Improving mortality rate estimation in 825 European cities: a linear mixed-effects modelling approach using Urban Audit data

Résumé fourni par la source

Assessing the health of urban populations and its determinants is crucial to ensure sustainable urban living. Burden of disease studies and health impact assessments are often used and city-level baseline mortality rates are a key parameter. However, studies tend to use unrepresentative estimates, with regional or national estimates typically applied at city-level, particularly for large-scale studies. We developed a standardized method for age- and sex-specific mortality rate estimation for 825 European cities, using open-source data. To fill for missing data, linear mixed-effects models with natural cubic splines modelled mortality and population data for 2011-19, based on city-specific trends. By generating a publicly available dataset of city-level mortality rates, and a replicable approach for other years, we sought to overcome many of the existing shortcomings in baseline mortality estimates. The highest age-standardized natural-cause mortality rates in 2018 were observed in cities in Hungary, Poland, Croatia, and Czech Republic while the lowest rates were in cities in Spain, France, and Italy. Variation between city-level and national rates was observed, with a median relative deviation of ∼16%. Our method addresses many of the common data quality and harmonization challenges associated with city-level mortality rate estimation. Advancement of effective policies and interventions necessitates robust baseline health data across cities, with good temporal coverage. To achieve this, requires strengthened efforts to improve data accuracy and consistency from regional sources to ensure accurate city-level representation. We fully documented the process and provide the code and data to be used in other studies.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.