Alcohol and pedestrians (No.20)
Table of contents
Executive Summary
As countermeasures against drink driving become more
successful, so the relative importance of the problem of
alcohol and the pedestrian tends to increase.
This report presents the results from three separate, but
linked studies, on various aspects of alcohol and adult
pedestrians.
A Controlled
Study of the Role of Alcohol in Fatal Adult Pedestrian
Accidents in the West Midlands
A study was undertaken of fatally injured adult (over 16
years) pedestrians who died in the West Midlands over a
seven-year period 19901996. With the permission of HM
Coroners, fatal accident files were searched to extract the
relevant information.
Of the 204 fatalities who died within 12 hours of the
accident, blood alcohol concentrations (BAC) were known for
134 (66%) pedestrians. Pedestrians were more likely to be
tested for alcohol if they were male and their accident
occurred late at night. Amongst tested pedestrians, 40% of
males and 13% of females had been drinking
(BAC>9mg/100ml). Nearly a third (29%) of male pedestrians
had BACs in excess of150mg/100ml. No female pedestrian had a
BAC in excess of 100mg/100ml. Significant differences in the
BAC distribution by age group were found with the highest
incidence of drinking being in the 3054 year age group.
Drinking pedestrians are predominantly a night-time
phenomenon; two-thirds of fatalities occurring between 10pm
and 8am had been drinking. No significant differences
occurred between days of the week and quarters of the year.
Control data was gathered at less than half the sites where
the BAC of the fatally-injured pedestrian was known. Despite
the low control rate, no bias amongst those accidents that
were controlled was observed. Relative accident risks were
calculated for male pedestrians. The rates tended to confirm
the findings of previous studies that the risk of accident
involvement increases significantly at BACs above
100-150mg/100ml.
A Controlled
Study of the Role of Alcohol in Adult Pedestrian Accidents in
Cardiff
A study was undertaken of all adult (16 years and over)
pedestrian casualties taken to the Emergency Unit of Cardiff
Royal Infirmary during the period 3 May 1997 to 22
February1999. During the triage process, casualties judged
fit to do so were asked to provide informed consent to
supplying a breath sample for analysis and agreeing to a home
interview at a later date.
Blood alcohol concentrations (BACs) were known for a quarter
(24.5%) of the 360 adult pedestrians taken to the Emergency
Unit. Over 40% of those tested had been drinking
(BAC>9mg/100ml); many had high BACs. Nearly half (48%) of
male pedestrians and over a third (36%) of female pedestrians
had been drinking. No pedestrians over 60 years of age had
been drinking. The incidence of drinking tended to be higher
amongst 30-59year-olds than amongst 16-29 year-olds. The
overall incidence of drinking pedestrians was higher at
weekends and between 10pm and 6am.
Control data obtained from roadside interviewing of
non-accident-involved pedestrians was used to produce
relative accident risks. For BACs above 200mg/100ml, the
relative accident risk was more than 20 times that for a
pedestrian at 0mg/100ml.
Amongst a small sample of drinking pedestrians who were
interviewed subsequent to their accident, many admitted that
the alcohol had affected their judgement of the speed or
distance away of the vehicle that hit them. Most had consumed
an amount of alcohol comparable to their usual consumption
per drinking occasion.
The incidence of adult pedestrian casualties with positive
BACs appears to be increasing, due at least in part to the
increased drinking habits of women.
A Footpath
Survey in Cardiff
Roadside surveys, although widely used for assessing the
level of driving after drinking, have rarely been used with
pedestrians. The prime aims of the present study were to:
determine the incidence of drinking pedestrians and the
effects of situational and temporal factors; determine the
demographic and drinking characteristics of drinking
pedestrians, and determine the proportion of potential
drivers amongst the drinking pedestrians.
Over an eight-month period, five sites in Cardiff were
intensively surveyed between 5pmand midnight on all days of
the week. Over 1,600 adults were interviewed. BAC readings
were obtained from 95% of those interviewed. The proportion
of pedestrians who had been drinking (BAC>9mg/100ml) was
heavily dependent upon time of day, ranging from4% at 17:00
to 71% at 23:00 hours.
The sample consisted mainly of young unmarried males who were
either unwaged (including students) or in socio-economic
group C2. These demographic variables tended to vary by time
of day. Controlling for time of day, age and sex of the
pedestrian tended not to be related to the likelihood of
their having been drinking. Over a third of the pedestrians
were potential (or actual) drivers. The commonest reason why
such individuals had not driven that evening was because they
intended to drink alcohol.
The study has provided a baseline of information on the BAC
distribution of non-accident-involved adult pedestrians in a
city centre during the evening. It has also provided some
evidence as to the effects of countermeasures against driving
after drinking.
Conclusions
The number of adult pedestrian fatalities in Britain is
declining. However, the BAC distribution of those fatalities
who were tested for alcohol showed no significant change
since the earlier study undertaken in the mid 1970s.
Amongst injured pedestrians (in Cardiff), the overall
proportion of pedestrians who had been drinking was higher
than amongst the fatalities in the West Midlands. This result
may be a function of the different age structures of the two
samples; the fatal sample being generally older.
The results of previous studies augmented by some evidence
from the present studies of both fatal and non-fatal
casualties suggest that, for males at least, the risk of
accident-involvement begins to increase dramatically at
levels above 150mg/100ml.
The footpath study was the first of its kind in Britain. It
provides a baseline of information on the incidence of
non-accident-involved pedestrians in a city centre during the
hours 5pm to midnight. Time of day is the most powerful
predictor of whether or not a pedestrian has been drinking
and the likely BAC. Approaching midnight, nearly
three-quarters of those tested had been drinking.
The effect of countermeasures against drinking and driving
was clear. Over a quarter of pedestrians who had been
drinking had taken the conscious decision not to drive to and
from their drinking venue.
It would be useful to monitor the drinking behaviour of
pedestrians at intervals to determine whether their incidence
is increasing. This study has shown that the footpath survey
can provide useful information.
Countermeasures to the drinking pedestrian may be best based
upon the general public health message of the dangers of
excessive alcohol consumption.
Chapter 1 -
Introduction
Inevitably, with walking being the oldest method of transport
and alcohol being one of the oldest known drugs, the problem
of alcohol and the pedestrian is one of the oldest road
safety issues. Despite its antiquity, however, it has
received comparatively little attention from either
researchers or legislators, possibly because of the inherent
difficulties involved in modifying, by legislation or
otherwise, the behaviour of people who generally have the
right to travel along the highway.
A few studies have used a case-control design in which the
blood alcohol concentrations of the pedestrian casualties are
matched to those of a control group of non-accident-involved
pedestrians passing the accident scene at the same time of
day and day of week as the accident. Comparisons between
studies are, however, made difficult by differences in the
way that the relative risk of accident with increasing blood
alcohol concentration was calculated; differences in the
selection of the control groups; and the use of fatal,
non-fatal or all pedestrian casualties.
One of the earliest studies of this type was undertaken by
Haddon et al (1961) in New York City. Of the
accident sample of 20 fatalities who died within six hours of
the accident, 79% had a positive BAC and 42% had a BAC in
excess of 100mg/100ml. This distribution was significantly
higher than that of a control group matched for age and sex.
Clayton et al (1977) collected data on all adult
pedestrian fatalities in the West Midlands Metropolitan area
during 1969-1975. Control data for those fatalities for whom
a BAC was known was collected by roadside interviewing. The
analysis showed the increased risk of fatal accident
involvement associated with BACs of 120mg/100ml or over. Such
pedestrians comprised 27 per cent of male and 7 per cent of
female pedestrians dying within 12 hours of the accident.
Alexander et al (1990) studied a sample of 114
pedestrian casualties aged 2079 taken to hospitals in
Melbourne over an 18-month period in 1985-86. In Victoria,
accident victims are required by law to have their blood
alcohol level measured. Using the same method for calculating
relative risk as that used by Clayton et al, but
with a less closely matched control group, they found a
virtually identical risk level (15x) for BACs
above150mg/100ml.
Blomberg et al (1979), in a study in New Orleans,
used BACs within the range 050mg/100ml as their baseline
(rather than the effective zero (0-9mg/100ml) used by Clayton
et al and Alexander et al). They reported a
fivefold increase in risk at BACs within the range
200-250mg/100ml using a control group matched by age, sex,
accident location and time of day.
Other studies have considered only the role of alcohol in
pedestrian casualties, both fatal and non-fatal.
Everest et al (1991) studied the incidence of
alcohol amongst road traffic accident victims taken to the
John Radcliffe Hospital Oxford in 1988-89. They found that
27% of all injured pedestrians (38% of males and 11% of
females) had a BAC above 80mg/100ml.Pedestrians had the
highest incidence of alcohol of any of the road user groups
studied.
National data for Great Britain shows that the incidence of
alcohol amongst fatally injured adult pedestrians is
increasing. In 1997, 46% of fatally-injured adult pedestrians
(who died within 12 hours of the accident and whose BAC was
measured) had BACs in excess of9mg/100ml and 37% had BACs in
excess of 80mg/100ml (Department of the Environment,
Transport and the Regions, 1999). A decade earlier, the
corresponding percentages were39% with BACs greater than
9mg/100ml and 34% with BACs greater than
80mg/100ml(Department of Transport, 1989). These incidences
are substantially higher than those for riders or drivers.
The data are not split by the gender of the pedestrian.
Fontaine and Gourlet (1997) studied 1,289 fatal pedestrian
accident reports in France for the period March 1990 to
February 1991. They reported that 35% of fatally injured
adult pedestrians (with a known BAC) had a BAC in excess of
80mg/100ml. The incidence was higher in rural areas, at
night, amongst 3044 year-olds, males and unemployed persons.
Holubowycz (1995) studied adult pedestrian fatalities in
South Australia (19811992) and those admitted to an Emergency
Unit in Adelaide during 19851987. Amongst fatalities, half
(50%) of the males and 38% of the females had BACs in excess
of 100mg/100ml.Amongst those taken to hospital, the
incidences were lower39% of males and 29% of females. High
BACs were commonest amongst the young and middle-aged.
Peden (1997) found that, of 227 injured pedestrians presented
to the Trauma Unit of a major hospital in Cape Town, South
Africa, 62% had positive BACs and over 40% had BACs in excess
of 200mg/100ml. She also reported that self-evaluation and
clinical assessment were poor screening tools in comparison
to breath test devices.
Roadside surveys, although widely used for assessing the
level of driving after drinking, have been rarely used for
pedestrians.
Foss et al (1999) studied the BACs of a sample of
students at the University of North Carolina returning home
to campus accommodation between 10pm and 3am. They reported
that, out of 1,790 students who provided a BAC sample, over
one quarter had been drinking and 10% had BACs of 100mg/100ml
or over. The incidence of a positive BAC as well as the
incidences of BACs above 80 and 150mg/100ml were equally
common above and below the legal drinking age (21 years).
This report covers the results of three separate but related
studies:
- A controlled study of the role of alcohol in fatal adult
pedestrian accidents in the West Midlands
- A controlled study of the role of alcohol in adult
pedestrian accidents in Cardiff
- A footpath survey in Cardiff.
The method, results and discussion are presented separately
for each study. The conclusions, references and appendices
are provided towards the end of the report.
Chapter 2 -
A Controlled Study of the Role of Alcohol in Fatal Adult
Pedestrian Accidents in the West Midlands
2.1
Introduction
The aims of the study were:
- To undertake a case-controlled study of the role of
alcohol in fatal adult pedestrian accidents in the West
Midlands
- To compare and contrast the results with a study
undertaken in the same area in the1970s.
2.2 Method
2.2.1 The
Accident Group
If an adult road user dies within 12 hours of a road traffic
accident, then it is normal for a sample of blood to be taken
at the post-mortem and sent to the local Home Office Forensic
Laboratory or hospital for analysis of its alcohol content.
Permission
was sought and obtained from all but one of the HM Coroners
covering the former West Midlands Metropolitan County
[ 1 ] to examine their
files and extract the relevant data.
The accident sample was taken from all fatal adult (16 years
or over) pedestrian accidents that occurred within the study
area over the period 1 January 1990 to 31 December 1996.Data
were gathered at regular intervals throughout the study. The
time gap between the date of the accident and the date of the
inquest varied considerably. In particular, an inquest
involving a road traffic accident cannot be completed until
any trial for a motoring offence has been held.
2.2.2 The
Control Group
The control sample was matched in terms of the location of
fatal accident, time of day, day of week and the sex of the
pedestrian. As the research team were not aware of the
existence of a fatality until the coroners records were
examined, all accidents were controlled on the same day of
the week in the month of the anniversary of the accident.
Details of the day of the week, time and location of the
accident together with the age and sex of the pedestrian were
sent to the interview team. The interviewers were never told
the BAC of the victim. Within the specified month, they
visited the locations of the accidents at the same time and
on the same day of the week. They then attempted to interview
six pedestrians who passed the location of the accident and
who were of the same sex and approximate age as the injured
pedestrian. The interviews started 15 minutes prior to the
given time and lasted for a maximum of 30 minutes or until
six pedestrians had been interviewed.
The questionnaire (Appendix 1) consisted of 19 questions
covering the reasons for the journey that day and the
frequency of doing the journey, whether or not they could
have used a car for the journey, demographic data and
information on their drinking habits. At the end of the
interview, they were requested to provide a breath sample.
Lion Laboratories SD400s were used to measure BAC.
Each interviewer wore an identity badge (with a photograph)
whilst working. Each pedestrian was given a thank you sheet
containing a contact number for further information about the
study.
The West Midlands Police were informed that the study was in
progress but no police presence was either requested or
provided at the interview sites.
2.3 Results
2.3.1 Fatal
Accident Sample
2.3.1.1 Availability of alcohol data
Of the 204 fatalities in this sample who died within 12 hours
of the accident, blood alcohol concentrations (BACs) were
known for 134 (66%) pedestrians.
Tables 2.1 and 2.2 compare tested and non-tested pedestrians
in terms of some of the factors likely to influence the
presence of alcohol. It may be seen clearly that pedestrians
were more likely to be tested for alcohol if they were male
and their accident occurred late at night (22:0007:59). The
relationship with age was of borderline significance, there
being some tendency for a lower testing rate for the older
age groups. The rate of testing appeared independent of age,
day of the week, month and year (for the period 1990-1996).
|
|
|
|
|
|
|
Table 2.1: Comparison of the age and sex
distributions of tested and non-tested pedestrians
|
|
Variable
|
|
N
|
Yes (%)
|
Tested No (%)
|
p*
|
|
Age Groups#
|
16-29
|
33
|
60.6
|
39.4
|
0.057
|
|
|
30-54
|
32
|
87.5
|
12.5
|
|
|
|
55-69
|
40
|
62.5
|
37.5
|
|
|
|
70-79
|
50
|
66.0
|
34.0
|
|
|
|
80+
|
48
|
56.3
|
43.8
|
|
|
|
|
Sex
|
Male
|
125
|
80.0
|
20.0
|
0.000
|
|
|
Female
|
79
|
43.0
|
57.0
|
|
|
* Significance level using chi-square
# The age of one pedestrian was unknown
|
2.3.1.2 BAC distribution
The BAC distribution by sex is shown in Table 2.3. Overall,
one third of pedestrians had been drinking
(BAC>9mg/100ml). A significantly higher proportion of
males (40%) than females (12%) had been drinking and their
BACs were higher. The highest recorded BAC for a female
pedestrian was 69mg/100ml. Thus no female pedestrian but over
a third of the male pedestrians tested had a BAC above the
legal limit for drivers (Table 2.4). However, the numbers
were small; only four of the 34 female pedestrians who were
tested had been drinking.
|
Table 2.2: Comparison of the date/times of accidents
to tested and non-tested pedestrians
|
|
Variable
|
|
Yes
|
Tested No
|
p*
|
|
Hour of Day
|
08:00-09:59
|
43.8
|
56.3
|
0.009
|
|
|
10:00-11:59
|
50.0
|
50.0
|
|
|
|
12:00-13:59
|
60.0
|
40.0
|
|
|
|
14:00-15:59
|
68.4
|
31.6
|
|
|
|
16:00-17:59
|
50.0
|
50.0
|
|
|
|
18:00-19:59
|
75.0
|
25.0
|
|
|
|
20:00-21:59
|
76.2
|
23.8
|
|
|
|
22:00-07:59
|
85.4
|
14.6
|
|
|
|
|
|
|
|
|
Day night
|
8am-10pm
|
60.2
|
39.8
|
0.003
|
|
|
10pm-8am
|
85.4
|
14.6
|
|
|
|
|
|
|
|
|
Day of week
|
Sunday
|
89.7
|
10.3
|
0.056
|
|
|
Monday
|
63.0
|
37.0
|
|
|
|
Tuesday
|
66.7
|
33.3
|
|
|
|
Wednesday
|
56.3
|
43.8
|
|
|
|
Thursday
|
68.8
|
31.3
|
|
|
|
Friday
|
54.4
|
48.6
|
|
|
|
Saturday
|
68.0
|
32.0
|
|
|
|
|
|
|
|
|
Weekend
|
Saturday-Sunday
|
79.6
|
20.4
|
0.012
|
|
|
Monday-Friday
|
60.4
|
39.6
|
|
|
|
|
|
|
|
|
Month
|
Jan
|
57.1
|
42.9
|
|
|
|
Feb
|
62.5
|
37.5
|
|
|
|
Mar
|
71.4
|
28.6
|
|
|
|
Apr
|
75.0
|
25.0
|
|
|
|
May
|
84.6
|
15.4
|
|
|
|
Jun
|
61.5
|
38.5
|
|
|
|
Jul
|
83.3
|
16.7
|
|
|
|
Aug
|
46.2
|
53.8
|
|
|
|
Sept
|
50.0
|
50.0
|
|
|
|
Oct
|
51.7
|
48.3
|
|
|
|
Nov
|
81.0
|
19.0
|
|
|
|
Dec
|
67.9
|
32.1
|
|
|
|
|
|
|
|
|
Quarter
|
Jan-Mar
|
63.6
|
36.4
|
0.555
|
|
|
Apr-Jun
|
73.8
|
26.2
|
|
|
|
Jul-Sep
|
59.0
|
41.0
|
|
|
|
Oct-Dec
|
65.4
|
34.6
|
|
|
|
|
|
|
|
|
Year
|
1990
|
69.2
|
30.8
|
0.596
|
|
|
1991
|
60.9
|
39.1
|
|
|
|
1992
|
67.6
|
32.4
|
|
|
|
1993
|
55.9
|
44.1
|
|
|
|
1994
|
74.2
|
25.8
|
|
|
|
1995
|
81.8
|
18.2
|
|
|
|
1996
|
60.0
|
40.0
|
|
|
* Significance level calculated using
chi-square
|
|
Table 2.3: BAC distributions by sex
|
|
Sex
|
N
|
BAC (mg/100ml)
|
|
|
|
0-9
|
10-50
|
51-100
|
101-150
|
151-200
|
201-250
|
251-300
|
301+
|
Total
|
|
|
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
|
Male
|
100
|
60.0
|
4.0
|
4.0
|
3.0
|
7.0
|
6.0
|
7.0
|
9.0
|
100
|
|
Female
|
34
|
88.2
|
5.9
|
5.9
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
100
|
|
Total
|
134
|
67.2
|
4.5
|
4.5
|
2.2
|
5.2
|
4.5
|
5.2
|
6.7
|
100
|
|
Using the two-sample Kolmogorov-Smirnov
Test2 , z = 1.813; 0.003.
|
|
Table 2.4: Percentage of male and female pedestrians
with BACs above 80mg/100ml
|
|
Sex
|
N
|
BAC>80
|
Mg/100ml
|
|
|
|
|
Yes %
|
No %
|
|
P*
|
|
Male
|
100
|
36
|
64
|
100
|
0.000
|
|
Female
|
34
|
0
|
100
|
100
|
|
|
Total
|
134
|
36.9
|
73.1
|
100
|
|
|
* Chi-square = 16.736 for df 1
|
2 .3.1.3 BAC by age
Significant differences in the BAC distributions by age group
were observed (Table 2.5). The highest incidence of drinking
pedestrians was amongst the 3054 year old age group where
two-thirds had been drinking. The lowest incidences were in
the 7079 and 80 and over age groups where less than 10% had
been drinking.
|
Table 2.5: BAC distribution by age
group
|
|
Age Groups
|
N
|
BAC (mg/100ml)
|
|
|
|
0-9
|
10-50
|
51.100
|
101-150
|
151-200
|
201-250
|
251- 300
|
301+
|
Total
|
|
|
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
%
|
|
16-29
|
20
|
60.0
|
5.0
|
0.0
|
0.0
|
10.0
|
5.0
|
5.0
|
15.0
|
100
|
|
30-54
|
28
|
32.1
|
0.0
|
3.6
|
7.1
|
14.3
|
14.3
|
10.7
|
17.9
|
100
|
|
55-69
|
25
|
56.0
|
8.0
|
8.0
|
4.0
|
4.0
|
4.0
|
12.0
|
4.0
|
100
|
|
70-79
|
33
|
90.9
|
6.1
|
3.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
100
|
|
80+
|
27
|
92.6
|
0.0
|
7.4
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
100
|
|
Total
|
133
|
67.7
|
3.8
|
4.5
|
2.3
|
5.3
|
4.5
|
5.3
|
6.8
|
100
|
|
Using the Kruskal-Wallis test3 ,
chi-square = 29.998 for df 4; p = 0.000
|
Figure 1 shows the interaction between age and sex in terms
of BAC. The high proportion of drinking pedestrians in the
3054 age group was entirely composed of men. Three-quarters
of the men (76%) in this age group had been drinking. The
proportions of male drinking pedestrians in the 1629 and 5569
age groups were comparable although the BACs of the younger
age group tended to be slightly higher (Table 2.5).
No female pedestrians below the age of 55 had been drinking.
Above that age, the proportions of male and female drinking
pedestrians were comparable.
2.3.1.4 BAC by time of day
Drinking pedestrians are predominantly a night-time
phenomenon. Between 10pm and 8am, two-thirds of pedestrians
had been drinking and over 50% had BACs in excess
of150mg/100ml. By contrast, around breakfast time (8am10am)
and at lunchtime (12 noon2pm), none of the pedestrians tested
had been drinking (Table 2.6).
|
Figure 1: Percentage of drinking
pedestrians by age and sex
|
|
|
Chapter
3 - A Controlled Study of the Role of Alcohol in Adult
Pedestrian Accidents in Cardiff
3.1
Introduction
The aims of the study were:
- To undertake a case-controlled study of the role of
alcohol in adult non-fatal pedestrian accidents in Cardiff
- To gain some understanding of the circumstances
surrounding the accident including the drinking behaviour
of the pedestrian immediately prior to the accident
- To undertake a brief study of the effects of a major
sporting event (Rugby World Cup) and Millennium Eve upon
the incidence of alcohol in pedestrian accidents in
Cardiff.
3.2
Method
3.2.1
Injury Data
The sample consisted of all adult (16 years and over)
pedestrian casualties taken to the Emergency Unit of
Cardiff Royal Infirmary during the period 3 May 1997 to 22
February1999. The Infirmary serves the whole of Cardiff and
its immediate surrounding area and treats around 100,000
people (adults and children) per year. The neighbouring
casualty units are at Swansea Royal Infirmary (to the West)
and the Royal Gwent Hospital Newport (to the East).
The Emergency Unit was supplied with two Lion Alcolmeter
SD400s for measuring breath alcohol levels. Training was
provided to one of the consultants who then trained the
appropriate nursing and medical staff within his Unit.
A combined information and consent form was designed and
approved (Appendix 2). It outlined the nature of the study,
provided a contact telephone number in case of query, and
noted the hospital admissions number, date and time of
admission, age and sex of the pedestrian and, where known,
their BAC and the location of their accident.
Pedestrians were asked to co-operate in the study by the
nursing staff during the triage process. Nursing staff made
the clinical judgement as to whether the patient was in an
appropriate state to provide informed consent to taking
part in the study.
Where the patients provided informed consent, they signed
the consent form. If they also agreed to being interviewed
at a later date, they ticked the appropriate box on the
form and a note of their name, address and telephone number
was made.
A member of the research team visited the hospital at
monthly intervals to collect the completed consent forms.
At that time, the records of all adult pedestrian
casualties taken to the Unit since the previous visit were
extracted manually and brief details taken of the date,
time and location of the accident and the age and sex of
the pedestrian.
Those patients who agreed to a subsequent home interview
were sent a letter saying that a member of the interview
team would be contacting them. The initial contact was
normally made by telephone and a mutually convenient date
and time arranged. Care was taken to ensure that no contact
was made until a week or more after the patient had been
discharged from hospital.
3.2.2
Control Data
To undertake the collection of control data, the date, time
and location of the accident and the age, sex and BAC of
the injured pedestrian were required.
Details of the day of the week, time and location of the
accident together with the age and sex of the pedestrian
were sent monthly to the interview team. The interviewers
were never told the BAC of the injured pedestrian. Within
the next four weeks, they visited the locations of the
accidents at the same time and on the same day of the week.
They then attempted to interview six pedestrians who passed
the location of the accident and who were of the same sex
and approximate age as the injured pedestrian. The
interviews started15 minutes prior to the given time and
lasted for a maximum of 30 minutes or until six pedestrians
had been interviewed.
The interviewers were not recruited until shortly after the
data collection had started. Control interviews for those
accidents occurring during the first two months of the
study were therefore undertaken a year later.
The questionnaire (Appendix 1) consisted of 19 questions
covering the reasons for the journey that day and the
frequency of doing the journey, whether or not they could
have used a car for the journey, demographic data and
information on their drinking habits. At the end of the
interview, they were requested to provide a breath sample.
Again, Lion Laboratories SD400s were used to measure BAC.
Each interviewer wore an identity badge (with a photograph)
whilst working. Each pedestrian was given a thank you sheet
containing a contact number for further information about
the study.
The South Wales Police were informed that the study was in
progress but no police presence was either requested or
provided at the interview sites.
3.3
Results
3.3.1
Injury
During the period 3 May 1997 to 22 February 1999, a total
of 360 adult pedestrians were taken to the Emergency Unit.
The average monthly intake was 16the range was 229patients.
Lowest throughputs tended to be in the summer months; the
highest in the winter months. Figure 3.1 shows the
percentage distribution by month.
3.3.2
Incidence of Testing
Overall, blood alcohol concentrations (BACs) were known for
a quarter (24.5%) of all pedestrians. As Table 3.1 shows,
the proportion of tested pedestrians appeared to be related
to three factors. The proportion was higher in May the
first month of the study. Partly as a consequence, it was
also higher in the first year of the study (33%). Women
(18%) were less likely to be tested than men (29%).
|
Figure 3.1: Percentage distribution of
intake by month
|
|
|
Chapter 4 - A Footpath Survey in Cardiff
4.1
Introduction
The prime aims of the study were to:
- Determine the incidence of drinking pedestrians and
the effects of situational and temporal factors
- Determine the demographic and drinking
characteristics of drinking pedestrians
- Determine the proportion of potential drivers amongst
the drinking pedestrians.
4.2
Method
A feasibility study (Clayton (1998)) was undertaken in
Cardiff in July to August 1998. On the basis of that
experience, it was agreed that two studies should be
undertaken - in Birmingham and Cardiff. The Birmingham
study was abandoned after three months as a large
proportion of the pedestrians who were approached
declined to co-operate. Others offered verbal and, in
some cases, physical abuse to the interviewers. The
remainder of this report refers solely to the Cardiff
study.
Five sites were chosen on the basis of reasonable
night-time pedestrian flow (Table 4.1 and Figure 4.1).
All were relatively close to the city centre and
attempted to reflect the variety of social environments
within the city.
|
Table 4.1: Interview sites
|
|
Ref. No.
|
Location
|
|
1
|
Albany Road at Angus Street
|
|
2
|
City Road at The Parade
|
|
3
|
Crwys Road at Woodville Road
|
|
4
|
Cowbridge Road at Wyndham Crescent
|
|
5
|
Newport Road at Fitzalan Place
|
The
study was designed to run for eight months (October 1998
to May 1999) and cover all nights of the week from 17:00
to 23:59. With five sites, seven days, seven one-hour
time periods and eight months, there were a potential
1,960 combinations. It was agreed that a 10% sample
should be undertaken using 196 sessions. This schedule
resulted in each day being sampled on 28 occasions, each
hour-slot 28 times, each month 24-25 times and each site
on around 38 occasions. The interview schedule is given
elsewhere (Appendix 4). The interviews were all
undertaken by the same husband and wife team. The
questionnaire (Appendix 1) was the same as that used to
collect control data in other parts of the study. It
consisted of 19 questions and was designed to be
completed in a few minutes. At the end of the interview,
respondents were asked to provide a breath sample. The
instrument used was the Lion Alcolmeter SD400 which
provides a quantitative measure of BAC.
|
Figure 4.1: Map of Cardiff showing
location of sites. NB: This map is subject to
Crown Copyright and is licensed solely for the
personal use of the senior author. A separate
licence is required for any other
purpose.
|
|
|
Chapter 5 - Conclusions
The number of adult pedestrian fatalities in Britain is
declining. However, the BAC distribution of those
fatalities who were tested for alcohol showed no
significant change since the earlier study undertaken
in the mid 1970s.
Amongst injured pedestrians (in Cardiff), the overall
proportion of pedestrians who had been drinking was
higher than amongst the fatalities in the West
Midlands. This result may be a function of the
different age structures of the two samples; the fatal
sample being generally older.
The results of previous studies augmented by some
evidence from the present studies of both fatal and
non-fatal casualties suggest that, for males at least,
the risk of accident-involvement begins to increase
dramatically at levels above 150mg/100ml.
The footpath study was the first of its kind in
Britain. It provides a baseline of information on the
incidence of non-accident-involved pedestrians in a
city centre during the hours 5pm to midnight. Time of
day is the most powerful predictor of whether or not a
pedestrian has been drinking and the likely BAC.
Approaching midnight, nearly three-quarters of those
tested had been drinking.
The effect of countermeasures against drinking and
driving was clear. Over a quarter of pedestrians who
had been drinking had taken the conscious decision not
to drive to and from their drinking venue.
It would be useful to monitor the drinking behaviour of
pedestrians at intervals to determine whether their
incidence is increasing. This study has shown that a
footpath survey can provide useful information.
Countermeasures to the drinking pedestrian may be best
based upon the general public health message of the
dangers of excessive alcohol consumption.
Alternatively, if appropriate and realistic,
consideration may be given to designing countermeasures
where the potential for contact between pedestrians and
vehicles is minimised.
Chapter 6 - References
Alcohol Concern (2000) Women's drinking: UK facts and
figures. www.alcoholconcern.org.uk/servlets/doc/785
Alexander, K, Cave, T and J Lyttle (1990) The role of
alcohol and age in predisposing pedestrian accidents.
Melbourne, Victoria: Road Traffic Authority, Pedestrian
Accident Project Report No. 6.
Allsop, R E (1966) Alcohol and road accidents.
Harmondsworth: Road Research Laboratory RRL Report No.
6.
Blomberg, R D, Preusser, D F, Hale, A and L G Ulmer
(1979) A comparison of alcohol involvement in
pedestrians and pedestrian casualties. Washington DC:
US Department of Transportation, National Highway
Safety Traffic Administration.
Cardiff City Council (2000) Key information.
http://www.cardiff.gov.uk/facts/english/FactsFrame.htm
Clayton, A B, Booth, A C and P E McCarthy (1977) A
controlled study of the role of alcohol in fatal adult
pedestrian accidents. Crowthorne: TRRL Report SR332.
Department of the Environment, Transport and the
Regions (1999) Road Accidents Great Britain 1998.
London: DETR.
Department of Transport (1989) Road Accidents Great
Britain 1988. London: DOT.
Everest, J T, Banks, S, Hewer, P A and J Mineiro (1991)
Drinking behaviour and breath alcohol concentrations of
road accident casualties. Crowthorne: TRRL Report RR
311.
Fontaine, H and Y Gourlet (1997). Alcohol involvement
in pedestrian accidents in France. In Mercier-Guyon, C
(Ed) Proceedings of the 14th International
Conference on Alcohol, Drugs and Traffic Safety,
Annecy, September 1997, 155-160.
Foss, R, Marchetti, L, Holladay, K and K Scholla (1999)
BACs of university students returning home at night.
Washington, DC: Paper presented at
78thAnnual Meeting of the Transportation
Research Board, 13 January 1999.
Haddon, W, Valien, P, McCarroll, J and C J Umberger
(1961) A controlled investigation of adult pedestrians
fatally injured by motor vehicles in Manhattan.
Journal of Chronic Diseases, 14, 655-678.
Holubowycz, O T (1995). Age, sex and blood alcohol
concentration of killed and injured pedestrians.
Accident Analysis and Prevention, 27, 417-422.
Peden, M M (1997) Adult pedestrian traffic trauma in
Cape Town with special reference to the role of
alcohol. PhD thesis, Faculty of Medicine, University of
Cape Town.
Chapter 7 - List of Appendices
Appendix 1
Pedestrian Questionnaire
Appendix 2
Information sheet and consent form - Cardiff Royal
Infirmary
Appendix 3
Alcohol and Adult Pedestrian Survey
Appendix 4
Interview Schedule - footpath survey