Adverse reactions to vaccines in the tropics [editorial]

This article focuses on the finding that the use of Edmonson Zagreb high-dose measles vaccine at the age of 4-6 months was associated with an excess mortality in girls during their second year of life. Although this phenomenon was also observed in Guinea Bissau Senegal Haiti and Sudan it has not been explained. The Edmonson Zagreb story makes the point that there needs to be a research framework for some adverse reactions to be detected. The rotavirus tale demonstrates that adverse reactions in the US may obstruct the use of highly valuable vaccines in the tropics unless there are mechanisms in place to show the true risk/benefit in the tropics. In view of such both a vaccination database for a defined population and a system of recording contacts with all health services for that population are needed to be developed. Although the establishment of these systems would be very expensive and difficult there are a number of situations that might allow such a system to be grafted on. These systems include: 1) a simple expansion of the range of health outcomes recorded and the establishment of methods of individually linking vaccine records to outcomes; 2) many of these trials are conducted in populations under continuous demographic surveillance where the addition of morbidity data could be undertaken within an established framework; and 3) computerizing existing information and tightening up on the quality of the information.