Introduction: Just under four million infants die each year before reaching one month of age; neonatal deaths now account for 38% of the 10.8 million deaths among children younger than 5 years of age. Tackling neonatal mortality is essential if the millennium development goal to reduce by 2015 overall child mortality by two-thirds from its levels in 1990 is to be achieved. Postnatal care for mothers and neonates in developing countries, particularly when deliveries occur at home, is either not available or is of poor quality. Trained community workers are considered by many to be pivotal to newborn care in the community, as they can act as catalysts for community actions and also be providers of care.Reductions in neonatal mortality have been slower in Sub-Saharan Africa than in any other region, and no evaluations of the effectiveness and feasibility of home visits in reducing neonatal mortality have been conducted.

Trial aim: To link with the Ghana Health Service to develop a feasible and sustainable intervention to improve newborn care practices and careseeking during pregnancy and childbirth, and to identify and refer very low birth weight and/or sick babies, through routine home-visits by community health workers (CHWs), and by so doing reduce neonatal mortality.

Study area: The NEWHINTS trial will be conducted in 6 contiguous districts with an area of 12,000km in the central Brong Ahafo Region of rural Ghana. The area is predominantly rural and has a total population of approximately 600,000 persons, with more than 100,000 women of reproductive age. The study area is multi-ethnic and education levels are low. It is served by 4 district hospitals, and a small number of additional government health centres and private facilities. More than 15,000 babies are born within the area each year; the neonatal mortality rate is about 30 per 1000 live births. Over 50% of births occur at home and these home deliveries account for a large proportion of all neonatal deaths. A potential cadre of CHWs, who are currently working in the region and whose capacity could be expanded to include home visits, are Community Based Surveillance Volunteers (CBSVs); CBSVs currently assist the DHMTs with the registration of births, disease detection and community mobilization.

Intervention: The intervention is being developed by a collaborative group from Kintampo Health Research Centre, the District Health Management Teams (DHMTs) of the 6 districts and the London School of Hygiene and Tropical Medicine. It will comprise of:

1. Training CBSVs to identify pregnant women in the community and to conduct 2 home visits during pregnancy and 3 in the first week of life of the infant on days 1, 3 and 7:

   • to promote birth preparedness and prompt careseeking during pregnancy and childbirth
   • to promote essential perinatal and neonatal care practices
   • to identify and refer very low birth weight or sick babies
   • to give advice on special treatment for low birth weight babies
   • to teach neonatal danger signs and encourage prompt careseeking
   • to promote key child survival interventions for infancy (breastfeeding, bednets, immunisations) CBSVs will provide support through dialogue and problem-solving. Other family members key in decision-making concerning childbirth and newborn care will be invited to take part in the discussions.

2. Developing a sustainable supervisory and remuneration structure for the CBSVs;

3. Various intervention support activities inc

   • Sensitisation of health facility staff to intervention messages and approach#
   • Sensitisation of traditional birth attendants (TBAs)
   • Meetings with village leaders to introduce intervention aims and activities
   • Community durbars

Study design: Formative research consisting of in-depth interviews, focus-groups and pilot trials of home visits/improved practices will be conducted to improve the feasibility and optimality of the planned intervention and to design training and intervention support materials. A cluster randomised controlled trial design with 98 health zones as units of randomisation will be used to evaluate the impact on neonatal mortality, careseeking during pregnancy and childbirth and newborn care practices. The intervention will be implemented in half of the zones, chosen at random, and evaluated over an 18 month period. Impact data will come from ongoing 4-weekly surveillance of all women of child-bearing age and their infants including verbal post mortems, conducted as part of the ongoing "ObaapaVitA" Vitamin A and maternal mortality trial. A process evaluation will be conducted over 3-monthly intervals to assess the coverage and quality of the services provided and the response to the services.

Inclusion Criteria:

• All live births in trial area

Exclusion Criteria:

• None