Driver Injury Patterns in the United States and Australia: Does Beltwearing or Airbag Deployment Make A Difference?

Abstract

Airbags in the United States and Australia are fundamentally different. In the U.S., airbags are designed as primary restraints to protect unbelted occupants while those in Australia are design as supplementary to the seatbelt. The deployment thresholds, power, and velocity of deployment therefore differ. Using a cohort comparison method, this study set out to determine if the injury patterns of belted and unbelted drivers in airbag deployed crashes were different in the United States and Australia. This study focussed only on intermediate and full-size passenger vehicles to allow comparability between popular Australian vehicles and US vehicles. The results suggest that US belted and unbelted drivers tend to suffer a higher percentage of injuries to the face, thorax, and upper extremity as a result of airbag contact compared with the Australian sample. In addition, unbelted US drivers appeared to suffer a higher percentage of head and facial injuries from non-airbag contact sources than the US belted and Australian samples.

AIRBAGS are designed primarily to protect occupants from serious head, neck, and chest injuries. In the United States, where safety belt use is about 68% (NHTSA, 1996), Federal Motor Vehicles Standards (FMVSS 208) require that they protect unrestrained occupants. In Australia, where safety belt use has reached 95%, airbags are designed as “Supplemental Restraint Systems”, that is, to supplement a 3-point seat-belt system. This difference in design philosophy has led to different airbag systems in the two countries. Airbags in the United States are larger and deploy faster at lower crash thresholds than in Australia.

Although there are differences between airbag systems in individual cars in both countries, US airbags tend to be more aggressive in the rate of inflation compared to the Australian airbag. In addition, the US bag is typically quite large (70–80litres), while the Australian bag is generally smaller (30–60litres). The driver’s airbag is in the same position in both countries, with the module centred in the steering wheel. The driver sits directly in front of the module with the seat and the steering wheel placed in any position that he or she deems comfortable.

There have been a number of reports of serious injuries attributed directly to the action of an airbag in the US, including injuries to the driver’s face, eyes, neck, and upper extremity (Beckerman and Elberger, 1991. Dumas, Kress, Porta, et al. 1996. Freedman, Safran, Meals, 1995. Huelke, Moore, Compton, et al. 1995. Marco, Garcia-Lopez, Leon, et al. 1996.). In addition, there are over 52 cases of air-bag-induced serious or fatal injury in crashes where fatal injury would not otherwise be predicted. Passenger side airbags have been implicated as the cause of serious or fatal injury 79 adult and child front seat passengers (NHTSA, 1999). No similar cases have been reported in Australia (Fildes, 1997).

In this paper, we set out to characterize driver’s injury patterns in non-fatal, air-bag-deployed crashes and to compare patterns between the US and Australia. Given the design differences, it was expected that the injury patterns would differ with fewer injuries generally to occupants of passenger cars in Australia.

METHOD

Two data sets were created and compared for this analysis. The US source data was derived from the publicly available National Accident Sampling System (NASS). Australian data was derived from the Australian Crashed Vehicle File developed and held at the Monash University Accident Research Centre. Both Australia and the US use a similar data collection procedure developed by the United States National Highway Traffic Safety Administration (NHTSA, 1988). Cases selected for this analysis were confined to frontal impact crashes where the driver’s airbag deployed and the driver suffered an injury. The weights within the NASS data were not used, as the specific injury data was most relevant. The Australian cases represent a random sample of popular large passenger vehicles.

Equivalent barrier speed (EBS) was selected as the marker for crash severity. As the determination of delta-V requires inspection of both vehicles in a two-car crash (which was not always possible), there were fewer values available for this factor than EBS. The NASS data included the years 1995 and 1996, while Australian cases were extracted from the Australian Crashed Vehicle File for the years 1993 to 1997. Crashes at an EBS of greater than 65km\h were excluded from the analysis as airbags would not be expected to be as effective in mitigating injury severity in very high impact crashes. The current FMVS208 test speed is 48km/h at which occupant protection is measured. Hence, there is no guarantee that the airbag or vehicle structure will offer protection in crashes that involves an energy level greater than that of the current standard. It is also important to note that airbag effectiveness also depends on the structural integrity of the vehicle within a crash. In an observation paper, Paine, McGrane and Haley (1998) concluded that the ANCAP 64km/h offset test “…places severe demands on the structure of the vehicle…some vehicles perform exceptionally well during this crash test but many exhibit excessive structural collapse and other undesirable characteristics” (p.314). Consequently, case inclusion was limited to crashes with an EBS between 1 and 65km/h.

In order to control for vehicle size, so as to allow for comparability of the data-sets, we included cars in vehicle market classes rated “full size” and “intermediate” in the US and “large” cars in Australia; wheelbases ranged from 265–291cm. Vehicle mass was restricted to 1250kg – 1750kg. Cases were further limited to that vehicle primarily sampled; i.e. “Vehicle One” from each crash. The data sets were limited to these criteria, and those crashes for which EBS, injury, driver height, weight, or age data were missing were excluded. In the NASS data, the total number of crashes involving driver air bag deployment in intermediate or full size cars with a known EBS was 165. Thirty-one crashes had an EBS of greater than 65 km/h and were excluded (12.7%), while the remaining 28 cases were excluded from our analysis due to missing data or vehicle mass out of the specified range (17%). The Australian data set contained 156 cases meeting entry criteria. Fifty-five cases were excluded due to incomplete data (35%) and there were three cases with an EBS greater than 65 km/h. The final sets comprised of 106 cases from the US and 98 Australian cases.

Data were analysed using the SPSS and SAS\STAT statistical packages. Statistical tests were parametric and non-parametric tests where appropriate (Siegal & Castellan, 1988). Type 1 error for planned comparisons was controlled for using the Modified Bonferroni method (Keppel, 1991).

RESULTS

OCCUPANT DATA

Seat belt use differed among the US sample and Australian sample, with fewer Americans using their seat belts than Australians (74.5% vs. 98%), χ²(1) = 3.2, p≥.05). Many more US drivers were classified as unbelted at the time of the accident than in the Australian sample, χ²(1) = 20.02, p≤.05). As there were only two unbelted Australian cases consequently all future analysis will refer to only three samples: US belted, US unbelted, and Australian.

There were significant differences between the US belted, US unbelted, and the Australian samples in terms of sex, height, and patterns of belt use between sexes (see Table 1). The samples did not differ in terms of injury severity scores (ISS) age, weight characteristics, and Equivalent Barrier Speed (EBS).

Table 1.

US vs. Australian Data (SD)

	Age (yrs)	Weight (kgs)	Height1 (cm)	Sex2		ISS
				Male	Female3
US Belted (n=79)	46 (20)	74 (16.6)	170 (10.4)	37 47%	42 53%	3.3 (3.1)
US Unbelted (n=27)	40.5 (21.0)	78.6 (14.9)	172.4 (8.0)	19 70.4%	8 29.6%	2.85 (3)
AUS (n=98)	41.0 (13.1)	76.2 (13.3)	173.9 (9.8)	72 73%	26 27%	2.9 (2.4)

¹ F(2,201)=4.17, p=.017;

²χ²(1)=8.74, p= .003;

³ χ²(1) = 6.5, p≤.025

The mean EBS was 34km\h (SD=14.3) for the Australian sample, 34km\h (SD=12) for the belted US sample, and 39.5 km\h (SD=12.1) for the unbelted US sample. It is also of note that the unbelted group tended to be younger, taller, and heavier than other the other groups. The gender differences more than likely account for these trends.

In the US belted sample, the number of females (42, 53%) was slightly greater than the number of males (37, 47.1%). In the US unbelted sample, the number of males (19, 70%) was greater, though not statistically significant, than the number of unbelted females (8, 30%). However, there were a significantly greater number of males (72, 73%) in the Australian sample than females (26, 27%), χ²(1) = 21.6, p≤.01.a

In examining the sex pattern across the samples, it was evident that there were more female drivers in the US belted sample (42, 53%) than in the Australian sample (26 women, 26.5%), and more females in the Australian sample than the US unbelted sample (8, 29.6%), χ²(1) = 9.53, p≤.01a. The number of males in the Australian sample was greater than the number of males in the US belted sample, χ²(1) = 11.3, p≤.01a, and the US unbelted sample, χ²(1) = 30.8, p≤.01a.

The mean height of the occupants was statistically different between the three samples, F(2,201)=4.17, p=.017 (see Table 1). However planned comparisons revealed that only the US belted sample (169.7cm) and the Australian sample differed statistically (173.9cm), t(175)=2.8, p≤ .03a. No other comparisons revealed any differences between the samples in mean occupant height.

The distribution of injuries, irrespective of contact source, for each of the sample groups by AIS severity level is presented in Figure 1. Figure 1 indicates that the distribution of injury severity is similar across the three sample groups. There were no injuries in any of the sample groups above AIS 3. AIS 1 injuries account for 83%, 86% and 90% of the injuries in the US belted, US unbelted, and Australian samples respectively. Likewise, AIS 2 injuries account for 13%, 10.5%, and 8.7% of injuries in the US belted, US unbelted, and Australian samples. A smaller percentage of injuries were AIS 3 severity level. That no statistical difference exists between the samples for injury severity distribution allows for a finer examination of the differences in injury patterns across the samples given airbag contact.

Figure 1.

Distribution of injuries by AIS severity level by sample group.

ISS AND THE SAMPLE

ISS AND EBS

Crashes were grouped into the following EBS categories: 5–15kph (3–9mph), 16–25kph (10–16mph), 26–35kph (16–22mph), 36–45kph (22–28mph), 46–55kph (29–34mph), and 56–65kph (35–41mph). Although there was a trend toward higher ISS in higher speed impacts in all three groups, this was significant only for the Australian drivers, χ²(1) = 16.4, p=.006. (See Table 2.). Low sample size in the US belted and unbelted sample precludes any meaningful analysis and conclusions to be drawn of the EBS category ISS data.

Table 2.

Mean Injury Severity Score by EBS Category (SD).

	5–15 km/h	15–25 km/h	25–35 km/h	35–45 km/h	45–55 km/h	55–65 km/h
US Belted (n=79)	1.3 (0.6) n=3	2.3 (2.3) n=19	3.3 (3.2) n=20	3.9 (4.0) n=22	4.5 (4.3) n=11	3 (4.0) n=4
US Unbelted (n=27)	1.0 (0.0) n=1	6.0 (7.1) n=2	4.0 (3.1) n=7	2.0 (2.5) n=10	3.7 (2.5) n=3	1.25 (0.5) n=4
AUS* (n=98)	1.4 (0.52) n=12	2.8 (2.4) n=16	2.4 (1.9) n=28	3.8 (3.7) n=18	3.2 (2.0) n=19	4.6 (2.8) n=5

^*Kruskall Wallis - χ²(1) = 16.4, p=.006

ISS AND SEX

Within each of the samples there was a trend for women to suffer more severe injuries indicated by ISS, and this effect was greater in the US samples compared to the Australia sample. The high degree of variance, possibly related to the small sample size, precludes any definitive conclusions to be drawn.

ISS, HEIGHT & WEIGHT

Drivers in the US and Australia were grouped into 3 categories to examine the effects of stature and sex further (see Table 4). The height categories were: short (less than 160cm, 5′3″), medium (158 to 183cm, 5′4″ to 6′), and tall (over 183cm, 6′). Similar categories were investigated for weight: light (less than 55kg, 121 pounds); medium (56–90kg, 122–198lbs), and heavy (greater than 90kg, 198lbs). The results are shown in Table 4 and Table 5. The small sample size, particularly in the ‘light’ weight category, prevents any meaningful analysis of these results. However, there was a trend toward higher injury severity in both the short and light groups that may account, at least in part, for sex differences in mean ISS (see Table 3).

Table 4.

Mean Injury Severity Score by Height (SD)

	Short < 160cm	Medium	Tall >183cm
US Belted (n=79)	3.4 (4.1) n=13	3.3 (3.4) n=53	3.1 (3.4) n=13
US Unbelted (n=27)	nil	2.7 (3.1) n=21	3.3 (3.1) n=6
AUS (n=98)	2.4(1.4) n=10	3.1 (2.7) n=69	2.3(1.5) n=19

Table 5.

Mean Injury Severity Score by Weight Category (SD)

	Light <55kg	Medium	Heavy >90kg
US Belted (n=79)	5.0 (5.3) n=10	2.7(2.8) n=55	4.2 (3.8) n=14
US Unbelted (n=27)	1.0 n=1	2.9 (3.0) n=20	3.0 (3.3) n=6
AUS (n=98)	3.1(2.4) n=8	2.9 (2.6) n=74	2.5(1.6) n=16

Table 3.

Mean Injury Severity Score (ISS) by Sex (SD)

	Female	Male
US Belted (n= 79)	3.7 (3.9) n=42	2.8 (2.8) n=37
US Unbelted (n=27)	3.75 (3.2) n=8	2.5 (2.9) n=19
AUS (n=98)	3.2 (2.2) n=26	2.8 (2.5) n=72

AIS INJURY DATA

There were a total of 407 injuries in the US sample and 366 injuries in the Australian sample. Each injury was coded using the Abbreviated Injury Scale (AIS) (AAAM, 1990), and by the source of the contact which caused that particular injury. In the first instance it was important to look at the MAIS pattern for the three groups.

Figure 2 shows the distribution of MAIS level by sample group. As with the distribution of all AIS injury levels shown in Figure 1, it is apparent that the maximum AIS injury level is consistent across the samples. This result indicates that occupants in any one sample group is not suffering more severe injuries in comparison to the other groups

Figure 2.

Distribution of injuries by MAIS by sample group.

Table 6 shows the number of injuries by contact source. There were significantly more injuries identified as caused by airbag contact in the US sample overall (belted and unbelted combined) compared to the Australian sample, χ²(1) = 44.26, p=.001. There was no statistical difference in the number of injuries caused by non-airbag contact sources between the US and Australian sample. There were a greater number of airbag contact injuries in the US belted sample compared to the US unbelted sample, χ²(1) = 58.18, p≤.001.

Table 6.

Frequency of Injury by Contact Source

	Total Injuries	Airbag contact	Non-airbag contact	Unknown
US Belted	312	30% (95)	69% (217)	1% (3)
US Unbelted	95	16% (15)	84% (80)	Nil
AUS	366	8.4% (31)	76.2% (279)	15% (56)

It is important to note that 30% of the injuries in the US belted sample were attributed to an airbag contact, while this figure was 8.4% in the Australian sample. However it is equally important to bear in mind the high number of injuries in the Australian sample where the contact source was unknown. Despite this, given the comparability of the samples in terms of vehicle and occupant characteristics, and injury severity patterns as indicated by AIS levels (see Figure 1 above) and MAIS levels (see Figure 2), this is an important result. It is also of note that all of the airbag contact injuries in the Australian sample and unbelted US drivers were AIS level 1 (minor) injuries. However in the US belted sample, 6.3% of airbag attributed contact injuries were AIS 2 injuries and 4.2% were AIS 3 injuries.

Examination of injuries attributed to a non-airbag contact source reveals that a higher percentage of injuries in the Australian sample (89.6%) were AIS1 compared to the US belted (80%) and US unbelted sample (84%). The percentage of AIS 2 injuries is similar, while the greatest difference lie in the percentage of total injuries classified as AIS 3. Approximately 0.7% of injuries in the Australian sample were AIS 3, while 4.2% and 3.75% of injuries in the US belted and US unbelted sample were AIS 3 injuries.

INJURIES BY AIS BODY REGION

In this final section of this paper, we set out to examine the injury patterns by body region to complement the findings discussed above. Figures 3 and 4 show the distribution of injuries by body region for airbag and non-airbag contact injuries as a function of all injuries sustained, respectively. There are some immediate patterns that emerge in both Figures, namely the high percentage of facial and upper limb injuries in the US belted and US unbelted sample compared to the Australian sample (Figure 3). Also of note are the high percent of head and facial injuries as a result of non-airbag contacts in the US unbelted sample (Figure 4.). Injury patterns for each body region across the three samples is discussed separately below. Percentage for each sample are expressed as a function of the total injuries sustained where the contact source was known (US belted – 309 injuries, US unbelted – 95, Australian – 310 injuries).

Figure 3.

Percent airbag contact injuries by total injuries by body region.

Figure 4.

Percent of non-airbag contact injuries by total injuries by body region.

INJURIES TO THE HEAD

Figure 3 shows the differences in the pattern of injuries to the head. Of the injuries where the airbag was the contact source, approximately 0.3% of total injuries sustained by US belted occupants were to the head, while for the US unbelted sample this figure was 1%. Of note is that there was only one injury to the head for the US belted and unbelted sample, both of which were AIS level 1 injuries. There were no injuries to the head region attributed to an airbag contact in the Australian sample. Given the low numbers of airbag contact injuries to the head, it not possible to draw any conclusions from these injury results.

Figure 4 shows the number of injuries resulting from a non-airbag contact source as a function of total injuries sustained. Approximately 6.5% of the total injuries sustained by US belted occupants were to the head, 19% of total injuries in the US unbelted sample, and 3.5% of the total injuries in the Australian sample. The majority of the injuries sustained were AIS level 1. The US belted and US unbelted samples had approximately the same percentage of AIS 2 injuries (~15%) with a smaller number of AIS 3 injuries (5% & 11%). Importantly the US unbelted group had twice as many AIS 3 injuries than the US belted group, while there were no injuries above AIS 2 in the Australian sample (11; 63% AIS1, 36% AIS2).

INJURIES TO THE FACE

The pattern of injuries to the face as a result of contact with the airbag can be seen in Figure 3. Approximately 9% of the total injuries sustained by US belted occupants were to the face, 8.5% in the US unbelted sample, and 4.2% of total injuries sustained by the Australian sample. All injuries in all three samples as a result of airbag contact were AIS1.

Figure 4 shows the number of injuries resulting from a non-airbag contact source as a function of total injuries sustained. Approximately 2.9% of the total injuries sustained by US belted occupants were to the head, 17% of injuries in the US unbelted sample, and 1.6% of injuries in the Australian sample. All injuries sustained were AIS level 1. Of note was the greater number of facial injuries due to non-airbag contact sources in the US unbelted sample compared to the Australian and US belted sample.

INJURIES TO THE NECK

Figure 3 demonstrates the differences in the pattern of injuries to the neck by sample group. Approximately 0.3% of total injuries sustained by US belted and US unbelted occupants were to the head, while for the Australian sample this figure was 1%. There was only one injury to the head in each of the samples, and were all AIS level 1 injuries. As with the airbag contact to the head region, low numbers precludes any meaningful conclusions to be drawn from the airbag related neck injuries presented.

Injuries resulting from a non-airbag contact source as a function of total injuries sustained is shown in Figure 4. As a consequence of non-airbag contact, 3% of the total injuries sustained by US belted occupants were to the neck, 2% of total injuries in the US unbelted sample, and 7.7% of the total injuries in the Australian sample. All but one injury (in the Australian sample, 1 injury was AIS2) were AIS level 1 injuries. There were a greater number of neck injuries in the Australian sample compared to the belted and unbelted US samples as a result of non-airbag contact. Within the Australian sample, there were a greater number of injuries to the head where the contact source was an aspect other than the airbag. The same result was also evident in the US belted sample.

INJURIES TO THE THORAX

There were differences in the pattern of injuries to the thorax across the samples and by contact source. In the US belted group, 4.2% of total injuries were thoracic injuries were caused by airbag contact, 1% of total injuries in the unbelted US group, and none in the Australian sample (see Figure 3). There were also a larger number of thoracic injuries as a function of total injuries in the Australian sample (15.8%) compared to the US belted (9.7%), and US unbelted sample (3.2%) as a result of a non-airbag contact source (see Figure 4). There were also differences in the number of airbag contacts vs. non-airbag contact sources within the samples.

In the belted and unbelted US samples, there were a greater number of injuries to the thorax due to non-airbag contact sources than with airbag contact. All airbag contact injuries in both US samples were AIS 1 injuries. In the US belted and unbelted samples, the majority of non-airbag contact injuries were AIS 1, while there were a smaller percentage of AIS 2 injuries. Of note was that there were no thoracic injuries in the Australian sample due to an airbag contact, however there were 49 (15.8%) injuries as a result of a non-airbag contact. Within the Australian sample where the injury resulted from a non-airbag contact, 90% (44) were AIS 1, 6% (3) were AIS 2, and 3% (2) were AIS 3 injuries. Note there were no AIS 3 injuries to the thorax in the US samples.

INJURIES TO THE SPINE

No injuries to the spine region were attributed to an airbag contact in any group. Figure 4 shows the number of injuries resulting from a non-airbag contact injury compared to total injuries sustained. Injuries to the spine region comprised approximately 2.6% (8) of the total injuries in the US belted sample, 2.1% (2) in the US unbelted sample, and 1% (3 injuries, 1 AIS1 & 2 AIS2) in the Australian sample. Injuries in the US samples were AIS 1 injuries.

INJURIES TO THE ABDOMEN AND PELVIS

There were no injuries to the abdomen and pelvis in the Australian sample, and none in the US unbelted sample as a result of an airbag contact (see Figure 3). However in the US belted group, 1.6% of the total injury count was attributed to the airbag resulting in an injury to the abdomen \ pelvic region. There were differences in the pattern abdominal and pelvic injuries in all groups as a result of contacting another aspect of the vehicle as opposed to the airbag (Figure 4). There were more injuries in the Australian sample (13.2%, 41) compared to both the US belted (3.9%, 12) and US unbelted sample (2.1%, 2). Of the non-airbag contact injuries in the Australian sample, 95% were AIS 1 and 5% were AIS 2 level injuries. Of the 12 injuries in the US belted sample, 75% (9) were AIS 1 and 25% (3) were AIS 2 injuries. Injuries in the unbelted US sample were AIS 2 injuries.

INJURIES TO THE UPPER EXTREMITY

In looking at injuries where the airbag was the contact source (Figure 3), there was a greater percentage of upper extremity injuries in the US belted sample (14.23%) than in the Australian sample (5.5%). Unbelted US drivers had the lowest proportion of lower extremity injuries from an airbag contact (4.25%). All airbag contact upper extremity injuries in the Australian and US unbelted sample were AIS level one (minor) injuries. Of the injuries to the US belted sample, 77.5% were AIS 1 injuries, 13.5% were AIS 2 injuries, and 9% were AIS 3 injuries.

Where the contact source was not the airbag (Figure 4), there were a slightly greater number of injuries to the upper extremity in the US belted sample (17%) compared to the Australian sample (16.7%). In turn there are more non-airbag upper extremity contact injuries in the Australian sample compared to the US unbelted sample (14.7%). Injuries in the Australian sample were predominantly AIS1 (92%) with the remaining 7% AIS 2 injuries. Within the US belted sample, 83% of injuries were AIS1, 9.5% were AIS 2 injuries, and the balance AIS 3 injuries. The US unbelted sample comprised 86% AIS1 injuries, 7% AIS 2 and 7% AIS 3 injuries.

INJURIES TO THE LOWER EXTREMITY

There were no injuries to the lower extremity in the Australian and US unbelted sample with the airbag attributed as the contact source (see Figure 3). There were two injuries (0.6% total injuries of belted US group) in the belted US sample attributed to an airbag contact source, both if which were AIS 1 injuries.

In terms of non-airbag contact injuries, there were a greater number of lower extremity injuries in the Australian belted sample (30.3% of total injuries) compared to the unbelted US sample (23.2% of total injuries), and the belted US sample (24%). Injuries in the Australian sample were predominantly AIS1 (88%) with the remaining 12% AIS 2 injuries. Within the US belted sample, 68% of injuries were AIS1, 26% were AIS 2 injuries, and the remaining 6% AIS 3 injuries. The US unbelted sample comprised 86.4% AIS1 injuries and 13.6% AIS 2 injuries.

DISCUSSION

As the deployment threshold of the Australian airbag is higher than that in the US, we suspected there would be a higher EBS and higher ISS for the Australian drivers in this study. This was not the case. There was an overall trend toward higher ISS in the US drivers compared with Australians injured in similar airbag deployed crashes. This occurred although wheelbase (as a marker for car size and mass) was a control variable and EBS was not significantly different between the groups.

SAFETY BELT USE was significantly different between the US and Australia as expected. However, the effect of this on injury severity and airbag contact injury was unexpected. The data suggest that belt wearing in the US is associated with higher ISS even with a trend toward lower EBS. This finding may be caused by sex differences in belt wearing and the increase in injury severity for women, particularly light women. In the US, belted drivers had a significantly greater number of airbag contact injuries than their unbelted counterparts, possibly because the belt restrains the driver from sliding into the passenger seat and avoiding the airbag. Australian airbags caused a much lower percentage of overall injuries, but the number of “unknown source” injuries in the Australian dataset make these results less likely to be significant.

INCREASED CRASH SEVERITY (as higher EBS) was associated with significantly higher ISS only for the Australian sample, and neither the US belted or unbelted sample reached significance for this comparison. That the mean ISS was not significant across the EBS categories may be a consequence of the sample broken into the six EBS categories with a subsequent loss of power.

WOMEN were more frequently involved as injured drivers in airbag deployed crashes in the US sample than in the Australian sample. This is possibly the result of more women driving cars fitted with airbags in the US. Driver airbags are now (1998 and later model years) standard on all passenger cars in the US but are still optional in Australia. In Australia, airbags are more likely to be installed in larger or fleet cars, which are more likely to be driven by men.

In this study, female drivers in the US were significantly more likely to have been wearing a safety belt than male US drivers (Table 2). However, there was a trend for both belted and unbelted female drivers to have a higher ISS than males in US crashes. This finding was also true of the Australian drivers (Table 4). It is difficult to make any definitive conclusions due to the small sample size resulting in high variance in the ISS data.

LIGHT DRIVERS (defined as less than 55kg) tended to have a higher ISS in all groups (Table 6). There appears to be an increased risk for injury from any contact source for light, belted drivers (who happened to be women) in US airbag-deployed crashes, but not Australian crashes. This study suggests that the more powerful US airbag may be the cause for this difference, but as the injuries were from both airbag and non-airbag contact sources further research is warranted. However, there were too few light, unbelted drivers for our injury data to reach significance for that group, and the high degree of variance makes any definitive conclusions difficult, and inappropriate, to make.

The analysis of BODY REGION and contact source data reveal that in both Australia and the US, airbags were most likely to cause injury by contact with the driver’s face and upper extremity. In both belted groups, the upper extremity was even more likely to be contacted by the airbag than in the unbelted group. However, all the airbag contact injuries in Australia were minor (AIS 1) while among belted US drivers there were 34 AIS 1 injuries (77%), 6 AIS 2 injuries (13.5%), and 4 AIS 3 (9%) upper extremity injuries. In addition, 4.2% of belted US drivers injuries were to the thorax and were attributed to the airbag while none of the Australian drivers had a chest injury from an airbag source. Hence, the larger more powerful US airbag was seen to be associated with more thoracic injuries, facial and upper extremity injuries than the smaller Australian bag. An interesting result is the high number of thoracic, neck, and abdomen \ pelvic injuries resulting from non-airbag contacts in the Australian sample. This increased percentage of injuries to the thorax most likely reflects a proportional shift in the injury pattern as a result of the substantial reduction in head and face injuries compared to the US belted and unbelted samples.

CONCLUSION

Crash and injury patterns were compared among three groups of drivers injured in airbag deployed crashes: US belted, US unbelted, and Australian (belted) drivers. The larger, more powerful US airbag was seen to be associated with a greater number of facial, thoracic and upper extremity compared to the Australian airbag. Hence, the Australian airbag was seen to protect against injuries to the head, face, and upper extremity relative to the US airbag. Belt wearing was not associated with less severe overall injury but was associated with lower percentage of injuries caused by contact with the airbag. In both the US and Australia, lighter (under 55kg) drivers were at increased risk for all injuries and for injuries from airbag contact. Given the small sample size, strict exclusion criteria and the focus on intermediate and full size passenger cars, the findings presented here should be treated with a degree of caution. Despite this qualification, the differences in injury patterns between the US and Australian samples presented in this paper regarding intermediate and full size passenger vehicles reveal important insights in differences in airbag aggressivity and seat-belt use and resultant injury outcomes. Hence, it is important to conduct further detailed research to ensure representativeness of injury patterns of all airbag-deployed crashes.