null  null
Injuries in children aged
0–14 years and inequalities
A report prepared for the Health Development Agency
Elizabeth Towner, Therese Dowswell, Gail Errington, Matthew Burkes, John Towner
Community Child Health, Department of Child Health
University of Newcastle upon Tyne
This document is also published on the
Health Development Agency website at
www.hda.nhs.uk
Acknowledgements
This research study has been funded by the Health Development Agency, London, and this support is gratefully acknowledged.
We wish to thank Mr Antony Morgan for his comments and support.
We are also grateful to Mrs Penny Giddings for retrieving the publications cited, and Mrs Lilian Simpson and Mrs Ann Pattison
for typing and preparing the report.
Copies of this publication are available to download from the HDA website (www.hda.nhs.uk).
Health Development Agency
Holborn Gate
330 High Holborn
London
WC1V 7BA
Email:
[email protected]
ISBN 1-84279-339-X
© Health Development Agency 2005
About the Health Development Agency
The Health Development Agency (www.hda.nhs.uk) is the national authority and information resource on what works to
improve people’s health and reduce health inequalities in England. It gathers evidence and produces advice for policy makers,
professionals and practitioners, working alongside them to get evidence into practice.
ii
Contents
Summary
1
A: Setting the scene
3
A1 Introduction
3
A2 Definitions
4
A3 Methods
5
A4 National data sources on injury and coverage of inequalities
6
A5 Key points
9
B: Why are inequalities associated with injury risk? Evidence from observational studies
10
B1 Introduction
10
B2 Age
12
B3 Gender
13
B4 Social and economic factors
15
B5 Culture and ethnicity
17
B6 Place
19
B7 Vulnerable groups
21
B8 Key points
23
C: Have injury intervention studies addressed inequalities?
24
C1 Introduction
24
C2 Pedestrian injuries
25
C3 Car occupant injuries: child restraints
29
C4 Bicycle injuries
31
C5 Injuries in the home
34
C6 Injuries in the leisure environment
41
C7 Mass media and training interventions
43
C8 Community-based interventions
44
C9 Key points from all the intervention studies
47
D: Discussion and recommendations
48
D1 Discussion
48
D2 Recommendations for practitioners, researchers and policy makers
51
References
53
References for Appendix Tables C2–C8
58
Contents
iii
Summary
All children are exposed to injury as part of their everyday
lives, but the burden is not evenly spread: injuries
disproportionately affect some children more than others.
Great variations occur in injury mortality and morbidity, which
reflect a child’s age, gender, socio-economic group, cultural
and/or ethnic group, and where they live.
In this report we attempt to describe and seek to understand
these variations. For each inequality in turn (age, gender,
social and economic factors, culture and ethnicity, place)
we question why it is associated with injury risk. We then
examine evaluated intervention studies identified in an
earlier review (Towner et al 2001; www.hda-online.org.uk/
downloads/pdfs/prevent_injuries.pdf) and ask whether injury
prevention studies have addressed these inequalities. The
information is presented in tabular and narrative form for
different injury types.
Key points identified in our description of inequalities are:
• There are great variations in injury mortality and morbidity,
reflecting children’s age, gender, socio-economic group,
cultural or ethnic group, and location
• These variations suggest there is considerable scope for
improvement and intervention
• A number of reports and policy initiatives are relevant
to this report: the Independent inquiry into inequalities
in health (Acheson, 1998); Saving lives: our healthier
nation (Secretary of State for Health, 1999); the report of
the Accidental Injury Task Force (Department of Health,
2002b); and the Economic and Social Research Council
(ESRC) Health Variations Programme
• Children as a group are particularly vulnerable to injury:
they have little say in how environments are planned, and
in recent years child poverty has increased
• National injury data systems provide good information
on age and gender and some information on social and
economic factors, and on place
Summary
• National injury data systems do not routinely provide
information on ethnic groups or on vulnerable groups
such as children in homeless families or disabled children.
Key points identified in relation to the causes of inequalities
are:
• The factors associated with injury inequalities are multifaceted and inter-related; the causal pathways linking
these factors to injury events remain uncertain
• Three tiers of factors need to be considered to clarify the
inter-relationships: (1) proximate tier, immediate conditions
that result in exposure to hazard; (2) intermediate tier,
eg childcare practices; (3) ultimate tier, the wider social,
economic, political and cultural processes
• These tiers fit into the concept of structure and agency;
the interplay of structure and agency is worked out in
specific geographic locations
• More direct causes of injury include exposure to hazards,
the ability of parents, carers and communities to protect
children, and children’s ability to manage hazards
• Greater knowledge about disparities between groups and
factors leading to increased risk is important in designing
interventions.
Key points relating to whether intervention studies have
addressed inequalities are:
• Few intervention studies explicitly address inequalities
– and still fewer attempt to set out the problem in relation
to that factor, take it into account in the design of the
intervention, and report on whether there has been any
differential impact in relation to that factor
• Age: many interventions target specific age groups of
children, but there are few examples of results comparing
the impact of an intervention on different age groups
• Gender: despite the great differences in injury rates in
boys and girls, few interventions have targeted this factor
1
• Social and economic factors: interventions have
increasingly been targeted at deprived individuals or
groups – the main strategy adopted in interventions is the
provision of safety equipment
• Culture/ethnicity: cultural differences are rarely addressed
(a few interventions have involved the target group in
the design of the intervention), and few studies compare
interventions in different ethnic groups
• Place: when communities have been matched in controlled
interventions, broad demographic variables have been
used and little attention has been paid to the context.
We discuss how definitions need to be improved and how
descriptive and analytical studies and intervention studies
could be improved to address inequalities. We also consider
the role of targeted and universal approaches to injury
prevention.
Recommendations are made that are of relevance to
practitioners, researchers and policy makers.
2
Injuries in children aged 0–14 years and inequalities
A: Setting the scene
A1 Introduction
Unintentional injury is the main cause of death in childhood
in the UK, and a major cause of ill health and disability. All
children are exposed to hazards as part of their everyday
lives as they play, travel around, work, and even (at times)
when they are asleep. The burden of injury is not evenly
spread – injuries disproportionately affect some children more
than others. Great variations occur in injury mortality and
morbidity which reflect a child’s age, gender, socio-economic
group, ethnic group, and where they live.
In this report we attempt to describe and seek to understand
these variations. Injuries are preventable, and the great
variations in injury mortality and morbidity between, for
example, different socio-economic groups suggest that there
is considerable scope for improvement and intervention.
Sections A (variations in injury risk) and B (factors associated
with risk) both relate to the UK context, while section C
(evaluated interventions) is drawn from the international
literature. For information on patterns of child injuries
and inequalities in other countries a special edition of the
journal Injury Control and Safety Promotion (2001), on social
inequalities in injury risks, is a useful source, particularly
for articles related to childhood injuries and inequalities in
Sweden and Germany.
What policy initiatives are relevant to childhood
injury and inequalities?
The Independent inquiry into inequalities in health (Acheson,
1998) reviewed and summarised inequalities in health
in England and identified priority areas for future policy
development. These areas included poverty, income
tax and benefits; education; employment; housing and
environment; mobility, transport and pollution; and
nutrition. Areas were also identified by life-course stage,
Setting the scene
and one important group identified was mothers, children
and families. There was also a focus on ethnic and gender
inequalities. The Independent inquiry into inequalities in
health was an important influence in the development of the
health strategy for England.
Saving lives: our healthier nation (Secretary of State for
Health, 1999) includes accidents as one of four priority areas.
The target set for 2010 was:
‘to reduce the death rate from accidents by at least a fifth
and to reduce the rate of serious injury from accidents by
at least a tenth – saving 12,000 lives’
Children were one of the groups highlighted. The national
strategy aimed to improve the health of everyone, but in
particular aimed to improve the health of the worst-off in
society. The report acknowledged the importance of social,
economic and environmental factors in poor health.
Our Healthier Nation also announced that a Task Force on
Accidental Injury would be set up. This was convened in
November 2000 and its terms of reference were:
‘To advise the Chief Medical Officer on the most important
priorities for action to meet the national target, to develop
an implementation plan and to develop and publicise
a more unified approach to accident prevention across
Government and the NHS’
The Task Force identified immediate priority areas using three
criteria: they are a substantial burden of injury; they represent
areas where marked social disadvantages are apparent; and
some evidence-based interventions are available. For children
aged 0–15 years the priority areas were pedestrian injuries;
fires and thermal injuries; and play and recreation (eg cycling
and swimming) (Department of Health, 2002a).
Also relevant to this review is the work of the Health
Variations Programme. This was established by the ESRC in
3
1996 to focus on the causes of health inequalities in Britain.
Publications stemming from this review have been used
throughout the report.
Why is inequality in childhood particularly
important?
In the last two decades of the twentieth century Britain
became richer and healthier, but inequalities in income
and health widened markedly. The benefits of increasing
prosperity were not equally shared, and there was a sharp
rise in the proportion of children living in households below
the poverty line (below 50% of average income after housing
costs). An increase in child poverty has profound implications
for health inequalities in both current and future generations
(Graham, 1999).
•
•
•
•
•
•
•
Children as a group are particularly vulnerable to injury
because they live in a world in which they have little power
or control. They live in environments constructed by and for
adults, and seldom have any input into how places, even
places such as playgrounds, are designed.
The inequalities we examine are age, gender, social and
economic factors, cultural and ethnic groups, and place.
Many of these factors are, however, inter-related and this
needs to be borne in mind throughout the report. Below, we
define how we have used these terms.
This report
This report is divided into four sections. In this section we set
out the definitions (A2) and methods employed in the review
(A3), and examine national sources of injury data for their
coverage of inequalities (A4).
Section B uses evidence from observational studies to
investigate the causes of injury. For each inequality in turn
(age, gender, social and economic factors, culture/ethnicity,
and place) we question why it is associated with injury risk.
Section C uses evidence from evaluated intervention studies
and asks whether injury intervention studies have addressed
inequalities. The information is presented in tabular and
narrative form for different injury types, eg pedestrian injuries
(C2), car occupant injuries (C3) etc.
At the end of each section A–C, the key points are
summarised. In Section D these key points are used as the
basis for discussion. Recommendations are made that are of
relevance to practitioners, researchers and policy makers.
A2 Definitions
In this report we examine childhood injuries and inequalities.
This is the focus of both Saving lives: our healthier nation
(Secretary of State for Health, 1999) and the Accidental Injury
Task Force (Department of Health, 2002a,b).
4
Unintentional injury can be defined as ‘injury occurring
as a result of an unplanned and unexpected event which
occurs at a specific time from an external cause’. The Office
for National Statistics (ONS) uses definitions based on the
International Classification of Diseases, ICD-9 (www.cdc.gov/
nchs/about/otheract/icd9/abticd9.htm) and ICD-10
(www.who.ict/whosis/icd10/). The types of injury included
are:
Transport – rail, road, air, water
Poisoning
Falls
Fire, flames and smoke
Natural and environmental factors
Submersion, suffocation and foreign bodies
Other accidents.
• Age: we concentrate in particular on the 0–14 age group
but on occasions data are given for 0–15 year olds. Where
possible, subgroups of 0–4, 5–9 and 10–14 years are used.
• Gender: we use the term ‘gender’ throughout the report.
• Social and economic factors: in section A4 on national
data sources, social class has been derived from the
classification of children’s parental occupation. Wider
definitions of social and economic factors are discussed in
section B4.
• Culture/ethnicity: in section A4 on national data sources,
definitions of ethnic groups as used in labour force surveys
are used for the UK (Scott et al 2001). In section B5 we
discuss wider definitions of ‘cultural group’ and ‘ethnic
group’.
• Place: can be considered as a portion of geographic
space such as a ward, enumeration district etc, and this
definition is mainly used in section A4. In section B6 a
broader definition is discussed – ‘space’ relating to physical
attributes and a site or area to which people attach certain
meanings.
Sections A4 and B focus on conditions in the UK. Some
statistics are collected and reported for the whole of the UK,
eg fire statistics; road traffic accidents are reported for Great
Britain (England, Wales and Scotland); and mortality data
for England and Wales. Section C on intervention studies
describes the international literature but is limited to studies
published in English.
Injuries in children aged 0–14 years and inequalities
A3 Methods
The aim of this report is to address the issue that
unintentional injuries disproportionately affect some
children more than others, and to consider the scope for
improvement and intervention.
The objectives are to:
• Describe inequalities related to age, gender, social and
economic factors, culture and ethnicity and place, using
national data sources relating to unintentional injury
• Seek to understand why inequalities are associated with
injury risk
• Examine whether, and to what extent, intervention studies
have taken inequalities into account in the selection of
injury target groups or settings, explicitly in the ways
interventions have been designed or results and outcomes
have been reported.
These three objectives are examined in sections A4, B and
C. The methods used in these three sections are described
below.
Description of inequalities and childhood injury
in national data sources (section A4)
The primary source for this section is the report of the
Measuring and Monitoring Injury Group to the Accidental
Injury Task Force (Department of Health, 2002b). In the
course of preparing this review we collected information
from different national data sources (eg ONS), but the Task
Force Report succinctly summarises these data and we
have consequently used this as the main source. For each
of the inequalities in the review – age, gender, social and
economic factors, cultural and ethnic group and place – we
provide illustrations of the varying patterns in relation to
unintentional injury. This is not a comprehensive review, but
is intended to show what information can be gleaned from
the different national data sources.
Explanation of why inequalities are associated
with injury risk (section B)
In this section we use a framework described by Thomson
et al (2001) in a review examining ethnicity and injury. This
suggests that the direct causes of injury are exposure to
hazards, parents, carers and communities’ ability to protect
children, and children’s capacity to manage hazards. We
apply this framework to the five inequality areas. The
literature for this section is primarily drawn from large cohort
studies and case–control studies examining the characteristics
Setting the scene
of injured children. Relevant cohort and case–control studies
were identified by searching electronic databases (eg
MEDLINE, Web of Science) and by hand-searching relevant
journals (eg Injury Prevention). Studies are included which
were carried out in the UK context. Additional literature
on injury risk was identified from the same sources and
is included where it sheds further light on the causes of
injuries related to inequalities. In view of the varied and
potentially large volume of literature in this area, the search
and reviewing strategies adopted for this section were not
systematic, instead illustrative examples of the literature
have been provided. The search strategies used mean that
many of the studies included were drawn from the medical/
epidemiological literature which is predominantly quantitative
in nature. This section does not explicitly identify qualitative
studies or examine the contribution they can make to
understanding inequalities in childhood injury. This would
be a valuable future area of work to extend the scope of the
review.
Examination of whether, and to what extent,
evaluated intervention studies have taken
inequalities into account (section C)
Section C uses as its source a systematic review of the
literature which attempted to answer the question ‘How
effective are health promotion interventions in preventing
unintentional injuries in childhood and young adolescents?’
(Towner et al 2001). It includes studies published between
1975 and 2000 and builds on two earlier reviews published
in 1993 and 1996. This review has been updated with studies
published in 2001.
In the 2001 systematic review, relevant literature was
identified by a variety of means. Computerised databases
including MEDLINE, BIDS (and more recently the Web of
Science), Excerpta Medica, and more specialised sources
such as the Transport and Road Laboratory databases were
searched. A full list of databases searched and search terms
used is available on request. This electronic search was
supplemented by hand-searching a number of relevant
journals such as Accident Analysis and Prevention and
Injury Prevention along with the reference lists of relevant
published articles and books. In addition, relevant informants
(researchers and specialists in the area of child injury
prevention) were consulted. The criteria for inclusion of
studies were that they should:
• Be written in English and published between 1975 and
2000
• Relate to the prevention of unintentional injuries (solely or
in part)
5
• Have a target population including children under 15 and
report results for this group
• Describe either a primary intervention measure to prevent
accidents occurring or a secondary measure to prevent or
reduce the severity of injuries
• Have been evaluated using some measure of outcome
or impact, eg changes in injury mortality or morbidity,
changes in observed or reported behaviour, environmental
change or hazard removal, or changes in knowledge or
attitudes.
Violence prevention studies were excluded except where they
were combined with unintentional injury studies.
All studies were read and reviewed independently by two
reviewers. Where statistical advice or other specialised
knowledge was required a third reviewer was consulted.
A standardised data extraction form was devised and used
to record details from each study (available on request).
Details recorded included the date and place of the study,
the injury target group, and the aim, content and setting
of the intervention. Where interventions had been targeted
at socially or economically disadvantaged groups this
was noted. In addition, details about the evaluation were
recorded. This included a brief description of the methods
used (study design, sample size, data collection methods,
outcome and impact measures), and a note was made of
strengths and weaknesses of the evaluation. The process
of assessing the quality of the evidence of the various
studies was informed by the NHS Centre for Reviews and
Dissemination guidelines on carrying out systematic literature
reviews (Arblaster et al 1995). The reviewers reached a
consensus decision on the quality of the evidence. Each
study was graded on a five-point scale ranging from weak to
good (weak, reasonable/weak, reasonable, reasonable/good,
good).
Key results were recorded and a consensus decision made
about the effectiveness of the intervention. Where results
were reported for subgroups in the sample (eg for children
attending schools in higher- or lower-income areas), these
were recorded.
Details from the data extraction forms were used to devise
summary tables for each study. Studies where the evidence
was rated as weak were excluded at this stage.
The 2001 systematic review has been supplemented by
studies published in 2001, using the methods outlined above.
This new database of 164 studies was then re-analysed to
see whether inequalities (age, gender, social and economic
6
groups, culture/ethnicity and place) have played any part in
the selection of injury target groups and settings, the design
of the interventions, and the results and outcomes reported.
Our research team extracted this additional information
from the original articles and amended the entries in the
tables. The additional findings were discussed by the team
and formed the basis for analysis of the tables, as presented
in section C. The tables themselves are available online
at www.hda.nhs.uk/evidence, along with a full list of the
studies included.
A4 National data sources on injury and
coverage of inequalities
There are six national data systems in the UK, or parts of the
UK, that are relevant to this review:
•
•
•
•
Mortality statistics – ONS
Road traffic accidents – STATS19 data
Hospital admissions data – Hospital Episode Statistics
A&E attendance statistics – Home and Leisure Accident
Surveillance Systems (HASS/LASS)
• Fire statistics
• Self-reported injury – Health Survey for England.
Detailed information on the first five of these systems is
available in Appendix 2 of the Measuring and monitoring
injury report to the Accidental Injury Task Force (Department
of Health, 2002b). The types of data collected by the
different data systems are summarised below.
Mortality statistics – ONS
The ONS compiles mortality statistics on unintentional injuries
based on details collected when deaths are certified and
registered. Deaths have been coded using the International
Classification of Disease coding, ICD-9 for the period 1979–
2000 and ICD-10 since 2001. In England and Wales, for each
major type of injury information is available on age, gender,
and cause and place of occurrence of the injury event.
Road traffic accidents – STATS19 data
In Great Britain information on mortality and morbidity
resulting from road traffic accidents, involving vehicles on
public highways, is collected by the police and collated by the
Department for Transport.
Information on three levels of severity is available: fatal,
serious and slight. Data are collected on the location and
circumstances of the accident, the vehicle involved and the
resulting injuries incurred.
Injuries in children aged 0–14 years and inequalities
Hospital admissions data – Hospital Episode
Statistics
The Department of Health collects data on hospital
admissions. Diagnosis and external cause are coded using
the International Classification of Disease coding, and ICD-10
codes have been in use since 1995. Data are collected on
personal, medical and administrative variables.
A&E attendance statistics – Home and Leisure
Accident Surveillance Systems (HASS/ LASS)
The Department for Trade and Industry collects information
about home and leisure accidents in the HASS and LASS
databases. Data are collected from 18 A&E departments
in the UK, representing about 5% of all UK attendances
for home and leisure accidents. Data are collected directly
from patients through an interview and supplemented by
information derived from medical records. Information is
available on the immediate circumstances of the accident,
accident mechanism, location, activity at the time, type of
injury and body part injured, products involved, outcome of
the initial A&E visit and duration of in-patient stay.
Fire statistics
The Office of the Deputy Prime Minister collates and analyses
fire statistics, including injuries from fires, which are collected
by local fire brigades. Records of deaths are checked against
certification from the ONS.
Self-reported injuries – Health Survey for
England
In the Health Survey for England self-reported data on
‘major’ and ‘minor’ accidents were collected for children
aged 2–15 years for the years 1995–97. ‘Major’ accidents
were considered to be those where a hospital was visited or a
doctor consulted, while ‘minor’ accidents were those causing
pain or discomfort for more than 24 hours (Purdon, 1997).
These systems are uncoordinated at present and it is possible
for one injured person to generate sets of unlinked records
in different systems. Some of the systems also achieve
incomplete coverage (Department of Health, 2002b). For
example STATS19 records omit many cyclist and pedestrian
injuries, even when casualties are admitted to hospital
(Department of Health, 2002b).
Two out of these six systems, the Health Survey for England
and HASS/LASS data, are sample surveys intended to provide
nationally representative data.
Table 1 summarises whether these national data systems
can provide information on unintentional injury and age,
gender, socio-economic group, ethnic group and place. All
six systems provide information on the age and gender of the
injured child. Five of the systems provide some information
on place. The ONS mortality data and Hospital Episode
Statistics data include the geographic area of residence
(postcode) of the injured person. STATS19 data has the
geographic area of the road accident, and since 1999 it
has also started to collect the geographic area of residence
(postcode) of the casualty. HASS/LASS data includes the
home address postcode of the casualty, but not of the
accident location if this is outside the home.
In relation to social and economic factors, ONS has done
limited analysis in relation to death registration data (see
below for details). STATS19 and Fire statistics do not collect
information on the socio-economic status of casualties.
HASS/LASS data include information on employment status.
Some studies have used STATS19, Fire statistics and HASS/
LASS data, and examined social and economic factors using
Table 1 National data sources on unintentional injury and coverage of health inequalities
Age
Gender
Social and
economic
factors
Ethnic
group
Place
ONS
ü
ü
(ü)
x
ü
STATS19
ü
ü
(ü)
x
ü
Hospital Episode Statistics
ü
ü
(ü)
x
ü
HASS/LASS
ü
ü
(ü)
x
x
Fire statistics
ü
ü
(ü)
x
ü
Health Survey for England
ü
ü
x
x
x
( ) Can be derived from data.
Setting the scene
7
the residential postcode of accident victims. The Health
Survey for England has some information on social class for
adults 16 and over, but there is no information for children.
None of the systems routinely provides information on ethnic
groups of injured children.
In relation to particular vulnerable groups, eg children living in
homeless families or disabled children, it is impossible to obtain
a national picture in relation to childhood injury (see B7).
Using examples from these data systems we illustrate the
patterns relating to age, gender, social and economic factors,
ethnic group and place. This is not a comprehensive review,
but is intended to give a flavour of the sources available.
Age
Patterns of injury can be identified that reflect children’s
age and stage of development, eg pre-school children are
more likely to have an unintentional injury in the home
environment. School-age children are much more likely to
have injuries in the road environment.
ONS mortality data for England and Wales for the threeyear period 1997–99 show that 1,071 children aged 0–14
years died of an unintentional injury (ICD9 E800-949). Forty
per cent of these deaths were in children aged 0–4 years.
For certain injuries such as cycle or pedestrian injuries, older
children aged 5–14 years predominated, accounting for 99%
of cycle deaths and 78% of pedestrian deaths. In the home
environment pre-school children accounted for 66% of the
deaths from suffocation and foreign bodies and 70% of the
deaths from fire and flames.
STATS19 data for the three-year period 1998–2000 for Great
Britain showed that 84 children were killed as pedal cyclists,
2,219 were seriously injured and 15,889 were slightly injured.
There was a steep age gradient, with older children (aged
10–14 years) having the greater number of casualties and
pre-school children (aged 0–4 years) having the fewest. A
more pronounced age gradient occurred for deaths than for
either serious or slight injuries.
HASS data for the four-year period 1996–99 for children
attending A&E departments as a result of a choking incident
showed that 62% of the incidents involved children aged
0–4 years, 24% children aged 5–9 years, and 15% children
aged 10–14 years. The peak age for choking incidents was
the second year of life, and from this age the number of cases
declined with increasing age (Towner and Errington, 2002).
Gender
The pattern of injury mortality and morbidity is very different
for boys and girls, with boys experiencing more injuries than
8
girls for most injury types. The gender differences in injury
mortality have remained constant over time (Jarvis et al,,
1995).
Of the 1,071 children aged 0–14 years in England and Wales
who died as a result of an injury in the period 1997–99,
66% were male and 34% female. In the road environment,
74% of cyclist deaths were male, as were 63% of pedestrian
deaths and 48% of passenger accidents. For drowning
deaths 81% were male, for falls 71%, for suffocation and
foreign bodies 64%, and for poisoning 61%. Deaths from
fire and flames showed fewer gender differences, with 54%
of fatalities in males.
The Health Survey for England provides evidence of the scale
of injury morbidity. Self-reported data on ‘major’ and ‘minor’
accidents were collected for children aged 2–15 years for the
years 1995–97. Major accident rates (where a hospital was
visited or a doctor consulted) were higher in boys than girls,
31 per 100 per year, compared with 22 per 100 per year.
Minor accidents (other accidents causing pain or discomfort
for more than 24 hours) were reported more by boys: 216
per 100 per year, compared with 144 per 100 for girls
(Purdon, 1997).
Socio-economic group
The decennial supplement of occupational mortality published
by the Office of Population in Censuses and Surveys (now
ONS) provides injury mortality data by occupational class for
children aged 1–15 years in England and Wales for 1979,
1980, 1982 and 1983 (data for 1981 are unavailable because
of an industrial dispute; OPCS, 1988). Direct comparisons with
more recently published data are made difficult by changes
in the reporting of class-based population data. Roberts and
Power (1996) were able to overcome these limitations by
obtaining anonymised records of all child deaths for 1989–92,
and parents’ occupational class.
In 1979–83 the injury death rate for children in social class V
was 3.5 times that in social class I (Roberts and Power, 1996).
For 1989–92 the injury death rate was five times that of
children in social class 1. As a result of the differential decline
in injury death rates, socio-economic mortality differences
increased.
Over this period, motor vehicle accident death rates declined
by 30 and 39% in social classes I and II, compared with 18
and 1% in social classes IV and V. For deaths due to fire and
flames, rates in social classes I and V declined by 28 and 5%,
respectively (Roberts and Power, 1996). The death rates for
children from social class V for fire and flames were 16 times
those for children from the most affluent backgrounds and
Injuries in children aged 0–14 years and inequalities
five times those for pedestrian injury (Roberts, 1997). Death
rates among the ‘unoccupied’ groups were even greater than
for social class V, with children from these backgrounds being
particularly vulnerable (Judge and Benzeval, 1993). Similar
trends have been identified in Scotland using alternative
methods of classifying social groups (Morrison et al 1999).
questionnaire survey of seriously injured pedestrians over
the same period. For the coroners’ data, ethnic origin was
derived from post-mortem descriptions of the body, language
used in interviewing next of kin, etc. For the STATS19 data
ethnicity was derived from victims’ surnames (Thomson et al
2001).
STATS19 data contain information on the postcode where
road traffic accidents take place. Kendrick (1993) has
analysed pedestrian accidents in children in Nottingham
using locally derived deprivation scores. She found a
significantly higher accident rate in deprived areas and a
dose–response relationship between areas of deprivation
and accident rates. Other studies have extracted data from
police files to obtain postcoded home addresses of children:
in an analysis of child pedestrian accidents in Manchester
and Salford (Preston, 1972), and in analysing road traffic
accidents in Newcastle (Walsh et al 1990). In the past few
years STATS19 data have included the casualty’s home
address postcode – when coverage rates are higher it will
be possible routinely to examine the effect of deprivation on
road traffic accidents.
Place
Ethnic groups
None of the national data sources on injury specifically
collects information about the ethnic group of injured
children. Raleigh and Balarajan (1995) noted that information
on ethnic origin had not been previously recorded in health
information systems and population-based statistics. In
particular they point to the paucity of information in relation
to ‘child abuse, accidental injury and levels of disability and
handicap’. As information is lacking about injury in different
ethnic groups at national level, in the following we describe
demographic patterns relating to children from minority
ethnic groups, then consider some small-scale studies of
childhood road accidents.
Estimates of the minority ethnic populations of Great Britain
in 2000 were derived from the Labour Force Survey and
relate to the population living in private households. Four
million people belonged to a minority ethnic group – 7.1%
of the total population. An estimated total of 1,065,000
children aged 0–14 years belonged to minority ethnic groups,
9.9% of the child population, and 90% of these children
were born in the UK (Scott et al 2001).
Three studies have been conducted by Lawson and
colleagues examining road accidents by ethnic origin
(Lawson and Edwards, 1991). These include an examination
of coroners’ records for a five-year period, an analysis of
all road casualties for Birmingham over 12 months, and a
Setting the scene
It is possible to examine OPCS/ONS data on child accident
death rates by health district. One of the earliest studies to
do this in England and Wales was conducted by Avery et al
(1990) who found higher rates of injury in the North and
West of England. Jarvis et al (1995) found similar variations
in their analysis of child accident death rates by regional
health authority for the period 1989–91. The death rate per
100,000 population was 7.2 in England and 8.1 in Wales.
Regions with the highest rates were North Western (9.8)
and Northern (9.5), and those with the lowest were South
Western (5.4) and Oxford (5.9).
Lynam and Harland (1992), in their analysis of STATS19 data
on child pedestrian casualties, found that casualty rates
varied twofold in different administrative areas in the UK.
Rates in major conurbations were 50% higher than the
average for the UK, and the casualty rate for inner London
districts was 50% higher than for the outer suburbs.
A5 Key points
• Great variations occur in injury mortality and morbidity
which reflect children’s age, gender, socio-economic
group, cultural or ethnic group and place where they live.
• These variations suggest there is considerable scope for
improvement and intervention.
• A number of reports and policy initiatives are relevant
to this report: the Independent inquiry into inequalities
in health (Acheson, 1998); Saving lives: our healthier
nation (Secretary of State for Health, 1999); the report of
the Accidental Injury Task Force (Department of Health,
2002b); and the Economic and Social Research Council
(ESRC) Health Variations Programme.
• Children as a group are particularly vulnerable to injury:
they have little say in how environments are planned, and
in recent years child poverty has increased.
• National injury data systems provide good information
on age and gender and some information on social and
economic factors and place.
• National injury data systems do not routinely provide
information on ethnic groups, or on vulnerable groups
such as children in homeless families or disabled children.
9
B: Why are inequalities associated with injury risk?
Evidence from observational studies
B1 Introduction
In this section we examine the pattern of injuries relating
to inequalities. We draw on evidence from large cohort
studies and case–control studies examining the characteristics
of injured children. Evidence from other observational
studies will also be considered, particularly where
population denominator data have been used to enable the
characteristics of injured children to be compared with those
of the child population as a whole. The literature in this area
is large, and we will concentrate on studies carried out in
the UK. While observational studies are not able to point to
the direct causes of injury, or to account for differences
in injury rates in different groups, they do provide
evidence of different patterns of injury and the range of
factors associated with increased risk. These studies on
injury will be supplemented by reference to the more
general literature on risk where this sheds light on factors
that might contribute to inequalities between groups.
Knowledge about disparities between groups and the
factors leading to increased risk is important in targeting
and designing interventions to prevent injuries and to
reduce inequalities in injury rates.
In the following sections different aspects of inequality are
addressed – age, gender, social and economic factors, culture
and ethnicity, and place – and their relationship with injury
risk is examined. Separating aspects of inequality in this way
allows us to examine the scale and nature of inequalities in
different groups, but does little to explain why disparities
arise. This is partly because, as many studies show, the
factors leading to injury are multi-faceted and inter-related.
Further, as Thomson et al (2001) note in relation to ethnicity
and social background, such factors are not in themselves
direct contributors to risk. The more direct causes of injury
10
are exposure to hazards; parents’, carers’ and communities’
ability to protect children; and children’s capacity to manage
hazards to avoid harm.
Figure 1 sets out the structure of section B.
The inter-relationships discussed above can be considered to
operate in a broader conceptual framework, as suggested by
Millard (1994).
Three tiers of factors are suggested:
• Proximate tier – includes immediate conditions that result
in exposure to hazard and an injury event
• Intermediate tier – includes factors such as childcare
practices and other behaviour that increases exposure in
the proximate tier
• Ultimate tier – embraces the wider social, economic,
political and cultural processes that lead to a differential
distribution of resources in society.
This framework also falls within the overall general concept
of structure and agency (Gesler and Kearns, 2002). This seeks
to understand social phenomena, such as inequality, in terms
of the interplay between underlying structures in society that
create divisions along the lines of age, gender, class, ethnicity,
etc and individual (human agency) actions and beliefs. This
interplay between structure and agency is worked out in
specific geographic locations or places.
In each of the following sections attention is drawn to how
those factors associated with inequality at each of the tiers
of causation may influence exposure and ability to deal with
hazards so as to influence injury risk. Boxes 1 and 2 contain
examples of major studies whose findings illustrate the
inter-relatedness and complexity of factors associated with
injury risk.
Injuries in children aged 0–14 years and inequalities
Figure 1 Childhood injuries and inequalities: structure of the review
CULTURE
AGE
INJURY EVENT
Exposure to hazardous activities
and environments
Opportunity/ability of children to
avoid harm
PLACE
Opportunity/ability of parents,
carers/communities to protect
children from harm
GENDER
SOCIAL/ECONOMIC
FACTORS
Box 1 British Births Survey
This large cohort included 98% of children born in Great Britain during a week in April 1970.
More than 13,000 children were followed up when they were five years old (just over 80% of the original cohort) and
information was collected from more than 10,000 parents when the children were 10 years old. At each interview
parents reported any medically attended injuries and hospital admissions children had suffered over the previous five
years. Data were also collected from health visitors. Information on socio-demographic and family factors (including
maternal psychological distress) was collected along with information on children’s development and behaviour.
Parents who were not able to speak English were excluded from interviews.
The results are summarised below.
• The study revealed age-related patterns in injury rates, eg young children under five were much more likely to
suffer burns and scalds or poison ingestion; older children (aged 5–10) were more likely to suffer fractures and
head injuries
• Between the ages of 0 and 5, boys suffered 30% more medically attended injuries
• Young maternal age was associated with increased injury rates
• Children living with both natural parents were less likely to suffer injuries
• Aggressive children were more likely to suffer injuries
• Social class was associated with repeated injury and hospital admission
• Frequent house moves were associated with increased risk.
(Taylor et al 1983; Wadsworth et al 1983; Bijur et al 1986, 1988a,b; Stewart-Brown et al 1986)
Why are inequalities associated with injury risk?
11
Box 2 A case–control study examining child pedestrian accidents
This case–control study included children aged 5–16 years injured as pedestrians in five areas in the UK between 1991
and 1994. Cases included 152 injured children; controls were 484 children attending schools in the catchment areas
of hospitals treating injured children. The study included interviews with parents and children and surveys of the road
environment near children’s homes. Information was collected on socio-demographic factors and parents’ attitudes.
Summary of findings
• Injured children were more likely to live on through roads
• Injured children were more likely to live in pre-war housing
• Injured children were more likely to live on roads without parking
• Injured children were less likely to be living with both natural parents
• Injured children were more likely to be from non-white backgrounds
• Children with hearing impairment were at particularly high risk.
(Christie, 1995a)
B2 Age
Most studies observing injury in childhood reveal distinct agerelated patterns, not only in relation to the overall rates of
injury in different age groups, but also in terms of the types
of injury that children suffer at different times in their lives.
For example, evidence from the British birth cohort study
(Box 1) revealed that medically attended injuries sustained
by children aged between birth and five years were different
in scale and type from those of children aged between five
and 10 years. Here, while poisoning, burns and scalds were
frequent among the younger children, they were relatively
rare among the older age group. Conversely, rates of
fractures and severe head injury were greater among older
children (Bijur et al 1988a,b). An earlier cohort study revealed
that for traffic-related injuries there were higher rates of
medically attended injuries in older children. Thus while rates
of injury were 24 per 1,000 in boys under seven years, they
were 65 per 1,000 in boys over 11 years (Pless et al 1989).
Similarly, hospital- and general practice-based observational
studies have revealed age-related patterns in injury (Walsh
and Jarvis, 1992).
How does age affect injury risk?
While it is clear that the risk of different types of injury varies
at different ages, the reasons for these differences are less
obvious. As Wilson et al (1991) note:
12
‘The kinds of events in which a child is likely to be injured
depend on the child’s abilities and on where, how, and
with whom the child spends time, all of which change as
a child grows and matures. Children develop at different
rates, and caretakers’ decisions to grant independence
are subjective, so age is only a crude correlate of
developmental stage.’
Box 3 sets out a range of age-related factors that may
influence injury risk.
Box 3 Age and injury risk
• Physical development (height, size of head in relation
to body)
• Motor coordination (using playground equipment, falls
etc)
• Perceptual development (immature search strategy in
traffic)
• Cognition and intellectual development (awareness of
hazards)
• Immature attitudes and behaviour (risk taking in
adolescents)
• Changing nature of play and increasing independence
• Changing levels of supervision and greater freedom of
activities (licence to travel independently)
• Exposure to different environments inside and outside
the home
• Changing health service use.
Injuries in children aged 0–14 years and inequalities
Several studies provide information on age-related hazard
exposure – where, how and with whom a child spends time.
In studies where the locations of injury events are identified,
it is evident that at different times during childhood children
are exposed to different types of environments and activities
which may give rise to increased risk. These exposures may
also relate to gender. For example, in an observational study
of A&E attendance the most likely location for injuries in
female children of all ages (0–14) was the home. For boys,
the location of injury events depended on age – pre-school
boys tended to be injured at home, while older boys were
more likely to be injured during a sport or leisure activity in a
public place (Gorman et al 1999).
Studies that focus on particular types of injury emphasise
this link between age and exposure to different activities
and environments (Campbell and O’Driscoll, 1995). A
study of fatal head injury in the Northern Region similarly
revealed that for pedestrian injuries (which accounted for
approximately half the total number of fatalities) children at
different ages were likely to be injured in different locations
and while undertaking different types of activities (Sharples
et al 1990). The median age of children fatally injured while
playing in the street was seven, for those injured on the
school journey the median age was 10, and for those injured
on other journeys the median age was 12.
Health service utilisation may also partly account for different
patterns of injury observed in different age groups. In a study
examining injury in the Northern Region, Walsh and Jarvis
(1992) revealed that the rate of deaths, moderate and serious
injuries all increased with age. However, the same pattern
was not observed for all injuries leading to hospital admission
as it seemed that younger children with less serious injuries
were more likely to be admitted to hospital than older
children. This study revealed some of the difficulties inherent
in interpreting epidemiological studies using health service
utilisation or treatment data to ascertain rates of injury. It
may well be that a child’s age influences whether or not
parents seek medical attention for their child, and a child’s
age may also influence their medical management.
More general surveys of children’s activities also show how
exposure is age-related. Towner et al (1994), in a study of
injury risk in children aged 11–14, revealed that younger
children on school journeys were less likely to walk, had
shorter journeys, crossed fewer roads and were more likely to
wear helmets while cycling than their older counterparts.
The types of injuries children experience at different stages in
childhood illustrate the interaction between developmental
Why are inequalities associated with injury risk?
factors and exposure. Thus children under five are at
increased risk of home injuries because they spend more
time at home than older children, and are unable to manage
potential hazards because of their developmental stage.
For example, bathtub drownings occur almost exclusively in
children under two, and occur in older children only where
they have a disorder that places them at particular risk.
Similarly, as injury studies demonstrate, poison ingestion
and bathroom scalds occur most frequently in the youngest
children. In all these examples older children would have
the motor coordination and intellectual skills to avoid injury
(Gallagher et al 1984; Kemp and Sibert, 1992). Agran et al
(1994) have shown the vulnerability of young children to
injury in parking places and driveways close to their home.
In these injury events a child’s small stature and the inability
of drivers to see them when reversing contributes to risk.
Children’s psychological and perceptual development may
also contribute to risk. Studies of children in traffic reveal that
young children may not have the knowledge, skills or levels
of concentration needed to manage in the road environment:
skills related to perception of movement, sound, speed
and distance are not well developed in children (Sandels,
1975; Klein, 1980; Ampofo-Boateng and Thompson, 1991;
Christie, 1995b; Thomson et al 1996).
B3 Gender
Just as children of different ages have different patterns of
injury, most studies reveal distinct patterns of injury for boys
and girls. For most (but not all) types of injury the number
of injuries sustained by boys is greater than for girls, and
the association between males and increased risk appears to
apply at all severity levels. Having said this, the gender ratio
in different studies varies considerably.
Several large cohort studies identify increased risk
associated with being male. Focusing on traffic injuries,
the National Child Development Study identified an almost
twofold increase in risk for boys, with injury rates for boys
and girls being 122 and 65 per 1,000 children, respectively
(Pless et al 1989). For all injury types the gender ratio
identified in the British Births Survey was less pronounced,
but single and recurrent medically attended injuries,
and injuries leading to hospital admission, were all more
common among boys (Wadsworth et al 1983). A cohort
study in Nottingham focusing on young children (3–12
months) with follow-up for two years also identified being
male as a factor associated with increased risk of medically
attended injury in these young children (Kendrick and
Marsh, 2001).
13
Case–control studies examining the characteristics of injured
children have also identified an excess of injuries among
boys. For example, in a study of children attending hospital
70% of cases (children with injuries) compared with 51% of
controls (age-matched children with non-injury conditions)
were boys (Ohn et al 1995). Similarly, in a case–control
study focusing on children with pedestrian injuries, 65%
were male compared to 47% of age-matched controls
(Christie, 1995a).
Other observational studies have revealed an increased risk
of death and morbidity among boys at all levels of severity
(Sharples et al 1990; Sellar et al 1991; Laing and Logan,
1999).
Why is gender associated with injury risk?
Exposure to different environments and the ability of children
to cope successfully with them may relate to the factors
associated with inequality. Box 4 sets out factors associated
with gender that may influence children’s injury risk.
Information about children’s activities and the location
of injury events demonstrates that boys and girls may be
exposed to different types of risks. A study of leisure injuries
in Scotland revealed that football injuries accounted for
the largest group of injuries among boys but were low for
girls, whereas netball and horse riding were an important
cause of injury for girls but were rare for boys (Campbell and
O’Driscoll, 1995). Where boys and girls share the same leisure
environment the disparity in injury rates may be reduced – a
study of playground injuries revealed only a slight increase in
risk for boys (Mott et al 1994).
The large cohort studies suggest other mechanisms that may
help explain the increased risk of injury among boys. The
findings of the National Child Development Study suggest
that a child’s physical coordination, intellectual development
Box 4 Gender and injury risk
•
•
•
•
•
Different rates of physical development
Motor coordination
Spatial ability
Cognition and intellectual development
Gender differences in behaviour (eg risk taking, peer
pressure)
• Different forms of play and levels of independence
• Different levels of supervision and freedom of
activities
• Exposure to different environments.
14
and behaviour may contribute to their risk of pedestrian
injury (Pless et al 1989). The findings were not simple to
interpret, as different risk factors appeared to operate for
boys and girls at different ages. This may not be surprising
– both age and gender influence exposure to different
hazards, and physical and intellectual development may
occur at different times for boys and girls. Thus for boys
aged between seven and 11 years, sensory deficit, fidgety
or sensitive behaviour, and poor maths and oral ability were
all significantly associated with increased injury risk. For girls
of the same age, poor gross motor control was associated
with increased risk, and in girls aged 12–16 years poor
coordination increased risk. For both genders previous injury
increased risk of subsequent injury.
The British Births Survey also identified behavioural factors
associated with increased risk. In this study risk was greater
among boys, and those with high aggression scores were
at particular risk. As part of this study children’s general
behaviour was rated, and again there was a linear association
between behaviour scores, injury occurrence, re-occurrence
and hospital admission. Boys whose behaviour was rated as
poor were at high risk of injury (Wadsworth et al 1983; Bijur
et al 1986, 1988a,b).
More general studies on risk provide more information that
may help explain the inequalities in injury rates between
boys and girls. For example, parents’ attitudes and children’s
risk-taking behaviour may be different for boys and girls,
and may influence exposure. Towner et al (1994), in a study
of travel and leisure activities among secondary school
children, reported that compared to girls, boys were more
likely to travel unaccompanied on their school journeys,
make longer leisure journeys, and were much more likely
to ride bicycles. Similarly, a study of pedestrian exposure
revealed that between the ages of five and 15 years, boys
had much greater exposure to the road environment than
girls. This difference in exposure can account for much
of the difference in casualty rates for children aged 5–9
years, with boys and girls having similar casualty rates per
kilometre walked. When exposure was taken into account,
girls aged between 10 and 15 were at higher risk per
kilometre walked compared to boys (Ward et al 1994).
There is some evidence that boys and girls of the same age
may differ in their ability to manage hazards. One study has
shown that boys have better motor coordination and faster
reaction times compared to girls (Grossman and Rivara,
1992). Boys and girls may also differ in their attitudes
towards taking risks and towards protective equipment
such as cycle helmets.
Injuries in children aged 0–14 years and inequalities
B4 Social and economic factors
Social and economic influences on injury and injury risk can
be considered as part of the wider structural influences on
health, part of the ultimate tier suggested by Millard (1994)
(page 10). Section A3 demonstrates that there is a strong
association between social class and injury. Thomson et al
(2001) make the critical point that it is much more difficult to
determine what gives rise to this association. They note that:
‘it seems certain that low socio-economic status only
correlates with accidents to the extent that it indicates the
presence of other factors which more directly confer risk
on members of low socio-economic status groups’ (p. 17)
Thomson et al (2001) suggest that more proximate factors
have to be explained to establish clear causal pathways.
Thus we must consider socio-economic factors not as a
source of explanation, but as something to be explained
(Gesler and Kearns, 2002). Understanding how social and
economic factors influence causal pathways is complex.
This complexity is compounded by the fact that a broad
range of quite different, but often related, factors have
been identified which have been associated with injury
risk. These include economic indicators such as family
income; social factors such as maternal education; factors
concerning family structure and characteristics including
lone parenting, maternal age, number in household and
number of children under 16; accommodation-related
factors such as tenure, overcrowding and housing type; and
indicators encompassing a range of factors measured at
either individual or neighbourhood level. Paternal occupation
has frequently been used as a means of stratifying social
groups, but many authors point to the weakness of this
means of classification in a context where a large proportion
of children are raised in households without a male
breadwinner. Overall, it is not clear how these many and
different factors actually increase risk, whether they have an
independent effect or are markers for one or other factor(s)
not measured, or whether they are incremental or interactive
in the ways they influence risk.
Despite these problems with definition and interpretation,
many studies at national level reveal that social and economic
factors are strongly associated with injury risk. Section
A3 illustrates the steep mortality gradient associated with
social class at national level; other studies at regional or
local level reinforce this. In a study of fatal head injuries in
the Northern Region, Sharples et al (1990) noted that for
pedestrian fatalities the rate of injury in the most deprived
neighbourhoods (electoral wards) was 9.1 per 100,000
Why are inequalities associated with injury risk?
Box 5 Social and economic factors and
injury risk
• Lack of money (ability to buy safety equipment)
• Exposure to hazardous environments inside and
outside the home (facilities for safe play; smoking
parents; older wiring; lack of garden; small, cramped
accommodation)
• Ability of parents/carers to supervise children (singleparent families; parents’ maturity, awareness and
experience; depression and family illness; large family
size)
• Children’s attitudes and behaviour (risk taking)
• Access to information and services.
children compared to 0.23 per 100,000 in the most affluent
areas. Eighty per cent of these injuries were sustained
within 1.6 km of children’s homes. Another study in the
Northern Region revealed that, while inequalities were most
pronounced for fatalities, they could also be identified for
non-fatal injuries. Here the relative risk of injury for the
most, compared to the least, deprived groups were 2.4 for
all injuries, 3.6 for severe injuries and 7.5 for deaths (Walsh
and Jarvis, 1992). At the local level, Laing and Logan (1999)
identified a correlation between A&E attendance and area
of residence, with this factor accounting for a third of the
variation in attendance rates.
How does socio-economic group influence risk?
Box 5 sets out some of those mechanisms whereby social
factors may influence the proximate tier (page 10) in the path
of causation by increasing children’s exposure to risk or their
ability to overcome hazards.
Evidence from cohort and case–control studies demonstrate
some of the associations between family background and
injury risk. The British Births Survey identified a number of
maternal factors associated with increased risk of sustaining
medically attended injury between birth and five years.
Young maternal age and maternal psychological distress
were both associated with increased risk. These factors, in
turn, were associated with high aggression scores in children
which had an independent effect on injury rates and hospital
admission (Bijur et al 1986). Further analysis suggested that
the influence of maternal age on injury may have related
to issues concerning supervision and maternal experience.
Thus, while children of young mothers had more injuries at
home, they were not at increased risk of injuries in traffic or
at nursery (Taylor et al 1983). The ability of families to provide
15
appropriate supervision for their young children may also
relate to the number of other children in the family, family
structure, housing type and neighbourhood characteristics.
In the same studies recurrent injury was associated with
larger families and frequent moves, while increased risk of
hospital admission was associated with atypical family type
(ie where the child was not living with both natural parents),
frequent moves, and the presence of younger siblings.
Maternal awareness of hazards may also influence risk; when
children were followed up at age 10, maternal education was
associated with risk of hospital admission (Stewart-Brown
et al 1986).
The association between young maternal age and increased
risk of admission to hospital was also identified in the cohort
study carried out by Kendrick and colleagues in Nottingham
(Kendrick and Marsh, 2001). For traffic-related injuries,
various family factors were associated with increased risk in
the National Child Development Study (Pless et al 1989). In
this study, girls in families described by teachers as ‘having
problems’ were at increased risk of injury, along with boys
who were not living with their own mothers or were in local
authority care. Again factors were inter-related: boys whose
behaviour was described as fidgety and who were in local
authority care were at 1.8 times the risk of injury compared
to boys without these characteristics. The many factors
associated with risk identified in that study led the authors
to conclude that there was a lack of strong predictors and
serious difficulties in adjusting individual behavioural risk
factors (Pless et al 1989). The links between behaviour, social
factors and injury were also recognised as being complex in
the British Birth Study cohort, where Bijur et al (1998a) note:
‘Thus, whereas the observed consistent nature of injury
occurrence may be due, in part, to the characteristics of
the child, it is also likely to be explained by a complex
interaction between the child and his physical and social
environment.’ (p. 711)
When children were followed up at age 10 as part of this
study, it was revealed that the area in which children lived
and the play facilities available to them were associated with
risk. Thus children living in poor urban and rural areas were
at increased risk, along with children who had no access to
unshared gardens or yards. This study also highlighted the
different factors associated with single and recurrent injuries
and those leading to hospital admission, again illustrating the
complex way in which factors increase risk (Stewart-Brown et
al 1986).
Case–control studies also identify risk factors that may partly
explain why children from deprived backgrounds are more
16
vulnerable to injury. Christie (1995a) examined social and
environmental factors associated with risk of pedestrian
injury. This study involved in-depth interviews and a survey
of the road environment around children’s homes. Features
of the environment associated with increased risk included
living in pre-war housing, on a road with through traffic
and/or without on-street parking. This finding suggests that
injured children lived in different types of traffic environment
than uninjured children. Christie concluded that many of the
factors associated with increased risk were highly correlated
with socio-economic group.
The environment in the home may also influence risk factors,
eg overcrowding may have an effect on supervision and on
the physical environment in the home. Alwash and McCarthy
(1988) noted that children from overcrowded homes were
at almost three times the risk of injury compared to children
from uncrowded homes. Sharing facilities with other families
was also associated with increased risk.
A study of injuries among older children in Scotland suggests
ways in which social and environmental factors inter-relate
so that more deprived children may have greater exposure
to risk and may be less equipped to deal with it (Williams
et al 1991). In this study, overall, medically attended injury
rates were not associated with the father’s occupational
group or family affluence scores, but specific injury types
were associated with social group. Lower occupational
group was associated with injuries occurring in the road
environment, while affluent children were more likely to be
injured in school. Bicycling and pedestrian injuries (together
with assault injuries) were associated with lower social
groups, whereas sports and car passenger injuries were
more prevalent among the more affluent. The use of safety
equipment also related to social factors, with those from the
higher occupational groups being more likely to wear cycle
helmets and protective sports equipment. Other behaviours
also appeared to be associated with social background, with
those from more deprived backgrounds reporting increased
risk taking.
Other more general studies illustrate how exposure and
safety behaviours may relate to social factors. Towner et al
(1994) reported that children from deprived backgrounds
were less likely to travel in cars and more likely to walk,
compared with their more affluent counterparts. DiGuiseppi
et al (1998) showed (not surprisingly) that primary school
children without access to a car may be more exposed to risk
as pedestrians, but that the majority of child pedestrians in
this age group were accompanied on their journey to school.
Kendrick and Marsh (1998) revealed that use of baby walkers
Injuries in children aged 0–14 years and inequalities
in the home was associated with social group, but that this
association was not strong. This study also showed that use
of baby walkers was associated with other unsafe practices
(eg absence of stair gates or fire guards) which are likely to
compound the risk associated with baby-walker use.
Studies investigating parents’ attitudes and injury risk
indicate that attitudes may vary in different social groups.
In a qualitative study involving interviews with parents of
injured children residing in affluent and deprived areas,
Sparks et al (1994) revealed that parents living in different
environments perceived risk and responsibility in different
ways. Affluent parents perceived that the low rates of
injury in their area were due to the quality of parenting and
adequate supervision, while parents living in deprived areas
perceived that the high injury rates in their area related to
the poor quality environment in which they lived. Affluent
parents regarded it as their own responsibility to keep their
children safe, whereas parents living in deprived surroundings
thought the council had a responsibility to keep children safe,
particularly for accidents outside the home. Parents in both
groups said they taught their children about safety.
One reason for increased injury rates in children from more
deprived backgrounds may be that parents are unaware
of hazards, as they do not have access to injury prevention
advice or other health promotion interventions. There is some
evidence to support this and we will return to this issue in
the section on interventions (page 24). Kendrick and Marsh
(2000) revealed that more deprived families were less likely
to receive health promotion interventions, but that this
depended to some extent on the nature of the intervention.
While families living in rented accommodation were less likely
to take up an offer of first aid training, they were more likely
than other groups to take up low-cost safety equipment. In
an earlier study by Kendrick et al (1995), deprived families
were less likely overall to take up a range of child health
services where safety advice was imparted, but there was
no evidence that children who had not attended these
appointments were any more likely to be injured.
While there is a great deal of evidence linking social factors
and injury risks, the mechanisms whereby children from more
deprived backgrounds are injured remain elusive. Injury risk
is likely to be mediated by a broad and inter-related set of
factors. These factors may operate differentially at different
ages, for different types of injury, and in relation to gender
and cultural group. There is evidence that children from more
deprived backgrounds may have individual characteristics
that increase risk by reducing their ability to deal with
hazards. Many of these factors are not easily addressed,
Why are inequalities associated with injury risk?
although educational interventions may modify some risks.
It is also evident that children in these groups face more
hazardous environments and are exposed to hazards more
frequently than their more affluent counterparts (see section
B6, page 19). Hence initiatives at community level, modifying
the environmental hazards in children’s neighbourhoods
or reducing the number of exposures to hazards, may be
effective in reducing inequalities.
In addition to trying to establish the links between broad
social and economic factors and the more proximate causes
of injury, a further challenge relates to the fact that the
various factors may not operate to the same degree for
different injury severity levels. For deaths and, to some
extent, severe injuries there is a recognised association with
socio-economic factors, but there is a less pronounced
association between minor injury and social and economic
factors. This may be an artefact of how data are collected
–while case ascertainment for fatalities is relatively simple,
the threshold at which parents of children with a minor injury
seek medical attention, or at which medical staff decide to
admit a child to hospital, may be influenced by social factors.
A third complexity is that the association between social
factors and injury does not appear to apply equally to all
types of injury. Thus while burns and pedestrian injury are
more frequent among some social groups, this relationship is
not apparent for other injury types. There is also controversy
about whether adolescents are less susceptible to the risks
associated with deprived social backgrounds.
B5 Culture and ethnicity
The links between culture and injury, like those between
social group and injury, are complex.
Culture is difficult to define. It can be taken to refer to the
values, rules and beliefs of social groups as they reflect
their view of the world. Culture can also be considered to
operate at different spatial scales, from the international to
the national, regional and local scales, and to include the
relationship between minority cultures and the culture of the
majority in a location. Ethnicity refers to groups of people
who share a relatively homogeneous culture through kinship
and/or particular situations.
As with socio-economic factors, culture should not be
seen as a source of explanation for inequalities, but as a
contributory factor that needs to be explained. In the same
vein, Thomson et al (2001), as with their consideration of
socio-economic processes, stress the need to search for
proximate factors of culture that increase exposure to injury
17
risk. The role of culture and ethnicity in the area of injury has
been little explored although its importance has often been
acknowledged. Berger and Mohan (1996) state:
‘Although little research has been done on the subject,
socio-cultural factors play a major role in shaping the
frequency and nature of injury events, and the intensity
and success of efforts to reduce injuries.’
However, out of a book of 278 pages they devote only 1.25
pages to this topic. What work has been done tends to view
culture/ethnicity as a facet of a minority group in society,
rather than how major cultures in society may vary between
countries. Thus we only have occasional glimpses into what
may be national cultural variables. Hillman et al (1990)
note differences between German and English children in
terms of exposure to risk which stem from variations in
culture. These include the greater supervision of children
on streets in Germany by all adults, not just parents; the
observation and guidance of children if their behaviour falls
below the standard expected; and the German law which
forbids children to use the street for play. Observations of
pedestrians in Denmark and Germany similarly show a great
difference from the UK in pedestrian crossing behaviour, the
former groups rarely crossing when the red man is shown.
In Nigeria the usual pattern of boys having more pedestrian
accidents than girls is reversed, because the cultural practice
is to keep boys at home and to send girls out on errands.
Boys in Nigeria, on the other hand, have more domestic
accidents (Berger and Mohan, 1996). We are not clear how
cultural variations and attitudes to what constitutes an
‘accident’ may have a bearing on exposure.
In trying to tease out the proximate factors for injury
among minority ethnic groups in Britain, Thomson et al
(2001) concluded that, despite the overlap between socioeconomic structural factors such as poverty/housing and
ethnicity, it is possible to identify specific ethnic factors that
influence exposure – these are supervision, socialisation, and
opportunities to learn.
In most studies that examine injury and culture, ‘culture’ has
been defined narrowly and tends to relate to differences in
injury rates between majority compared to minority racial or
ethnic groups. Even with this more limited view of culture,
definitions have varied across studies so that comparing
and interpreting findings is not simple. In some studies all
‘non-whites’ are grouped, implying that the many subgroups
in the ‘non-white’ category can somehow be regarded as
similar to each other but different from the ‘white’ group.
It is known that in the UK different cultural groups tend
to be localised and findings cannot easily be transferred
18
to other settings. Some groups are from long-established
communities where children and their parents were born
in the UK; other groups have arrived more recently. In
many studies it is difficult to disentangle the influence of
deprivation on injury risk from that associated with cultural
group. Yet, as noted above, to understand whether cultural
background is associated with risk we must understand how
culture influences exposure to risk and ability to cope with
risk independently of, or in addition to, any influence exerted
by social and economic factors.
There is some evidence that children from minority groups
in Britain have different injury experiences from the majority
group. However, findings relating to the scale and types of
injury suffered by children in different cultural groups are
mixed and, as with other aspects of inequality, the influence
of culture may be mediated by age, gender, social and other
factors.
In a study examining the links between ethnicity and injury,
Alwash and McCarthy (1988) used census data to calculate
injury rates for Caribbean, Asian, British and other children
attending a London hospital. Here, the ethnic group was
related to the father’s country of birth, or mother’s where the
parents were not living together. Rates and types of injury
suffered by children from different ethnic groups were similar.
Other factors, including distance from hospital, social class
and housing-related factors, were more important predictors
of hospital attendance.
Other studies have pointed to increased risk for children from
minority ethnic backgrounds, but in many cases there has
been no adjustment for social factors so it is difficult to know
whether ethnicity exerts any independent effect or is simply
a marker for more general deprivation. For example, an
analysis of national mortality data revealed that compared to
injury death rates overall, the death rate for Pakistani children
was 2.55 times that of all children (Balarajan, 1995). It is not
clear why these differences occurred.
In a case–control study where ethnicity was ascertained
by children’s names, there was no increased risk of A&E
attendance for injury among those children defined as being
from minority ethnic groups (Ohn et al 1995). Here more
than two thirds of all children, in both case and control
groups, were from deprived backgrounds.
For road injuries there may be more evidence of an
association between ethnic group and injury risk. In a case–
control study of pedestrian injuries, Christie (1995a) noted
that cases were more likely to be from non-white families.
Further analysis revealed that non-white parents tended to
Injuries in children aged 0–14 years and inequalities
have lower scores on a measure examining risk perception
compared to white parents.
Other studies have indicated that young children of Asian
descent had an increased risk of injury on the road, but
that among children aged 10 years and over there was no
increased risk (Lawson and Edwards, 1991).
Why is culture/ethnicity associated with injury
risk?
It needs to be emphasised again that few studies examine
culture in the wider sense as a possible influence on injury,
but focus rather on minority ethnic groups. The possible
reasons for increased risk of injury among children from
different cultural or ethnic groups are set out in Box 6.
As suggested in other sections of this report, it is likely
that the way that culture influences risk is by influencing
exposure and the ability of children to deal with hazards.
As Thomson et al (2001) note, however, separating the
influence of culture from that of social deprivation is
complex, and if social factors are taken into account many
of the differences between cultural groups disappear. Having
said this, exposure, attitudes and behaviour may relate to
cultural group, and these may increase risk for some children.
The study by Christie (1995a) suggests that parents from
non-white groups may not perceive risk in the same way as
white parents. This may, in turn, influence attitudes towards
exposing their children to the road environment and the level
of supervision they feel is appropriate. Similarly, Hapgood et
al (2000) revealed that ethnicity was negatively associated
with safety practices (eg use of safety equipment), but
that this factor explained only a small part of the variation
between groups as, overall, unsafe practices were common
among all groups. Thomson et al (2001) suggest that recent
immigrants may be unfamiliar with the road environment
and that parents may have difficulty in acting as role
models and teaching their children appropriate road safety
Box 6 Culture/ethnicity and injury risk
• Exposure to different environments inside and
outside the home (food preparation practices,
different activities)
• Access to information and services
• Barriers relating to language
• Ability of parents/carers to supervise children
(attitudes to childcare, lack of familiarity with traffic
conditions for first-generation immigrant families).
Why are inequalities associated with injury risk?
behaviour. Many of these theories to explain increased risk
among different cultural groups, although plausible, remain
untested.
One possible explanation for differences between groups is
access to services which may influence levels of awareness.
There is some evidence that parents from minority ethnic
groups are less likely to attend child health surveillance
appointments at which safety advice is offered (Kendrick
and Marsh, 2000). The reasons explaining this variation are
complex, and outside the scope of this review.
B6 Place
‘Place’ has been described above as where the interplay
between structural factors in society and human agency is
worked out. The notion of place requires some elaboration
– here we consider three of its attributes which are relevant
to this study:
• Place as a portion of geographic space
• Physical/environmental attributes of places
• Place as an area to which people attach certain meanings
and actions.
Place can, at one level, be considered simply as a portion
of geographic space. Often it is conceived as an official
geographic entity, eg enumeration districts, wards, boroughs,
counties or health authority regions (Gesler and Kearns,
2002). In epidemiological studies the use of statistics
gathered at some areal scale generally accords to this idea of
place (section A3). Important patterns of risk and injury can
be discussed in this way.
Places can also be seen as possessing certain physical and
environmental attributes such as topography, land use,
the layout of roads and buildings etc. These features may
influence exposure to injury by affecting the speed and
volume of traffic, visibility, and the ratio of pedestrians
to traffic. Exposure to risk may reflect a number of these
attributes coming together in certain places, eg a lack of
garden play space and busy roads. The physical environment
of a place can be one of the proximate factors in injury risk,
closely linked to social and economic factors (see section B1,
page 10).
However, ‘place’ in the geographic literature is taken to
mean more than a portion of space or a physically different
area (Shaw et al 1988; Curtis and Rees-Jones, 1998; Gesler
and Kearns, 2002). Places can be sites or areas to which
people attach certain meanings, giving them a ‘sense of
place’ or belonging, such as a neighbourhood (Spencer
and Blades, 1993). Equally, it could be an area from which
19
people feel alienated. Godkin (1980) showed that it was
beneficial for alcoholics to have a feeling of ‘rootedness’ in
places where they lived, or went for help, in their treatment.
Alternatively ‘placelessness’, a lack of belonging to an
area, may contribute to unhealthy behaviour. In addition,
Matthews (1995) points to the important inter-relationships
between culture (section B5) and environment in influencing
children’s behaviour in particular places. Places, in this sense,
are created by the intersections of local and wider factors
where individual agency and society’s structures are acted
out. And these intersections also have the dimension of
historical time, creating layers of varying experience for
people in different areas. Thus Phillimore and Morris (1991)
compared two towns with similar levels of social deprivation
but different premature mortality rates. They concluded
that closer examination of the social and economic histories
of places was needed as part of the explanation for these
health variations. Epidemiological studies that simply try to
‘control’ certain social and economic factors when comparing
different places will not take account of this notion of place.
Ultimately these concepts of place feed into the debate over
whether health variations are because of people or places
– compositional or contextual factors. The former view sees
area health patterns as the aggregate effect of personal
factors, whereas the latter stresses the context in which a
person lives. Health is thus not only a product of individual
characteristics but is influenced by where a person lives
(Macintyre et al 1993). So far, few studies have developed
these ideas of place in relation to injury.
How does place affect injury risk?
The three conceptions of place discussed above can be
traced in a number of works related to injury and inequality.
By far the largest corpus of material relates to the use of
space as an areal unit for statistics in describing the varying
patterns of injury at national, regional and local scale. They
help us answer the question of where injury inequalities
can be found in the UK and make broad correlations with
structural factors such as social and economic spatial
variations (section A3).
Differences in injury rates associated with children’s houses
and the road environment near children’s homes have
been explored already in section A3, relating to social and
economic factors, as these broad themes are inevitably
inter-connected. As Christie (1995a,b) and others have
noted, characteristics of the environments in which children
live relate closely to economic and social indicators. Where
children spend their time influences not only exposure to
20
hazard, but also the ability of children to avoid harm and the
ability of parents and others to protect children. Children may
be more at risk in cramped living accommodation, especially
where this problem is compounded by overcrowding. In
section B7 (page 21) we consider the home as a place where
limited and cramped accommodation, with crowded kitchens
and little space for children to play, can expose children to
more physical hazards. Parents may have more difficulty
providing adequate supervision in housing that lacks safe
play areas visible from the house, and where children use the
street environment for play. King et al (1987) found higher
child accident casualties in streets of Victorian terraced
houses with little or no gardens, few garages and much
on-street parking, compared with post-war housing estates.
The way in which many physical and environmental factors
can come into play in particular places is emphasised by
Whitelegg’s (1987) description of a road traffic accident as
being:
‘a function of the land use system, residential patterns,
population densities, street geometry, location of
workplace, shopping precinct, health centre or other
traffic generators.’
The third concept of place, as a site or area to which
people attach certain meanings and actions, has been little
researched so far in the field of injury and inequality. A few
studies have attempted to tease out some of the implications
of place and identity. A recent study by Grundy et al (2002)
examines the ways in which schoolchildren identified
‘safe’ and ‘dangerous’ places in their neighbourhoods as
settings for their leisure time. Their personal and group
experiences and perceptions, parental rules, the behaviour
of others, and the physical environment combined to
produce a kaleidoscope of places which differed in many
ways from the ‘official’ view of hazardous places. The study
revealed through qualitative work how children experience
exposure to risk in their everyday lives, and how human
agency combines with social and economic structures often
determined by others.
Sparks et al (1994) have some interesting findings relating to
place and injury. Parents in an affluent area saw exposure to
risk as their responsibility, and the nature of the environment
or place around them was not in itself seen as critical. Parents
in a poor area defined exposure to risk very much in the
terms of their surrounding environment, and responsibility
for making that place safe was seen as the responsibility
of the authorities rather than the inhabitant. Implicitly, this
study raises issues about attachment to place and ‘sense of
place’ and injury risk.
Injuries in children aged 0–14 years and inequalities
Box 7 Place and injury risk
• Different places lead to differing exposures to injury
risk through their varying physical environment,
eg road layout, housing density.
• People’s relationship to the place they live in can
vary, eg lack of identity with a community or strong
community spirit.
The work of Roberts et al (1995) also explored how human
agency works out in particular places. Rather than viewing
injury and inequality from an official injury statistics point
of view, the authors considered the overall lives of children
and adults in their environments in the Corkerhill area of
Glasgow, and how these compromised their safety. They
were interested to find out how human life ‘works’ in an
area in its wider social and environmental context. However,
although place plays a considerable part in this study, a
number of the ideas of place discussed above are not
explicitly explored.
Finally, the definition of place as an area to which people
relate is closely linked with an emerging concept in the field
of health promotion – social capital (Hawe and Shiell, 2000).
Social capital can be defined as:
‘the resources within a community that create family and
social organisation ... These resources, which arise out
of activities such as civic engagement, social support or
participation, benefit individuals, but are developed in
relationship to and with others, for example within groups
or communities.’ (Swann and Morgan, 2002)
The concept of social capital allows us to focus on social
context to explore the importance of place in people’s
daily lives and the relevance of those places for their social
relationships (Morrow, 2002).
The possible reasons for increased risk of injury among
children from different places are given in Box 7.
B7 Vulnerable groups
This report adopts a framework for considering inequalities
in injury which is encompassed by the broad concept of
structure and agency in society, together with a three-tier
ordering of factors related to exposure to injury ranging
from proximate to underlying. These processes are acted
out in particular places. In this framework we can envisage
inequalities generally ranging along a continuum from those
least exposed to those who are most exposed to injury risk.
Why are inequalities associated with injury risk?
However the idea of a continuum is probably too simple
– there will be groups in society who are in extreme positions
of vulnerability, those who are marginal to, or excluded from,
the mainstream of social life. Here the interplay between the
socio-economic structures of society and certain attributes of
human agency will ‘ratchet up’ all the general problems of
inequality to produce groups who may be exposed to much
higher levels of injury risk. In these groups risk factors may be
additive and cumulative over time. In this section we consider
two examples of vulnerable groups: homeless families, and
families with disabled children.
Homeless families
Homeless families are likely to represent some of the most
materially and socially deprived families in the country.
Homelessness can be viewed as people sleeping rough on the
streets, but this is only the most visible form of homelessness.
The charity Shelter advocates the use of a broader definition
which encompasses:
‘those living in temporary accommodation, those living
in poor or overcrowded conditions and people forced to
sleep on friends’ floors because they have no home of
their own.’ (Stone, 1997)
Widdowfield (1999) examines the difficulties involved in
trying to measure homelessness and the limitations of official
homelessness statistics, and believes that homelessness
counts are getting further from reality. Official homelessness
figures simply record the number of households accepted
as homeless by local authorities under the provision of the
1985 and the 1996 Housing Acts. Shelter estimates that
about 65% of households accepted as homeless by local
authorities are families with children (Shelter, 2002). For the
year 2000–01 this represented about 65,670 households
with dependent children, and over 100,000 children lived in
these households (Shelter, 2002).
In a study in central London, Lissauer et al (1993) reviewed
70 children admitted to hospital who were living in
temporary accommodation, and found a significantly higher
rate of admissions for accidental injury for the former
compared with permanently housed children. Richman et
al (1991) examined the use of hospital services by children
in B&B accommodation, but found no difference in A&E
attendance for unintentional injury for children from B&B
accommodation and other children. There was, however,
an interesting difference in the type of injury experienced:
children from B&B accommodation had more burn and
scald injuries, while other children had more bruises, soft
tissue injuries and fractures. How can factors associated
21
with homelessness increase injury risk in children? The
physical conditions of temporary accommodation – cramped
cooking and laundry facilities, limited indoor spaces to play
and absence of safe outdoor places to play – can expose
children to more physical hazards. In Lissauer et al’s (1993)
study in central London of hospital admissions of children in
temporary and permanent accommodation, 70% of mothers
from temporary accommodation complained about lack of
space, compared with 54% of permanently housed families;
68% said there was nowhere safe where their children
could play, compared with 14% of controls. An independent
assessment of children’s play facilities by the interviewer in
this study found that only 2% of the children in temporary
accommodation had satisfactory play facilities, compared
with 96% of the children in permanent accommodation. The
following quotes from mothers of homeless families serve to
illustrate the conditions that homelessness imposes:
‘There’s only one cooker between about 130 people ...
and if you do start cooking and go back to your room
then your dinner either gets burnt or stolen.’
‘My daughter has been scalded and other children have
had electric shocks. In your own home you can be careful,
have gates across doorways, etc, but in a hotel you’ve got
to put up with other people being careless.’
(Spencer, 1993).
These examples also illustrate how multiple occupation can
reduce the power of parents to protect their children, eg in
the use of safety equipment. Stress can be caused by poor
living conditions, insecurity and enforced mobility. Homeless
families can often be placed in accommodation far from
the area they know. In London, one in seven households
accepted as homeless were in temporary accommodation
outside their local authority area. Informal support networks
of friends and family can be severely disrupted by distance.
Barry et al (1991) have shown that families living in B&B
accommodation often suffer from isolation, boredom and
loneliness.
In the study by Lissauer et al (1993) many of the homeless
families were recent immigrants, including refugees from
various countries and ‘many of the homeless families were
further disadvantaged by their inability to speak fluent
English’. Maternal depression was significantly increased
among the homeless families. Homeless mothers had
experienced a median of three significant life events in the
previous year compared with mothers in the control group
who had experienced a median of one significant life event.
All but one of the homeless mothers had moved in the
previous year, compared with a quarter of the controls.
22
Families with disabled children
Disabled children and their families can be exposed to
additional burdens associated with poverty, and have fewer
resources than other families. Specific types of disability, such
as those relating to hearing, seeing or communication, may
expose children to specific risks.
Botting and Crawley (1995) note that there are no routine
national data on disability in childhood that can be used to
monitor trends on conditions such as cerebral palsy, deafness
or blindness. A major survey of disability in childhood carried
out by OPCS in 1985–86 estimated that 3% of children
aged 0–15 years in Great Britain had a disability (Bone and
Meltzer, 1986). For each age group, boys had a higher level
of disability than girls, and prevalence was higher in schoolage groups than in younger children. Disabilities included
those related to behaviour, communication, locomotion,
continence, intellectual functioning, personal care,
hearing, consciousness, dexterity, reaching and stretching,
seeing, disfigurement and eating, drinking and digestion.
Behavioural disability was most commonly identified in 2% of
children, but a given disability seldom occurred in isolation.
In relation to hearing loss, eight per 1,000 children in Great
Britain aged 5–9 years and six per 1,000 aged 10–15 years
had some level of hearing loss, with two per 1,000 aged 5–9
and one per 1,000 aged 10–15 years having severe hearing
loss. Two per 1,000 children under 16 years had some seeing
disability, including children who could not see enough to
recognise someone they knew across the road.
Gordon and Heslop (1999) point out that:
‘None of the specifically designed poverty surveys in
Britain have had a sufficiently large sample size to provide
direct evidence of the levels of poverty experienced by
households with disabled children. However, all the
indirect evidence available indicates that as a group these
households are among the “poorest of the poor”.’
The presence of a disabled child in a family may not only limit
the earning power of parents but can also alter their patterns
of expenditure. For example for families with a disabled child,
a car is often a necessity rather than a luxury.
In their case–control study of child pedestrian accidents
(Box 2), Christie (1995a) reported that hearing impairment
among accident victims was noticeably high compared to
control children. Of the nine accidents involving children
reporting hearing difficulties, three were described as having
slight hearing difficulties, three were deaf in one ear, one was
hard of hearing, one suffered from ‘glue ear’, and one was
Injuries in children aged 0–14 years and inequalities
profoundly deaf. Although the numbers were small in that
study, they support previous findings indicating that accident
risk for child pedestrians was higher for those with hearing
difficulties (Scottish Development Department, 1989).
The Department for Transport, Local Government and the
Regions commissioned a review of the road safety of children
with special needs (DTLR, 2002). The review identifies the
characteristics and lifestyles of children with special needs
that may affect their ability to be safe pedestrians.
B8 Key points
• The factors associated with injury inequalities are multifaceted and inter-related. The causal pathways linking
these factors to injury events remain uncertain.
• The three tiers clarify these inter-relationships:
(1) proximate tier, immediate conditions that result in
exposure to hazard; (2) intermediate tier, eg childcare
practices; (3) ultimate tier, the wider social, economic,
political and cultural processes.
• These tiers fit into the concept of structure and agency.
The interplay of structure and agency is worked out in
specific geographic locations.
• More direct causes of injury include exposure to hazards;
the ability of parents, carers and communities to protect
children; and children’s abilities to manage hazards.
• Greater knowledge about disparities between groups and
factors leading to increased risk is important in designing
interventions.
Why are inequalities associated with injury risk?
23
C: Have injury intervention studies addressed
inequalities?
C1 Introduction
be designed and targeted in a way that may reduce risk
among the most vulnerable.
The findings of studies of risk described in section B have
implications for those involved in preventing injury and
attempting to reduce inequalities in injury rates. However
there are some difficulties in responding appropriately to
differences in risk. First, many risk factors are not easy to
change (eg a child’s gender and age). Second, single risk
factors alone often do not account for much variation in
injury risk. Rather, it appears that combinations of factors
and possible interactions between them influence risk.
Further, it is not differences in age, gender, social and
cultural background that lead to injury events, but the
influence of these factors on exposure and ability to deal
with hazards. How, then, can those involved in health
promotion use information on inequalities to reduce risk
among the most vulnerable groups? Box 8 sets out some
examples of how health promotion interventions can
In this section we examine a broad range of health
promotion interventions to reduce injury. Apart from broad
targeting in terms of children’s ages (and this does not
always reflect injury risk), relatively few interventions are
designed and targeted in a way that takes account of injury
inequalities. Some studies may increase risk in some groups
by denying them the intervention, eg in many research
studies those unable to speak and/or read English are
excluded.
In sections C2–C8 we analyse the information related to
inequalities that is reported in the 164 intervention studies
reviewed. The full tables are online at www.hda.nhs.uk/
evidence.
Box 8 Examples of interventions to address inequalities
• Age: it is known that children’s developmental stage influences their ability to deal with hazards. Child road safety
education must build on what is known about children’s age-related attitudes, perceptions, knowledge and skills.
• Gender: boys are at increased risk of most types of injury. Promotion of safety equipment (eg cycle helmets) must
build on what is known about boys’ attitudes to helmet wearing.
• Social and economic factors: families living in multi-occupancy dwellings or in low-income neighbourhoods
may be particularly vulnerable to injury from house fires. Fire safety campaigns could target high-risk areas. Smoke
alarms must be fitted and working: campaigns to promote smoke alarms could include the provision of free devices,
installation and maintenance. Parents need to be aware of the vulnerability of young children to injury from smoke
and fire.
• Culture/ethnicity: educational materials must be available in a range of languages to reflect local diversity. The target
group should be involved in the development of the intervention, or consulted about its contents.
• Place: families in rural areas may have different needs from those in urban areas. Community-based interventions
need to take account of whether people in that community feel a sense of communality, or a sense of belonging to
that place. Transport is important, particularly in relation to people’s access to facilities.
24
Injuries in children aged 0–14 years and inequalities
C2 Pedestrian injuries
C2(i) Transport policy and traffic calming
INTRODUCTION
Seven studies were identified and reviewed relating to
transport policy and traffic calming. Of these seven, only
one was related to the evaluation of transport policy – this
study was conducted in the UK. Two studies relating to traffic
calming were conducted in the UK, two in Denmark and two
in the Netherlands.
These interventions took place in the road environment and
generally used road traffic accident data as their outcome
measures. Other outcome measures employed by some
studies included costs, observed speed reduction, traffic
volume and residents’ opinions.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
Most studies targeted the general population; three either
targeted children or presented a detailed breakdown of
data with regard to age. The other studies did not present
any breakdown with respect to age, but as children (along
with older people) are the most vulnerable road users it
was felt that reported outcomes for interventions targeting
the general population would be particularly applicable to
children. It was therefore decided to include these studies here.
Intervention design
In the road transport policy intervention, Boxall (1988)
measured differences in road traffic accidents following the
introduction of school crossing patrols. The other studies
describe the impact of introducing packages of engineering
measures on recorded injuries in the road environment. These
measures included the introduction of cycling tracks, speed
reduction measures (20 mph zones, speed humps etc) and
traffic redistribution schemes.
Key results
Where considered, the results of the studies generally report
favourable results for the reduction of child casualties. Boxall
(1988) suggests that the presence of school crossing patrols
can reduce the number of injuries to children. Webster and
Mackie (1996) report that the introduction of traffic-calming
measures specifically designed to lower speeds reduced child
pedestrian injuries by 70%, showing child pedestrians to be
the greatest beneficiaries of these engineering measures.
Ward and colleagues report that child cyclists as well as
Have injury intervention studies addressed inequalities?
pedestrians can benefit from engineering measures aimed at
increasing road safety (Ward et al 1989a–c).
Gender
All the studies reviewed targeted both genders. None of the
studies considered the possible differential impact of the
interventions on children of different genders, or reported
injury rates separately for males and females.
Social and economic groups
As the studies generally involved area-wide interventions,
it is likely that social and economic factors were at work in
the interventions. However, none of the studies specifically
considered these factors in either the design or execution of
the work.
Culture/ethnicity
Similarly, no specific targeting of cultural groups or reporting
of results in relation to cultural groups was reported in the
studies. It could be noted though that the engineering
measures appeared to be broadly successful in reducing
casualty rates whether they were implemented in the UK, the
Netherlands or Denmark.
Place
Place can be an important factor in accident involvement. All
seven of the studies involved manipulating the environment
in some way, whether by providing school crossing patrols
or through more engineering-related measures. Trafficcalming measures were typically undertaken in urban settings
(Webster and Mackie, 1996). We are not aware of direct
comparisons between the effectiveness of the interventions
in rural and urban settings. Where the location of the
intervention had been chosen on the basis of the perceived
need for accident reduction, observed speeds dropped and
accident rates were considerably reduced (Webster and
Mackie, 1996). Where reported, local residents’ reactions to
traffic-calming schemes were favourable.
Summary
• Six studies focused on interventions relating to trafficcalming measures. One study evaluated transport policy.
• Age: interventions were targeted either at the general
population, or specifically at children. Where considered,
the results of the studies generally report reductions in
child casualties.
• Gender: all the studies targeted both males and females.
None of the studies reported injury rates separately for
males and females.
25
• Social and economic factors: none of the studies
specifically considered these factors in either design or
execution.
• Culture/ethnicity: no specific targeting of cultural
groups, or reporting of results in relation to cultural
groups, was noted in the studies.
• Place: all seven of the studies involved modifying the
environment. We are not aware of comparisons between
rural and urban settings. Where the location of the
intervention was chosen on the basis of accident risk,
observed speeds decreased and accident rates were
considerably reduced.
C2(ii) Pedestrian and traffic education
INTRODUCTION
Twenty-three studies were identified that dealt with
pedestrian and traffic education. These studies were
concerned with either pedestrian skills training (11 studies),
traffic clubs (five) or other forms of traffic education (seven).
Ten studies were conducted in the UK, five in the USA,
three in Australia, two in Canada, and one each in the
Netherlands, Japan and Sweden.
The studies focusing on pedestrian skills training were
generally school-based. Six were experimental programmes
involving small numbers of children. The other five were
operational programmes targeting a wider range of
pedestrian skills and were incorporated in the school
curriculum and/or other community action. In terms of
methodology, three of these studies employed a simulated
road in training and testing children, one used videotaped
performance feedback, and seven used some form of
training that took place in the road environment, often
supplemented or compared with performance following
training using table-top models. Methods of intervention
included individual training (Ampofo-Boateng et al 1993),
group training (Thomson et al 1992), class training (Antaki
et al 1986) and mass media approaches (Preusser and
Blomberg, 1984).
The majority of studies relating to traffic clubs were homebased (although one was school-based). All studies assessed
the effectiveness of the traffic clubs by using some measure
of knowledge or behaviour. These measures were always
self-reported. Two studies also took measures of parental
knowledge and attitudes. One study also took measures
of mortality and morbidity to assess the impact of the
intervention.
26
All seven of the other traffic education papers involved
interventions that were school-based. Two of the studies
also included mass media campaigns, one of which was
aimed at the general population, not just children. Two
of the interventions were combined with engineering
measures, and one study used a traffic simulation game
to teach children the rules of appropriate safe traffic
behaviour.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
Only one of the 23 studies targeted the general population,
and this study also included some school training. All others
targeted children of various ages. Seven studies targeted five
year olds (including one school-based traffic club study). The
other four traffic club studies generally targeted the underfives, although one intervention did follow children up to the
age of seven. Of the remaining studies, six targeted children
under 10 years old, three studies targeted children under 13,
and two studies targeted children under 18.
Intervention design
When designing the interventions, a large proportion of
the studies paid specific attention to ensuring the materials
used, and the design as a whole, were appropriate to the
age group targeted. For example, Tucker (1992) reports a
traffic club intervention that mailed a series of five books
to parents of targeted children. The contents of the books
were synchronised with the target age at each mailing. In a
number of school-based studies training was integrated into
the curriculum for children of varying ages (Rivara et al 1991;
Harland and Tucker, 1994). In these studies safety training
was delivered using age-appropriate methods and materials,
ranging from colouring-in tasks to class discussions and
workbooks. Cross et al (2000) developed materials designed
to be read aloud in class to assist less-able students.
The use of simulated roads instead of real roadside training
was often seen as a way to train young children to cross
the road safely, without their lack of ability causing them
increased risk in the process. Those studies that did involve
roadside training were well supervised and used trained
individuals to reduce risks to children arising from their
lack of knowledge and poor performance on crossing tasks
(Thomson and Whelan, 1997).
As well as developing interventions that considered the
abilities of the target age group, a number of studies also
took into account the differing needs of children at different
Injuries in children aged 0–14 years and inequalities
ages. For example, Preusser and Blomberg (1984) reported
epidemiological data suggesting that 71% of mid-block
dart-and-dash accidents involved children under nine. In their
intervention they specifically targeted this age group and
accident type, using a film featuring a cartoon character. In
a later study, Preusser and Lund (1988) targeted 9–12 year
olds. Recognising that older children were more independent
and interacted with the road environment in a different
way, and were often involved in more complex road traffic
situations, they developed a new safety film aimed at the
needs and attitudes of older children.
Key results
In general, the interventions reported had a positive effect
on children’s road safety behaviour. One study reported
that following the intervention the five year old recipients
were performing at the level of 11 year olds (AmpofoBoateng et al 1993). Antaki (1986) reported an unsuccessful
intervention, but noted that children improve during
their first year at school. In a study looking at pre-school
membership of a traffic club for children aged 2.5–5 years,
Downing (1981) reported that membership was highest
when children were four years old. It is possible that marked
increases in ability at five years may lead parents to move
away from traffic club membership.
Although most studies were targeted at particular age
groups, relatively few compared results for children of
different ages. Rivara et al (1991) reported improved
performance across all age groups taking part in the
intervention. Preusser and Blomberg (1984) reported accident
reductions of 21% in children under 15 years, and 31% in
children aged 4–6 years. Similarly, Tziotis (1994) reported
that benefits were larger for primary school-aged children.
A study by Wright et al (1995) found that benefits in terms
of increased knowledge were present in only younger
children, with no increases observed for older children. This
is a worrying finding as studies focusing on older children
suggest that older children were more likely to exhibit risky
behaviour (Burke et al 1996). Preusser and Lund (1988)
showed post-intervention increases in knowledge and
reductions in accident involvement in children aged 9–12
years, but on balance the studies suggest that older children
and young people are harder to reach.
Gender
INJURY TARGET GROUPS AND SETTINGS
Boys are more likely than girls to be involved in road traffic
accidents (Burke et al 1996). This effect also interacts with
age – peak accident involvement for boys occurs between
Have injury intervention studies addressed inequalities?
7–8 years, and for girls occurs around 12 years (Tucker, 1992).
Accidents for young boys are particularly high during the
afternoon, in the post-school 3–5 pm period (Tziotis, 1994).
Intervention design
Despite the increased risk for boys, no studies directly address
this gender difference in their interventions. This is can be
seen as a failure to target the most vulnerable road users,
and warrants further research.
Key results
West et al (1993) found that young boys were more likely
than their female counterparts to run into the road, less
likely to stop when told by a parent, and less likely to hold
a parent’s hand. They were more likely to cross roads on
their own and, possibly as a result of this, were more likely
to have knowledge of roadside objects. These results may
go some way to explaining boys’ increased level of accident
involvement. It is possible that different age peaks noted
in male and female accident involvement may be due to
differences in the level of independence at different ages.
In studies that compared impacts in relation to gender, it
was found that intervention programmes were effective
in improving safety performance in both boys and girls
(Thomson et al 1992; Burke et al 1996; Thomson and
Whelan, 1997).
Social and economic groups
Injury target groups and settings
Thomson and Whelan (1997) specifically targeted children
attending local schools which served a poor housing estate
with high levels of pedestrian injury. Rivara et al (1991)
studied a school in which 40% of students were recipients
of free school meals. Some studies reported wide socioeconomic spread among their target populations (Downing,
1981; Harland and Tucker, 1994). But the majority of studies
did not take socio-economic status into consideration.
It was suggested that traffic club members may have
a better socio-economic situation than non-members
(Gregersen and Nolen, 1994), but Downing (1981) ensured
that representatives of all social groups were included in a
UK-based study of traffic clubs. Similarly, West et al (1993)
included measures of social factors in their assessment of
knowledge, behaviour and attitudes following participation
in a traffic club. Cross et al (2000) compared the impact of
interventions on a school of mainly middle-to-high socioeconomic status with that of a mainly lower status school.
Intervention design
It was not clear that traffic club materials are designed
27
to take into account the socio-economic background of
participants. However, the provision of free materials may
help overcome the barriers of cost.
Key results
Children in more disadvantaged social groups were more
likely to cycle unsupervised, play in the street unsupervised
and cross roads by themselves (West et al 1993). They were
also less likely to recognise roadside objects or to know
safe places to play. It is encouraging to note, then, that the
programme implemented by Thomson and Whelan (1997)
proved effective in reducing ‘very unsafe’ crossing behaviour
in their targeted high-risk setting. Cross et al (2000) reported
increased safer behaviour in both high and low socioeconomic intervention groups.
Downing (1981) found that membership of traffic clubs
was indeed lower in lower socio-economic groups, and that
mothers in these groups were less likely to have understood
the material and less likely to have acted on it. In terms
of recruiting membership, parents in lower social groups
were more likely to have heard of the programme following
television advertising. Press and radio advertising had little
impact on reaching them.
Culture/ethnicity
Few studies considered cultural factors in their interventions.
Rivara et al (1991) reported an intervention in an ethnically
diverse school. Parental involvement was specifically
encouraged through the shared completion of parent–child
activity workbooks. This approach was found to improve
safe crossing performance in younger children. Thomson and
Whelan (1997) involved parents by training volunteers to
act as trainers for the children involved in the intervention,
with notable success. Cross et al (2000) also involved the
wider community by enlisting community participation in
determining road safety priorities to be addressed in the
intervention.
Place
Injury target groups and settings
Place can have a large impact on road injuries. Factors such
as traffic volume, local speed restrictions and engineering
measures all play a part in accident figures. The majority of
studies were carried out in an urban setting, although two
interventions were in rural, urban and suburban settings
(Harland and Tucker, 1994; Wright et al 1995). A number
of studies compared different geographic regions, although
this was generally done to compare an intervention region
with a similar non-intervention region acting as a control,
28
rather than to address differences between regions (Tucker,
1992, West et al 1993). At a more local level, seven studies
employed some training in the local road environment.
Thomson and Whelan (1997) specifically chose to perform
roadside training in the local environment to provide a
familiar backdrop for the trained children.
Intervention design
Two studies were specifically designed to be implemented
in regions with high levels of pedestrian injury (Tziotis,
1994; Thomson and Whelan, 1997). In the Tziotis study
municipalities were selected for the intervention on this basis.
In an attempt to compare knowledge and attitudes across
three regions of a single city, Downing (1981) measured
these factors in three regions of London: inner London, outer
north London and outer south London.
Key results
Few reported results were related to place. No differences
were observed in performance between rural, urban and
suburban settings. Both studies targeted at regions of high
accident incidence showed benefits for at least some of the
target groups.
SUMMARY
• Twenty-three studies focused on pedestrian and traffic
education, including pedestrian skills training (11 studies),
traffic clubs (five) and other forms of traffic education
(seven).
• Age: studies targeted different age groups. Traffic clubs
generally targeted pre-school children. Studies comparing
the performance of different age groups suggest that
some interventions were more successful at reaching
younger rather than older children.
• Gender: boys were more likely than girls to engage
in risky behaviour. Where gender was considered,
interventions were effective in improving performance in
both boys and girls.
• Social and economic factors: few studies addressed
socio-economic factors, but those studies that did so
proved to be effective interventions.
• Culture/ethnicity: few studies considered cultural or
ethnic factors. Three focused on community approaches
to training/education.
• Place: few studies related to place. No differences were
observed in performance between rural, urban and
suburban settings. Two studies targeted regions of high
accident incidence. Both showed benefits for at least some
of the target groups.
Injuries in children aged 0–14 years and inequalities
C3 Car occupant injuries: child restraints
INTRODUCTION
Thirty-nine studies relating to use of child restraints were
identified and reviewed. These studies cover child restraint
loan schemes (nine studies), educational campaigns designed
to increase the use of child restraints and seatbelts (16
studies), legislation requiring the use of restraints (nine) and
enforcement of this legislation (five). The majority of the
studies (33) were conducted in the USA. Of the remaining
six, two were conducted in Sweden and one each in
Australia, New Zealand and the UK.
In relation to child restraint loan schemes and education, 11
studies were hospital-based, seven were based at the primary
healthcare level, and seven were school-based. Studies
examining restraint legislation generally looked at effects of
state-wide legislation (seven studies), although two studies
examined the impact of legislation on groups of states. In
the examination of enforcement of restraint legislation, four
studies were community-based and one was school-based.
A number of methodologies were employed to measure
the impact of interventions. Twenty-two studies used
observation of restraint use, six used reported use, four used
a combination of observation and report, and six used injury
outcome measures.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
Three studies targeted the general population; the remaining
studies targeted specific age groups. Most frequently
targeted were infants under one year old, with 11 studies
examining interventions aimed at this age group. Two studies
targeted children under two years old, six targeted children
aged 0–3, two targeted children aged 0–4, two targeted
children aged 0–5, and one targeted children aged 0–10.
The remaining studies targeted a range of different ages,
generally from 3–5 to between seven and 19. In most cases
the targeted age group was dictated by the setting of the
study, eg hospital-based studies were more likely to target
infants under one, whereas school-based studies were more
likely to target children over three.
Intervention design
Intervention design and implementation were generally
specific to the age of the child. Interventions relating to
newborn infants generally involved provision of pre- or postnatal educational materials to parents (particularly mothers).
Have injury intervention studies addressed inequalities?
Nine of these interventions included provision of free child
seats. Studies targeting older children generally employed
school-based educational materials aimed at children and
adults. Two of these studies employed reward schemes for
correct use of child restraints.
In the nine studies examining the impact of state-wide
legislation, seven compared morbidity rates before and
after legislation, one compared observed seatbelt wearing
rates before and after legislation, and one compared
reported wearing rates between states with different laws.
In the five studies of legislation enforcement, educational
and publicity campaigns of a mass media or communitybased nature were employed before or concurrently with
the intervention.
Key results
Results relating to correct restraint use were moderated by
child age. A number of studies concluded that restraint use
was more prevalent among younger children (particularly the
0–3 age range) than older children (Miller and Pless, 1977;
Streff et al 1992). This effect was mitigated if the adult driver
was wearing a seatbelt (Russell et al 1994).
Scherz (1976) suggested that the decision to obtain a car
seat was frequently taken when the infant was 7–8 months
old, which is of some concern as Reisinger et al (1981) note
that children are particularly vulnerable to injury at 0–3
months. Scherz (1976) goes on to state that most infants
who are secured in safety restraints at eight weeks continued
to be transported safely at 9–12 months, suggesting that
interventions aimed at early infancy can provide later
benefits.
Legislation requiring seatbelt use was shown to have a
significant positive effect on the number of children observed
wearing restraints, and a significant reduction in the number
of children killed or seriously injured as passengers in motor
vehicle accidents. In states where legislation targeted only
very young children (typically 0–3 years old) a small but
significant ‘spillover’ effect was seen in the increased rates of
seatbelt wearing for older children (Wagenaar and Webster,
1986; Agran et al 1987). Following the introduction of
legislation aimed at all children under 16, Margolis et al
(1996) report an immediate increase in wearing rates for
young children, but a more gradual change for older children
(and adults).
Both legislation and enforcement studies noted greater
increases for child restraint use if accompanied by
educational methods rather than by coercion alone
(Bowman et al 1987).
29
Gender
All studies targeted both boys and girls, and only one
considered gender differences in its outcomes. Morrow
(1989) reported that boys started with higher rates of
seatbelt wearing than girls but, unlike girls, did not
significantly increase wearing rates following an educational
intervention. With regard to adults, a number of studies were
targeted specifically at mothers. Wagenaar and Webster
(1986) reported that the impact of a seatbelt law did not vary
by the gender of the adult driver.
Social and economic groups
Injury target groups and settings
Relatively few studies specifically considered social and
economic status. Where this was considered, a few studies
commented only that the target group was predominantly
well educated and relatively affluent (Greenberg and
Coleman, 1982; Sowers-Hoag et al 1987). However a few
studies did specifically target low-income areas (Liberato et al
1989; Hazinski et al 1995; Hanfling et al 2000) while others
obtained socio-economic information (Miller and Pless, 1977;
Reisinger and Williams, 1978).
Intervention design
Few studies were specifically designed with socio-economic
status in mind. Social group was typically measured by
parental education, number of children, marital status and
free school meal provision. It could be argued that provision
of free child restraints was an economic intervention,
however only two studies where free seats had been
provided included the collection of information on social
group (Reisinger and Williams, 1978; Colletti, 1986).
Key results
The implementation of free or loaned child seat schemes
for newborn infants or very young children appears to
be effective in increasing the likelihood of children being
restrained, at least in the short term. Use of provided
restraints was more likely if mothers were older, had fewer
children, were married and were not receiving medical
assistance. In terms of the effectiveness of educating people
on the use of restraints, Miller and Pless (1977) report no
difference in restraint use between social groups, while
other studies suggest that those of a higher social group are
more likely to use restraints (Goodson et al 1985; Hazinski
et al 1995). Hazinski and colleagues report that with good
implementation of education programmes, increased
restraint usage among more disadvantaged social groups can
be achieved. No clear-cut picture emerges following studies
30
of legislation and enforcement, with restraint use variously
reported as being unrelated to social group (Wagenaar and
Webster, 1986) or related to some aspects of social group
(Russell et al 1994). Hanfling et al (2000) suggest that higher
social group is related to safer behaviour, but that interventions
are effective across all levels of socio-economic status.
Culture/ethnicity
Injury target groups and settings
Again, relatively few studies specifically considered cultural
factors. One (Reisinger and Williams, 1978) specifically
excluded non-English-speaking participants from the
intervention.
Intervention design
Hanfling et al (2000) attempted to promote a communitybased approach to increased seatbelt wearing by involving
community and church leaders and local political figures in
the educational intervention. Geddis and Pettengell (1982)
specifically tried to change the cultural perception of the
special nature of the newborn’s ride home from maternity
hospitals which suggested that the mother should carry the
infant, rather than using a restraint.
Key results
Of the studies that noted participants’ ethnicity, most did
not perform any comparisons of different ethnic groups.
Hanfling et al (2000) did examine the impact of ethnicity
on restraint use and found no differences between ethnic
groups. Hanfling et al (2000) also found that good results
could be obtained by tailoring an integrated intervention to
the community’s needs.
Place
Location was rarely specifically targeted, and was mainly
noted only in relation to other factors such as social group.
Location was associated with the catchment area of the
intervention base, rather than the content or targeting of
the intervention. Hazinski et al (1995) did compare rural and
urban settings, and suggested that behavioural change was
observed only in the urban setting.
SUMMARY
• Thirty-nine studies focused on the use of child restraints,
including restraint loan schemes (nine studies), educational
campaigns (16), legislation (nine) and enforcement (five).
• Age: target age group varied, with the majority targeting
0–1 year olds. Restraint use decreased in older children
relative to younger children. Legislation increased restraint
use.
Injuries in children aged 0–14 years and inequalities
• Gender: all the studies targeted both males and females.
One study reported higher initial levels of restraint use
for males, although this did not increase following
intervention.
• Social and economic factors: few studies targeted social
and economic factors. Free or loaned child seat schemes
were effective in increasing the likelihood of children
being restrained. Some evidence suggests that restraint
use is higher among higher social groups, but that lower
social groups can be effectively targeted.
• Culture/ethnicity: few studies considered cultural or ethnic
factors. Those that did found no difference in restraint
use.
• Place: few studies considered place. One study suggested
that behavioural change following intervention took place
in urban, but not rural settings.
C4 Bicycle injuries
INTRODUCTION
Twenty-eight studies relating to bicycle safety, helmet
promotion and bicycle helmet legislation are reported. These
studies relate to bicycle training (three studies), educational
programmes designed to increase the use of bicycle helmets
(19) and the impact of legislation requiring the use of
helmets (6).
The majority of the studies (14) were conducted in the USA.
Of the remaining studies, six were conducted in Canada, two
in each of the UK, Australia and New Zealand, and one in the
Netherlands and Sweden.
In relation to helmet promotion and cycle training, 17
studies were school-based (with one relating to a pre-school
enrichment programme), two were hospital-based, one
involved a primary healthcare setting, and two were mass
media campaigns aimed at communities. The six studies
examining bicycle helmet legislation looked at effects of the
introduction of state-wide legislation.
Most of the interventions included an educational
component. Three of the legislation interventions specifically
examined the interaction between legislation and education
and how this had an impact on helmet-wearing rates.
Several of the interventions included the provision of free or
discounted bicycle helmets.
The interventions were evaluated by a range of study
designs, including seven randomised controlled trials, 13
controlled trails without randomisation, and seven time-series
observation studies. Twenty studies included the collection of
Have injury intervention studies addressed inequalities?
data on observed helmet use; the remaining studies collected
self-reported measures of helmet wearing, hospital records
and helmet sales.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
A number of different age groups were targeted in the
studies. Four targeted children of primary school age or
younger, four targeted children aged 5–12 years, and five
targeted children aged between nine and 12/13 years. In
terms of legislation, four studies detailed legislation requiring
all children under 16 to wear a bicycle helmet, and two
described legislation requiring helmet use by all riders of any
age (including adults). The concentration of studies targeting
school-age children was in response to epidemiological
evidence suggesting that these children were the most at risk
of bicycle-related injuries. The majority of the studies were
school-based, so differences in specifically targeted ages were
often due to the intake of the particular school.
Intervention design
Most of the studies reported making use of age-appropriate
materials in the education of children with regard to helmet
use. Also, a number of studies attempted to address the
attitudes and concerns of children specific to their age.
The principal area addressed in this way was peer pressure
and the social acceptability of bicycle helmet wearing, an
important factor in helmet use.
Key results
A frequently reported finding was that older children
were far less likely to own or wear helmets than younger
children (Kim et al 1997; Ni et al 1997). Older children were
also less likely to believe that a helmet could be effective
in preventing head injury (Logan et al 1998). Moore and
Adair (1990) note that older children are more susceptible
to peer pressure, and it may require legislation as well as
education to get this age group to increase their rate of
helmet wearing. Following legislation and a decade of
education in the Australian state of Victoria, an increase in
teenage helmet-wearing rates was observed – however this
may have been due to a 46% decrease in the number of
teenage cyclists seen following the introduction of the law,
suggesting that a good number of young people would
rather not cycle than wear a helmet. Peer pressure is an
important barrier to the use of bicycle helmets, particularly
among older children, and this should be taken into
consideration when designing interventions.
31
Gender
Injury target groups and settings
Boys aged 10–14 years are at a higher risk than girls of death
as a result of a cycling injury. While a number of studies report
helmet-wearing rates differentiated by gender, few studies
have been carried out to address this inequality of injury.
Intervention design
Only one study specifically addressed gender in its design.
Wood and Milne (1988) particularly targeted negative
attitudes to bicycle helmet use among older boys in their
educational intervention. Many other studies reported helmet
use by gender, but did not tailor the intervention specifically
to target either boys or girls.
Key results
Reported gender differences in helmet-wearing rates vary
in the different studies. Some studies report no difference
in helmet-wearing rates between boys and girls (Logan et al
1998). One study reported that helmet use was more likely
in boys, but presented no evidence to substantiate this claim
(Puczynski and Marshall, 1992). Several studies reported that
post-intervention, girls were more likely to wear helmets
than boys (Kim et al 1997). Two studies (Parkin et al 1993,
1995) presented an interesting finding in relation to gender
differences and social group: in observation studies carried
out in high-income areas there was no difference in helmetwearing rates between boys and girls, but when observations
were carried out in low-income areas boys were significantly
less likely than girls to wear a helmet.
In a study relating to cycle training, Savill et al (1996) found
that boys scored significantly higher than girls in tests of
cycling knowledge. However, this increased knowledge
appears to be insufficient to prevent injury. It is possible that
increased knowledge is due to increased cycling exposure for
boys relative to girls.
Social and economic groups
Injury target groups and settings
A number of studies targeted groups with some form
of social or economic disadvantage. Most of the studies
targeted areas or schools with disadvantage based on
average family income. Farley et al (1996) defined a
municipality as poor if more than 20% of families were
below the poverty threshold. Other measures of social
disadvantage taken were parental education (Logan et al
1998), lunch assistance (Puczynski and Marshall, 1992),
number of lone parents (Parkin et al 1995) and percentage
of owned dwellings (Parkin et al 1995). A number of studies
32
compared the impact of interventions in low- and higherincome areas (Farley et al 1996).
Not all studies worked with disadvantaged social groups;
some were carried out in middle-class communities
(Cushman et al 1991a; Macknin and Medendorp, 1994).
Dannenberg et al (1993) reported a study comparing helmetwearing rates in middle-class communities with those
observed in upper-middle-class communities.
Intervention design
As well as addressing inequality of social group by targeting
schools or areas of low income and using them as bases
of education, or regions in which to base observational
studies, many studies also specifically addressed the barrier
of helmet cost. The issue of cost was flagged as being
an important potential barrier to helmet use by lowincome parents (Towner and Marvel, 1992). To address
this, helmets were given away free of charge, or discount
coupons were provided to reduce the cost of purchase
(Mock et al 1995).
It is possible that the provision of free helmets may in some
way reduce their perceived worth, and may possibly lead to
less use. An interesting variation on the give-away schemes
was described by Kim et al (1997) who suggested that, while
cost may be a barrier to low-income groups, placing a small
cost on the helmet may lead to greater value being placed
on it and therefore greater use . This was described as ‘copaying’ and generally involved a large subsidy to bring the
cost of the helmet down to US$3–5. In this study no eligible
child was refused a helmet, and those who could not afford
to co-pay were given helmets free of charge.
Key results
Trying to understand the role of social group in bicycle
helmet use is a complex task. The evidence from the studies
reviewed does not present a clear overall picture. Several
studies reported that helmet use was associated with higher
socio-economic group (Towner and Marvel, 1992), however
two reported that this was not the case and that helmet use
was not related to social economic group (Pendergrast et al
1992). Interestingly, Ni et al (1997) reported that helmet use
was correlated with household income before, but not after,
the intervention.
In terms of effectiveness of the interventions described, again
there are equivocal results. Logan et al (1998) found that
helmet use could be increased through intervention in a lowincome setting, unlike Parkin et al (1993) who reported no
increases in helmet use in low-income areas. In a later study
Parkin et al (1995) noted that in low-income areas, post-
Injuries in children aged 0–14 years and inequalities
intervention helmet ownership increased but helmet wearing
did not. Farley et al (1996) reported a programme that was
three times more effective in average/rich neighbourhoods
compared with poor ones.
These results suggest that although the cost of a helmet is a
real barrier to low-income families in lower social groups, it is
not the only factor at play in determining levels of helmet use.
Culture/ethnicity
Injury target groups and settings
The study of Abularrage et al (1997) specifically addressed
issues of culture and ethnicity. The study attempted
to develop a helmet-promotion programme that was
appropriate and implementable in a multi-racial community.
The study was carried out in a county of New York State,
selected because it was the most racially diverse county in
the USA. Similarly, Hendrickson and Becker (1998) used
schools’ level of minority representation as a central criterion
for inclusion in their intervention. Britt et al (1998) describe
a study in which 18% of children taking part had a primary
language other than English. A few other studies, which
targeted low-income areas rather than focusing on ethnicity,
noted that in their target communities there were high levels
of minority students (Puczynski and Marshall, 1992) or nonEnglish speakers (Parkin et al 1995).
Intervention design
The main design facet employed was the use of multi-lingual
materials translated into a language appropriate to the target
community. For example, Cushman et al (1991a) report a
Canadian study using educational materials in both English
and French; Logan et al (1998) report a US-based study in
which parental questionnaires were translated into Spanish
and Vietnamese. Perhaps the most surprising study is that
of Abularrage et al (1997) which used only English language
materials – while many other studies also employed Englishonly materials, it seems particularly strange that a study
deliberately targeting a community so ethnically diverse
should do so. They report that multi-lingual materials were
not available at the time of the study.
Key results
Where results are presented broken down by cultural/ethnic
group, differences are generally found. However, few
studies report their findings in this way. DiGuiseppi et al
(1989) and Dannenberg et al (1993) report that whites are
more likely to be observed wearing bicycle helmets than
cyclists of other ethnic groups. Similarly, Hendrickson and
Becker (1998) found some relationship between ethnicity
and helmet use. Again, Abularrage et al (1997) reported
Have injury intervention studies addressed inequalities?
that, overall, white children were more likely to be observed
wearing bicycle helmets, however they also present an
interesting finding relating to the differential impact of the
intervention. Following the intervention increased helmetwearing rates were observed in black and in white children,
but no increases were observed in Hispanic or Asian children.
Whether this is due to the use of English language materials
only, or to other cultural factors, is not clear.
In terms of ‘culture of safety’ it is worth noting that a
number of studies reported that children were more likely to
wear bicycle helmets if cycling with other helmeted children,
and particularly likely to be wearing them if accompanied by
an adult wearing a helmet. Hendrickson and Becker (1998)
noted that a factor that significantly increased helmetwearing rates in children was a mother’s encouragement to
do so. Finally, mention should be made of the introduction
of helmet legislation in Victoria, Australia which was
introduced in the context of promotion of a culture of safety.
The legislation was introduced following a decade spent
creating a favourable climate through educational measures,
parliamentary statements and related safety laws.
Place
Injury target groups and settings
Place was rarely the target of the intervention studies.
Hendrickson and Becker (1998) reported a study in which
they specifically targeted a rural setting because access to
trauma services was restricted, and because of this injury
prevention was particularly important. A study by Logan et al
(1998) also had a rural setting. The majority of other studies
specified urban or suburban locations, or compared these
with rural settings (Britt et al 1998).
Intervention design
Savill et al (1996) describe a study that ran training courses
over a variety of different areas to include a wide geographic
spread. A number of other studies used location as a basis
for comparisons of intervention effectiveness. This included
educational studies that used similar, non-intervention areas
as controls, and legislation studies comparing the observed
helmet rates or injury outcomes in areas where legislation
had been enacted with those in areas without legislation.
This latter study type was sometimes used to compare the
impact of legislation alone with regions where legislation had
been accompanied with educational measures (Macknin and
Medendorp, 1994).
Key results
The study by van Schagen and Brookhuis (1994) found that
children’s behaviour related to the road conditions they were
33
in. DiGuiseppi et al (1989) reported that helmet-wearing
rates were higher in parks than on streets. Two studies noted
that the geographic differences were moderated by social
and economic factors. Parkin et al (1995) found that in highincome areas helmet use increased in schoolyards relative
to recreational sites, but that this was not the case in lowincome areas. Farley et al (1996) reported an intervention
that was effective in raising helmet-wearing rates in all areas
of average-rich municipalities, but effective only on local
streets (a higher-risk environment) in low-income areas.
Wood and Milne (1988) suggested that helmet-wearing rates
were lower in rural environments, and Ekman et al (1997)
reported greatest post-intervention decreases in head injuries
in rural locations, suggesting that rural environments are
worthy of targeting.
In terms of legislation, Macknin and Medendorp (1994)
reported that legislation alone leads to higher observed
helmet-wearing rates relative to areas without legislation.
However, greater impact still was attained in areas that
combined legislation with a programme of education.
SUMMARY
• Twenty-eight studies focused on interventions relating to
training (three studies), educational programmes (19) and
legislation (6).
• Age: a number of different age groups were targeted.
Older children were less likely to wear bicycle helmets and
were less likely to have a positive attitude towards them.
• Gender: several studies reported that girls were more
likely to wear helmets, although a number of studies
found no gender difference. Two studies suggested that
gender differences may be moderated by social and
economic factors.
• Social and economic factors: a number of studies
addressed helmet cost. Results suggest that while cost
is a barrier to low-income families, other factors are also
involved.
• Culture/ethnicity: where results are reported in relation
to ethnicity, minority children are less likely to wear bicycle
helmets. Other cultural factors that effect helmet use
include mother’s encouragement and peer helmet use.
• Place: children’s behaviour and likelihood of helmet use
varies depending on the environment.
C5 Injuries in the home
C5(i) Prevention of falls and general injuries in
the home
INTRODUCTION
In this section we examine 18 studies focusing on
interventions to prevent a range of injuries in the home,
including falls, suffocation, scalds and bath-tub drowning.
In addition to this literature on general injuries there is a
relatively large number of studies focusing specifically on the
prevention of burns and scalds and poisonings. These studies
are considered in sections C5(ii) and C5(iii), respectively.
Of the 18 studies evaluating interventions to prevent general
home injuries, six were carried out in the UK, 10 in the USA
and one each in Canada and Australia.
Most of the interventions included parent counselling on
home safety, and 11 involved home visits with individual
advice on hazards. Several interventions included the
provision and/or fitting of home safety devices such as
electric socket covers. While most of the interventions
targeted a range of injuries, the intervention described by
Spiegel and Lindaman (1977) focused specifically on falls
from windows.
The interventions were evaluated by a range of study
designs. Twelve studies included random allocation to
intervention and control groups, however in some studies
randomisation was at group rather than individual level.
Seven of the studies included the collection of injury outcome
data, the remainder evaluated the impact of interventions
by using data on observed home hazards, tests of parents’
knowledge, or questionnaires eliciting information on
attitudes and behaviour.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
Three of these studies targeted children of all ages, in the
remainder specific age groups were selected. Children under
five years of age were specifically targeted in 13 studies.
This concentration on the youngest age group reflects the
epidemiological evidence which shows that pre-school
children are most at risk of injuries in the home.
In a study focusing on suffocation, which targeted all
children under 15 years, the relative age-related risk of
different types of suffocation and entrapment injuries was
34
Injuries in children aged 0–14 years and inequalities
set out (Kraus, 1985). Thus for fridge entrapments, children
were at greatest risk between the ages of two and seven
years. For suffocation from plastic bags children under two
were recognised at being at high risk, while babies aged 6–8
months were at high risk of cot suffocation.
The selection of target age groups related not only to relative
risk, but also to the settings in which the interventions were
delivered. Several interventions were delivered in well child
clinics or as part of routine child surveillance appointments
(Kelly et al 1987; Kendrick et al 1999). Here the target age
groups would reflect the ages of the clinic population, eg
babies and children under two years.
Intervention design
The design of interventions in these studies frequently related
to the age of the children targeted, although this may not
have been made explicit. In several studies, free devices were
provided which reflected the vulnerability of children under
five to specific types of injuries, eg the provision of cupboard
locks, poison stickers and an emetic for the prevention and
treatment of poisoning. In the intervention described by
Kendrick et al (1999), where advice was offered at specific
child surveillance appointments, advice was designed to
reflect the child’s developmental stage.
Most of these interventions targeted parents rather than
children, and offered counselling and advice which frequently
formed a part of existing service provision. The intervention
described by Johnston et al (2000) involved parents of
children enrolled on Head Start programmes, while another
study involved health visitor counselling and the loan of
home safety equipment (Thompson et al 1998). Home visits
to offer safety advice occurred as part of several of these
interventions. For parents of children under five years, advice
on child health, surveillance and home visits are a routine
part of a child’s healthcare, and therefore these approaches
may have been more acceptable to these families than they
would have been for parents of older children.
Key results
None of these studies reported results specifically relating to
the ages of children included. Results of the studies suggest
that counselling parents of pre-school children and home
assessments can achieve some positive results.
Gender
All these studies targeted both genders, and there were no
studies where outcomes for boys and girls were reported
separately. Relatively few of the studies collected information
on injury outcomes and where outcomes were collected, eg
Have injury intervention studies addressed inequalities?
parent knowledge and/or behaviour or hazard removal, the
gender of the child may not have been perceived as relevant.
Some authors recognised that boys and girls had different
levels of risk for some types of injury, but this was not taken
into account in the way the interventions were designed or in
the measurement of outcomes.
Social and economic groups
Injury target groups and settings
A relatively large proportion of the studies reviewed (14
out of 18) targeted groups with some type of social or
economic disadvantage. The way these groups were defined
and selected varied across studies. However, most targeted
groups or neighbourhoods with disadvantage, rather than
selecting individuals. For example, one of the UK studies
(Colver et al 1982) described a home safety campaign aimed
at families living in a deprived area of Newcastle. Relative
deprivation was defined by a number of factors including
the numbers in the area unemployed and receiving benefits.
In another UK study the target area was described as having
a large proportion of local authority housing. In a US study
involving counselling at home visits, teenaged, unmarried
mothers were targeted (Olds et al 1994).
In contrast, in an early study by Dershewitz and colleagues
the target group was described as well educated and affluent
(Dershewitz, 1979; Dershewitz and Williamson, 1977).
Intervention design
A relatively large proportion of the interventions evaluated
(9 out of 18) included the provision or offer of free safety
equipment. In several studies the equipment included lowcost items such as electric socket covers or cupboard locks.
Smoke alarms were offered as part of some interventions.
In the study described by Spiegel and Lindaman (1977)
16,000 window guards were distributed free to 4,200
families. In three of the UK studies involving the provision
of free or low-cost equipment the intervention was tailored
to the circumstances of the families involved. For example,
in the studies by Clamp and Kendrick (1998) and Kendrick
et al (1999) low-cost safety equipment was made available
to those in the intervention group receiving means-tested
state benefits. Similarly, the study by Thompson et al
(1998) involved the free loan of equipment to families
on low incomes. Two studies which did not target lowincome families also provided free devices (Dershewitz and
Williamson, 1977; King et al 1999).
While the provision of free or low-cost equipment may
address economic barriers to their purchase, it does raise a
number of issues. It is possible that providing free devices
35
may undermine the way they are perceived by parents.
Parents may be less motivated to use them than if they had
bought them themselves (even at a subsidised cost). Further,
it is not known whether several of the devices offered (eg
electric socket covers and cupboard locks) are effective in
reducing injuries. It is possible that providing such devices
may increase risk if parents believe they offer protection and
reduce their level of supervision.
In two studies that included advice on home safety (Colver
et al 1982; Campbell et al 2001) there was specific emphasis
on recommending changes that families could achieve at
no or low cost. In the study by Colver there was also advice
to families on their entitlement to state benefits to allow
them to purchase safety equipment. In a study targeting
teenaged mothers advice was given on employment
opportunities (Olds et al 1994). A study from Ireland used
lay workers from the community to visit homes, enabling
trust to be established and relationships built up (Mullan
and Smithson, 2000).
Key results
Understanding the results of these studies in relation to
health inequalities is not simple. The results of the studies
were mixed, however many reported some success in
reducing the number of hazards in children’s homes and in
increasing families’ awareness of safety issues and first aid.
Given that many of the studies were targeted at families
living in deprived areas, it is possible that health inequalities
can be addressed by targeting interventions at such groups.
As Colver put it:
‘Even severely disadvantaged families will respond to
health education if the education is appropriate’.
(Colver et al 1982)
Having said this, only one of the studies examined the
impact of the intervention in different social groups. Clamp
and Kendrick (1998) examined the effectiveness of the
intervention in families either receiving or not receiving
state benefits, and found that the intervention was equally
if not more effective in families in receipt of benefits. In the
remaining studies it is not clear whether the interventions
were more or less effective in different social groups.
Gallagher et al (1985) reported that the intervention was
more effective where families had to make no, or only
minor, behavioural changes (eg where devices were fitted)
when an intervention was targeted at a neighbourhood of
poor housing. However, a similar finding was reported by
Dershewitz (1979) in relation to an affluent, well educated
target group. Mullan and Smithson (2000) found that the
36
home-visiting programme appeared more successful in more
stable communities, where fewer tensions existed among
residents.
Culture/ethnicity
Injury target groups and settings
Relatively little attention was paid in these studies to cultural
issues. One intervention was specifically targeted at Hispanic
migrant families (Campbell et al 2001). The participants in
another study were described as predominantly of black
origin (74%) (Kelly et al 1987). In the study of window falls
(Spiegel and Lindaman, 1977) it was noted that fall victims
were frequently non-white. In a small number of other
studies the ethnic breakdown of the samples was described
but there was no explicit attempt to target any particular
cultural groups.
Intervention design
Little attention was paid to issues concerning language
and culture in most of these studies. In the study targeted
at Hispanic young people the programme was delivered
by bilingual, bicultural students. In the remaining studies
language barriers were not specifically mentioned. In five
studies there was a general recognition that, for health
promotion interventions to be acceptable to participants,
they had to be delivered sensitively by trusted workers
(Colver et al 1982; Gallagher et al 1985; Olds et al 1994;
Johnston et al 2000; Mullan and Smithson, 2000).
Key results
Little reference was made to culture or ethnicity in the results
of these studies. None of the studies compared penetration
and impact of the interventions in different cultural groups.
In the study targeted at Hispanic young people it was
noted that 79% of the follow-up questionnaires had been
completed in Spanish, and the authors concluded that the
programme had resulted in improved confidence and first-aid
skills among participants.
Place
Injury target groups and settings
In a sense, all the studies reviewed related to place in that
the focus was on the environment in the home and many of
the studies aimed to reduce environmental hazards. At the
macro-level there was reference to place in that interventions
were targeted at particular neighbourhoods, and this
frequently related to the relative deprivation of families living
in these areas. All but two of the studies were carried out in
urban areas – often in built up inner-city locations. The study
Injuries in children aged 0–14 years and inequalities
by Olds et al (1994) was carried out in a semi-rural area, and
those by Paul et al (1994) and Mullan and Smithson (2000)
included both urban and rural areas.
Intervention design
Many of the studies included the provision of devices and
equipment or home safety checks aiming to improve the
environment in the home. In the study by Spiegel and
Lindaman (1977) the intervention aimed to prevent window
falls, and occupants of high-rise tenement buildings had
window guards fitted to improve safety. The study by
Gallagher et al (1985) alluded to the difficulties some families
face in altering their environment to reduce risk. They gave
as an example the inability of tenants in rented properties to
change hot-water temperatures to reduce the risks of hotwater scalds. In one study the aim of the intervention was
to increase the availability of safety equipment in local retail
outlets (Paul et al 1994). However in the majority of these
studies little attention was paid to broader environmental
issues.
Key results
The results of many of these studies suggest that
interventions can achieve behavioural changes resulting in
environmental changes in the home that may reduce injury
risk. No studies compared families living in different types
of accommodation or in different types of neighbourhood.
Mullan and Smithson (2000) found that the programme was
more successful in rural than in urban settings. It was not
clear whether differences in the built environment made it
easier or more difficult for families to improve safety in the
home.
In a study that attempted to improve access to safety
equipment, there was no reduction in the number of hazards
recorded after the intervention period (Paul et al 1994).
C5(ii) Prevention of burns and scalds
Nineteen studies specifically focused on the prevention of
burns and scalds, and three of these evaluated the impact
of regulations regarding smoke alarms or product design.
Most of the studies (14) were carried out in the USA. Seven
focused on all age groups (including adults), the remainder
specifically targeted children.
The studies were carried out in a variety of settings: four in
childcare centres and six in schools, the remainder targeting
general population or neighbourhood groups. Eleven of the
studies included an educational component. The duration
of these educational sessions ranged from brief, one-off
counselling sessions to intensive, structured, school-based
Have injury intervention studies addressed inequalities?
campaigns with activities spread over several weeks. In several
studies free devices were provided. Free smoke alarms were
distributed in four major campaigns. One study included
provision of a free thermostatic device that controlled bath
hot-water temperatures (Fallat and Rengers, 1993).
The interventions were evaluated using a variety of
approaches. Seven of the studies were randomised trials.
Injury outcome data were collected in four studies. In
the remaining studies outcome data included hazard
identification (eg observed presence of operational smoke
alarms). Data on knowledge and awareness were collected in
seven studies.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
These studies targeted a range of age groups. The three
studies evaluating regulations had a population focus, but
may have had relevance to children in a specific age group.
For example, the product regulations examined by Sorensen
(1976) applied in particular to injuries in young children
under five.
The studies included several school-based campaigns which
all focused on children over five years, although the target
age group varied. Three studies aiming to prevent scald
injuries targeted very young children (under three), the group
most at risk of this type of injury (Fallat and Rengers, 1993;
Waller et al 1993; King et al 1999). In a study in a childcare
centre, children aged 3–5 years underwent an intensive
programme on burn and scald prevention which involved
30 hours of classroom-based activities over 18 weeks. An
intervention involving nurse counselling at well baby clinics
was aimed at parents of babies aged 6–12 months.
Intervention design
Few studies explicitly referred to how the age of the target
group had influenced the design of the intervention. Eckelt
et al (1985) mentioned that materials had been designed
to reflect the existing level of knowledge and the attention
span of the target audience. Similarly, in a programme
targeting very young children, materials and exercises had
been designed to take account of the abilities of children at
this age (McConnell et al 1996). A community-based study
involved school-based activities for children aged 5–8 years
and young people 15–17 years (McLoughlin et al 1982).
Here teachers, students, parents and psychologists were
all involved in designing materials to ensure that they were
appropriate for the age groups concerned.
37
Key results
Few studies examined the impact of the various interventions
with children of different ages. Harré and Coveney (2000)
revealed that while the effectiveness of their programme was
most pronounced with older children (who showed greater
increases in their knowledge scores), the programme also had
an effect on younger children. These authors concluded that
even children aged 7–8 years have the ability to encourage
changes in the home to increase safety. McConnell et
al (1996), in a study of young children (aged 3–5 years),
revealed that while pre-test scores were higher in older
children the impact of the programme was more pronounced
in younger children. In the school-based campaigns described
by McLoughlin et al (1979, 1982) children aged both 5–8
and 15–17 years increased their knowledge scores. Here the
campaign seemed equally effective for different age groups.
The results of the study by McLoughin et al (1985) examining
the impact of legislation requiring the installation of smoke
alarms revealed that the presence of a child under the age of
10 had no impact on whether or not families had a working
smoke alarm in their homes.
No other studies compared results of interventions for
children in different age groups. Under these circumstances it
is not known whether the positive results reported following
some interventions would be achieved with children of
different ages.
Gender
There was little information on the gender of participants
in these studies, or on whether gender had any influence
on the outcome of campaigns. Harré and Coveney (2000)
revealed that a school-based education programme was
equally effective with boys and girls.
Fallat and Rengers (1993) noted that while the gender of
children did not influence the results of a campaign involving
the fitting of a tap thermostat, their fathers were likely to
object to the devices and remove them.
Social and economic factors
Target groups and settings
Two of these studies evaluated the impact of interventions
with more affluent groups (McLoughlin et al 1985; Waller
et al 1993). Several of the remaining studies specifically
targeted more deprived groups or attempted to include a
range of social groups. Three studies including schools or
childcare centres attempted to select schools/centres serving
both higher- and lower-income populations (Eckelt et al
38
1985; McConnell et al 1996; Harré and Coveney, 2000).
Two of the smoke alarm give-away campaigns specifically
focused on more deprived communities (Mallonee et al 1996;
DiGuiseppi et al 1999).
Intervention design
Four studies provided smoke alarms free as part of
burn-prevention campaigns. In one UK study more than
20,000 alarms were distributed free to families in London
(DiGuiseppi et al 1999). Eighty per cent of those receiving
alarms were living in rented accommodation. In the study
described by Thomas et al (1984) families were provided
with discount coupons for smoke alarms to overcome some
of the economic barriers to their purchase. Free devices
were also provided as part of scald-prevention campaigns
(eg thermometers to check bath-water temperatures or
thermostats for bath taps). In a study involving the fitting
of bath-tap devices to control temperature the devices were
inefficient and unpopular – bath taps became clogged up
and devices were removed.
Smoke alarms are associated with reductions in burn injuries;
the effectiveness of some other devices (eg tap thermostats)
is not known.
Key results
Relatively few studies examined results in different social
groups. In those campaigns targeted at low-income groups
campaigns achieved some success, eg families provided with
smoke alarms tended to use them. The study of a smoke
alarm give-away campaign where more than 10,000 devices
were distributed revealed that 45% of alarms were still
functioning at follow-up four years after the initial campaign
(Mallonee et al 1996). A study examining legislation requiring
smoke alarms revealed that the presence of operational
devices was positively related to income and property values
(McLoughlin et al 1985). Waller et al (1993) reported similar
findings with regard to hot-water temperatures, with more
affluent homes having safer water temperatures.
One study compared the effects of a school-based campaign
in private and state schools and revealed modest increases in
knowledge in both schools, with slightly greater gains among
private school students (Eckelt et al 1985).
Culture/ethnicity
Target groups and settings
Relatively few studies targeted specific cultural groups
or provided any information on the cultural background
of participants. There were some exceptions. First, in a
campaign to reduce scald injuries in Australia, Vietnamese,
Injuries in children aged 0–14 years and inequalities
Arabic and Chinese families were specifically targeted (King
et al 1999). In a New Zealand study a scald-prevention
campaign was targeted at what was described as an
ethnically diverse population, which included people from
European, Maori, Pacific Island and Asian backgrounds
(Waller and Marshall, 1993; Waller et al 1993). A smoke
alarm give-away campaign in London was similarly targeted
at a multi-cultural population (DiGuiseppi et al 1999).
Intervention design
Those studies targeting culturally diverse groups attempted
to design the intervention to take account of this diversity.
For example, in the Australian study including a range
of ethnic groups materials were developed after initial
consultation with local community groups. Materials were
prepared in a variety of languages and the communication
channels used related to the cultural practices of different
ethnic groups (King et al 1999).
Several interventions were delivered in a way so as to be
acceptable to the target group and to avoid any stigma
associated with being selected for intervention. For
example, in the study examining the effect of smoke alarm
regulations, free devices were provided to non-compliant
households (McLoughlin et al 1985). Smoke alarm give-away
programmes utilised local community workers, members of
tenants’ groups or other trusted local people (rather than
health professionals) to distribute alarms (Mallonee et al
1996; DiGuiseppi et al 1999).
Key results
There was little information regarding the impact of
programmes in different cultural groups. Shults et al (1998)
reported that while a smoke alarm give-away campaign
included an area with a variety of cultural groups, nonEnglish-speaking families were excluded from the evaluation.
Harré and Coveney (2000) reported that a scald-prevention
campaign was effective in all cultural groups targeted.
One study employed bilingual interviewers to collect outcome
data and made separate analyses of results for different
cultural groups (King et al 1999). Here it was revealed that
different groups had different baseline levels of knowledge.
Examination of the penetration of the programme revealed
that Vietnamese and Arabic groups were more likely to have
seen campaign information in newspapers rather than via
other media (eg posters or television). Vietnamese families
were more likely than others to be aware of the campaign,
and this group had the greatest increases in knowledge after
the campaign.
Have injury intervention studies addressed inequalities?
None of the other studies provided any separate breakdown
of results for different cultural groups; therefore it is not
known whether programmes would be equally as effective
with people with different cultural backgrounds and
practices.
Place
The built environment is likely to have an important
bearing on burn and scald injuries, although this was not
made explicit in most of the studies reviewed. Regulations
relating to the environment in the home, eg requiring the
installation of smoke alarms, are difficult to enforce. The
study of regulations by McLoughlin et al (1985) revealed how
regulations were limited (eg applied to new housing only) to
increase compliance.
The ability of individuals to change their homes to
prevent burns and scalds may also be limited by the built
environment. For example, hot-water systems relying on
fire back boilers may be difficult to control. While tap
thermostats may offer a means of controlling temperatures
at specific points in the home, devices must be easy to use
and efficient, otherwise they will be removed.
None of the studies reported results for different locations
(eg urban versus rural) or for families living in different types
of accommodation. Without this knowledge it is difficult
to appreciate the importance of the built environment in
encouraging or inhibiting safe behaviour.
C5(iii) Prevention of poisoning
INTRODUCTION
In this section we examine eight studies focusing on the
prevention of poisoning. Six of these studies examined
educational approaches (aimed at children, parents or
health professionals) to increase knowledge and awareness
of poisoning hazards. Five of those six were carried out in
the USA and one in South Africa. Two studies focused on
the impact of poison packaging regulations on injury, one
carried out in the UK (Sibert et al 1985) and one in the USA
(Rodgers, 1996).
The packaging regulations were applied at population
level. The educational studies were carried out in a range
of settings. One intervention was delivered in schools (Liller
et al 1998), one in childcare centres (Krenzelok and Garber,
1981), and the remainder via healthcare settings. The
interventions ranged from an intensive programme with
lessons on poisoning spread over many weeks, to a single
lecture on hazards. The school and child-centre studies
39
aimed to raise awareness among children, those based in
healthcare settings were aimed at parents. One intervention
was designed to increase knowledge of poison prevention
among healthcare staff (Eaton-Jones et al 2000). Another
intervention focused on the hazards associated with storage
of paraffin for heaters, and involved education on the safe
storage of paraffin and the distribution of child-resistant
containers (Krug et al 1994).
Two studies involved random allocation to control and
intervention groups (Woolf et al 1987, 1992) to evaluate the
interventions. The remaining studies used before-and-after
designs. Injury outcome data were collected in the studies by
Krug et al (1994) and Woolf et al (1992). In the remaining
studies the impact of the intervention was assessed by
examining participants’ knowledge and behaviour following
the intervention.
Overall, the results of these studies suggest that educational
interventions can improve knowledge among children and
parents.
INEQUALITIES
Age
Injury target groups and settings
Both studies involving evaluation of poison packaging
regulations focused on the prevention of injury in children
under five years. The educational studies also tended to
focus on young children under five, the group most at risk
of poisoning. One exception was a study that involved a
single lecture to school-age children aged 5–9 years (Liller et
al 1998). In the studies focusing on younger children, one
intervention was aimed directly at children (Krenzelok and
Garber, 1981). Here children aged 30–60 months attending
day-care centres were taught to recognise poison stickers
and to be aware of what poison meant. Four studies aimed
to raise awareness among parents of young children, and
one was aimed at health professionals caring for children
between six months and five years (Eaton-Jones et al 2000).
Intervention design
In the studies focusing on safe packaging of poisons the
nature of the intervention was specifically designed to take
account of young children’s developmental stage. Similarly in
the study in South Africa, paraffin storage containers were
specifically designed so that young children would have
difficulty opening them (Krug et al 1994). The educational
interventions also recognised age-related factors in the way
the interventions were designed. Liller et al (1998) reported
using age-appropriate educational materials, and in the study
40
aimed at children in childcare centres lessons were structured
to take account of the cognitive abilities of these pre-school
children.
Key results
Little information is provided in these studies relating to
the age-related impact of interventions. The school-based
study by Liller et al (1998) involved testing knowledge in two
separate age groups (children aged 5–6 and 8–9 years), and
results suggested that knowledge scores increased in both
groups. The study in childcare centres also reported increases
in knowledge in young children.
Gender
The gender of children was not mentioned in these studies
in relation to the targeting and design of interventions or
in the results. It is not clear whether poison-prevention
interventions have a different impact on boys and girls.
Social and economic factors:
Injury target groups and settings
There was relatively little information given in these studies
on the social and economic background of participants. The
studies focusing on regulations included all social groups.
In the study described by Liller et al (1998) six schools were
included which were selected to represent a range of social/
economic groups, but little detail was provided. In one
study (Woolf et al 1987) the hospital served a low-income
population, with 42% of those attending being in receipt
of Medicaid and a further 18% with no medical insurance.
In contrast, in a study of poison-centre callers by the same
team the participants were described as being predominantly
well educated and affluent (Woolf et al 1992). The study of
an educational intervention for health professionals reported
that the participants mainly cared for families receiving
Medicaid (Eaton-Jones et al 2000).
Intervention design
Several of these studies involved distribution of free devices
to prevent poisoning. For example, in the study by Krug et al
(1994) 20,000 child-resistant containers for the safe storage
of paraffin were distributed to overcome economic barriers
to their purchase. One study involved the provision of free
cupboard locks to intervention families (Woolf et al 1992).
Three studies included free or discounted distribution of
ipecac, an emetic drug for administration after the ingestion
of certain types of poison. A problem with ipecac is that it
may be harmful if administered inappropriately. The study by
Eaton-Jones et al (2000) suggests that many parents would
use ipecac without consulting a health professional. As
Injuries in children aged 0–14 years and inequalities
discussed above, the provision of devices or equipment free
to families, while overcoming economic barriers to purchase,
is useful only if these devices are of proven effectiveness and
if families are motivated and have the knowledge and skills
required to use devices appropriately.
Key results
None of these studies examined the impact of interventions
in different social groups. It is not known whether poisonprevention interventions are more or less effective for
children from deprived or more affluent backgrounds.
Culture/ethnicity
Injury target groups and settings
There was little information on culture in these studies. The
study in South Africa targeted black families where families
used paraffin heaters in their homes (Krug et al 1994). In
other studies little information on the ethnic background of
participants was provided.
Intervention design
The provision of paraffin storage containers (Krug et al 1994)
was a response to a particular poisoning hazard faced by
children in families where paraffin was stored in the home
and where paraffin ingestion was known to be a cause of
serious injury. In this respect the intervention was designed to
take account of the practices of participants.
The hospital-based educational intervention reported by
Woolf et al (1987) reported that educational materials and
counselling were provided in both English and Spanish.
Key results
The South African study suggested that the intervention was
effective in reducing paraffin ingestion among the group
targeted (Krug et al 1994).
None of the studies provided any information on whether the
interventions were more or less effective in different cultural
groups.
Place
None of these studies provided information on the location
or environment that helps us understand the impact of place
on injury. The safe storage of poisons in the home is likely to
relate to environmental considerations.
Several studies give information on where they were carried
out but provide little knowledge on how the setting of the
study, or the setting in which injuries occur, was likely to have
influenced results.
Have injury intervention studies addressed inequalities?
SUMMARY
• Forty-five studies examined the prevention of injuries in
the home environment. Interventions included parent and
child education, smoke alarm give-away campaigns and
poison-prevention packaging.
• Age: most interventions targeted children under five
years. Little information is available on the effectiveness of
interventions in different age groups.
• Gender: little information is provided on the gender of
participants in these studies. It is not known whether
interventions are equally effective with boys and girls.
• Social and economic factors: a relatively large
proportion of these studies targeted groups with some
sort of social or economic disadvantage. Interventions
designed to take account of economic disadvantage
included the provision of free devices. Smoke alarm
give-away programmes targeted at disadvantaged
neighbourhoods achieved some success. There is little
information on the effect of programmes in different
social groups.
• Culture/ethnicity: little attention has been paid to
cultural issues.
• Place: most of these studies focused on the environment
in the homes of individual participants and aimed to
modify the environment at this level. There is little
information on the hazards associated with particular
types of housing or neighbourhoods. Some vulnerable
groups, such as those in shared accommodation, have
limited ability to change some features of the home
environment.
C6 Injuries in the leisure environment
INTRODUCTION
Here we examine seven studies focusing on injuries occurring
outdoors in the leisure environment. Three of these
studies examined the impact of interventions designed to
reduce injuries in children’s playgrounds, one examined an
intervention to reduce mushroom poisoning, one focused
on fireworks injuries, one on drowning, and one on sports
injuries.
The approaches used in the different studies varied
considerably. The three studies aiming to improve the safety
of playgrounds focused on environmental modifications
(changes to equipment or surfacing) to reduce hazards.
The remaining studies used educational approaches. An
intervention to reduce rugby injuries included the provision
of mouth guards, and the one to reduce drowning promoted
41
the use of life jackets bought through bulk-purchase schemes
to reduce cost. In the evaluations only two of the seven
studies included control groups, the rest used simple beforeand-after designs.
All the studies reported positive outcomes, with injury
reductions being reported in four studies and increases in
knowledge or hazard reduction in the remaining three.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
The nature of the interventions meant that they targeted
a range of different age groups. The target groups reflect
the settings in which the interventions took place as well
as children’s exposure to particular injury risks. In two of
the studies focusing on the reduction of school playground
hazards, the target groups were children attending primary
school (Withaneachi and Meehan, 1998; Roseveare et al
1999). The study aiming to reduce rugby injuries focused
on older children attending secondary school (Morton
and Burton, 1979). The target group identified as being
at risk in a study to prevent fireworks injuries was children
aged 10–14 years, while a mushroom poisoning campaign
(partly delivered through schools) aimed to reduce injuries
in children aged 6–11 years. Finally, a study to prevent
drowning by the promotion of life jackets focused on
children aged 1–14 years.
Intervention design
Little information was given on how the intervention was
designed to take account of the ages of children targeted.
In some studies (eg those focusing on school playground
injuries) the target age group was relatively narrow and the
interventions involved environmental modifications to reduce
the exposure of primary school-age children to hazards.
Exposure to play equipment and the ways children use it are
likely to be age-related.
In the study aiming to reduce mushroom poisoning the
materials were specifically designed for children aged 6–11
years.
Key results
Several of these studies provided information on results
relating to different age groups. A study to prevent
mushroom poisoning reported that children’s ages and
where they lived had an impact on their knowledge scores
(Malinowska-Cieslik and van der Borne, 1998). The greatest
increases in knowledge were observed in younger children
(6–7 years) living in villages and towns and in older children
42
(10–11 years) living in cities. The study of fireworks injuries
in Italy reported reductions in injury rates for all age groups,
but the decline in injuries was most dramatic in the younger
children (aged 10–12 years) targeted by this campaign
(D’Argenio et al 1996). Similarly, Bennett et al (1999)
reported more success with younger children in a campaign
to promote the use of life vests. Here, while there was a
general modest increase in the number of children reported
using life vests, younger children were more likely than older
children to use them.
Other studies did not provide a breakdown of results in relation
to children’s ages so it is not known whether the campaigns
were more or less effective with different age groups.
Gender
Little information relating to the gender of participants was
provided in these studies. In the study by Morton and Burton
(1979), focusing on rugby injuries, the target group was
exclusively boys attending secondary school. The study of
fireworks injuries in Italy also drew attention to the relatively
high risk to boys of sustaining fireworks injuries. Following
this campaign, while the overall number of injuries was
reported to have declined, the ratio of boys’ to girls’ injuries
worsened. Before the campaign the male to female injury
ratio was 9 : 1, following the campaign it was 15 : 1. This
suggests that the campaign may have had more impact
on girls. In contrast, the results of a campaign in Poland to
reduce mushroom poisoning revealed that at post-test boys
had greater awareness of the hazards associated with picking
and eating wild mushrooms compared to girls (MalinowskaCieslik and van der Borne, 1998).
Social and economic factors
There was little information on the social and economic
background of participants in these studies. In one study
aiming to reduce playground hazards, schools were selected
to include different social groups. The sample of schools
was stratified by size and the socio-economic status of the
areas they served, and both control and intervention schools
included schools serving relatively deprived as well as affluent
areas. No other details were provided, and it was not clear
whether the intervention was more or less successful in
schools in more or less affluent areas.
None of the remaining studies provided information on the
social background of participants, it was not clear whether
any interventions had been designed to take account of
the social background of participants, nor whether the
campaigns were more or less effective with different groups
of children.
Injuries in children aged 0–14 years and inequalities
Culture/ethnicity
Two of the studies were specifically designed to reflect the
cultural practices in particular areas. In a study in Naples, Italy
the focus was specifically on injuries occurring from fireworks
during New Year celebrations (D’Argenio et al 1996). Here a
range of community groups were involved, and in addition
to educational activities, street cleaning was carried out to
remove unexploded fireworks. This campaign was launched
in response to a local injury problem. Similarly, the campaign
in Poland (Malinowska-Cieslik and van der Borne, 1998) was
launched to reflect the particular problem associated with
poisoning due to the common practice of foraging for wild
mushrooms. Again, a range of community organisations was
involved and a variety of mechanisms used to promote safety
messages, eg buses on routes to popular mushroom-picking
areas displayed posters with safety messages. Both these
campaigns responded to local injury surveillance and results
were positive for the groups targeted.
The campaign to promote life jackets described by Bennett et
al (1999) included information in both Spanish and English.
No other studies provided information on the culture of
participants or provided a breakdown of findings for different
cultural groups.
•
•
•
•
•
to reduce playground injuries, drowning, fireworks injuries
and mushroom poisoning.
Age: target age groups varied; interventions to reduce
playground injuries targeted primary school-age children.
Information on effectiveness in different age groups
suggested that some interventions were most effective in
younger rather than older children.
Gender: little information was provided on the gender
of study participants. A campaign on firework injuries
may have been more effective with girls. In a study of
mushroom poisoning, increases in knowledge were
greater in boys.
Social and economic factors: no information was
provided on whether these interventions were more or less
effective in children from different social groups.
Culture/ethnicity: two studies described programmes
designed to address local injury issues. No information
was provided in any of the studies reviewed on whether
interventions were equally effective with children from
different cultural backgrounds.
Place: studies were carried out in a range of countries.
Three studies specifically addressed environmental hazards
in playgrounds.
C7 Mass media and training interventions
Place
These studies were carried out in a range of different
countries including Poland, Australia, the USA and New
Zealand. The focus of campaigns reflected the different
leisure activities and environments in different areas.
Geographical factors are known to influence injury exposure.
For example, the climate in some countries means that
outdoor swimming is relatively unpopular, exposure is
relatively low and the number of deaths correspondingly
small. In other areas, outdoor swimming is a common leisure
pursuit. Mushroom foraging is popular in woods, forests and
pastures in some parts of Europe but is less likely in more
built-up areas.
Only one of these studies compared the impact of the
intervention in different places. The Polish study of
mushroom poisoning (Malinowska-Cieslik and van der Borne,
1998) revealed that young children living in rural areas and
small towns increased their knowledge more than those
living in cities.
SUMMARY
• Seven studies focused on interventions to prevent injuries
in the leisure environment. These included interventions
Have injury intervention studies addressed inequalities?
INTRODUCTION
In this section we examine five studies using mass media
or professional training to promote general safety. Three of
these were carried out in the UK.
The interventions varied and included television safety
campaigns (Williams and Sibert, 1983; Sundelin et al 1996),
a children’s safety centre (Gielen et al 1996), a school-based
education programme (Frederick et al 2000), and training
in injury prevention for healthcare professionals (Marsh and
Kendrick, 1998). One study included the collection of injury
outcome data (Williams and Sibert, 1983). In the remaining
studies, information on knowledge and attitudes was used to
evaluate the impact of programmes.
Overall, studies reported some increases in knowledge
following interventions but the study involving the collection
of injury data revealed no impact on injury rates.
INEQUALITIES ADDRESSED
Overall these studies provided little information on
inequalities.
43
Age
SUMMARY
The ages targeted by the studies varied. School-based studies
targeted school-age children. Sundelin et al (1996) described
a mass media campaign aimed at children under six years.
The intervention aimed at healthcare professionals targeted
those working with young children (under five years).
• Five studies used mass media or training approaches to
promote safety. These included TV campaigns and schoolbased programmes.
• Age: programmes were targeted at a range of age
groups. No information was provided on the impact of
interventions on children of different ages.
• Gender: none of these studies provided information on
the gender of participants or on whether interventions
had a different impact on boys and girls.
• Social and economic factors: little information was
provided on the social background of participants, or on
whether programmes were equally effective in different
social groups.
• Culture/ethnicity: none of these studies addressed
cultural issues.
• Place: one study included children from urban and rural
schools, but results were not reported separately for
children from different areas.
There was little information on how interventions had been
designed to take account of children’s ages. One exception
was the description of the school-based programme
described by Frederick et al (2000) aimed at 10–11 year
old children. Here the programme was designed to form
an integrated part of the school curriculum. Delivered by
teachers using specially designed materials, the programme
aimed to meet age-specific learning objectives. The
intervention described by Gielen et al (1996) – a children’s
safety centre – was also designed to meet specific learning
objectives. None of the studies described the impact of the
programmes in different age groups.
Gender
None of these studies provided information on the gender of
participants or whether programmes had a different impact
on boys and girls.
Social and economic factors
Again little information was provided. As part of the
intervention described by Gielen et al (1996), discount
coupons for smoke alarms were offered to overcome
economic barriers to their purchase.
The impact of a mass media intervention in different social
groups was described by Sundelin et al (1996). Here it was
reported that the programmes on safety were more likely to
have been watched by two-parent rather than lone-parent
families. However, parents’ unemployment was positively
and significantly associated with the number of programmes
watched.
None of the other studies provided a breakdown of results in
relation to different social groups.
Culture/ethnicity
None of these studies addressed cultural issues.
Place
In the study reported by Gielen et al (1996) there was
an attempt to include schools from both urban and rural
locations in the evaluation. However, results were not
reported for different groups.
44
C8 Community-based interventions
INTRODUCTION
In this section we examine 10 studies that have evaluated the
effectiveness of community-based injury programmes and
that included elements related to child safety. Six of these
programmes were inspired by the ‘safe communities’ model
developed in Falköping, Sweden (Svanström et al 1995)
and include the use of community diagnosis using a local
surveillance system and a reference group to coordinate a
range of injury-prevention approaches.
The community-based approach can combine both ‘passive’
environmental measures and ‘active’ behavioural solutions.
The stimulation of a ‘common culture of safety allows the
adoption of complementary solutions which should allow a
multiplier effect to be achieved.’ (Moller, 1992b)
Three of the community-based programmes took place in
Scandinavia, three in the USA, two in Australia, and one
each in New Zealand and Greece. Important elements of
community-based approaches are a long-term strategy,
effectively focused leadership, multi-agency collaboration,
tailoring to the needs of the local community, the use of
local injury surveillance systems, and time to coordinate
existing networks and develop new ones (Towner and
Dowswell, 2002). Community-based approaches allow
injury-prevention messages to be repeated in different
forms and contexts.
Injuries in children aged 0–14 years and inequalities
The 10 studies all attempted to deliver a range of different
interventions in a defined geographic locality. In the Safe
Kids/Healthy Neighbourhoods Program, interventions
included environmental measures (renovation of playground),
the provision of safety equipment, violence and injury
prevention education, and the involvement of young people
in safe, supervised activities that taught them useful skills
(Davidson et al 1994; Kuhn et al 1994). The Shire of Bulla
Safe Living Program (Hennessey et al 1994) documented the
implementation of 113 programme components comprising
nine major objectives: traffic safety, children’s safety, schools,
sports safety, occupational safety, seniors’ safety, farm safety,
public places and community awareness.
None of the community-based studies reviewed included a
randomised controlled design, one used several interventions
and control communities, eight used one control community,
and one study compared outcomes in targeted and nontargeted injuries in the same community.
INEQUALITIES ADDRESSED
Age
Injury target groups and settings
Six studies were targeted at all age groups, with specific
components targeted at children. The Shire of Bulla Program
in Australia had 62 of 113 programmes specifically related
to children’s safety, and a higher proportion of these were
targeted at pre-school children (Hennessey et al 1994). The
Greek island programme targeted two vulnerable age
groups: young people and older adults (Petridou et al 1997).
The remaining three programmes targeted children: the
Lidköping Accident Prevention Programme (Svanström et al
1995) targeted children aged 0–14 years, the Safe Kids/
Healthy Neighbourhoods Program (Davidson et al 1994)
children aged 5–16 years, and the Statewide Childhood
Injury Prevention Program children under five (Guyer et al
1989).
Intervention design
All the studies targeted specific components of their
overall programme at children or their parents or carers,
and the content of the different interventions reflected the
predominant injury type affecting the age group in question.
The Statewide Childhood Injury Prevention Program in the
USA, targeted at pre-school children, selected the main injury
types affecting this group for which proven countermeasures
were available, and thus developed programmes aimed at
the prevention of burns, poisoning, falls, suffocation and
passenger road traffic accidents (Guyer et al 1989). The
Have injury intervention studies addressed inequalities?
Safe Block Project included specific interventions for preschool children such as syrup of ipecac (for poisoning) and
medication with childproof caps (Schwarz et al 1993). The
Safe Kids/Healthy Neighbourhoods Program (Davidson et
al 1994) concentrated on interventions to reduce outdoor
injuries and assaults in children aged 5–10 years, and the
Harstad WHO Safe Community Programme had a burnprevention programme targeted at pre-school children
(Ytterstad and Sogaard, 1995).
Key results
The Harstad programme reported that certain interventions
were more effective in specific age groups, eg in children
under 10 years the greatest reductions were seen in traffic
injury rate, pedestrian injury rate and increased use of cycle
helmets. The Latrobe Valley Better Health Project in Australia
(Day et al 1997) found that playground injuries declined in
0–5 year olds, and that sports injuries attending emergency
rooms increased for 0–14 year olds.
The Waitakere Community Injury Prevention Project aimed
to cover all age groups, but in practice focused on children.
Coggan et al (2000) concluded that ‘community-based injury
prevention projects should consider fewer, more targeted
interventions’.
Gender
All the studies reviewed targeted both genders, but none
reported that interventions had been specifically designed to
take account of the differences between boys and girls.
The Harstad WHO Safe Community Programme (Ytterstad
and Sogaard, 1995) reported that reduction in injury rates
among cyclists was greater for boys (43%) than girls (23%).
The Lidköping Accident Prevention Programme (Svanström
et al 1995) also reported a greater decrease in hospitalised
injuries in boys compared to girls, but it was unclear whether
the rates were restricted to target injuries.
Social and economic groups
Community-based injury-prevention programmes
have the potential to be particularly effective in more
deprived communities through their ability to strengthen
neighbourhoods. Both top-down and bottom-up approaches
can be used in such programmes. Bottom-up approaches
involve the target population in defining the problem and
finding solutions. Moller (1992b) believes that:
‘the aim is not only to produce effective solutions to a
single problem but to develop skills and commitment to
generalising what is learnt to other problems.’
45
He also suggests that bottom-up approaches are valuable
in dealing with diffuse problems in environments where
regulation and enforcement are difficult to achieve.
Injury target groups and settings
Two studies, both conducted in the USA, focused specifically
on disadvantaged communities. The Safe Block Project
(Schwarz et al 1993) was based in a poor African–American
inner-city community in Philadelphia, and the Safe Kids/
Healthy Neighbourhoods Program (Davidson et al 1994) in
an inner-city black community in Harlem. In the latter nearly
40% of residents were below the poverty level compared
with 19% in the city as a whole (Davidson et al 1994).
Intervention design
The interventions in the two disadvantaged communities
described above specifically attempted to increase access to
safety equipment by providing free or reduced-cost items.
For example the Safe Block Project provided smoke detectors,
batteries for smoke detectors, bathwater thermostats and
syrup of ipecac (for poisonings), and the Safe Kids/Healthy
Neighbourhoods Program provided more than 500 bicycle
helmets to reduce the barriers of cost and access to safety
equipment.
Key results
In the Safe Block Project (Schwarz et al 1993) the intervention
was partially effective for home hazards requiring minimal to
moderate effort to correct, but there were fewer differences
for home hazards that required a major effort to correct.
Culture/ethnicity
The central ethos of the community-based model is that
those most able to solve community problems are people
who live in that particular community (Coggan et al 2000).
Such an approach also provides an opportunity to change the
whole culture of safety in a community.
Injury target groups and settings
The two studies targeting disadvantaged communities
described above also related to communities with a high
proportion of minority ethnic groups. The Safe Block Project
(Schwarz et al 1993) was conducted in a predominantly
African-American community. The Safe Kids/Healthy
Neighbourhoods Program (Davidson et al 1994) included
a predominantly black community as its intervention
community and a predominantly Hispanic community as
a control. The design of this study meant that it was not
possible to determine differences between ethnic groups
in the two communities. The Waitakere Community Injury
Prevention Project (Coggan et al 2000) was based in a
46
multi-cultural community where Pakeha/European people
represented 67% of the population, Maori people 14%, and
Pacific people 11%. In the Australian Shire of Bulla study
(Hennessey et al 1994) 77% of residents in the community
were born in Australia.
Intervention design
The Waitakere Community Injury Prevention Project in New
Zealand (Coggan et al 2000) included different programmes
targeting different indigenous or ethnic groups. After
community consultation seven priority areas were identified,
two of which related to Maori and Pacific people, and working
parties were formed to identify priorities for each group. A
Maori coordinator was employed who was experienced in the
culture and perspectives of the target group. Injury prevention
was incorporated into a holistic view of health.
The Safe Block Project (Schwarz et al 1993) used a network
of community volunteers to conduct the intervention, and
used cascade training to disseminate the messages widely.
The Greek islands community injury prevention project
(Petridou et al 1997) also used local women to perform
weekly home visits and to advise on safety.
Many of the other projects actively involved the target groups
in a range of activities (Schelp, 1987; Davidson et al 1994).
Key results
The Safe Block Project showed that the approach of many
community representatives was acceptable to the community,
and community representatives were recruited for 88% of
blocks. The case study of the Waitakere Community Injury
Prevention Project showed that the approach adopted was
acceptable to community members and a considerable range
of projects was stimulated.
Place
Injury target groups and settings
Moller (1992a) defines community injury prevention as:
‘any project which seeks to work with a community,
whether it is defined by geography or common interest,
to identify injury problems and seek to find ways of
addressing these problems using means which are
generalised or at least agreed to, by the community.’
The notion of identification with a community or sharing
a feeling of commonality is an important factor (Moller,
1992b). All of the 10 studies reviewed took place in specific
communities, and in most one intervention and one control
group were selected.
Injuries in children aged 0–14 years and inequalities
In the one study with multiple intervention and control
communities, the Statewide Childhood Injury Prevention
Program in Massachusetts (Guyer et al 1989), nine cities were
selected from a potential 351 cities and towns and matched
for a number of relevant variables with five control cities.
However, the nature of the different intervention and control
communities was not elaborated on. In the Shire of Bulla
study (Hennessey et al 1994) a control community, the Shire
of Melton, was selected. This was a geographically separate
community served by different newspapers, with little
crossing of municipal boundaries for school, work, shopping
etc. The Swedish programme in Falköping (Schelp, 1987) had
one control community, Lidköping, an area of similar size and
socio-demographic mix to Falköping.
The existence of similar data collection systems in other areas
may govern the selection of control communities. In these
cases controls may be very different from intervention areas.
The Norwegian Harstad Programme (Ytterstad and Sogaard,
1995) was conducted in a small town of 20,000 people
north of the Arctic Circle, while the control community of
Trondheim was much larger and 1,000 kilometres away.
SUMMARY
Ten studies focused on community-based injury-prevention
programmes. These combined a range of passive
environmental measures and active behavioural solutions. A
coordinating group brought together a range of agendas.
This approach can allow messages to be repeated in different
forms and contexts.
• Place: all 10 studies took place in specific communities.
Intervention and control communities are often
demographically matched, but the nature of the
community is rarely given much attention.
C9 Key points from all the intervention
studies
• Few intervention studies explicitly address inequalities;
even fewer attempt to take account of inequalities in the
design of the intervention and report on any differential
impact in relation to them.
• Age: many interventions target specific age groups of
children, but there are few examples of results comparing
the impact of an intervention on different age groups.
• Gender: despite the great differences in injury rates in
boys and girls, few interventions have targeted this factor.
• Social and economic factors: interventions have
increasingly been targeted at deprived individuals or
groups. The main strategy adopted in interventions is the
provision of safety equipment.
• Culture/ethnicity: cultural differences are rarely
addressed. A few interventions have involved the target
group in the design of the intervention. Few studies
compare interventions in different ethnic groups.
• Place: when communities have been matched in
controlled interventions, broad demographic variables
have been used and little attention paid to the context.
• Age: six studies were targeted at all age groups, with
special programmes directed at children. One programme
targeted two vulnerable age groups and three were
directed at children. Two programmes provided some
evidence that they had been effective with specific age
groups.
• Gender: all 10 programmes targeted both genders, but
none reported different tailoring of content for boys and
girls. Two studies reported greater improvements for boys
compared with girls.
• Social and economic factors: two studies focused
particularly on disadvantaged neighbourhoods. Free or
reduced-cost safety equipment was provided.
• Culture/ethnicity: three studies targeted communities
with high priorities for ethnic/indigenous groups. One
project based in a multi-cultural community included a
separate intervention designed with the Maori community
in mind and with a Maori coordinator.
Have injury intervention studies addressed inequalities?
47
D: Discussion and recommendations
D1 Discussion
The key points are drawn together at the end of each section
A–C. Below we discuss how definitions need to be improved,
and how observational and intervention studies could be
improved to address inequalities. Finally, we consider the role
of targeted and universal approaches to injury prevention.
Improving definitions
48
intervention studies, there is little consensus about what
social deprivation means. Definitions of social deprivation
can be based on the characteristics of geographic areas
such as inner-city wards; on the characteristics of individuals
themselves or the places where they live, such as ‘lowincome’ families or those living in multi-occupancy housing;
or on the characteristics of schools, childcare centres
or medical settings serving more deprived communities
(Dowswell and Towner, 2002).
A recurring theme of this review relates to the lack of
consistency of definitions, both in how injuries are defined,
and in the definition of other terms. The way injuries
themselves are defined may reflect health service utilisation
rather than a particular severity level of an injury. A child’s
age may influence whether or not parents seek medical
attention for injury, and also the medical management of
that injury. Thus a pre-school child with a mild head injury
may be admitted to hospital for observation, while an
older child may not. Similarly, a child from a disadvantaged
background may be admitted to hospital, while a child
with a similar injury from a more affluent background
may not. Clear definitions of injury are fundamental to
understanding inequalities between groups. The report to
the Accidental Injury Task Force, Measuring and monitoring
injury (Department of Health, 2002b), recommends that a
minimum data set should be established to include both the
accident and personal characteristics of each person injured.
We support this recommendation for clearer definitions so
that agencies collecting data relevant to unintentional injury
can adopt a more coordinated approach.
More attention needs to be paid to definitions and how
they affect study findings. Greater consistency in definition
across studies, particularly in the international literature, may
help us to understand why a programme works in specific
circumstances or localities and not in others, and how
transportable it is to other settings.
This report also comments on the problems of defining
terms such as ‘ethnic group’ and ‘homelessness’. Even in
one country, these definitions can vary over time. When
comparing studies from the international literature it may
be even more difficult to obtain comparable definitions.
In section B on the causes of injury and section C on
Children of different ages and stages of development have
different patterns of injury types and severities. Thomson et
al (1996) reviewed the role of child development in relation
to pedestrian injuries, but a similar detailed review has not
been conducted for other injury types.
Observational studies related to inequalities in
childhood injury
This report examines a range of factors that may influence
injury risk. For some factors – age, gender and socioeconomic group – the patterns have been well described
at a variety of geographic scales. For other factors – ethnic
groups, vulnerable groups such as children from homeless
families, disabled children – it is difficult to obtain a national
picture in the UK and evidence can be obtained only from
small-scale studies. Few studies have examined broader
cultural variations. For all factors, the reasons why the
differences occur are even less clear.
Age
Injuries in children aged 0–14 years and inequalities
Gender
Place
Boys and girls have distinct patterns of injury, with boys
generally having much higher rates of injury. A detailed
review drawing on other areas of health promotion in
relation to gender differences needs to be conducted.
The idea of ‘place’ adopted by most epidemiological studies
of injury is that of place as a portion of geographical space.
Patterns of injury can be identified that relate to factors such
as the distribution of more affluent or more deprived wards
or enumeration districts. Fewer studies have incorporated the
physical or environmental attributes of places (eg different
layouts of roads or housing types). The identification and
mapping of specific hazards in the environment could be
conducted by members of the local community and by
professionals, eg checklists of a child’s journey to school
or of playgrounds. The third idea of place, as an area to
which people attach certain meanings or actions, has been
little explored in the field of analytical injury epidemiology
research. Section C8, on community-based injury-prevention
programmes, shows that place has a potential role to
play in understanding the differences between different
communities.
Social and economic factors
Increasing attention is being paid to why children from
different social and economic groups have very different
rates of injury, but the complex causal pathways are being
unravelled only slowly. Many indices of social and economic
deprivation correlate with accidents because they increase
exposure, decrease supervision and undermine children’s
ability to deal with hazards.
The evidence related to causal mechanisms is patchy and
unbalanced, with more studies being conducted on the
individual characteristics of children and parents, and fewer
on social and environmental influences. Broader studies are
needed in this area, such as the case–control study of child
pedestrian injuries conducted by Christie (1995a,b) dissecting
both behavioural and structural components.
From these broader studies it would be possible to examine
whether factors such as family income, maternal education
and overcrowding independently predict risk. Further
application of multivariate analyses which have been
conducted in some studies could also help in identifying
dimensions of socio-economic position which had particularly
powerful effects on injury risk.
Culture/ethnicity
In relation to minority ethnic groups, the reasons why
children from these groups suffer a disproportionately
higher injury rate than majority culture peers has often been
explained in terms of socio-economic differences. Thomson
et al (2001) examined the involvement of children from
minority ethnic groups in road accidents, and hypothesise
that other factors may come into play that cause additional
risk for children from these groups. These issues need to be
explored more fully in relation both to the road environment
and other settings in which children are exposed to hazards.
For such studies to provide meaningful results they need
to recognise that factors that make up ethnicity, such as
language, religion, ancestry and forms of identity, are
constantly changing and that it may not be possible to draw
a distinct boundary around categories such as ‘Indian’ or
‘Asian’.
Discussion and recommendations
What emerges in section B of this review is that the causes
of injury are multi-faceted and inter-related. Factors such
as age, gender, social and economic groups, cultural and
ethnic groups and place do not in themselves cause injury.
Instead they are closely associated with more direct factors
that include children’s exposure to hazards, parents/carers’
and communities’ ability to protect children, and children’s
capacity to manage and avoid hazards.
The lack of clarity in the examination of causal pathways
has had an impact on the development and evaluation of
interventions in the field of childhood injury prevention.
Developing effective interventions to address
inequalities in childhood injury
How is it possible to develop injury interventions when
the causes of injury are multi-factoral and inter-related?
Graham (1999) asserts that ‘multiple causes point to the
need for wide ranging policy solutions’. In section B2 we
discussed three tiers of factors: the proximate, intermediate
and ultimate tiers that underlie injury. The Independent
inquiry into inequalities in health (Acheson, 1998), discussing
the ultimate tier of social, political, cultural and economic
factors, argues that the more a potential intervention relates
to the wider determinants of inequalities in health, the less
possibility there is of using the methodologies of a controlled
trial to evaluate it. This is equally true in the field of injury
prevention, and the intervention studies discussed in section
C relate to the proximate and intermediate tiers, not the
ultimate tier. The latter has been little addressed in injury
49
prevention through controlled experiments. Whitehead et al
(2000) emphasise that macro-level policies cannot normally
be subjected to controlled experiments and that tools and
methodologies to do so are underdeveloped. They suggest
some pragmatic solutions to this problem, including studying
a cross-section of localities in which different levels of policy
on a specific issue are in operation; studying changes over
time in one country where different levels of policy on a
specific issue are in operation; and studying the period before
and after the introduction of a new policy in several countries.
In section C it is clear that few intervention studies
systematically set out the problem in relation to age or
social and economic group or design the intervention with
this factor in mind, then report whether there has been a
differential impact in relation to that factor. Instead what
emerges is a patchwork of examples – the threads are not
woven throughout the studies.
Many evaluated studies do not provide a full description
of participants to allow us to interpret results in relation
to different subgroups (children of different ages or social
groups). Information on programme reach is also limited –
are all groups exposed to interventions in a similar way? Such
process information is essential in helping us to understand
what elements of an intervention are successful, and how
positive results in one particular area could be transferred
to other settings. Pawson and Myhill (2001) believe that
instead of simply asking what works in injury prevention, we
should pose the question ‘What works for whom, in what
circumstances, and in what particular respects?’. It is thus
essential that more attention is paid to the context of where
and how an intervention takes place.
In this review there are few examples of the relative
effectiveness of injury interventions on different subgroups
of the population. Moller (1997) reminds us that ‘Marketing
strategies are aware that products need to be designed to
appeal to different target groups’.
In section C, when considering whether interventions have
been designed with, for example, social and economic
factors in mind, relatively few strategies emerge. The main
emphasis is on the use of low-cost or free items of safety
equipment, but even in this case, what is the most effective
approach? Free items of safety equipment may not be
valued by recipients, and a small charge may be more
effective (Kim et al 1997).
Involving different target groups in the development of
an intervention has been under-utilised in the injury field.
Programmes may be piloted with the target audience, but
50
more actively involving children and their parents in the
design of interventions and their evaluation needs to be
greatly encouraged. Roberts (2000), in her review of What
works in reducing inequalities in child health, believes that
‘parents and children living in particular environments are
experts in identifying local risks’ and that effective injury
prevention needs to draw more on ‘the specialist local
knowledge of children and parents’. The role of qualitative
studies in contributing to the evidence base in this area is
important. This report does not systematically review the
contribution that qualitative studies could make in the field
of inequalities and injury, but a number of examples are
quoted. Qualitative research studies could help illuminate
some of the mechanisms whereby age, gender, social and
economic position, culture/ethnicity and place influence risk.
The experiences of parents and children are also fundamental
in designing and implementing interventions to reduce
risk. Such perspectives would provide valuable insights into
why potentially useful interventions work or do not work in
specific contexts.
Targeted and universal approaches to injury
prevention
Finally, in a review of inequalities in childhood injury
prevention we must consider whether injury prevention
should be targeted at high-risk groups, or whether it should
focus on the wider population.
There has been considerable debate about universal
(population-based) versus targeted or high-risk approaches
in relation to health and welfare, but little in the injury field
until recently. The journal Injury Prevention (Volume 3, Issue
3, 1997) contains three articles and an editorial comment on
this subject. Roberts (2000), in her review of inequalities in
child health, also discusses the different approaches. Kendrick
and Marsh (1997), in a study on general practice in the UK,
suggest that injury prevention targeted at higher-risk groups
is not worthwhile and that a universal approach is preferable.
Roberts (2000) agrees, arguing that universal approaches are
non-stigmatising and affect a great number. Moller, on the
other hand, disagrees, asserting:
‘A universal approach to injury prevention will only
exacerbate the problem of the disadvantaged. Targeted
programs are necessary to ensure that those health
differences between the rich and poor do not increase.’
(Moller, 1997)
This debate has mainly been conducted in relation to social
and economic factors, but it is equally relevant to the other
inequalities considered in this review.
Injuries in children aged 0–14 years and inequalities
Pless (1997), in his editorial comment in Injury Prevention,
suggests that both population and high-risk approaches
could co-exist, citing the example of:
‘A nationwide campaign to reduce speeds, which if
successful would benefit everyone, alongside more
rigorous enforcement of reduced speed limits in poor
areas.’
We agree with Pless’s dual approach, and in this review we
highlight ways in which inequalities can be better described
and understood, and how interventions could more
effectively address inequalities in childhood injuries.
D2 Recommendations for practitioners,
researchers and policy makers
Practitioners
What can practitioners glean from this review? How is it
possible for practitioners to address the issues of inequalities
in their injury-prevention work?
The following suggestions are drawn from examples of good
practice in the intervention studies reviewed in this report.
No one study, however, combines all these elements. Other
potential themes that should be considered for good practice
have not been rigorously tested and are summarised in the
recommendations for research.
Age
How can programmes be designed specifically to target the
issue of a child’s age and stage of development?
• Educational materials need to be age-specific, developed
and piloted for particular ages of children
• Age-paced educational materials can be developed
that are related to different stages of development and
distributed at appropriate intervals (eg children’s traffic
club materials distributed every six months)
• Attitudes and concerns of children and young people of a
particular age need to be addressed (eg peer pressure and
social acceptability of bicycle helmets)
• Home visits as a routine part of child surveillance can be
used for tailored safety advice.
Gender
How can programmes be designed specifically to target the
issue of gender?
Discussion and recommendations
• Attitudes and concerns of boys and girls need to be
addressed (eg boys and girls have different attitudes
towards wearing bicycle helmets)
(Few interventions have targeted gender issues.)
Social and economic factors
How can programmes be designed specifically to target the
issues related to social and economic groups?
• Free or low-cost safety equipment needs to be provided
• Equipment needs to be installed by professionals
• Specific advice focused on the needs of target groups is
required
• Materials should be easy to read, taking literacy levels into
account
• There should be less reliance on written media
• No-cost or low-cost interventions are more likely to be
implemented
• Free crèche facilities, provision of transport etc should be
considered for families attending training
• Home visits can be utilised to provide support and advice
• Lay workers from the same community as the target group
can be recruited and trained
• Specific groups and injuries need to be targeted (eg
window guards provided for tenants of high-rise flats to
prevent window-fall injuries)
• Some programmes need to be modified for specific groups
(eg equipment loan schemes with free equipment for
those receiving state benefits).
Culture/ethnicity
How can programmes be designed specifically to target the
issues relating to culture/ethnicity?
• Materials can be produced in different languages, or
translators made available
• Programme coordinators can be drawn from the same
ethnic group as the target promotion
• Community volunteers can be trained
• Community, church or political leaders need to be involved
in the development of programmes
• Target groups should be involved in designing interventions
• Programmes need to be delivered sensitively by trained
investigators
• Programmes may need to emphasise different cultural
values (eg a holistic view of health)
• Specific injuries affecting members of a particular ethnic
group need to be targeted.
(Few interventions have addressed variations in majority
cultures.)
51
Place
How can programmes be designed specifically to target the
issues related to place?
• Programmes need to draw on local data when available
• Context of the intervention area needs to be addressed
as fully as possible (eg physical characteristics of area,
housing type, road layout, social and economic, cultural
and ethnic composition).
(Few interventions have specifically addressed variations
related to place.)
Researchers
• National injury data systems do not routinely provide
information on ethnic groups or vulnerable groups such as
children in homeless families or disabled children. Research
is required to obtain a clear picture of the scale of injuries
in these groups.
• There are considerable gaps in research relating to the
causal mechanisms of injury. For some factors, such as
gender, ethnicity and culture, little substantial research
has been conducted. These gaps in research need to be
addressed.
• The ideas and methodologies being explored in the
wider ESRC Health Variations Programme are only now
starting to enter the field of injury research. The field
has until recently been dominated by engineering and
medical approaches, with little input from behavioural
and social sciences. National, regional and local spatial
variations in injury patterns tend to be descriptive, rather
than a starting point for understanding injury variations.
A broader view of place needs to be explored in relation
to injury, eg the physical and environmental attributes
of places, and the idea of place to which people attach
certain meanings or actions. The issue of majority cultures,
how they vary in and between countries and how they
relate to injury, has rarely been addressed. The concept of
social capital needs to be explored in relation to childhood
injury, particularly in relation to its contribution to
understanding the context in which an intervention takes
place. Qualitative studies have an important role to play, in
addition to quantitative studies.
• Programmes designed specifically to target the issues
related to inequalities need to be evaluated within the UK
context. There should be an emphasis on understanding
the process of how an intervention develops, and the
programme’s effect on specific subgroups (eg boys/girls,
older or younger children) needs to be incorporated into
programme evaluation.
52
• Programmes in which there is greater emphasis on
involvement of the target group need to be evaluated.
Policy makers
• Injury-prevention initiatives have rarely addressed
inequalities (age, gender, social and economic factors,
culture and ethnicity, place) in a systematic manner. These
factors merit greater attention.
• Some groups of children and young people are in
positions of extreme vulnerability and exposed to much
higher levels of injury risk where risk factors are additive
and cumulative over time. More information is needed on
these groups as specific interventions may be required to
address their needs.
• This review highlights the fact that few policy-level
interventions have been explored. This important gap
in the evidence needs to be addressed by innovative
initiatives and research.
Injuries in children aged 0–14 years and inequalities
References
Acheson, D. (1998). Report of the independent inquiry into
inequalities in health. London: The Stationery Office.
Agran, P.F., Winn, D.G. and Anderson, C.L. (1994).
Differences in child pedestrian injury events by location.
Pediatrics 93 (2): 284-8.
Alwash, R. and McCarthy, M. (1988). Accidents in the
home among children under 5: ethnic differences or social
disadvantage? British Medical Journal 296: 1450-3.
Ampofo-Boateng, K. and Thompson, J.A. (1991). Children’s
perception of safety and danger on the road. British Journal
of Developmental Psychology 82: 487-505.
Arblaster, L., Entwistle, V., Lambert, M., Forster, M., et al.
(1995). Review of the research on the effectiveness of health
service interventions to reduce variations. CRD Report 3.
York: NHS Centre for Reviews & Dissemination, University of
York.
Bijur, P., Golding, J. and Haslum, M. (1988a). Persistence of
occurrence of injury: can injuries of preschool children predict
injuries of school-aged children? Pediatrics 82: 707-12.
Bijur, P., Golding, J., Haslum, M. and Kurzon, M. (1988b).
Behavioural predictors of injury in school-age children.
American Journal of Diseases of Children 142: 1307-12.
Bone, M. and Meltzer, H. (1986). The prevalence of disability
among children. London: Office of Population Censuses and
Surveys.
Botting, B. and Crawley, R. (1995). Trends and patterns in
childhood mortality and morbidity. In: Botting, B. (ed.) The
health of our children: a review in the mid-1990s. Decennial
supplement OPCS DS No. 11, 62-81. London: Office of
Population Censuses and Surveys.
Campbell, H. and O’Driscoll, S. (1995). The epidemiology of
leisure accidents in Scotland. Health Bulletin 53 (5): 280-3.
Avery, J.G., Vaudin, J.N., Fletcher, J.L. and Watson, J.M.
(1990). Geographical and social variations in mortality due to
childhood accidents in England and Wales 1975–84. Public
Health 104: 171-82.
Christie, N. (1995a). The high risk child pedestrian: socioeconomic and environmental factors in their accidents.
Project Report 117. Crowthorne, Berkshire: Transport
Research Laboratory.
Balarajan, R. (1995). Ethnicity and variations in the nation’s
health. Health Trends 27: 114-9.
Christie, N. (1995b). Social, economic and environmental
factors in child pedestrian accidents: a research review.
Project Report 116. Crowthorne, Berkshire: Transport
Research Laboratory.
Barry, A., Carr-Hill, R. and Glanville, I. (1991). Homelessness
and health: What do we know? What should be done? York:
University of York.
Berger, L. and Mohan, D. (1996). Injury control: a global
view. Delhi: World Health Organization.
Bijur, P., Stewart-Brown, S. and Butler, N. (1986). Child
behavior and accidental injury in 11,966 preschool children.
American Journal of Diseases of Children 140: 487-92.
References
Curtis, S. and Rees-Jones, I. (1998). Is there a place for
geography in the analysis of health inequality? Sociology of
Health and Illness 20 (S): 645-72.
Cushman, R., Down, J., MacMillan, N. and Waclawik, H.
(1991). Helmet promotion in the emergency room following
a bicycle injury: a randomized trial. Pediatrics 88 (1): 43-7.
53
Department of Health, D. (2002a). Preventing accidental
injury – priorities for action. Report to the Chief
Medical Officer from the Accidental Injury Task Force.
www.doh.gov.uk/accidents. London: Department of Health.
Grundy, S., Towner, E., Sparks, G., Hughes, K., et al. (2002).
Taking chances: the lifestyles, leisure worlds and leisure risks
of young people. Report prepared for the National Lotteries
Charity Board. London: Child Accident Prevention Trust.
Department of Health. (2002b). Measuring and monitoring
injury. Report to the Accidental Injury Task Force from
the Measuring and Monitoring Injury Working Group.
www.doh.gov.uk/accidents. London: Department of Health.
Hapgood, R., Kendrick, D. and Marsh, P. (2000). How well
do sociodemographic characteristics explain variation in
childhood safety practices? Journal of Public Health Medicine
113: 285-9.
DiGuiseppi, C., Roberts, I., Li, L. and Allen, D. (1998).
Determinants of car travel on daily journeys to school: cross
sectional survey of primary school children. British Medical
Journal 316: 1426-8.
Hawe, P. and Shiell, A. (2000). Social capital and health
promotion: a review. Social Science and Medicine 51: 87185.
Dowswell, T. and Towner, E. (2002). Social deprivation
and the prevention of unintentional injury in childhood: a
systematic review. Health Education Research 17 (2): 221-37.
DTLR (2002). Road safety research. Compendium of research
projects 2001/2001. London: Department for Transport, Local
Government and the Regions.
Gallagher, S., Finison, K., Guyer, B. and Goodenough,
S. (1984). The incidence of injuries among 87,000
Massachusetts children and adolescents: results of the
1980–81 Statewide Childhood Injury Prevention Program
Surveillance System. American Journal of Public Health 74
(12): 1340-7.
Gesler, W. and Kearns, R. (2002). Culture/place/health.
Critical geographies. London: Routledge.
Godkin, M. (1980). Identity and place: clinical applications
based on notions of rootedness and uprootedness. In:
Buttimer, A. and Seamon, D. (eds) The human experience of
space and place. London: Croom Helm.
Gordon, D. and Heslop, P. (1999). Poverty and disabled
children. In: Dorling, D. and Simpson, S. (eds) Statistics in
society. The arithmetic of politics. London: Arnold.
Gorman, D., Ramsay, L., Wilson, G. and Freeland, P. (1999).
Using routine accident and emergency department data to
describe local injury epidemiology. Public Health 113: 285-9.
Graham, H. (1999). Inquiry into inequalities in health. Health
Variations 3: 2-4.
Grossman, D. and Rivara, F. (1992). Injury control in
childhood. Pediatric Clinics of North America 39 (3): 471-85.
54
Hillman, M., Adams, J. and Whitelegg, J. (1990). One false
move: a study of children’s independent mobility. London:
Policy Studies Institute.
Jarvis, S., Towner, E. and Walsh, S. (1995). Accidents. In:
Botting, B. (ed.) The health of our children: a review in the
mid 1990s. Decennial Supplement OPCS DS No. 11, 95-112.
London: HMSO.
Judge, K. and Benzeval, M. (1993). Health inequalities:
new concerns about the children of single mothers. British
Medical Journal 306: 677-80.
Kemp, A. and Sibert, J. (1992). Drowning and near
drowning in children in the United Kingdom: lessons for
prevention. British Medical Journal 304: 1143-6.
Kendrick, D. (1993). Accidental injury attendances as
predictors of future admission. Journal of Public Health
Medicine 15 (2): 171-4.
Kendrick, D. and Marsh, P. (1997). Injury prevention
programmes in primary care: a high risk group or a whole
population approach? Injury Prevention 3: 170-5.
Kendrick, D. and Marsh, P. (1998). Babywalkers: prevalence
of use and relationship with other safety practices. Injury
Prevention 4: 295-8.
Kendrick, D. and Marsh, P. (2000). Inequalities in receipt of
injury prevention in primary care. Health Education Journal
59: 150-6.
Kendrick, D. and Marsh, P. (2001). How useful are sociodemographic characteristics in identifying children at risk of
unintentional injury? Public Health 115: 103-7.
Kendrick, D., West, J., Wright, S. and Presbury, M. (1995).
Does routine child health surveillance reach children most at
risk of accidental injury? Journal of Public Health Medicine 17
(1): 39-45.
Injuries in children aged 0–14 years and inequalities
King, D., Lawson, S., Proctor, S., et al. (1987). Child
pedestrian accidents in inner city areas: patterns and
treatment. In: Proceedings of the PTRC Summer Annual
Meeting, University of Bath, 7–11 September 1987.
Crowthorne, Berkshire: Transport Research Laboratory.
Klein, D. (1980). Societal influences on childhood accidents.
Accident Analysis and Prevention 12: 275-81.
Laing, G. and Logan, S. (1999). Patterns of unintentional
injury in childhood and their relation to socio-economic
factors. Public Health 113: 291-4.
Lawson, S.D. and Edwards, P.J. (1991). The involvement of
ethnic minorities in road accidents: data from three studies of
young pedestrian casualties. Traffic Engineering and Control
(Jan): 12-19.
Lissauer, T., Richman, S., Tempia, M., Jenkins, S., et al. (1993).
Influence of homelessness on acute admissions to hospital.
Archives of Disease in Childhood 69: 423-8.
Lynam, D. and Harland, D. (1992). Child pedestrian safety
in the UK. Proceedings of the Conference ‘Road Safety in
Europe’, Berlin, Germany, 30 September–2 October 1992.
VTI Report No. 380A. Linkoping, Sweden: Swedish Road and
Transport Research Institute (VTI).
Macintyre, S., Maciver, S. and Soomans, A. (1993). Area,
class and health: should we be focusing on places or
people? Journal of Social Policy 22 (2): 213-34.
Matthews, H. (1995). Culture, environmental experience and
environmental awareness: making sense of young Kenyan
children’s views of place. Geographical Journal 161: 285-95.
Millard, A. (1994). A causal model of high rates of child
mortality. Social Science and Medicine 38 (2): 253-68.
Moller, J. (1992a). Major issues and debates in injury control.
In: Community based injury prevention. A practical guide, 915. South Australia: National Safety Council of Australia.
Moller, J. (1992b). Community based injury prevention. A
practical guide. South Australia: National Safety Council of
Australia.
Moller, J. (1997). Population strategies for prevention? If only
it were that simple! Injury Prevention 3: 162-4.
Morrison, A., Stone, D., Redpath, A., Campbell, H., et al.
(1999). Trend analysis of socioeconomic differentials in
deaths from injury in childhood in Scotland, 1981-95.
British Medical Journal 318: 567-8.
References
Morrow, V. (2002). Children’s experiences of ’community’:
implications of social capital discourses. In: Swann, C. and
Morgan, A. (eds) Social capital for health. London: Health
Development Agency.
Mott, A., Evans, R., Rolfe, K., Potter, D., et al. (1994).
Patterns of injuries to children on public playgrounds.
Archives of Disease in Childhood 71 (4): 328-30.
Ohn, T., Gilmour, W. and Stone, D. (1995). Pattern and risks
of accidental injuries in children presenting to a paediatric
accident and emergency department. Maternal and Child
Health Journal (Dec): 404-7.
OPCS (1988). Occupational mortality: childhood supplement.
Registrar General’s decennial supplement for England and
Wales, 1979-80, 1982-83. London: Office of Population
Censuses and Surveys.
Pawson, R. and Myhill, A. (2001). Learning lessons:
enhancing evaluation through research review. Crowthorne,
Berkshire: Transport Research Laboratory.
Phillimore, P. and Morris, D. (1991). Discrepant legacies:
premature mortality in two industrial towns. Social Science
and Medicine 33 (2): 139-52.
Pless, I. (1997). Difficult choices. Editorial. Injury Prevention 3:
153-4.
Pless, I., Peckham, C. and Power, C. (1989). Predicting traffic
injuries in childhood: a cohort analysis. Journal of Pediatrics
115: 932-8.
Preston, B. (1972). Statistical analysis of child pedestrian
accidents in Manchester and Salford. Accident Analysis and
Prevention 4: 323-32.
Purdon, S. (1997). Non fatal accidents. In: Prescott-Clarke, P.
and Primatesta, P. (eds) The Health Survey for England 1995,
Vol 1: Findings. London: The Stationery Office.
Raleigh, V. and Balarajan, R. (1995). The health of infants and
children among ethnic minorities. In: Botting, B. (ed.) The
health of our children: a review in the mid-1990s. Decennial
Supplement OPCS DS No. 11. London: Office of Population
Censuses and Surveys.
Richman, S., Roderick, P., Victor, C. and Lissauer, T. (1991).
Use of acute hospital services by homeless children. Public
Health 105 (4): 297-302.
Roberts, H. (2000). What works in reducing inequalities in
child health. Ilford, Essex: Barnardo’s.
55
Roberts, H., Smith, S. and Bryce, C. (1995). Children at risk?
Safety as a social value. Buckingham: Open University Press.
Roberts, I. (1997). Cause specific social class mortality
differentials for child injury and poisoning in England and
Wales. Journal of Epidemiology and Community Health 51:
334-5.
Roberts, I. and Power, C. (1996). Does the decline in child
injury mortality vary by social class? A comparison of class
specific mortality in 1981 and 1991. British Medical Journal
313: 784-6.
Sandels, S. (1975). Children in traffic. Revised. London: Elek
Books.
Scott, A., Pearce, D. and Goldblatt, P. (2001). The sizes and
characteristics of the minority ethnic populations of Great
Britain – latest estimates. Population Trends 105: 6-15.
Scottish Development Department (1989). Must do better:
a study of child pedestrian accidents and road crossing
behaviour in Scotland. Consultants’ report to Scottish
Development Department. London: MVA Consultancy.
Secretary of State for Health (1999). Saving lives: our
healthier nation. London: The Stationery Office.
Sellar, C., Ferguson, J. and Goldacre, M. (1991). Occurrence
and repetition of hospital admissions for accidents in
preschool children. British Medical Journal 302: 16-19.
Sharples, P., Storey, A., Aynsley-Green, A. and Eyre, J. (1990).
Causes of fatal childhood accidents involving head injury
in Northern region, 1979–86. British Medical Journal 301:
1193-7.
Shaw, K., McCormick, M., Kustra, S., Ruddy, R., et al. (1988).
Correlates of reported smoke detector usage in an innercity population: participants in a smoke detector give-away
program. American Journal of Public Health 78 (6): 650-3.
Shelter (2002). Growing up homeless. London: Shelter.
Sparks, G., Craven, M.A. and Worth, C. (1994).
Understanding differences between high and low childhood
accident rate areas: the importance of qualitative data.
Journal of Public Health Medicine 16 (4): 439-46.
Spencer, C. and Blades, M. (1993). Children’s understanding
of places: the world at hand. Geography 78: 367-72.
Spencer, N. (1993). Commentary on Lissauer et al. (1993).
Archives of Disease in Childhood (69): 428-9.
56
Stewart-Brown, S., Peters, T., Golding, J. and Bijur, P. (1986).
Case definition in childhood accident studies: a vital factor in
determining results. International Journal of Epidemiology 15
(3): 352-60.
Stone, D. (1997). All about homelessness. London: Shelter.
Swann, C. and Morgan, A. (2002). Social capital for health:
insights from qualitative research. In: Swann, C. and
Morgan, A. (eds) Social capital for health. London: Health
Development Agency.
Taylor, B., Wadsworth, J. and Butler, N. (1983). Teenage
mothering, admission to hospital, and accidents during the
first 5 years. Archives of Disease in Childhood 58: 6-11.
Thomson, J., Tolmie, A., Foot, H. and McLaren, B. (1996).
Child development and the aims of road safety education
– a review and analysis. DoT Road Safety Research No. 1.
Norwich: The Stationery Office.
Thomson, J., Tolmie, A. and Mamoon, T. (2001). Road
accident involvement of children from ethnic minorities:
a literature review. Road Safety Research Report No. 19.
London: Department of the Environment, Transport and the
Regions.
Towner, E. and Dowswell, T. (2002). Community-based
childhood injury prevention interventions: what works? Health
Promotion International 17 (3): 273-84.
Towner, E. and Errington, G. (2002). The epidemiology of
choking in childhood and implications for prevention. A
report prepared for the Department of Health. London:
Department of Health.
Towner, E., Jarvis, S., Walsh, S. and Aynsley-Green, A. (1994).
Measuring exposure to injury risk in schoolchildren aged
11–14. British Medical Journal 308: 449-52.
Towner, E., Dowswell, T., Mackereth, C. and Jarvis, S. (2001).
What works in preventing unintentional injuries in children
and young adolescents? An updated systematic review.
London: Health Development Agency.
www.hda-online.org.uk/downloads/pdfs/prevent_injuries.pdf
Wadsworth, J., Burnell, I., Taylor, B. and Butler, N. (1983).
Family type and accidents in preschool children. Journal of
Epidemiology and Community Health 37: 100-4.
Waller, A.E. and Marshall, S.W. (1993). Childhood thermal
injuries in New Zealand resulting in death and hospitalization.
Burns 19 (5): 371-6.
Injuries in children aged 0–14 years and inequalities
Walsh, S. and Jarvis, S. (1992). Measuring the frequency
of ‘severe’ accidental injury in childhood. Journal of
Epidemiology and Community Health 46: 26-32.
Walsh, S., Barton, S., Jarvis, S. and Clark, W. (1990). Children
injured in road traffic accidents: a collaborative project with
Northumbria Police. NERRL Research Report 90/3. Newcastle:
North East Regional Research Laboratory.
Ward, H., Cave, J., Morrison, A., Allsop, R., et al. (1994).
Pedestrian activity and accident risk. Basingstoke, Hants: AA
Foundation for Road Safety Research.
Whitehead, M., Diderichson, F. and Burström, B. (2000).
Researching the impact of public policy on equalities
in health. In: Graham, H. (ed.) Understanding health
inequalities, 203-18. Buckingham: Open University Press.
Whitelegg, J. (1987). A geography of road traffic accidents.
Transactions of the Institute of British Geographers 12: 16176.
Widdowfield, R. (1999). The limitations of official
homelessness statistics. In: Dorling, D. and Simpson, S. (eds)
Statistics in society. The arithmetic of politics, 181-8. London:
Arnold.
Williams, F., Lloyd, O. and Dunbar, J. (1991). Deaths from
road traffic accidents in Scotland: 1979–1988. Does it matter
where you live? Public Health 105: 319-26.
Wilson, M.H., Baker, S.P., Tenet, S.P., Shock, S., et al. (1991).
Saving children. A guide to injury prevention. Oxford: Oxford
University Press.
References
57
References for Appendix Tables C2–C8
C2(i) Pedestrian injuries – transport policy
and traffic calming (N = 7)
1
2
3
4
58
Engel, U. (1982). Short-term and area-wide evaluation
of safety measures implemented in a residential area
named Osterbro. A case study. Seminar on short-term
and area-wide evaluation of safety measures, Amsterdam.
Amsterdam: Netherlands Institute for Road Safety
Research.
Jorgensen, E. (1985). Bicycle tracks in urban areas in
Denmark. Evaluation of the effect on safety. In: Biecheler,
M., Lacombe, C., Muhlrad, N. (eds) Evaluation 85:
International meeting on the evaluation of local traffic
safety measures, Proceedings of the Paris conference,
20–23 May 1985, pp. 755-61.
Boxall, J.A. (1988). School crossing patrols: how effective
are they? Traffic Engineering and Control 29 (11): 586.
i Walker, R.T., Gardner, G., McFetridge, M. (1989).
Urban safety project: the Nelson scheme. Contractor’s
Report CR191. Department of Transport, Transport
and Road Research Laboratory. Crowthorne,
Berkshire: Transport and Road Research Laboratory.
ii Ward, H., Norrie, J., Sang, A., Allsop, R. (1989a).
Urban safety project: the Reading scheme.
Contractor’s Report CR138. Crowthorne, Berkshire:
Transport and Road Research Laboratory.
iii Ward, H., Norrie, J., Sang, A. and Allsop, R.
(1989b). Urban safety project: the Sheffield scheme.
Contractor’s Report CR134. Crowthorne, Berkshire:
Transport and Road Research Laboratory.
iv Ward, H., Norrie, J., Allsop, R., Sang, A. (1989c).
Urban safety project: the Bristol scheme. Contractor’s
Report CR192. Crowthorne, Berkshire: Transport and
Road Research Laboratory.
v
5
6
7
Walker, R., McFetridge, M. (1989). Urban safety
project: the Bradford scheme. Contractor’s Report
CR190. Crowthorne, Berkshire: Transport and Road
Research Laboratory.
vi Mackie, A., Ward, H., Walker, R. (1988). Urban safety
project. 2. Interim results for area wide schemes.
Research Report 154. Crowthorne, Berkshire:
Transport and Road Research Laboratory.
vii Mackie, A., Ward, H., Walker, R. (1990). Urban safety
project. 3. Overall evaluation of area wide schemes.
Research Report 263. Crowthorne, Berkshire:
Transport and Road Research Laboratory.
viii Tillman, M. (1992). A study of the longer term effects
of the Urban Safety Project. The case of Reading,
Berkshire. MSc, Newcastle upon Tyne: Transport
Operations Research Group.
Janssen, S. (1991). Road safety in urban districts: final
results of accident studies in the Dutch demonstration
projects of the 1970s. Traffic Engineering and Control
32: 292-6.
Vis, A., Dïjkstra, A., Slop, M. (1992). Safety effects of 30
km/h zones in the Netherlands. Accident Analysis and
Prevention 24 (1): 75-86.
Webster, D., Mackie, A. (1996). Review of traffic calming
schemes in 20 mph zones. TRL Report 215. Crowthorne,
Berkshire: Transport Research Laboratory.
C2(ii) Pedestrian injuries – pedestrian and
traffic education (N = 23)
1
Downing, C. (1987). Evaluation of the impact and
penetration of a children’s traffic club. In: Rottengatter,
J., De Bruin, R. (eds) Road User Behaviour. Theory and
Practice. The Netherlands: Van Gorcum.
Injuries in children aged 0–14 years and inequalities
2
Preusser, D., Blomberg, R. (1984). Reducing child
pedestrian accidents through public education. Journal
of Safety Research 15: 47-56.
3 Antaki, C., Morris, P.E., Flude, B.M. (1986). The
effectiveness of the ‘Tufty Club’ in road safety education.
British Journal of Educational Psychology 56: 363-5.
4 Young, D., Lee, D. (1987). Training children in road
crossing skills using a roadside simulation. Accident
Analysis and Prevention 19 (5): 327-41.
5 Preusser, D., Lund, A. (1988). ‘And keep on looking.’ A
film to reduce pedestrian crashes among 9 to 12 year
olds. Journal of Safety Research 19 (4): 177-85.
6 van Schagen, I. (1988). Training children to make safe
crossing decisions. In: Rothengatter, J., de Bruin, R. (eds)
Road user behaviour: theory and research. Assen, the
Netherlands: Van Gorcum.
7 Malenfant, L., Van Houten, R. (1989). Increasing the
percentage of drivers yielding to pedestrians in three
Canadian cities with a multifaceted safety program.
Health Education Research 5 (2): 275-9.
8 Renaud, L., Suissa, S. (1989). Evaluation of the efficacy
of simulation games in traffic safety education of
kindergarten children. American Journal of Public Health
79 (3): 307-9.
9 Nishioka, N., Ieda, S., Takahashi, H., Okajima, Y. et al.
(1991). An experimental study on the safety behavior
of children in a dashing-out situation – effects of verbal
instructions and traffic conditions on safety behavior.
International Association of Traffic and Safety Sciences
(IATSS) Research 15 (1): 39-45.
10 Rivara, F.P., Booth, C.L., Bergman, A.B., Rogers, L.W.,
et al. (1991). Prevention of pedestrian injuries to
children: effectiveness of a school training program.
Pediatrics 88 (4): 770-5.
11 Thomson, J., Ampofo-Boateng, K., Pitcairn, T., Grieve,
R., et al. (1992). Behavioural group training of children
to find safe routes to cross the road. British Journal of
Educational Psychology 62: 173-83.
12 i Tucker, S. (1992). The operation of the Eastern
Region traffic club. Unpublished. WP/RUS/116.
Department of Transport, Transport and Road
Research Laboratory.
ii Bryan-Brown, K. (1994). The effectiveness of the
General Accident Eastern Region children’s traffic
club. Project Report 99. Crowthorne, Berkshire:
Transport Research Laboratory.
iii Bryan-Brown, K. (1995). The effects of a children’s
traffic club. Road accidents: Great Britain 1994.
London: Department of Transport.
References: Appendix Tables
13 Ampofo-Boateng, K., Thomson, J.A., Grieve, R.,
Pitcairn, T., et al. (1993). A developmental and training
study of children’s ability to find safe routes to cross the
road. British Journal of Developmental Psychology 11:
31-45.
14 Demetre, J., Lee, D., Grieve, R., Pitcairn, T. et al. (1993).
Young children’s learning on road-crossing simulations.
British Journal of Educational Psychology 63 (2): 349-59.
15 West, R., Sammons, P., West, A. (1993). Effects of a
traffic club on road safety knowledge and self-reported
behaviour of young children and their parents. Accident
Analysis and Prevention 25 (5): 609-18.
16 Gregersen, N.P., Nolen, S. (1994). Children’s road
safety and the strategy of voluntary traffic safety clubs.
Accident Analysis and Prevention 26 (4): 463-70.
17 Harland, G., Tucker, S. (1994). ‘Let’s Decide Walk Wise’
– the development and testing of a pedestrian training
resource. 14th Annual Conference of the British Health
and Safety Society. Crowthorne, Berkshire: Transport
Research Laboratory.
18 Penna, C. (1994). ‘Streets ahead’ evaluation. GR 94-13
VIC ROADS KEW.VIC.3101. Victoria, Australia: Monash
University.
19 Tziotis, M. (1994). Evaluation of the ‘Safe routes to
schools’ and ‘Walk with care’ programs. Melbourne, VIC,
Australia: Road Safety Department, VicRoads.
20 Wright, M., Rivara, F., Ferse, W. (1995). Evaluation of
the ’Think First’ head and spinal cord injury prevention
program. Injury Prevention 1: 81-5.
21 Burke, G., Lapidus, G., Banco, L., Zavoski, R., et al.
(1996). Evaluation of the effectiveness of a pavement
stencil in promoting safe behavior among elementary
school children boarding school buses. Pediatrics 97:
520-3.
22 Thomson, J., Whelan, K. (1997). A community approach
to road safety education using practical training
methods. The Drumchapel project. Road Safety Research
Report No. 3. London: Department of Transport.
23 Cross, D., Stevenson, M., Hall, M., Burns, S., et al.
(2000). Child pedestrian injury prevention project:
student results. Preventive Medicine 30: 179-87.
C3 Car occupant child restraints (N = 39)
1
2
Scherz, R. (1976). Restraint systems for the prevention
of injury to children in automobile accidents. American
Journal of Public Health 66 (5): 451-6.
Miller, J., Pless, I. (1977). Child automobile restraints:
evaluation of health education. Pediatrics 59 (6): 907-11.
59
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
60
Reisinger, K., Williams, A. (1978). Evaluation of programs
designed to increase the protection of infants in cars.
Pediatrics 62 (3): 280-7.
Reisinger, K., Williams, A., Wells, J., John, C., et al.
(1981). Effect of pediatricians’ counseling on infant
restraint use. Pediatrics 67 (2): 201-6.
Christophersen, E., Sullivan, M. (1982). Increasing the
protection of newborn infants in cars. Pediatrics 70: 215.
Geddis, D., Pettengell, R. (1982). Parent education: its
effect on the way children are transported in cars.
New Zealand Medical Journal 95: 314-6.
Greenberg, L., Coleman, A. (1982). A prenatal and
postpartum safety education program: influence
on parental use of infant car restraints. Journal of
Developmental and Behavioral Pediatrics 3 (1): 32-4.
Berger, L.R., Saunders, S., Armitage, K., Schauer, L.
(1984). Promoting the use of car safety devices for
infants: an intensive health education approach.
Pediatrics 74 (1): 16-19.
Decker, M., Dewey, M., Hutcheson, R., Schaffner, W.
(1984). The use and efficacy of child restraint devices.
The Tennessee experience, 1982 and 1983. Journal of
the American Medical Association 252 (18): 2571-5.
Chang, A., Dillman, A., Leonard, E., English, P. (1985).
Teaching car passenger safety to preschool children.
Pediatrics 76 (3): 425-8.
Christophersen, E., Sosland-Edelman, D., LeClaire, S.
(1985). Evaluation of two comprehensive infant car seat
loaner programs with 1-year follow-up. Pediatrics 76 (1):
36-42.
Goodson, J., Buller, C., Goodson, W. (1985). Prenatal
child safety education. Obstetrics and Gynaecology 65:
312-5.
Guerin, D., MacKinnon, D. (1985). An assessment of
the California child passenger restraint requirement.
American Journal of Public Health 75 (2): 142-4.
Colletti, R. (1986). Longitudinal evaluation of a statewide
network of hospital programs to improve child passenger
safety. Pediatrics 77 (4): 523-9.
Roberts, M., Fanurik, D. (1986). Rewarding elementary
school children for their use of safety belts. Health
Psychology 5 (3): 185-96.
Sewell, C., Hull, H., Fenner, J., Graff, H. et al. (1986).
Child restraint law effects on motor vehicle accident
fatalities and injuries: the New Mexico experience.
Pediatrics 78 (6): 1079-84.
Wagenaar, A., Webster, D. (1986). Preventing injuries to
children through compulsory automobile safety seat use.
Pediatrics 78 (4): 662-72.
18 Agran, P.F., Dunkle, D.E., Winn, D.G. (1987). Effects of
legislation on motor vehicle injuries to children. American
Journal of Diseases of Children 141: 959-64.
19 Bowman, J., Sanson-Fisher, R., Webb, G. (1987).
Interventions in preschools to increase the use of safety
restraints by preschool children. Pediatrics 79 (1): 103-9.
20 Macknin, M., Gustafson, C., Gassman, J., Barich, D.
(1987). Office education by pediatricians to increase seat
belt use. American Journal of Diseases of Children 141:
1305-7.
21 Sowers-Hoag, K., Thyer, B., Bailey, J. (1987). Promoting
automobile safety belt use by young children. Journal of
Applied Behavior Analysis 20: 133-8.
22 Jamark, S., Ljungblom, B.A., Turbell, T. (1988). Infant
carriers – a trial in two counties. Linkoping, Sweden:
Swedish Road and Traffic Research Institute.
23 Margolis, L., Wagenaar, A., Liu, W. (1988). The effects
of a mandatory child restraint law on injuries requiring
hospitalization. American Journal of Diseases of Children
142: 1099-103.
24 Roberts, M.C., Fanurik, D., Wilson, D. (1988). A
community program to reward children’s use of seat
belts. American Journal of Community Psychology 16 (3):
395-407.
25 Seekins, T., Fawcett, S., Cohen, S., Elder, J. et al. (1988).
Experimental evaluation of public policy: the case of state
legislation for child passenger safety. Journal of Applied
Behavior Analysis 21 (3): 233-43.
26 Downing, C., Franklin, J. (1989). An evaluation of
two local infant restraint loan schemes. First World
Conference on Accident and Injury Prevention; secondary
conference on Child Accident Prevention, Stockholm.
Stockholm: Karolinska Institute.
27 Liberato, C., Eriacho, B., Schmiesing, J., Krump, M.
(1989). SafeSmart Safety Seat Intervention Project: a
successful program for the medically indigent. Patient
Education and Counseling 13: 161-70.
28 Lund, A., Stuster, J., Fleming, A. (1989). Special publicity
and enforcement of California’s belt use laws: making a
secondary law work. Journal of Criminal Justice 17: 32941.
29 Morrow, R. (1989). A school-based program to increase
seatbelt use. Journal of Family Practice 29 (5): 517-20.
30 Robitaille, Y., Legault, J., Abbey, H., Pless, I. (1990).
Evaluation of an infant car seat program in a low-income
community. American Journal of Diseases of Children
144: 74-8.
Injuries in children aged 0–14 years and inequalities
31 Streff, F., Molnar, L., Christoff, C. (1992). Increasing
safety belt use in a secondary enforcement state:
evaluation of a three-county special enforcement
program. Accident Analysis and Prevention 24 (4): 36983.
32 Lindqvist, K. (1993). Does the use of child safety seats
increase as a result of loan schemes? Accident Analysis
and Prevention 25 (4): 421-9.
33 Stuy, M., Green, M., Doll, J. (1993). Child care centers:
a community resource for injury prevention. Journal of
Developmental and Behavioral Pediatrics 14 (4): 224-9.
34 Decina, L., Temple, M., Dorer, H. (1994). Increasing
child safety-seat use and proper use among toddlers.
Evaluation of an enforcement and education program.
Accident Analysis and Prevention 26 (5): 667-73.
35 Russell, J., Kresnow, M.-J., Brackbill, R. (1994). The effect
of adult belt laws and other factors on restraint use for
children under age 11. Accident Analysis and Prevention
26 (3): 287-95.
36 Hazinski, M., Eddy, V., Morris, J. (1995). Children’s traffic
safety program: influence of early elementary school
safety education on family seat belt use. Journal of
Trauma 39: 1063-8.
37 Margolis, L., Bracken, J., Stewart, J. (1996). Effects of
North Carolina’s mandatory safety belt law on children.
Injury Prevention 2: 32-5.
38 Williams, A., Wells, J., Ferguson, S. (1997). Development
and evaluation of programs to increase proper child
restraint use. Journal of Safety Research 28: 197-202.
39 Hanfling, M., Mangus, L., Gill, A., Bailey, R. (2000).
A multifaceted approach to improving motor vehicle
restraint compliance. Injury Prevention 6: 125-9.
4
5
C4 Bicycle injuries (N = 28)
1
2
3
Wood, T., Milne, P. (1988). Head injuries to pedal cyclists
and the promotion of helmet use in Victoria, Australia.
Accident Analysis and Prevention 20 (3): 177-85.
i DiGuiseppi, C., Rivara, F., Koepsell, T., Polissar, L.
(1989). Bicycle helmet use by children. Evaluation of
a community-wide helmet campaign. Journal of the
American Medical Association 262 (16): 2256-61.
ii Bergman, A., Rivara, F., Richards, D., Rogers,
L. (1990). The Seattle children’s bicycle helmet
campaign. American Journal of Diseases of Children
144: 727-31.
Moore, D., Adair, V. (1990). Effects of a school-based
education programme on safety helmet usage by 11- to
13 year old cyclists. Educational Psychology 10 (1): 73-8.
References: Appendix Tables
6
7
8
9
i
Cushman, R., Down, J., MacMillan, N., Waclawik,
H. (1991a). Helmet promotion in the emergency
room following a bicycle injury: a randomized trial.
Pediatrics 88 (1): 43-7.
ii Cushman, R., James, W., Waclawik, H. (1991b).
Physicians promoting bicycle helmets for children: a
randomized trial. American Journal of Public Health
81 (8): 1044-6.
i Leicester, P., Nassau, F., Wise, A. (1991). The
introduction of compulsory bicycle helmet wearing in
Victoria. Report GR 91-4. Melbourne, VIC, Australia:
Road Safety Department, VicRoads.
ii Cameron, M., Heiman, L., Neiger, D. (1992).
Evaluation of the Bicycle Helmet Wearing Law in
Victoria during its first 12 months. Report No. 32.
Victoria, Australia: Accident Research Centre, Monash
University.
iii Vulcan, A., Cameron, M., Watson, W. (1992).
Mandatory bicycle helmet use: experience in Victoria,
Australia. World Journal of Surgery 16: 389-97.
iv Finch, C., Newstead, S., Cameron, M., Vulcan, A.
(1993). Head injury reductions in Victoria two years
after introduction of mandatory bicycle helmet use.
Report No. 51. Victoria, Australia: Accident Research
Centre, Monash University.
v Cameron, M., Vulcan, A., Finch, C., Newstead, S.
(1994). Mandatory bicycle helmet use following a
decade of helmet promotion in Victoria, Australia
– an evaluation. Accident Analysis and Prevention 26
(3): 325-37.
vi McDermott, F. (1995). Bicyclist head injury prevention
by helmets and mandatory wearing legislation in
Victoria, Australia. Annals of the Royal College of
Surgeons of England 77: 38-44.
Morris, B., Trimble, N. (1991). Promotion of bicycle
helmet use among schoolchildren: a randomized clinical
trial. Canadian Journal of Public Health 82 (Mar/Apr): 924.
Pendergrast, R., Ashworth, C., DuRant, R., Litaker, M.
(1992). Correlates of children’s bicycle helmet use and
short-term failure of school-level interventions. Pediatrics
90 (3): 354-8.
Puczynski, M., Marshall, D. (1992). Helmets! All the pros
wear them. American Journal of Diseases of Children
146: 1465-7.
Towner, P., Marvel, M. (1992). A school-based intervention
to increase the use of bicycle helmets. Family Medicine
24: 156-8.
61
10 i
11
12
13
14
15
16
17
18
19
20
62
Dannenberg, A., Gielen, A., Beilenson, P., Wilson,
M., et al. (1993). Bicycle helmet laws and educational
campaigns: an evaluation of strategies to increase
children’s helmet use. American Journal of Public
Health 83 (5): 667-74.
ii Coté, T., Sacks, J., Lambert-Huber, D., Dannenberg,
A. et al. (1992). Bicycle helmet use among Maryland
children: effect of legislation and education.
Pediatrics 89 (6): 1216-20.
Parkin, P., Spence, L., Hu, X., Kranz, K., et al. (1993).
Evaluation of a promotional strategy to increase bicycle
helmet use by children. Pediatrics 91 (4): 772-7.
Macknin, M., Medendorp, S. (1994). Association
between bicycle helmet legislation, bicycle safety
education, and use of bicycle helmets in children.
Archives of Pediatric and Adolescent Medicine 148: 2559.
van Schagen, I., Brookhuis, K. (1994). Training young
cyclists to cope with dynamic traffic situations. Accident
Analysis and Prevention 26 (2): 223-30.
Mock, C., Maier, R., Boyle, E., et al. (1995). Injury
prevention strategies to promote helmet use to decrease
severe head injuries at a Level 1 Trauma Centre. Journal
of Trauma 39: 29-33.
Parkin, P., Hu, X., Spence, L., Kranz, K. (1995). Evaluation
of a subsidy program to increase bicycle helmet use by
children of low-income families. Pediatrics 96 (2): 283-7.
Farley, C., Haddad, S., Brown, B. (1996). The effects of
a 4-year program promoting bicycle helmet use among
children in Quebec. American Journal of Public Health
86: 46-51.
Savill, T., Bryan-Brown, K., Harland, G. (1996). The
effectiveness of child cycle training schemes. TRL
Report 214. Crowthorne, Berkshire: Transport Research
Laboratory.
Abularrage, J., De Luca, A., Abularrage, C. (1997).
Effect of education and legislation on bicycle helmet
use in a multiracial population. Archives of Pediatric and
Adolescent Medicine 151: 41-4.
Ekman, R., Schelp, L., Welander, G., Svanstöm, L. (1997).
Can a combination of local, regional and national
information substantially increase bicycle-helmet wearing
and reduce injuries? Experiences from Sweden. Accident
Analysis and Prevention 29: 321-8.
Kim, A., Rivara, F., Koepsell, T. (1997). Does sharing
the cost of a bicycle helmet help promote helmet use?
Injury Prevention 3: 38-42.
21 Ni, H., Sacks, J., Curtis, L., Cieslak, P., et al. (1997).
Evaluation of a statewide bicycle helmet law via multiple
measures of helmet use. Archives of Pediatric and
Adolescent Medicine 151: 59-65.
22 Britt, J., Silver, I., Rivara, F. (1998). Bicycle helmet
promotion among low income preschool children. Injury
Prevention 4: 280-3.
23 Hendrickson, S., Becker, H. (1998). Impact of a theory
based intervention to increase bicycle helmet use in low
income children. Injury Prevention 4: 126-31.
24 Logan, P., Leadbetter, S., Gibson, R., Schieber, R. (1998).
Evaluation of a bicycle helmet giveaway program – Texas,
1995. Pediatrics 101: 578-82.
25 Macarthur, C., Parkin, P., Sidky, M., Wallace, W. (1998).
Evaluation of a bicycle skills training program for
young children: a randomized controlled trial. Injury
Prevention 4: 116-21.
26 George, A., Koerner, C. (2000). Childhood injury
prevention services (CHIPS) bicycle helmet intervention:
save your head by knowing the rules. Journal of
Investigative Medicine 48 (1).
27 Lee, A., Mann, N., Takriti, R. (2000). A hospital led
promotion campaign aimed to increase bicycle helmet
wearing among children aged 11–15 living in West
Berkshire 1992–98. Injury Prevention 6: 151-3.
28 Scuffham, P., Alsop, J., Cryer, C., Langley, J.D. (2000).
Head injuries to bicyclists and the New Zealand bicycle
helmet law. Accident Analysis and Prevention 32: 56573.
C5(i) Prevention of falls and general injuries
in the home (N = 18)
1
2
3
Spiegel, C., Lindaman, F. (1977). Children can’t fly: a
program to prevent childhood morbidity and mortality
from window falls. American Journal of Public Health 67
(12): 1143-7.
i Dershewitz, R., Williamson, J. (1977). Prevention of
childhood household injuries: a controlled clinical
trial. American Journal of Public Health 67 (12):
1148-53.
ii Dershewitz, R. (1979). Will mothers use free
household safety devices? American Journal of
Diseases of Children 133: 61-4.
Colver, A., Hutchinson, P., Judson, E. (1982). Promoting
children’s home safety. British Medical Journal 285:
1177-80.
Injuries in children aged 0–14 years and inequalities
4
5
6
7
8
9
10
11
12
13
14
15
16
Minchom, P., Sibert, J., Newcombe, R., Bowley, M.
(1984). Does health education prevent childhood
accidents? Postgraduate Medical Journal 60 (Apr): 2602.
Gallagher, S., Hunter, P., Guyer, B. (1985). A home injury
prevention program for children. Pediatric Clinics of
North America 32 (1): 95-112.
Kraus, J. (1985). Effectiveness of measures to prevent
unintentional deaths of infants and children from
suffocation and strangulation. Public Health Reports 100
(2): 231-40.
Kelly, B., Sein, C., McCarthy, P.L. (1987). Safety education
in a pediatric primary care setting. Pediatrics 79 (5): 81824.
Paul, C.L., Sanson-Fisher, R.W., Redman, S., Carter, S.
(1994). Preventing accidental injury to young children
in the home using volunteers. Health Promotion
International 9 (4): 241-9.
Olds, D., Henderson, C., Kitzman, H. (1994). Does
prenatal and infancy nurse home visitation have enduring
effects on qualities of parental caregiving and child
health at 25 to 50 months of life? Pediatrics 93: 89-98.
Bablouzian, L., Freedman, E., Wolski, K., Fried, L. (1997).
Evaluation of a community based childhood injury
prevention program. Injury Prevention 3: 14-16.
Clamp, M., Kendrick, D. (1998). A randomised controlled
trial of general practitioner safety advice for families
with children under 5 years. British Medical Journal 316:
1576-9
Thompson, R., Edwards, P., Jarvis, S., Avery, A., et al.
(1998). Childhood accidents: is it time to prescribe safety
equipment? Community Practitioner 71 (4): 138-40.
Kendrick, D., Marsh, P., Fielding, K., Miller, P. (1999).
Preventing injuries in children: cluster randomised
controlled trial in primary care. British Medical Journal
318: 980-83.
Johnston, B., Britt, J., D’Ambrosio, L., Mueller, B., et
al. (2000). A preschool program for safety and injury
prevention delivered by home visitors. Injury Prevention
6: 305-9.
Mullan, C., Smithson, R. (2000). Community
childhood accident prevention project: using home
visits to promote child safety in deprived areas. Belfast:
Co-operation and Working Together.
King, W., Klassen, T., LeBlanc, J., Bernard-Bonnin, A.C.,
Robitaille, Y., Pham, B., Coyle, D., Tenenbein, M., Pless,
I. (2000). The effectiveness of a home visit to prevent
childhood injury. Pediatrics 108: 382-8.
References: Appendix Tables
17 Gielen, A., Wilson, M., McDonald, E., Serwint, J.,
Andrews, J., Hwang, W.T., Wang, M.C. (2001).
Randomised trial of enhanced anticipatory guidance for
injury prevention. Archives of Pediatric and Adolescent
Medicine 155: 42-9.
18 Campbell, N., Ayala, G., Litrownik, A., Slymen, D., et al.
(2001). Evaluation of a first aid and home safety program
for Hispanic migrant adolescents. American Journal of
Preventive Medicine 20 (4): 258-65.
C5(ii) Prevention of burns and scalds (N = 19)
1
2
3
4
5
6
7
8
9
Sorensen, B. (1976). Prevention of burns and scalds in a
developed country. Journal of Trauma 16 (4): 249-58.
Linares, A.Z., Linares, H.A. (1979). Burn prevention
programmes for children: are they effective? Burns 6 (2):
73-9.
i McLoughlin, E., Healer, C., Crawford, J.D. (1979). Burn
education intervention: a controlled study. Burns 6 (1):
26-9.
ii McLoughlin, E., Vince, C., Lee, A., Crawford, J. (1982).
Project burn prevention: outcome and implications.
American Journal of Public Health 72 (3): 241-7.
iii MacKay, A., Rothman, K. (1982). The incidence
and severity of burn injuries following Project Burn
Prevention. American Journal of Public Health 72 (3):
248-52.
Miller, R., Reisinger, K., Blatter, M., Wucher, F. (1982).
Pediatric counseling and subsequent use of smoke
detectors. American Journal of Public Health 72 (4): 3923.
Thomas, K., Hassanein, R., Christophersen, E. (1984).
Evaluation of group well-child care for improving burn
prevention practices in the home. Pediatrics 74 (5): 87982.
Eckelt, K., Fannon, M., Blades, B., Munster, A. (1985).
A successful burn prevention program in elementary
schools. Journal of Burn Care and Rehabilitation 6 (6):
509-10.
McLoughlin, E., Marchone, M., Hanger, L., German,
P., et al. (1985). Smoke detector legislation: its effect
on owner-occupied homes. American Journal of Public
Health 75 (8): 858-62.
Varas, R., Carbone, R., Hammond, J. (1988). A one-hour
burn prevention program for grade school children:
its approach and success. Journal of Burn Care and
Rehabilitation 9 (1): 69-71.
Katcher, M., Landry, G., Shapiro, M. (1989). Liquid-crystal
thermometer use in pediatric office counseling about tap
water burn prevention. Pediatrics 83 (5): 766-71.
63
10 Erdmann, T., Feldman, K., Rivara, F., Heimbach, M. et
al. (1991). Tap water burn prevention: the effect of
legislation. Pediatrics 88 (3): 572-7.
11 Grant, E., Turney, E., Bartlett, M., Winbon, C. et al.
(1992). Evaluation of burn prevention program in
a public school system. Journal of Burn Care and
Rehabilitation 13 (6): 703-7.
12 Fallat, M.E., Rengers, S.J. (1993). The effect of education
and safety devices on scald burn prevention. Journal of
Trauma 34 (4): 560-3.
13 Waller, A., Clarke, J., Langley, J. (1993). An evaluation of
a program to reduce home hot tap water temperatures.
Australian Journal of Public Health 17 (2): 116-23.
14 Mallonee, S., Istre, G., Rosenberg, M., Reddish-Douglas,
M., et al. (1996). Surveillance and prevention of
residential-fire injuries. New England Journal of Medicine
335: 27-31.
15 McConnell, C., Leeming, F., Dwyer, W. (1996).
Evaluation of a fire-safety training program for
preschool children. Journal of Community Psychology
24: 213-27.
16 Shults, R., Sacks, J., Briske, L., Dickey, P., et al. (1998).
Evaluation of three smoke detector promotion programs.
American Journal of Preventive Medicine 15 (3): 165-71.
17 DiGuiseppi, C., Slater, S., Roberts, I., Adams, L., et al.
(1999). The ’Let’s Get alarmed’ initiative: a smoke alarm
giveaway programme. Injury Prevention 5: 177-82.
18 King, L., Thomas, M., Gatenby, K., Georgiou, A., et al.
(1999). ’First Aid for Scalds’ campaign: reaching Sydney’s
Chinese, Vietnamese, and Arabic speaking communities.
Injury Prevention 5: 104-8.
19 Harré, N., Coveney, A. (2000). School-based scalds
prevention: reaching children and their families. Health
Education Research 15 (2): 191-202.
4
5
6
7
8
C6 Injuries in the leisure environment
(N = 7)
1
2
3
4
C5(iii) Prevention of poisoning (N = 8)
1
2
3
64
Krenzelok, E., Garber, R. (1981). Teaching poison
prevention to preschool children, their parents, and
professional educators through child care centers.
American Journal of Public Health 71: 750-2.
Sibert, J.R., Clarke, A.J., Mitchell, M.P. (1985).
Improvements in child resistant containers. Archives of
Disease in Childhood 60: 1155-7.
Woolf, A., Lewander, W., Filippone, G., Lovejoy, F.
(1987). Prevention of childhood poisoning: efficacy of an
educational program carried out in an emergency clinic.
Pediatrics 80 (3): 359-63.
Woolf, A.D., Saperstein, A., Forjuoh, S. (1992). Poisoning
prevention knowledge and practices of parents after a
childhood poisoning incident. Pediatrics 90 (6): 867-70.
Krug, A., Ellis, J.B., Hay, I.T., Mokgabudi, N.F., et al.
(1994). The impact of child-resistant containers on the
incidence of paraffin (kerosene) ingestion in children.
South African Medical Journal 84 (11): 730-4.
Rodgers, G. (1996). The safety effects of child-resistant
packaging for oral prescription drugs: two decades of
experience. Journal of the American Medical Association
275: 1661-5.
Liller, K., Craig, J., Crane, N., McDermott, R. (1998).
Evaluation of a poison prevention lesson for kindergarten
and third grade students. Injury Prevention 4: 218-21.
Eaton-Jones, S., Fields, K., Rasco, T., Akintemi, O. (2000).
Preventing childhood poisoning: an intervention in a
family medicine residency program. North Carolina
Medical Journal 61 (6): 333-5.
5
6
7
Morton, J.G., Burton, J.F. (1979). An evaluation of the
effectiveness of mouthguards in high-school rugby
players. New Zealand Dental Journal 75 (July): 151-3.
D’Argenio, P., Cafaro, L., Santonastasi, F., Taggi, F., et
al. (1996). Capodanno Senza Danno: the effects of an
intervention program on fireworks injuries in Naples.
American Journal of Public Health 86: 84-6.
Malinowska-Cieslik, M., van der Borne, B. (1998).
Prevention of mushroom poisoning of children:
effectiveness of a community-based school education
programme. Health Education Research 13: 13-23.
Withaneachi, D., Meehan, T. (1998). Promoting safer
play equipment in primary schools: evaluation of an
educational campaign. Health Promotion Journal of
Australia 8: 125-9.
Bennett, E., Cummings, P., Quan, L., Lewis, F. (1999).
Evaluation of a drowning prevention campaign in King
County, Washington. Injury Prevention 5: 109-13.
Roseveare, C., Brown, J., Barclay-McIntosh, J., Chalmers,
D. (1999). An intervention to reduce playground
equipment hazards. Injury Prevention 5: 124-8.
Sibert, J., Mott, A., Rolfe, K., James, R. et al. (1999).
Preventing injuries in public playgrounds through
partnership between health services and local authority:
community intervention study. British Medical Journal
318: 1595.
Injuries in children aged 0–14 years and inequalities
C7 Mass media and training intervention
(N = 5)
1
2
3
4
5
Williams, H., Sibert, J. (1983). Untitled. (Play it Safe
campaign and children’s injuries). British Medical Journal
286: 1893.
Gielen, A., Dannenberg, A., Ashburn, N., Kou, J., et al.
(1996). Teaching safety: evaluation of a children’s village
in Maryland. Injury Prevention 2: 26-31.
Sundelin, C., Rasmussen, F., Berfenstram, R., Troedsson,
K. (1996). Information through television: does it
promote child safety? Injury Prevention 2: 36-40.
Marsh, P., Kendrick, D. (1998). Injury prevention training:
is it effective? Health Education Research 13: 47-56.
Frederick, K., Bixby, E., Orzel, M., Stewart-Brown, S.,
et al. (2000). An evaluation of the effectiveness of the
Injury Minimization Programme for Schools (IMPS). Injury
Prevention 6: 92-5.
C8 Community-based intervention (N = 10)
1
2
3
4
i
Schelp, L. (1987). Community intervention and
changes in accident pattern in a rural Swedish
municipality. Health Promotion 2 (2): 109-25.
ii Svanström, L., Schelp, L., Ekman, R., Lindstrom, A.
(1996). Falköping, Sweden, ten years after: still a
safe community? International Journal for Consumer
Safety 3 (1): 1-7.
Guyer, B., Gallagher, S., Chang, B., Azzara, C., et al.
(1989). Prevention of childhood injuries: evaluation of
the Statewide Childhood Injury Prevention Program
(SCIPP). American Journal of Public Health 79 (11): 15217.
Schwarz, D., Grisso, J., Miles, C., Holmes, J., et al.
(1993). An injury prevention program in an urban
African-American community. American Journal of Public
Health 83 (5): 675-80.
i Davidson, L., Durkin, M., Kuhn, L., O’Connor, P.,
et al. (1994). The impact of the Safe Kids/Healthy
Neighborhoods Injury Prevention program in
Harlem, 1988 through 1991. American Journal of
Public Health 84 (4): 580-86.
ii Kuhn, L., Davidson, L.L., Durkin, M.S. (1994). Use
of Poisson regression and time series analysis for
detecting changes over time in rates of child injury
following a prevention program. American Journal of
Epidemiology 140 (10): 943-55.
References: Appendix Tables
5
i
Ozanne-Smith, J., Sherrard, J., Brumen, I., Vulcan,
P. (1994). Community based injury prevention
evaluation report: Shire of Bulla Safe Living Program.
Victoria, Australia: Accident Research Centre, Monash
University.
ii Hennessey, M., Arnold, R., Harvey, P. (1994). The first
three years: final report of the first three years of the
Shire of Bulla’s Safe Living Program (1991–1993).
Victoria, Australia: Shire of Bulla.
6 i Ytterstad, B. (1995). The Harstad Injury Prevention
Study: hospital-based injury recording used for
outcome evaluation of community-based prevention
of bicyclist and pedestrian injury. Scandinavian
Journal of Primary Health Care 13: 141-9.
ii Ytterstad, B., Sogaard, A. (1995). The Harstad Injury
Prevention Study: prevention of burns in small
children by a community-based intervention. Burns
21 (4): 259-66.
iii Ytterstad, B., Wasmuth, H. (1995). The Harstad
Injury Prevention Study: evaluation of hospital-based
injury recording and community-based intervention
for traffic injury prevention. Accident Analysis and
Prevention 27 (1): 111-3.
iv Ytterstad, B., Smith, G., Coggan, C. (1998). Harstad
injury prevention study: prevention of burns in
young children by community based intervention.
Injury Prevention 4 (4): 176-80.
7 Svanström, L., Ekman, R., Schelp, L., Lindstrom, A.
(1995). The Lidköping accident prevention programme
– a community approach to preventing childhood injuries
in Sweden. Injury Prevention 1: 169-72.
8 Day, L., Ozanne-Smith, J., Cassell, E., McGrath, A.
(1997). Latrobe Valley Better Health Project. Evaluation of
injury prevention program 1992–1996. Report No. 114.
Victoria, Australia: Accident Research Centre, Monash
University/Victorian Health Promotion Foundation.
9 Petridou, E., Tolma, E., Dessypris, N., Trichopoulis, D.
(1997). A controlled evaluation of a community injury
prevention project in two Greek islands. International
Journal of Epidemiology 26 (1): 173-9.
10 i Coggan, C., Patterson, P., Brewin, M., Douthett,
M. et al. (1998). Process evaluation report of the
Waitakere Community Injury Prevention Project.
Auckland, New Zealand: Injury Research Centre,
University of Auckland.
ii Coggan, C., Patterson, P., Brewin, M., Hooper, R., et
al. (2000). Evaluation of the Waitakere Community
Injury Prevention Project. Injury Prevention 6: 130-4.
65
Notes
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement