Odds ratios for reclined seating positions in real-world crashes

Abstract It is widely believed that with higher levels of vehicle automation and especially with the advent of fully automatic vehicles, the currently typical forward-facing, upright position will give way to a more relaxed and reclined seating posture. Therefore, the current study investigates the influence of a reclined sitting position on crash injury severity by analyzing real-world crash data from the German in-depth accident study (GIDAS). We compared reclined to upright occupants and focused on effect sizes regarding odds ratios at different injury severity levels. We used the abbreviated injury scale (AIS 2015) for injury scaling and the maximum AIS (MAIS) at the levels 2+, 3+, and 4+ to convert injury severity into a dichotomous metric. Two different analyses were conducted, one looking at the occupant MAIS and one focusing on selected body regions. The body regions investigated are head/face/neck (HFN), thorax, abdomen, pelvis/hip/lower extremities (PHL), and upper extremities. We computed odds ratios greater than one indicating a higher odds of injury at a given injury severity level in the reclined group compared to the upright group. The odds ratios for belted, reclined occupants compared to belted, upright sitting occupants are 2.07, 3.09, and 3.66 for the injury severity levels MAIS 2+, MAIS 3+, and MAIS 4+, respectively. When looking at the body regions, the spread of the odds ratios is wider: At the MAIS 2+ level, the odds ratios range between 1.6 and 7.1; at the MAIS 3+ level, the odds ratios span from 1.5 to 8.7, with the latter value representing the PHL region. No odds ratio could be computed for the upper extremity injuries at this level. At the MAIS 4+ injury severity level, only the HFN odds ratio was statistically significant with a value of 5.6. This study is among the first to show an association between body posture and injury severity at MAIS 3+ and MAIS 4+ injury level in real-world crashes for reclined seating postures.