Introduction

According to the United Nations, people aged 60 years and over are predicted to reach a share of 34% of Europe’s population in 2050 [2]. They still actively take part in daily life, including working and traffic participation. Regarding the time they spend in traffic daily (on average more than 1 h), the German mobility panel 2013 [3] shows that around 45% of the time are traveled in a car and around 30% are spent as pedestrians.

The increased share of elderly in the population and their vulnerability are reflected in the increase of geriatric trauma as several studies have shown [4–8]. Trauma patients, including road traffic accident (RTA) casualties, aged 65 or older, have a higher mortality [5, 9]. Age alone [10] but also preexisting diseases, functional decline, and co-morbidities coming with age [11–13] seem to increase mortality rates in the elderly.

In comparison to gender and body mass index, age was shown to be a more significant predictor for Abbreviated Injury Scale 3+ (AIS3+) injury risk for more body regions by crash type, especially for thorax and head injuries [14]. In car passengers, the influence of airbag and steering wheel on the cause of injury seems to decrease for the elderly in favor of seat belts. Furthermore, the increasing risk of thoracic injury with age was shown to be significantly higher for women than for men [15]. It was also demonstrated that from 65 years on osteoporosis was the most common contributing factor for injury risk, followed by obesity [15].

Age-dependent changes in bone substance and thorax geometry like single rib and sternum morphology and rib angles have been investigated extensively [16–24]. However, their influence on rigidity and response in terms of deflection and deformation and their contribution to rib fracture risk are not finally solved [25, 26]. Accordingly high-standard finite element (FE) human models are developed, improved, and used for biomechanical impact and injury simulations [17, 22–31]. Load limiters are established, and seat belt pretensioners have been shown to protect elderly car occupants from Maximum Abbreviated Injury Scale 3+ (MAIS3+) thorax injuries [32]. Age elevates the risk for rib fractures, and increasing numbers of rib fractures are contributing to worse outcomes [33–35]. Kent et al. [19] reveal that older drivers are significantly more likely to die of a chest injury and less likely to die of a head injury.

In-hospital mortality is higher for elderly patients due to frailty [36], co-morbidities, and probably due to a so-called inflamm-aging phenomenon [37]. Currently, only little is known about death on scene [38]. In many countries, including Germany, a 60% rate of motor vehicle collision and or trauma fatalities are estimated not to reach hospital [39–43], and therefore, those patients will not occur in clinical trauma registries. Preventing death on scene by focusing on these 60% of RTA fatalities could contribute to the reduction of RTA fatalities as claimed by the Decade of Road action 2011–2020 [44]. It might be assumed that death on scene occurs in cases with critical and maximum injury severity, however, in the elderly frailty (bones, but probably also lung capacity and circulation) also might contribute to mortality. In general, trauma deaths are reported to be preventable in around 40% [39, 45]. The rate within the deaths on scene remains unknown. The question arises, if identifying elderly high-risk patients [46–51] is complicated by a minor manifestation of severe injuries, in addition to the higher risk of dying from comparatively minor injuries.

It is widely accepted that age increases the injury risk and the severity of injuries in comparable accident situations. Furthermore, elderly with the same injury severity as younger patients show a higher mortality. What we do not know is the following:

  • Do elderly RTA fatalities show a different injury pattern compared to the younger?

  • Are injury patterns of fatalities different for death on scene and death in hospital, depending on age?

  • Are rib cage fractures a frequent cause of death in elderly fatalities?

Material and methods

Sample definition

The sample consists of traffic accident fatalities, including death after 30 days, who were autopsied at the Institute of Forensic Medicine, Ludwig-Maximilians-University (LMU) in Munich, in the years 2004 and 2005. Fatalities with death due to other reasons than traffic accidents (e.g., suicide) or with an unclear causal relationship between accident and death were excluded, as well as children younger than 15 years. The capture area roughly covers Southern Bavaria. The autopsy rate is known to be higher for the area of Munich as the capital city, compared to the surroundings.

Data extraction and categorization

Extracted data from autopsy reports and public prosecution files were sex (male/female), age in years, site of accident (urban/rural/autobahn/unknown), traffic participation (bicycle/pedestrian/powered two-wheeler/passenger car/other), time to death (exactly in hours and days), site of death (on scene/in hospital), cause of death, and all injuries. Postmortem computer tomography had not been performed in 2004 and 2005, so that possibly, additional injuries that can better be traced by imaging as for example the facial skeleton, spine, pelvis, extremities [52, 53], and bone bruises [54] might be underrepresented.

Injuries were coded according to the Abbreviated Injury Scale, AIS© 2005 update 2008 [1], by trained but not certified coders and control readers (C.H., A.F., D.H.). In case of need, coding was discussed and decisions were made with a supervisor (S.S.). The MAIS, the Injury Severity Score (ISS) (grouped to <15, 16 to 32, 33 to 66, and 75), and the region in which the MAIS occurred (detailed and categorized to Head/Thorax/HeadandThorax/other(s)) were documented, as well as the MAIS in every body region (named HeadMAIS, ThoraxMAIS, etc.), the appearance of body region AIS3+ (yes/no), and the number of body regions with AIS3+, AIS3+ indicating an AIS of 3 points or higher and being at least a serious injury. The body regions we used represent the AIS body regions 1 to 9 [1], except for the spine where we coded cervical spine to neck, thoracic spine to thorax, and lumbar spine to abdomen and regarded the pelvic bone separately. AIS3+-injured body regions as yes/no were calculated as sample percentages.

Bony thorax injuries were analyzed in detail. The number of rib fractures and the number of fractured ribs were documented. In case of expressions like multiple fractures of all ribs on the left side/all ribs on the left side are multiply fractured in the autopsy report, the minimum number of fractures mentioned was counted. In the example given, multiply fractured is translated to at least two fractures per rib and therefore counted as 24 fractures (on 12 ribs). Rib fractures clearly attributable to resuscitation were not taken into account.

In addition, we created a variable indicating that a rib cage fracture (excluding thoracic spine) is the most severe injury of the regarded person and therefore leads to the ThoraxMAIS and overall MAIS. Hence, any other injuries occurring in this person have lower AIS severities.

Data analysis

This descriptive but exploratory cross-sectional study includes only fatalities, and therefore, mortality cannot be derived. There is no case control study performed; neither risk factors for fatal outcome nor survival are analyzed. The age group of the elderly (65+ years) is compared to the younger age group (up to 65 years, excluding children younger than 15 years) to see if there are any differences in the injury patterns. As different accident circumstances might contribute to different injury patterns, the variables mentioned above are analyzed for both age groups. This comparison is done separately for death on scene and for death in hospital to be able to identify any possible differences and to compare the results with literature.

For data analysis, Microsoft Office © Excel 2010 was utilized and statistical testing was done with IBM ® SPSS ® Statistics v23.0.0.0. For sample description, a cross tabulation of sex, site of accident, and traffic participation for both sites of death are performed and only the results of the chi-squared tests will be reported. Bivariate frequency distributions for listed variables (see above) for both age groups are calculated. Chi-squared tests are applied for nominal and ordinal variables; Mann-Whitney U test for independent samples is used to compare frequencies of rib fractures and of fractured ribs. The association between number of rib fractures and number of fractured ribs with the ThoraxMAIS is tested by median and Kruskal-Wallis test. As this study should be seen as a first screening analysis, the significance level is set to 5% and Bonferroni correction is not applied. The results of the bivariate analyses are presented in tables (numbers and percent), results towards rib cage fractures as most severe injury of the person are presented in absolute numbers, and the numbers of rib fractures and fractured ribs are displayed by box plots.

The percentage of fatalities showing an AIS3+ injury in a given body region is grouped to percentage ranges covering 0 to <20, 20 to <40, 40 to <60, 60 to <80, and 80 to 100%, depicted in different shades from white to black and graphically presented in a body scheme. This analysis is performed for young and old at both sites of death and each for passenger car occupants and pedestrians, urban and rural, and male and female. For other groups of participation (bicyclists, powered two-wheelers, and others) and accident site autobahn, the case numbers are too low. Statistical tests are not performed.

Representativeness

We compared data from the Bavarian State Office for Statistics and Data Processing of traffic fatalities (age >14 years) within the capture area with our autopsied fatalities. Traffic fatalities are defined by death within 30 days; therefore, we excluded autopsied fatalities that died after 30 days for the representativeness analysis. The differences between age groups (up to 65/65+ years), sex, and traffic participation are investigated. As the site of death (on scene/in hospital) is not documented by the statistical office, representativeness cannot be assessed for this aspect. It is analyzed whether sex and traffic participation are distributed accordingly within each age group and furthermore, if the age groups are equally represented. The findings will be used for discussion, not for extrapolation calculations.

Results

The LMU autopsy material of RTA victims from 2004 to 2005 consists of 309 fatalities of which 89 (29%) are equal or older than 65 years. Three-hundred-one persons with death within 30 days are regarded for representativeness checks. In the capture area, there had been 1093 RTA fatalities in the years 2004–2005 of which 255 were 65 years or older (23.3%). Thirty-four percent of the 65+ years were autopsied at our institute but only 26% of the group up to 65 years (see Electronic Supplementary Material (ESM) Tables 3 and 4).

Both age groups of our sample show a similar sex distribution as found in the traffic fatalities population (see ESM Tables 1 and 2); however, concerning the mode of traffic participation, both age groups show an overrepresentation of pedestrians. The autopsy rates for males and females are comparable. In the younger age group, 79% of pedestrian fatalities were autopsied but only 22% of the passenger car fatalities. In the group of the elderly, it was 57% of the pedestrians and 27% of the passenger car fatalities. Detailed numbers and frequency distributions can be found in the ESM Tables 1–4.

All 187 victims that died on scene (61% of all fatalities) were declared dead by an emergency physician either with or without resuscitation on scene. The maximum time between accident and declaration of death on scene was 1.17 h. Of each age group, 95% were declared dead within 35 min. The surviving time after the accident of in-hospital deaths varied widely; the median for the up to 65-year olds was 14.7 h (range half an hour up to 123 days). The group of 65+ year olds had a median survival time of 27.7 h (range 47 min up to 59 days).

Accident-, person-, and injury-related characteristics of both age groups with death on scene

Of all 187 victims dying on scene, 39 (21%) are in the age group 65+ years. There are more male fatalities in the younger group (see Table 1).

Table 1 Injury characteristics of different age groups of traffic accident fatalities with death on scene
Full size table

The cross tabulation shows a significant difference (p = 0.009) for the combination of age group, sex, site of accident, and traffic participation mode with a high share of young male motorcycle riders in rural accidents. Seven male and seven female elderly fatalities happened at urban sites, among them five females and two males as pedestrians. Due to low case numbers, a significant difference cannot be found for this subgroup.

In the bivariate analysis, a higher share of urban accidents is seen in the older age group (36 vs. 9%) and a higher share of powered two-wheelers in the younger group. Still, the fatal accidents most frequently occurred on rural roads in both age groups (74 and 56%; see Appendix Table 3).

An ISS of less than 16 is seen in 3.4% of the younger and none of the older fatalities. The most severely injured body region for age less than 65 years most frequently is the head (40%) and for 65+ year olds the thorax (39%). The elderly are discovered to have fewer seriously injured body regions. In the group up to 65 years, 23% only have up to two body regions with AIS3+, and in the 65+ year olds, there are 34%.

Thirty-six percent of the elderly show a MAIS of six in the thorax region, whereas in the group of up to 65-year olds, there are only 20% (n. sign; not shown in table). The person with MAIS2 (Table 1) remained with an unclear death, possibly already dying from internal reasons before crashing into a heavy goods vehicle at 30 km/h. In both age groups, the cause of death was documented by the forensic pathologist to be multiple trauma most frequently (55% of the younger, 62% of the older).

Figure 1 visually represents the frequencies of victims showing at least seriously injured (AIS3+) body regions. Of the two age groups, 85 and 95%, respectively, suffer from AIS3+ thorax injuries. Serious to maximum head injuries are seen in 76% of the younger but only in 59% of the elderly. Nearly half of the younger (46%) have AIS3+ abdominal injuries, but only 31% of the elderly do. Further, 49% of the elderly show AIS3+ pelvic fractures but only 39% of the younger.

In urban as well as in rural accidents, the younger more often show serious injuries to the head, abdomen, and lower extremities and less often serious thorax injuries and pelvic fractures compared to the elderly (see Appendix Fig. 7). When separately regarding passenger car occupants and pedestrians, the differences also remain; however, the elderly pedestrians more often show lower leg injuries (see Appendix Fig. 6).

This pattern also remains for males, whereas elderly females additionally show serious abdominal and lower leg injuries more frequently than the younger females (see Appendix Fig. 5).

Rib fractures in both age groups and rib cage fractures as most severe injury; death on scene

In the age group up to 65 years, there is one person with rib fractures being the most severe injuries and leading to the ThoraxMAIS5 and overall MAIS5. In the age group 65+ years, there is no one to whom that applies. Yet, the number of rib fractures and the number of affected ribs are strongly associated with the ThoraxMAIS (all tests p = 0.000).

The number of rib fractures (median 20 vs. 11, ranges from 0 to 64 and 0 to 47, respectively, p = 0.000) and the number of affected ribs (median 18 vs. 10, both ranging from 0 to 24, p = 0.000) are higher for the elder group compared to the younger (see Fig. 2).

Fig. 1
figure 1

Percentage of victims showing AIS3+ injury in different body regions for both age groups, death on scene (no visual representation of external body region), and body regions with higher percentages in comparison to the other age group are marked as shaded cells in the tables

Full size image
Fig. 2
figure 2

Distribution of rib fractures (number and number of ribs affected) for both age groups with death on scene

Full size image
Fig. 3
figure 3

Percentage of victims showing AIS3+ injury in different body regions for both age groups, death in hospital (no visual representation of external body region), and body regions with higher percentages in comparison to the other age group are marked as shaded cells in the tables

Full size image
Fig. 4
figure 4

Distribution of rib fractures (number and number of ribs affected) for both age groups with death in hospital

Full size image

Accident-, person-, and injury-related characteristics of both age groups with death in hospital

Of all 122 victims dying in hospital, 72 are up to 65 years old and 50 are in the age group 65+ years. There are more male than female fatalities in the younger age group (64%) and more females in the elderly (54%, p = 0.050; see Table 2).

Table 2 Injury characteristics of different age groups of traffic accident fatalities with death in hospital
Full size table

The cross tabulation does not show a significant difference (p = 0.187) for the combination of age group, sex, site of accident, and traffic participation mode. Still, it seems noteworthy that in urban accidents, 11 out of 13 elderly females are pedestrians, and in the elderly males, there are only 9 out of 15.

However, the site of accident shows a higher share of urban accidents in the older age group, and the traffic participation shows a high share of elderly pedestrians (see Appendix Table 4).

More than 50% of the elderly show an ISS 16 to 32 and more than 50% of the younger an ISS 33 to 66 (see Table 2). It seems remarkable that 14% in the younger group are attributed an MAIS of 6. The person with MAIS2 died from a paralytic ileus after 14 days.

The most severely injured body region for age less than 65 is the head (50%) and for 65+ year olds in 36% the head and in 30% the thorax. The elderly have fewer seriously injured body regions. Fifty-two percent of the younger and 72% of the elderly only have up to two AIS3+-injured body regions.

The younger in-hospital deaths die from multiple trauma and central regulation failure in 82%, the elderly only in 64% (see Table 2).

About 75% of both age groups suffer from AIS3+ thorax injuries (see Fig. 3). Eighty-two percent of the younger show serious to maximum head injuries, but only 66% of the elderly do. Further, 29% of the younger show AIS3+ abdominal injuries, and in the elderly, it is only 14%.

The tendency of the young showing serious head, abdominal, and lower extremity injuries more frequently, and a similar share of pelvic fractures compared to the elderly remains the same when looking at passenger car occupants and pedestrians separately (see Appendix Fig. 9). Only exception is serious thorax injuries which shift to higher shares in elderly pedestrians compared to the young.

The described injury pattern for young and old stays the same when looking at rural accidents. However, in urban accidents, serious pelvic fractures are found more often in the young, and a higher share of serious thorax injuries is seen in the old, whereas head and lower extremity injuries are found about equally frequent (see Appendix Fig. 10).

There is a change in the injury pattern when separating by sex; AIS3+ head injuries are seen more often in the young females compared to the old, but for male fatalities, the frequencies are about the same for both age groups. Serious thorax injuries are rather found in the elderly female and in the young males. Serious pelvic fractures appear more frequently in young males compared to the old and in the old females compared to the younger (see Appendix Fig. 8).

The abdominal injury frequency remains as described, higher shares for the younger ones in males and females, on both sites of accident, and for both analyzed modes of traffic participation.

Rib fractures in both age groups and rib cage fractures as most severe injury; death in hospital

In the age group up to 65 years, there are seven persons with rib cage fractures being their most severe injury and leading to their ThoraxMAIS and overall MAIS, one person with MAIS4 and six persons with MAIS5. In the age group 65+ years, there are five persons, three with MAIS4 and two with MAIS5. Yet, the number of rib fractures and the number of affected ribs are strongly associated with the ThoraxMAIS (all tests p = 0.000).

Comparing the age group up to 65 years with the age group 65+, the number of rib fractures (median 2 vs. 9, ranges from 0 to 48 and 0 to 54, respectively, p = 0.095) and the number of affected ribs (median 2 vs. 7, ranges both from 0 to 24, p = 0.041) are higher for the elderly (see also Fig. 4).

Discussion

For the RTA fatalities with death on scene, it firstly can be seen that young and old do not show significant differences in the injury severity. In more than 50% of both age groups, non-survivable conditions (AIS6—currently untreatable and ISS 75, respectively) were present.

Yet, the most severely injured body region most frequently was the head for age less than 65 years and the thorax for the 65+ year olds. The elderly show a MAIS of six in the thorax region in 36% of cases, the younger only in 20%. The thorax as body region with the MAIS is seen in 39% of the elderly and 23% of the younger. These findings match the age-dependent injury patterns described by Kent et al. in 2005 [19], who report 47.3 and 24.0%, respectively. For the in-hospital deaths, we see 30% of the elderly and 18% of the younger with the MAIS in the thorax region.

The elderly with death on scene show higher shares of at least serious thorax injuries compared to the young, still when separated to sex, accident site, and traffic participation. For the in-hospital fatalities, there is no clear difference between old and young in the frequency distribution of AIS3+ thorax injuries. Only in the elderly females, higher shares are seen compared to the young, as well as in the elderly fatalities participating as pedestrians and in urban accidents. These findings could be explained by the high percentage of females in pedestrian and urban accidents. The different trends for males and females were also reported before [14, 15], yet, we can only confirm this for the in-hospital deaths.

The analysis whether the rib cage injury was the most severe injury and lead to the MAIS shows low numbers for both age groups with death on scene. In the group of the 15 elderly who died of an MAIS in their thorax region, there was no one dying from rib cage fractures, which is remarkably lower than the 56% of in-hospital fatalities, older than 60 years dying solely from rib fractures if the MAIS was in the thorax region, that are reported by Kent et al. in 2008 [34]. Also, our in-hospital deaths show only 5 out of 15 (one third) in the elderly with bony thorax fractures leading to the MAIS. One reason for this strong deviation might be that they [34] included cases in which body regions other than the thorax additionally showed the same MAIS value. The rib cage fractures as cause of death do not seem to be very important for death on scene but more relevant for the in-hospital fatalities. On both sites of death, the rib cage fractures as most severe injury of the person are of severities AIS4 and AIS5, implying that neither sternum fractures nor simple rib fractures of any number, nor unilateral flail including up to five ribs, contributed to a fatal outcome. This is in line with Borman 2006 [55] and Huber et al. 2014 [56], who confirm a higher mortality for in-hospital patients by thorax injuries only for the bilateral flail chest (AIS5).

Still, rib fractures are frequently named as being associated to a higher mortality risk, especially for the elderly [33, 34, 57–60]. As we cannot derive mortality risks from our study, we can only confirm higher rates of rib fractures in the elderly compared to the younger for both in-hospital fatalities and also for the deaths on scene; a causal relationship between rib fractures and death should be doubted [56, 61]. Further, we can fortify the association between number of fractured ribs and number of rib fractures to the thorax injury severity in general.

We found serious to maximal abdominal injuries more frequently in the young fatalities. For death on scene, this can only partly be explained by a higher share of males, as this relation is not found in females. An explanation could be the young males’ participation as motorcycle riders where high shares of abdominal injuries are known as contributing to mortality [62]. However, the difference between old and young for abdominal injury is also found in passenger car occupants, pedestrians, and also in the in-hospital deaths with a stable relation throughout all subdivision to sex, accident site, and traffic participation. Concerning the elderly with comparably lower abdominal injury rates, yet higher thorax injury rates, this finding could be interpreted as kind of a trade-off. One geometric element that might explain the differences for abdomen and thorax injuries is the comparatively more pronounced kyphosis of the elderly thoracic spine. However, this hypothesis needs further research.

The pelvic bone is seen to be more affected in the elderly with death on scene compared to the younger (49 vs. 38%), even when separating to traffic participation and sex, except for urban accidents. Seat belt geometry and load limiters could help to protect the car occupants [32] by addressing thorax, abdomen, and pelvic bone. However, this difference is also seen in the pedestrians. Especially in pedestrians, a high pelvic fracture frequency in general has been described [63–65]. The relevance of pelvic fractures is frequently addressed [66–68]; however, we can only confirm an age effect like Kimbrell et al. and Toth et al. [69, 70] for the deaths on scene, where the high share of rural accidents might contribute to this finding. Although we found a higher share of urban accidents in the older age group, still, 22 of the 39 accidents (56%) occurred on rural roads; for the younger aged, the share was 72%. In Germany, 58% of the traffic fatalities stem from rural roads in 2015 (see Tables 3.1_(4) and 2_(2) in “Verkehrsunfälle Zeitreihen” of the Statistic Office [71]). Rural accidents are thought to be more severe [72] due to higher velocities than in urban areas, and in contrast to motorways, most often, there are oncoming traffic and unfavorable roadside conditions present [73]. The 65+-year-old drivers in Germany were more frequently involved in turning (especially turning left), crossing (especially collisions with bicycles from the right when entering a priority road), and pedestrian accidents compared to the younger [74]. Only the first situation (turning left) might lead to a fatal outcome for the older driver; therefore, it seems obvious that fatal accidents need to be analyzed separately. From our data, we see higher numbers of pelvic bone injuries in the elderly passenger car occupants with death on scene; however, to conclude a higher share of side impacts would be too far-fetched. The direction of impact influences the injury pattern and needs to be taken into account in a following analysis with increased case numbers.

More than 50% of both age groups with death on scene show an ISS of 75. This confirms the hypothesis that a death on scene is mostly due to non-survivable injuries. When looking at the deaths in hospital, in the group of fatalities up to 65 years, 14% had a MAIS of 6, while in the group of elderly, there were none. A possible explanation could be that intensified resuscitation measures are taken to save children and younger adults or that they are able to survive longer with a maximum injury (MAIS6), and therefore, more of them manage to leave the site of accident and can be brought to hospital. It needs to be analyzed if undertriage and less aggressive treatment could be a reason for elderly casualties not to reach hospital, as is reported in different studies [46–51, 75].

Our data reveals the tendency that in traffic fatalities aged 65 and older, fewer body regions with an AIS3+ are found than in the group up to 65-year olds. That observation applies for both death sites, on scene and in hospital. Already, Osler et al. [76] found that elderly people more often have fatal outcomes than the younger but are less likely to be injured in the first place. Looking at the ISS of deaths in hospital, there are significantly more severely injured fatalities in the younger group than in the elderly (p = 0.003). This finding is consistent with Chiang et al. [77], who use an ISS cutoff value of 15 to classify trauma as major or minor and show that compared to the younger patients (ISS = 17) among the elderly (65 years and older), a lower ISS cutoff value predicts a higher mortality rate. Werman et al. developed geriatric trauma triage criteria [78], and Calland et al. request to treat all elderly trauma patients (65+ years) with one body region of AIS3+ in a trauma center [79]. Pape et al. [80] include the factor age in the “Berlin definition” as one parameter (age ≥70 years), contributing to the definition of polytrauma. According to them, still two or more AIS body regions need to be injured with a severity of at least AIS3. Ninety-two percent of our fatalities on scene in both age groups show at least two body regions with AIS3+ (in hospital around 75%).

On scene, multiple trauma was the most frequent cause of death (57%), followed by central regulation failure (28%). Three studies report multiple trauma as a cause for on scene deaths in 59% [41] and for early deaths in 40% [81] and 16% [82]. The central nervous system as the cause of death is reported by different authors in 49% [83, 84], 21% [81], 27% [82], 46% [85], and 73% [42]. Especially for in-hospital early deaths, bleeding seems to be a common cause of death today [86, 87] and also still the central nervous system [42, 85, 88]. In our sample, bleeding as cause of death (in hospital) is only found in 8% (young) and 12% (old). In all studies mentioned, different definitions of early and late death as well as different inclusion criteria (all trauma including falls or only multiple trauma patients as well as all fatalities or all hospital admittances) and different coding habits are found, so that a direct comparison is not possible and might explain some of the deviances between our findings and literature. Yet, it indicates that when focusing on death on scene of RTA victims, the frequency distributions of causes of death are shifting towards multiple trauma.

From the results found when focusing on death on scene in comparison to deaths in hospital, it would be advisable to perform future studies in order to analyze risk factors for mortality and detect possible preventive factors. These studies could combine and match data, e.g., from the TraumaRegister DGU® for our capture area; increase case numbers; and draw a representative sample for conducting, e.g., a case control study.

The selection of cases is conditioned by the decision of the responsible state attorneys whether an autopsy has to be performed. Yet, comparable to the traffic fatalities in our capture area with a share of 23% elderly, we find a share of 21% in our sample with death on scene, however a share of 40% with death in hospital. The pedestrian is more frequently autopsied than any other traffic participant. As pedestrians clearly are vulnerable road users, it is necessary to perform an autopsy in addition to accident reconstruction for determining the accident causes as it has legal consequences for the involved opponent. The Statistical Office presents a share of 24% pedestrians and 47% car occupants in the elderly traffic fatalities. The time they spend in traffic shows comparable frequencies [3], so that an increased risk by the mode of traffic participation cannot be concluded for our capture area.

The representativeness checks revealed that traffic fatalities’ analysis of our sample needs to be performed by taking age, sex, site of accident, site of death, and traffic participation into account.

To our knowledge, this is the first study presenting injury pattern separately to age groups and taking site of death into account. Following analyses with increased case numbers also regarding direction of impact, impact speed, and opponent will lead to results that are more robust.

Conclusion

Do elderly RTA fatalities show a different injury pattern compared to the younger?

Yes, elderly RTA fatalities do show different injury patterns compared to the younger; these differences can only partly be explained by different distributions of sex, traffic participation, or site of accident within the age groups.

Are injury patterns of fatalities different for death on scene and death in hospital, depending on age?Yes, it seems that elderly with death on scene more often show serious to maximal thorax injuries and pelvic fractures than the younger. Elderly with death in hospital less frequently show AIS3+ abdominal and head injuries compared to the younger. A lower injury severity in the elderly is seen more clearly for in-hospital fatalities.

Are rib cage fractures a frequent cause of death in elderly fatalities?

Bony thorax injuries and rib fractures are very seldom the cause of death for elderly RTA fatalities, although we did find the thorax to be the most severely injured body region most frequently and observed rib fractures to occur more often in the elderly. Rib fractures should be regarded as an indicator for the injury severity of the thoracic organs.

Our study shows that for developing new ideas for prevention measures, the RTA fatalities that die on scene must not be neglected. Death on scene should be analyzed apart from death in hospital and should at least be stratified to age, sex, accident site, and traffic participation mode. It is advisable to take further steps, increase the data material, and analyze the deaths on scene more in-depth, with a view on both, injury, and accident mechanisms.