| Health Survey for England - The Health of Minority Ethnic Groups '99 | |||||
|
12 Non-fatal accidents Sally McManus and Susan Purdon
12.1.1 Coverage This chapter presents findings from the 1999 Health Survey on the incidence of non-fatal accidents within minority ethnic groups. The main focus of the chapter is on differences between minority ethnic groups and the general population in terms of overall accident rates, but some analyses within minority ethnic groups by age, social class and income are also presented. For these latter analyses, the small sample sizes involved mean that only general patterns can be identified. The accidents covered by the Health Survey are self-reported accidents to the population living in private households. By definition fatal accidents are excluded. In addition, some accidents that result in long-term stays in hospitals or other institutions are excluded. 12.1.2 Accident classification and time periods Informants were asked about two types of accident: 'major accidents', which cause the informant to see a doctor or go to hospital, and 'minor accidents' which include all other accidents causing pain or discomfort for 24 hours or more. Informants were asked to recall how many major accidents they had had in the six months prior to interview and in which months these accidents had occurred. Data on the characteristics of the most recent major accident were collected but are not reported on here, the size of the sample of reported accidents by minority ethnic group being too small for reliable analysis. For minor accidents, the reference period was four weeks prior to interview. Informants were asked how many minor accidents occurred during each of the preceding four weeks. No data was collected on the characteristics of these minor accidents. In keeping with recent Health Survey reports, accident rates for minor accidents have been derived using just a two week recall period, since this reduces the observed recall bias in the data. (See the 1995 Health Survey1 report for a discussion of recall bias in the data on minor accidents.) 12.1.3 Calculation of accident rates Informants may have had more than one accident during the reference period asked about, and so comparisons between subgroups are based upon the mean number of reported accidents during the reference period. The figures involved in such an approach tend to be very low so, to aid comparisons, the means for major accidents have been multiplied by 200 (two 6-month periods for each of 100 persons). This allows the incidence of major accidents to be presented in terms of 'annual accident rates per 100 persons'. Minor accidents are also given as an annual rate per 100 persons, derived by multiplying the mean rate by 2600 (26 two-week periods x 100 persons). 12.1.4 Interpretation The definition of accidents given to informants was 'accidental events which resulted in injury or physical harm to you personally'. An 'accidental event' was not defined, but left to the informant to interpret. In most cases identifying which events are appropriate would be unproblematic, but there may have been cases where inclusion or exclusion was not obvious and depended on the informant's subjective perception. In particular, cultural or, in some cases, linguistic factors could have affected interpretation of the term 'accidental event', and thus contributed to variation in reported accident rates between minority ethnic groups. It is not possible to check whether or not the questions have been interpreted differently by different informants or groups of informants, and it has been assumed in the reporting of this chapter that the accidents recalled by different informants are comparable. 12.2 Accident rates by minority ethnic group 12.2.1 Accident rates among men For the general adult population the major accident rate for men was estimated to be 19 accidents per 100 persons. This was a higher rate than for any of the minority ethnic groups, although the rates for Black Caribbean men and Irish men were close at 15 and 16 per 100 persons respectively. The lowest accident rates were found amongst Bangladeshi and Chinese men, both at 7 accidents per 100. The major accident rates for the remaining two groups, Indian and Pakistani men, were both estimated at 12 accidents per 100. Minor accident rates were considerably higher than major accident rates, the general population minor accident rate for adult men being 200 accidents per 100 persons. As with major accidents, the minor accident rate within minority ethnic groups was lower than for the general population, although for the Irish the difference was very small, and certainly not statistically significant (at 191 minor accidents per 100). The lower rates for both major and minor accidents within minority ethnic groups strongly suggest that the lower major accident rates cannot be attributed to an unwillingness of those in minority ethnic groups to seek medical attention for accidents. If this had been the case we would expect to observe minor accident rates at the same or slightly higher rates within minority ethnic groups than in the general population. The variation by minority ethnic group for minor accidents broadly followed
that for major accidents, the Irish and Black Caribbeans having relatively
high accident rates (the rate for Black Caribbean men being 147 minor
accidents per 100 men), Bangladeshi men having the lowest reported accident
rate at 32 minor accidents per 100 men, and Indian and Pakistani men
being between these extremes at 69 per 100 and 55 per 100 respectively.
In contrast to their relative position on major accidents, the minor
accident rate for Chinese men was relatively high at 142 accidents per
100. There were considerable differences in the age-profiles of the minority
ethnic groups and, in theory at least, some of these differences could
account for all or some of the observed differences in accident rates.
Figures 12A and 12B show age-standardised accident rates presented as
rate ratios (i.e. the age-standardised accident rate for a minority
ethnic group as a ratio of the age-standardised accident rate for the
general population), with their 95% confidence intervals. These figures
suggest, however, a pattern across minority ethnic groups very similar
to that for the non-standardised rates. Taking major accidents as an
example, the age-standardised accident rate ratios for Black Caribbean
and Irish men were fairly high at 0.73 and 0.90 respectively, whereas
the rate ratios for Bangladeshi and Chinese men were much lower at 0.31
and 0.34 respectively. The rate ratios for Indian and Pakistani men
were 0.54 and 0.50 respectively.
12.2.2 Accident rates among women Previous Health Survey reports have shown women to have lower average accident rates than men, (although after retirement age accident rates for women start to exceed those for men). The 1999 Health Survey confirms this general pattern, both for the general population and within minority ethnic groups. The one exception to this rule is the Chinese population, Chinese men and women having an equal major accident rate, at 7 accidents per 100 persons. This 'exception' may be attributable to sampling error rather than to a genuine difference between the Chinese and other minority ethnic groups. The major accident rate among the general population of women was 15 accidents per 100. As in the case of men, the rate for the minority ethnic groups was consistently lower than the general population rate. Again, as for men, Black Caribbean and Irish women had the highest rates, both at 12 per 100. Indian, Pakistani and Chinese women had slightly lower rates at 10, 9 and 7 per 100 respectively, and Bangladeshi women had the lowest major accident rate at 3 per 100. This variation is broadly similar to that observed for men. Minor accident rates followed the same broad pattern. The minor accident rate for the general population was 120 accidents per 100 women. The accident rates for Black Caribbean and Irish women were fairly close to this at 134 and 172 per 100 respectively. Indian, Pakistani, Bangladeshi and Chinese women had lower rates at 76, 46, 28 and 28 per 100 respectively. As was found for men, standardising the accident rates for women by age does not significantly alter the relative size or ordering of the rates. The age-standardised major accident rate ratios for Black Caribbean and Irish women were relatively high at 0.80 and 0.88 respectively. The rate ratios for Indian, Pakistani and Chinese women were 0.64, 0.68 and 0.46 respectively. The rate ratio for Bangladeshi women was 0.32. 12.2.3 Types of major accident For major accidents, details about the most recent accident were collected through a series of specific questions. Informants were asked whether:
Using the responses to these questions, accident rates have been derived for the following (overlapping) categories of accident: falls (excluding those occurring whilst playing sport or exercising), sports accidents, accidents involving a tool or other implement, accidents involving a moving vehicle (including bicycles) and accidents at work. The estimated rates are shown on Table 12.2, both in absolute terms and as age-standardised rate ratios. After standardising for age, the major accident rate within minority ethnic groups was estimated to be consistently below the rate for the general population for all types of accident considered. This suggests that none of the accident types can alone account for the generally lower accident rates among minority ethnic groups. The few cases where the accident rates were not lower among minority ethnic groups tended not to be statistically significant. The main trends by each type of accident are summarised, in turn, below. Falls The rate of major accidents involving a fall was very similar for men and women for each of the minority ethnic groups, at between four and six falls per 100 persons for Black Caribbean, Indian, and Irish men and women, and for Pakistani women. The rate of falls was around two per 100 for Pakistani men and for Bangladeshis and Chinese of either sex. The age-standardised rate ratios ranged from 1.27 for Irish men (which is not significantly different from 1), to 0.15 for Chinese men. Sports accidents Black Caribbean and Irish men reported accidents when playing sport at only slightly lower rates than the general population (the age-standardised rate ratios being 0.86 and 0.92 respectively). Men in the other minority ethnic groups experienced considerably fewer sporting accidents than the general population, the rate ratios for Indian, Pakistani, Bangladeshi and Chinese men being 0.59, 0.37, 0.11 and 0.07 respectively. Women from all minority ethnic groups were significantly less likely to experience major sporting accidents than women in the general population, the largest of the age-standardised rate ratios being just 0.25. These findings parallel the different levels of participation in sport reported by minority ethnic groups (reported in Chapter 8), with South Asians and Chinese having much lower participation rates than the general population. Accidents involving a moving vehicle Although the data shows some variation in accident rates involving moving vehicles by minority ethnic group, there were no significant differences between those in minority ethnic groups and the general population, with the single exception that Bangladeshis of both sexes had significantly lower rates of accident. Accidents caused by the use of a tool, implement or piece of electrical or mechanical equipment All minority ethnic groups reported fewer accidents caused by the use of a tool or implement than the general population. Comparing age-standardised rate ratios, reported accidents were particularly low for Pakistanis and Bangladeshis. Accidents at work Work-based major accident rates are calculated per 100 persons in work at the time of interview. The work-based accident rate for men followed the same general pattern by minority ethnic group as the overall accident rate: Black Caribbean and Irish men had work-based accident rates lower, but close to, the rate for the general population (the age-standardised rate ratios were 0.84 and 0.86 respectively). The rate ratios for Indian, Pakistani, Bangladeshi, and Chinese men were lower at 0.49, 0.51, 0.51 and 0.62 respectively. At 3 accidents per 100 women in work, the work-based major accident rate for
women was only about a third of that for men. With the exception of
Black Caribbean women, for whom the work-based accident rate was estimated
at just 1 accident per 100, none of the rates within minority ethnic
groups were significantly different to the general population rate. 12.3 Major and minor accidents, by age Table 12.3 shows major and minor accident rates, for each sex, within three broad age groups: 16-34, 35-54 and 55 and over. In the adult male general population, both major and minor accident rates decreased rapidly with age. For women in the general population, there was a similar, but shallower, decline with age for minor accidents, whereas major accidents tended to decrease until the age of about 60 after which they slowly rose again. The general population figures of Table 12.3 confirm these age trends, although the very broad age groups conceal much of the detail. These general population age trends largely hold true for the minority ethnic groups but there are some notable exceptions. Some exceptions are reported on here, but it should be borne in mind that at least some of the apparent exceptions might be due to sampling error. The main divergences from the age trend are:
12.4 Major and minor accidents, by social class of head of household Table 12.4 shows major and minor accident rates by social class of the head of household. Because of small sample sizes social class has been divided into two broad categories: non-manual (social classes I, II and IIINM) and manual (social classes IIIM, IV and V). Age-standardised rate ratios are shown together with the absolute rates, the rate ratios being calculated as the age-standardised rate within a social class and minority ethnic group divided by the overall age-standardised rate for the general population. This allows for comparisons to be made between minority ethnic groups within social classes as well as between social classes within minority ethnic group. The analysis reported here concentrates on social class differences within minority ethnic group. After standardising for age, major accidents to men in the general population were found to be significantly higher for those with a manual head of household than for those with a non-manual head of household. Minor accident rates were also found to be higher than average amongst those with a manual head of household, but the difference was not statistically significant. For major accidents a similar association was found in the Black Caribbean, Indian, Pakistani, and Irish minority ethnic groups, although only for the Black Caribbeans was the difference statistically significant. In contrast, for Bangladeshi and Chinese men, higher rates of major accidents were found for those in non-manual households, the difference for Chinese men reaching statistical significance. No significant associations between social class and the minor accident rate were found within the minority ethnic groups. For women in the general population there was no significant association
between accident rates, either major or minor, and social class. Nor
were there any significant associations within minority ethnic groups,
with the possible exception of Chinese women in manual households for
whom the minor accident rate was higher than the average for Chinese
women. However, no difference in the major accident rate for
Chinese women was observed, which suggests the observed difference for
minor accidents may be attributable to sampling error. 12.5 Major and minor accidents, by equivalised household income For consistency with other chapters of this report, Table 12.5 shows major and minor accident rates by equivalised household income divided into three tertiles. Unlike social class, there is no statistically significant relationship between
major or minor accident rates and household income for the general population.
Given this context, it is not surprising that no clear relationship
between accidents and income is to be found within minority ethnic groups
(after controlling for age). A few of the trends observable in the age-standardised
rate ratios of Table 12.5 are statistically significant, such as the
trend in the minor accident rate for Indian men. However, since the
nature of the trend appears to differ from ethnic group to ethnic group
any apparently significant findings should be interpreted with caution. References and notes 1 Purdon S. Chapter 5: Non-fatal accidents in Prescott-Clarke P and Primatesta P (eds) The Health Survey for England 1995, The Stationery Office, London 1997. 12.1 Annual accident rates per 100 adults, by minority ethnic group 12.2 Annual major accident rates per 100 adults for types of accident, by minority ethnic group 12.3 Annual accident rates per 100 adults, by age within minority ethnic group 12.4 Annual accident rates per 100 adults, by social class of head of household within minority ethnic group 12.5 Annual accident rates per 100 adults, by equivalised household income tertile within minority ethnic group
| |||||
