A comparison of hospital and police road injury data

摘要: In order to gather as much information possible on road crashes and outcomes, routinely collected police reports of traffic accidents hospital discharge files were individually matched or "linked" using a computerised iterative procedure name-identified data from both sources. The two groups linked unlinked records compared. Within the dataset, comparison like variables was made showed good agreement between sources accident type user type. However, police-reported levels injury severity shown be less reliable. addition, proportion inpatient that record found influenced by several factors. overall linkage rate 64% but varied 29% for motorcyclists in single-vehicle 79% motor vehicle drivers. increased with increasing substantially lower casualties certain ethnic groups. It deduced most instances where did not link record, crash had been reported. These findings confirm there considerable underreporting hospitalised extent greater those severely injured.