Injury Severity Score Inflation Resulting From Pan-Computed Tomography in Patients With Blunt Trauma

All articles that demonstrate a mortality benefit from liberal pan-computed tomography (CT) use in patients with blunt trauma have relied on Injury Severity Score (ISS) to control for morbidity. This mortality benefit may be artifact, the result of an increased use of a sensitive diagnostic modality rather than a true benefit. We quantify the magnitude of ISS inflation in patients with blunt trauma who are undergoing routine pan-CT compared with patients who receive more selective scanning.

This study re-analyzes data collected from a previous study of pan-CT use in patients with blunt trauma in which surveyed emergency physicians prospectively indicated which portion of a pan-CT they wished to obtain. The trauma surgeons who jointly managed all patients in this study ultimately decided which CTs to obtain. We recalculated the ISS excluding injuries found on the undesired CT scans that did not lead to a predefined set of critical actions and compared original and recalculated ISS.

There were 701 study subjects who received a total of 2,615 scans. Of these, there were 992 undesired scans. Ninety-nine of the obtained undesired scans, performed in 92 patients, had noncritical abnormalities. The original ISS for these 92 patients was 10 (IQR 5, 18); the recalculated ISS was 5 (interquartile range 1, 10), a 50% decrease.

Although the median ISS for our study was lower than that of previous studies claiming a mortality benefit, ISS inflation appears to be a real phenomenon and may confound studies that use ISS to control for morbidity.