School Attendance, School Dropouts, School Enrollment, School Feeding Programs, Child Nutrition
Conditions
Brief summary
The goal of this clinical trial is to test daily provision of peanut paste-based milk-containing ready-to-use school food (PM-RUF) in children 5-17 years of age in Ghana . The main question it aims to answer is: \- Will provision of PM-RUF as a daily school meal improve attendance, improve matriculation, and/or reduce dropouts among Ghanaian schoolchildren 5-17 years of age in Mion District as compared with provision of a common local flour made of rice/millet?
Detailed description
School feeding programs offer an opportunity to advance individual and community health and well-being, especially in contexts of poverty and limited diet quality. When compared with no school feeding, school feeding has been shown to improve attendance, reduce dropouts, and strengthen household food security. In many low- and middle-income countries (LMICs), a model called Home-Grown School Feeding (HGSF) has been employed in an effort to increase the coverage of school feeding programs. This model involves local growing, procurement, and cooking of school meals, thereby aiming to boost local economies and improve sustainability. A potential drawback of this approach is the nutritive quality of the school meals, which will depend on what is typically grown and procured in the program. In areas of higher food insecurity, such a program may largely provide the foods to which children already have access, rather than nutrients their diets may be missing. Additional concerns include costs related to decentralized procurement and cooking, as well as food safety, which is more challenging to monitor in such a program. Peanut paste-based school meals might offer benefits as an alternative. Local production is possible, as is done for ready-to-use supplementary and therapeutic foods. Local ingredient sourcing could offer similar economic and sustainability advantages. Peanut pastes are food safe with long storage capabilities. They also provide a matrix into which varied ingredients can be added while maintaining organoleptic acceptability to children. Finally, in regions where current government-run HGSF programs suffer from inadequate funds and instability, local production and distribution of RUFs (ready to use foods) might improve reliability and impact school attendance, dropouts, and matriculation. This study is a cluster-randomized, controlled, investigator-blinded superiority trial. Schools will be randomized to receive PM-RUF or local rice/millet flour for porridge. PM-RUF will contain peanut, palm oil, sugar, fat-free milk powder, and 0.5-1 RDA (recommended dietary allowances) of 14 micronutrients. Attendance will be tracked with the use of fingerprint biometric scanning each day.
Interventions
DIETARY_SUPPLEMENTReady-to-use school food
80 grams of peanut paste-based food containing skim milk powder, cowpea, peanuts, vegetable oil, and multiple micronutrients
DIETARY_SUPPLEMENTStandard meal
300ml of porridge made with rice/millet flour
Sponsors
Washington University School of Medicine
University of Ghana
United States Agency for International Development (USAID)
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
SINGLE (Investigator)
Masking description
Neither participants nor outcomes assessors (who perform biometric attendance tracking) will be blinded to allocation to PM-RUF vs. rice/millet flour. The trial statistician responsible for final analyses will remain blinded by use of letter assignment to each school. The code linking letter/school to intervention vs. control will remain locked and inaccessible to the trial statistician.
Intervention model description
Cluster-randomized, controlled, investigator-blinded superiority trial. Schools will be randomized to receive PM-RUF or local flour for porridge. Randomization will be stratified / covariate-constrained by primary vs. junior secondary school, school size, and distance from a main road.
Eligibility
Sex/Gender
ALL
Age
5 Years to 17 Years
Healthy volunteers
Yes
Inclusion criteria
* Provision of signed (or thumb-printed) and dated informed consent form by parent/guardian * Enrolled at level Basic 1 or higher in a participating school * Parent/guardian stated willingness to comply with all study procedures and availability for the duration of the study, including no plan to move from the catchment area of a participating school * 5 - 17 years of age
Exclusion criteria
* Known allergy to components of intervention or control study food or medications * Condition requiring immediate hospitalization
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Attendance percentage | 11-30 months from enrollment | Percent attendance will be compared between PM-RUF and rice/millet flour groups using ordinal logistic regression with school as a random effect to account for clustered randomization. Odds ratios with 95% confidence interval (CI) will be reported as well as model-derived median of differences with 95% CI. Higher numbers are better. Maximum is 100%. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Dropout from school | 11-30 months | Dropout is a binary outcome, defined as no attendance for 3 consecutive months, and the time from enrollment to last day of school attended will define time-to-dropout. Dropouts will be analyzed using time-to-event analysis with Cox proportional hazards regression. A random effect for school will be included to account for clustered randomization. The reported effect measure will be a hazard ratio with 95% CI. Censoring will occur at graduation from the school and moving away from school's catchment area. |
| Matriculation | 11-30 months | Matriculation is a binary outcome defined by attendance at a subsequent grade after completing a school year. Logistic regression with a random effect for school will be used to analyze matriculation to subsequent school grade. This will be repeated for each school year included in the study. The reported effect measure will be an OR with 95% confidence interval. |
| Afternoon attendance percentage | 11-30 months | Defined as attendance registered following lunch break. Percent afternoon attendance will be compared between PM-RUF and rice/millet flour groups using ordinal logistic regression with school as a random effect to account for clustered randomization. Odds ratios with 95% CI will be reported as well as model-derived medians of difference with 95% CI. Higher numbers are better. Maximum is 100%. |
| New Attendees | 11-30 months | New attendees are defined as new enrollments in school after initiation of school feeding. The number of new attendees as a percent of the school population they join will be analyzed using ordinal logistic regression with school as random effect. The outcome will be reported as an OR with 95% CIs. Higher numbers are better. |
Countries
Ghana
Contacts
CONTACTMark Manary, MD
CONTACTPeter Kidd, BS
PRINCIPAL_INVESTIGATORMark Manary, MD
Washington University School of Medicine
Outcome results
None listed