COVID-19 burden differed by city district and ethnicity during the pre-vaccination era in Amsterdam, the Netherlands 3 83 Netherlands found that whilst migration background was associated with an increased risk of COVID-19 hospitalisation, 21-day mortality and ICU admission rates were comparable between different ethnic groups once hospitalised(28). These findings suggest relative rates of hospitalisation and mortality should be similar between ethnic groups. In our study, however, we observed greater differences in hospitalisation rates compared to mortality rates. This adds further evidence to suggest that COVID-19 mortality in the community may be underreported in those with a non-European migration background dying at home without having been tested, as opposed to ethnic-Dutch older adults residing in nursing homes. When restricting to individuals aged younger than 60 years, however, we found the relative risk for COVID-19 hospitalisation among non-European individuals to be no different than in the primary analysis (RR 4.86, 95%CI=4.59-5.12 < 60 years; compared to RR 4.51, 95%CI=4.37-4.65 overall). The interplay between predisposing social, economic, cultural and medical factors for COVID-19 is therefore complex. To alleviate COVID-19 burden, targeted information and testing campaigns as well as long-term, parallel investments to reduce both socioeconomic deprivation and non-communicable disease prevalence are crucial. Following the first wave of COVID-19, several targeted prevention programs were undertaken by the Public Health Service of Amsterdam(36). Firstly, in partnership with existing research projects, focus groups were conducted in October 2020 with numerous key non-European migrant groups in order to evaluate the key barriers to adhering to COVID-19 regulations, thus increasing exposure to and onward transmission of the virus. Secondly, a Corona Prevention Team was set up in order to improve inclusivity of COVID-19 public health communication tools, and initiate specific interventions such as mobile testing services in high-incidence city districts(37) given that access to testing services has been previously reported to be lower among ethnic minority groups(38, 39). It is remarkable that despite these programs, differences in risk of hospitalisation between groups appear to have increased in the current analysis compared to our previous analysis. This suggests that these programs might not have been completely effective in preventing further spread and infection among ethnic groups, although we cannot be sure if the differences observed may have been even greater without these interventions. Further efforts therefore need to be made to understand the drivers of and identify effective interventions to combat the disproportionate burden of COVID-19 in Amsterdam. This will allow policy-makers to invest in ways to prevent these inequalities in a future outbreak or pandemic. Our study’s strengths are the use of notification data in the period prior to the availability of vaccinations and self-tests, thus allowing for analysis of COVID-19 burden without
RkJQdWJsaXNoZXIy MTk4NDMw