Assessing The Impact of Social and Behaviour Change Communication and Social Accountability on Nutrition-Related Outcomes in Jamalpur and Sherpur Districts, Bangladesh

The BIeNGS project aims to improve maternal and child nutrition in two districts of Bangladesh (Jamalpur and Sherpur) by promoting multi-sector pro-poor governance models and nutrition interventions. The BIeNGS model is based on four interrelated components comprising a number of different nutrition-specific and nutrition-sensitive, pro-poor governance interventions: (1) social and behaviour change communication, (2) health system strengthening through social accountability, governance improvements and, capacity development, (3) productive and economic empowerment, and (4) multisector coordination.

The evaluation focuses on two key activities: (i) Social and Behaviour Communication Change and (ii) Social Accountability. The SBCC component is nested within a wider component which aims to improve nutrition and hygiene practices among caregivers of children under five (U5), pregnant and lactating women (PLW) and adolescent girls through promotion within various community delivery platforms across various nutrition specific and sensitive sectors. These interventions will be delivered by local healthcare providers (particularly those charged with delivering the NNS), agriculture, livestock and fisheries extension workers and teachers and will focus on promoting effective behaviour change, based on a range of community and household mobilization and training methods within the SBCC model. This component aims to reach 100% of eligible households with children U5, PLWs and adolescent girls. WVB’s SBCC strategy was developed based on a socio-ecological model of SBCC across multiple influences on health behaviours at individual, family network, community and society levels. It uses multiple channels to disseminate information including face to face household visits and courtyard meetings, community outdoor media, community outreach activities, and print and electronic media. Frontline workers (healthcare providers, extension workers and teachers) will receive training based on a module to be developed with government representatives and rolled out via a training of trainer (ToT) model, reaching 1426 government workers. They will be supported by a coordinated network of community volunteers, promoters and facilitators who reach households and caregivers directly. Government health workers and Community Nutrition Promoters (CNPs) will conduct Timed and Targeted Counselling (TTC) for PLW at HH level. Allpregnant women will be registered and followed up by a CNP until the child is two through 11 scheduled visits with appropriate nutrition messages

The SBCC component is nested within a wider component which aims to improve nutrition and hygiene practices among caregivers of children under five (U5), pregnant and lactating women (PLW) and adolescent girls through promotion within various community delivery platforms across various nutrition specific and sensitive sectors. These interventions will be delivered by local healthcare providers (particularly those charged with delivering the NNS), agriculture, livestock and fisheries extension workers and teachers and will focus on promoting effective behavior change, based on a range of community and household mobilization and training methods within the SBCC model. This component aims to reach 100% of eligible households with children U5, PLWs and adolescent girls.

WVB’s SBCC strategy was developed based on a socio-ecological model of SBCC across multiple influences on health behaviours at individual, family network, community and society levels. It uses multiple channels to disseminate information including face to face household visits and courtyard meetings, community outdoor media, community outreach activities, and print and electronic media. Frontline workers (healthcare providers, extension workers and teachers) will receive training based on a module to be developed with government representatives and rolled out via a training of trainer (ToT) model, reaching 1426 government workers. They will be supported by a coordinated network of community volunteers, promoters and facilitators who reach households and caregivers directly. Government health workers and Community Nutrition Promoters (CNPs) will conduct Timed and Targeted Counselling (TTC) for PLW at HH level. Allpregnant women will be registered and followed up by a CNP until the child is two through 11 scheduled visits with appropriate nutrition messages delivered each time. Husbands and mothers-in-law are also targeted in home visits as CNPs will be tasked withfacilitating families to identify barriers to best nutrition practices and agree actions to address factors that hinder maternal and child health. CNPs will also work with health workers to implement a ‘Positive Deviance Hearth’ model, which will identifypositive deviant households (households with well-nourished children) and their practices, to serve as a model for households with children who are underweight). Monthly Growth Monitoring and Promotion (GMP) sessions will also be conducted at Extended Programme on Immunization (EPI) centres, supported by Government health and family planning staff. These direct contacts will be supported by a broader media campaign via print and cable TV media (continuous broadcast on local cable tv channels plus 20 nutrition messages annually in local print media) and taskforces taking place in 36 schools across the districts (including students and teachers promoting nutrition and WASH messaging).

Social Accountability (SA) arm
The SA component aims to improve local level social accountability by facilitating social monitoring of nutrition services and local level advocacy by educating citizens and particularly the most vulnerable groups about their rights, while equipping them with tools designed to empower them to protect and enforce those rights. World Vision’s Citizen Voice and Action(CVA) social accountability model will be used to engage citizens to increase awareness of the services they are entitled to and advocate for improved delivery of nutrition related services, including quality rollout of nutrition protocols, policy implementation gaps, budget allocations, coordination and referral mechanisms, and facility monitoring by health authorities.

The CVA process is summarized as follows: civic education is provided about rights to services under local law. Communities learn what their national and/or local governments set as standards. These standards are then compared to the reality that exists in individual delivery platforms, including in this case community clinics. Communities are then introduced to a scorecard system that enables them to rate the services provided by the clinic or other delivery platform and provide their own qualitative performance measures, choosing relevant indicators themselves.

This component will also provide capacity development to local health service providers and authorities to collect and utilize nutrition data to be used for upward and downward service decision making and link it to the Government’s Nutrition Information System and Health Management Information System (HMIS). Local civil society organizations (CSOs) and community members will facilitate the approach, targeting local policy-makers and service providers, encouraging participation of the most vulnerable and marginalized, including ethnic minorities, women, and people with disabilities.

This approach is expected to improve community participation in local governance processes and in other programme aims, including strengthening National Nutrition Service delivery, andnutrition-sensitive agriculture, including biofortification, local value chain development, and effective targeting and delivery of social safety net programs. The intended added value of CVA will be giving voice to community groups that have traditionally not been able to participate in nutrition governance, creating local level platforms of engagement between communities and service providers and the creation of CVA groups for lower level accountability that can be sustained beyond the project’s life.

2019-04-30

2021-04-30

CHILDREN’S ANTHROPOMETRY

Height-for-age (HAZ)

Weight for height (WZH)

Weight for age (WAZ)

MUAC

FOOD SECURITY AND DIETARY DIVERSITY

Food Insecurity Experience Scale (FIES)

Food Consumption Score (FCS)

IYCF knowledge Index

IYCF PRACTICE

Proportion of children under 6 months who are exclusively breastfed

Proportion of children between 6 and 24 months who are given complementary food

Proportion of children under 24 months who are still breastfed.

Dietary diversity of children 6-24 months.

ACCESS TO HEALTH SERVICES

Proportion of women who attended at least four antenatal care sessions

Proportion of women provided with iron tablets

Number of Tetanus Toxioids (TT) vaccinations received during pregnancy.

Immunisation status of children

Proportion of mothers provided with vitamin A after delivery

Proportion of mothers provided with iron supplements during the 6 months after delivery.

SANITATION AND HYGIENE PRACTICES

Sanitation and hygiene practices will be gauged using the standard UNICEF monitoring tool.

PERCEPTIONS OF HEALTH AND NUTRITION SERVICES INDEX

COMMUNITY ENGAGEMENT AND CIVIC LIFE ATTITUDES INDEX

The intervention will be evaluated using a cluster randomized trial (CRT) design. The unit of randomization will be community clinics (CC). CCs cover about 6,000 people on average and they correspond to the level of implementation of the SBCC and SA interventions. The study will include 156 community clinics. The evaluation design features two treatment groups and one control group. The first treatment group will receive the SBCC intervention and the second treatment group will receive both the SBCC and SA interventions. The control group will receive neither SBCC nor SA. All 156 clinics (including the control clinics) will receive other interventions in the BIeNGS package on health system strengthening, productive and economic empowerment and multisector coordination.

In a first stage, we will randomly exclude 68 of the 224 implementation clinics from the scope of the evaluation. This randomization process will be stratified by upazila so that the proportion of excluded (or non-evaluation) clinics is the same across upazilas.

In a second stage, we will randomly assign each evaluation clinic into one of three groups: SBCC clinic, SBCC+SA clinic and control clinic. The randomization process will be stratified by upazila so that the proportion of each type of clinic is the same across subdistricts. We will not account for the treatment status of neighboring clinics during the randomization plan. The risk of contamination across CCs is real, but the BIeNGS project will be rolled out in all community clinics of the selected upazilas. To create buffer zones around the evaluation clinics, the study would require a high number of community clinics to be denied the BIeNGS intervention. This would be unacceptable from an ethical point of view, and it would compromise the statistical power of the study as the total number of clinics that can be part of the evaluation is fixed. Instead, potential contamination bias will be addressed by (i) including a detailed health-seeking module to monitor people’ choice of community clinics, and by (ii) accounting for the status of neighboring clinics in the estimations of the treatment effect.

In a third stage, one village will be randomly selected with probability proportional to size among villages covered by each community clinic, based on the latest population figures at village-level. When villages are very large, they will be split in a number of equal sized segments and one segment will be selected at random for the survey.

In a fourth stage, in each village (or segments), we will list every households and collect basic demographic characteristics of the household. Households with a pregnant or lactating woman or with a child under the age of 5 will be eligible for inclusion in the survey. In the unlikely case the quota of eligible households is inferior to the sample size target, we will include the closest neighboring village and fill up the quota there.

In a fifth stage, 15 eligible households will be randomly selected in each village (or segments) for inclusion in the survey.
In each household, the pregnant or lactating woman or the mother of the child below 5 will be the primary respondents. To minimize the burden on these respondents – and wherever it is justified to obtain accurate information – some modules will be answered by male household members instead. These include modules on assets and agriculture/fishery.

In each selected household with at least one child below the age of 5, the latest born child will be the index child and some modules will ask questions about this child only. Anthropometric measurements, however, will be conducted on all children below 5 in the household.

The unit of randomization is the Community Clinic. Within each community clinic catchment areas, we randomly select one village with PPS. Households are then randomly selected from each sampled village.

Yes

156 community clinics

2340 households

The study will include 46 clusters in each of the two treatment arms plus 64 control clusters.

The evaluation will aim to detect with a statistical power of 80% a change of 0.2 standard deviation (SD) in the height-for-age z-score (HAZ) of children under 5.