Citation

Bota GW, Therrien SA, Rowe BH. Acad. Emerg. Med. 1997; 4(4): 291-296.

Affiliation

Sudbury General Hospital, Emergency Department, Ontario, Canada.

Copyright

(Copyright © 1997, Society for Academic Emergency Medicine, Publisher John Wiley and Sons)

DOI

unavailable

PMID

9107328

Abstract

OBJECTIVES: ED injury surveillance requires accurate information about mechanism. This study explored the clinometric properties of an E-code system specifically designed to track ED injuries. METHODS: All patients assessed in the ED had cause-of-injury information documented using a truncated E-code system. Patient records were hand-searched to determine coding compliance. A selection of 98 charts (50 injury/48 noninjury) were coded by 7 physicians, 2 nurses, and 2 nosologists. Agreements (interrater and intrarater) on the diagnosis of trauma and exact E-codes were determined (using kappa; kappa). RESULTS: E-coding compliance was high (overall 90%: 95% CI: 85-93%), and accuracy of injury classification was 99%. Compared with an expert's coding, agreement on injury classification was excellent for physicians (kappa = 0.91; 95% CI: 0.80-1.0), nurses (kappa = 0.88; 95% CI: 0.75-1.0), and nosologists (kappa = 0.92; 95% CI: 0.81-1.0). Agreement was substantial for the exact E-codes between physicians (kappa = 0.77; 95% CI: 0.60-0.94) and nurses (kappa = 0.72; 95% CI: 0.54-0.90). Recode reliability was also excellent for physicians (kappa = 0.88; 95% CI: 0.75-1.0) and nurses (kappa = 0.96; 95% CI: 0.88-1.0). CONCLUSIONS: Injury coding using a truncated E-code system can provide valid and reliable data from the ED. Differences between nurses, physicians, and nosologists in the ability to accurately code using this system were minimal, thus eliminating the need for additional staff and resources.

Language: en