DIFFERENCES BETWEEN DRIVERS INJURED AND NOT INJURED IN COLLISIONS IN VICTORIA, AUSTRALIA JANE HENDTLASS, INGRID BOCK AND MARY RYAN ROAD SAFETY AND TRAFFIC AUTHORITY**, VICTORIA, AUSTRALIA. 3122. 801 GLENFERRIE ROAD, HAWTHORN, Apart from three in-depth studies of casualty crashes which occurred in the State capital cities of Melbourne, Sydney and Adelaide (1,2,3), there seems to be no information about the use of alcohol by both uninjured drivers and those who were taken to the hospital or the morgue from collisions in Australia. This situation has developed because drivers who are killed and those who have been treated in hospitals in Victoria now routinely have a blood sample analysed for its alcohol content and the reading may have subsequent legal implications. However, drivers who are not killed or taken to hospital from road crashes are breath tested at the discretion of the police who attend the scene; in Victoria, less than 3% of uninjured drivers in metropolitan casualty crashes and 11% of uninjured drivers in fatal crashes were tested for their blood alcohol level (4). Drivers who were taken to hospital from Melbourne metropolitan crashes have been shown to differ from the drivers who were not taken to hospital from these collisions in sex, age, occupation and alcohol use (5) and alcohol use was related to treatment in hospital for the men in the two groups and not for the women or for any particular age group or occupational category. * Present address - Police Headquarters, 380 William Street, Melbourne, Australia, 3000. **The opinions expressed in this paper do not necessarily reflect those of the Road Safety and Traffic Authority, the Commonwealth Department of Transport or Victoria Police. 124
In this paper, drivers and motor cycle riders involved in fatal crashes will be looked at from two angles: 1) Drivers who were killed in road crashes in Victoria in 1977 will be compared with Drivers who were not killed in the same collisions; 2) The characteristics of the drinking and non-drinking drivers involved in collisions in which at least one driver was killed will be compared. A drinking driver has been defined as one who has been reported to have been drinking at the time of the collision, that is someone with a blood alcohol concentration over zero g/100 ml or someone reported to have been drinking within four hours of the collision. METHOD All information reported in this paper relates to drivers and motor cycle riders involved in collisions in which at least one driver was killed. These crashes include those in which a driver died from injuries received in the crash, those in which a driver died from natural causes while he was driving and those where the driver died for other reasons such as drowning or incineration. Study, Population - The Study Population consists of all 687 drivers involved in collisions in Victoria during 1977 where a driver was killed as recorded by the Statistics Section of the Traffic Operations Group of Victoria Police. These drivers have been divided into four categories according to their degree of injury: CATEGORY I - Drivers Killed in Single-Vehicle Crashes - 30% CATEGORY II - Drivers Killed in Multi-Vehicle Crashes - 35% CATEGORY III - Drivers taken to Hospital from the Collision, not including those who subsequently died - 10% CATEGORY IV - Drivers not Killed or Taken to Hospital - 25%. Data Collection - Information about each driven was collected from Police, licence, accident, criminal and coronial records. Alcohol use was assessed from blood ro breath analysis where this information existed. Statements from those involved were used where analytical data did not exist and these estimates agreed with the analytical data in fifteen out of sixteen drivers not killed for whom both pieces of information.were available, Data Analysis - Data was analysed using the Statistical Package for Social Sciences (6). Information about place of residence has been grouped into five 125
geographic areas which have been associated with particular socioeconomic status (7) and occupational groupings agreed with those published by the Australian Bureau of Statistics (8). Complete information was not available for all drivers involved in the study and this accounts for differences in sample sizes. Statistical Analyses - Chi-square tests of association were used for comparisons of unweighted data. In comparisons between drinking and non-drinking groups, data from each driver group has been weighted according to the contribution that that group makes to the Study Population. The statistical significance of the difference between two driver samples in terms of the proportion of drivers in a particular categori (e.g. Sex equals Male) were judged by the Normal distribution approximation for the difference of two proportions (9). The finite population correction was employed in estimating the sample variance for each proportion (10) and where a sample proportion was based on samples from more than one stratum (i.e. Drinkers and Non-Drinkers) the sample variance was calculated from Cochrans (10) approximation for the variance of a proportion from a stratified sample. Differences between groups are reported significant (*) from the 5% level, very significant (**) from the 1% level and highly significant (***) from the.1% level. RESULTS 1) Comparison of Drivers Killed in Motor Vehicle Collisions with the Drivers not Killed in the Same Collisions. In 1977, 687 drivers were involved in crashes in which 448 drivers died; this means that one third of the drivers involved were not killed themselves. The characteristics of four different categories of drivers have been campared. General Characteristics (Table 1) - Overall, 87% of the Study Population were men and 41% were under 26 years old. As well, over 60% belonged to blue collar occupations and 45% lived in Rural Victoria. Forty nine per cent of drivers involved in the Study had been drinking before the collision. There were significant differences between the people in each of the four injury categories with drivers killed in single vehicle crashes having a greater proportion of drinkers than any of the other groups. 126
People aged under 26 years and country dwellers were over represented in the fatality groups compared with the drivers who were not killed while city people from areas associated with higher socioeconomic status made a larger contribution to the survivor categories. TABLE 1. General Characteristics of Drivers in the Study Population CATEGORY CATEGORY CATEGORY CATEGORY SIGNIFICANCE I II III IV OF DIFFERENCE Number in 1977 207 241 69 171 SEX (%) (N=207) (N=241) (N=6$) (N=171) * ** - Men 90 82 82 94 - Women 10 18 18 6 AGE (%) (N=207) (N=241) (N=67) (N=157) ieie"k - Under 26 years 43 45 39 32-26 to 40 years 29 25 42 38 - Over 40 years 28 30 19 30 OCCUPATION (%) (N=167) (N=189) (N=62) (N=157) - White Collar 20 23 31 20 - Blue Collar 64 52 53 73 - Pensioner 9 13 2 3 - Other 7 11 14 4 POSTAL ADDRESS (%) (N=207) (N=241) (N=68) (N=171) ** - Urban Area lower SES. 14 11 9 10 higher SES. 26 28 35 36 other 4 7 4 5 - Rural Area 50 47 40 41 - Outside Victoria 6 7 11 8 ALCOHOL USE (%) (N=175) (N=198) (N=42) (N=68) *** - No Use Reported 33 56 57 62 - Use Reported 67 44 42 38 (SES - Socioeconomic Status) 127
Driving Characteristics (Table 2) Information which relates to the crash rather than to the driver includes vehicle type and licence status. About 65% of the Study Population were driving cars or station wagons at the time of the collision, 70% drove a utility or panel van, 12% a motor car and 16% a heavy vehicle such as a truck, train or bus. TABLE 2. Driving Characteristics of Drivers in Study Population CATEGORY CATEGORY CATEGORY CATEGORY SIGNIFICANCE I II III IV OF DIFFERENCE VEHICLE TYPE (%) (N==205) (N=239) (N=67) (N=171) - Car or s/wagon 74 66 72 51 "kie"k - utility or p/van 8 7 5 - Motor cycle 14 23 1 6 - Truck 3 18 29 - Other hf - 1 14 LICENCE STATUS (%) (N=194) (N=230) (N=64) (N=147) - Full licence 54 58 75 80 "k4tk - Learner 5 6 - - - Probationary 24 23 19 17 - Conditional 5 5 2 1 - Unlicenced 11 5 5 3 - Cancelled licence 2 2 - - Similarly, 64% of the drivers involved were fully licenced and one quarter held learners or probationary licences. Seven percent were not licenced to drive. In general, the contribution which each vehicle type made to each of the different driver categories reflected the vehicle's injury potential: motor cyclists were predominant among drivers killed while truck drivers and other such as tram and bus drivers were more numerous among those who survived the collision. Differences in licence type between the four injury categories can be partly explained by the ages of the drivers in each group. There were more learner and probationary drivers, as there were more people aged under 26 years, among these people who were killed. As well, motor cyclists were much more prevalent in the driver fatility groups 128
and those people seem more likely to be young and to hold learner and probationary licences (1). Prior Convictions (Table 3) - At least 30% of the drivers involved in collisions in which a driver was killed had come to notice before for traffic, drink-drive or criminal offences. Thirty one percent had been convicted for traffic offences, 20% had criminal convictions and 6% had previous drink-drive convictions. There were considerably more drivers with prior convictions among those killed than among those not killed. Over one half of the people killed in single vehicle crashes had at least one prior traffic conviction, 28% had previous criminal convictions and 9% had drinkdrive convictions. TABLE 3. Prior Convictions of Drivers in Study Population PRIOR TRAFFIC CONVICTIONS (%) - Nil - One or Two - More than Two CATEGORY CATEGORY CATEGORY CATEGORY SIGNIFICANCE I II III IV OF DIFFERENCE (N=159) (N=197) (N=64) (N=147) 49 58 97 97 26 25 25 17 DRINK-DRIVE CONVICTIONS (%) (N=161) (N=195) (N=64) - Nil 91 91 97 - One 7 7 3 - Two or Three 2 2 - (N=147) 100 PRIOR CRIMINAL CONVICTIONS (%) (N=108) (N=113) (N=68) (N=171) - Nil 72 78 90 90 - One or Two 15 17 9 6 - More than Two 13 5 1 4 By comparison, less than 3% of the drivers who survived had traffic convictions, 10% had a criminal record and less than 2% had previous drink drive convictions. 129
To summarize, drivers who were killed in motor vehicle crashes in Victoria in 1977 were very different from the drivers who survived the same collisions. The dead drivers had a higher proportion of people aged under 26 years, pensioners and country people. They were more likely to ride a motor cycle, to hold a learners permit or a probationary licence and to have prior convictions for traffic drink-drive or criminal offences. Drivers killed in single vehicle crashes had the highest proportion of drinkers but the incidence of alcohol use among drivers involved in multi-vehicle crashes in which a driver was killed was relatively constant across each of the injury categories. 2) The Characteristics of Drinkers and Non-Drinkers involved in Collisions in which at least One Driver was Killed. Half of the drivers involved in collisions in which a driver was killed appear to have been drinking at the time of the crash. This figure comprises 67% of those killed in single-vehicle crashes and 42% of drivers involved in multi-vehicle collisions. The characteristics of the drinking and non-drinking drivers from each injury category have been weighted according to the group's contribution to the Study population and the results have been compared in an effort to try and identify the sorts of people who are more likely to contribute to alcohol-related fatal collisions. General Characteristics (Table 4) - There were three times as many women in the non-drinking group as the drinking group while people aged under 26 years were more numerous among the drinkers. People who lived in rural Victoria were over represented in the drinking group and those with addresses associated with lower socioeconomic status or from outside the State were more numerous among the non-drinkers. 130
TABLE 4. General Characteristics of Drinkers and Non-Drinkers. DRINKERS NON-DRINKERS SIGNIFICANCE Estimated Number in 1977 337 350 OF DIFFERENCE SEX (%) (N=242) (N=231) - Men 93 78 - Women 7 22 AGE (%) (N=242) (N=231) - Under 26 years 45 37-26 to 40 years 27 30 - Over 40 years 28 33 OCCUPATION (%) (N=198) (N=196) - White Collar 18 22 - Blue Collar 61 60 - Pensioner 10 9 - Other 10 9 POSTAL ADDRESS (%) (N=239) (N=226) - Urban Area lower SES 7 13 higher SES 30 32 other 3 7 - Rural Area 57 42 - Outside Victoria 3 5 (SES - Socioeconomic Status) Driving Characteristics (Table 5) - There were significantly more heavy vehicle drivers among non-drinkers compared with drinkers while a greater proportion of the drinking group held a probationary licence or were driving whilst disqualified. The large proportion of probationary licence - holders among the drinking drivers may partly reflect the larger proportion of drivers aged under 26 years in that group compared with the non-drinking group. 131
TABLE 5. Driving Characteristics of Drinkers and Non-Drinkers DRINKERS VEHICLE TYPE (%) (N=241) - Car or Station Wagon 75 - Utility or Panel Van 9 - Motor Cycle 11 - Truck 5 - Other NON-DRINKERS LEVEL OF SIGNIFI- CANCED DIFFERENCE (N=231) 60 *** 7 12 19 2 LICENCE STATUS (%) (N=253) - Current, Full Licence 51 - Learners Permit 3 - Probationary Licence 34 - Cancelled 2 - Unlicenced/Never Licenced 7 - Conditional 3 (N=228) 69 *** 3 1 y * * * 6 6 * * * Prior Convictions (Table 6) - In all cases it was found that drinking drivers involved in a collision in which a driver was killed had a significantly higher proportion of people with longer traffic, drinkdriving and criminal records compared with the non-drinking group. Over one third of the drinking group had previous traffic or criminal convictions and 11% had been convicted for at least one prior drinkdrive offence. On the other hand, only 25% of the non-drinkers had been convicted of traffic or criminal offences before and 3% had prior drink-drive convictions. 132
TABLE 6. Prior Convictions of Drinking and Non-Drinking Drivers PRIOR TRAFFIC CONVICTIONS (%) - Nil - One or Two - More than Two DRINKERS (N=200) 62 22 15 NON-DRINKERS (N=202) 75 *** *** 11 * * SIGNIFICANCE OF DIFFERENCE PRIOR DRINK-DRIVE CONVICTIONS (%) - Nil - One - Two or Three (N=198) 89 7 4 (N=204) 98 3 PRIOR CRIMINAL CONVICTIONS (%) - Nil - One or Two - More than Two (N=202) 80 13 7 (N=231) 91 7 2 In brief, it seems that there was a greater proportion of men, under 26 year olds, people from rural areas, car or station wagon drivers and probationary and cancelled drivers licence holders among drinking drivers involved in a collision in which a driver was killed compared with the non-drinking drivers involved in the same collisions. As well, the drinkers have been shown to have a higher incidence of people with previous traffic, drink-drive and criminal convictions. These factors are heavily influenced by the particular contribution made by drivers killed in single vehicle collisions, the group with a high proportion of drinkers. The factors are also inter related through the large proportion of young people among probationary licensees, among drivers killed in single vehicle crashes and among drinkers. It is difficult to know whether alcohol use in these drivers is a function of age, licence type that is experience, or crash type. 133
DISCUSSION When considering the four categories of drivers involved in collisions in which a driver was killed in Victoria in 1977, it has been shown that dead drivers differed significantly from the survivors of the same collisions in: Sex, Age, Occupation, Place of Residence, Alcohol Use, Vehicle Type, Licence Status, and the number of Prior Convictions. These factors seem generally similar to those which are associated with different injury categories of drivers involved in casualty crashes (4,5). Drivers killed in motor vehicle crashes have different characteristics from the overall group of drivers involved in these fatal collisions in the same way that drivers taken to hospital are not representative of all the drivers involved in injury crashes. As well, when the figures from this study are compared with those from a similar study of metropolitan casualty crashes, it seems that men were more predominant in these fatal collisions (87%) than they were in casualty crashes (77%), there were also more young people in fatal crash group, and there were differences between the two groups in the distributions of occupational categories and places of residence. Alcohol was implicated in half of the drivers in this fatal crash population and in only 24% of the drivers involved in crashes from which a driver was taken to hospital. When the drinking and non-drinking driver groups are compared this paper has confirmed again that the drinking drivers involved in crashes are more likely than non-drinking drivers to be men under 26 years old; and lower socioeconomic status in this case, as measured by place of residence. These characteristics of drinking drivers involved in fatal crashes have been identified in American studies before (e.g. 11, 12) where it has been shown that the alcohol-involved drivers were also more likely to have been at fault in the crash situation. To add to these factors, over 35% of drinking drivers involved in a collision in which a driver was killed had already come to notice for traffic or criminal offences. Eleven percent of the drinking drivers and 9% of all the drivers killed in Victoria had been previously convicted of drink-drive offences. 134
While it has been shown that three quarters of the drinking drivers were killed themselves, with or without another associated death, it could be argued that effective intervention at the time of the earlier drink-drive offences could have saved the lives of at least 32 drivers killed in Victoria in 1977. In dollar terms this is somewhere around $125,000 in State-funded Accident Compensation alone (13). This outlay could be seen to be sufficient to justify further investigation into more cost effective intervention programmes the first time a drink-drive offence is detected. ACKNOWLEDGEMENTS This study was funded by the Australian Commonwealth Department of Transport. The study team is grateful for the active help and co-operation of Victoria Police and the Melbourne Coroner. REFERENCES!'.«Ryan, G.A. Report of Road Accident Research Unit. Consultative Council of Road Accident Mortality. Health Commission of Victoria, 1978. 2. McLEAN, A.J., and Robinson, G.K., Adelaide In-Depth Accident Study 1965-1979, Part I: An Overview. University of Adelaide, 1979. 3. Humphreys, M. In Depth Crash Investigation. Problems and Potentials. Australian Road Research Board Symposium, Melbourne, 1980. 4. Hendtlass, J., Bock, I., and Ryan, M. Drink-Drive Countermeasures in Victoria, in preparation. 5. Hendtlass, J., Bock, I., and Ryan M. Drivers Involved in Metropolitan Casualty Accidents. 1st. Pan Pacific Conference on Alcohol and Drugs, Canberra, Australia, March, 1980. 6. Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K, and Bent, D.H., Statistical Package for Social Sciences. McGraw. Hill N.Y. 1975. 7. Little, F.M., Morozow, W., Rawlings, S.W., and Walker, J.R., Social Dysfunction and Relative Poverty in Metropolitan Melbourne, prepared for Melbourne Metropolitan Board of Works, 1974. 135
8. Australian Bureau of Statistics. The Labour Force, May, 1978. Catelogue No. 6203. 0 Canberra, 1978. 9. Howell, P.G., Introduction to Mathmetical Statistics 3rd edition. Wiley, New York, 1962. 10. Cochran, W.G., Sampling Techniques. 2nd edition, Wiley, New York, 1963. 11. Fell, J.C., Profice of Fatal Accidents Involving Alcohol. Proc. American Association for Automotive Medicine, 21st Conference, Vancouver, 1977, P. 197. 12. May, G.W., Baker, W.E., A Multi-Disciplinary Study of Alcohol-related Accidents University of Mexico, 1977. 13. Motor Accidents Board. Statistics of People Killed or Injured in Road Accidents occurring year ended 30th June, 1978, for which claims were registered with the Board. Preliminary. 1979. 136