The specific objectives of the baseline survey are to:

• Identify coverage and timing estimates of pregnancy registration, clinic-and home-based antenatal care (ANC) and Post-natal care (PNC) visit
• Identify estimates of ante-natal care and newborn care knowledge, attitude and practices among mothers and health workers
• Identify estimates of bulging fontenelle at birth and 
• Identify knowledge, skill and behavior levels of health workers elated to pregnancy surveillance/registration, ante-natal services and advice and post-natal services and advice.
• Explore possible confounders, like, socio-economic characteristics of households, contact with CHWs before and after delivery, perception of vitamin A, presence of CHW in the community, distance to the nearest clinic, coverage of essential health services (such as EPI, Tetanus Toxoid), births attended by a skilled health worker, ratio of clinic to population, ratio of health workers to population, and other variables.

Sample design and Sample Size

The Baseline Survey is conducted in the NBVAS 6 upazilas, namely,Sadar and Swarupkathiupazilas in Pirojpur district, Modhupur and Shakhipur upazilas in Tangail district, and Jaldhaka and Dimlaupazilas in Nilphamari district. The survey is done interviewing a sample of 5,400 mothers of children aged 0-6, selected from 180 clusters (with30 clusters included from each of the 6 upazilas) drawing 30 mothers from a cluster. A cluster is village/mauza in rural areas and a mohollah in urban areas. The sample is selected in two stages.

The number of recently delivered women are sampled that is required to discern whether the mother-family delivery model has at least 10% higher coverage compared with CHW model. Sample size estimates require that we consider a plausible intervention effect (i.e. the difference between the two delivery models with respect to the proportion of newborns being given 50,000 IU vitamin A within two days of birth), acceptable probabilities for Type (α=0.05) and Type II (β=0.20 errors) and a design effect of 2.0.   It is expected that the coverage of newborn Vitamin A supplementation at endline survey is approximately 50 percent in the mother-family member dosing model areas (assuming that all the 50 percent of the women receiving antenatal care during pregnancy receive and dose their newborns with Vitamin A supplements). To detect a 10 % change, a sample size of 816 respondents per delivery model (CHW or mother-family dosing model) and program setting would be required. Allowing for a non-response rate of 10 %, a sample size of 900 per model and per program setting would be required. This baseline survey is interviewing 5400 respondents in the 6 upazilas covered by the program.