Background: In areas where vitamin A deficiency (VAD) is prevalent, vitamin A repletion reduces child mortality by 23% on average. Objectives: To estimate the potential child survival benefits of policies and programmes aimed at controlling VAD in Mozambique, and to make policy and programme recommendations. Methods: The potential contribution of VAD to child mortality in Mozambique was estimated by combining the observed VAD prevalence in the under-5s (71.2%), the measured child mortality effects of VAD (risk of death in children with VAD=1.75 times higher than in children without VAD) and the observed under-5 mortality rate in the country (210 per 1000 live births). Results: In Mozambique, an estimated 2.3 million children below the age of 5 years are vitamin-A-deficient. In the absence of appropriate policy and programme action, VAD will be the attributable cause of over 30 000 deaths annually in the under-5s. This represents 34.8% of all-cause mortality in this age group. Discussion: Vitamin A supplementation (VAS) has been adopted as a short- to medium-term strategy to control VAD in children, and is integrated into routine child health services. However, the last VAS coverage survey showed that only 46% of children received a vitamin A supplement in the 6 months preceding the survey. If VAS coverage is to increase significantly in the foreseeable future, four areas appear to be of paramount importance: (1) reduce missed opportunities for VAS such as visits of sick children to child health services and community outreach activities; (2) take advantage of all potential opportunities for accelerating VAS coverage, such as additional vaccination campaigns and emergency response activities; (3) strengthen health workers' training, supervision and monitoring skills; and (4) increase community demand for VAS of children. Biannual VAS, as the primary component of an integrated strategy for VAD control in children, has the promise to be among the most cost-effective/high-impact child survival interventions in Mozambique.
