Chapter 3 84 biases from pharmaceutical interventions. Moreover, the vast majority of COVID-19 cases were linked to the municipality register, resulting in few missing data on city district and migration background. However, our study also has several drawbacks. First, as we used confirmed COVID-19 notifications, a detection bias exists, driven by access to testing facilities. Second, imputing the date of illness onset based on the distribution of time from illness onset to notification date for individuals missing this information may have led to a small proportion of asymptomatic contacts being erroneously allocated an earlier date of illness onset. However, given that this misclassification is unlikely to be differential and only applicable to a small proportion of individuals, no impact on our conclusions is expected. Third, hospitalisation status might not have been accurately recorded for all individuals, for instance if contact-tracing teams were unable to contact the case or their relatives and hospital reporting of admission was incomplete. Furthermore, some hospitalisations may have been due to reasons other than COVID-19. However, given that under- or overreporting of hospitalisation status are unlikely to be linked to city district or migrations status, any influence on the measures of effect of these factors is likely to have been minimal. Additionally, we did not have access to individual-level data on possible causal or confounding factors such as comorbidities and occupation. Further studies should focus on studying individual and community-level drivers of COVID-19 disparities. CONCLUSIONS Despite the initiation of public health programs to reduce disparities in COVID-19 burden, the risk of SARS-CoV-2 infection, hospitalisation and deaths continued to be higher in peripheral city districts with a lower SES as well as in individuals with a non-European migration background during the second wave. Socio-economic, cultural and medical factors may account for a large part of these differences. Prospective observational studies are required to evaluate existing strategies and successfully identify and serve groups at increased risk of SARS-CoV-2 infection, severe outcomes and post-acute COVID-19 syndrome as we move forward from the pandemic.
RkJQdWJsaXNoZXIy MTk4NDMw