Serious-to-fatal injury to second-row occupants in rear impacts using 1994–2020 field data

Abstract Objective Serious-to-fatal injury was analyzed for second-row children aged 0–14 years and adults aged 15 and older in rear impacts by body region, restraint use, and injury mechanism using field data collected by NHTSA. Method 1994–2015 NASS-CDS and 2017–2020 CISS data were used to investigate the rate for Maximum Abbreviated Injury Scale (MAIS) 3 + F injury in rear crashes involving 1994+ model year vehicles. All second-row occupants were included, irrespective of restraint use and ejection status. The data were analyzed by group: children (0–14 years old) and adults (15+ years old). All available electronic files for seriously injured second-row occupants in the rear impacts were reviewed for mechanism of injury. Results The rate of serious injury (MAIS 3 + F) for second-row occupants was 0.93% ± 0.36% in rear crashes; it was 0.76% ± 0.39% for children and 1.22% ± 0.40% for adults. There were 2.8 AIS 3+ injuries per seriously to fatally injured occupant on average. Most serious injuries occurred to the head in children and to the head and chest in adults. Restraint use was only 31.3% for all seriously injured second-row occupants in the rear impacts. It was 45.1% for children and 17.8% for adults. The overall rate of serious injury in rear impacts was 10.0 times higher when unrestrained than restrained overall; it was 5.6 times higher for children and 20.2 times higher in adults. The case review indicated that many young children were improperly restrained or placed in the incorrect child seat. More than 17% of second-row adults were ejected; all were unrestrained. The primary mechanism for child injury was related to intrusion (86.0%). About 14% was not related to intrusion; 12.3% involved the front seat rotating rearward into the child. The primary mechanisms for adult injury differed from those for children; 68.0% was related to intrusion, 21.6% was not related to intrusion, and 10.4% involved ground impact with ejection. Of the non-intrusion-related cases, 19.1% involved acceleration forces injuring the adult and 2.5% involved the front seat rotating rearward. Conclusions The primary mechanism for serious injury to second-row occupants in rear crashes was intrusion either by direct force, compression into front components, or acceleration into forward components. The front seat moving rearward was an infrequent cause for injury.


Introduction
NHTSA and automakers have improved rear impact protection (Cammisa and Strassburger 2016;Viano and Parenteau 2016). Despite those improvements, the interest in secondrow occupant protection in rear crashes has increased, as evidenced by Senate letters of inquiry (Markey and Blumenthal 2016) and petitions for rulemaking (Whitman and Sicher 2015;Ditlow 2016) to strengthen front seats, and this interest is primarily related to second-row children in rear impacts. Viano and Parenteau (2009) analyzed 1991-2006 NASS-CDS rear impact data involving 1991þ model year vehicles. Only cases where there was a second-row occupant seated behind an occupied front seat were chosen. The risk of serious injury (Maximum Abbreviated Injury Scale [MAIS] 3 þ F) was assessed by crash severity and front seat performance for second-row outboard occupants. The overall risk was 1.0% for second-row occupants. The risk was 1.6% in crashes at 25-40 km/h and 7.7% in crashes at 40-72 km/h, although the sample size was sometimes limited. The authors compared the data to first-row occupants with and without seat deformation and reported that the relative risk for serious injury to the second-row occupant was 2.3 overall and 1.3 with seat deformation. Front seat deformation occurred due to occupant loading or intrusion in severe rear impacts.
Crashes involving the front seat rotating rearward into a second-row child causing serious injury are rare in NASS-CDS. Viano and Parenteau (2008a) found that there was only one unweighted NASS-CDS case (5.3%, 1 of 19) where the front seatback rotated rearward, contacting a second-row child. The case involved an AIS 3 facial injury to the child occupant. Based on field accidents, this kinematic is not the prevalent means of injury to children in the second row in rear impacts. Parenteau and Viano (2020) updated the Viano and Parenteau (2008a) study with more recent NASS-CDS data but limited the analysis to 1994þ model vehicles. Seven additional unweighted cases were found. More than half of the electronic cases in the 1997-2015 NASS-CDS involved significant intrusion (30þ cm) directly injuring the child, compressing them or pushing them forward into the back of the front seat or component.  analyzed 1994-2015 NASS-CDS and reported that rear crashes accounted for 8.7% of exposed and 5.4% of seriously to fatally injured rear seat occupants. Parenteau, Viano, and Campbell (2022) analyzed 1989-2015NASS-CDS and 2017-2019 data. They reported that second-row occupants were 15.5 years old, weighed 45.4 kg, and were 137.7 cm tall on average. The objective of this study was to update the Parenteau and Viano (2020) study with CISS data and include all second-row occupants, irrespective of age, restraint, and ejection status. The data were analyzed for children (0-14 years old) and adults (15þ years old). Injury rate and frequency by body regions were determined. All available cases were reviewed to determine injury mechanisms.

Methods
Field accident data from 1994 to 2020 on second-row occupants with serious (MAIS 3 þ F) injury in rear impacts were analyzed. The data were subdivided by children aged 0-14 years old and adults aged 15 or older. Appendix A (see online supplement) provides additional information on the methodology.
1994-2015 NASS-CDS: The CDS is a stratified sample representing vehicle crashes selected for in-depth investigation. Vehicle researchers use the data to assess crash performance and overall safety of vehicles.
2017-2020 CISS: CISS replaced NASS-CDS in 2016 and became publicly available in 2017. Similar to CDS, CISS has a stratified multiple-stage sample design. Eligible crashes must involve at least one towed passenger vehicle and must be selected for in-depth investigation.
Vehicles included towed passenger cars, SUVs, minivans, and pickup trucks with 1994þ model year. Rear crashes were defined using the principal impact location and excluded rollovers. Non-rear crashes included other planar crashes and rollovers. Second-row occupants were identified using seating location. Injury severity of the occupant was assessed using the MAIS and fatality. Serious (AIS 3þ) injuries were classified as serious-to-maximum (AIS 3-6) injuries. Different versions of AIS coding were used over the many years of field data.
Electronical cases were reviewed for most serious injury and mechanism involving MAIS 3 þ F to second-row occupants in rear crashes. All calculations were based on the weighted sample. Cases with an inflation ratio equal to 0 or with a negative inflation ratio were excluded from the analysis.

Results
The analysis included weighted estimates of 8,938,864 second-row occupants in all crashes and 721,394 in rear crashes (Appendix A, Table A1, see online supplement). Children accounted for 59.3% of second-row occupants in all crashes and 60.2% in rear crashes. Figure 1 shows the rate of serious injury (MAIS 3 þ F) for second-row occupants for all, rear, and non-rear crashes by age groups. The overall rate was 0.93% ± 0.36% (95% confidence interval [CI] 0.20-1.65) for rear crashes and 1.81% ± 0.20% (95% CI 1.40-2.21) in all crashes. The corresponding rates were 0.76% ± 0.39% (95% CI 0.00-1.55) and 0.98% ± 0.18% (95% CI 0.61-1.35) for children (0-14 years old) and 1.22% ± 0.40% (95% CI 0.42-2.01) and 3.29% ± 0.48% (95% CI 2.32-4.26) for adults (15þ years old). The rate of serious injury (MAIS 3 þ F) was higher in non-rear than in rear crashes for adults (15þ years old) seated in the second row (P ¼ .0057). Table 1 shows the distribution of serious (AIS 3þ) injury to second-row occupants in rear crashes by age group and body region. Table A2 (see online supplement) provides additional information. Serious head injury accounted for 49.5% of all occupant injuries, 73.0% of child injuries, and 26.5% of adult injuries. The corresponding frequency of serious thoracic injury was 21.8%, 10.5%, and 33.0%. These results suggest different loading patterns between children and adults. Seriously (MAIS 3 þ F) injured second-row occupants had 2.8 serious (AIS 3þ) injuries on average; the Grey cell: Unwgt sample < 10. corresponding ratio of serious injuries and seriously injured occupants was 3.2 in children and 2.4 in adults. Tables A3 and A4 (see online supplement) respectively show the number and frequency of contact sources for AIS 3þ head injuries and AIS 3þ chest injuries to second-row occupants involved in rear impacts. The child restraint system was the most common source of serious (AIS 3-6) head injury in rear impacts overall; however, most of the head injuries were to young children. About 74% of serious thoracic injuries occurred from contact with the seatback. Table 2 summarizes second-row occupants involved in rear impacts by restraint use. The data are tabulated by injury severity and age group. There was a weighted estimate of 579,955 second-row occupants with known injury severity (MAIS 0-6 þ F); 60.6% were children and 38.4% were adults. There were 2,661 (31 unweighted) children and 2,716 (43 unweighted) adults seriously injured in the second row in rear impacts. Restraint use was 31.3% for all, 45.1% for children, and 17.8% for adults among seriously injured occupants. The overall rate of serious injury was 10.0 times higher when unrestrained than restrained (3.53% ± 1.72 vs. 0.35% ± 0.12%); the injury rate was 5.6 times higher for children and 20.2 times higher in adults. For crash-exposed second-row occupants (MAIS 0 þ F), restraint use was more than 82%.
The available electronic cases involving serious injury (MAIS 3 þ F) were downloaded and reviewed. More than 69% of children and 46% of adults were involved a multiimpact collision. The majority involved 3 or more occupants in the vehicle. The case review also indicated that many young children were improperly restrained, including some 0-to 1-year-old children placed in forward-facing child seats and some children unrestrained in the child seat. Table 3 summarizes the mechanisms for serious injury (MAIS 3 þ F) from the individual electronic case reviews. Based on weighted data, the primary mechanisms for child injury were related to intrusion (86.0%), with 55.3% related to direct force, 27.5% to compression into the front interior, and 3.2% to being accelerated into an interior impact. Fourteen percent were not related to intrusion, with 12.3% involving the front seat rotating rearward into the child and 1.7% being accelerated into an interior impact. About half of the children were restrained, although most children <1 years old were in forward-facing child seats. The primary mechanisms for adult injury differed from those for children. The primary mechanisms for serious injury to adults seated in the second row were related to intrusion (68.0%), with 46.2% related to direct forces, 15.3% to compression into the front interior, and 6.4% to being accelerated into and interior impact. For this group, 21.6% were not related to intrusion, with 19.1% involving forces of vehicle acceleration and 2.5% the front seat deforming rearward. About 17% of adults were ejected in the collision based on the weighted data; all were unbelted.

Discussion
This study provides insight into second-row occupant injury frequency, rate, and injury mechanisms in rear impacts. The  results indicated that serious head injuries were most common in children, whereas head and thoracic injuries were most frequent in adults. Head and thoracic injuries were also frequent among adult front seat occupants in rear crashes (Tatem and Gabler 2017).
Second-row occupant kinematics and biomechanical loads have been assessed in rear sled and crash tests using anthropomorphic test devices (Viano and Parenteau 2018;Stephens et al. 2020). Occupants move rearward relative to the vehicle interior in rear crashes. Occupant loading generally occurs first from the pelvis and then to the torso, shoulders, and head. Significant head impacts with the child restraint have been observed when the occupant is initially leaned forward (Stephens et al. 2020). Blunt impacts can occur when the head and upper torso contact the vehicle interior and/or intruded structures . Intrusion dynamics are an important factor in child and adult injury.
More than 70% of serious thoracic injuries were from contact with the seatback. The location of the seatback (i.e., front, rear, or adjacent seatbacks) is unknown in NASS-CDS coding. Seatback can be at either the seated location or "other" location in CISS. The case review showed that most thoracic injuries resulted from second-row intrusion at the occupant location. Rear torso loading has been observed in field and crash test data. Viano and Parenteau (2008b) reviewed NASS-CDS data involving a seriously injured driver in 56þ km/h rear crashes. The authors noted that second-row intrusion was a factor in the injury mechanism.  analyzed 4 rear vehicle-to-vehicle crash tests with an instrumented lap-shoulder-belted 5th percentile female Hybrid III dummy in the right rear. They observed that second-row intrusion occurred early and pushed the right-rear occupant forward. The peak in chest acceleration coincided with the second-row intrusion. There are biomechanical data on blunt impact to the back of the chest. Viano et al. (2001) reported on 23.4 kg pendulum impacts at 6.7 m/s to the back of 8 postmortem human subjects and observed significant injury to the rib cage and thoracic spine.
The field data on serious injury to second-row occupants in rear impacts included a mix of vehicle types, age, and design. Some cases involved heavy trucks as the striking vehicles. The risk for serious injury varies with the type of vehicle or object impacting the rear. Viano and Parenteau (2022) found that impacts with a tractor-trailer or heavy truck involved 8.2 times the relative risk and impacts with fixed objects involved 4.5 times the risk. Striking vehicle weight likely reduces the effectiveness of countermeasures, including restraint.
This study included 1994þ model year vehicles to increase the sample size. This involves older vehicles. Parenteau, Viano, and Lau (2022) found that half of fatally injured second-row children in rear impacts were in vehicles that were 10 years old or older. Changes have been made in the second row as a result of FMVSS 301R, head restraint, lap-shoulder belt, and cargo retention requirements. The second-row seats in passenger vehicles are often integrated into the vehicle structures. Many modern vehicles have second-row seats certified to the European cargo retention requirements. This has resulted in stiffer rear seatbacks. In SUVs, second-row seats are often stand-alone bucket seats with options for movement to facilitate cargo or removal from the vehicle.
Various factors affect the force on an occupant, including the velocity of intrusion, velocity change of the occupant compartment, compatibility between the vehicles, restraint use, and initial occupant posture. The case review identified ejection and lack of restraint as frequent factors in the adult population. Out of the 43 adult unweighted cases, 9 were with complete ejection. O' Day and Scott (1984) found 40 times greater risk of fatal injury in ejected compared to nonejected occupants. However, the analysis was conducted in the early 1980s and involved largely unbelted occupants. In this study, all of the seriously injured second-row occupants with complete ejection were unbelted; 4 were fatally injured, reflecting a continued high risk of severe injury with ejection. Part of the effectiveness of seat belt use is the prevention of ejection provided by the lap belt (Evans 1991).
Restraint use was greater than 80% for second-row crashexposed (MAIS 0 þ F) occupants, irrespective of age group. Restraint use for seriously injured second-row occupants in rear crashes was much lower at 31.3% overall, 45.1% for children, and 17.8% for adults. Restraint systems are effective in the reduction of acceleration forces from interior and exterior impacts. Most of the benefit is in reducing occupant movement in the vehicle. The case study of the collisions indicated that many involved multiple impacts. Belts are less effective in multiple crashes than in single events, and when improperly used.
The case review indicated that many young (0-2 years old) children were improperly restrained. Some 0-to 1-yearold children were in forward-facing child seats, some were not secured to the child seat, and 1 was on the lap of another occupant. One 21-month-old child was in a booster seat.

Limitations
Some of the analysis was based on a small number of unweighted data. Generally, more than 10 unweighted cases give a reasonable estimate. When the number is less than 10, the weighted estimates are highlighted in gray and should be viewed with caution. In our experience, the national estimates are reasonable and stable, but there are high uncertainties, and the data should be used for trends.
This study identified seriously to fatally injured secondrow occupants using various variables, including MAIS (see Appendix A). The data included calendar years 1994 to 2020. Various AIS coding versions were used by crash investigators over the years. The effect of AIS coding variations was not analyzed in this study and may influence the results.

Funding
The author(s) reported there is no funding associated with the work featured in this article.