Article pubs.acs.org/est
Cite This: Environ. Sci. Technol. 2018, 52, 14402−14410
Effectiveness of Multilevel Risk Management Emergency Response Activities To Ensure Free Chlorine Residual in Household Drinking Water in Southern Syria Mustafa Sikder,*,† Umar Daraz,‡ Daniele Lantagne,† and Roberto Saltori§ †
Civil and Environmental Engineering, Tufts University, Medford, Massachusetts 02155, United States UNICEF Middle East and North Africa Regional Office, Amman 11181, Jordan § Independent Consultant, Amman 11181, Jordan Environ. Sci. Technol. 2018.52:14402-14410. Downloaded from pubs.acs.org by IOWA STATE UNIV on 01/13/19. For personal use only.
‡
ABSTRACT: To provide safe drinking water and reduce the risk of disease, emergency responders in southern Syria are implementing a multilevel risk reduction strategy with the aim of ensuring free chlorine residual (FCR) in household drinking water. Responders implemented activities across the water chain (from chlorination station and well operators to water vendors to household members), including distribution of supplies for chlorination and training on chlorine use; activities varied by responder. We evaluated the effectiveness of these interventions in a cross-sectional observation study including interviews and observations with 24 chlorination station operators and 63 well owners/managers; interviews, observations, and water quality testing with 220 water truckers; and surveys and water quality testing with 1006 households. Across all responders, activities successfully ensured FCR in household drinking water (61−96% of households with FCR ≥ 0.1 mg/L compared to 21% in nonintervention households, p < 0.001). Centralized interventions led to the highest FCR results. Household FCR was associated with access to piped water systems (aOR 3.5, 95% CI 1.8−6.7) and chlorine distribution (aOR 6.1, 95% CI 3.4−11.0). We recommend continuing activities, emphasizing central-level activities, and supplementing with household-level activities. These results will help to optimize current interventions and guide future response design in similar contexts.
■
INTRODUCTION Access to safe water supply, sanitation, and hygiene (WASH) are immediate priorities for human survival and dignity in emergencies.1 Efficacious, effective, and context-appropriate WASH interventions have been shown to reduce both the risk of disease and risk of disease transmission in emergencies.2−8 WASH interventions are commonly implemented by local governments, United Nations (UN) agencies, and local and international responders. The Syrian emergency presents a unique, yet emerging, context for WASH response. Traditional WASH emergency response activities include (1) increasing water quantity; (2) improving water quality; (3) isolating feces from the environment; (4) promoting (and facilitating with materials) personal hygiene; and (5) reducing environmental risks.1,6,9 However, these interventions, and usual funding levels, are often calibrated on contexts of population displacement and camps. Traditional programs are not easily scalable to cover large resident populations in middle income contexts with local functioning markets and urban settings10 (such as preconflict Syria where the majority of the population had access to piped water supply before the conflict).11,12 To support the affected © 2018 American Chemical Society
population inside Syria, the UN Security Council authorized a cross-border humanitarian assistance program operated from Turkey and Jordan.13 The activities covered by this paper have been carried out in opposition-controlled southern Syria by humanitarian actors based in Amman, Jordan, as part of the humanitarian response mechanism termed Whole of Syria (WoS).13 The two main drinking water sources in southern Syria are piped water from the pre-existing infrastructure network which has been compromised by the conflict, and a privately operated tanker truck network which emerged as a local alternative after the conflict began.10 Households receive water from either or both systems. The piped water system is intermittent, which can negatively impact water quality.14,15 In trucked water, contamination can come from the source, during filling, transport, and distribution.16,17 As households can have piped and/or trucked water, and roof-top water storage tanks, there Received: Revised: Accepted: Published: 14402
June 25, 2018 September 25, 2018 November 13, 2018 November 13, 2018 DOI: 10.1021/acs.est.8b03487 Environ. Sci. Technol. 2018, 52, 14402−14410
Article field supervisors, outreach hygiene promoters field supervisors, outreach hygiene promoters field supervisors, outreach hygiene promoters
no supervisors, fewer promoters than standard-east
get chlorinated water from the vendors
rehab accessible chlorination stations, training operators, and supply liquid chlorine disinfect water with a specific amount of chlorine rehab accessible chlorination stations, training operators, and supply liquid chlorine chlorinate water
activities with network key recommendations by promoters field implementation personal involved
activities with water vendors
rehab accessible chlorination stations, training operators, and supply liquid chlorine disinfect water with a specific amount of chlorine
All vendors received chlorine, tester, water disinfection training, and signed agreement to distribute chlorinated water; local authorities and promoters routinely check FCR in trucked water limited network functionality ( 0.2 mg/L in unadjusted analysis. In the best-fit multivariate mixed-effect logistic model, water collection from public wells serving both piped network and vendors was positively associated with FCR ≥ 0.2 mg/L (aOR 32.6, 95% CI 7.5−141.0) (Table 3). Cleaning the filling-pipe before water collection (aOR 9.8, 95% CI 3.0−32.7) and reported receipt of chlorine from promoters (aOR 6.0, 95% CI 1.8−20.0) were also positively associated with FCR ≥ 0.2 mg/ L in trucked water samples. Household Survey. A total of 1006 households were surveyed from the four intervention, one no-intervention, and one distribution-only groups (Table 4). The average number of people in surveyed households was 6 (range 5−7), with 68% (range = 66−72%) of the households having at least one child < 5 years old, and 9% (range = 2−13%) of households headed by a female. Both reported main water source (piped = 10−95%; trucked = 5−89%) and access to secondary water source (range 10− 54%) in the preceding 30 days before the survey varied among intervention groups (p < 0.001, p < 0.001) (Table 4). Almost all households (96−99%) had a water tap inside the house, which was connected to an average of 1.1−1.8 roof-tanks. The number of households reporting treating water (14−86%) varied among groups (p < 0.001). The reasons for not treating the water in the household also varied among groups, with most commonly reported reason being the water is already treated (29−100%), not having the means to treat water (0− 64%), and not liking chlorine taste (0−29%). The number of households reported receiving drinking water risk mitigation training (2−85%) differed among groups (p < 0.001). Almost all respondents from every group reported drinking water in the household was safe to drink (94−100%) and that chlorine gives a bad taste only when dosage is too high (95−100%). The number of households reporting not being able to drink water with chlorine taste was also high (84−98%) and varied among groups (p < 0.001). Respondents’ knowledge score calculated form the eight questions differed significantly (p < 0.001), however not meaningfully, among the groups, from 3.5 to 5.4). Please note only 6% respondents report unsafe water could cause diabetes. The household drinking water FCR ≥ 0.1 mg/L (21−96%) varied among groups, and all intervention groups had more households with FCR ≥ 0.1 m/L compared to the nointervention group (Figure 1). When we compared the overall group performance adjusting for water source, the water-voucher-east group had the highest adjusted odds ratio (aOR 73.9, 95% CI 33.8−177.9) followed by limited east group (aOR 28.3, 95% CI 15.0−56.0), distribution-only group (aOR 15.2, 95% CI 7.6−31.9), standard-east group (aOR 9.5, 95% CI 5.3−17.5), and standard-west group (aOR 5.6, 95% CI 3.3−10.1). The rooftank FCR ≥ 0.1 mg/L (21−94%) also varied among groups with all interventions more likely to have FCR compared to no-intervention group (p < 0.001 for all). However, within the same household, the drinking water and roof-tank water FCR ≥ 0.1 mg/L did not differ significantly (p = 0.073), suggesting that household chlorination occurred in the roof-top tank, not
Table 3. Water Trucker Results and Unadjusted and Adjusted Mixed-Effect Logistic Regression Model Results Affecting the Likelihood of FCR ≥ 0.2 mg/L in Trucked Watera
Water collected from Private well Public well serves both piped water and truck Public well serves only truck Filling pipe was cleaned before use Received chlorine from promoter Received water disinfection training The filling pipe had leakage Capacity of the truck (cubic meter) Communities served in past three days Number of client served in past three days Thinks chlorination is better in delivered water Reported chlorination Observed chlorination Knowledge score FCR ≥ 0.2 mg/L
number (%)
unadjusted OR (95% CI)b
105 (48%) 68 (31%) 48 (21%) 67 (30%) 61 (28%) 84 (38%) 75 (34%) 5.5 (2.1) 1.1 (0.3) 10 (12.8) 203 (92%) 83 (38%) 94 (43%) 4.8 (1.8) 98 (45%)
ref
ref
30.4 (11.0−84.2)c 4.0 (1.6− 10.2)c 5.2 (2.4−11.2)c 1.9 (0.9− 4.1)
32.6 (7.5−141.0)c 2.9 (0.8−10.2)
3.0 (1.3−6.7)c
adjusted OR (95% CI)b
7.8 (2.1−29.3)c 5.3 (1.2−24.4)c 3.1 (0.9−11.2)
1.1 (0.6−2.1) 1.0 (0.9−1.2) 0.9 (0.3−2.4)
0.4 (0.1−2.3)
1.0 (
