Article Text
PDF
Abstract
Objectives HIV diagnoses among men who have sex with men (MSM) in several high-income countries, including Australia, have increased substantially over recent years. Australia, in line with global prevention strategies, has emphasised a ‘test and treat’ HIV prevention strategy which relies on timely detection of HIV through frequent testing by those at risk. We examined trends in repeat testing among MSM defined as ‘high-risk’ according to Australian testing guidelines.
Methods HIV test records from MSM attending high caseload clinics in Melbourne 2007–2013 and classified as high-risk were analysed. Binary outcomes of ‘test within 3 months’ and ‘test within 6 months’ were assigned to tests within individuals’ panel of records. Negative binomial regressions assessed trends in overall HIV testing and returning within 3 and 6 months. Annualised proportions of return tests (2007–2012) were compared using two-sample z tests.
Results Across 18 538 tests among 7117 high-risk MSM attending primary care clinics in Melbourne (2007–2013), the number of annual HIV tests increased (p<0.01). Between 2007 and 2012 annualised proportions of tests with a subsequent test within 3 and 6 months also increased (p<0.01); however, by 2012 only 36.4% and 15.1% of tests were followed by another test inside 6 and 3 months, respectively.
Conclusions Repeat testing among high-risk MSM in Australia remains unacceptably low, with recent modest increases in testing unlikely to deliver meaningful prevention impact. Removing known barriers to HIV testing is needed to maximise the potential benefit of test and treat-based HIV prevention.
- HIV TESTING
- MEN
- PREVENTION
- SEXUAL BEHAVIOUR
Statistics from Altmetric.com
Introduction
In several high-income countries, HIV cases among men who have sex with men (MSM) have increased, while HIV transmission among other populations has declined.1 Alongside a range of complex drivers of HIV transmission (eg, transmission probabilities for specific risk behaviours, sexual role versatility, sexual networks, other sexually transmitted infections and adoption of non-condom risk reduction),1 community-level factors such as increasing HIV prevalence and suboptimal testing, diagnosis and treatment coverage2 further mediate risk. HIV prevention in Australia, in line with global prevention strategies, emphasises ‘test and treat’ approaches that encourage timely HIV detection, postdiagnosis behaviour change and early treatment. Australian gay community behavioural surveys consistently show high self-reported lifetime (∼90%) and annual (∼60%) HIV testing among MSM and high self-reports of antiretroviral treatment (ART; ∼77%) and viral suppression (∼74%) among MSM diagnosed with HIV.3 Australia could be considered closest to achieving the UNAIDS 90-90-90 elimination goals,4 yet HIV diagnoses in Australia continue to increase and are now at a 20-year high.5
Increasing testing frequency has the potential to significantly reduce the transmission of HIV among MSM. Understanding of the contribution of undiagnosed primary HIV infection (first 3 months) to transmission more specifically suggests that targeting high-frequency testing among high-risk individuals is likely to yield substantial epidemic impact.2 Increasing annual HIV screening to testing every 3 months among high-risk individuals has also shown to be cost-effective in modelling of the UK epidemic.6 The importance of frequent testing for individuals with high partner turnover and/or engaging in condomless sex is reflected in Australian guidelines which recommend ‘high-risk’ MSM test for HIV up to four times a year.7 In contrast to high self-reported rates, a 2010 Australian study8 provided the first objective estimates of annual and 6 month repeat testing among MSM, finding low rates of 12 (35%) and 6 month (15%) HIV testing using linked clinical test records. This study provided important insights into testing patterns among MSM, though was limited by available data and did not examine 3-month testing.8 We determined repeat (three and six monthly) HIV testing among MSM classified as high-risk according to Australian guidelines to inform targeted HIV testing and prevention strategies.
Methods
This analysis included linked HIV tests records from all HIV-negative MSM who were aged ≥16 years and attended one metropolitan sexual health centre (SHC) and two general practice (GP) clinics specialising in gay men's health in Melbourne between 2007 and 2013. Surveillance questionnaires self-completed by patients at each presentation classified MSM as high-risk if they reported >10 male partners in 6 months or >20 in 12 months and/or recent inconsistent condom use, in accordance with Australian testing guidelines.7 Repeat tests within 30 days (n=1203) were considered the same testing episode and excluded, as were tests performed as part of post-exposure prophylaxis (n=359) as these tests were not considered part of routine testing.
Individuals’ records within-clinics were organised as panels, with ‘test within 6 months’ (182 days) and ‘test within 3 months’ (91 days) created as binary outcomes (yes/no) and assigned at every record within each panel of data. Each test record was considered an index and test within 6 months or test within 3 months was assigned ‘yes’ to an index if subsequent tests occurred within 182 or 91 days, respectively, at the same clinic.
Negative binomial regressions assessed trends in overall testing and 6 and 3 months returning for testing. A negative binomial approach was preferred over a Poisson model due to overdispersion of the testing count data. A two-sample z test assessed differences in the annual proportion of return tests between 2007 and 2012. A Wilcoxon rank-sum test for equality of medians tested for a difference in median number of tests per individual per year between 2007 and 2013. p<0.05 was considered statistically significant and analysis was conducted using Stata V.13.1 (StataCorp, College Station, Texas, USA).
Data were collected as part of the Victorian Primary Care Network for Sentinel Surveillance on sexually transmissible infections and bloodborne viruses, which is approved by six human research ethics committees including the Victorian Department of Health and Human Services (Project number 52/05) Alfred Health (Project number 213/05).
Results
Of 46 600 HIV tests among 17 852 sexually active MSM from 2007 to 2013, 18 538 tests were conducted among 7117 MSM classified as high-risk. Among high-risk MSM, the number of annual HIV tests increased each year (p<0.01); however, there was no significant change in the median number of tests per individual (p=0.17; table 1). Between 2007 and 2012, 31.5% of HIV tests among high-risk MSM were followed by a test within 6 months and 12.6% were followed by a test within 3 months at the same clinic. Annualised proportion of tests with a subsequent test within 6 and 3 months increased (p<0.01), but by 2012 still only represented 36.4% and 15.1% of tests, respectively (table 1). At GPs the overall proportion of return tests within 6 and 3 months was 27% and 9.4%, and at the SHC, 32% and 12.4%, respectively (data not shown).
Description of HIV tests and returning for testing within 3 and 6 months among ‘high-risk’ MSM, January 2007 to December 2013, n=18 358
Discussion
Despite increasing HIV test numbers over time among high-risk MSM attending the major gay and bisexual men primary healthcare clinics in Melbourne, a large majority were not returning to the same clinic for testing within 3 or 6 months as recommended by Australian guidelines.7 We found only approximately 1 in 10 high-risk MSM returned to test at the same clinic within 3 months, similar to a recent US finding (13% of MSM testing quarterly).9 Our findings provide the first estimate of 3-month repeat testing among Australian MSM and highlight a disparity between testing patterns, clinical recommendations7 and levels of testing likely to be needed to maximise the impact of HIV test and treat prevention strategies.
Interrupting HIV transmission through timely diagnosis, especially during primary infection is likely to return a significant prevention impact and underpins high-frequency testing guidelines for high-risk individuals. Modelling of the Australian epidemic has reported a disproportionate contribution of undiagnosed HIV to transmissions; ∼9% of MSM with undiagnosed HIV accounting for ∼30% of new infections and ∼3% HIV-infected MSM with primary HIV infection accounting for ∼19% of the new infections.2 Community-based bio-prevalence studies and recent modelling of surveillance data estimated 10%–30% of HIV-infected MSM remain undiagnosed in Australia,5 therefore, the relative contribution of undiagnosed infections to transmission may be higher than estimated by Wilson et al.2 Given ∼80% of high-risk MSM in this large clinic-based sample did not reattend for 3 month HIV testing, substantial gains in case detection could be made in Australia through targeted promotion and facilitation of frequent testing in this population.
New strategies are now needed to increase HIV testing and diagnosis among high-risk individuals. Australia lags behind comparable countries in providing regulatory approvals and implementing alternate models of HIV testing. Rapid point-of-care HIV testing was only recently introduced and has limited reach. The overwhelming majority of HIV testing in Australia is provided through clinic presentations and laboratory testing, a model reported by MSM as creating a range of barriers to testing.10 Given evidence suggesting high acceptability of alternate HIV testing models, including self-testing among Australian MSM11 the lack of repeat testing reported through this analysis suggests an urgent need for Australia to implement new and innovative models of testing as a mechanism to address increasing HIV diagnoses.
Some caution is required when interpreting our findings. It is unknown if MSM attending the sites in our study-high caseload primary care clinics- test more or less frequently than high-risk MSM attending low caseload clinics. HIV tests among MSM outside the clinical network or repeat tests at different sites within the network will be missed. This study does have significant strengths, including a large amount of testing denominator data from clinics which provide the majority of sexual healthcare in the state. Although a previous study provided valuable empirical estimates of testing patterns among Australian MSM,8 we have built on this study with additional data and rolling index tests (the previous estimate used calendar years to restrict cohort entry and returning for testing) providing updated 6-month estimates and the first estimates of three monthly testing rates. The data also provides objective intraclinical testing frequency data to reliably indicate testing frequency trends and is the only ongoing linked surveillance system monitoring HIV testing in Australia.
Despite high and increasing treatment coverage for people living with HIV, HIV diagnoses continue to increase in Australia. High frequency testing is central to HIV prevention and repeat testing reported here among Australia's key risk population remains unacceptably low. Implementing new testing models to reduce known structural barriers to frequent testing is critical if Australia is to take advantage of low cost and highly accessible HIV treatment and move towards HIV control.
Acknowledgments
The authors would like to gratefully acknowledge the patients, notifying medical practitioners and the laboratories for the ongoing participation and contribution to data collection. The authors acknowledge Dr Norm Roth and Dr BK Tee for their provision of clinical advice and ongoing participation of their respective clinics in the Victorian Primary Care Network for Sentinel Surveillance (VPCNSS). The VPCNSS has significant ongoing support from surveillance officers at the Burnet Institute.
Footnotes
Handling editor Jackie A Cassell
Contributors ALW completed data analysis, interpreted the results and drafted the manuscript. CE-H contributed to data management and analysis. TS contributed to data analysis. CKF, DL and MH contributed to data collection. ESM critically reviewed the manuscript. MS conceptualised the study and oversaw data analysis. All authors participated in the interpretation of the data, critically reviewed the manuscript and authors approved final manuscript.
Funding The Victorian Department of Health funds ongoing surveillance projects within the Burnet Institute. The authors would like to acknowledge the NHMRC who provide funding to MH as a Senior Research Fellow (1062877), ALW as a public health scholarship recipient (1055196) and MS (1090445) and EM (1034464) through a Career Development Fellowships. The authors gratefully acknowledge the contribution to this work of Victorian Operational Infrastructure Support Programme received by the Burnet Institute.
Competing interests None declared.
Ethics approval Six Human Research Ethics Committees including the Victorian Department of Health and Human Services approved the VPCNSS project.
Provenance and peer review Not commissioned; externally peer reviewed.