Interobserver Variability in Injury Severity Scoring After Combat Trauma: Different Perspectives, Different Values?

Abstract

Anatomic measures of injury burden provide key information for studies of prehospital and in-hospital trauma care. The military version of the Abbreviated Injury Scale [AIS(M)] is used to score injuries in deployed military hospitals. Estimates of total trauma burden are derived from this. These scores are used for categorization of patients, assessment of care quality, and research studies. Scoring is normally performed retrospectively from chart review. We compared data recorded in the UK Joint Theatre Trauma Registry (JTTR) and scores calculated independently at the time of surgery by the operating surgeons to assess the concordance between surgeons and trauma nurse coordinators in assigning injury severity scores. Trauma casualties treated at a deployed Role 3 hospital were assigned AIS(M) scores by surgeons between 24 September 2012 and 16 October 2012. JTTR records from the same period were retrieved. The AIS(M), Injury Severity Score (ISS), and New Injury Severity Score (NISS) were compared between datasets. Among 32 matched casualties, 214 injuries were recorded in the JTTR, whereas surgeons noted 212. Percentage agreement for number of injuries was 19%. Surgeons scored 75 injuries as "serious" or greater compared with 68 in the JTTR. Percentage agreement for the maximum AIS(M), ISS, and NISS assigned to cases was 66%, 34%, and 28%, respectively, although the distributions of scores were not statistically different (median ISS: surgeons: 20 [interquartile range (IQR), 9-28] versus JTTR: 17.5 [IQR, 9-31.5], p = .7; median NISS: surgeons: 27 [IQR, 12-42] versus JTTR: 25.5 [IQR, 11.5-41], p = .7). There are discrepancies in the recording of AIS(M) between surgeons directly involved in the care of trauma casualties and trauma nurse coordinators working by retrospective chart review. Increased accuracy might be achieved by actively collaborating in this process.