Improved chlorination and rapid water quality assessment in response to an outbreak of acute watery diarrhea in Somali Region, Ethiopia

Somali Region of Ethiopia has been affected by drought for several years. Drought conditions have led to food and water scarcity and a humanitarian crisis in the region. In January 2017, an outbreak of acute watery diarrhea (AWD) was declared in the region. AWD prevention and control activities include strengthening water, sanitation, and hygiene (WASH) services. Access to safe drinking water is critical in preventing transmission of AWD and chlorine is an effective chemical to disinfect water supplies. The US Centers for Disease Control and Prevention collaborated with the WASH Cluster and the United Nations Children ’ s Fund, Ethiopia, to provide technical assistance to the Somali Regional Water Bureau to improve chlorination of drinking water supplies and quickly assess water quality improvements in Jijiga town, Fafan Zone. Timely sharing of surveillance and case investigation data allowed for the identi ﬁ cation of gaps within the water supply system in Jijiga and implementation of centralized and decentralized chlorination interventions and monitoring systems. Pilot use of a rapid assessment to determine residual chlorine levels at various points in the city helped improve chlorination intervention impact. This work illustrates that rapid community-level water quality improvements can be implemented and assessed quickly to improve interventions during outbreaks.

cultivation has led to food and water scarcity, population displacement, and a humanitarian crisis in the region. This crisis was further exacerbated by an outbreak of AWD that started in January 2017. From 1 January to 10 July 2017, there were a total of 33,993 cases in the region including over 1,000 cases in Jijiga, the capital of the region.
Jijiga is the largest urban area in Somali Region and a transportation hub, with population movement in and out of the region and country. Thus, improving access to safe water in Jijiga was considered critical to responding to the outbreak in the region. A variety of water sources are used to provide drinking water to the approximate 159,300 residents of Jijiga. Water supplied by the piped system is sourced from deep boreholes and stored in two reservoirs 1,000-3,000 cubic meters in volume before release. The Somali Regional Water Bureau (RWB) estimates that the piped network reaches 80% of residents in Jijiga primarily through communal tapstands. However, piped water is only available for several hours a day resulting in queuing at tapstands. In addition, the piped system supplied by the reservoirs does not reach all areas of Jijiga. The remainder of residents collects water from a variety of sources, including donkey carts, water trucks, and local boreholes, not connected to the piped network. The water sources for the donkey carts and water trucks vary. Some collect directly from the piped network, others from filling stations, and still others fill with surface water from a nearby dam. Initial FRC testing conducted in Jijiga before the intervention, determined that water supplied from the reservoirs was not consistently chlorinated and that water from donkey cart filling stations and truck filling stations was not chlorinated.
We identified two main challenges that compromised the drinking water quality in Jijiga. First, although water supplied by the reservoirs was chlorinated at times, no consistent treatment and monitoring plan were in place.
Second, donkey carts and trucks that filled from filling stations were not chlorinated and those that filled from the piped network did not receive booster chlorination doses, despite longer transport times and infrequent tank flushing.
To address these challenges, the US Centers for Disease Control and Prevention (CDC), in collaboration with the Somali Region WASH Cluster and the United Nations Children's Fund, Ethiopia (UNICEF), provided technical assistance to the Somali RWB to improve chlorination of drinking water supplies in Jijiga and water quality monitoring. Our report details the chlorination interventions implemented and focuses on the use of a rapid assessment exercise to quickly collect data to improve intervention impact.

METHODS
The CDC, WASH Cluster, and UNICEF supported the Somali RWB in several areas. First, partners identified and mapped priority water points in AWD affected areas of Jijiga by using AWD surveillance and case investigation data. Second, appropriate chlorination methods were implemented at water points. Centralized chlorination was used to improve chlorination through the piped network, and trucks and donkey carts were chlorinated at filling stations using decentralized chlorination methods. Third, the CDC and the Somali RWB trained all attendants in charge of releasing water at the city's reservoirs, truck, and donkey cart filling stations on chlorination procedures. Attendants at the reservoirs were trained how to chlorinate daily during reservoir filling using high-test hypochlorite (HTH) with a target dose of approximately 1.0 mg/L. Filling station attendants at truck and donkey cart filling stations were trained to dose trucks and carts at 2.0 mg/L. Fourth, to supplement the chlorination efforts, a water quality monitoring phone application was created for RWB in collaboration with UNICEF, CDC, and IRC to monitor FRC levels at different water points serving the community, including tapstands, truck filling stations, and donkey cart filling stations. The application was designed so that water quality data could be entered over time and tracked by location. Fifth, a Somali RWB staff member was trained to conduct regular monitoring visits to the reservoirs and filling stations to ensure chlorination efforts continued. Lastly, in partnership with the WASH Cluster, a free residual rapid assessment was conducted to quickly assess chlorination interventions. The implementation and findings of the rapid assessment are the focus of this short communication.
On 15-16 June, approximately, 1 month after the initiation of chlorination activities in Jijiga, the WASH Cluster conducted an FRC rapid assessment to evaluate in real time the chlorination status of different water points in the city's water supply system. Representatives from nine WASH Cluster partner organizations assembled into ground storage tanks, and trucks that filled from surface water sources were not targeted by chlorination interventions, due to a lack of feasibility. In addition, data from two kebeles could not be collected due to time restrictions.

DISCUSSION
Approximately 1 month after chlorination activities started, we were able to detect FRC at the majority of water points tested and targeted with centralized or decentralized chlorination interventions. In addition, we also detected FRC in | Spatial distribution of FRC data collected from tapstands, donkey carts, water trucks, and underground storage tanks during the rapid assessment in Jijiga, Somali Region. The figure also includes points not targeted by centralized and decentralized chlorination efforts, including tapstands connected to boreholes and water trucks and underground storage tanks filled with surface water. This spatial information was used to find five additional donkey cart filling stations in the north of Jijiga within the red circle. Decentralized chlorination was implemented at these points.

Other water points not targeted by chlorination efforts
Tapstandborehole 7 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) Trucksurface water 4 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) Underground tank (Birkat)surface water strates that water quality improvements can be implemented and documented quickly during an outbreak. Moreover, the FRC rapid assessment described in this report allowed for the timely collection of water quality data that resulted in targeted action in the north of Jijiga where several water points were not chlorinated. After this assessment, we identified five previously unknown donkey cart filling stations servicing these areas. Staff at these stations were subsequently trained and provided supplies for chlorination. This rapid assessment was successful in documenting improvements in water quality and identifying gaps in treatment. For sustained impact, periodic water quality monitoring will be essential to maintain consistent chlorination levels.
While free chlorine was detected in the majority of truck and donkey cart water tested, only a small proportion of these levels were >1.0 mg/L, the recommended level.  The work in Jijiga is an example of a rapid targeted WASH intervention and assessment to respond to an ongoing outbreak and prevent further cases of AWD. This targeted response and assessment highlights that during an outbreak, community-level chlorination improvements can be rapidly implemented and assessed to ensure treated water is reaching affected communities.

DISCLAIMER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or the United States Agency for International Development.