-
Challenges in modelling the proportion of undiagnosed HIV infections in Sweden
-
View Affiliations Hide AffiliationsEmmi Anderssonemmi.andersson ki.se
-
View Citation Hide Citation
Citation style for this article: . Challenges in modelling the proportion of undiagnosed HIV infections in Sweden. Euro Surveill. 2019;24(14):pii=1800203. https://doi.org/10.2807/1560-7917.ES.2019.24.14.1800203 Received: 18 Apr 2018; Accepted: 28 Nov 2018
Abstract
Sweden has a low HIV prevalence. However, among new HIV diagnoses in 2016, the proportion of late presenters and migrants was high (59% and 81%, respectively). This poses challenges in estimating the proportion of undiagnosed persons living with HIV (PLHIV).
To estimate the proportion of undiagnosed PLHIV in Sweden comparing two models with different demands on data availability and modelling expertise.
An individual-based stochastic simulation model of HIV positive populations (SSOPHIE) and the incidence method of the European Centre for Disease Prevention and Control (ECDC) HIV Modelling Tool were applied to clinical, surveillance and migration data from Sweden 1980–2016.
SSOPHIE estimated that the proportion of undiagnosed PLHIV in 2013 was 26% (n = 2,100; 90% plausibility range (PR): 900–5,000) for all PLHIV, 17% (n = 600; 90% PR: 100–2,000) for men who have sex with men (MSM), 35% in male (n = 300; 90% PR: 200–700) and 34% in female (n = 400; 90% PR: 200–800) migrants from sub-Saharan Africa (SSA). The estimates for the ECDC model in 2013 were 21% (n = 2,013; 95% confidence interval (CI): 1,831–2,189) for all PLHIV, 15% (n = 369; 95% CI: 299–434) for MSM and 21% (n = 530; 95% CI: 436–632) for migrants from SSA.
The proportion of undiagnosed PLHIV in Sweden is uncertain. SSOPHIE estimates had wide PR. The ECDC model estimates were unreliable because migration was not accounted for. Better migration data and estimation methods are required to obtain reliable estimates of proportions of undiagnosed PLHIV in similar settings.
Full text loading...