Injury surveillance at emergency departments in an urban area. Methodology and quality control

An injury surveillance system has been developed and tested in a pilot study at two emergency departments in the Stockholm area. The object was to develop an all-age, all-injury registration at all hospital emergency departments and primary care units treating injured patients. The information will be used at the local level for preventive measures, and on a national level for comparison with other areas. All injured patients were given a record-sheet that served both as a registration form and a medical record. The NOMESKO-code was used as a basis for classifying intent, activity, type of location, and injury mechanism. Altogether, 11,327 injured patients were registered. One day each month was randomly selected from the registration for control of registry completeness. The drop-out rate was on average 13%. The reasons were that patients had not been provided with the injury form, or that the copy of the form had not been sent to, or had not reached the Epidemiological Unit where data entry was performed. 6% of the injuries occurred in patients living outside the Stockholm area. The average shortfall-rate in filling in the NOMESKO-code on the registration form was 10%. The rate of registry drop-outs and incomplete forms should decrease when registration has become a routine procedure, provided that the staff can be engaged in the preventive work and the registration procedure can be adjusted to the routines of trauma management in each emergency department.