Public Health Research Unit

Transcription

1 Scores for Scottish Postcode Sectors from the 1991 Census Public Health Research Unit Public Health Research Unit, University of Glasgow, 4 Lilybank Gardens, Glasgow, G 12 8RZ. Tel: Philip McLoone Public Health Research Unit First Published May 1994 Republished Nov 2000

2 Scores for Scottish Postcode Sectors from the 1991 Census Philip McLoone Public Health Research Unit, University of Glasgow First Published May 1994 Republished Nov 2000 The tables in this report are based on data provided by the Census Branch of the General Register Office for Scotland, the Vital Statistics Branch of the General Register Office for Scotland and the Information and Statistics Division of the NHS Common Services Agency. Census data were provided as part of the ESRC purchase of data from the 1991 Census. The Public Health Research Unit acknowledges their Assistance in the provision and analysis of the data. We wish to thank Vera for her helpful comments on an earlier draft of this report. This version of the report has been made available from the MRC Social & Public Health Sciences Unit, University of Glasgow About PHRU On the 1st October 1998 the Public Health Research Unit and the MRC Medical Sociology Unit were merged to form the MRC Social & Public Health Sciences Unit.

3 MRC Social and Public Health Sciences Unit Scores for Scottish Postcode Sectors from the 1991 Census Table of Contents Page INTRODUCTION 1 The Score 1 Variables used in Constructing the Scores 1 Creating the score 2 Defining Postcode Sectors 3 The DEPCAT variable 3 METHODOLOGICAL ISSUES 5 Problems of categorising small area populations 5 What do deprivation scores identify? 6 Robustness 7 Problems with interpreting changes between Censuses 10 An alternative formulation of the score 11 Comparing 1991 scores with those for ASSOCIATIONS WITH MORTALITY & MORBIDITY 14 Mortality 14 Morbidity 16 CONCLUSION 17 REFERENCES 18 APPENDICES 19 Appendix I 20 Appendix II 21 Appendix III 68 Appendix IV 71 Appendix V 72 Page i - Scores for Scottish Postcode Sectors from the 1991 Census

4 INTRODUCTION The Score scores are derived by combining variables taken from small area Census data and are described as a measure which reflects access to those material resources which provide access to "those goods and services, resources and amenities and of a physical environment which are customary in society" ( and Morris, 1991). The scores are not a measure of the extent of material well-being or relative disadvantage experienced by individuals but are rather a summary measure applied to populations contained within small geographic localities. The scores have generally been applied to the populations of postcode sectors, and are derived by manipulating selected Census variables in order to create a composite score. They are thus a method of quantifying levels of relative deprivation or affluence in different localities. Variables used in Constructing the Scores The scores have been calculated from the combination of four variables derived from the Small Area Statistics Tables of the 1991 Census using SASPAC (SASPAC manual, 1992). As nearly as possible, the variables used were the same as those employed in the creation of scores for postcode sectors based on the 1981 Census by and Morris and were defined as: Overcrowding: the proportion of all persons living in private households with a density of more than one person per room. Male unemployment: the proportion of economically active males seeking or waiting to start work Low social class: the proportion of all persons in private households with an economically active head with head of household in social class 4 or 5. No car: the proportion of all persons in private households which do not own a car. The cells of the Small Area Statistics Tables for 1991 used to create these variables are listed in Appendix I and are the equivalents of those used in creating the 1981 scores. Table 1 compares the 1981 and 1991 population weighted means and standard deviations for each variable. Page 1 - Scores for Scottish Postcode Sectors from the 1991 Census

5 Mean Standard Standard Mean deviation deviation No car ownership Male unemployment Over crowding Social class 4 & score Table 1: Population weighted mean percentages and standard deviations for each component variable used to create the score values are based on 1011 postcode sectors, 1991 values are based on 1001 postcode sectors. Three variables have the same definition as in In the 1991 Census, however, the definition of a room was different from that used in The computation of the number of rooms in a household in the 1981 Census excluded rooms used as ancillary kitchens, in 1991 kitchens of at least two metres wide were included in the definition with the effect that these kitchens increased the number of rooms reported*. In 1981, the mean value for the "overcrowding" variable was 25.3% but only 7.4% in 1991; a change of this order could only come about as the consequence of the change in definition. Comparisons of this measure between the two Censuses must therefore be made with caution. For the remaining three variables, there has been a small decrease in the percentage of those in households whose head of household was in social class 4 or 5 (a reduction of 3 %), and in the percentage of the population without access to a car ( a reduction of 7%). The mean proportion of unemployed males rose only slightly (by 0.4%), but the variation between postcode sectors for this measure increased (a standard deviation of 7.3 in 1981 and 8.4 in 1991). Creating the score Following the method of and Morris (1991), each variable was standardised to have a population weighted mean of zero and a variance of one. The reason for this procedure was in order to control the relative contribution of each variable in the score; a variable with a relatively large variance would result in the eventual score being unduly influenced by that variable. Standardising each to have a variance of one means that each variable has an equal influence on the resultant score. Standardisation was achieved by subtracting the all-scotland population mean from each variable and dividing the result by the standard deviation of that variable. This is called the z- score method and the sum of these new standardised variables produces the deprivation score for each postcode sector. The lower part of Table 1 illustrates how the all-scotland mean deprivation score is zero for both 1981 and 199 1, with a standard deviation of 3.61 in 1981 and a slightly smaller value of 3.51 in As an example of how the score for a single postcode sector is derived, Table 2 lists the observed percentage values and standardised values for postcode sector G12 8 which has a deprivation score of The score is the sum of the standardised values (z-scores): 1.27, 1.41, 0.96 and = Page 2 - Scores for Scottish Postcode Sectors from the 1991 Census

6 Observed percentage Mean Standard deviation Standardisation (z-score) No car ( )/17.76 = 1.27 Male unemployment ( )/8.35 = 1.41 Overcrowding ( )/4.38 = 0.96 Social class 4& ( )/8.63 = Table 2: Percentages and z-scores for postcode sector G12 8 The values of each of the four component variables and the resultant score for each of the 1001 postcode sectors in 1991 are listed in Appendix II by Health Board and Government District. For comparison, the scores calculated from the 1981 Census data are included. Appendices III & IV list the mean values of the deprivation scores for Government Districts and Health Boards. Defining Postcode Sectors In Appendix II, postcode sectors in 1981 and in 1991 were matched using a combination of the district identifier and the 2 character sector identifier taken from the header record of the matrix output file format of SASPAC (SASPAC manual, 1992). A difficulty in making direct comparisons is that a restructuring of postcode sectors has occurred in the Grampian Region and so, in some instances, the matching of these sectors may be incorrect. For example, although the matching of postcode sector AB4 0 in 1981 to AB41 0 in 1991 appears to be reasonable (with populations of 99 and 103 respectively), the matching of AB2 0 with AB2 0 does not (the population was 531 in 1981 but 5017 in 1991). Care must be taken when comparing Grampian postcode sectors between the two Censuses. The DEPCAT variable The 1981 distribution of deprivation scores for postcode sectors was restructured as a categorical variable, DEPCAT, which ranged from DEPCAT 1 (the most affluent postcode sectors) to 7 (the most deprived). This division of the scores for 1981 was performed on an arbitrary basis and, for this reason, a comparable division of the 1991 scores could only be achieved by dividing the new distribution into a similar number of DEPCAT groups containing the same proportions of the whole population as in The DEPCAT category of each postcode sector is included in Appendix H. Table 3 summarises the distribution of the population and postcode sectors in each of the 7 DEPCAT categories for 1981 and *This change in definition brings the reporting of rooms into line with England & Wales. Page 3 - Scores for Scottish Postcode Sectors from the 1991 Census

7 DEPCAT Population % Postcode sectors % Population % Postcode sectors % 1 305, , , , ,095, ,090, ,282, ,270, , , , , , , Total 5,032, ,998, Table 3: Frequency distribution of populations and postcode sectors contained within each DEPCAT category (1981 & 1991). Page 4 - Scores for Scottish Postcode Sectors from the 1991 Census

8 METHODOLOGICAL ISSUES Problems of categorising small area populations Several interrelated problems suggest that a measure of area deprivation derived from Census variables in this way should be used cautiously. They include: the way in which variables are chosen and the reasons for choosing some and not others; technical problems in creating the scores; in particular the way in which the selected variables interact, and the consistency of these interactions across the whole range of different population sub-groups; the a priori assumption that, within the selected or available data, there exists a simple univariate level of deprivation that allows localities and their populations to be quantified in these terms. The choice of variables used in the score has been justified on the basis of theoretical assumptions about their relationship to unmeasured or unmeasurable concepts such as wealth, or poor access to material resources (see and Morris, 1991; pp.4-11). This means that the derived census variables become empirical indicators of a poorly defined concept. Such assumptions are difficult to refute or validate and the chosen variables could also be associated with other variables that differentiate areas, for example, their demography. This comment introduces the problem of knowing whether, in different types of localities, the chosen variables consistently reflect the concepts with which they have been associated. Car ownership, for example is associated with population density, and varies with level of urbanisation after controlling for household income (Bates et al, 1981). These observations cast doubt on the assumption that lack of a car can be equated with low income in simple or direct ways. The variables used in the index are more strongly associated with socio-economic circumstances than, for example, demographic variables such as those describing age groups or family structures which may also vary between localities. However, other variables available from the Census - such as housing tenure - are also strongly associated with the socio-economic characteristics of small areas. There is no valid statistical method for choosing some Census variables in preference to others as a measure of the relative affluence or deprivation of small areas. However, statistical methods such as principle components analysis or factor analysis are useful in allowing the relationships between sub-groups of variables to be explored. These methods make it possible to identify different dimensions by which areas vary while still taking account of the multivariate nature of the Census data. Principle components analysis of the 1981 Census identified five such dimensions, of which one could be described as a "socioeconomic component" with similarities to the scores (McLoone and Boddy, 1992). A second principal component was associated with population density, transport to work, and types of housing typically found in rural areas (such as tied cottages); a third with young families and children; a fourth with older people and pensioners living alone and a fifth with Page 5 - Scores for Scottish Postcode Sectors from the 1991 Census

9 larger families, female unemployment and overcrowded housing. The socio-economic dimension on which the index ranks areas is thus only one of several ways in which localities differ and which may have important implications for the planning of health services and the provision of health care. There is a tautological danger in selecting variables which best predict outcomes since these chosen variables may not necessarily reflect the relationship between outcomes and what the derived index is believed to represent. What do deprivation scores identify? Deprivation indices (such as the scores) rank small geographical areas along a univariate dimension by using aggregated information about the individuals living in these areas expressed as a proportion of individuals with one or other attribute. Figure 1 represents three hypothetical localities (the three polygons) ranked on a deprivation scale. Within each, the circles represent households with/without one of the variables employed in constructing the score. In the left and right polygons - the extremes of the scores - the proportion of households with or without this attribute is greater; that is, their internal populations are increasingly homogeneous. In the centre of the scale, the mix of characteristics is relatively heterogeneous and it is this heterogeneity that locates the score in the middle of the scale. An important feature of scales derived in this way, therefore, is that areas need to be internally homogeneous - at least for some variables - for that area's score to be extreme, and thus identified as being affluent or deprived. Figure 1: Illustration of three areas ranked on a deprivation scale. Circles represent the presence or absence of an indicator of deprivation within households. Scores at the middle of the scale result from the mix of household types contained within these areas and define the majority of postcode sectors; 62% of the Scottish population live in areas designated as DEPCAT3, 4 or 5. As Figure 1 suggests, however, these areas contain deprived households and, because the population of postcode sectors varies Page 6 - Scores for Scottish Postcode Sectors from the 1991 Census

10 considerably, it could be that these "middle" areas contain more deprived households than some of the sectors identified at the deprived extreme. This problem of heterogeneity arises as a result of the way in which postcode sectors are defined. Postcodes boundaries are devised to enable the Post Office to deliver mail; the problem of their heterogeneity is a consequence of three factors; their geographical size, the size of their populations and the spatial location of postcode sectors within Scotland. For these reasons, the scores are likely to provide a better account of affluence or deprivation in urban rather than rural areas. The former are geographically smaller and socially more homogeneous; the latter are geographically larger and include populations with much more heterogeneous socio-economic characteristics. It might be argued that if smaller geographical units - such as Census enumeration districts - were used, more internally homogeneous areas would be result. However, the boundaries of enumeration districts are not defined by the characteristics of the populations which they include and there is no reason to assume that greater homogeneity would result from their use. Robustness The score is derived by summing the standardised value of each variable taken from the Census. Standardisation between the variables is used to control the relative contribution of each variable so that the eventual score is not unduly influenced by a large variance in one of them. The reason for doing so concerns the relative importance of one variable or another in defining deprivation on an area basis. The method thus involves a judgement about the relative weights of the variables used and has important, but uncertain, implications for the resultant distribution of the composite scores. The practical importance of this question has to do with the stability of the scores (are they a true measure?) and with Figure 2: Randomly weighted scores (100 per postcode sector) plotted against the 1981 scores. Page 7 - Scores for Scottish Postcode Sectors from the 1991 Census

11 their ranking one against another (does this postcode sector really belong at this point in the scale?). Figure 2 illustrates what happens when the weight given to each variable is varied. Four random weights, constrained to sum to four, were generated and assigned to each variable as it was measured in The resultant scores, derived from the summation of the four weighted standardised variables, and the rankings that they produced were noted for each postcode sector. This procedure was carried out one hundred times using different sets of randomly generated weights on each occasion thus producing a set of one hundred scores and ranks for each postcode sector. The 101,000 scores obtained by this method were plotted against the original 1981 scores and are shown in Figure 2. The broad diagonal in this figure reflects a measure of the agreement/disagreement between the scores and the randomly generated scores. What is important in Figure 2 is the variation within the scores for each postcode sector - this can be seen on the vertical axis of the figure. It is evident that some postcodes can have widely varying scores, this is especially apparent in the middle of the scale where the scores for some areas range from less than -6 to more than 14. In this illustration, the overall mean difference between the minimum and maximum scores for each sector was about 2.5 The use of different weights for the variables influences the ordering of postcode sectors along the horizontal axis. Figure 3 plots the rankings obtained from each of the randomly weighted scores plotted against the original 1981 ranking and presents a similar picture. There is again a large amount of variation with some areas changing rank quite dramatically. The importance of this figure is in demonstrating that varying the weights given to individual variables creates a considerable potential for change in the ranking of areas along the horizontal axis. Figure 3: Ranks derived from randomly weighted scores (100 per postcode sector) plotted against ranks obtained from the 1981 scores. Page 8 - Scores for Scottish Postcode Sectors from the 1991 Census

12 Figure 4: Plot of the standard deviations of the ranks obtained from randomly weighted scores for each postcode sector against the ranks obtained from the 1981 scores (postcodes with populations < 2000 are excluded). Figure 4 shows the scatter of the standard deviations derived from the distribution of ranks for each sector plotted with the ranking of the original score. Postcode sectors with populations of less than 2000 have been excluded. It is apparent from this figure that for those areas ranked lowest (the most affluent) and for those areas ranked highest (the most deprived) the standard deviations of ranks generated by assigning random weights are very small compared to those in the middle of the range. Areas in the middle (the majority) can have large variation in their relative ranking. The data presented in Figures 2-4 suggest that the scores generated by the method are not robust because changes in weights can have large effects on the scores or the rankings (and, thus, on their DEPCAT categories). Those areas that are ranked "worst" or "best" are, however, generally more consistent in their ranks. Around 50 areas were consistently ranked among the most deprived 100 and a similar proportion for the most affluent. Figure 5 suggests one of the reasons for this instability in scores and ranks. This graph shows the standard deviation of the randomly generated scores for each postcode sector plotted against the population size of each sector. It is apparent that those sectors with small populations can have large standard deviations. Although not all postcode sectors with small populations behave in this way, this observation questions the argument that smaller geographical units would provide more reliable scores. Page 9 - Scores for Scottish Postcode Sectors from the 1991 Census

13 Figure 5: Standard deviations of randomly weighted scores for each postcode sector plotted against population size Problems with interpreting changes between Censuses Because of the way the scores are constructed, the 1991 scores have the same mean and variance as in 1981 unless the covariance between each of the four variables used has changed dramatically; in other words, unless the relationships between the four variables used to construct the scores has changed. Table 4 shows that only a slight change has occurred in correlations with low social class and virtually no change in the correlations between the other variables over the 10 year inter-censal period. This means that the variance of the distribution of scores in 1991 and 1981 is very similar: 13.0 in 1981 and 12.3 in 1991 (Table 1 ). The constraint placed on the variance of the scores by the z-score method creates difficulties in assessing whether area inequalities have increased or decreased between the two Censuses. Although the score provides a relative measure of deprivation within each year, a comparison of the scores for 1991 with those for 1981 does not provide an interpretable account of changes between the two periods. By standardising each variable to have unit variance within a year, the method effectively irons out changes occurring between the two dates; this can mean, for example, that a sector's score has decreased, even though the corresponding proportions for its component variables have increased. Such a situation could occur if a few sectors have increased their scores substantially. An increase in the score may occur because the internal structure of an area has become more homogeneous in its characteristics but this does not necessarily mean that individuals within that population have become more deprived. Because the score measures Page 10 - Scores for Scottish Postcode Sectors from the 1991 Census

14 deprivation by the presence or absence of certain criteria, it simply means that the relative proportions of the component variables have increased. This could happen for at least two potential reasons: one is due to changes in the population of the area (that is, the populations in 1981 and 1991 are not the same, perhaps due to migration), and the other as the consequence of a real increase in the inequality experienced by individuals within the area. Both explanations would indicate the increasing marginalisation of the locality, but only the latter explanation would indicate increasing disadvantage. For these reasons, changes in scores between 1981 and 1991 should be interpreted with caution. Correlations in 1981 Correlations in 1991 Male Unemployment Overcrowding Social Class 4 & Cairstairs Score No car Unemployment Overcrowding Social Class 4 & 5 No car Unemployment Overcrowding Social Class 4 & 5 Table 4: Correlation between component variables and the score in 1981 and 1991 An alternative formulation of the score The difficulties described above can be overcome by using a simpler alternative to the scores. This involves adding each of the original variables together, without standardisation, and then dividing the sum for each area by the all Scotland mean value. This method produces a ratio in which the whole of Scotland has a value of one, and postcode sectors have a range of values which can be interpreted in relation to this Scottish average. For example, a postcode sector with a value of 2 would indicate "deprivation" that is 100% above the Scottish mean, and a value of 0.5 would indicate "deprivation" 50% below the Scottish mean value. Such a ratio can be interpreted in much the same way as standardised mortality ratios and allows comparisons between Censuses. There are, of course, difficulties with such a ratio. A major problem has to do with the fact that a score produced in this manner will be weighted in favour of those variables which demonstrate the greatest variability between areas. In the case of the four variables used in the score, this would mean that car ownership has greatest weight in deriving the ratio. This is another expression of the problem of choosing weights when constructing scores of this kind. The index has been described as an unweighted score but this is a misleading description because each variable receives a weight of one. The suggested reason is that, without further knowledge, no judgement can be made about the relative importance of one variable over another. This apparent "non-choice" is in fact a real choice of weights and the decision to assign equal weights to each variable is as specific as a decision to assign different weights to them. The fact that all the weights are the same does not make this choice of weights less contentious. Page 11 - Scores for Scottish Postcode Sectors from the 1991 Census

15 By simply adding each variable together, without standardisation, a decision about weighting does not have to be made. The resultant score simply weights itself in favour of those variables which demonstrate greatest variability between areas. In the case of the four variables used in the score, each probably does relate to a measure of the socioeconomic status of localities. It follows from this argument that the variables which show greatest variability are those which should receive greatest weight. Setting these considerations aside, the ratio created in the above manner has a correlation of and with the respective scores from the 1991 and 1981 Censuses. This ratio provides a simple and readily interpretable alternative to the score Comparing 1991 scores with those for 1981 Figure 6 is a plot of the scores calculated in 1991 against those derived from the 1981 census. There is a strong association between them (r = 0.958) and most areas do not show a substantial change in their scores. Areas with small populations, however, do tend to show large changes between 1981 and 1991; the explanation probably relates to the earlier observation that the robustness of scores is affected by the size of postcode sector populations. Figure 6: 1991 scores plotted against the 1981 scores The slight curve in the line fitted in Figure 6 is highlighted in Figure 7 which plots the difference between the 1991 and 1981 scores against the mean of the 1981 and 1991 scores. For the reasons described above, it is difficult to interpret what these differences might mean. A simple interpretation of the two figures would be that deprived areas have become more deprived and that affluent areas have also become more deprived. Because the scores for the two years are not directly comparable, this may not be a valid inference. Page 12 - Scores for Scottish Postcode Sectors from the 1991 Census

16 Figure 8 is a similar picture but uses the ratio score described above. The line fitted to this plot show a more pronounced curve but suggests that areas at the more affluent end of the range have generally remained the same (with differences at about zero) but that deprived areas have become more deprived. The correlation between the ratios calculated from the 1981 and 1991 Censuses was Figure 7: The difference between the scores in 1991 and 1981 plotted against the mean of the 1981 & 1991 scores (postcode sectors with populations < 2000 are excluded) Figure 8: Difference in ratio scores ( ) plotted against the mean of the 1981 & 1991 ratio scores (postcode sectors with populations < 2000 are excluded) Page 13 - Scores for Scottish Postcode Sectors from the 1991 Census

17 ASSOCIATIONS WITH MORTALITY & MORBIDITY All-cause standardised mortality ratios (SMRs) and all-cause standardised hospital discharge ratios (SDRs) for postcode sectors divided by Health Board and Government District, together with their 1991 score and DEPCAT category, are listed in Appendix V. Mortality The strong association between deprivation and mortality is demonstrated in the scatter diagram in Figure 9. This figure plots age and sex standardised mortality ratios (SMRs) for all causes in the age range * for the years ** against the 1991 score. There is a linear trend in mortality with increasing deprivation and the correlation between the two measures is The figure includes some areas with very large SMRs (more than 200) but this is simply due to areas with small populations producing SMRs with large standard errors. The corresponding correlation with mortality during and the scores for 1981 was Figure 9: Age & sex standardised mortality ratios (ages, years ) plotted against scores for All Scotland = 100 (Note: the figure shows some sectors with very large SMRs due to small populations producing ratios with large errors). * The 1991 Census contains an element of under counting. This age range was chosen because population estimates for younger age groups are less reliable. * * The original intention was to use the peri-censal years However, Grampian postcodes were restructured in 1991, and as no simple method could be found for assigning deaths in 1990 to the new sector boundaries, it was decided to exclude Page 14 - Scores for Scottish Postcode Sectors from the 1991 Census

18 Table 5 shows the SMRs for ages in and in for each DEPCAT grouping. In , the SMRs ranged from 63 in DEPCAT 1 to 139 in DEPCAT 7. For , the SMRs exhibit a wider range - from 61 for DEPCAT 1 to 159 in DEPCAT 7. The areas categorised as most deprived have increased their SMRs by 20 percentage points in the 10 year period between Censuses. This increasing mortality differential is illustrated in Figure 10 which shows the linear relationship between mortality and the deprivation index in and When compared to , the slope of the line for has increased with some improvement in SMRs for more affluent localities and a comparative worsening of those for deprived areas. The extent of these changes for the DEPCAT categories is indicated in Table 6; those to the left of the mid-point (the more affluent) show small decrements in their SMRs whereas those to the right (the more deprived) show rather larger increments. The most striking feature of the Table, however, is the substantial increase (+23) in the SMR for those areas categorised as DEPCAT 7 in Deprivation category SMRs ( ) SMRs ( ) Difference 90s - 80s Table 5: Age & sex standardised mortality ratios within each DEPCAT category. Ages All Scotland = 100. Figure 10: Linear regression lines showing the relationship between the scores in 1981 and 1991 with standardised mortality ratios for and Ages. Page 15 - Scores for Scottish Postcode Sectors from the 1991 Census

19 Difference in SMRs (1990s-1980s) Deprivation category in Table 6: Difference in age & sex standardised mortality ratios ( minus ) within each DEPCAT grouping as classified in 1981 for ages. Morbidity In their account of the 1981 scores, and Morris provide very detailed analyses of their relationship to measures of hospital use in Scotland. These analyses have not been repeated for this Report but Table 7 sets out standardised discharge ratios (SDRs) derived from continuous inpatient stays for the seven DEPCAT categories for ages in Discharge counts were obtained from the linked SMR1 database *. As with mortality, there is a strong gradient across the range; DEPCAT 7 (the most deprived group) had an SDR of 135 compared to an SDR of 69 for DEPCAT 1. For technical reasons, a direct comparison of these ratios with the data reported by and Morris for 1981 has not been possible. The correlation coefficient of SDRs for postcode sectors with their scores for 1991 was Deprivation category in SDRs (1991) Table 7: Age & sex standardised discharge ratios within each DEPCAT category. Ages. All Scotland = 100. * The linked SMR1 database links together all SMR1 records for a patient. A new SMR1 record is completed every time a patient is transferred between consultants, specialties or hospital. The Medical Record Linkage Team at ISD has developed rules for combining the separate SMR1 records generated by an individual during a stay in hospital into what is termed a continuous inpatient stay. Page 16 - Scores for Scottish Postcode Sectors from the 1991 Census

20 CONCLUSION The scores in this Report provide a relative measure of deprivation or affluence when this is judged on the basis of a combination of selected Census variables standardised to their mean for Scotland. The score for a particular postcode sector is thus a summary measure of its socio-economic status relative to the average for Scotland as a whole. It is important to appreciate that they refer to the populations of the postcode sectors and that they are based on the proportions of individuals within them who have reported a particular attribute at the time of the Census. Scores based on the 1991 Census are essentially the same as the scores from the 1981 Census. However, the relative ranking they produce has changed and this has resulted in some postcode sectors changing DEPCAT category. This is partly due to the poor robustness of the scores. The scores have been constructed in the same way as the 1981 scores and, as in 1981, there is a good measure of correlation between the component variables. Although the definition of overcrowding changed between the two Censuses, the correlation between this variable and the other variables remains high. A difference between the two years, however, is that the correlation between the proportions in social classes 4 and 5 with other variables is less in 1991 than in This change could reflect the changing validity of this occupational classification as a measure of socio-economic status. Any classification of this kind depends on the nature of the data from which it is constructed. In the case of the scores, it is important to keep in mind that, although they reflect the socio-economic characteristics of localities, there are other features of small areas - such as their demography and urban/rural differences - that are also important for health planning. Because the scores employ the proportions of people with a particular attribute within a population, the question of the heterogeneity or homogeneity of these populations is an important one. Urban localities are probably more homogeneous than rural areas and so the fact that rural postcode sectors are more likely to have middle-range scores does not necessarily mean that their populations can be regarded simply as "average". The size of postcode sectors varies considerably and, as Figure 5 demonstrates, the scores for those with small populations can have large standard deviations; this observation has implications for their rankings. Simple interpretations of the change in scores between 1981 and 1991 are not possible because of the way they are constructed. Comparisons of an alternative method of ranking, however, suggests that the socio-economic circumstances of localities at the "affluent" end of the range remained much the same when compared to the Scottish average and that there has been a deterioration of the relative circumstances of the most deprived areas in the decade between 1981 and 1991 (Figure 8). A similar worsening of standardised mortality ratios is observed, especially in DEPCAT7 where the SMR at ages for those areas categorised as DEPCAT7 in 1981 has increased from 139 to 162. The linear regression lines in Figure 10 illustrate this increasing divergence across the range of postcode sectors. Page 17 - Scores for Scottish Postcode Sectors from the 1991 Census

21 REFERENCES Bates J, Roberts M, Lowe S, Richards P. The factors affecting household car ownership. Gower Publishing Company Limited, London, V, Morris R. Deprivation and health in Scotland. Aberdeen University Press, Aberdeen, McLoone P, Boddy FA. Categorising small geographical areas. Public Health Research Unit, SASPAC Manual vv. 1-2, London Research Centre and Government Management Board, London, Page 18 - Scores for Scottish Postcode Sectors from the 1991 Census

22 APPENDICES I Cells taken from the 1991 Small Area Statistics Tables to create the component17 variables used in the construction of the score. II Values of each of the four component variables used to construct the score, the derived scores, DEPCAT values and population counts for 1981 & 1991 for each postcode sector, indicating Government District and Health Board. III Mean values of each of the four component variables used to construct the score, mean deprivation scores and population counts for 1981 & 1991 for each Government District, indicating Health Board. IV Mean values of each of the four component variables used to construct the score, mean deprivation scores and population counts for 1981 & 1991 for each Health Board. V 1991 scores, DEPCAT values, observed deaths ( ) & hospital 58 discharges (1991) with standardised mortality & hospital discharge ratios showing 95% confidence intervals for each postcode sector, indicating Government District and Health Board. Page 19 - Scores for Scottish Postcode Sectors from the 1991 Census

23 Appendix I Census count Cell Derived variable Formula Total persons in private households S Population weight S Total persons in households with no car Total residents in households S S Proportion of all persons in private households which do not own a car S210045/S Total residents in households with >1 and up to 1.5 persons per room Total residents in households with >1.5 persons per room Total residents in households SS SS SS Proportion of all persons living in private households with a density of more than one person per room (SS SS230048)/SS Unemployed males Total economically active males S S Proportion of economically active males seeking work S080078/S Total residents in households with head of household in social class 4 Total residents in households with head of household in social class 5 Total residents in households with an economically active head S S S Proportion of all persons in private households with an economically active head, with head of household in social class 4 or 5 (S S900032)/S Page 20 - Scores for Scottish Postcode Sectors from the 1991 Census

24 Appendix II Health Board district Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population Argyll & Clyde Argyll & Bute FK208 FK G83 7 G PA200 PA PA209 PA PA212 PA PA223 PA PA237 PA PA238 PA PA248 PA PA258 PA PA268 PA PA278 PA PA286 PA PA296 PA PA308 PA PA318 PA PA328 PA PA331 PA PA344 PA PA345 PA PA351 PA PA PA371 PA PA384 PA PA417 PA Page 21 - Scores for Scottish Postcode Sectors from the 1991 Census

25 Appendix II (continued) Health Board district Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population PA427 PA PA437 PA PA447 PA PA457 PA PA467 PA PA477 PA PA487 PA PA497 PA PA607 PA PA617 PA PA626 PA PA PA646 PA PA656 PA PA666 PA PA676 PA PA PA PA706 PA PA PA726 PA PA736 PA PA PA756 PA PA766 PA PA776 PA PA786 PA Page 22 - Scores for Scottish Postcode Sectors from the 1991 Census

26 Appendix II (continued) Health Board district Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population Argyll & Clyde Dumbarton G60 5 G G G82 1 G G82 2 G G82 3 G G82 4 G G82 5 G G83 0 G G83 7 G G83 8 G G83 9 G G84 0 G G84 7 G G84 8 G G84 9 G Argyll & Clyde Inverclyde PA PA113 PA PA134 PA PA145 PA PA146 PA PA151 PA PA152 PA PA153 PA PA154 PA PA160 PA PA167 PA Page 23 - Scores for Scottish Postcode Sectors from the 1991 Census

27 Appendix II (continued) Health Board district Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population PA168 PA PA169 PA PA186 PA PA191 PA Argyll & Clyde Renfrew G G G53 7 G G G78 1 G G78 2 G G78 3 G G78 4 G PA1 1 PA PA1 2 PA PA1 3 PA PA102 PA PA113 PA PA124 PA PA PA146 PA PA2 0 PA PA2 6 PA PA2 7 PA PA2 8 PA PA2 9 PA PA3 1 PA PA3 2 PA Page 24 - Scores for Scottish Postcode Sectors from the 1991 Census

28 Appendix II (continued) Health Board district Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population PA3 3 PA PA3 4 PA PA4 0 PA PA4 8 PA PA4 9 PA PA5 0 PA PA5 8 PA PA5 9 PA PA6 7 PA PA7 5 PA PA8 6 PA PA8 7 PA PA9 1 PA Ayrshire & Arran Cumnock & Doon DG DG KA KA181 KA KA182 KA KA183 KA KA184 KA KA4 8 KA KA5 5 KA KA5 6 KA KA6 6 KA Page 25 - Scores for Scottish Postcode Sectors from the 1991 Census

29 Appendix II (continued) Health Board Ayrshire & Arran district Cunninghame Post code sector score DEPCAT No car Unempl -oyment Over crowding Social class 4&5 Population G KA111 KA KA112 KA KA113 KA KA114 KA KA115 KA KA120 KA KA128 KA KA129 KA KA136 KA KA137 KA KA143 KA KA151 KA KA152 KA KA2 0 KA KA203 KA KA204 KA KA215 KA KA216 KA KA227 KA KA228 KA KA239 KA KA244 KA KA245 KA KA256 KA KA257 KA Page 26 - Scores for Scottish Postcode Sectors from the 1991 Census