Sand barriers around latrine pits reduce fecal bacterial leaching into

3 days ago - Research of the year C&EN's most popular stories of the year Molecules of the year Science that delighted... SCIENCE CONCENTRATES ...
0 downloads 0 Views 573KB Size
Subscriber access provided by Iowa State University | Library

Ecotoxicology and Human Environmental Health

Sand barriers around latrine pits reduce fecal bacterial leaching into shallow groundwater: a randomized controlled trial in coastal Bangladesh Abu Mohd Naser, Solaiman Doza, Mahbubur Rahman, Kazi Matin Uddin Ahmed, Mohammed Shahid Gazi, Gazi Raisul Alam, Mohammed Rabiul Karim, Golam Kibria Khan, Mohammed Nasir Uddin, Mohammed Ilias Mahmud, Ayse Ercumen, Julia Rosenbaum, Jonathan Annis, Stephen P. Luby, Leanne Unicomb, and Thomas F. Clasen Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b04950 • Publication Date (Web): 17 Jan 2019 Downloaded from http://pubs.acs.org on January 18, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

Environmental Science & Technology

1

Sand barriers around latrine pits reduce fecal bacterial leaching into shallow groundwater:

2

a randomized controlled trial in coastal Bangladesh

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Abu Mohd Naser1,2*, Solaiman Doza3, Mahbubur Rahman3, Kazi Matin Uddin Ahmed4, Mohammed Shahid Gazi3, Gazi Raisul Alam3, Mohammed Rabiul Karim3, Golam Kibria Khan3, Mohammed Nasir Uddin3, Mohammed Ilias Mahmud5, Ayse Ercumen6,7, Julia Rosenbaum8, Jonathan Annis9, Stephen P. Luby10, Leanne Unicomb3, Thomas F. Clasen1 1Department

of Environmental Health Sciences, Rollins School of Public Health, Emory University, Atlanta, GA, 30322, United States 2 Emory Global Diabetes Research Center, Hubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, GA, 30322, United States 3International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b), Dhaka, Dhaka, 1212, Bangladesh 4Department of Geology, University of Dhaka, Dhaka, 1000, Bangladesh 5Department of Geology and Mining, University of Barisal, Barisal, 8200, Bangladesh 6Division of Epidemiology & Biostatistics, University of California Berkeley School of Public Health, Berkeley, CA, 94720, United States 7Department of Forestry and Environmental Resources, North Carolina State University, Raleigh, NC, 27695, United States 8FHI360, Washington, DC, 20009, United States 9Tetra Tech, Arlington, VA, 22202, United States 10Woods Institute for the Environment, Stanford University, Stanford, CA, 94305, United States

*Corresponding author: Abu Mohd Naser, MPH, PhD Rollins School of Public Health, Emory University 1518 Clifton Rd NE, CNR 7040-G Atlanta, GA 30322, USA email: [email protected]; Cell: 470-279-1624

1

ACS Paragon Plus Environment

Environmental Science & Technology

33

Page 2 of 29

Abstract

34

We evaluated the effectiveness of a sand barrier around latrine pits in reducing fecal indicator

35

bacteria (FIB) leaching into shallow groundwater. We constructed 68 new off-set single pit pour

36

flush latrines in Galachipa sub-district of coastal Bangladesh. We randomly assigned 34 latrines

37

to include a 50-cm thick sand barrier under and around the pit and 34 received no sand barrier.

38

Four monitoring wells were constructed around each pit to collect water samples at baseline and

39

subsequent nine follow-up visits over 24 months. Samples were tested using the IDEXX Colilert

40

method to enumerate E. coli and thermotolerant coliforms most probable number (MPN). We

41

determined the difference in mean log10MPN FIB counts/100ml in monitoring well samples

42

between latrines with and without a sand barrier using multilevel linear models and reported

43

cluster robust standard error. The sand barrier latrine monitoring well samples had 0.38 mean

44

log10MPN fewer E. coli (95% CI: 0.16, 0.59; p = 0.001) and 0.38 mean log10MPN fewer

45

thermotolerant coliforms (95% CI: 0.14, 0.62; p = 0.002), compared to latrines without sand

46

barriers, a reduction of 27% E. coli and 24% thermotolerant coliforms mean counts. A sand

47

barrier can modestly reduce the risk presented by pit leaching.

2

ACS Paragon Plus Environment

Page 3 of 29

48

Environmental Science & Technology

Abstract Art

49

50

3

ACS Paragon Plus Environment

Environmental Science & Technology

51

Page 4 of 29

Introduction

52

Improved sanitation is the primary barrier to human fecal organisms contaminating the

53

environment. Pit latrines are the most commonly used human excreta disposal systems in low-

54

income countries.1 Globally, an estimated 1.8 billion people use pit latrines as the primary means

55

of sanitation.2 There are concerns, however, about the extent to which pit latrines may introduce

56

fecal contamination into shallow aquifers and pose a health risk to households relying on shallow

57

groundwater reserves for drinking and livelihood activities. Pit latrines that are not shared with

58

other households count toward the Sustainable Development Goals (SDG) sanitation target and

59

are considered a solution for “safely managed sanitation” if excreta is routinely emptied,

60

transported and treated off-site.3 However, there is evidence that leaching from the latrine pit can

61

occur prior to emptying, particularly when constructed in high-water table environments.2 In

62

many rural areas, pit latrines are located in close proximity to community groundwater sources

63

used for drinking and household chores.2 Latrine pits are either unlined or lined with bricks,

64

concrete rings or other materials with gaps in the liner to allow liquids to leach out into

65

surrounding soil. Latrine leachate is a known source of groundwater contamination.4

66

Bangladesh has seen dramatic increases in reduction of open defecation over the past

67

decade. Direct pour-flush latrine with a one-meter pit containing four concrete rings and a plastic

68

pan fitted into a concrete slab is the most common latrine technology in rural Bangladesh. Much

69

of the southern coastal Bangladesh is at sea level and groundwater table levels are between 1-4

70

meters below the surface.5 Latrines in southern coastal Bangladesh are often installed in a way

71

that untreated effluent discharges directly into unconsolidated shallow groundwater.6,7 Flooding

72

due to heavy rainfall, tidal surges or cyclones increases the threat of surface and groundwater

73

contamination from traditional pit latrines.8,9 4

ACS Paragon Plus Environment

Page 5 of 29

74

Environmental Science & Technology

Slow or intermittent sand filters have been used for centuries to treat wastewater from

75

sewer systems and septic tanks.10 Generally, sand filters are superior to native soils at filtering

76

septic effluent.11 In areas where naturally occurring underlying soils are ill-suited for percolation

77

of septic effluent, the construction of sand-lined trench has proven a viable alternative.11 When

78

septic effluents percolate into the sand, a layer of bacteria (biomat) is formed in the sand surface

79

exposed to the effluent.12 Once formed, the biomat reduces bacteria concentration in the filtered

80

effluent; it can operate long-term if the hydraulic conductivity within the biomat is maintained.13

81

We hypothesized that sand barriers constructed around pit latrines could act like a sand filter and

82

reduce the spread of bacteria from latrine pits in high groundwater areas. Currently sand barriers

83

surrounding the pit latrines are not used in Bangladesh, and there is a lack of field research data

84

to fully understand whether sand filters can reduce pathogen leaching from pit latrines into

85

shallow groundwater in tropical climates. We evaluated whether constructing a 50-cm sand

86

barrier below and around the walls of an offset pour-flush latrine pit reduced leaching of fecal

87

bacteria into the surrounding environment in a high groundwater coastal region of Bangladesh.

88 89

.

5

ACS Paragon Plus Environment

Environmental Science & Technology

Methods

90 91

Page 6 of 29

Study site- The study was conducted in Galachipa Upazilla (sub-district), Patuakhali

92

district, Bangladesh (Figure 1). Galachipa (between 21°48' and 22°21' north latitudes and 90°15'

93

and 90°37' east longitudes; area 1268 square km) is a low-lying coastal sub-district with many

94

rivers and canals. The communities in these areas have limited livelihood opportunities; this is

95

aggravated by regular flooding, poor infrastructure and underdeveloped communication systems.

96

The groundwater table level is variable in Galachipa depending on rainfall and tidal patterns, but

97

is typically very shallow, 5 meters from existing unimproved latrines, and (iii) the

119

land donated by the household for pit latrine construction was not adjacent to surface water

120

bodies. A statistician generated a unique household ID along with block randomization and

121

prepared sealed-envelopes coded for a latrine with a sand barrier and without a sand barrier.

122

Block randomization was chosen to ensure equal numbers of latrines with and without a sand

123

barrier among the geographically clustered households to minimize confounding from local

124

geological factors, and to ensure an equal number of latrines with and without a sand barrier in

125

case construction crews had to stop installing latrines due to flooding and heavy rainfall during

126

the wet season.

127 128

Latrine, sand barrier and monitoring well construction- We selected three local contractors

129

to construct the latrines, sand barriers and monitoring wells and closely supervised their work to

130

conform with specifications (Supporting Information1). For all 68latrines, contractors used

131

five concrete liner rings of 300 mm height for the pit (Figure S1-S4). For both latrines with and

132

without sand barrier, they also placed four monitoring wells at even intervals around the pit at a

133

distance of 1 meter from the outer side of the concrete rings. One meter distance was considered

134

due to space limitation in the homesteads of the participating households. All monitoring wells 7

ACS Paragon Plus Environment

Environmental Science & Technology

135

had 6-meter depths  the top 3 meters comprised a PVC casing and the bottom 3 meters

136

included a screen for the accumulation of groundwater from the surrounding soil layers. The top

137

of the monitoring well was encased with a 0.3 meter x 0.3 meter concrete pad and capped to

138

firmly secure the well in place and prevent any surface water runoff from leaking (Supporting

139

Information1). For the latrines with a sand barrier, the contractors constructed a 50 cm sand

140

barrier using locally available sand around and below the concrete rings. Therefore, the distance

141

between outer side of the sand barrier and monitoring wells were 50 cm for the sand barrier

142

latrines. The cost for materials for construction was US$ 282 per sand barrier latrine, and US$

143

257 per latrine without sand barrier.

Page 8 of 29

144 145

We randomly selected two sand barrier latrines from each of the four unions (geographical

146

administrative unit) and collected the sand samples used for the barriers for sieve analysis at the

147

laboratory of Department of Geology, Dhaka University to determine the textural composition.

148

Field staff also collected soil samples from the bottom of all 68 study latrines during pit hole

149

excavation for analysis at the laboratory of Department of Geology, Dhaka University.

150

Hydrometer analysis was used to determine the distribution of the finer soil particles at the

151

bottom of the pits and assess whether the soil texture varied geographically across the study area.

152 153

Blinding-The data collectors and the investigators were blinded to the household intervention

154

status during study implementation. We had two groups of field staff. The first team was

155

involved in securing informed consent, ensuring construction of the latrines as per

156

randomization, and monitoring well installation. After construction, we replaced this team with a

157

second team to collect study data. 8

ACS Paragon Plus Environment

Page 9 of 29

Environmental Science & Technology

158 159

Baseline water sample collection: Baseline groundwater samples were collected from

160

monitoring wells of all 68 study latrines before household members started using the latrine

161

(Supporting Information2). Field staff purged each monitoring well 24 hours prior to sample

162

collection by pumping out well water using a portable hand pump until clear water was

163

discharged. 18,19 Purging was done in order to obtain aquifer water samples that were not affected

164

by the conditions created by the well.18 They used a sterile disposable 1.9cm polyethylene bailer

165

for collecting 500 ml water samples from each monitoring well.

166 167

Promotion of latrine use: We selected 11 local community health promoters (CHPs) and

168

trained them to deliver messages to target household members. They promoted regular latrine

169

use and maintenance, and promoted protection of the monitoring wells from children and

170

household activities. CHPs instructed households to demolish any previously used latrines. CHPs

171

visited participating households weekly to deliver latrine use and maintenance messages. CHPs

172

were not apprised of the study objectives or methods.

173 174

Follow-up water sample collection: Following baseline groundwater sample collection, field

175

staff collected water samples from each monitoring well of all 68 study latrines in months 1

176

through 5, and then at 12, 15, 18 and 24 months after initiation of latrine use using the procedure

177

described above for the baseline sample (Figure 2).

178

Water sample transport, process and analysis- Upon collection, water samples were

179

immediately poured into sterile Whirlpak® bags and transported on ice in an insulated cool box

180

(at 2 to 8°C) to a field laboratory for processing within six hours of collection. 100 mL aliquots 9

ACS Paragon Plus Environment

Environmental Science & Technology

Page 10 of 29

181

of samples were processed with Colilert®-18 media and incubated for 18 hours at 44.5°C to

182

enumerate the most probable number (MPN) of E. coli and thermotolerant coliforms using the

183

IDEXX QuantiTray®-2000 MPN table based on the number of fluorescent and yellow cells. 20,21

184

One lab blank was run per day of sample testing. Field blanks were collected once per week per

185

sample collector.

186

Data analysis- We replaced the E. coli and thermotolerant coliforms MPN values where no

187

contamination was detected with 0.5 (half the lower detection limit) and then converted the

188

counts into log10MPN for analysis. We used multilevel mixed effect linear models using two

189

random effects to address clustering at latrine and monitoring well levels to determine the

190

difference in mean log10MPN counts between water samples from latrines with and without a

191

sand barrier. In secondary analyses, we investigated the effect modification of sand barrier on

192

water quality by including an interaction term with season (dry vs. wet) in the regression

193

models. Follow-up visits conducted between May and November were considered as wet season

194

visits. 14We categorized water sample contamination according to the WHO thresholds of low

195

risk (100 MPN/100 ml).22 We used the cumulative approach of ordered

197

logistic regression23 with the same multilevel mixed effect approach to determine the odds ratio

198

of higher WHO risk categories between latrines with and without a sand barrier. In all models,

199

we reported the cluster robust standard error. We performed statistical analyses in Stata, version

200

15.0.

201 202

Ethics- Selected households were informed about our study objectives and their right to discontinue participation at any point of the study period. Informed written consent was taken 10

ACS Paragon Plus Environment

Page 11 of 29

Environmental Science & Technology

203

from all household heads prior to installing the latrines. The study protocol was approved by the

204

Ethical Review Committee of icddr,b (PR-14117) and FHI360.

11

ACS Paragon Plus Environment

Environmental Science & Technology

Results

205 206

Page 12 of 29

Pre-intervention characteristics of study latrines: The mean members per household was

207

five for latrines with and without sand barrier (Table 1). Seventy-six percent of the sand barrier

208

latrine households had an unimproved private latrine compared to 79% of households without

209

sand barrier. The median nearest distance between the study latrine and a previous unimproved

210

latrine was 18 meters in both groups. A surface water source was present within 10 meters of

211

50% of constructed sand barrier latrines, and 56% of latrines without sand barrier. We collected

212

1,220 monitoring well water samples from the sand barrier latrines, and 1,212 from latrines

213

without a sand barrier (Figure 2). Four monitoring wells were damaged and we were unable to

214

collect water samples from them in the last four visits.

215

Soil composition of latrines and sand barriers: Latrines with and without sand barriers

216

had similar soil compositions at the bottom of the pits (Supporting Information Figure S5).

217

The median proportion of silt was 86% for latrines with sand barrier and 84% for latrines without

218

sand barrier and the median proportion of clay was 11.5% for both groups (Supporting

219

Information Figure S5). The median grain size of the pit bottom soil was 0.12 mm for latrines

220

with a sand barrier and 0.11 mm for latrines without a sand barrier. Sieve analysis suggests

221

similarly homogenous composition of the locally produced sand barrier in all four unions. The

222

proportion of sand was 87 – 93% and the grain size was 0.09 – 0.12 mm (Supporting

223

Information Table S1).

224

Fecal contamination of water from monitoring well: The pre-intervention (baseline)

225

mean log10MPN E. coli per 100 ml was 2.3 for latrines with sand barrier and 2.0 for latrines

226

without (Figure 3). The pre-intervention mean log10MPN thermotolerant coliforms per 100 mL

227

was 2.6 for latrines with sand barrier and 2.4 for latrines without. Thereafter, mean log10MPN E. 12

ACS Paragon Plus Environment

Page 13 of 29

Environmental Science & Technology

228

coli and thermotolerant coliforms counts in water samples from latrines with sand barrier were

229

lower than E. coli and thermotolerant coliforms counts in samples from latrines without sand

230

barrier (Figure 3). The proportions of water samples < 1 log10MPN/100 mL were 28% (95% CI:

231

25.1, 30.2) for latrines with sand barrier and 21% (18.3, 22.9) for latrines without for E. coli; the

232

corresponding proportions for thermotolerant coliforms were 22% (95% CI: 20.0, 24.8) and 17%

233

(14.6, 18.8) (Table 2). The proportions of water samples >10 log10MPN/100 mL were 48%

234

(45.4, 51.0) for latrines with sand barrier and 59% (95% CI: 56.3, 62.0) for latrines without for

235

E. coli; the corresponding proportions for thermotolerant coliforms were 54% (95% CI: 50.7,

236

56.4) and 65% (95% CI: 62.6, 68.0) (Table 2).

237

In monitoring wells from latrines with sand barrier, we detected 1.06 mean

238

log10MPN/100 mL E. coli and 1.25 mean log10MPN/100 mL thermotolerant coliforms at follow

239

up visits. In monitoring wells from latrines without sand barrier, we detected 1.43 mean

240

MPN/100 mL E. coli and 1.62 mean MPN/100 mL thermotolerant coliforms during follow-up.

241

Compared to water samples from latrines without sand barrier, those from latrines with sand

242

barrier had a 0.38 (95% CI: 0.16, 0.59; p = 0.001) mean log10MPN reduction in E. coli and 0.38

243

(95% CI: 0.14, 0.62; p = 0.002) mean log10MPN reduction in thermotolerant coliforms in 100 ml

244

water (Table 3), translating to 27% reduction in E. coli and 24% reduction in thermotolerant

245

coliforms provided by the sand barrier.

246

Compared to latrines without sand barrier, latrines with sand barrier had 0.46 (95% CI:

247

0.23, 0.69; p 1 m

322

distance from the pit latrine as households were unwilling to provide extra space due to space

323

limitation in their homesteads. However, a similar study identified that 90% water samples in

324

coastal Bangladesh had fecal contamination when collected two meter laterally from the latrine

325

pit and in 86% samples had fecal contamination when collected 4 meters laterally. 26 With our

326

study design, we were unable to assess whether the reduction in pit latrine leaching would 17

ACS Paragon Plus Environment

Environmental Science & Technology

Page 18 of 29

327

improve the microbiological water quality of the water sources used by the households. We did

328

not assess whether the sand barrier can reduce leaching of chemical contaminants such as

329

nitrogenous and carbon compounds.

330

The SDGs unify sanitation and safe water supply under a common goal; promoting pit

331

latrines with sand barriers can only modestly reduce shallow groundwater contamination. Further

332

modification of the sand barrier such as increasing the thickness or altering sand composition

333

such as modifying pore size may increase its effectiveness and further reduce leaching.

334

Environmental engineering laboratory exploration of such sand barrier modifications could be

335

assessed at low cost prior to field testing. Careful modeling studies to understand the role of pit

336

latrine leaching on community fecal exposure and health burden can inform whether

337

interventions to lower leaching could be a pragmatic approach to support the SDG goals.

18

ACS Paragon Plus Environment

Page 19 of 29

Environmental Science & Technology

338

Acknowledgements: This research was funded by the United States Agency for

339

International Development (USAID) and the Swedish International Development Cooperation

340

Agency (SIDA). icddr,b acknowledges with gratitude the commitment of the funders to its

341

research efforts. icddr,b is also grateful to the Governments of Bangladesh, Canada, Sweden and

342

the UK for providing core/unrestricted support. We are grateful to the study households who

343

provided space for latrine construction. Supporting Information: Schematic cross-sectional and longitudinal view of latrines

344 345

with and without sand barrier. Summary findings on the composition of the bottom soil of

346

latrines with and without sand barrier. Sieve analysis findings of the composition of sands used

347

for constructing sand barriers. Field activities and process for pit latrine, monitoring well

348

installation. Methodologies for water sample collection from monitoring wells. Study dataset

349

(.xls)

350 351 352 353

Sources of Funding: United States Agency for International Development (USAID) and Swedish International Development Cooperation Agency (SIDA).

354 355

Authors conflict of interst: None.

19

ACS Paragon Plus Environment

Environmental Science & Technology

Page 20 of 29

References:

356 357 358

1.

Cairncross S, Bartram J, Cumming O, Brocklehurst C. Hygiene, sanitation, and water: what needs to be done? PLoS Medicine 2010; 7(11) : e1000365.

359 360

2.

Graham JP, Polizzotto ML. Pit latrines and their impacts on groundwater quality: a systematic review. Environmental Health Perspectives 2013; 121(5): 521.

361 362 363 364

3.

World Health Organization, UNICEF. Progress on drinking water, sanitation and hygiene: 2017 update and SDG baselines. 2017. Website; http://apps.who.int/iris/bitstream/handle/10665/258617/9789241512893eng.pdf?sequence=1 (accessed November 24, 2018)

365 366

4.

Tillett T. Pit latrines and groundwater contamination: negative impacts of a popular sanitation method. Environmental Health Perspectives 2013; 121(5): a169.

367 368 369

5.

Zahid A, Ahmed SRU. Groundwater resources development in Bangladesh: Contribution to irrigation for food security and constraints to sustainability. Groundwater Governance in Asia Series-1. 2006: 25-46.

370 371 372

6.

Knappett PS, Escamilla V, Layton A, McKay LD, Emch M, Williams DE, Huq R, Alam J, Farhana L, Mailloux BJ. Impact of population and latrines on fecal contamination of ponds in rural Bangladesh. Science of the Total Environment 2011; 409(17): 3174-82.

373 374 375

7.

Knappett PS, McKay LD, Layton A, Williams DE, Alam J, Huq R, Mey J, Feighery JE, Culligan PJ, Mailloux BJ. Implications of fecal bacteria input from latrine-polluted ponds for wells in sandy aquifers. Environmental Science & Technology 2012; 46(3): 1361-70.

376 377 378

8.

Luby SP, Gupta SK, Sheikh MA, Johnston RB, Ram PK, Islam MS. Tubewell water quality and predictors of contamination in three flood-prone areas in Bangladesh. J Appl Microbiol 2008; 105(4): 1002-8.

379 380

9.

Brammer H. Bangladesh’s dynamic coastal regions and sea-level rise. Climate Risk Management. 2014; 1: 51-62.

381 382

10.

Healy MG, Rodgers M, Mulqueen J. Treatment of dairy wastewater using constructed wetlands and intermittent sand filters. Bioresource Technology 2007; 98(12): 2268-81.

383 384 385

11.

United States EPA. Septic Tank - Soil Absorption Systems. Website; https://www.hgac.com/community/water/ossf/DSTFS_Septic-Tank_Soil-Absorption-Systems.pdf (accessed November 24, 2018)

386 387

12.

Oliver DM, Clegg CD, Haygarth PM, Heathwaite AL. Assessing the potential for pathogen transfer from grassland soils to surface waters. Advances in Agronomy 2005; 85(85): 125-80.

388 389

13.

Beal C, Gardner E, Kirchhof G, Menzies N. Long-term flow rates and biomat zone hydrology in soil columns receiving septic tank effluent. Water Research 2006; 40(12): 2327-38. 20

ACS Paragon Plus Environment

Page 21 of 29

Environmental Science & Technology

390 391 392 393 394

14.

Qureshi AS, Ahmed Z, Krupnik TJ. Groundwater management in Bangladesh: an analysis of problems and opportunities. Cereal Systems Initiative for South Asia -Mechanization and Irrigation, 2014. Website; https://repository.cimmyt.org/xmlui/bitstream/handle/10883/4273/56862.pdf?sequence=1 &isAllowed=y

395 396 397

15.

Ercumen A, Naser AM, Unicomb L, Arnold BF, Colford Jr JM, Luby SP. Effects of sourceversus household contamination of tubewell water on child diarrhea in rural Bangladesh: a randomized controlled trial. PLoS One 2015; 10(3): e0121907.

398 399 400

16.

fhi360. WASHplus: water, sanitation and hygiene programs lead to healthier households and communities. 2016. Website; https://www.fhi360.org/news/water-sanitation-and-hygieneprograms-lead-healthier-households-and-communities (accessed August 06 2018).

401 402

17.

Matin I, Hulme D. Programs for the Poorest: Learning from the IGVGD Program in Bangladesh. World Development 2003; 31(3): 647-65.

403 404

18.

Keely JF, Boateng K. Monitoring Well Installation, Purging, and Sampling Techniques-Part 1: Conceptualizations. Groundwater 1987; 25(3): 300-13.

405 406

19.

Keely JF, Boateng K. Monitoring well installation, purging, and sampling techniques—Part 2: Case histories. Groundwater 1987; 25(4): 427-39.

407 408 409

20.

Yakub GP, Castric DA, Stadterman-Knauer KL, Tobin MJ, Blazina M, Heineman TN, Yee GY, Frazier L. Evaluation of Colilert and Enterolert defined substrate methodology for wastewater applications. Water Environment Research 2002; 74(2): 131-5.

410 411 412 413 414 415

21.

Ercumen A, Pickering AJ, Kwong LH, Mertens A, Arnold BF, Benjamin-Chung J, Hubbard AE, Alam M, Sen D, Islam S, Khalil MMR, Kullmann C, Chase C, Ahmed R, Parvez SM, Unicomb L, Rahman M, Ram P, Clasen TF, Luby SP, Colford JM. Do Sanitation Improvements Reduce Fecal Contamination of Water, Hands, Food, Soil, and Flies? Evidence from a ClusterRandomized Controlled Trial in Rural Bangladesh. Environmental Science & Technology 2018; 52(21): 12089-97.

416 417 418 419

22.

World Health Organization. Guidelines for Drinking-water Quality 2nd Edition. Geneva, Switzerland: WHO; 1997. Website; http://www.who.int/water_sanitation_health/publications/gdwq2v1/en/index1.html (accessed November 24, 2018)

420 421

23.

Fullerton AS. A conceptual framework for ordered logistic regression models. Sociological Methods & Research 2009; 38(2): 306-47.

422 423 424 425

24.

Ercumen A, Pickering AJ, Kwong LH, Arnold BF, Parvez SM, Alam M, Sen D, Islam S, Kullmann C, Chase C, Ahmed R, Unicomb L, Luby SP, Colford JM Jr. Animal feces contribute to domestic fecal contamination: evidence from E. coli measured in water, hands, food, flies, and soil in Bangladesh. Environmental science & technology 2017; 51(15): 8725-34.

21

ACS Paragon Plus Environment

Environmental Science & Technology

Page 22 of 29

426 427 428

25.

Rahe T, Hagedorn C, McCoy E, Kling G. Transport of Antibiotic-resistant Escherichia coli Through Western Oregon Hillslope Soils Under Conditions of Saturated Flow 1. Journal of Environmental Quality 1978; 7(4): 487-94.

429 430 431

26.

Ravenscroft P, Mahmud ZH, Islam MS, Hossain AKMZ, Zahid A, Shaha GC, Zulfiquar Ali AHM, Islam K, Cairncross S, Clemens JD, Islam MS. The public health significance of latrines discharging to groundwater used for drinking. Water Research 2017; 124: 192-201.

432 433 434 435

27.

Benneyworth L, Gilligan J, Ayers JC, Goodbred S, George G, Carrico A, Karim MR, Akter F, Fry D, Donato K, Piya B. Drinking water insecurity: water quality and access in coastal southwestern Bangladesh. International Journal of Environmental Health Research 2016; 26(5-6): 508-24.

436 437

28.

Islam MA, Sakakibara H, Karim MR, Sekine M. Potable water scarcity: options and issues in the coastal areas of Bangladesh. Journal of Water and Health 2013; 11(3): 532-42.

438 439

29.

Ali M, Emch M, Donnay J-P, Yunus M, Sack R. The spatial epidemiology of cholera in an endemic area of Bangladesh. Social Science & Medicine 2002; 55(6): 1015-24.

440 441

30.

Emch M. Diarrheal disease risk in Matlab, Bangladesh. Social Science & Medicine 1999; 49(4): 519-30.

442 443 444 445

31.

Sharmin A. Water and wastewater in Bangladesh, current status and a design of a decentralized solution. Master Thesis, Lund University, Lund, Sweden, 2016. Website; https://lup.lub.lu.se/student-papers/search/publication/8895656 (accessed November 24, 2018)

446 447

22

ACS Paragon Plus Environment

Page 23 of 29

Environmental Science & Technology

448 449

Figure 1: Study sites in coastal Bangladesh.

23

ACS Paragon Plus Environment

Environmental Science & Technology

Page 24 of 29

450 451 452

Figure 2: Trial profile and sample collection during field visits. Four monitoring wells were damaged and we were unable to collect water samples from them in the last four visits

24

ACS Paragon Plus Environment

Page 25 of 29

Environmental Science & Technology

453 454 455

Figure 3: Temporal trend in mean log10MPN thermotolerant coliforms and E. coli in water samples from the latrines with and without sand barrier during all visits

25

ACS Paragon Plus Environment

Environmental Science & Technology

456

Page 26 of 29

Table 1: Pre-intervention characteristics of the households with and without sand barrier latrines Characteristics

Study latrines Sand barrier latrine Latrine without sand barrier

Household members Living in compound, mean (Standard deviation) Living in household, mean (Standard deviation) latrine users per household, median (Inter-quartile range) Household environments

14 (12)

13 (10)

5 (2)

5 (2)

5 (4-6)

5 (4-6)

Households with an existing unimproved latrine, n/N (%) Distance from study latrine to nearest unimproved latrine in meters, median (IQR) Presence of surface water (e.g. pond) within 10 m of study latrine, n/N (%) Reported household feces disposal, n/N (%) Existing pit latrine

26/34 (76)

27/34 (79)

18 (10-20)

18 (10-25)

17/34 (50)

19/34 (56)

22/34 (65)

18/34 (53)

Surface water

9/34 (26)

13/34 (38)

Open place

3/34 (9)

2/34 (6)

Deep tube well

22/34 (65)

26/34 (76)

Shallow tube well

11/34 (33)

8/34 (24)

Primary household drinking water source, n/N (%)

26

ACS Paragon Plus Environment

Page 27 of 29

Environmental Science & Technology

458 459

Table 2: Proportions of shallow groundwater samples for microbiological risks as per WHO categories

460

E. coli WHO Microbiological risk categories Low risk 10 MPN/100 mL, % (95% CI) (n/N) Very high risk >100 MPN/100 ml, % (95% CI) (n/N)

Latrines without sand barrier 21 (18.3 – 22.9) (248/1209) 79 (77.1 – 81.7) (961/1209) 59 (56.3 – 62.0) (716/1209) 34 (31.6 – 37.0) (414/1209)

Sand barrier latrines 28 (25.1 – 30.2) (337/1220) 72 (69.8 – 74.9) (883/1220) 48 (45.4 – 51.0) (588/1220) 23 (20.7 – 25.5) (281/1220)

Thermotolerant coliforms Latrines without sand barrier 17 (14.6 – 18.8) (201/1209) 83 (81.2 – 85.4) (1008/1209) 65 (62.6 – 68.0) (790/1209) 40 (36.7 – 42.3) (478/1209)

Sand barrier latrines 22 (20.0 – 24.8) (270/1220) 78 (75.4 – 80.2) (950/1220) 54 (50.7 – 56.4) (653/1220) 27 (24.8 – 29.8) (333/1220)

27

ACS Paragon Plus Environment

Environmental Science & Technology

461 462

Page 28 of 29

Table 3: Post-intervention mean log10MPN thermotolerant coliforms and E. coli in water samples from latrines with and without sand barriers. Baseline data not included in the analyses. Latrines without sand barrier

Thermotolerant coliforms

N

Difference between latrines with and without sand barrier Sand barrier latrines

mean log10MPN

N

1.62

1220

mean log10MPN

p value

Difference in mean log10MPN₤

95% CI

p value

-0.62, -0.14

0.002

--

--

--

Difference in mean log10MPN*

95% CI

-0.38

Primary analysis 1209

All visits

1.25

Interaction with season Wet season

402

2.15

406

1.96

-0.19

-0.47, 0.10

0.194

ref

ref

ref

Dry season

807

1.36

814

0.89

-0.47

-0.73, -0.22

< 0.001

-0.28

-0.53, -0.02

0.032

1.43

1220

1.06

-0.38

-0.59, -0.16

0.001

--

--

--

E. coli

Primary analysis All visits

1209

Interaction with season Wet season

402

1.95

406

1.74

-0.21

-0.49, 0.07

0.149

ref

ref

ref

Dry season

807

1.17

814

0.71

-0.46

-0.69, -0.23

< 0.001

-0.25

-0.51, 0.01

0.063

* Difference in mean log10MPN refers to comparison between latrines without sand barrier

463 464

₤Difference

in mean log10MPN refers to comparison to reference category

Note- random effect models were used to generate the 95% confidence intervals.

465

28

ACS Paragon Plus Environment

Page 29 of 29

466

Environmental Science & Technology

Table 4: Monitoring well sample WHO risk categories for thermotolerant coliforms and E. coli across latrines with and without sand barriers

467 Indicator bacteria

Latrines without sand barrier

Sand barrier latrines

Difference between latrines with and without sand barrier

N of wells

Low risk

Intermediate risk

High risk

Very high risk

N of wells

Low risk

Intermediate risk

High risk

Very high risk

Change in ordered logodds of higher risk category

E. coli

1209

20.5%

20.3%

25.0%

34.2%

1220

27.6%

24.2%

25.2%

23.0%

-0.51

-0.84, -0.18

0.006

Thermotolerant coliforms

1209

16.6%

18.0%

25.8%

39.5%

1220

22.1%

24.3%

26.2%

27.3%

-0.54

-0.91, -0.17

0.006

95% CI

p value

468

29

ACS Paragon Plus Environment