimating equations (GEE) to fit 228 Poisson regression models with robust standard errors. Our GEE models accounted for clustering 229 at the compound level because it was the highest level of nested data and the level of the 230 intervention allocation. We estimated the effect of the intervention as the interaction of variables 231 representing treatment status (intervention versus control) and phase (pre-or post-intervention). 232 Therefore, effect estimates from our DID analysis are presented as ratio measures (ratio of 233 prevalence ratios) instead of absolute differences. Multivariable models were adjusted for 234 covariates determined a priori as potentially predictive of our outcomes, including child age and 235 sex, caregiver's education, and household wealth. Additional covariates (supplemental 236 information p 10-11) were considered for inclusion in multivariable models if they were 237 imbalanced between arms at baseline (>0.1 standardized difference in prevalence or mean) and 238 resulted in a meaningful change in the DID effect estimate (±10% change in 12-month DID 239 prevalence ratio). We used the same statistical approach for sub-group analyses. All analyses 240 were performed on complete case data, and missing data tables are presented in the supplemental 241 information (p 12-13). We performed all statistical analyses with Stata version 16 (StataCorp,242 College Station, USA). 243 The trial was pre-registered at ClinicalTrials.gov (NCT02362932). 244 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020 Role of the Funding Source 245 The funders had no role in study design, data collection and analysis, decision to publish, or 246 preparation of the manuscript. The corresponding author had full access to all study data and had 247 final responsibility for the decision to submit for publication. 248 and control compounds were enrolled at approximately the same rate during each phase 255 (supplemental information p 2-3). Due to migration out of the compound, we collected longitudinal 256 data from 62% of children (59% controls, 67% interventions) between baseline and 12-month and 257 51% of children (46% controls, 58% interventions) between baseline and 24-month. 258 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020

263
At baseline enrolment, intervention compounds had more residents, households, and on-premise 264 water taps than controls, though the number of shared latrines was similar (Table 1). Animals were 265 observed in over half of compounds. Intervention and control households had similar wealth 266 scores, though intervention households had more members and were more crowded while control 267 households more often had walls made of sturdy materials. All households used a municipal water 268 tap as their primary drinking water source with 78% reporting use of a tap on the compound 269 grounds. At baseline, latrines used by intervention households more often had pedestals or slabs, 270 drop-hole covers, and sturdy walls compared with controls. Baseline characteristics of intervention 271 and control children were similar: the average age at enrolment was 23 months (SD = 13), 51% 272 were female, and 32% were still breastfeeding (Table 1). 273 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020

277
We detected ≥1 bacterial or protozoan enteric pathogen in 78% (591/753) of stools and ≥1 STH in 278 45% (308/698) of stools at baseline. The prevalence of predefined outcomes, individual pathogens, 279 and pathogen types were similar between arms at baseline ( Table 2). The prevalence of most 280 bacterial, protozoan, and STH infections increased with age while the prevalence of enteric viruses 281 decreased with age (supplemental information p 14-16). 282 The characteristics of children with repeated observations (including baseline) were similar to 283 characteristics of children measured at baseline only and similar to characteristics of children 284 measured at 12-month and/or 24-month only with the exception of age-related characteristics 285 (supplemental information p 17-26). Over half of the children enrolled after baseline were born 286 into study sites (336/622 [54%], Figure 1). 287 Our main analyses included observations from all eligible children enrolled at baseline (mean 288 sampling age 664 days, SD=393) and the 12-month (940 days, SD=498) and 24-month (1137 days, 289 SD=603) follow-up visits. We found no evidence that the intervention had an effect on the 290 prevalence of any bacterial or protozoan infection (primary outcome, adjusted PR 1·05, 95% CI 291 [0·95 -1·16]), or any STH re-infection (secondary outcome, 1·11 [0·89 -1·38]) 12 months after 292 implementation ( Table 2). The prevalence of diarrhoea remained fairly constant in both arms in 293 all three phases with the exception of the 12-month measure in the control arm which was lower, 294 resulting in a larger effect estimate with low precision (1·69 [0·89-3·21]). 295 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020

21
*Analysis includes all children measured at baseline and 12-month visits. ⁑Analysis includes all children measured at baseline and 24-month visits. ‡Outcome was 297 pre-specified in trial registration. All other outcomes are exploratory. †Pathogen outcomes adjusted for child age and sex, caregiver's education, and household 298 wealth index. Reported diarrhoea was also adjusted for baseline presence of a drop-hole cover and reported use of a tap on compound grounds as primary drinking 299 water source. Sample sizes for adjusted analyses are slightly smaller than numbers presented in prevalence estimates due to missing covariate data. Y. enterocolitica, 300 V. cholerae, E. histolytica, and rotavirus A detected in <2% of samples in each arm at each phase. Descriptive data for these pathogens are available in the 301 supplemental information.

302
The intervention had no relevant effect at 12 months on the prevalence of infection with any of the 303 three pathogen types measured by the GPP (bacterial, protozoan, viral), pathogen coinfection, or 304 on any individual pathogen (Table 2). There was poor precision in the effect estimates for 305 infrequently detected pathogens, evident from their wide confidence intervals. Therefore, some 306 estimates suggestive of a large protective or detrimental effect (Campylobacter, C. difficile, E. coli 307 O157, STEC, Norovirus GI/GII, Adenovirus 40/41) may have arisen by chance. While the NDC 308 provided albendazole to all compound members following baseline, during 12-month visitation 309 only 58% of caregivers (56% control, 60% intervention) confirmed that their child was dewormed 310 during these visits. A sensitivity analysis restricted to children confirmed to have been dewormed 311 produced similar results to the main analysis (supplemental information p 27). 312 There was no evidence that the intervention had an effect on the prevalence of any bacterial or 313 protozoan infection, any STH reinfection, or diarrhoea after 24-months (Table 2). We also found 314 limited evidence of effect on the prevalence of any pathogen type or coinfection with ≥2 GPP 315 pathogens 24-months after intervention. Results for several individual outcomes were suggestive 316 of a protective (STEC, E. coli O157, Cryptosporidium, STH coinfection) or adverse 317 (Campylobacter, C. difficile) effect but evidence was weak as estimates were accompanied by 318 wide confidence intervals. At 24-months, caregivers confirmed baseline and/or 12-month 319 deworming more frequently for intervention children ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint met our inclusion criteria. Point estimates of effect and associated confidence intervals were 325 largely similar in unadjusted and adjusted models with few exceptions (e.g. ETEC at 24-month) 326 (Table 2). For diarrhoea, two additional variables met our inclusion criteria and were included in 327 adjusted models: presence of a latrine drop-hole cover at baseline and reported use of a water tap 328 located within the compound grounds at baseline. The effect estimates were larger and confidence 329 intervals wider for diarrhoea in adjusted versus unadjusted models in the 12-month and 24-month 330 analyses (Table 2). 331 In sub-group analyses comparing children born into study compounds before the 24-month visit 332 with children of similar ages at baseline (<2 years old), there was suggestive evidence that the 333 intervention reduced the prevalence of infection with any STH by half (n= 522; adjusted 334 prevalence ratio 0·51, [95% CI 0·27 -0·95]), Trichuris by 76% (n=522; 0·24, [0·10 -0·60]), and 335 Shigella by 51% (n=630; 0·49, [0·28 -0·85]) (Table 3). We did not observe similar results among 336 children born into the study by the 12-month visit (supplemental information p 28-29), but the 337 sample size was small, resulting in high uncertainty in effect estimates. 338 Longitudinal sub-group analyses explored the effect of the intervention on children with repeated 339 measures at baseline and 12-month (for unadjusted analyses: n=572 for Kato-Katz outcomes, 340 n=868 for GPP outcomes, and n=1112 for diarrhoea) and at baseline and 24-month (n=402, n=716, 341 n=834). Effect estimates were consistent with results from the main analyses (supplemental 342 information p 30-35) but less precise due to the reduced sample numbers. 343 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. Intervention 18/183 (9·8%) 1/75 (1·3%) 0·13 (0·02 -1·08) 0·12 (0·01 -1·02) Analysis includes children <2 years old at baseline and children born into the study after baseline and <2 years old at 347 the time of the 24-month visit. ‡Outcome was pre-specified in trial registration. All other outcomes are exploratory. 348 †Pathogen outcomes adjusted for child age and sex, caregiver's education, and household wealth index. Reported 349 diarrhoea was also adjusted for baseline presence of a drop-hole cover and reported use of a tap on compound 350 grounds as primary drinking water source. Sample sizes for adjusted analyses are slightly smaller than numbers 351 presented in prevalence estimates due to missing covariate data. ⁂Models would not converge due to sparse data. 352 Y. enterocolitica, V. cholerae, E. histolytica, and rotavirus A were detected in <2% of samples in each arm at each 353 phase and excluded. Descriptive data for these pathogens are available in the supplemental information.

354
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint

355
We found no evidence that this urban, onsite shared sanitation intervention was protective for our 356 pre-specified child health outcomes of enteric infection, STH reinfection, or diarrhoea. We also 357 found no strong evidence that the intervention affected prevalence of any individual pathogen, 358 pathogen type, or coinfection with ≥2 enteric pathogens or STH. In exploratory sub-group 359 analyses, we found suggestive evidence that the intervention may have reduced the prevalence of is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10. 1101/2020 It is unlikely that our findings are due to poor intervention fidelity or use, an issue encountered in 398 trials of rural sanitation interventions. ( to detect an effect on STH outcomes. Sensitivity analyses adjusting for caregiver-confirmed 413 deworming and estimated time between deworming and re-infection measurement produced 414 similar results to the main analysis (supplemental information p 28). 415 There are several important limitations of this study. As the intervention was pre-planned and not 416 implemented by the study team, we could not randomize its allocation, increasing the risk of 417 confounding. We assessed potential confounding variables at baseline and used a DID analysis, 418 which accounts for baseline outcome measures, to limit the effect of unmeasured, residual 419 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint confounding. To assess the validity of the parallel trend assumption, a key assumption of DID 420 analyses, we ran "placebo tests" by replacing outcomes with variables unrelated to the 421 intervention, such as child age, respondent role, and presence of animals. Placebo tests showed no 422 effect of the intervention on these variables. 423 It was not possible to mask participants to their intervention status, and our measure of caregiver-424 reported diarrhoea could be subject to respondent and recall biases. To reduce the risk of 425 respondent bias, the MapSan field enumerator team and implementation team were different, and 426 respondents were not informed explicitly that the MapSan team was evaluating the health effect 427 of the intervention. To limit recall bias, we used a 7-day recall period. Due to the greater than expected losses to follow-up in both study arms, we were not able to follow 431 all children enrolled at baseline through time as expected, but we still achieved our target 432 enrolment numbers due to migration and births into study compounds. We conducted the originally 433 planned longitudinal analysis as a sub-group analysis. It also served as a sensitivity analysis to 434 estimate the impact of migration on our effect estimates. Results from this sub-group analysis were 435 largely similar to results of the main analysis which treated measures as repeated cross-sections, 436 though the reduction in sample size led to wider confidence intervals (supplemental information p 437 30-35). Measures of outcomes and covariates in children with and without repeated measures were 438 mostly similar, further limiting the likelihood that changes in the study population biased our 439 results. 440 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. Results from analyses including and excluding Salmonella were similar. 451 We had limited ability to evaluate the impact of seasonality or weather-related trends on our effect 452 estimates due to drought conditions during the 2015/2016 rainy season. We assessed cumulative 453 30-day rainfall as a potential confounder but excluded it because it did not meet our inclusion 454 criteria for adjusted models. 455 Our results demonstrate that access to hygienic, shared onsite sanitation systems was not sufficient 456 to reduce enteric infection or diarrhoea in children aged 6 years or younger (≤4 at baseline) 12-24 457 months after implementation. Results from our sub-group analysis of children born into 458 intervention sites showed a substantial reduction in the prevalence of any STH, Trichuris, and 459 Shigella infection, suggesting that children may require protection from birth to delay infection. 460 Our results do not suggest that shared sanitation is inadvisable in this setting, as we did not compare 461 against household-level sanitation improvements, nor do they account for the many non-health 462 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint 00-09-00015-00) and the Bill and Melinda Gates Foundation (www.gatesfoundation.org) grant 504 OPP1137224. The funders had no role in study design, data collection and analysis, decision to 505 publish, or preparation of the manuscript. 506

Data sharing 507
De-identified participant data which underlie the results reported in this manuscript will be made 508 available beginning three months after publication and ending 60 months after publication. A 509 corresponding data dictionary, the study protocol, and analysis plan and code will also be 510 available for the stated period. Data will be available to researchers who provide a 511 methodologically sound proposal. Proposals for data use can be submitted to 512 joe.brown@ce.gatech.edu and authors of approved proposals will need to sign a data access 513 agreement. Once approved, requestors will be able to access data on the MapSan trial Open 514 Science Forum website (https://osf.io/p5shk). The published trial protocol can be accessed at: 515 https://bmjopen.bmj.com/content/5/6/e008215 . 516

Competing interests 517
All authors have completed the ICMJE uniform disclosure form 518 at www.icmje.org/coi_disclosure.pdf and declare: no support from any organisation for the 519 submitted work; no financial relationships with any organisations that might have an interest in 520 the submitted work in the previous three years; no other relationships or activities that could 521 appear to have influenced the submitted work. 522 523 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint community-led sanitation intervention on child diarrhoea and child growth in rural Mali: a 649 cluster-randomised controlled trial. The Lancet Global Health, 3(11), e701-e711. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. Consent procedures, survey administration, and specimen collection 9      Table 6: Balance of characteristics measured at 12-month visits between children with repeat observations at BL and 12-month and children with only 12-month observations. 23 Table 8: Sensitivity analysis of deworming scenarios on STH effect estimates 12 and 24 months after the intervention. 27 Table 9: Effect of intervention on bacterial, protozoan, and STH infection and reported diarrhoea in children born into study sites post implementation (post-baseline) and before 12-month visit compared with children of a similar age at baseline (<1 year old). 28 Table 10: Effect of the intervention on children with repeated observations at baseline and 12-month visit. 30 Table 11: Effect of the intervention on children with repeated observations at baseline and 24-month visit. 33 Table 12: Sensitivity analysis assessing impact of independent upgrading of control sanitation facilities on effect estimates. 36 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint Consent procedures, survey administration, and specimen collection 704 Enumerators visited households with enrolled children at least twice at each point of follow-up. On the first visit of 705 each phase enumerators completed consent procedures, administered child-, household-, and compound-level 706 surveys, and delivered stool specimen collection supplies. The child's mother was the target respondent for child 707 and household surveys, though the father or another guardian was also eligible. For compound-level surveys, the 708 head of the compound or his or her spouse was the preferred respondent. We sought written, informed consent from 709 the parent or guardian of each eligible child prior to initial enrolment. We sought verbal assent from parents or 710 guardians at each follow-up visit. Consent procedures, surveys, and all study-related verbal communication was 711 performed in Portuguese or Changana as requested by the participant. Written materials were provided in 712 Portuguese. Enumerators provided each caregiver with stool collection supplies, including disposable diapers, a 713 plastic potty if the child was no longer wearing diapers, and a pre-labeled sterile sample bag. Enumerators returned 714 the next day to collect the specimens. If a specimen was unavailable during the scheduled pickup, caregivers called 715 the field team, using phone credit provided by the study, as soon as one was available or if fresh collection supplies 716 were needed. If field enumerators were unable to collect a stool sample after multiple attempts, a registered nurse 717 used an anatomically designed rectal swab (Copan Diagnostics Inc, Murrieta, CA, USA) to collect fecal material.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint *Standardized difference between arms in baseline covariates. † Compared with 12-month unadjusted prevalence ratio. ‡ Compared with 12-month prevalence 747 ratio adjusted for a priori covariates child age, sex, caregiver education, and poverty. ⁑ Compared with 12-month prevalence ratio adjusted for covariates child 748 age, sex, caregiver education, and poverty.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

749
The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020  Results are presented as prevalence (n/N (%)) or mean (standard deviation) at baseline. * Prevalence (or mean (SD)) for children with repeated observations at 757 baseline and 12-month visits. † Prevalence (or mean (SD)) for children with observations from the baseline visit and not the 12-month visit. ‡ Standardized mean 758 difference between observations of children with and without repeated measures at baseline and 12-month visits. ⁑ Could not be calculated. 759 760 20 Results are presented as prevalence (n/N (%)) or mean (  Results are presented as prevalence (n/N (%)) or mean (standard deviation) at 12-month visit. * Prevalence (or mean (  Results are presented as prevalence (n/N (%)) or mean (standard deviation) at 24-month visit. * Prevalence (or mean (SD)) for children with repeated 776 observations at baseline and 24-month visits. † Prevalence (or mean (SD)) for children with observations from the 24-month visit and not the baseline visit. ‡ 777 Standardized mean difference between observations of children with and without repeated measures at baseline and 24-month visits. ⁑ Standardized mean 778 difference between observations from control and intervention children measured at 24-month visit only. estimates at 24-month are adjusted for baseline and/or 12-month deworming confirmation. ‡Analysis is restricted to children whose caregivers confirmed 783 baseline deworming. ⁑ Adjusted for time between 12-month deworming and 24-month sample collection, with deworming considered as intention-to-treat and 784 time broken into 3 intervals: 0-3 months, 4-6 months, and >6 months. NDC 12-month deworming activities at the end of the 12-month phase instead of 785 concurrent to 12-month sample collection resulting in some variation in the amount of time between 12-month deworming and 24-month sample collection 786 among participants. All samples during 12-month phase were collected >6 months after deworming and no adjustment for time since deworming was made. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

787
The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint 30 Analysis includes children with complete observations at baseline and 12-month visits. Results are presented as prevalence (n/N (%)). †Pathogen outcomes 797 adjusted for child age and sex, caregiver's education, and household wealth index, reported diarrhoea also adjusted for baseline presence of a drop-hole cover and 798 reported use of a tap on compound grounds as primary drinking water source. ‡ Models would not converge due to sparse data. 799 800 33 Analysis includes children with complete observations at baseline and 24-month visits. Results are presented as prevalence (n/N (%)). †Pathogen outcomes 802 adjusted for child age and sex, caregiver's education, and household wealth index, reported diarrhoea also adjusted for baseline presence of a drop-hole cover and 803 reported use of a tap on compound grounds as primary drinking water source. ‡ Models would not converge due to sparse data. 804 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. 13a For each group, the numbers of participants who were randomly assigned, received intended treatment, and were analysed for the primary outcome For each group, the numbers of clusters that were randomly assigned, received intended treatment, and were analysed for the primary outcome  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10.1101/2020.08.20.20178608 doi: medRxiv preprint 812 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted August 23, 2020. . https://doi.org/10. 1101/2020