Evaluation of the Dubai Cares WASH in School Initiative

Overview

This study will use longitudinal data collection to quantify the impact of a school-based water, sanitation and hygiene (WASH) program in Mali, West Africa, on pupil absenteeism, diarrheal illness, and respiratory illness. Data will be collected from 100 intervention and 100 control schools across 4 regions in Mali. At each school research staff will randomly select 60 pupils in grades 3-6 (typically ages 7-14). Research staff will ask school directors to provide written consent for pupil participation in place of parents, a procedure that has been approved by the Malian Ministry of Education. Research staff will approach selected pupils at school to request informed oral assent for participation. The informed assent process and data collection will be conducted by local trained enumerators in the child's preferred language of French or Bambara. Oral rather than written assent was chosen due to varied literacy capabilities and to limit paperwork that identifies participants. Data collection will take place at each school once every six to eight weeks during the duration of two school years. Research staff will conduct observations of school facilities, observations of handwashing behavior of children upon leaving latrines, and an interview with the school director about WASH conditions and school enrollment information. None of these school activities collect personal data or identifiers. Data collection with pupils will be done with the same pupils every 6-8 weeks and entails a 5-minute interview asking about recent absence and symptoms of diarrhea and respiratory illness. All data collection will occur at the school, will be conducted by trained local enumerators, and will be stored on password-protected mobile data collection devices. Data will be uploaded to a password-protected server and will not contain personal identifying information. Enumerators will maintain a separate form that links pupil IDs with names, which will be kept confidential. There are no direct benefits other than contribution to general knowledge that will inform future school WASH projects. There are no risks to participation other than a small amount of class time missed by pupils, and great efforts will be made to minimize time outside of class.

BACKGROUND AND SPECIFIC AIMS Household-based water, sanitation and hygiene (WASH) improvements have demonstrated reductions in diarrheal. However, there is little academically sound research on the impact of school-based programs. The small body of current evidence is promising; school-based WASH interventions have been associated with reductions in absenteeism, particularly among girls. School WASH programs have also demonstrated decreases in diarrheal disease and helminth reinfection rates among school children. In 2010, a consortium between UNICEF, CARE Mali, Oxfam GB, WaterAid Mali and Save the Children was launched to implement WASH programming in over 700 primary schools across Mali under the Dubai Cares WASH in Schools Initiative (DC Mali Initiative). Core activities of the project include: installing or rehabilitating school water points, latrines and hygiene kits; promoting good WASH practices and behavior change at school and within the community; and establishing management systems to ensure sustained financing, monitoring and maintenance. The specific aim of this study is to quantify the impact of the DC Mali Initiative on pupil learning and health. Using longitudinal data collection, investigators will assess absenteeism, diarrhea, and respiratory infections in 100 intervention schools and compare the results to those observed in 100 control schools. DATA COLLECTION ACTIVITIES Facility Observations will be conducted by trained study enumerators using electronic data collection devices. The observations will be of the school facilities specifically, not of the students. Structured Director Interviews will be administered to the school director. In the case where the school director is not available, an adjunct director with knowledge of school WASH facilities will be interviewed. Responses will be recorded on electronic data collection devices by trained enumerators. The questionnaire will address presence and functionality of school WASH facilities, and should take no more than 20 minutes. The information collected in this interview is a report on the state of facilities at the school and is not sensitive personal information. Structured Pupil Interviews will be administered to a random sample of 40 pupils at each school. Pupils in grades 3-6 (typically ages 7-14) will be eligible for the sample. As this is a longitudinal study, the same pupils will be interviewed during each visit. New pupils will be selected at the start of the second academic year in September 2013 to replace those pupils who graduated or dropped out during the 2012-2013 academic year. Upon initial recruitment and after obtaining informed oral assent, pupil names will be recorded on a master list of participants, which will assign an ID to each participant. This list will be kept confidential by each enumerator. Enumerators will only record the corresponding ID number on data collection forms, and no names will be permanently recorded in the database. Pupil interview responses will be recorded on electronic data devices by trained enumerators. The interview will address recent history of school absenteeism, diarrheal illness, and respiratory illness. Pupils will also provide information about their age, grade, presence of a latrine at home, and academic grade the previous year. The pupil interview will take no more than five minutes per pupil. Interviews will be conducted in either French or the local language, according to the preference of the pupil. A Roll Call will be taken of all pupils in the school on the day of each visit. Upon the first visit, enumerators will record a list of all pupils enrolled in the school. These lists with data will be kept confidential by enumerators and will not be shown to school administrators or others not involved in the research. At each data collection visit, enumerators will call the name of each pupil in the class and mark whether the pupil is absent. This method will be used to triangulate attendance data with school roster records. Aggregate data from all visits will be recorded in the database without any personal identifying information. Handwashing observations will be made of pupils upon exiting latrines. Enumerators will observe the latrine area for 20 minutes and tally whether or not pupils use water and/or soap after leaving the latrine. No names or other personal or identifying information will be collected. An Academic Data Review will be conducted three times throughout the 2012-13 and 2013-14 school years to determine percentage of children obtaining a passing average, percentage of children advancing to the next class, and percentage of drop-outs at each school. Only school-level data, disaggregated by gender and grade, will be collected during this review. No individual or personal data will be collected for this activity. DATA MANAGEMENT AND MONITORING This trial does not entail a medical intervention of any kind; therefore, a typical data safety and monitoring plan is not necessary. Study staff will be instructed to monitor any unforeseen situations that suggest an increase in risk to participants and immediately inform the study manager so that a decision can be made on how to mitigate those risks. RISKS AND BENEFITS OF PARTICIPATION Participation in the study will not subject participants to any excess risk other than a brief five-minute interruption of class time during classroom roll call, and a five-minute break from class for pupil participants to leave class for their interviews. There are no direct benefits to participants other than knowing that information gained from this study will contribute to knowledge in the WASH sector worldwide about the educational and health impacts of improved WASH in schools. Schools participating as controls will be placed in high consideration by DC Mali Initiative partners for future interventions. CONFIDENTIALITY Names of pupil participants will be recorded separately from survey forms and will be destroyed after data collection has been completed. This information will only be retained on a separate ID linking form to ensure the same pupils are contacted in repeat visits to the school. Private health data relating recent history of diarrhea and respiratory illness will be linked to demographic information including age, gender, school, and school grade; however, this information will be insufficient to serve as an identifier. No identifiers will leave Mali INFORMED CONSENT The Malian Ministry of Education has granted a waiver of parental consent for pupil interviews. At each school, school directors will be asked to sign in loco parentis ("in the place of parent") on behalf of the pupil participants. These forms inform the director of the activities and purpose of the research and give permission to the researchers to carry out short interviews at the school with consenting pupils. In addition to the loco parentis form, all pupil subjects selected to respond to the pupil interview will be read a full informed oral assent script the first time the interview is conducted. They will provide oral assent if they choose to participate in the longitudinal study. Continued assent will be obtained from pupils before commencing all subsequent data collection rounds. Oral assent rather than written has been chosen due to varying levels of literacy at the school as well as a desire to minimize paperwork that contains participants' names. Investigators expect to sample 60 pupils per school. Therefore, school directors would sign the in loco parentis form for up to 60 pupils at each school. This form will be signed at the first visit, and will serve as consent for all subsequent visits. A total of 200 in loco parentis forms will be collected from school directors. A maximum of 12,000 oral assents will be recorded from the pupils during the first round of data collection, with an equal number of assents during each of the following data collection rounds.