• Hideo Tohira, Ian Jacobs, Tetsuya Matsuoka, and Kazuo Ishikawa.
• School of Primary, Aboriginal and Rural Health Care, University of Western Australia, Crawley, Western Australia, Australia. soba@tororo.net
• J Trauma. 2011 Jul 1;71(1):56-62.

BackgroundThe Abbreviated Injury Scale (AIS) was updated in 2008 (AIS 2008). We aimed to investigate the impact of AIS 2008 on the characterization of injury severity and quality assessment of trauma care.MethodsWe identified all blunt trauma patients in the Japan Trauma Data Bank. First, we converted AIS 98 codes to AIS 2008 codes using a mapping table. Next, we compared Injury Severity Scores (ISSs) and New ISSs (NISSs) based on AIS 98 and AIS 2008. We compared the proportion of major trauma (ISS >15) between the two AISs. We derived risk-adjusted models using the two AISs and separately ranked hospitals according to the observed-to-expected death (OE) ratio. We counted the number of performance outliers for the two rankings. We analyzed the association between the percent change in OE ratios and the proportion of NISS outliers (change in NISS of <-12).ResultsThere were 19,899 subjects. The ISSs and NISSs based on AIS 2008 were significantly less than those based on AIS 98. The proportion of major trauma was 46.3% and 38.9% for AIS 98 and AIS 2008, respectively (p < 0.001). The numbers of performance outliers were different between the two rankings. There was a significant positive linear relationship between the percent change in the OE ratio and the proportion of NISS outliers.ConclusionThe use of different AIS versions influenced the selection of major trauma patients and affected the quality assessment of the trauma care. Researchers should be aware of these findings when selecting the version of the AIS.

30 keywords...

hide…