How important is it to avoid indices of deprivation that include health variables in analyses of health inequalities?

OBJECTIVES: This study aimed to quantify the difference in mortality inequalities using the Scottish Index of Multiple Deprivation (SIMD) and the Income and Employment Index (IEI; a subindex of SIMD, which excludes health) as ranking measures in Scotland. STUDY DESIGN: This ecological study was a cross-sectional analysis of routine administrative data. METHODS: Data from the 2020 SIMD and the subindex using data from only the Income and Employment domains, the IEI, were obtained. The correlation between data zones, percentage of data zones that changed deprivation tenth and differences in the Slope Index of Inequality (SII) and Relative Index of Inequality (RII) for Standardised Mortality Ratios (SMRs) across tenths were compared when data zones were ranked by SIMD and IEI. RESULTS: There was a close correlation between data zones ranked by SIMD and IEI (R2 = 0.96). When data zones were ranked by IEI, 18.7% of data zones moved to a lower deprivation tenth, and 20.8% of data zones moved to a higher deprivation tenth, compared with SIMD. However, only a negligible number of data zones moved two or more tenths. The SMRs across deprivation tenths were very similar between the SIMD and IEI, as were the summary health inequality measures of SII (87.3 compared with 85.7) and RII (0.88 and 0.86). CONCLUSION: Although there is a logical problem in using deprivation indices that include health outcomes to rank areas to calculate the scale of health inequalities, the impact of using an alternative subindex containing only data from the income and employment domains is minimal. For population-wide analyses of health inequalities in Scotland, the SIMD does not introduce a substantial bias in the health inequalities summary measures despite substantial movement of small areas between ranked population tenths. Although not examined here, this is likely to be relevant to other similar indices across the United Kingdom.

How well do area-based deprivation indices identify income- and employment-deprived individuals across Great Britain today?

BACKGROUND: Area-based deprivation indices are used in many countries to target interventions and policies to populations with the greatest needs. Analyses of the Carstairs deprivation index applied to postcode sectors in 2001 identified that less than half of all deprived individuals lived in the most deprived areas. OBJECTIVE: This article examines the specificity and sensitivity of deprivation indices across Great Britain in identifying individuals claiming income- and employment-related social security benefits. STUDY DESIGN: This was a descriptive analysis of cross-sectional administrative data. METHODS: The data sets for the 2020 Scottish Index of Multiple Deprivation, Scottish Income and Employment Index, the 2019 English Index of Multiple Deprivation and the 2019 Welsh Index of Multiple Deprivation were obtained. For each data set, small areas were ranked by increasing overall deprivation, and the cumulative proportions of individuals who were income and employment deprived were calculated. Receiver operating characteristic curves were plotted to show the sensitivity and specificity of each index, and the percentages of income- and employment-deprived individuals captured at different overall deprivation thresholds were calculated. RESULTS: Across all indices, the sensitivity and specificity for detecting income- and employment-deprived individuals were low, with less than half living in the most deprived 20% of areas. Between 55% and 62% of income-deprived people and between 56% and 63% of employment-deprived people were missed across the indices at the 20% deprivation threshold. The sensitivity and specificity were slightly higher for income deprivation than employment deprivation across indices and slightly higher for the Scottish Index of Multiple Deprivation and Scottish Income and Employment Index than for the English Index of Multiple Deprivation and Welsh Index of Multiple Deprivation. CONCLUSION: Area-based deprivation measures in Great Britain have limited sensitivity and specificity for identifying individuals who are income or employment deprived. Place-based policies and interventions are unlikely to be effective at reducing inequalities as a result. Creation of individually linked data sets and interventions that recognise the social and economic relationships between social groups are likely to be more effective.

How well does the Scottish Index of Multiple Deprivation identify income and employment deprived individuals across the urban-rural spectrum and between local authorities?

BACKGROUND: Area-based indices of deprivation are used to identify populations at need, to inform service planning and policy, to rank populations for monitoring trends in inequalities, and to evaluate the impacts of interventions. There is scepticism of the utility of area deprivation indices in rural areas because of the spatial heterogeneity of their populations. OBJECTIVE: To compare the sensitivity of the Scottish Index of Multiple Deprivation (SIMD) for detecting income and employment deprived individuals by urban-rural classification and across local authorities. STUDY DESIGN: Descriptive analysis of cross-sectional data. METHODS: Data from the 2020 Scottish Index of Multiple Deprivation (SIMD) were used to calculate the number and percentage of income and employment deprived people missed within each of the six-fold urban-rural classification strata and each local authority using areas ranked by the national SIMD, within local authority rankings, and within urban-rural strata rankings, for deprivation thresholds between the 5% most deprived areas and the 30% most deprived areas. The Slope Index of Inequality (SII) and Relative Index of Inequality (RII) were calculated within local authorities and urban-rural classification strata to estimate the concentration of deprivation within ranked data zones. RESULTS: The number and percentage of income and employment deprived people is higher in urban than rural areas. However, using the national, local authority, and within urban-rural classification strata rankings of SIMD, and under all deprivation thresholds (from the 5%-30% most deprived areas), the percentage of income and employment deprived people missed by targeting the most deprived areas within urban-rural strata is higher in more remote and rural areas, and in island local authorities. The absolute number of income and employment deprived individuals is greater in urban areas across rankings and thresholds. CONCLUSION: The SIMD misses a higher percentage of income and employment deprived people in remote, rural and island areas across deprivation thresholds and irrespective of whether national, local or within urban-rural classification strata are used. However, the absolute number of people missed is higher in urban areas.