Sign the Resolution Contents | Feedback | Search
DRCNet Home | Join DRCNet
DRCNet Library | Schaffer Library | Drugs and Driving
Sign the Resolution
Contents | Feedback | Search
DRCNet Home | Join DRCNet
DRCNet Library | Schaffer
Library | Drugs and Driving
Department of Orthopaedics and Trauma, Royal Adelaide Hospital, North Tce, Adelaide 5000, Australia
Australian studies of alcohol involvement in pedestrian crashes are reviewed. The paper tabulates the results of these studies with respect to distribution of blood alcohol concentration (BAC), age and sex of intoxicated pedestrians, relationship between age and BAC among both males and females, time of day and day of week of crashes involving intoxicated pedestrians, pedestrian movement and pre-crash drinking patterns.
Overall only seven studies were identified. These indicated that 20-30% of pedestrian casualties had a BAC in excess of 150 mg/dL, with alcohol involvement being greater among fatalities. Few of the seven studies presented results pertaining to the other factors listed above. Those that did found that males comprised 60-70% of casualties with a known BAC and 80-90% of those with a BAC exceeding 150 mg/dL. Crashes occurring at night or on weekends were most likely to involve intoxicated pedestrians. Intoxicated pedestrians were hit most commonly as they crossed a road some distance from a traffic control; in those instances where the site had a traffic control, it was rarely used correctly. The one study that reported pre-crash drinking behaviour of intoxicated pedestrians found that about two-thirds had been drinking beer and about one-half had been drinking in a hotel.
The paper concludes that knowledge of the characteristics of alcohol-related pedestrian crashes in Australia is inadequate; recommendations are made to address these inadequacies.
Pedestrians constitute a significant proportion of crash-related deaths and injuries in Australia. Police data indicate that 419 pedestrians were killed in 1990 and 3,283 were hospitalized, accounting for 18.0% and 13.1% of crash-related deaths and hospitalizations respectively (FORS, 1993). Pedestrian fatality rates for 1990 were 2.5 per 100,000 population in Australia, 2.6 in the USA, 2.2 in Canada and 3.1 in the UK (FORS, 1991; Economic Commission for Europe, 1992).
Intoxication among pedestrians has long been recognized as a risk factor in pedestrian crashes (eg Haddon et al., 1961; Honkanen et al., 1976; Struik & Rogerson, 1988). Indeed, comparisons of the extent of alcohol involvement among fatally injured pedestrians, drivers, passengers and motorcycle riders showed that alcohol involvement, particularly at the higher levels of BAC, was most prevalent among pedestrians (Holubowycz, in press). Recent studies both in Australia (Holubowycz, in press) and the USA (Centers for Disease Control, 1993) indicate that although the numbers and rates of adult pedestrian fatalities have decreased over the last decade, the significant decline in the extent of alcohol involvement observed among fatally injured drivers has not been evident among fatally injured pedestrians.
There are few published studies of alcohol involvement among pedestrians in Australia; these are summarized in Table 1. These studies, however, provide little information on factors other than the distribution of BAC; the BAC results are summarized in Table 2.
Table 1
Description of Australian Studies on Alcohol Involvement among Pedestrians
| Study | Sample | Period | Ages |
|---|---|---|---|
| Ryan & Salter, 1977 | treated at or admitted to Alfred Hosp, Melb | 1975 | >=15 |
| McLean et al., 1979 | crash to which ambulance called, Adelaide | 1975-79 | >=14 |
| Jordan, 1981 | reported crashes, including fatalities | 1977-79 | >=18 |
| Struik & Rogerson, 1988 | treated at or admitted to five Melbourne hospitals | 1985-86 | 20-79 |
| Attewell & Dowse, 1992 | FORS Fatality File | 1988 | all |
| Holubowycz, 1994a | admitted to Royal Adelaide Hospital | 1985-87 | >=16 |
| Holubowycz, 1994b | SA fatalities | 1981-92 | >=16 |
Table 2
Distribution of BAC among Pedestrian Casualties
| Study | BAC (mg/dl) | Number with BAC | ||||
|---|---|---|---|---|---|---|
| zero | 1-150 | >150 | known | unknown (%) | ||
| Ryan & Salter | 74.6% | 12.3* | 13.2** | 114 | unk | (unk) |
| McLean et al | 77.8 | 11.1 | 11.1 | 27 | 5 | (15.6) |
| Jordan | 59.5 | 18.0 | 22.4 | 1690 | 2576 | (60.4) |
| Struik & Rogerson | 59.6*** | 16.7**** | 23.7 | 114 | 42 | (26.9) |
| Holubowycz (a) | 62.3 | 15.6 | 22.1 | 199 | 18 | (8.3) |
| Attewell & Dowse | 55 | 30 | 350 | 192 | (35.4) | |
| Holubowycz (b) | 54.3 | 15.8 | 29.9 | 369 | 33 | (8.2) |
Few studies provide information on the demographic characteristics of intoxicated pedestrians. The available data on sex are shown in Table 3, on age in Table 4 and on the relationship between age and sex in Tables 5 and 6 for males and females, respectively.
Table 3
Sex of Intoxicated Pedestrians
| Study | % males among those with BAC: | % BAC >150 among: | ||
|---|---|---|---|---|
| known | >150 | males | females | |
| Jordan | 65.1 | 90.5 | 31.2 | 6.1 |
| Struik & Rogerson | 69.3 | 81.5 | 27.8 | 14.3 |
| Holubowycz (a) | 60.8 | 86.4 | 31.4 | 7.7 |
| Holubowycz (b) | 69.9 | 90.0 | 39.0 | 9.6 |
Table 4
Age of Intoxicated Pedestrians
| Study | Of those with BAC >150, % aged: | |||||
|---|---|---|---|---|---|---|
| 18-25 | 26-35 | 36-45 | 46-55 | 56-65 | >65 | |
| Jordan | 19 | 14 | 15 | 23 | 15 | 15 |
| Struik & Rogerson* | 41 | 26 | 4 | 18 | 7 | 4 |
| Holubowycz (a) | 39 | 23 | 13 | 7 | 13 | 5 |
| Holubowycz (b) | 29 | 18 | 15 | 16 | 16 | 6 |
Table 5
Relationship between Age and Blood Alcohol Concentration among Males
| Study | % of age group with BAC >150: | |||||
|---|---|---|---|---|---|---|
| 18-25 | 26-35 | 36-45 | 46-55 | 56-65 | >65 | |
| Jordan | 32 | 36 | 36 | 44 | 33 | 18 |
| Struik & Rogerson* | 33 | 45 | 9 | |||
| Holubowycz (a) | 36 | 45 | 36 | 27 | 26 | 11 |
| Holubowycz (b) | 47 | 44 | 46 | 59 | 41 | 12 |
Table 6
Relationship between Age and Blood Alcohol Concentration among Females
| Study | % of age group with BAC >150: | |||||
|---|---|---|---|---|---|---|
| 18-25 | 26-35 | 36-45 | 46-55 | 56-65 | >65 | |
| Jordan | 5 | 6 | 11 | 12 | 7 | 2 |
| Struik & Rogerson* | 25 | 14 | 6 | |||
| Holubowycz (a) | 14 | 11 | 20 | 0 | 11 | 0 |
| Holubowycz (b) | 24 | 40 | 80 | 13 | 0 | 0 |
Table 7
Time of Day and Day of Week of Crashes Involving Intoxicated Pedestrians
| Study | Findings |
|---|---|
| McLean et al | all crashes with pedestrian BAC >=100 were at night |
| Jordan | 78.5% of crashes with ped BAC >150 were between 6pm & midnight Thursdays, Fridays, Saturdays |
| Struik & Rogerson | 81.5% of crashes with ped BAC >150 were between 6pm & 6am 51.9% of crashes with ped BAC >150 were on a weekend |
The movements of intoxicated pedestrians within the road environment have not been commonly documented; the findings are presented in Table 8 below.
Table 8
Pedestrian Movement in Crashes Involving Intoxicated Pedestrians
| Study | BAC | Findings |
|---|---|---|
| Jordan | >150 | 38.4% far side; 27.8% near side; 10.8% playing or lying on road; 76.5% occurred with no traffic control at site |
| Struik & Rogerson | >150 | 25.9% far side; 66.7% near side; 0% standing or lying on road; 85.2% occurred >21 m from traffic control |
| Attewell & Dowse | >50 | 30.7% were far side; 29.3% near side; 33.6% on carriageway (playing, lying, working, standing, walking with or against traffic), 3.6% emerging (came from in front of stationary or parked vehicle), 2.9% other |
Only two studies have documented the drinking context of intoxicated pedestrians prior to their crash involvement, the findings of which are shown in Table 9 below.
Table 9
Pre-Crash Drinking Patterns of Intoxicated Pedestrians (if known)
| Study | BAC | Findings |
|---|---|---|
| McLean et al | >0 | predominantly in hotels beer drinkers regularly consumed substantial amounts |
| Struik & Rogerson | >150 | 55% drinking in hotel 66.7% drinking beer 65% drinking with family, friends |
A number of other reports have included some data on pedestrian alcohol involvement, although these have been either incomplete or reported in more detail elsewhere (eg Aylward & O'Connor, 1988; Roads and Traffic Authority, 1994; Road Accident Research Unit, 1978).
Table 10 below is taken from O'Connor and Trembath (1995) and shows the legislative requirements for the taking of blood samples for blood alcohol analysis from individuals presenting at hospital emergency departments following involvement in a road crash.
Table 10
Legislative Requirements for the Taking of Blood from Road Trauma Victims for Road
Trauma Victims Attending or Admitted to Hospital (from O'Connor & Trembath, 1995)
| NSW | VIC | QLD(a) | SA | WA | TAS | NT | ACT | |
|---|---|---|---|---|---|---|---|---|
| Hospital to take blood samples from all road trauma victims | * | . | . | * | . | . | * | . |
| Blood sample taken at police request/ discretion | . | * 1 | * | . | * | * | . | * |
| Blood sample waived if: | ||||||||
| - prejudicial to medical condition | * | * | * | * | . | * | * | * |
| - BAC known from other measure | . | * | . | . | . | . | . | * |
| - not driver/rider | * 2 | * 4 | . | . | . | . | . | * |
| Consent not given | . | * 3 | . | * | . | * 3 | . | * |
| Time limits | 12 hrs | . | 2 hrs | 8 hrs | 4 hrs | 3 hrs | 12 hrs to, + 4 hrs from, admis | 2 hrs from incid or admis |
| Minimum age for sample | 15 yrs | 15 yrs | . | 14 yrs | . | . | 15 yrs | . |
As seen in the table only three states/territories, namely New South Wales, South Australia and the Northern Territory, have legislation allowing the mandatory taking of blood samples for blood alcohol analysis from injured pedestrians who present to hospital (subject to the waivers noted in the table). However, it is worth noting that the BACs of seriously injured (ie admitted to hospital) pedestrians are not reported in the routine annual crash reports of these three states, even though the data are presumed to be available. FORS (1993) reported that Australia wide police data on hospitalised (ie presumed admitted) pedestrians indicated that BAC was unknown for 57.2% of pedestrians aged 16 years or older.
According to O'Connor and Trembath (1995), for fatal cases, the coroner can be advised that a sample for blood alcohol analysis should be taken but it is then up to the coroner to request that this is done. The rate at which pedestrian fatalities are tested varies between states. For example, for those states where BAC data have recently been reported for a longer period of time, 52% of pedestrian fatalities of all ages were tested over the period 1986-1993 in Queensland (Fraine, 1994), BACs were known for 92% of pedestrian fatalities aged 16 or over in South Australia from 1981 to 1992 (Holubowycz, in press) and for 75% of pedestrian fatalities from 1983 to 1993 in Victoria (Vic Roads, 1994). It should also be noted that because some pedestrians are likely to die after admission to hospital, the rates of known BACs are likely to be higher in those states with mandatory blood alcohol testing in hospitals.
Haworth & Rechnitzer (1993) reported that BAC data were missing for 55% of 7498 persons included in the 1988 FORS Fatality File and that missing data were more likely for non-drivers/riders and for those not severely injured. They also noted that the percentage of missing data varied substantively between states. Unfortunately their report did not provide any straightforward data on alcohol involvement in pedestrian crashes, as alcohol crashes were defined according to alcohol being coded as the first or second most important contributing factor to the crash and according to the maximum BAC of persons responsible for that crash (ie >150 or 50-150 mg/dl).
The Australian studies indicate that 20-30% of pedestrian casualties have a BAC in excess of 150 mg/dl. These findings support the results of case-control studies which have shown that alcohol plays a signifcant role in the aetiology of pedestrian crashes. The extent of alcohol involvement is highest among fatalities.
The few studies that have reported the sex of pedestrian casualties have indicated that males comprise 60-70% of casualties with a known BAC, and 80-90% of casualties with a BAC in excess of 150 mg/dl.
The available data on age and BAC among males shows that in most instances at least one-third of every age group with the exception of the elderly had a BAC in excess of 150 mg/dl. However, the results of the studies did not suggest that any particular age group was consistently more likely to show an elevated level of alcohol involvement. The data on females is less conclusive given the relatively small numbers of casualties.
Alcohol involvement appears to be highest at night and on weekends.
Only three studies examined pedestrian movement in relation to BAC. Crashes involving intoxicated pedestrians predominantly occurred whilst crossing roads at a site some distance from a traffic control. However, if the site had some form of traffic control, this was rarely used correctly.
There was little agreement with respect to the side of the road on which crashes involving intoxicated pedestrians were most likely to occur: one study reported equal representation between near and far-sided crashes, one reported a much higher prevalence of the former and the third reported a somewhat higher prevalence of the latter.
Only one study examined where, what and with whom the intoxicated pedestrians had been drinking prior to their crash. About two-thirds had been drinking beer, a similar proportion had been drinking with family and/or friends and just over one-half had been drinking in a hotel.
The relatively low number of studies and, in particular, recent studies, coupled with the relatively high rates of unknown BACs in most of the studies, limit our knowledge of the true extent of alcohol involvement among pedestrian casualties.
Our knowledge of the characteristics of alcohol-related pedestrian crashes is totally inadequate. This is of major concern because the successful design and implementation of countermeasures requires a sound knowledge of such characteristics.
To address these inadequacies, this report makes the following recommendations:
That selected samples of crash-involved pedestrians are interviewed to ascertain the following characteristics:
These samples should be derived from:
(a) prospective studies of crash-involved pedestrians admitted to hospital
It is recommended that samples be selected on the basis of hospital. Although admissions to any one hospital are unlikely to be representative of all serious pedestrian casualties, the ease of accessing such a sample probably justifies the resulting selection bias, especially as the nature of this bias can be estimated.
(b) retrospective follow-up studies of crash-involved pedestrians
Pedestrians could be identified retrospectively from existing site-specific studies of pedestrian crashes such as that of Toorak Road, South Yarra (Diamantopoulou & Corben, 1994). No alcohol data are available for that study although the crash characteristics suggest a high level of alcohol involvement. Interviewing these pedestrians about their drinking behaviour prior to crash involvement would provide a valuable supplement to the data already available from that study.
That each state routinely collects and reports the following data on fatally and seriously injured (ie admitted to hospital) pedestrians:
That all Australian states adopt a uniform method of reporting alcohol use among killed and injured road users including pedestrians, particularly with respect to BAC categories reported (eg zero, 1-49, 50-79, 80-149, 150-249, >=250 mg/dl) and the ages of pedestrians for whom BACs are reported (eg >=15 years old).
It is recommended that tables be of the following format:
BAC distribution of fatally injured pedestrians aged >=15 yo, by sex
| BAC (mg/dl) | Males | Females | Total |
|---|---|---|---|
| zero | xx.x% | . | . |
| 1-49 etc | xx.x | . | . |
| . | . | . | . |
| Total no. BAC known | yyy (100%) | . | . |
| Total no. BAC unknown | zzz (xx.x%)* | . | . |
| Total no. aged <15 yo | . | . | . |
The important feature of the above table is the reporting of the number and percentage of pedestrians with unknown BACs within the age group for which BAC should be reported, as well as the number below that age limit.
That individual sites at which more than a predetermined number of pedestrian crashes occur over a certain period of time are subjected to a safety audit to determine whether specific environmental modifications would reduce crash numbers.
That in selected hospitals around Australia, data on emergency room attendances (including admissions) resulting from vehicular trauma are linked with police crash reports and blood alcohol data to enable the monitoring of trends in pedestrian crash patterns.
That funding for the implementation of countermeasures is dependent on the inclusion of an evaluation of outcome.
This paper was prepared as part of the Intoxicated Pedestrian Project funded by Austroads; the assistance of Rob Klein, John Collis and Steve Brown is gratefully acknowledged. The views expressed are those of the author and do not necessarily represent those of Austroads.
Attewell RG, Dowse MJ. Fatal crash types. Analysis of 1988 Fatality File. Canberra: Federal Office of Road Safety, Department of Transport and Communications, March 1992.
Aylward P, O'Connor P. Pedestrian safety in South Australia. Adelaide: Road Safety Division, Department of Transport, Report No. 15/88, November 1987.
Centers for Disease Control Alcohol involvement in pedestrian fatalities - United States, 1982-1992. MMWR 42: 716-719; 1993.
Diamantopoulou K, Corben B. Pedestrian safety issues in victoria. Report on investigation of pedestrian crashes in commercial locations. Case No: P3 - Toorak Road between Chapel Street and Punt Road, South Yarra. [draft report, 1994]
Economic Commission for Europe. Statistics of road traffic accidents in Europe. Vol 37. New York: United Nations, 1992.
Federal Office of Road Safety. 1990 road fatality statistics, Australia. Canberra: Federal Office of Road Safety, Department of Transport and Communications, 1991.
Federal Office of Road Safety. Road crashes resulting in hospitalisation, Australia, 1990. Statistical summary, year ended 31 December 1990. Canberra: Federal Office of Road Safety, Department of Transport and Communications, October 1993.
Fraine G. Pedestrian safety and travel in Queensland. Road User Behaviour Section, Queensland Transport. Draft paper, December 1994.
Haddon W Jr, Valien P, McCarroll JR, Umberger CJ. A controlled investigation of the characteristics of adult pedestrians fatally injured by motor vehciles in Manhattan. J Chronic Dis 14: 655-678; 1961.
Haworth N, Rechnitzer G. Description of fatal crashes involving various causal variables. Canberra: Federal Office of Road Safety, Department of Transport and Communications, February 1993.
Holubowycz OT. Age, sex and blood alcohol concentration of killed and injured pedestrians. Accid Anal and Prev (in press).
Holubowycz OT. 1994(a). Unpublished data.
Holubowycz OT. 1994(b). Unpublished data.
Honkanen R, Ertama L, Kuosmanen P, Linnoila M, Visuri T. A case-control study on alcohol as a risk factor in pedestrian accidents. a preliminary report. Mod Probl Pharmacopsychiatry 11: 46-50; 1976.
Jordan PW. Alcohol in pedestrian accidents. Department of Civil Engineering, Monash University, January 1981.
McLean AJ, Brewer ND, Sandow BS. Adelaide In-Depth Accident Study, 1975-1979. Part 2: Pedestrian accidents. adelaide: Road Accident Research Unit, University of Adelaide, 1979.
O'Connor PJ, Trembath RF. An investigation of missing values of BAC in road crash databases. Road Injury Information Program Report Series. Adelaide: National Injury Surveillance Unit, 1995.
Road Accident Research Unit. Multidisciplinary In-Depth Accident Survey (MIDAS) report. Road Accident Research Unit of the Health Commission of Victoria, December 1978.
Roads and Traffic Authority. Submission to the Joint Standing Committee on Road Safety for the Inquiry into Pedestrian Safety. Rosebery, NSW: Road User Safety Branch, Road Safety Bureau, Roads and Traffic Authority, January 1994.
Ryan GA, Salter WE. Blood alcohol levels and drinking behaviour of road crash victims. Department of Social and Preventive Medicine, Monash University, June 1977.
Struik M, Rogerson P. Pedestrian Accident Project. Report No. 5. Methodology and general results. Hawthorn, Vic: Road Safety Division, road Traffic authority, Report No. GR/88/7, July 1988.
Vic Roads. Road Traffic Accidents involving serious casualties, Victoria 1993. Melbourne: Information Services Department, Vic Roads, 1994.