Abstract Objective To examine temporal changes in late diagnosis of HIV (LD) among migrant and non‐migrant people with HIV in Sweden 2003–2023 and to assess demographic and socioeconomic risk factors for LD in these two populations. Methods People with HIV diagnosed with HIV‐1 in Sweden 2003–2023 were included ( n = 6278). LD was defined as a first CD4+ T‐cell count <350 cells/μL or an AIDS‐defining event within 3 months of diagnosis. People with HIV with evidence of recent infection were reclassified as non‐late. Temporal changes in LD were examined using descriptive statistics and regression analyses. To assess risk factors for LD, modified Poisson regression was employed. Risk factor analyses were restricted to 2010–2020 when complete sociodemographic data were available ( n = 2778). Data were obtained from Swedish national registries. Results The absolute incidence of total and late HIV diagnoses decreased over the study period, whereas the annual proportion of LD varied between 46% and 60% and trended upwards. LD occurred in 41% of non‐migrant people with HIV and 58% of migrant people with HIV. Among non‐migrant people with HIV, having an upper secondary education or less was associated with LD compared to post‐secondary education, as was male sex with heterosexual HIV acquisition and higher age. For migrant people with HIV, neither lower education nor income was statistically significantly associated with LD. Instead, higher age, certain birth regions, heterosexual acquisition and male sex with acquisition through injection drug use were associated with LD. Conclusions LD declined in absolute terms yet constituted a high and increasing proportion of new HIV cases in Sweden 2003–2023, with differing sociodemographic determinants by migrant status.
Falk et al. (Thu,) studied this question.