Causes of injuries resulting in hospitalisation in Australia: assessing coder agreement on external causes.

OBJECTIVE: To assess extent of coder agreement for external causes of injury using ICD-10-AM for injury-related hospitalisations in Australian public hospitals. METHODS: A random sample of 4850 discharges from 2002 to 2004 was obtained from a stratified random sample of 50 hospitals across four states in Australia. On-site medical record reviews were conducted and external cause codes were assigned blinded to the original coded data. Code agreement levels were grouped into the following agreement categories: block level, 3-character level, 4-character level, 5th-character level, and complete code level. RESULTS: At a broad block level, code agreement was found in over 90% of cases for most mechanisms (eg, transport, fall). Percentage disagreement was 26.0% at the 3-character level; agreement for the complete external cause code was 67.6%. For activity codes, the percentage of disagreement at the 3-character level was 7.3% and agreement for the complete activity code was 68.0%. For place of occurrence codes, the percentage of disagreement at the 4-character level was 22.0%; agreement for the complete place code was 75.4%. CONCLUSIONS: With 68% agreement for complete codes and 74% agreement for 3-character codes, as well as variability in agreement levels across different code blocks, place and activity codes, researchers need to be aware of the reliability of their specific data of interest when they wish to undertake trend analyses or case selection for specific causes of interest.

Accuracy of external cause-of-injury coding in hospital records.

OBJECTIVE: To appraise the published evidence regarding the accuracy of external cause-of-injury codes in hospital records. DESIGN: Systematic review. DATA SOURCES: Electronic databases searched included PubMed, PubMed Central, Medline, CINAHL, Academic Search Elite, Proquest Health and Medical Complete, and Google Scholar. Snowballing strategies were used by searching the bibliographies of retrieved references to identify relevant associated articles. SELECTION CRITERIA: Studies were included in the review if they assessed the accuracy of external cause-of-injury coding in hospital records via a recoding methodology. METHODS: The papers identified through the search were independently screened by two authors for inclusion. Because of heterogeneity between studies, meta-analysis was not performed. RESULTS: Very limited research on the accuracy of external cause coding for injury-related hospitalisation using medical record review and recoding methodologies has been conducted, with only five studies matching the selection criteria. The accuracy of external cause coding using ICD-9-CM ranged from approximately 64% when exact code agreement was examined to approximately 85% when agreement for broader groups of codes was examined. CONCLUSIONS: Although broad external cause groupings coded in ICD-9-CM can be used with some confidence, researchers should exercise caution for very specific codes until further research is conducted to validate these data. As all previous studies have been conducted using ICD-9-CM, research is needed to quantify the accuracy of coding using ICD-10-AM, and validate the use of these data for injury surveillance purposes.