Abstract
BACKGROUND: Mortality due to measles is often under-reported. Traditional methods of measuring mortality can be time and resource-intensive. We describe the implementation of a community-based method to monitor measles mortality. FINDINGS: Using standardized questionnaires in the midst of a measles outbreak, a community-based network of volunteers recorded a much larger number of deaths (376) than deaths recorded in health centres (27). Deaths were predominantly (93.5%) among children aged less than 5 years; 54.5% of measles deaths reported antecedent measles vaccination. CONCLUSIONS: In this setting, the number of deaths due to measles reported in community-based surveillance was much higher than deaths reported in health centres. Lack of reliable population data and incomplete coverage of the surveillance system make it impossible to calculate overall attack rates and cause-specific mortality rates. Similar systems could be rapidly implemented in other difficult outbreak settings.