The public health utility of assays to test for recent HIV infection: an evaluation on UK case-based surveillance data
File(s)
Author(s)
Aghaizu, Adamma Rose-Linda
Type
Thesis or dissertation
Abstract
Determining accurate, real-time epidemic trends for HIV is an ongoing challenge due to the lengthy asymptomatic period of infection. Current available methods to determine the number of new infections are based on back calculation models of diagnosis data and/or simulation models of behavioural data. Both approaches do not provide timely estimates for recent years or estimates for risk groups other than gay, bisexual and men who have sex with men (MSM), for whom there are less published data on risk behaviours. The aim of this thesis is to explore the public health utility of serological HIV incidence assays applied to case-based surveillance data in the UK.
For the first five years of Public Health England’s surveillance programme, I determined demographic predictors for a recent infection diagnosis and estimated HIV incidence in both sexual health clinic attendees and the general population. I also undertook a feasibility study for enhanced behavioural surveillance among MSM with incident infection to explore if this could highlight new trends in risk behaviours or if more traditional infectious disease control methods, such as active case finding, could become more applicable to HIV.
Between 2009 and 2013, I found predictors for a recent infection diagnosis to have been younger age (15-24 years compared to + 50 years) (adjusted odd ratio (AOR) 2.8 95% C.I 2.2-3.7), the UK as probable country of infection (AOR 1.4 95% C.I 1.2-1.6) and higher CD4 counts (>1000 cells/mm3 compared to >50≤200 cells/mm3, AOR: 14.3, 95% C.I. 8.9-22.8) in MSM, and UK country of birth (AOR: 1.7, 95% C.I. 1.2-2.3) and UK country of infection (AOR: 1.4 95% C.I. 1.1-1.8) in heterosexuals. HIV incidence was up to 30-fold higher in sexual health clinic attendees (130 per 100,000 person years (pys) in 2009 increasing to 200 per 100,000 pys in 2013) compared to the general population (between 6 and 6.5 per 100,000 over the years), with little change over the period. The two key populations most affected were MSM, with approximately 300 infections per 100,000 pys, and black African heterosexuals, with between 45 and 70 infections per 100,000 pys. The number of new HIV infections was five-fold higher in London compared to outside London. The behavioural surveillance data showed that nearly all men had exhibited high risk behaviours in the six months before diagnosis; half had had a sexually transmitted infection (STI) in the previous year. Men had met partners mainly via mobile phone dating apps. Despite two thirds of sexual partners having been contactable, only one in five had been contacted with men indicating preference to notify partners themselves.
Findings from this thesis show serological HIV incidence assays applied to case-based surveillance data in the UK can produce timely estimates of HIV incidence for the whole population. It is currently the only method allowing comparisons by geography which may enable prevention resources to be targeted more effectively. In light of the ongoing decline in new HIV diagnoses and likely transmission, and the roll out of a new biomedical intervention (pre-exposure prophylaxis (PrEP)), all sources of HIV epidemic intelligence will be crucial to work towards the elimination of HIV.
Whilst the enhanced behavioural surveillance was feasible in this group, it is unlikely to discover new risk behaviours or facilitate active case finding. However, there is a role for this approach of data collection among recent seroconverters; the surveillance scheme, now referred to as SHARE (Surveillance of HIV Acquired Recently: Enhanced), has been modified and rolled out on a national scale to obtain insights into how new infections may or may not relate to exposure of PrEP in light of the ongoing PrEP trial (https://www.prepimpacttrial.org.uk/). Findings of this new initiative will feed into future evaluations of PrEP use in the UK.
For the first five years of Public Health England’s surveillance programme, I determined demographic predictors for a recent infection diagnosis and estimated HIV incidence in both sexual health clinic attendees and the general population. I also undertook a feasibility study for enhanced behavioural surveillance among MSM with incident infection to explore if this could highlight new trends in risk behaviours or if more traditional infectious disease control methods, such as active case finding, could become more applicable to HIV.
Between 2009 and 2013, I found predictors for a recent infection diagnosis to have been younger age (15-24 years compared to + 50 years) (adjusted odd ratio (AOR) 2.8 95% C.I 2.2-3.7), the UK as probable country of infection (AOR 1.4 95% C.I 1.2-1.6) and higher CD4 counts (>1000 cells/mm3 compared to >50≤200 cells/mm3, AOR: 14.3, 95% C.I. 8.9-22.8) in MSM, and UK country of birth (AOR: 1.7, 95% C.I. 1.2-2.3) and UK country of infection (AOR: 1.4 95% C.I. 1.1-1.8) in heterosexuals. HIV incidence was up to 30-fold higher in sexual health clinic attendees (130 per 100,000 person years (pys) in 2009 increasing to 200 per 100,000 pys in 2013) compared to the general population (between 6 and 6.5 per 100,000 over the years), with little change over the period. The two key populations most affected were MSM, with approximately 300 infections per 100,000 pys, and black African heterosexuals, with between 45 and 70 infections per 100,000 pys. The number of new HIV infections was five-fold higher in London compared to outside London. The behavioural surveillance data showed that nearly all men had exhibited high risk behaviours in the six months before diagnosis; half had had a sexually transmitted infection (STI) in the previous year. Men had met partners mainly via mobile phone dating apps. Despite two thirds of sexual partners having been contactable, only one in five had been contacted with men indicating preference to notify partners themselves.
Findings from this thesis show serological HIV incidence assays applied to case-based surveillance data in the UK can produce timely estimates of HIV incidence for the whole population. It is currently the only method allowing comparisons by geography which may enable prevention resources to be targeted more effectively. In light of the ongoing decline in new HIV diagnoses and likely transmission, and the roll out of a new biomedical intervention (pre-exposure prophylaxis (PrEP)), all sources of HIV epidemic intelligence will be crucial to work towards the elimination of HIV.
Whilst the enhanced behavioural surveillance was feasible in this group, it is unlikely to discover new risk behaviours or facilitate active case finding. However, there is a role for this approach of data collection among recent seroconverters; the surveillance scheme, now referred to as SHARE (Surveillance of HIV Acquired Recently: Enhanced), has been modified and rolled out on a national scale to obtain insights into how new infections may or may not relate to exposure of PrEP in light of the ongoing PrEP trial (https://www.prepimpacttrial.org.uk/). Findings of this new initiative will feed into future evaluations of PrEP use in the UK.
Version
Open Access
Date Issued
2018-11
Online Publication Date
2022-05-23T11:00:26Z
Date Awarded
2022-04
Copyright Statement
Creative Commons Attribution NonCommercial Licence
Advisor
Ward, Helen
Delpech, Valerie
Sponsor
Public Health England
Publisher Department
School of Public Health
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)