[Previous] [Contents] [Next]
13.1 INTRODUCTION AND BACKGROUND
Reducing the number of accidents is a priority for improving health in Scotland.1 However, achieving a reduction in the number of accidents needs some understanding of how accidents in general occur and to whom. In particular information is needed on the frequency and nature of accidents, about which very little is currently known. A module of questions in the Health Survey fills part of this gap, providing information about the number and causes of non-fatal accidents, the location of accidents and the injuries incurred, and providing detail about the personal characteristics of accident victims.
13.1.1 Accident classification and recall period
The term accident covers a very broad range of events ranging from, say, serious car accidents to, at the other extreme, accidents resulting in minor cuts or bruises. Some means is needed of identifying those of sufficient severity to be of importance in a health context. For the purpose of the Health Survey, accidents were considered to fall into the category of interest if advice about the accident was sought from either a doctor or at a hospital.2
Informants were asked to recall how many accidents they had had in the 12 months prior to interview about which they saw a doctor, nurse or other health professional. For those having had at least one accident, detailed information about the nature and cause of the most recent accident was collected.
The reference period of 12 months prior to interview was chosen so as to be sufficiently long to generate details of enough accidents for analysis purposes yet at the same time short enough for informants to accurately remember all, or at least a high proportion of, their accidents. The choice of reference period can be critical when making inferences from recalled events such as accidents. If the reference period is too long then there is a possibility of bias since some accidents from the earlier part of the reference period may be forgotten and, perhaps, other accidents from before the start of the reference period wrongly remembered as being within the reference period. An assumption is made throughout this chapter that the estimates presented are approximately unbiased: there is no internal evidence in the data about the validity of this assumption, and there is no satisfactory external data against which to validate the results.
13.1.2 Coverage of accidents
The survey does not cover all accidents. Fatal accidents are, by definition, excluded. In addition there will be under-representation of accidents that lead to long-term hospitalisation - in particular all accidents leading to twelve months hospitalisation will be excluded. Accidents that lead to a move out of the household population into the institutional population will also be excluded. Consequently, the accident rates presented in this chapter will be slight under-estimates of true accident rates, although, given that the great majority of accidents do not lead to death, long-term hospitalisation or a move to an institution, the downward bias should be small.
13.1.3 Derivation of accident rates and weighting to compensate for selection bias
Because any one informant may have had more than one accident during the reference period of twelve months, comparisons between sub-groups are based upon the mean number of reported accidents over the reference period. Such an approach tends to give very low figures (typically around 0.2) so, to aid comparisons, the means have been multiplied by a factor of 100 to give an annual accident rate per 100 persons.
Although informants were asked to specify the total number of accidents they had had over the reference period, detailed information was collected on only the most recent accident. Using the raw data to derive estimates about the kinds of accidents people have would lead to biases in the estimates in that this selection process leads to the over-representation of accidents to people for whom accidents are relatively uncommon events. To avoid this bias, analyses that use the detailed data on individual accidents use weighted data. The weight per informant is calculated as equal to the total number of accidents in the reference period.3 On average, the weights will be largest for those groups with the highest accident rates.4
13.2 THE INCIDENCE OF ACCIDENTS AND THEIR CHARACTERISTICS
13.2.1 The incidence of accidents by age and sex
Overall, 18% of men and 13% of women of working age had at least one accident about which they saw a doctor or visited a hospital in the twelve months prior to interview. Three percent of men and 1% of women reported having had more than one such accident.
For men, there was a very marked inverse relationship between age and the likelihood of having had an accident in the reference period, with 30% of men aged 16-24, and 22% of men aged 25-34, having had at least one accident. A similar, but weaker, relationship was observed for women, with 17% of women aged 16-24 having had at least one accident compared with a figure of around 11% for women aged 25 and over. Table 13.1
13.2.2 Estimated annual accident rates by age and sex
Table 13.1 also gives estimates of the annual accident rate per 100 persons by age and sex. The average accident rate was 24 per 100 for men and 15 per 100 for women. With the exception of those aged 55-64, men consistently had a higher accident rate than did women of the same age, but the difference was most marked for those aged 16-34, where the accident rate for men was double that for women of the same age. The trends are illustrated in Figure 13A. Table 13.1, Figure 13A
Figure 13A: Annual accident rates, by age and sex
13.2.3 Comparison with England
A comparison with data on accidents from the 1995 Health Survey for England5 showed very little difference in the age-sex specific accident rates between the two countries. Table 13.2
13.2.4 Cause of accident
Informants who had at least one accident in the reference period were asked to describe the cause of the accident. Their responses were coded by interviewers into nine categories of accident:
hit by a falling object
fall, slip or trip
road traffic accident
sports or recreational accident
use of tool or implement, or piece of electrical or mechanical equipment
burn or scald
animal or insect bite or sting
another person (including attacks)
If other cause was coded then the interviewer recorded a short description of the accident. On subsequent inspection of these descriptions it was found that accidents caused by lifting heavy objects were mentioned on enough occasions to merit being listed as an extra category of accident. Consequently, Table 13.3 includes lifting as a separate category.
Some caution is needed in the interpretation of the data on accident type and circumstances derived from this coding. What is coded in individual cases will depend firstly upon how the informant describes the accident, and secondly on how the interviewer interprets that description. For example, an accident in which a person sprains their ankle when jogging may be described as a fall by one informant (I fell and sprained my ankle) and as a sporting accident by another (I sprained my ankle when I was out jogging). If the informant describes the accident to the interviewer as I fell and sprained my ankle when I was running, then some interviewers may code this as a fall automatically whereas others may probe further, establish that running was actually jogging, and code it as a sporting accident. Interviewers were briefed to code more than one category per accident if appropriate, the intention being to collect as full a description of the accident as possible in order to avoid these sorts of classification problems.
One obvious implication of this ambiguity in coding is that rates cannot be reliably derived for different types of accident. For example, it would not be realistic to assume that all sporting accidents were coded as such and estimates of the number of sporting accidents occurring per 100 persons would be an underestimate.
Table 13.3 gives, by age and sex, the categories of accident recorded by interviewers. Some clear trends emerge. For example, accidents were more likely to be recorded as sports or recreational for men than for women, and for younger rather than older informants. The most commonly coded type of accident for women, across all age groups, was a fall or trip. Mens accidents were more evenly divided between falls, sports accidents and accidents caused by the use of a tool or other implement. Table 13.3
13.2.5 Location of accidents
The location of accidents also differed by age and sex. A higher percentage of the accidents suffered by women happened in the home or garden - 39% of accidents to women happened in the home or garden, compared with 24% of accidents to men. In contrast, a higher percentage of mens than womens accidents happened at a workplace or other public building, or in a place for sports or recreation.
As well as these differences by sex, there were also trends by age. In particular the percentage of accidents happening in a place for sports or recreation was significantly higher for men aged 16-34 and for women aged 16-24 than for those in older age-groups. Table 13.4
Table 13.5 shows the location of falls, slips and trips (the most common type of accident) by age. Table 13.5
13.2.6 Type of injury
Informants were asked to describe the injuries incurred in their most recent accident using twelve categories of injury:
straining or twisting a part of the body
cutting, piercing or grazing a part of the body
bruising, pinching or crushing a part of the body
swelling or tenderness in some part of the body
something stuck in the eye, throat, ear or other part of the body
burning or scalding
other injury to internal parts of the body
animal or insect bite or sting
For each accident, all relevant categories were recorded.
Table 13.6 gives the distribution of types of injury by sex and broad age band. Few, if any, of the differences were significant, the main types of injury recorded by both sexes and by the different age-groups being swelling or tenderness, straining or twisting, cuts and grazes, and bruising. Compared with accidents to women, accidents to men were slightly more likely to have resulted in cuts or grazes. For both sexes, around 15% of accidents resulted in broken bones. Table 13.6
13.2.7 Source of help or advice about accidents
By definition advice was given by either a doctor, nurse or other health professional for all of the accidents reported on (although in a proportion of cases the informant saw a nurse at a doctors surgery rather than the doctor). In 58% of cases help or advice was given at a hospital and in 54% of cases help or advice was given by a GP or a nurse at a GPs surgery. In 20% of cases advice was given by both a GP and at a hospital, although the order of visit was not established. Only in 8% of cases was advice given only by some other medical professional (such as a doctor at a workplace).
Table 13.7 gives the source of advice by age and sex. Some differences are apparent but there are no clear patterns. Table 13.8, however, which shows the source of advice by the type of area informants lived in,6 gives a much clearer pattern. For example, advice was sought from a hospital in respect of 71% of the accidents to those living in urban areas, compared with around half of those living in small country towns, villages or other rural areas. Suburban areas fell between these two extremes at 60%. Furthermore, advice was sought from a hospital but not a GP (thus excluding those referred to a hospital by a GP) in respect of almost half of the accidents suffered by those in urban areas, compared with less than one-third in rural areas. These figures suggest that where advice is sought from was dependent not only upon the severity and type of accident but also upon geographical location. Table 13.7, Table 13.8
Table 13.9 shows the relationship between type of injury and the source of advice. The patterns are largely as might be expected. Advice was sought from a hospital for most accidents that resulted in broken bones (88%) or that caused consciousness loss (83%). In contrast, for accidents resulting in straining or twisting of a part of the body, informants most commonly sought advice only from their GP (49%). Table 13.9
13.2.8 Prevention of accidents
Informants were asked to judge whether or not anything could have been done to prevent their accident. For 41% of accidents, informants judged that nothing could have been done. In a further 41% of accidents, informants judged they could have prevented it themselves, and in 21% of accidents informants judged that others could have prevented the accident. For a small proportion of accidents it was thought that the accident could have been prevented by both the informant themselves and by others.
To an extent, judgements about preventability differed according to the location of the accident. Informants judged that just 40% of accidents occurring in a place for sport or recreation could have been prevented, compared with around 60% for accidents occurring elsewhere. Table 13.10
13.3 ACCIDENTS AT WORK
13.3.1 Accidents at work by age and sex
For those in a paid job (71% of men and 58% of women), 46% of the accidents to men and 29% of the accidents to women, happened when they were at work. The sample numbers are too small to detect any trends by age in these figures. Table 13.11, Table 13.12
Rather more informative than the ratio between work-based accidents and non-work based accidents are the work-based accident rates themselves. These are calculated as the total number of work-based accidents in a twelve month period divided by the number of people in work, giving an accident rate per person in work. The denominator can be estimated from the survey data as the number whose main activity status at the time of interview was in paid employment. This leads to a slight upward bias in the estimates of work-based accident rates since the denominator excludes some persons whose main activity was other than in paid work but who nevertheless had a paid job. Some students for example might fall into this category. Nevertheless the biases should be small.
Table 13.13 gives the estimated work-based accident rates per 100 persons in work by age and sex. A very clear pattern emerges: men had higher work-based accident rates than women across all age groups, and, for both sexes, the work-based accident rate was particularly high for those aged 16-24. The average work-based accident rate for men was 12 per 100 and for women was 4 per 100. Table 13.13, Figure 13B
Figure 13B: Annual work-based accident rates, by age and sex
13.3.2 Causes of accidents at work
The most common cause of work-based accidents (as recorded by the interviewer - see Section 13.2.4 for a discussion of the interpretation of cause) was the use of a tool, implement or piece of electrical or mechanical equipment (35% of work-based accidents). One in five (21%) work-based accidents involved a fall or slip and 11% were recorded as being caused by lifting some object. There were no major differences in the distribution of causes by sex. Table 13.14
13.3.3 Accidents at work by SIC group7
Table 13.15 gives the estimated work-based accident rates by the SIC group of the informant. As might be expected, accident rates were significantly higher than the overall average for those working in the agriculture, forestry and fishing, manufacturing, and construction industries, at over 10 accidents per 100.8 Table 13.15, Figure 13C
13.3.4 Location of work-based accidents
Overall, 65% of all work-based accidents happened indoors. Not surprisingly this percentage differed by SIC group, from 42% for those working in agriculture, hunting and fishing, mining, electricity, gas and water supply, and construction combined, to 79% for those working in wholesale and retail trades. Table 13.16
13.3.5 Time off work
For 53% of accidents at work, informants reported having to take time off work. The sample numbers are too small for any differences by age or sex to be detected. (Table not shown.)
Figure 13C: Annual work-based accident rates, by SIC group
13.4 accidents outside of work
13.4.1 Accident rates for non-work accidents by age and sex
Although a fairly high proportion of accidents to informants occurred at work, the majority of accidents occurred outside of the workplace (64% of accidents to men and 83% of accidents to women, these percentages including informants who were not in work). The rates for accidents outside of work, by age and sex, are given in Table 13.17.
The pattern by age and sex was broadly the same for non-work accidents as for work-based accidents, with the accident rate being significantly higher than average for those aged 16-24 of both sexes. However, the difference between the sexes, both overall and within age-groups, was less marked, the non-work accident rates being only slightly lower for women (at 12 per 100) than for men (at 15 per 100). The figures suggest that women aged 55-64 actually had a slightly higher non-work accident rate than did men of the same age. Table 13.17, Figure 13D
13.4.2 Accidents outside of work and sports participation
As might be expected, given that a fairly high proportion of accidents happen in places for sports or recreation (see Table 13.4), there is an association between frequency of participation in sports and non-work accident rates. The correlation is particularly high for those aged 16-44, the accident rate for men of this age group rising from 14 per 100 for those who, in an average week, undertook no sports of a moderate or vigorous intensity, to 33 per 100 for those who, in a average week, undertook sports of moderate to vigorous intensity three or more times a week. The figures for women of this age group were 11 per 100 and 23 per 100 respectively. Table 13.18
Figure 13D: Annual non-work accident rates, by age and sex
13.5 accidents and social class
Table 13.19 gives work-based accident, non-work accident and all accident rates by age group, sex and social class.
There is no obvious relationship between social class and the non-work accident rate, the accident rate within groups defined by age and sex being broadly constant across the social classes.
In contrast, there is a very clear relationship between social class and the work-based accident rate, the accident rate being consistently above average for those in Social Classes IIIM, IV and V. The difference between the social classes was particularly marked for men aged 16-44: those in manual occupations had a work-based accident rate more than three times that for those in non-manual occupations. Table 13.19
13.6 regional variations in accident rates
Work-based accident, non-work accident and all accident rates by age group, sex and region are given in Table 13.20. Table 13.20
Although there appears to be some variation in the accident rates by region none of the observed differences were statistically significant. Nevertheless, there are differences between the regions that partially explain even the small variations observed. For example, men in Highland & Islands and in Borders, Dumfries & Galloway were more likely than average to be working in agriculture, forestry and fishing and in construction (the SIC groups with the highest work-based accident rates), whereas men in Greater Glasgow were more likely to be working in the service sectors where the work-based accident rate was lower than average. These differences partially explain the observed differences in work-based accident rates by region. (Table not shown.)
[Previous] [Contents] [Next]