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