ABSTRACT
Road traffic injuries (RTIs) are increasing in developing countries where accurate routine data are usually not available. Although a capture-recapture technique has increasingly been employed in studies of human populations to provide reliable estimates of the magnitude of problems, it has rarely been used in road traffic injury research. We applied two sample capture-recapture methods using hospital and traffic police records to estimate non-fatal road traffic injuries in Thai Nguyen City during the years 2000-2004. We generated a conservative adjusted estimate of non-fatal RTIs using data from the two sources matched by name, surname, sex of victims and at least one of the other matching variables, of age, address of victim and date of injuries. We then compared the estimated rates with those reported based on police and hospital data. The results show that during years 2000-2004, the police reported 1,373 non-fatal RTIs, while hospital records revealed 6,069 non-fatal RTIs. Most reported victims on both hospital and police reports were males (67.3 % and 74.4%, respectively). More than half the victims on both hospital and police reports were drivers (77.5% and 66.1%, respectively) or pedestrians (10.6% and 7.1%, respectively). Youth and young adults (ages 15-34) constituted the majority of the victims on the hospital and police reports (52.8% and 63.7%, respectively). The capture-recapture analysis estimated that 11,140 (95% CI: 10,626-11,654) subjects were involved in RTIs during the study period. In comparison to the estimated figure, official sources accounted for only 21.9 to 60.1% of total non-fatal RTIs. Estimated rates of non-fatal RTIs were 105.5 injuries/10,000 population per year and 393 injuries/10,000 vehicles. Given the fact that under reporting of RTIs has been a major limitation of routine official data sets in developing countries, we suggested the capture-recapture method be used as a tool to provide affordable and reliable estimates of RTIs in resource-poor countries.