Accident Analysis and Prevention 43 (2011) 18931900

Contents lists available at ScienceDirect

Accident Analysis and Prevention

Motorcycle protective clothing: Protection from injury or just the weather?

The George Institute for Global Health, The University of Sydney, Australia Medical School, Australian National University, Australia c Accident Research Centre, Monash Injury Research Institute, Monash University, Australia d CARRS-Q (Centre for Accident Research and Road Safety Queensland), Queensland University of Technology, Australia
b

a r t i c l e

i n f o

a b s t r a c t
Background: Apart from helmets, little is known about the effectiveness of motorcycle protective clothing in reducing injuries in crashes. The study aimed to quantify the association between usage of motorcycle clothing and injury in crashes. Methods and ndings: Cross-sectional analytic study. Crashed motorcyclists (n = 212, 71% of identied eligible cases) were recruited through hospitals and motorcycle repair services. Data was obtained through structured face-to-face interviews. The main outcome was hospitalization and motorcycle crash-related injury. Poisson regression was used to estimate relative risk (RR) and 95% condence intervals for injury adjusting for potential confounders. Results: Motorcyclists were signicantly less likely to be admitted to hospital if they crashed wearing motorcycle jackets (RR = 0.79, 95% CI: 0.690.91), pants (RR = 0.49, 95% CI: 0.250.94), or gloves (RR = 0.41, 95% CI: 0.260.66). When garments included tted body armour there was a signicantly reduced risk of injury to the upper body (RR = 0.77, 95% CI: 0.660.89), hands and wrists (RR = 0.55, 95% CI: 0.380.81), legs (RR = 0.60, 95% CI: 0.400.90), feet and ankles (RR = 0.54, 95% CI: 0.350.83). Non-motorcycle boots were also associated with a reduced risk of injury compared to shoes or joggers (RR = 0.46, 95% CI: 0.280.75). No association between use of body armour and risk of fracture injuries was detected. A substantial proportion of motorcycle designed gloves (25.7%), jackets (29.7%) and pants (28.1%) were assessed to have failed due to material damage in the crash. Conclusions: Motorcycle protective clothing is associated with reduced risk and severity of crash related injury and hospitalization, particularly when tted with body armour. The proportion of clothing items that failed under crash conditions indicates a need for improved quality control. While mandating usage of protective clothing is not recommended, consideration could be given to providing incentives for usage of protective clothing, such as tax exemptions for safety gear, health insurance premium reductions and rebates. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 16 November 2010 Received in revised form 24 April 2011 Accepted 26 April 2011 Keywords: Motorcycle Rider Protective clothing Crash Injury Accident Personal protective equipment

1. Introduction Motorcycles are the fastest growing sector of motor vehicles globally and comprise the majority of the eet in many low and middle income countries. Their presence on the roads is mirrored in the increasing proportion of motorcycle crash casualties in both low and high income countries (WHO, 2004). Compared to car drivers, motorcyclists are more likely to be killed or severely injured in crashes (DFT, 2008; Lee et al., 2010). For many years motorcycle

Corresponding author at: Injury Division, The George Institute for Global Health, PO Box M201, Missenden Road, Sydney, NSW 2050, Australia. Tel.: +61 2 9657 0361; fax: +61 2 9657 0301. E-mail addresses: [email protected], [email protected] (L. de Rome). 0001-4575/$ see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2011.04.027

safety research has been dominated by debate about the effectiveness of helmets (Lawrence et al., 2002; Liu et al., 2008). There has been less focus on other protection for the riders body, although the protective value of materials such as leather have been known for at least 30 years (Feldkamp et al., 1977; Zettas et al., 1979; Aldman et al., 1981; Hurt et al., 1981a; Schuller et al., 1982, 1986; Otte and Middelhauve, 1987; Hell and Lob, 1993). In Europe, standards have been developed for motorcycle protective clothing based on two mechanisms for protecting the motorcyclists body (EEVC, 1993). The rst requires protection of soft tissues by material and construction that is abrasion, cut, tear and burst resistant (EU, 2002). The second requires the use of body armour or impact protectors (high-density foam shields) which absorb and distribute the force of direct impacts to exposed areas, e.g. elbows (EU, 1998). There are now separate standards

1894

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900

for motorcycle protective gloves, boots, one piece suits, jackets and pants and body armour for the limbs and back. While only enforceable in Europe, the standards have provided benchmarks for manufacturers across the international market (de Rome, 2006). The result has been the emergence of a new generation of protective clothing products, however to date their performance in real world crashes has not been examined. While there are limits to the extent clothing can prevent injury in high impact crashes, it is in low impact crashes that protective clothing is thought to offer the greatest injury reduction (Hell and Lob, 1993). The majority of motorcycle crashes do not involve high speeds nor impacts with xed objects (EEVC, 1993; Noordzij et al., 2001; ACEM, 2004). However it is apparent that many riders who wear helmets do not fully protect the rest of their bodies (Hurt et al., 1981a; Reeder et al., 1996; ACEM, 2004; de Rome and Stanford, 2006; Wishart et al., 2009). Given the increasing human and economic costs of motorcycle injuries around the world, there is a clear need for research to establish the effectiveness of motorcycle protective clothing.

2. Methods The aim of the study was to examine the association between use of motorcycle protective clothing and risk of injury in crashes. A 12 month prospective cohort study of motorcycle crashes was conducted from June 2008. Eligible participants were residents of the study area, aged 1770 years, who were riders or passengers involved in motorcycle crashes causing injury or vehicle damage, on public roads within the Australian Capital Territory (ACT). Motorcyclists were excluded if they scored <13 on the Glasgow Coma Scale (GCS), sustained severe head (3+) or spinal injuries (4+) on the Abbreviated Injury Scale (AIS), or were otherwise unable to provide informed consent (Teasdale and Jennett, 1974; AAAM, 2005). The ACT consists of an urban centre with a population of 354,900 (ABS, 2010) surrounded by a rural region. Potential participants were identied through the two hospitals and/or thirteen motorcycle crash repairers servicing the area. Repair services received a recruitment fee to obtain written consent and contact details from customers who had been involved in road crashes. Potential participants were mailed information and telephoned to invite their participation in the study. Participants were interviewed faceface approximately two weeks after their crash and surveyed by mail after six weeks and six months. The interview format was based on the OECD methodology for motorcycle crash investigation (OECD, 2001). Information about the crash, the type and speed of impact, clothing worn and injury details were collected, in addition to demographic details and information relating to their riding experience and exposure. External clothing worn and any evidence of crash impact was photographed. Participants also completed six questionnaires about their general health prior to the crash. These were repeated in the follow up surveys on the longer term consequences of the crash, the ndings of which are not reported here. Injury details were recorded by location: hands/wrists, upper body (including arms, shoulders, chest, abdomen and upper back), lower body (including hips, lower back and legs) and feet/ankles. Injuries were classied by type: soft tissue (abrasion, cuts, laceration, bruises and burns), joint damage (sprains and dislocations), fractures and internal injuries. Injury details, including location and dimensions, were recorded on a body outline diagram by the interviewer and subsequently independently scored on the AIS scale by a trained assessor. The medical records of participants who attended hospital were used to corroborate interview reports on injuries and admissions details.

The current article is a cross-sectional analysis of the baseline data. The main outcomes reported are injury by body zone and admission to hospital. Self-reported injury data obtained at interview was used for this analysis, as it was available for all participants and provided more detail on minor injuries than available in emergency department records. The main exposure was use of motorcycle protective clothing with or without body armour. Motorcycle clothing not designed for injury protection (e.g. weather protection) was not included. Sample size estimates were based on the expected exposure (protection) and outcome (injury) rates for each body zones. The expected protection rates for each body zone was based on a recent Australian survey of riders (de Rome and Brandon, 2007). The expected injury rates in unprotected riders were based on those reported by Hurt et al. (1981b). The largest sample was required for the upper body based on an expected 12% unprotected (de Rome and Brandon, 2007) and 58% of unprotected being injured (Hurt et al., 1981b). A sample of 201 was required to detect a 30% difference in the proportion of protected and unprotected riders with injuries to their upper body, based on a test for two independent proportions with a target level of = 0.05 and power of 80%. Age was categorised into identied motorcycle crash risk age groups (ATSB, 2002). The impacts causing injury to the motorcyclists body were classied into 5 types and up to 4 separate impacts were recorded per motorcyclist. Ethical approval for the study was obtained from the Human Research Ethics Committees (HREC) for ACT Health and Calvary Health Care. The study was endorsed by the main local motorcycle community organisations. Relative risk ratios were selected as the appropriate means of comparison as this was a cross-sectional study and the outcomes of interest were relatively common (incidence greater than 10%) (Zou, 2004). To avoid convergence issues encountered with the log-binomial model, the Poisson model with sandwich variance estimator was used to compute adjusted relative risks (Spiegelman and Hertsmark, 2005; Lumley et al., 2006). The analysis computed the relative risk for injury to each part of the body by level of protection. Poisson regression models were used to estimate the relative risk (RR) and corresponding 95% condence intervals for hospitalization and, separately, for injury to each body zone by injury type. Tests of association were adjusted for potential confounders of injury identied from the literature including the age and gender of the motorcyclist; type of motorcycle; type of crash (single or multi-vehicle); type of impact (e.g. road surface, other vehicle or xed object); and the estimated speed of impact (Lin et al., 2003; Zambon and Hasselberg, 2006; Pai and Saleh, 2007). As it was likely that riders who were not injured were underrepresented in the study, a sensitivity analysis was conducted. The analysis was re-run on the subset of participants who were recruited from hospital, excluding those recruited via other means. Analysis was conducted in SAS 9.1 (SAS, 2008).

3. Results Over the study period the researchers identied 618 crashed motorcyclists, including 298 who had been involved in a road crash within the ACT and met all eligible criteria for inclusion in the study. Of these 298 eligible motorcyclists 71.1% (n = 212) participated, 20% (n = 59) could not be contacted, 8% (n = 24) declined and 1% (n = 3) were excluded on medical grounds. The 212 participants included 126 (59.4%) identied from hospital presentations, 75 (35.4%) from crash repair services and 9 self-referred (4.2%). There were no signicant differences in age or sex between eligible participating and non-participating riders identied through hospitals.

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900 Table 1 Key features of participants, clothing worn and the crash event. Factor n (%) Gender Age group Motorcycle licence Type of motorcycle Hospitalization MAIS (Most serious injury) Motorcycle clothing Worn (n) Impact sustained Garment damaged Clothing failure Materiala Seams/fastenings Impact protection Worn (n) Impact sustained Crash type Type of impact Impact speedb Sustained direct impact to body Injuries due toc
a

1895

Male 184 (86.8) 1725 61 (29.3) None 5 (2.4) Standard/Scooter 36 (17.0) No medical treatment 39 (18.4) 0 3 (1.4) Helmet 209 (98.6) 126 (60.3) 139 (66.5) NA NA Upper limb 153 (72.2) 103 (67.3) Single vehicle 89 (42.0) Loss of control 100 (47.2) 020 56 (26.4) Yes 162 (76.4) Road surface 178 (84.0)

Female 28 (13.2) 2639 90 (42.5) Learner 54 (25.5) Touring/Cruiser 60 (28.3) Medical treatment only 124 (58.5) 1 156 (73.6) Jacket 175 (82.6) 142 (81.1) 115 (65.7) 52 (29.7) 18 (10.3) Lower limb 21 (9.9) 10 (47.6) Multi vehicle 104 (49.1) Collision (object/vehicle) 112 (52.8) 2140 65 (30.7) No 50 (23.6) Slid on road 139 (65.6)

4075 60 (28.3) Provisional 25 (11.8) Supersport/Sport/ 106 (50.0) Admitted to hospital 49 (23.1) 2 36 (17.0) Pants 74 (34.9) 55 (74.3) 42 (56.8) 52 (28.1) 2 (2.7) Hand/wrist 109 (51.4) 120 (64.9) Animal on road 19 (9.0)

Full 128 (60.4) Off-road 10 (4.7)

3 12 (5.7) Gloves 185 (87.3) 120 (64.9) 101 (54.6) 19 (25.7) 14 (7.6) Foot/ankle 66 (31.1) 54 (66.7)

4 3 (1.4) Boots 81 (38.2) 54 (66.7) 56 (69.1) 7 (8.6) 5 (6.2) Back/spine 40 (18.9) 22 (55.0)

5 1 (0.5)

6 1 (0.5)

4160 39 (18.4)

61120 35 (16.5)

Unknown 17 (8.0)

Own motorcycle 70 (33.0)

Other vehicle 350 (16.5)

Fixed object 31 (14.6)

Material failure is dened by holes in the protective layer. Speed of impact is based on motorcyclist estimations. c Percentages refer to the proportion of motorcyclists who sustained each type of impact and are not additive. Up to four different impacts causing injury were recorded for each motorcyclist.
b

The 320 motorcyclists classied as ineligible either lived (n = 66, 21%) or crashed outside the study area (n = 63, 20%), crashed offroad (n = 159, 50%), or were misclassied or double counted by crash repairer and hospital (n = 32, 9%). Study participants included 49 (23.1%) admitted to hospital, 124 (58.8%) who required some medical treatment and 39 (18.4%) who did not seek medical treatment. The majority (73.6%, n = 156) sustained minor injuries (AIS 1) with only 1% (n = 3) uninjured (AAAM, 2005). Almost half of the crashes (49%) involved another vehicle, 42% were single vehicle and 9% involved an animal, usually a kangaroo on the road. Over the same period, 283 motorcycle road crashes were reported to ACT police, these included 100 injury crashes and 183 involving property damage only (ACT Roads, 2010). Table 1 summarises key demographic and crash characteristics of the sample. Nearly all participants wore helmets (98.6%), motorcycle jackets (82.5%) and motorcycle gloves (87.3%). Fewer wore motorcycle pants (34.9%), motorcycle boots (38.2%) or other heavy boots (25.9%). Body armour was worn over shoulders and elbows (71.7%), hands (50.9%), feet/ankles (29.7%) backs (18.9%), knees 9.9% and hips (7.6%). Almost half (45.8%) wore foam inserts in their jacket backs. A high proportion of the motorcycle clothing showed signs of crash impact. Over a quarter of the pants (28.1%), jackets (29.7%) and gloves (25.7%) were assessed as having failed because the protective layer was holed potentially exposing the wearer to injury. The most common form of damage was due to material erosion with little evidence of tear, cut, sharp penetration or crush damage. Table 2 shows the proportions of motorcyclists by types of injury and type of protection worn. The most common injuries were cuts, abrasions and bruises followed by sprains mostly to the

upper torso. Fractures were less common and more likely to affect upper limbs. Hospital records showed a close correspondence with injury reports obtained at interview for those participants who attended hospital with 88% agreement on the number with long bone (legs and arms) fractures. There was less agreement in relation to small bone fractures with hospital records accounting for only 57% (n = 4/7) of the foot and ankle fractures and 39% (11/28) of the hand and wrist fractures that were reported at interview. Overall riders were signicantly less likely to be admitted to hospital if they crashed while wearing a motorcycle jacket (RR = 0.79, 95% CI: 0.690.91), motorcycle pants (RR = 0.49, 95% CI: 0.250.94), or motorcycle gloves (RR = 0.41, 95% CI: 0.260.66). The effect of motorcycle boots on hospitalization was not signicant (RR = 1.04, 95% CI: 0.591.83). These ratios are adjusted for age, gender, motorcycle type, single or multivehicle crash, impact type and the riders estimate of impact speed. Table 3 presents the unadjusted and adjusted relative risk (RR) for motorcyclist injuries associated with motorcycle clothing with and without body armour, compared to non-motorcycle clothing. The RR represents the benet, if any, of the particular item of protective clothing in reducing the risk of each type of injury to the relevant part of the body. Motorcyclists wearing motorcycle protective clothing tted with body armour, were signicantly less likely to sustain injuries to the protected areas compared to those wearing non-motorcycle clothing. Specically, when body armour was tted, there was a 23% lower risk of injury associated with motorcycle jackets (RR = 0.77, 95% CI: 0.680.86), 45% for motorcycle gloves (RR = 55, 95% CI: 0.370.81), 39% for motorcycle pants for leg injuries only (RR = 0.61, 95% CI: 0.410.91 and 45% by motorcycle boots (RR = 0.55, 95% CI:

1896

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900

Table 2 Proportion of different types of injuries by type of protection worn. Motorcycle clothing worn Motorcycle jacket No Yes Yes with body armour Motorcycle gloves No Yes Yes with body armour Motorcycle pants No Yes Yes with body armour Motorcycle boots No (shoes/joggers) No (other boots) Yes Yes with body armour Helmet No Open face Full face Motorcycle back protector No Foam insert Back armour Total n = 212 37 23 152 27 77 108 138 53 21 76 57 17 62 3 26 183 75 97 40 Any injury (%) 91.9% 78.3% 69.7% 66.7% 49.4% 35.2% 92.0% 84.9% 71.4% 55.3% 26.3% 35.3% 32.3% 60.0% 19.2% 15.5% 10.7% 21.6% 7.5% Bruised (%) 43.2 39.1 31.6 25.9 26.0 12.0 61.6 64.2 61.9 28.9 14.0 23.5 16.1 0.0 11.5 9.3 5.3 3.1 2.5 Abrasions/cuts (%) 56.8 26.1 25.7 55.6 18.2 14.8 72.5 49.1 33.3 30.3 8.8 5.9 4.8 33.3 15.4 2.7 2.7 3.1 2.5 Fracture (%) 21.6 13.0 9.9 7.4 13.0 8.3 5.1 1.9 4.8 9.2 3.5 5.9 6.5 33.3 3.8 0.0 1.3 2.1 2.5 Sprain (%) 29.7 52.2 35.5 18.5 18.2 11.1 19.6 18.9 4.8 15.8 14.0 17.6 14.5 33.3 11.5 8.7 6.7 16.5 0.0 Internal (%) 0.0 0.0 0.7 0.0 0.0 0.0 2.9 1.9 9.5 1.3 0.0 0.0 4.8 66.7 19.2 15.8 0.0 0.0 0.0

0.350.85). The risk of any foot or ankle injuries was reduced 53% by non-motorcycle boots (RR = 0.47, 95% CI: 0.280.77) when compared to shoes or joggers, a risk reduction similar to motorcycle boots. The results for motorcycle jackets and gloves were conrmed in the sensitivity analysis when non-hospital recruits were excluded,

however the analysis for motorcycle pants and boots failed to converge, most likely due to insufcient numbers. There was an increased risk of back injury associated with usage of foam inserts in the backs of jackets (RR = 2.16, 95% CI: 1.084.36). Table 4 presents the adjusted relative risk (RR) for types of injury according to the protection worn.

Table 3 Unadjusted and adjusted relative risk for any injuries sustained by level of protection. Type of clothing Injuries sustained (n) No Upper body No motorcycle jacket Motorcycle jacket Motorcycle jacket + BAb Hand/wrist injuries No motorcycle gloves Motorcycle gloves Motorcycle gloves + BAb Lower body injuries No motorcycle pants Motorcycle pants Motorcycle pants + BAb Leg injuries only No motorcycle pants Motorcycle pants Motorcycle pants + BAb Feet/ankles Shoes/joggers Non-motorcycle boots Motorcycle boots Motorcycle boots + BAb Back/Spine No back protection Foam insert in jacket Separate back armour 3 5 46 9 39 70 11 8 6 21 13 10 34 42 11 42 67 76 37 Yes 34 18 106 18 38 38 127 45 15 117 40 11 42 15 6 20 8 21 3 Reference 0.85NS 0.76e Reference 0.73NS 0.53e Reference 0.92NS 0.77NS Reference 0.88NS 0.61c Reference 0.47d 0.63NS 0.58d Reference 2.00NS 0.69NS Reference 0.83NS 0.77f Reference 0.69NS 0.55d Reference 0.93NS 0.79NS Reference 0.89NS 0.61d Reference 0.47d 0.56NS 0.55d Reference 2.16c 0.77NS Unadjusted RR 95% Condence interval Adjusted RRa 95% Condence interval

0.671.08 0.660.87

0.651.05 0.680.86

0.511.04 0.360.76

0.461.04 0.370.81

0.811.04 0.591.01

0.821.05 0.611.02

0.751.05 0.410.93

0.751.06 0.410.91

0.290.76 0.321.24 0.380.87

0.280.77 0.271.17 0.350.85

0.944.27 0.192.47

1.084.36 0.222.64

NS = non-signicant. a Adjusted for age, gender, motorcycle type, crash type (single or multivehicle), impact type (slid, object, other road user) and impact speed. b BA garment had body armour incorporated. c p 0.05. d p 0.01. e p 0.001. f p < 0.0001

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900 Table 4 Adjusted relative risk of soft tissue, open wound and fracture injuries by level of protection. Injuries sustained Area/type of clothing Upper bodyb No motorcycle jacket Motorcycle jacket Motorcycle jacket + BA Hand/wrist injuries No motorcycle glove Motorcycle gloves Motorcycle gloves + BA Lower body injuriese No motorcycle pants Motorcycle pants Motorcycle pants + BA Leg injuries No motorcycle pants Motorcycle pants Motorcycle pants + BA Feet/ankles Shoes/joggers Non-motorcycle boots Motorcycle boots Motorcycle boots + BA Back/Spine No back protection Foam insert in jacket Separate back armour All soft tissuec Adj. RR
a

1897

Open woundsd 95% CI Adj. RR

Fractures 95% CI Adj. RRa Reference 0.59NS 0.48NS Reference 1.42NS 1.24NS Reference 0.51NS 1.15NS 95% CI

Reference 0.73NS 0.67h Reference 0.60f 0.38i Reference 0.92NS 0.66f Reference 0.87NS 0.53f Reference 0.39g 0.35f 0.31i Reference 0.99NS 0.96NS

0.491.07 0.530.84

Reference 0.42f 0.37i Reference 0.30h 0.27i Reference 0.62g 0.41f Reference 0.63f 0.09g Reference 0.24g 0.17NS 0.10h

0.171.01 0.250.55

0.162.16 0.221.06

0.380.96 0.240.60

0.150.59 0.150.49

0.297.08 0.265.90

0.781.08 0.440.97

0.420.90 0.200.85

0.122.23 0.393.42

0.681.09 0.310.90

0.420.95 0.010.60

NA

0.220.70 0.130.91 0.170.55

0.100.58 0.021.47 0.030.34

Reference 0.27NS 0.48NS 0.43NS

0.061.32 0.073.58 0.121.51

0.283.50 0.204.65

NA

NA

BA Garment had impact protection incorporated. NA data not available due to small numbers and convergence issues. a Adjusted for age, gender, motorcycle type, crash type (single or multivehicle), impact type and impact speed. b Upper body fractures to arms, elbows or shoulders only. c All soft tissue injuries include bruises, abrasions, cuts and lacerations. d Open wounds include abrasions, cuts and lacerations but excludes bruises. e Lower body fractures adjusted for age gender only due to convergence issues. f p 0.05. g p 0.01. h p 0.001. i p < .0001

There was a signicant reduction in the risk of any soft tissue injuries (including bruises, abrasions, cuts and lacerations) associated with all forms of motorcycle clothing tted with body armour. Where body armour was not tted, motorcycle gloves (RR = 0.60, 95% CI: 0.380.96); and motorcycle boots (RR = 0.35, 95% CI: 0.130.91) still provided a reduced risk of soft tissue injuries as did non-motorcycle boots (RR = 0.39, 95% CI: 0.220.70), however no signicant reduction was observed for motorcycle jackets or pants. When bruises were excluded from the list of soft tissue injuries, there was a signicant reduction in the risk of open wounds (abrasions, cuts and lacerations) associated with all forms of motorcycle clothing tted with body armour, and for gloves and pants when body armour was not present (see Table 4). However there was no evidence of a reduction in the risk of fractures associated with body armour for any area of the body. 4. Discussion These ndings conrm earlier reports on the value of abrasion resistant materials in the reduction of soft tissue injuries in motorcycle crashes (Aldman et al., 1981; Hurt et al., 1981a; Schuller et al., 1982; Otte et al., 2002b). This study takes that work further and is the rst detailed examination of the performance of motorcycle clothing in crashes since standards for protective clothing were established in Europe (CEC, 1989). These results suggest considerable potential to reduce the human costs of non-fatal motorcycle crash injuries. In Australia motorcyclist account for 22% (n = 6270) of all seriously injured road casualties each year, with an estimated human costs for a hospi-

talized injury of approximately AUS$214 000 per injury (including disability-related costs) (AIHW, 2009; BITRE, 2009). By comparison in the US, the total lifetime costs for all non-fatally injured motorcyclists in 2005 are estimated to be US$3992 million for hospitalized injuries and US$1046 million for ED treated only (Naumann et al., 2010). The reduced risk of hospitalization observed in this study suggests that motorcycle clothing can signicantly reduce the severity of injures in crashes. While the greatest benets observed were in relation to the prevention of soft tissue, and particularly open wound, injuries, this is not a trivial outcome. Such injuries are rarely life threatening, but can have serious consequences for the motorcyclist such as opportunistic infections, scarring, loss of mobility and longer term disability. A New Zealand study of disabled motorcyclists found a high proportion (80%) had impairments due to disguring and scarring from soft tissue injuries (Clarke and Langley, 1995). The most important result relates to the contribution of body armour, which was associated with substantial reductions in the risk of any injury in crashes when other factors such as speed and type of impact were controlled. This is the rst evidence of the effectiveness of body armour from crash studies, although it has previously shown promise in laboratory tests (Otte and Middelhauve, 1987; Otte et al., 2002a). The reduced risk of injuries to the legs in motorcycle crashes is particularly important because legs are most likely to be injured but least likely to be protected (Hurt et al., 1981b; Lateef, 2002; ACEM, 2004; Chen, 2006; de Rome, 2006; Watson et al., 2008). While motorcycle boots were not associated with a signicant reduction in risk of hospitalization, the results did conrm the ben-

1898

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900

ets of motorcycle boots and, in fact, any type of sturdy boots compared to shoes such as joggers, It would appear that the basic elements of protection are not unique to motorcycle boots, but can be provided by other boots. Whether this is due to the additional coverage for the ankles, or because shoes are more likely to be torn off during a crash is unclear and requires further investigation. Motorcycle gloves and pants not tted with body armour were also associated with a reduced risk of open wound injuries. The absence of effect for motorcycle jackets (n = 23) and motorcycle boots (n = 17) was unexpected but may be due to a lack of power given the small numbers not tted with body armour. It may also be due to a lack of quality in some products, noting the material erosion of 29% of jackets or to a lack of differentiation from nonmotorcycle garments which did include some heavy jackets. While material erosion also occurred in jackets that were tted with body armour, the armour is likely to have provided an additional shield from cuts and abrasions quite apart from its primary function of impact protection. Small numbers may also explain the lack of effect for back armour. However, other research suggests that most motorcyclecrash back injuries are caused by bending and torsional forces, not direct impacts to the spine (EU, 2003). The back sprain injuries in those wearing foam inserts may be due to such bending and twisting forces, but it is hard to explain why that group should be more at risk than motorcyclist without any protection. This certainly warrants further investigation in laboratory-based studies and future in-depth research. Despite the reduced risk of any injuries when wearing body armour, the benets could not be detected specically in relation to fractures. Given the relatively low occurrence of fractures (15%), compared to soft tissue injuries (71%), in unprotected motorcyclists (Duffy and Blair, 1991) the sample size was likely too small to be able to detect any such difference. Further research is necessary to explore the benets of products such as body armour for different parts of the body in crashes in relation to specic injury types, particularly fractures. The injury reduction effects observed here gives rise to the question of mandatory use of protective clothing in addition to helmets. However, mandating use of protective equipment is unlikely to be either feasible or effective, given known ergonomic issues, the lack of global standards and the lack of quality control in motorcycle protective clothing as evidenced by the failure rates in this study. Conict between primary safety (accident avoidance) and secondary safety (injury protection) is associated with protective clothing in many industries. This is because the materials required to provide injury protection tend to be heavy and may negatively affecting the operators ability to perform safely (Nunneley and Myhre, 1976; Bittel et al., 1992; EEVC, 1993; Koch, 1996; James, 2002). Earlier reports on motorcycle protective clothing cautioned that such materials may increase riders crash risk due to discomfort and heat fatigue, however to date there does not appear to have been any research into this (EEVC, 1993; Koch, 1996). The challenge for industry has been to provide protection from injury and the weather without restricting the motorcyclists ease of movement nor creating discomfort or fatigue. The European Standards for motorcycle protective clothing require that garments withstand the forces of a crash within set limits (EU, 2002). While the forces involved in some crashes will exceed these limits, the clothing failure rate found in this study suggests a need for improved quality control. Particularly as over half of the crashes (57.1%) involved estimated impact speeds less than 40 kmh with only 16.5% over 60 kmh. The failure rate is also consistent with reports from independent tests of motorcycle clothing conducted in the United Kingdom (Ride, 2009a,b). As the market for motorcycle clothing involves a diverse international industry, such consumer driven information systems may more viable and more effective

than regulation in the short-to-medium term. They could also provide manufacturers with the incentives and market certainty to improve product quality and discourage the production of inferior products. Preliminary work has recommended using the European Standards as the benchmark for independent evaluations in Australia (de Rome, 2005; Haworth et al., 2007). The strength of this study was the attempt to obtain a representative sample by the inclusion of both injury and non-injury motorcycle crashes. Previous studies have focused on injury and/or police reported crashes thus biasing their sample towards more severe crashes and potentially excluding those where protective clothing had proved effective. Helmet usage by participants was high a nding indicative of high levels of compliance with mandatory helmet laws in Australia. Other motorcycle protective clothing is not required by law, but usage rates in the study were consistent with those reported in an observational study of riders in the same region which included both commuter and recreational riding routes (Watson et al., 2008). The use of face to face interviews to obtain injury details allowed for comprehensive itemization of soft tissue injuries, which are less likely to be documented in Emergency Department records. A limitation of the study was the reliance on self-report without independent investigation of the crash scenes or vehicles. Factors such as impact speed were therefore uncorroborated; however the distribution of the estimated impact speed reported by participants (Table 1) is consistent with that reported in studies where the crash speed was estimated objectively (ACEM, 2004). A further limitation is that the participation rate for crashes where the rider did not attend a hospital cannot be determined. The number of referrals provided by each of the nine crash repair services, which actively participated, was consistent with pre-study estimates of their turnover. The number of possible referrals not referred by the four non-participating companies is not known, although some of their customers were recruited through hospital presentations. While the number of injury crashes included in the study is greater than that recorded by police, the number of non-injury crashes is substantially less. The consequences of less serious crashes being under represented in the sample, means the size of the reported benets may be underestimated. However, a sensitivity analysis which included only those who were admitted to hospital showed similar results to the main results, so any effect is likely to be minimal. In addition, there were no signicant differences in the age or sex of eligible participating and non-participating riders identied through hospitals. Finally as a cross-sectional study, the design is not ideal to evaluate the effectiveness of interventions, so these results require conrmation by other studies.

5. Conclusions This study demonstrates that motorcycle protective clothing is associated with a signicantly reduced risk of injury in crashes, particularly when body armour is tted. While the most substantial effect was observed for open wound injuries, crashed motorcyclists who were wearing motorcycle clothing were also signicantly less likely to require admission to a hospital. These ndings have implications for policy decisions related to encouraging the use of motorcycle protective clothing, however mandating use is not recommended at this stage. The failure rate for jackets, pants and gloves indicates a need to establish systems to ensure such products are t for the purpose. Consideration could be given to providing incentives for the use of recommended gear, such as tax waivers for safety gear, health insurance premium reductions and rebates. Future research would be well served to examine the association, if any, between body armour for different parts of the body and

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900

1899

injury, particularly fractures, however the ndings here suggest a large-scale study would be required. Finally, it is also recommended that the association between motorcyclists crash risk and usage of protective clothing in hot weather be investigated.

Acknowledgements This study was funded by Swann Insurance Australia and supported by The Canberra Hospital and Calvary Health Care. Liz de Rome is supported by an NRMA ACT Road Safety Trust Scholarship. Rebecca Ivers and Wei Du receive fellowship support from the National Health and Medical Research Council of Australia. Narelle Haworths position is supported by the Motor Accident Insurance Commission of the Queensland State Government. We acknowledge and thank research assistants: Judy Perry who managed recruitment and conducted the majority of interviews, Thomas Brandon who developed the data management program, and David Tesolin who provided independent scoring of participants injury severity. We also thank the staff at Canberra and Calvary Hospitals and all the motorcycle crash repairers who supported the study. We would particularly like to thank the motorcyclists who took part in the study.

References
AAAM, 2005. In: Gennarelli, T.A., Wodzin, E. (Eds.), Abbreviated Injury Scale 2005. Association for the Advancement of Automotive Medicine, Barrington, IL, USA. ABS, 2010. 3101.0 - Australian demographic statistics, December 2009. Australian Bureau of Statistics, Canberra. ACEM, 2004. MAIDS in-depth investigation of accidents involving powered two wheelers: Final report 1.2. Association of European Motorcycle Manufacturers (ACEM) http://www.maids-study.eu/Last accessed 11 October 2010, Brussels, pp. 173. ACT Roads, 2010. Accidents involving a motorcycle: June 20082009, (unpublished data). ACT Roads, Canberra, Australia. AIHW, 2009. Serious injury due to land transport accidents, Australia 200607. Injury research and statistics series No. 53. Australian Institute of Health and Welfare, Canberra. Aldman, B., Cacciola, I., Gustafsson, H., Nygren, A., Wersall, J., 1981. The protective effect of different kinds of protective clothing worn by motorcyclists. In: Proceedings of the 6th International Conference of the Research Council on the Biomechanics of Impact, IRCOBI , Salon-de-Provence (France), pp. 19. ATSB, 2002. Motorcycle rider age and risk of fatal injury. Motorcycle Safety Monograph 12, Australian Transport Safety Bureau, http://www.infrastructure. gov.au/roads/safety/publications/2002/Mcycle Age 1.aspx Canberra. BITRE, 2009. Cost of road crashes in Australia 2006. Report 118, http:// www.Bitre.Gov.Au/publications/48/les/cost of road crashes in Australia.Pdf, Bureau of Infrastructure Transport and Regional Economics, Canberra, Australia. Bittel, J., Hanniquet, A.M., Forssard, H., 1992. Thermal constraints related to the wearing of protective clothing: Body ventilation by fresh air. In: McBriarty, J.P., Henry, N.W. (Eds.), Performance of Protective Clothing. American Society for Testing and Materials, Philadelphia, pp. 597603, fourth volume, stp 1133. CEC, 1989. The Council of the European Communities directive on the approximation of the laws of the member states relating to personal protective equipment 89/686/EEC. Ofce for Ofcial Publications of the European Communities, 1989L0686-08/10/1996 CONSLEG. Chen, H.B., 2006. Injury patterns and risk factors of motorcycle crashes. Chin. J. Clin. Rehabil. 10 (40), 187190. Clarke, J.A., Langley, J.D., 1995. Disablement resulting from motorcycle crashes. Disabil. Rehabil. 17 (7), 377385. de Rome, L., 2005. A product safety assurance system for motorcycle protective clothing. In: Proceedings of the Australasian Road Safety Research Policing Education Conference, 2005 , Wellington, New Zealand, p. 10, Last accessed 5 October 2010 http://www.rsconference.com/roadsafety/detail/522. de Rome, L., 2006. The injury reduction benets of motorcycle protective clothing. NTSB Public Forum on Motorcycle Safety, September 1213 2006. National Transport Safety Board http://www.ntsb.gov/events/symp motorcycle safety/symp motorcycle safety.htm, Washington DC. de Rome, L., Brandon, T., 2007. Survey of motorcyclists in NSW, 2006 Motorcycle Council of NSW, Inc. www.roadsafety.mccofnsw.org.au Last accessed 5 October 2010, Sydney. de Rome, L., Stanford, G., 2006. Motorcycle protective clothing: Fashion or function. The 2006 International Motorcycle Safety Conference http://www.msfusa.org/imsc/index.html, Motorcycle Safety Foundation, Long Beach. DFT, 2008. Road Casualties in Great Britain: 2007 Annual Report. Transport Statistics. Department for Transport, UK, London.

Duffy, E.M., Blair, A., 1991. Motorcycle accidents: the leg injury problem in perspective. In: Proceedings of the International Conference of the Research Council on the Biomechanics of Impact, IRCOBI , Berlin, pp. 231246. EEVC, 1993. Report on motorcycle safety. An ESV Paper, Report of the Ad-hoc Group on Motorcycle Safety, 1993. European Experimental Vehicles Committee Brussels, pp. 75. EU, 1998. Motorcyclists Protective Clothing Against Mechanical Impact: Requirements and Test Methods for Impact Protectors. European Committee for Standardization, Brussels. EU, 2002. Protective Clothing for Professional Motorcycle Riders: Jackets, Trousers and One Piece or Divided Suits General Requirements. European Committee for Standardization, Brussels. EU, 2003. Motorcyclists Protective Clothing Against Mechanical Impact: Motorcyclists Back Protectors Requirements and Test Methods. European Committee for Standardization, Brussels. Feldkamp, G., Prall, W.D., Buehler, G., Junghanns, K., 1977. Motocycle accidents; epidemiology, clinical aspects and protective measures, a followup and prospective study. Unfallheilkunde 80 (1), 119. Haworth, N., de Rome, L., Varnsverry, P., Rowden, P., 2007. Motorcycle protective clothing: are stars better than standards? In: Australasian Road Safety Research, Policing and Education Conference , Melbourne, 2007, Last accessed 5 October 2010, Melbourne http://www.rsconference.com/roadsafety/detail/820. Hell, W., Lob, G., 1993. Typical injury patterns of motorcyclists in different crash types - effectiveness & improvement of countermeasures. In: Proceedings of the 37th Annual Conference of the American Association for the Advancement of Automotive Medicine, AAAM , San Antonio, pp. 7786. Hurt, H.H., Ouellet, J., Wagar, I., 1981a. Effectiveness of motorcycle safety helmets and protective clothing. In: Proceedings of the 25th Annual Conference of the American Association for the Advancement of Automotive Medicine, AAAM , San Francisco, pp. 223235. Hurt, H.H., Ouellet, J.V., Thom, D.R., 1981b. Motorcycle accident cause factors and identication of countermeasures. University of Southern California Trafc Safety Centre Technical Report to the National Highway Trafc Safety Administration, Washington, DC. James, P.Z., 2002. The hidden hazard of protective apparel. Occup. Health Saf. 71 (1), 55. Koch, H., 1996. Improvement of motorcycle riders secondary safety by protectors tted to riders clothing. In: Proceedings of the 15th International Technical Conference the Enhanced Safety of Vehicles , Melbourne, pp. 1160 1166. Lateef, F., 2002. Riding motorcycles: is it a lower limb hazard? Singapore Med J 43 (11), 566569. Lawrence, B.A., Max, W., Miller, T.R., 2002. Costs of Injuries Resulting from Motorcycle Crashes: A Literature Review. National Highway Trafc Safety Administration, Washington DC (accessed June 11, 2010) http://www.nhtsa.dot.gov/ PEOPLE/INJURY/pedbimot/motorcycle/Motorcycle HTML/toc.html. Lee, H.Y., Chen, Y.H., Chiu, W.T., Hwang, J.S., Wang, J.D., 2010. Quality-adjusted lifeyears and helmet use among motorcyclists sustaining head injuries. Am. J. Public Health 100 (1), 165170. Lin, M.R., Chang, S.H., Huang, W., Hwang, H.F., Pai, L., 2003. Factors associated with severity of motorcycle injuries among young adult riders. Ann. Emerg. Med. 41 (6), 783791. Liu, B., Ivers, R., Norton, R., Boufous, S., Blows, S., Lo, S., 2008. Helmets for preventing injury in motorcycle riders (review update). Cockrane Database of Systematic Reviews Issue 1, pp. 37. Lumley, T., Kronmal, R., Ma, S., 2006. Relative risk regression in medical research: models, contrasts, estimators and algorithms. In: UW Biostatistics Working Paper Series. University of Washington. Naumann, R.B., Dellinger, A.M., Zaloshnja, E., Lawrence, B.A., M.T, R., 2010. Incidence and total lifetime costs of motor vehicle-related fatal and nonfatal injury by road user type, united states, 2005. Trafc Inj. Prev. 11 (4), 353360. Noordzij, P., Forke, E., Brendicke, R., Chinn, B., 2001. Integration of needs of moped and motorcycle riders into safety measures. Review and statistical analysis in the framework of the European research project PROMISING, Workpackage 3, Report D-2001 5. SWOV Institute for Road Safety Research, Leidschendam, The Netherlands, pp. 212. Nunneley, S.A., Myhre, L.G., 1976. Physiological effects of solar heat load in a ghter cockpit. Aviat. Space Environ. Med. 47 (9), 969973. OECD, 2001. Motorcycles: common international methodology for on-scene, indepth accident investigation. In: Development, O.F.E.C.A. (Ed.), Technical Expert Group of the Coordinating Group for Motorcycle Accident Investigations. Road Transport Research Program of the Directorate for Science, Technology and Industry, OECD, Paris. Otte, D., Middelhauve, V., 1987. Quantication of protective effects of special synthetic protectors in clothing for motorcyclists. In: Cesari, D., Charpenne, A. (Eds.), Proceedings of the International Conference of the Research Council on the Biomechanics of Impact, IRCOBI. Birmingham, UK, pp. 118. Otte, D., Schroeder, G., Richter, M., 2002a. Possibilities for load reductions using garment leg protectors for motorcyclists a technical, medical and biomechanical approach. In: Proceedings of the 46th Annual Conference of the American Association for the Advancement of Automotive Medicine, AAAM , pp. 367 385. Otte, D., Schroeder, G., Richter, M., 2002b. Possibilities for load reductions using garment leg protectors for motorcyclists a technical, medical and biomechanical approach. In: Annual Proceedings of the Associaton for the Advancement of Automotive Medicine, vol. 46 , pp. 367385.

1900

L. de Rome et al. / Accident Analysis and Prevention 43 (2011) 18931900 Teasdale, G., Jennett, B., 1974. Assessment of coma and impaired consciousness: a practical scale. Lancet 2 (42), 8184. Watson, B., Wishart, D., Christie, T., 2008. Canberra motorcycle apparel observation study. NRMA ACT Road Safety Trust, http://www.roadsafetytrust.org.au/ c/rtt?a=da&did=1004593. Last accessed 5 October 2010, Canberra. WHO, 2004. In: Peden, M., Scureld, R., Sleet, D., Mohan, D., Hyder, A., Jarawan, E., Mathers, C. (Eds.), World Report on Road Trafc Injury Prevention. World Health Organisation, Geneva. Wishart, D., Watson, B., Rowden, P., 2009. Motorcycle rider protective apparel wearing: observational study results from the Brisbane and Canberra regions. J. Australasian College Road Saf 20 (4), 5259. Zambon, F., Hasselberg, M., 2006. Factors affecting the severity of injuries among young motorcyclists a Swedish Nationwide Cohort Study. Trafc Inj. Prev. 7 (2), 143149. Zettas, J., Zettas, P., Thanasophon, B., 1979. Injury patterns in motorcycle accidents. J. Trauma 19 (11), 833836. Zou, G., 2004. A modied poisson regression approach to prospective studies with binary data. Am. J. Epidemiol. 159 (7), 702706.

Pai, C.W., Saleh, W., 2007. Exploring motorcyclist injury severity resulting from various crash congurations at t-junctions in the United Kingdom an application of the ordered probit models. Trafc Inj. Prev. 8 (1), 6268. Reeder, A.I., Chalmers, D.J., Langley, J.D., 1996. The risky and protective motorcycling opinions and behaviours of young on-road motorcyclists in New Zealand. Soc. Sci. Med. 42 (9), 12971311. Ride, 2009a. Textile suits. Ride, http://www.motorcyclenews.com/MCN/Products/ productsresults/Last accessed 24 April 2011. Bauer Media, UK. Ride, 2009b. Winter gloves. Ride, http://www.motorcyclenews.com/MCN/Products/ productsresults/Last accessed 24 April 2011. Bauer Media, UK. SAS, 2008. 9.1. SAS Institute Inc., Cary, NC, USA. Schuller, E., Beier, G., Spann, W., 1982. Effectiveness of protective clothing in Munich area motorcycle accidents. In: Proceedings of the 26th Stapp car crash conference , Warrendale, PA, Ann Arbor, pp. 259267. Schuller, E., Beir, G., Spann, W., 1986. Disability and impairment of protected and unprotected motorcycle riders. In: SAE International Congress and Exposition Crash Injury Impairment and Disability: Long Term Effects , Warrendale, PA, Detroit, MI, pp. 5156. Spiegelman, D., Hertsmark, E., 2005. Easy SAS calculation for risk or prevalence ratios and differences. Am. J. Epidemiol. 162 (3), 199200.