Limitations of secondary data sets for road traffic injury epidemiology: A study from Karachi, Pakistan

Objectives. To evaluate the completeness of descriptive information and accuracy of ambulance data when compared with two other secondary sources of data, namely, medical examiners' reports and medical records, in the city of Karachi, Pakistan. Methods. A retrospective chart review was carried 7ut of all road traffic injury (RTI) victims transported by an ambulance service in Karachi from January 1 to December 31, 2003. Information on the name, age, and gender of the victim, date, time, and place of injury, and type of injury and its outcome was obtained, and then compared with medical examiners' data and inpatient medical records. The variables missing in the ambulance service database were identified. Results. The ambulance service transported 1,245 RTI patients during 2003, 81% of whom were taken to one of three trauma centers. Twenty percent died before reaching hospital. There were missing variables in 340 (27%) cases, the majority of which were the unique identifiers such as name and age (67%). Data on other variables were reported in 95% or more cases. None of the medical examiners' reports or inpatient medical records contained a description of event of injury. The agreement rate for the variables available in the three data sets ranged from 61% to 100%. Conclusion. Secondary data on injuries, such as ambulance reports, medical examiners' reports, and medical records, have significant limitations. In Karachi, except with regard to unique identifiers, ambulance data seem to provide information just as accurate as that in medical examiners' and inpatient medical records.