The aim of this study was to estimate the series of tobacco smoking prevalence year-by-year in Spain, by sex and age group, for the period 1991–2020. Based on smoking prevalence obtained from national surveys and smoking-related auxiliary information from public statistics, we fitted a multinomial logistic mixed model with random area and time effects. Joinpoint regression was used to identify changes in the prevalence series across the period. To analyse the precision of the model-based estimates, we calculated the coefficients of variation. Between 1991 and 2020, the prevalence of smoking in Spain decreased in both sexes. In the 15–24 age group, the prevalence of smokers showed no differences by sex until 2007, after which prevalence in men exceeded that of women. However, in women aged 55 and over prevalence of smoking has been rising since 1991. After applying the model, the precision of smoking prevalence estimates improved. The reconstruction of a detailed series of tobacco smoking prevalence provides insight into the evolution of the tobacco epidemic in Spain. A detailed analysis by sex and age shows different trends in the prevalence of smoking among women that should be considered when control measures are formulated.
Introduction: Small-area estimation methods are an alternative to direct surveybased estimates in cases where a survey’s sample size does not suffice to ensure representativeness. Nevertheless, the information yielded by small-area estimation methods must be validated. The objective of this...
12 October 2022 in Scandinavian Journal of Statistics
This paper introduces empirical best predictors of small area bivariate parameters, like ratios of sums or sums of ratios, by assuming that the target unit‐level vector follows a bivariate nested error regression model. The corresponding means squared errors are estimated by parametric bootstrap. Several simulation experiments empirically study the behavior of the introduced statistical methodology. An application to real data from the Spanish household budget survey gives estimators of ratios of food household expenditures by provinces.
Under a unit-level bivariate linear mixed model, this paper introduces small area predictors of expenditure means and ratios, and derives approximations and estimators of the corresponding mean squared errors. For the considered model, the REML estimation method is implemented. Several simulation experiments, designed to analyze the behavior of the introduced fitting algorithm, predictors and mean squared error estimators, are carried out. An application to real data from the Spanish household budget survey illustrates the behavior of the proposed statistical methodology. The target is the estimation of means of food and non-food household annual expenditures and of ratios of food household expenditures by Spanish provinces.
Indoor radon is a human lung carcinogen. The objective of this research is to describe residential radon distribution in a European region with high geogenic radon potential using cross‐sectional data. This study is also aimed at describing the use of census tracts as a geographical unit in radon maps and describing radon distribution according to floor level. To do this, 6080 radon measurements were carried out by the Galician Radon Laboratory (GRL, https://www.radon.gal ) with alpha track detectors located at each dwelling for at least 3 months. The dwellings measured were the primary residences of volunteer participants. All participants filled out a questionnaire on the dwelling characteristics, and a follow‐up call assured the appropriate return of radon devices to the GRL. A descriptive analysis was performed, including radon concentration distribution per floor level measured and a Galician map of indoor radon using both municipalities and census tracts as geographic units was made. In Galicia, a median radon concentration of 132 Bq/m 3 (interquartile range: 66–239 Bq/m 3 ) and a geometric mean (GM) of 126 Bq/m 3 were observed. Eighteen percent of dwellings exceeded 300 Bq/m 3 . Radon levels vary widely across and within municipalities. The latter can be observed with census tract‐level maps. A height gradient for radon concentration was observed from a GM of 289 Bq/m 3 in cellars to a GM of 89 Bq/m 3 on the fifth floor and above. Remarkably, 14% of dwellings on the second floor exceeded 300 Bq/m 3 , thus challenging the paradigm that indoor radon is only a priority for underground and ground levels in radon priority areas. Galicia is mostly composed of radon priority areas, with high heterogeneity in indoor radon distribution. Census tracts can be used in indoor radon maps as the smallest geographic unit to increase resolution and acknowledge variations within municipalities. In radon priority areas, radon surveillance should include second‐floor levels.
Background: Europe has had a large variability in COVID-19 incidence between and within countries, particularly after June 2020. We aim to assess the variability between European countries and regions located in a given country.Methods: We used ECDC information including countries having 7 regions or more. The metric used to assess the regional variability within a country was the intercuartilic range in a weekly basis for 32 weeks between June 29th 2020 and February 1st 2021. We also calculated each country's overall variability across the 32 weeks using the distances from the regional curves of the 14-day incidence rates to the corresponding national curve, using the L2 metric for functional data. We afterwards standardised this metric to a scale from 0 to 100 points. We repeated the calculations excluding island regions.Results: The variability between and within countries was large. Slovenia, Spain and Portugal have the greatest variability. Spain and Slovenia held also the top three places for the greatest number of weeks (Spain for 19 weeks and Slovenia for 10) with the highest variability. For variability among the incidence curves across the 32-week period, Slovenia, Portugal and Spain ranked first in functional variability, when all the regions were analysed but also when the island regions were excluded.Conclusions: These differences might be due to how countries tackled the epidemiological situation. The persistent variability in COVID-19 incidence between regions of a given country suggests that governmental action may have an important role in applying epidemiological control measures.