Incidence and seasonal variation of injury in rural Vietnam: A community-based survey

Abstract

Study objective: To describe seasonal effects on injury incidence at the community level and to assess the relative merits of cross-sectional or longitudinal surveillance for injuries in such a setting. Population and methods: This study took place in Bavi district, northern Vietnam, in the context of a longitudinal community surveillance site called FilaBavi. All non-fatal unintentional injuries which occurred in a sample of 24,776 people living in 5801 households were recorded during 2000. Four interview surveys per household were conducted continuously during 2000, each covering a 3-month period of recall. Injury morbidity data were analysed according to gender, age and circumstances of injury. Statistical analyses were based on monthly, quarterly and annual incidence rates with 95% confidence intervals. Significant differences between incidence rates were noted where the 95% confidence interval of the rate ratio excluded unity. Results: There were 1917 persons who experienced a total of 2079 new non-fatal injuries during the period of observation, corresponding to an incidence of 89/1000 person-years. Seasonal variations were found in all types of injury. Overall, the highest incidence rates were observed in July and April, while the lowest monthly rates were found in May and November. Peaks were observed in February and April for traffic injuries, June for work-related injuries, July, August and October for home injuries. A significantly higher incidence rate was found in the third quarter survey (103/1000 person-years, p<0.05). Conclusion: There can be interesting and significant seasonal variations in injury incidence within a community such as seen here in rural Vietnam and these variations have important implications for the design and planning of injury surveillance activities. Seasonal effects may cause invalid assessments of an injury problem if only cross-sectional household surveys are used, demonstrating the need for longitudinal approaches to injury incidence surveillance. © 2003 Elsevier Ltd. All rights reserved.