Proper coding of the Abbreviated Injury Scale: can clinical parameters help as surrogates in estimating blood loss?

Abstract

The Abbreviated Injury Scale (AIS) requires the estimation of the lost blood volume for some severity assignments. This study aimed to develop a rule of thumb for facilitating AIS coding by using objective clinical parameters as surrogate markers of blood loss. Using the example of pelvic ring fractures, a retrospective analysis of TraumaRegister DGU(®) data from 2002 to 2011 was performed. As potential surrogate markers of blood loss, we recorded the hemoglobin (Hb) level, systolic blood pressure (SBP), base excess (BE), Quick's value, units of packed red blood cells (PRBCs) transfused before intensive care unit (ICU) admission, and mortality within 24h. We identified 11,574 patients with pelvic ring fractures (Tile/OTA classification: 39% type A, 40% type B, 21% type C). Type C fractures were 73.1% AISpelvis 4 and 26.9% AISpelvis 5. Type B fractures were 47% AISpelvis 3, 47% AISpelvis 4, and 6% AISpelvis 5. In type C fractures, cut-off values of <7g/dL Hb, <90mmHg SBP, <-9mmol/L BE, <35% Quick's value, >15 units PRBCs, and death within 24h had a positive predictive value of 47% and a sensitivity of 62% for AISpelvis 5. In type B fractures, these cut-off values had poor sensitivity (48%) and positive predictive value (11%) for AISpelvis 5. We failed to develop a rule of thumb for facilitating a proper future AIS coding using the example of pelvic ring fractures. The estimation of blood loss for severity assignment still remains a noteworthy weakness in the AIS coding of traumatic injuries.