[Construction of a deprivation index by Basic Healthcare Area in Aragon using Population and Housing Census 2011]

Rev Esp Salud Publica. 2018 Dec 10;92:e201812087.
[Article in Spanish]

Abstract

Objective: The measurement of inequalities using composite indicators facilitates the prioritization and implementation of public health actions. The most commonly source of information used for this has been the Population and Housing Census of 2011 (PCH_2011).The objective of this study was to evaluate the use of PHC_2011 and develop a deprivation index (DI) by Basic Healthcare Area (BHA) and to analyse its association with mortality in Aragon.

Methods: Ecological study by BHA. Since PHC_2011 was a sample of the population it was validated by the Chi-square test for homogeneity. 26 socioeconomic indicators were calculated. Spearman correlation coefficients were used to evaluate the relationship between socioeconomic indicators and Standardized Mortality Ratios (SMR). Principal Component Analyses (PCA) were conducted using the indicators in which a significant correlation was found. Components with eigenvalues higher than 1 were extracted, and the rotated matrix (Varimax) was obtained. PCA from each component were conducted, extracting only one factor. BHA were grouped into, according to the deprivation index values. Mortality rates adjusted to the European Standard Population by age, sex and quartile were calculated. The most discriminant factor by quartiles was considered DI. A different DI for urban areas was obtained from the same variables.

Results: The validation of PHC sample detected 4 underrepresented BHA. 17 socioeconomic indicators were significatively correlated with SMR. From the first PCA, 3 components were obtained. The DI included %unemployment, %eventual workers, % insufficient education 16-64 years old and %foreigners. The % of variance explained by the DI was 59.7% and 73.8% in urban areas. In men, mortality in the quartile with the lowest deprivation (544,7 per 105; CI95%: 515,8-573,6) was significatively lower than in the most deprivated areas(618,7 per 105;CI95%:589,4-648,0).

Conclusions: This new DI allows us to identify deprived BHA. This is a useful tool to bring to light health inequalities and to plan interventions according to population´s needs.

Objetivo: La medición de las desigualdades mediante indicadores compuestos facilita la priorización y puesta en marcha de acciones de salud pública. La fuente de información más comúnmente utilizada para ello ha sido el Censo de Población y viviendas de 2011 (CPV_2011). El objetivo fue validar la utilización del CPV_2011 por Zona de Salud (ZBS) y construir un índice de privación (IP) por ZBS así como analizar su asociación con la mortalidad en Aragón.

Metodos: Estudio ecológico por ZBS. El CPV_2011, con diseño muestral, se validó mediante un test de homogeneidad de Chi_cuadrado y se calcularon 26 indicadores socioeconómicos. Se obtuvo el coeficiente de correlación de Spearman entre indicadores socioeconómicos y Razones de Mortalidad Estandarizadas (REM). Se realizó un análisis de componentes principales (ACP) con los indicadores correlacionados significativamente, extrayendo los componentes con autovalores mayores a 1 y se obtuvo la matriz rotada (Varimax). Se realizaron ACP con las variables de cada componente extrayendo un único factor. Se agruparon las ZBS en cuartiles, según el factor calculando tasas de mortalidad ajustadas a población estándar europea por edad, sexo y cuartil. El factor que más discrimina por cuartiles se consideró IP y se recalculó para ZBS urbanas con idénticas variables.

Resultados: La validación de la muestra del CPV_2011, detectó cuatro ZBS infrarrepresentadas. 17 indicadores socioeconómicos se correlacionaron con REM. Del primer ACP se extrajeron 3 componentes, eligiendo como IP, el formado por %Desempleo, %Asalariados eventuales, %Instrucción Insuficiente 16-64 años y %Extranjeros. Las varianzas explicadas fueron 59,7% y 73,8% en el IP urbano. En hombres, la mortalidad en el cuartil menos privado (544,7 por 105; IC95%:515,8-573,6), fue inferior a la del más privado (618,7 por 105; IC95%:589,4,648,0).

Conclusiones: El IP permite identificar ZBS desfavorecidas constituyendo una herramienta para evidenciar desigualdades y planificar intervenciones según necesidades.

Keywords: Census; Factor analysis; Health inequalities; Socieconomic factors; Spain; Statistical.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Censuses
  • Chi-Square Distribution
  • Cross-Sectional Studies
  • Female
  • Health Status Disparities*
  • Housing
  • Humans
  • Male
  • Middle Aged
  • Mortality
  • Poverty Areas
  • Socioeconomic Factors*
  • Spain / epidemiology
  • Unemployment / statistics & numerical data
  • Young Adult