STUDY PURPOSE: Little information exists on the cause of injury for patients who are treated and discharged from emergency departments; these patients comprise approximately 90% of all injured patients requiring medical care. A method is described to assign external cause of injury codes (E codes) prospectively to all injured patients seen in a large-volume ED. METHODS: E code assignment was performed by the ED triage nurses on entrance to the ED. A checklist was used that contained frequently occurring codes as identified in a pilot study. E codes were entered into the patients' records on the hospital mainframe computer by a medical records technician. These were acquired for a nine-month period to determine feasibility, accuracy, and ease of use. Accuracy was verified retrospectively. RESULTS: During the nine-month period, 67,495 patients were treated. Acute injury accounted for 16,186 of the visits, and 2,085 were return visits for a previously treated injury. The majority of patients with any injury type were treated and discharged and would not have been included in traditional data sets of injured patients. Overall admission rate for injury was 13%. Accuracy of prospectively assigned E codes was 98%, and this method of assigning E codes resulted in no additional effort by the triage nurse. CONCLUSION: E coding is a valuable method for injury surveillance, easily performed in EDs with high volume. Its value is essential for injury prevention research on injuries from any cause.
STUDY PURPOSE: Little information exists on the cause of injury for patients who are treated and discharged from emergency departments; these patients comprise approximately 90% of all injured patients requiring medical care. A method is described to assign external cause of injury codes (E codes) prospectively to all injured patients seen in a large-volume ED. METHODS: E code assignment was performed by the ED triage nurses on entrance to the ED. A checklist was used that contained frequently occurring codes as identified in a pilot study. E codes were entered into the patients' records on the hospital mainframe computer by a medical records technician. These were acquired for a nine-month period to determine feasibility, accuracy, and ease of use. Accuracy was verified retrospectively. RESULTS: During the nine-month period, 67,495 patients were treated. Acute injury accounted for 16,186 of the visits, and 2,085 were return visits for a previously treated injury. The majority of patients with any injury type were treated and discharged and would not have been included in traditional data sets of injured patients. Overall admission rate for injury was 13%. Accuracy of prospectively assigned E codes was 98%, and this method of assigning E codes resulted in no additional effort by the triage nurse. CONCLUSION: E coding is a valuable method for injury surveillance, easily performed in EDs with high volume. Its value is essential for injury prevention research on injuries from any cause.