• Samantha L Schoell, Ashley A Weaver, and Joel D Stitzel.
• Virginia Tech - Wake Forest University.
• Biomed Sci Instrum. 2013 Jan 1; 49: 274-80.

AbstractDetection of occult injuries, which are not easily recognized and are life-threatening, in motor vehicle crashes (MVCs) is crucial in order to reduce fatalities. An Occult Injury Database (OID) was previously developed by the Center for Transportation Injury Research (CenTIR) using the National Automotive Sampling System Crashworthiness Data System (NASS-CDS) 1997-2001 which identified occult and non-occult head, thorax, and abdomen injuries. The objective of the current work was to develop an occult injury model based on underlying injury characteristics to derive an Occult Score for common MVC-induced injuries. A multiple logistic regression model was developed utilizing six injury parameters to generate a probability formula which assigned an Occult Score for each injury. The model was applied to a list of 240 injuries comprising the top 95 percent of injuries occurring in NASS-CDS 2000-2011. The parameters in the model included a continuous Cause MRR/year variable indicating the annual proportion of occupants sustaining a given injury whose cause of death was attributed to that injury. The categorical variables in the model were AIS 2-3 vs. 4-6, laceration, hemorrhage/hematoma, contusion, and intracranial. Results indicated that injuries with a low Cause MRR/year and AIS severity of 4-6 had an increased likelihood of being occult. In addition, the presence of a laceration, hemorrhage/hematoma, contusion, or intracranial injury also increased the likelihood of an injury being occult. The Occult Score ranges from zero to one with a threshold of 0.5 as the discriminator of an occult injury. Of the considered injuries, it was determined that 54% of head, 26% of thorax, and 23% of abdominal injuries were occult injuries. No occult injuries were identified in the face, spine, upper extremity, or lower extremity body regions. The Occult Score generated can be useful in advanced automatic crash notification research and for the detection of serious occult injuries in MVCs requiring prompt treatment at a trauma center.

3 keywords...

hide…