Given the wide usage and potential value of the AIS and ISS in rating severity of trauma, it is essential that certain reliability issues concerning application of the indices be resolved. At present, different data sources (e.g., Emergency Department forms, inpatient charts, autopsy reports) are being used to code AIS with little information available about the reliability of each. Furthermore, individuals with varying levels of training and clinical experience are employed as AIS coders;​ until now, no systematic comparisons have been made of the coding capabilities of these different individuals. Finally, the generalizability of the scale to non-vehicular, penetrating trauma has not been established.

This paper describes a study which was designed to address these reliability issues. Each of 15 raters with varying qualifications (i.e. two research assistants, two medical records technicians, four nurses, four paramedics, and three physicians) was presented with the Emergency Department control sheets and inpatient charts (at appropriately spaced intervals) of 175 trauma cases, and asked to identify and AIS code all injuries. Three months after the first coding of the inpatient chart, the raters were asked to review and severity code the same 175 inpatient charts plus an additional 200 charts. The 375 trauma cases were admissions sampled from four hospitals in the Baltimore area and represent the full range of trauma severity and etiology .