Spatial variation in tobacco smoking among pregnant women in South Limburg, the Netherlands, 2016–2018: Small area estimations using a Bayesian approach. (August 2022)

• Record Type: Journal Article
• Authors: Wang, Haoyi; Smits, Luc; Putrik, Polina
• Is Part Of: Spatial and spatio-temporal epidemiology. Volume 42(2022)
• Publication Date: 2022-08
• Subjects: Bayesian inference -- Spatial analysis -- Tobacco smoking -- Pregnancy -- Small area estimation; Epidemiology -- Statistical methods -- Periodicals; Epidemiology -- Periodicals; 614.4072
• DOI: 10.1016/j.sste.2022.100525
• Languages: English
• ISSNs: 1877-5845

Abstract

Highlights: Maternal tobacco smoking is heterogenous in South Limburg, the Netherlands. Bayesian spatial analysis provided robust estimations over the frequentist approach. Areal SES proxies can impact the spatial distribution of maternal tobacco smoking. Refining the geographical scale can lead to enhanced insights to support local prevention. Abstract: The aim of this study was to provide small area estimations (SAE) of smoking prevalence during pregnancy in South Limburg, the Netherlands. To illustrate improvements in accuracy and precision of estimates compared to traditional frequentist analyses, we used Bayesian inference with the Integrated nested Laplace approximation to account for spatial structures and area-level proxies. Results revealed a heterogenous prevalence of smoking with a range between 6.7% (95% credible interval 4.7, 8.7) and 16.7% (14.3, 19.2) among municipalities; and an even more heterogenous prevalence among neighbourhoods a range from 0 (-14.9, 6.5) to 32.1 (20.3, 46.8). Clusters with significant lower- and higher-than-average risk were identified (RR between 0.6-1.4 and 0.0-2.4 for municipality- and neighbourhood-level, respectively). Higher proportion of non-western migrants and lower average income were associated with higher prevalence of tobacco smoking. The obtained estimates should inform local prevention policies, as well as provide methodological example for public health researchers on application of Bayesian methods for SAE.