A spatiotemporal analysis of socioeconomic and demographic inequality in COVID-19 infections and hospitalisations in the Netherlands

Abstract Background The COVID-19 pandemic particularly affected already-vulnerable social groups. We quantified area-level socioeconomic and demographic inequalities in SARS-CoV-2 infections and COVID-19 hospitalisations across 3,228 Dutch neighbourhoods during 2020–2021. Methods We developed and fitted a hierarchical Bayesian spatiotemporal model incorporating spatial proximity, human mobility and temporal dependence to estimate time-varying associations between neighbourhood socioeconomic status (SES) and migration background and the incidence of notified SARS-CoV-2 infections and COVID-19 hospitalisations. Results Higher shares of low-SES residents and residents with a migration background are associated with increased risk for infection (ORs up to 1.21 [95%CrI: 1.19-1.22] and 1.62 [1.57-1.66]) and hospitalisation (ORs up to 1.34 [1.24-1.45] and 1.68 [1.55-1.82]). Hospitalisation-risk is consistently elevated in low-SES and high-migration neighbourhoods across pandemic phases while infection risk varies in both magnitude and direction across phases (ORs ranging from 0.83 [0.82-0.85] to 1.54 [1.49-1.59]), reflecting differential testing behaviour rather than true differences in exposure. Conclusions Disadvantaged neighbourhoods are disproportionately affected throughout the pandemic, with inequalities varying over space and time. The divergence between infection and hospitalisation gradients underscores that notified infection data incompletely captures true disease burden in socio-demographically disadvantaged groups. Integrating fine-scale socioeconomic indicators into routine surveillance can support more equitable, targeted outbreak responses.