THE PROTECTIVE PERFORMANCE OF BICYCLE HELMETS INTRODUCED AT THE SAME TIME AS THE BICYCLE HELMET WEARING LAW IN VICTORIA

Transcription

1 THE PROTECTIVE OF BICYCLE S INTRODUCED AT THE SAME TIME AS THE BICYCLE WEARING LAW IN VICTORIA by MaxCameron Caroline Finch Peter Vulcan Accident Monash University Research Centre July 1994 Report No. 59

2 MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE REPORT DOCUMENTATION PAGE Report No. Report Date ISBN Pages 59 July Title and sub-title: The Protective Performance of Bicycle Helmets Introduced at the Same Time as the Bicycle Helmet Wearing Law in Victoria Author(s) Type of Report & Period Covered Cameron, M.H. Finch, C.F. Vulcan, A.P. General, Sponsoring Organisation: Roads Corporation (Vic Roads) Abstract: This project aimed to examine any changes in helmet performance due to the amendment of the Australian Standard for bicycle helmets, which was made at essentially the same time as the introduction of the bicycle helmet wearing law in Victoria on I July There was concern that the deletion of the penetration test from the Standard may have resulted in reduced protection to the heads of cyclists involved in crashes. Forty helmets sustaining impacts in crashes were collected from cyclists who were killed or treated at selected Melbourne hospitals during These helmets were predominantly "foam-only" (a foam helmet often with a material cover), "micro-shell" (a foam helmet with a thin plastic shell), or light weight "hard-shell" (a foam helmet with a hard plastic shell) allowed under the amended Standard. The new helmets were tested, and information on the bicyclists' injuries obtained, so that comparison could be made with similar information previously obtained for older-design, heavier hard-shell helmets. It was concluded that the new helmets transmit a lower level of peak acceleration to the cyclist's head inside the helmet, for a given severity of impact on the external surface of the helmet. There was no evidence of a real difference in protective performance between the older and new helmets so far as actual head injury risks are concerned. This may have been due to the absence of a difference or due to the relatively small number of helmets considered in the two helmet groups. Key Words: (IRRD except when marked*) bicycle, crash helmet, cyclist, evaluation (assessment), injury, statistics, safety, collision, head. Disclaimer: This report is disseminated in the interests of information exchange. The views expressed are those of the authors and not necessarily those of Monash University. Reproduction of this page is authorised.

3 EXECUTIVE SUMMARY This project was commissioned by VicRoads to examine any changes in helmet performance due to the amendment of the Australian Standard for bicycle helmets, which was made at essentially the same time as the introduction of the bicycle helmet wearing law in Victoria on 1 July There was concern that the deletion of the penetration test from the Standard may have resulted in reduced protection to the heads of cyclists involved in crashes. The Royal Australasian sustained College of Surgeons (RACS) had collected 64 helmets which had an impact in a crash which resulted in the helmet wearer being admitted to or treated at hospital during These helmets were predominantly heavy weight hardshell type. The helmets were submitted to the testing laboratory of Technisearch Limited, who simulated the principal helmet damage by impact tests on new helmets of the same make, model and size. The test results included the drop height (a measure of the impact severity) and the peak acceleration of the headform inside the helmet. Information on head injuries was also obtained. Forty helmets sustaining impacts in crashes were collected from cyclists who were killed or treated at selected Melbourne hospitals during These helmets were predominantly "foam-only" (a foam helmet often with a material cover), "micro-shell" (a foam helmet with a thin plastic shell), or light weight "hard-shell" (a foam helmet with a hard plastic shell) allowed under the amended Standard. The new helmets were also tested by Technisearch, and information on the bicyclists' injuries obtained, so that comparison could be made with the information obtained by the RACS. The conclusions regarding the protective performance of the new helmets, in comparison with the older design, heavier hard-shell helmets, were: 1. The new helmets transmit a lower level of peak acceleration to the cyclist's head inside the helmet, for a given severity of impact on the external surface of the helmet, for a range of impact types representative of those occurring in real bicycle crashes (the majority resulting in blunt impacts to the helmets). 2. There was no evidence of a real difference in protective performance between the older and new helmets so far as actual head injury risks are concerned. This may have been due to the absence of a difference or due to the relatively small number of helmets considered in the two helmet groups. It was also concluded that the specified drop height of 1500 mm for the impact energy attenuation test in the Australian Standard has been set too low if the intention is to cover closer to the full range of impact severities experienced by the helmets of cyclists involved in crashes resulting in severe injury. In addition, since one-third of the major impacts on the new helmets occurred below the test line, consideration could be given to lowering the line to ensure that helmets provide protection against a larger proportion of impacts sustained in real crashes.

4 ACKNOWLEDGMENTS A project as long and as complex as this could not have been carried out without the help and cooperation of a number of people. The authors particularly wish to acknowledge: VicRoads (Roads Corporation of Victoria) for sponsoring the project Mr Ron Christie, Ms Fairlie Nassau and Ms Andrea Anderson of VicRoads Road Safety Department who supported and provided advice for the project The management, staff and Human Ethics Committees of the following hospitals who provided access to bicyclist patients for interview and to their medical records: Royal Children's Hospital Westem Hospital Dandenong and District Hospital Preston and Northcote Community Hospital Box Hill Hospital Alfred Hospital Dr John Lane, Member of the Victorian Road Trauma Committee, Royal Australasian College of Surgeons (RACS), and Principal Research Fellow at MUARC, who provided valuable guidance throughout the study Dr Joan Ozanne-Smith, Director of the Victorian Injury Surveillance System and Senior Research Fellow at the Monash University Accident Research Centre (MUARC), who prepared the submissions to the hospital Human Ethics Committees and provided advice on the collection of patient injury data SRNs Barbara Fox and Di Holtz, Research Nurses at MUARC, who conducted the interviews with injured cyclists, arranged collection of their helmets where,appropriate, and extracted and coded details of cyclists' injuries from medical and Coroners' records Mr George Rechnitzer, Senior Research Fellow at MUARC, who investigated crashes resulting in cyclists being killed and arranged the collection of their helmets Mr Martin Williams, Manager, Engineering and Scientific Services of Technisearch Limited, who diligently undertook the impact testing of the helmets collected from the killed and injured cyclists Manufacturers and importers who provided new helmets (some free of charge or at a discount rate) for use in the impact testing program Professor Frank McDermott, Chairman of the Victorian Road Trauma Committee, RACS, for providing access to the impact test results and injury information collected during the College's study of bicycle helmets impacted in crashes Ms Anne Tremayne of the State Coroner's Office, Victoria, who provided data collected during the RACS's study Mr Tri Le, Computer Systems Officer at MUARC, who entered the new data, established the database for comparing the helmets, and provided assistance with'the statistical analysis last, but not least, the injured cyclists who provided information about their crashes and their helmets for testing

5 THE PROTECTIVE OF BICYCLE S INTRODUCED AT THE SAME TIME AS THE BICYCLE WEARING LAW IN VICTORIA Table of Contents Page No. 1. BACKGROUND 1 2. PREVIOUS RESEARCH 1 3. COLLECTION Helmet collection 3.2 Patient interview 3.3 Patient injury information 3.4 Helmet impact tests ANAL YSIS AND RESULTS 4.1 Principal points of impact 4.2 Distribution of drop heights from impact tests 4.3 Head acceleration related to impact severity 4.4 Head injuries DISCUSSION CONCLUSIONS REFERENCES 13 APPENDICES A. Summary of helmets collected B. Patient Interview form C. Patient Information form D. Helmet impact test reports from Technisearch

6 THE PROTECTIVE OF BICYCLE S INTRODUCED AT THE SAME TIME AS THE BICYCLE WEARING LAW IN VICTORIA 1. BACKGROUND In September 1989, the Victorian Government announced that the wearing of approved bicycle helmets would become mandatory in that State from 1 July At the same time the Government moved to permit the wearing of the lighter and better-ventilated helmets then existing, as well as helmets approved under the then current Australian Standards AS and AS As an interim measure prior to the introduction of the new Australian Standard, Vic Roads established an approval system for helmets satisfying the impact energy attenuation test and the helmet stability test of the 1986 Standards. The amended Standard introduced in April 1990, AS , confirmed these test requirements, and specifically deleted the requirements for a hard shell, maximum size of ventilation openings, and resistance to a penetration test. The VicRoads interim approval system was phased out in favour of approval to the new Standard on 1 August This project was commissioned by VicRoads to examine any changes in helmet performance due to the change in the Australian Standard for bicycle helmets, which was made at essentially the same time as the introduction of the bicycle helmet wearing law in Victoria. There was concern that the deletion of the penetration test from the Standard may have resulted in reduced protection to the heads of cyclists involved in crashes. 2. PREVIOUS RESEARCH The Royal Australasian College of Surgeons (RACS) collected 64 helmets which had sustained an impact in a crash, as part of a larger study of bicyclist injuries (McDermott et al 1993). These crashes had resulted in the helmet wearer being admitted to or treated at one of 11 hospitals in Victoria during The helmets were submitted to the testing laboratory of Technisearch Limited, who simulated the principal helmet damage by impact tests on new helmets of the same make, model and size. The test results included the drop height (measuring the impact severity) and the peak acceleration of the headform inside the helmet. Hospital records were interrogated to obtain details of the actual head injuries sustained (if any) and descriptions of the circumstances of the crashes were obtained (Williams 1991). The majority of the helmets (61, or 95%) consisted of a hard shell with an expanded polystyrene (EPS) foam impact-absorbing liner. Fifty-three (85%) were designed to meet the requirements of the Australian Standard before its 1990 amendment. The remainder were imported helmets which had not been submitted for Australian Standards approval. Thus the group of 64 helmets were representative of the range of helmets being worn and involved in crashes prior to the bicycle helmet wearing law, ie. they were mainly heavy weight hard-shell helmets approved under the old Australian Standard. The data set of impact test results and head injury information collected by the RACS represented a valuable basis for a comparison of the new, lighter, "foam-only" and "micro-shell" helmets permitted under the amended Standard.

7 3. COLLECTION Forty helmets sustaining impacts in crashes were collected from cyclists who were killed or treated at selected Melbourne hospitals during These helmets were predominantly "foam-only" (a foam helmet often with a material cover), "micro-shell" (a foam helmet with a thin plastic shell), or light weight "hard-shell" (a foam helmet with a hard plastic shell) allowed under the amended Standard (or the VicRoads interim approval system). The collection of these helmets and associated data, including the results of impact testing by Technisearch, will be described in the following sections. The information collected was intended to be comparable with information obtained by the RACS in their study of bicyclist helmets impacted in crashes during Helmet collection Arrangements were established with six Melbourne hospitals to be advised of bicyclists who had been admitted or otherwise medically treated and to obtain access for initial interviews. Permission was granted from the Royal Children's Hospital, Western Hospital, and Dandenong and District Hospital in April 1991, Preston and Northcote Community Hospital in May 1991, and Box Hill Hospital in June The Alfred Hospital was added to the group in May At the interview (usually in a hospital ward), the patient or his/her parents were asked whether the patient was wearing a bicycle helmet at the time he/she was injured and, if so, whether the patient's helmet struck the ground or another object. If the helmet had been impacted, the interview continued and the patient was asked to supply the helmet for testing in exchange for a voucher to purchase a new helmet up to a value of $50. Informed consent to access the patient's medical record to obtain details of the injuries sustained was also obtained directly from the patient. Three helmets worn by fatally injured cyclists were also collected. In these cases the helmets were sought and obtained by the Police, and the information on the crash circumstances and the cyclist's injuries was obtained from Coroner's records. Thirty-seven helmets were obtained which were representative of helmets approved under the amended Australian Standard, AS , plus three hard-shell helmets representative of the trend towards lighter weight helmets before the Standard was amended (Table 1 and Appendix A). All were considerably less massive than the heavy weight hardshell helmets approved under the old Standard, which typically weighed nearly 600 grams. Table 1: Types of helmets collected from cyclists who were killed or treated in hospital. The helmets were considered to have been impacted in the crash. shell ) Probably Capable Category collected Comment No ofcapable passing Mass by40 Technisearch of AS passing AS and AS

8 An additional 25 helmets of the heavy weight hard-shell type approved under the old Standard were also collected, but these have been held in reserve and were not submitted for impact testing. 3.2 Patient interview If the patient's helmet was impacted, the patient was interviewed to obtain some details of the circumstances of the crash. The information was recorded on the Patient Interview form in Appendix B (or on a variation of this form if the interview was with the patient's parent). Thirty (75%) of the patients' crashes involved a collision with a motor vehicle. The remainder were single bicycle crashes. This contrasts with the crashes which involved the wearers of the 64 helmets collected by the RACS, where 52% involved a collision with a motor vehicle, 39% involved a single bicycle, and in the remainder the wearer's bicycle collided with another bicycle or a jogger (Williams 1991). All but three of the patients were riding on a bitumen road and two of the remainder crashed on a concrete driveway or cycle track. The exception was a cyclist riding on a grass track when he crashed. Thirty-five of the patients (88%) were certain or probably certain that their helmets were retained on their heads at the time it was impacted in the crash. Another two patients did not know. The remaining three patients considered that their helmets came off before impact. It should be noted that many of the patients suffered concussion and that their opinions on this subject may not be reliable. In two of the 40 cases, Technisearch considered that the patient's helmet could not have been retained on hislher head at the time of the impact. Because there could be no clear association between the impact severity and the patient's, (head) injuries in these circumstances, it was decided not to subject these helmets to impact testing. For similar reasons, only 58 ofthe 64 helmets collected by the RACS were impact tested; the remainder had not been retained on the cyclist's head, had been run over by a motor vehicle, or more than one impact had occurred on the same site (Williams 1991). Of the 38 helmets tested, the principal point of impact in 25 cases (66%) was with a bitumen roadway or concrete surface. Eleven (29%) principal points of impact were with a vehicle metallic surface or windscreen. The remaining two cases involved an impact of the helmet with a flat electricity pole and an impact with a bicycle pedal. In comparison, 62% of the impacts on the helmets collected by the RACS were with a bitumen road and 13% were with a sand or dirt path or track. Only 20% were with a vehicle panel or windscreen (Williams 1991). Visual inspection indicated that none of the helmets had sustained a penetrating had none of the helmets collected by the RACS. impact, as 3.3 Patient injury information The medical records of interviewed patients were interrogated to determine the injuries they sustained and the duration of any loss of consciousness. Coroner's records were accessed in 3

9 the cases of killed bicyclists. in Appendix C. The information was recorded on the Patient Information form The recorded injuries were coded on the Abbreviated Injury Scale (AIS) using both the 1985 and 1990 versions (AAAM 1985, 1990). The AIS measures the threat-to-life of individual injuries on an internationally recognised scale. While the 1990 version reflects finer levels of severity of head injuries, it was not available when the head injuries bicyclists included in the RACS data series were coded. Thus the 1985 coding of the bicyclists injuries included in this new study was necessary to allow a comparison injuries in the two data sets. of of head Information was recorded to allow coding of the Glasgow Coma Scale of conscious state for those hospitalised patients who had sustained a head injury. While 15 out of the 40 cases had sustained a head injury, only 10 had sufficient complete information to code the Glasgow Coma Scale. This was considered too few cases to make analysis of the Glasgow Coma Scale worthwhile as an additional measure of head injury. 3.4 Helmet impact tests Thirty-eight newly collected helmets were tested by Technisearch Limited, who have had extensive experience in impact performance testing. The test procedure closely followed the impact energy attenuation test in the Australian Standard AS (Williams 1991), which is also required in The same procedure had been used to test the helmets collected by the RACS during The main points of impact on each helmet were determined by the depth, area and shape of permanent crushing that remained on the surfaces of the expanded polystyrene (EPS) energy absorbing material from which the helmet was constructed. Up to three such points were located, but in the majority of tested cases (24) only one major point was found. A total of 54 main points of impact were found. Technisearch provided the location of the impacts on the helmets in relation to the test line specified in the Australian Standard AS The Standard specifies that helmets must satisfy the performance tests when impacted anywhere above the test line, but does not necessarily require satisfactory performance below the line. Eighteen (33%) ofthe 54 major points of impact on the newly collected helmets occurred below the test line. Among the helmets collected by the RACS, 63% of the major impacts occurred below the test line (Williams 1991). Four new helmets of the same make, model and size as each impacted helmet were obtained from manufacturers and importers and then passed to Technisearch. Technisearch simulated the damage at each main impact point on the impacted helmets by dropping the new helmets, strapped to an instrumented aluminium headform, in guided free-fall onto a steel anvil. The new helmets were dropped from progressively greater heights until the damage sustained by the test helmet was similar to that produced on the impacted helmet during the crash. The shapes of the steel anvils used were chosen to represent approximately the shape of the surface which the impacted helmet hit at each impact point during the crash. For all but five 4

10 of the impact points (91%), a flat anvil was used. In two cases the anvil was a 50 mm round cross-section bar, and the other anvils were a 20 mm square cross-section rod, a 20 mm "H"-section rod, and the end of a 12 mm round rod. It is understood that a flat anvil was used for each impact test of the helmets collected by the RACS, because the surface struck in the impact was generally flat or there was insufficient information to determine its shape. The drop height obtained from the impact test was considered tq be a measure of the impact severity to which the helmet was exposed in the crash. Instruments within the headform measured the peak acceleration which was transmitted through the helmet structure. Further details of the test procedure and the accuracy and reliability of the results are given by Williams (1991). The test reports provided by Technisearch are given in Appendix D. It should be noted that the injury information recorded on the report forms was preliminary information provided to Technisearch to assist in locating the main points of impact and is not necessarily the same as the information extracted and coded from medical and Coroner's records (see Section 3.3). 4. ANALYSIS AND RESULTS 4.1 Principal points of impact The analysis was focused on the test results for the principal point of impact, chosen as the point where the impact tests had suggested that the greatest impact severity (highest estimated drop height) had been applied during the crash. It was presumed that the most severe head injury (if any), measured on the AIS scale, was related to the impact at this point. Only the maximum AIS of head injury was available for the bicyclists with tested helmets included in the RACS series. 4.2 Distribution of drop heights from impact tests Sixteen of the 38 newly collected helmets (42%) had impact test results suggesting they were exposed to impact severities equivalent to drop heights below 250 mm. In contrast, none of the helmets collected in the RACS series had estimated drop heights below 250 mm. It was considered that the newly collected helmets, being predominantly foam-only or micro-shell type, were more likely to display external damage than the hard-shell helmets. Thus they were more likely to have been considered by the cyclist to have been impacted during the crash, and thus to have been included in the study after sustaining low impact severities, than the hard-shell helmets. This difference between the two helmet collections made it imperative that the impact severities to which each group were exposed should be taken into account in the analysis. Four (or 11%) of the newly collected helmets had drop heights estimated as exceeding 1500 mm, the height from which the impact energy attenuation test in the Australian Standard AS is performed. Ten per cent of the drop heights estimated for the helmets collected in the RACS series also exceeded this level. Thus a significant proportion of helmets involved in real crashes leading to severe cyclist injury appear to be exposed to more severe impacts than the test in the Standard requires. 5

11 4.3 Head acceleration related to impact severity Figure 1 shows that there were strong, but different, relationships between the impact severity (measured by the drop height) applied to each helmet, and the resulting peak acceleration experienced by the head form (as a proxy for the cyclist's head), for each of the two sets of helmets. The helmets in the newly collected set (1991/92) appear to produce lower peak head accelerations for a given impact severity, compared with the helmets in the older set ( ). FIGURE 1 RELATIONSHIP BETWEEN HEAD ACCELERATION AND DROP HEIGHT " s: ll.. u ~ 100 "" Gi l: o o 0 o 9.j). o Drop height (mm) 1991/92 o regression /92 regression Linear regressions were fitted to the relationships shown in Figure 1 for each of the helmet sets. The 95% confidence limits for the estimated regression slopes and intercepts did not overlap when the two helmet sets were compared. Thus there was a statistically significant difference between the two relationships. The higher head accelerations, for a given impact severity, apparently experienced by the wearers of the older helmets, which were predominantly hard-shell types, may be due to the plastic shell deflecting elastically giving a non-negligible rebound velocity and hence a higher velocity change to the cyclist's head. It may also be due to the thick shell, when impacting a flat surface, spreading the load sufficiently widely for too little of the crushable EPS foam to be engaged in absorbing energy. 4.4 Head injuries The AIS of the most severe head injury sustained by each injured cyclist, plotted against the drop height, is shown for each of the two sets of helmets in Figure 2. Many of the cyclists did not sustain any head injury (AIS = 0), but sustained injuries to other body regions requiring treatment in hospital. There is a general tendency for a greater proportion to have sustained a head injury when their helmets have sustained a greater impact severity. 6

12 FIGURE 2 MAXIMUM AIS OF HEAD vs DROP HEIGHT 3+ o 0 0 o ~" :? 2 'tl I'll Gl.c '0 III C( E" E 1 ';( I'll ::E ElOO o 00 e 0 o O. o 0 o 0 o o I_ _~~ ----~.~.~--~--~.~~I----~---~----~ o Drop height (mm) 1991/92 " Figures 3 and 4 show the percentage distribution of head injuries at different severity levels in 200 mm ranges of drop height. There was no clear pattern of increasing frequency or severity of head injury with the drop heights of the new helmets, possibly due to the relatively small number of these helmets (N = 38), but there was some indication of a trend among the head injuries sustained by wearers of the older helmets (N = 56) (Figure 4). FIGURE 3 DISTRIBUTION OF HEAD SEVERITY ACCORDING TO DROP HEIGHT -1991/92 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Hoo Drop height (mm) I 0 No head injury 0 AIS=1 AIS=2.AIS=3+ 7

13 FIGURE 4 DISTRIBUTION OF HEAD SEVERITY ACCORDING TO DROP HEIGHT % % 70% 60% 50% 40% 30% 20% 10% 90% Drop height (mm) I 0 No head injury 0 AIS=1 AIS=2 AIS=3+ In order to smooth the trends in the data on the limited number of helmets available, Figures 5 and 6 show the percentages of injured cyclists with head injuries in cumulative ranges of drop height. It can be seen that there is a tendency for a reducing percentage of cyclists to have escaped any head injury as the range of drop heights increases. There is also a tendency for the percentage sustaining severe injury (AIS at least 2) to increase. FIGURES PROPORTION OF INJURED CYCLISTS WITH HEAD INJURIES ACCORDING TO CUMULATIVE RANGES OF DROP HEIGHT -1991/92 100% 80% 60% 40% 20% 0% <=200 <=400 <=600 <=800 <=1000 <=1200 <=1400 <=1600 <=1800 <=2000 <=2200 <=2400 Drop height (mm) I 0 No head injury 0 AIS=1 AIS=2 AIS=3+ 8

14 FIGURE 6 PROPORTION OF INJURED CYCLISTS WITH HEAD INJURIES ACCORDING TO CUMULATIVE RANGES OF DROP HEIGHT % 80% 60% 40% 20% 0% <=200 <=400 <=600 <=800 <=1000 <=1200 <=1400 <=1600 <=1800 <=2000 <=2200 <=2400 Drop height (mm) I D No head injury 61AIS=1 AIS=2 AIS=3+ Figure 7 compares the cumulative percentages of the cyclists in each group who sustained any head injury (ArS = 1 or more), as a function of increasing drop height, for the two sets of helmets. It can be seen that, in general, the new helmets were associated with a lower proportion of cyclists sustaining head injuries than the older helmets, for impact severities up to any given level (the exception being for impacts equivalent to drop heights below 200 mm). However the difference between these two cumulative distributions was not statistically significant when tested by the Kolmogorov-Smimov (K-S) test (Neave 1981) (K-S test statistic = 78; p > 0.05). FIGURE 7 CUMULATIVE PROPORTION OF CASES WITH A HEAD ACCORDING TO DROP HEIGHT " ~" ~ c <= ~ 0" o'. J- -c:-.. _()II - _0.0-0 _.. 0"" - _ <=400 <=600 <=800 <=1000 <=1200 <=1400 <=1600 <=1800 <=2000 <=2200 <=2400 Drop height (mm) data data I 9

15 Figures 8 and 9 show the same comparison for the more severe head injuries, defined as those with AIS greater than 2 and 3, respectively. Figure 8 shows that the new helmets were associated with a lower proportion of cyclists sustaining AIS 2 and above head injuries, for impacts equivalent to drop heights greater than 800 mm. This difference was not statistically significant (K-S test = 88; p > 0.05). FIGURES CUMULATIVE PROPORTION OF CASES WITH AT LEAST AN AIS 2 LEVEL HEAD ACCORDING TO DROP HEIGHT ~ ~ 50 l:! er , 10, o 6', I 4=- _.0. _.0- - ()o 00 -c <=200 <=400 <=600 <=800 <=1000 <=1200 <=1400 <=1600 <=1800 <=2000 <=2200 <=2400 Drop height (mm) data -.()o dataI In contrast, Figure 9 shows that the new helmets were associated with a higher proportion of cyclists sustaining the more life-threatening AIS 3 and above head injuries. However, this difference was also not statistically significant (K-S test = 5; p> 0.05). FIGURE 9 CUMULATIVE PROPORTION OF CASES WITH AN AIS 3 LEVEL OR ABOVE ACCORDING TO DROP HEIGHT HEAD ~ i 50 l:! er I() ~ <5.....:J :J-... -<:) _0_ o ~ <=200 <=400 <=600 <=800 <=1000 <=1200 <=1400 <=1600 <=1800 <=2000 <=2200 <=2400 Drop height (mm)! data dataI 10

16 5. DISCUSSION In reviewing the results of the comparison of the performance of the older and new helmets which have sustained impacts in crashes, it needs to be noted that the two sets of helmets were involved in somewhat different types of crashes. The new helmets were more often involved in collisions on the road with a motor vehicle and, as a result, the impacts on the helmets were more often from contacts with hard surfaces such as bitumen roads and parts of vehicles. As a further result of these different crash circumstances, the helmet impact test program found it appropriate to use non-flat anvils for some (five) of the drop height tests of the new helmets, because these were considered to represent the actual contact surface better than a flat anvil. A flat anvil had been used in all drop height tests of the older helmets. It should be noted that Williams (1990) had found that foam-only helmets had generally performed better (ie. lower surface forces and peak accelerations of the headform), for a given drop height, than hard-shell helmets when flat anvils were used, and that the foam-only helmets had been disadvantaged when tested on non-flat anvils, especially those with smaller radii of curvature. Notwithstanding this, it is believed that the impact testing provided a reliable estimate of the impact severity (measured as equivalent drop height) applied to the cyclist's helmet at each major point of impact during the crash. The testing program also provided a measure of the output from this severity of impact in terms of the forces on the cyclist's head inside the helmet, measured by the peak acceleration of the headform to which the test helmet was strapped. It was found that the new helmets appeared to have been exposed to a greater proportion of contacts with impact severities at low levels than the older helmets. However, at higher levels of impact, both sets of helmets had been exposed to a broad range of impact severities up to levels equivalent to drop heights exceeding 2000 mm. Nevertheless, the disparity in the distribution of impact severities to which the two sets of helmets were exposed made it imperative that impact severity was taken into account in the analysis, and this was subsequently done throughout. It was noted that in none of the cases did a new helmet appear to have sustained a penetrating impact (this was also the case for the helmets collected by the RACS). This is noteworthy because it suggests that few, if any, penetrating impacts occur in real bicycle crashes. This suggests that the deletion of the penetration test from the Australian Standard may not be considered to have relaxed or weakened the Standard, because in fact the test has little or no relevance. More than 10% of helmets collected in each series appeared to have sustained contacts with impact severities of greater magnitude than the equivalent drop height of 1500 mm specified for the impact energy attenuation test in the Australian Standard. This suggests that the specified drop height has been set too low if the intention is to cover closer to the full range of impact severities experienced by the helmets of cyclists involved in crashes resulting in severe Injury. 11

17 In addition, it was found that one-third of the major impacts on the new helmets occurred below the test line (over 60% of the major impacts on the helmets collected by the RACS fell below the line). Since the Standard currently does not require satisfactory performance below the test line, consideration could be given to lowering th~ line to ensure that helmets provide protection against a larger proportion of impacts sustained in real crashes. The relationship between the estimated impact severity on the helmet and the estimated peak acceleration of the cyclist's head inside it was clear and direct for each of the two helmet sets. There was also statistically significant evidence that the two relationships different, with the new helmets apparently transmitting lower accelerations to the cyclist's head, for a given impact severity, than the older helmets. This suggests that the new helmets, which were predominantly absorbing and distributing their predominantly were foam-only (or had light weight shells), were better at blunt impacts than the older helmets, and is consistent with Williams (1990) finding. This generally superior performance was observed even though five of the new helmets had been exposed to significant impacts with non-flat surfaces and were tested with appropriate non-flat anvils. When the incidence and severity of the head injuries of the cyclists wearing the tested helmets was analysed, taking the impact severities into account, the results suggested that, in comparison with the older helmets, the new helmets displayed: (a) better protection against head injuries of minor or moderate severity (AIS of 1 or 2), and (b) worse protection against severe to critical head injuries (AIS of 3 and above). However, none of the analyses comparing the head injuries of the cyclists wearing the two groups of helmets were statistically significant. Thus there was no evidence of a real difference in protective performance between the older and new helmets so far as actual head injury risks are concerned. This may have been due to the absence of a difference or due to the relatively small number of helmets considered in the two helmet groups. It should be noted that maximum AIS was the only measure of head injury available for comparison of the two groups of helmets. An analysis based on other measures of head injury severity such as the number of head injuries or the Glasgow Coma Scale, had they been available, may have displayed different results. 6. CONCLUSIONS New, lighter bicycle helmets, manufactured entirely of expanded polystyrene foam or covered with a light weight plastic shell, have become common in Victoria following the deletion of the penetration test from the Australian Standard for bicycle helmets at essentially the same time as the introduction of the mandatory' requirement for cyclists to wear approved helmets. The conclusions from this study of the protective performance of the new helmets, in comparison with the older design, heavier hard-shell helmets, were: 1. The new helmets transmit a lower level of peak acceleration to the cyclist's head inside the helmet, for a given severity of impact on the external surface of the helmet, for a range of impact types representative of those occurring in real bicycle crashes (the majority resulting in blunt impacts to the helmets). 12

18 2. There was no evidence of a real difference in protective performance between the older and new helmets so far as actual head injury risks are concerned. This may have been due to the absence of a difference or due to the relatively small number of helmets considered in the two helmet groups. 3. The specified drop height of 1500 mm for the impact energy attenuation test in the Australian Standard has been set too low if the intention is to cover closer to the full range of impact severities experienced by the helmets of cyclists involved in crashes resulting in severe injury. 4. Since one-third of the major impacts on the new helmets occurred below the test line, consideration could be given to lowering the line to ensure that helmets provide protection against a larger proportion of impacts sustained in real crashes. REFERENCES ASSOCIA non FOR THE ADVANCEMENT OF AUTOMOTIVE MEDICINE (AAAM) (1985), The Abbreviated Injury Scale: 1985 Revision. AAAM, Illinois. ASSOCIA non FOR THE ADVANCEMENT OF AUTOMOTIVE MEDICINE (AAAM) (1990), The Abbreviated Injury Scale: 1990 Revision. AAAM, Illinois. AS (1986), Lightweight protective helmets (for use in pedal cycling, horse riding and other activities requiring similar protection), Part 1 - Basic performance requirements, AS Standards Association of Australia, Sydney. AS (1986), Lightweight protective helmets (for use in pedal cycling, horse riding and other activities requiring similar protection), Part 2 - Helmets for pedal cyclists, AS Standards Association of Australia, Sydney. AS (1990), Lightweight protective helmets (for use in pedal cycling, horse riding and other activities requiring similar protection), Part 2: Helmets for pedal cyclists, Standards Australia, Sydney. McDERMOTT, FT, LANE, lc, BRAZENOR, GA, and DEBNEY, EA (1993), The effectiveness of bicycle helmets: a study of 1710 casualties. Journal of Trauma, Vol. 34, pp NEA VE, HR (1981), Elementary Statistical Tables For All Users of Statistical Techniques. Aldren Press, London. WILLIAMS. M. (1990), Evaluation of the penetration test for bicyclists' helmets: comparative performance of hard shell and foam helmets. Accident Analysis and Prevention, Vol. 22, No. 4, pp WILLIAMS, M. (1991), The protective performance of bicyclists' helmets in accidents. Accident Analysis and Prevention, Vol. 23, Nos. 2/3, pp

19 APPENDIX A SUMMARY OF S COLLECTED

20 SUMMARY OF S COLLECTED B1/01/91 ATOM ROSEBANKChallenger ROSEBANKUltralite PRO MET NOLAN EQUINE BELL KIN HEADWAY701 HEADWAYJoey DAVIES HEADWAYFreestyle ATOM YONG HEADGEAR BANKGrand CRAIG SCIENCE Triat V1-Pro Cyclone Zephyr TCB Airlite Pro-Ultracool888 Hartop Pro-Ultracool252 Lucci PB2 LUNG Streamlight Itracool Prix 888 New 501Max 252 Australia USA Australia Italy Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Italy Australia USA Australia Taiwan ABS/EPS Thermoplastic/EPS PBT/EPS EPS PBT/EPS PBT/EPS EPS EPS EPS EPS PBT/EPS EPS ABS/EPS PBT/EPS EPS EPS EPS ABS/EPS EPS EPS EPS EPS EPS EPS ABS/EPS EPS EPS EPS EPS EPS Plastic/EPS PBT/EPS ABS/EPS

21 PATIENT INTERVIEW FORM APPENDIXB

22 Interview Date. MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE PATIENT INTERVIEW MUARC Case No. PATIENT DEI' Aill) Age Sex. ACCIDENT DEI'AILS Date of accident. Where did the accident occur? (e.g.road, footpath) Was another vehicle involved? YES (NO If YES, what type of vehicle was it? Do you know how fast you were travelling? YES / NO If YES, estimated speed. Do you know how fast the other vehicle was travelling? YES / NO If YES, estimated speed. Describe what happened (including why you think the accident happened) :

23 What do you think: caused the injuries? car) (e.g. hitting the roadway/part of the bicycle/part of a Any other comments about the accident : Were you wearing any protective clothing (apart from a bicycle helmet)? YES/NO If YES, please give details: DErAILS What did the helmet strike in the accident? Did the helmet stay on with the impact? YES/NO If NO, was the helmet done up properlylcorrectly fitted? (give details) Were there any problems with the helmet? etc.) (e.g. uncomfortable, too loose, How old was the helmet?. How often was the helmet used?. Where was it stored when not in use?. Are there any visible signs of damage which occurred prior to this accident? YES/NO If YES, describe what happened to cause the damage?

24 Was the helmet ever dropped/thrown/etc.? YES / NO If YES, give details (including how often) Helmet may be collected from : Name ~. Addresss. Day Time. Telephone number.

25 PATIENT INFORMATION FORM APPENDIXC

26 MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE VICTORIAN SURVEIlLANCE SYSTEM MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE BICYCLE EVALUATION PATIENT INFORMATION MUARC case No. HOSPITAL PATIENT DETAILS Age. Sex '". 1. INJURIES BY BODY REGION HeadlNeck Injuries(ISS Body Regions, maximum AIS) 1. AIS90 90 AIS 85 Face Injuries(ISS Body Region, maximumais) 1. AIS AIS AIS AIS AIS AIS 85 AIS AIS AIS90 AIS90 AIS AIS RAILWA Y AVENUE, CAULFIELD EAST 3145 (po BOX 191, CAULFIELD EAST, MELBOURNE, VICTORIA 3145) AUSTRALIA FAX: (61X3) TELEPHONE: (03) loo:

27 Chest Injuries (ISS Body Region, maximum AIS) AIS AIS AIS AIS AIS AIS 85 AIS90 AIS90 AIS AIS 90 AIS 90 AIS 90 Abdomen and Pelvic Content (ISS Body Region, maximum AIS) 1. AIS 85 AIS 90 _ AIS 85 AIS 90 _ 3. AIS 85 AIS AIS 85 AIS AIS 85 AIS AIS 85 AIS 90 _ Extremity and Pelvis (ISS Body Region, maximum MS) MS MS MS85 _ 4. MS AIS MS 85 MS MS90 AIS MS MS90 MS 90 External Injuries (ISS Body Region, maximum MS) 1. MS MS MS MS MS85 6. MS 85 MS AIS90 MS90 MS MS MS90 _ 2. HEAD IDENTIFIED 1 = yes 2=no D

28 3. WAS THE CASUALTY RENDERED UNCONSCIOUS IN THE ACCIDENT? FOR THE PURPOSES OF TIllS STUDY, A CASUALTYIS SAID TO BE UNCONSCIOUS IF: FROM THE POINT OF VIEW OF A BYSTANDE~ THE CASUALTY IS UNROUSABLEAND UNRESPONSIVE, AND/OR FROM THE POINT OF VIEW OF THE VICTIM, HE OR SHE IS ABSOLUTELY UNAWARE OF THEIR SURROUNDINGS,AS IF ASLEEP. 1 = Yes, the casualty was unconscious 2 = No, there was no loss of consciousness 3 = Don't know D 4. DURATION OF UNCONSCIOUSNESS 1 = Only a second or two - transient, momentary 2 = Less than a minute 3 = More than a minute 4 = Less than 1 hour 5 = More than 1 hour D 5. DOES THE CASUALTY REMEMBER BEING AT THE SCENE OF THE ACCIDENT, BEFORE IT OCCURRED? 1 = Yes 2 = No 3 = Not sure 4 = No information available D 6. HEAD CIRCUMFERENCE.. ems 7. WAS AN OPERATION PERFORMED ON THE HEAD? 1 = Yes 2 = No D Describe briefly if yes:

29 8. GLASGOW COMA CHART READINGS (HEAD CASES ONLY) I.E. ACTUAL TRANSCRIBINGS FROM THE HOSPITAL OBSERVATION CHARTS IF AVAILABLE AT THE FOLLOWING TIMES: mms GLASCOW 4 mins 24 injwy 72 Time ADMISSION POSTE.D. hours PRE- since... VERBAL MOTOR 2 4 RESPONSE Words =To = Withdraw Extension fuappropriate Confused Incomprehensible Reflection None To Oriented Obey Localises spontaneous pain voice RESPONSE command (pain) TOTAL 9. LENGTH OF STAY IN EMERGENCY DEPARTMENT... Hours 10. LENGTH OF STAY IN ACUTE HOSPITAL Days 11. SURVIVAL 1 = Alive 2 = Dead from Head Injury Only 3 = Dead from Multiple Injuries (Including Head) 4 = Dead from Complications of Treatment for Head Injury 5 = Dead from Other Injuries 6 = Death Unrelated to Accident D 04107/9416:12 HELM-PAT.DOe

30 APPENDIXD REPORTS FROM TECHNISEARCH

31 Technlsearch,~ I PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists I helmets. CODE Bl/l/91. Atom. Australia. PB2. M57. DATE OF PROD'N lid NUMBER F ABS/EPS. (1) (li) SURFACE RESULT Centre left a* Roadway. 1620mm, 140g. * a = above test line; b = below test line EPS foam fusion failure. Contributed to injury? Possibly. RETENTION Fatal. Extensive distributed brain damage. No skull fracture. Bruised scalp top of head. NUMRER 16653,

32 Technisearch I PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. I MEmOD CODE B5/2/9l. Bell. USA. VI-Pro. S/M. Snell. DATE OF PROD'N lid NUMBER B249l254. Thermoplastic/ EPS. (1) (ll) SURFACE RESULT Front left a* Roadway. l690mm, 134g. * a = above test line,' b = below test line Brittle failure of shell. RETENTION Webbing ends not doubled over to prevent removal from fittings. Fatal. No head injury. Severe chest injuries and fractured spine. NUMBER 16653,

33 Technisearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. phthalate. R CODE DATE OF PBT*/EPS. Australia. 60cm. G Grand AS B6/3/9l. Rosebank. PROD'N Prix. (I) (H) SURFACE RESULT No impact damage a* Nil. N/A. N/A. * a = above test line; b = below test line Helmet not damaged while rider sustained severe head injury. Helmet not on rider's head at impact. RETENTION Retention webbing habitually worn very loose. Indicated by creased chinstrap webbing and wear at crease. Contributed to injury? Yes. Fatal. Comminuted fracture of left parietal and occipital bones with a transverse extension across the middle cranial fossa involving both petrous temporal bones. Subgaleal haematoma over the left parieto-occipital region. Subarachnoid haemorrhages. NUMBER 16653,

34 Technisearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. CODE 01. Nolan. Italy. Zephyr New Max. VicRoads Part DATE OF PROD1N Large. 4/90. lid NUMBER EPS. (1) (ll) SURFACE RESULT Rear left a* Roadway. 540mm,60g. * a = above test line; b = below test line RETENTION Webbing ends were not doubled over to prevent removal from fittings. Male clip held in place by granny knot in webbing. Contributed to injury? Yes. No head injury. Severely grazed L knee, L elbow. Bruised ribs. NUMBER 16653,

35 Technlsearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. CODE 02. Davies Craig. Australia. Hartop. 57. AS DATE OF PROD'N 10/87. lid NUMBER G PBT/EPS. 360mm; Rear (ll) Roadway. left 31g. rim b* Front Rear 20mm 1730mm; s/wagon left square 181g; b* XS pillar. 925N. rod. (1) SURFACE RESULT * a = above test line,' b = below test line EPS foam severely cracked at thin cross-sections - old cracking. RETENTION No head injury. NUMIlER 16653,

36 Technisearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists I helmets. CODE 03. Davies Craig. Australia. Hartop. 53. AS DATE OF PROD'N 12/87. lid NUMBER F PBT/EPS. (I) (11) SURFACE RESULT Front left rim b* Vehicle panel. 80mm, 18g. * a = above test line; b = below test line EPS foam severely cracked at thin cross-section - old cracking. Broken into five pieces. RETENTION Female portion of clip fractured prior to accident. Material fault. Crazed left forehead. Broken R tibia and fibula. Grazed knee and upper L leg. NUMBER 16653,

37 Jechnisearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. CODE 07. Atom. Australia. Airlite Part AS DATE OF PROD'N lid NUMBER EPS foam. (I) (11) Right rear top a* Roadway. SURFACE RESULT 950mm,83g. * a = above test line; b = below test line RETENTION CONOITION Webbing ends were not doubled over to prevent removal from fittings. No head injury. Fractured L femur (transverse, closed). L thigh swollen, pain L shoulder and both elbows. NUMBER 16653,

38 Jechnisearch PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. CODE 08. Pro Headgear. Australia. Pro-Ultracool VicRoads, Part AS DATE OF PROD'N 11/90. lid NUMBER No EPS foam. (I) (ll) SURFACE Rear centre top a* Windscreen. Rear left rim b* Roadway. RESULT 140mm,30g. 280mm; 38g. * a = above lest line; b = below test line RETENTION L.O.C. time not specified. LI vertebra crushed. Lacerations skin and hip. Bruised legs and feet. NUM8ER 16653,

39 i Technisearch I~--- i! PROTECTIVE EVALUATION REPORT Post-crash evaluation of bicyclists' helmets. CODE 10. Atom. Australia. Airlite DATE OF PROD'N 2/91. lid NUMUER EPS foam. (I) (II) SURFACE RESULT Front centre upper a* Rear panel of truck. 280mm,42g. * a = above test line; b = below test line RETENTION Lacerated upper and lower lip. Grazed nose. PRO.JECTNuMRER 16653,