A comparison of small-area deprivation indicators for public-health surveillance in Sweden. Issue 4 (June 2023)

• Record Type: Journal Article
• Authors: Strömberg, Ulf; Baigi, Amir; Holmén, Anders; Parkes, Brandon L.; Bonander, Carl; Piel, Frédéric B.
• Is Part Of: Scandinavian journal of public health. Volume 51:Issue 4(2023)
• Page Start: 520
• Page End: 526
• Publication Date: 2023-06
• Subjects: Factor analysis -- mortality -- spatial smoothing -- small-area characteristics -- sociodemographic factors -- Sweden; Public health -- Periodicals; Social medicine -- Periodicals; 614
• DOI: 10.1177/14034948211030353
• Languages: English
• ISSNs: 1403-4948

Abstract

Aims: The aims of this study were to construct a small-area index of multiple deprivation (IMD) from single deprivation indicators (SDIs) and to compare the explanatory power of the IMD and SDIs with regard to mortality. We considered a small-area division of Sweden consisting of 5985 DeSO ( Demografiska statistikområden ), each with a population size between 653 and 4243 at the end of 2018. Methods: Four SDIs were provided by open-source data: (a) the proportion of inhabitants with a low economic standard; (b) the proportion of inhabitants aged 25–64 years with ⩽12 years of schooling; (c) the proportion of inhabitants aged 16–64 years who were not in paid employment; and (d) the proportion of inhabitants who lived in a rented apartment/house. A four-indicator IMD was constructed using factor analysis. As a validation, the IMD and SDIs were compared by exploring their DeSO-level associations with spatially smoothed death rates, with robustness checks of associations across different small-area contexts defined by degree of urbanisation and distribution of immigrants from non-Western countries. Results: The constructed IMD and SDI1 performed essentially equally and outperformed SDI2, SDI3 and SDI4. Associations between IMD/SDI1 and the spatially smoothed death rates were most pronounced within the age range 60–79 years, showing 5–8% lowered rates among those categorised in the least deprived quintiles of IMD and SDI1, respectively, and 7–9% elevated rates among those categorised in the most deprived quintiles. These associations were consistent within each small-area context. Conclusions: We suggest prioritisation of SDI1, that is, a DeSO-level deprivation indicator based on open-access data on economic standard, for public-health surveillance in Sweden.