Advertisement
Short Communication| Volume 66, P131-134, January 2018

Estimating the annual risk of HIV transmission within HIV sero-discordant couples in sub-Saharan Africa

  • Susanne F. Awad
    Affiliations
    Infectious Disease Epidemiology Group, Weill Cornell Medicine-Qatar, Cornell University, Qatar Foundation – Education City, Doha, Qatar

    Population Health Research Institute, St George’s, University of London, London, UK
    Search for articles by this author
  • Hiam Chemaitelly
    Affiliations
    Infectious Disease Epidemiology Group, Weill Cornell Medicine-Qatar, Cornell University, Qatar Foundation – Education City, Doha, Qatar
    Search for articles by this author
  • Laith J. Abu-Raddad
    Correspondence
    Corresponding author at: Infectious Disease Epidemiology Group, Weill Cornell Medicine–Qatar, Qatar Foundation – Education City, P.O. Box 24144, Doha, Qatar.
    Affiliations
    Infectious Disease Epidemiology Group, Weill Cornell Medicine-Qatar, Cornell University, Qatar Foundation – Education City, Doha, Qatar

    Department of Healthcare Policy and Research, Weill Cornell Medicine, Cornell University, New York, NY, USA
    Search for articles by this author
Open AccessPublished:November 09, 2017DOI:https://doi.org/10.1016/j.ijid.2017.10.022

      Highlights

      • The annual risk of HIV transmission within sero-discordant couples varied across 23 African countries and hovered around 10% per person-year.
      • HIV prevalence was strongly associated with HIV transmission risk within sero-discordant couples, and explained most of its variation.
      • Heterogeneity in HIV infectiousness may explain, partly, the differences in the scales of HIV epidemics in sub-Saharan Africa.

      Abstract

      Objective

      To estimate the annual risk of HIV transmission (ϕ) within HIV sero-discordant couples in 23 countries in sub-Saharan Africa (SSA), by utilizing newly available national population-based data and accounting for factors known to potentially affect this estimation.

      Methods

      We used a recently developed pair-based mathematical model that accommodates for HIV-dynamics temporal variation, sexual risk-behavior heterogeneity, and antiretroviral therapy (ART) scale-up.

      Results

      Estimated country-specific ϕ (in absence of ART) ranged between 4.2% (95% uncertainty interval (UI): 1.9%-6.3%) and 47.4% (95% UI: 37.2%-69.0%) per person-year (ppy), with a median of 12.4%. ϕ was strongly associated with HIV prevalence, with a Pearson correlation coefficient of 0.92, and was larger in high- versus low-HIV-prevalence countries. ϕ increased by 1.31% (95% confidence interval: 1.00%-1.55%) ppy for every 1% increase in HIV prevalence.

      Conclusions

      ϕ estimates were similar to earlier estimates, and suggested considerable heterogeneity in HIV infectiousness across SSA. This heterogeneity may explain, partly, the differences in epidemic scales.

      Keywords

      Introduction

      Quantifying the annual risk of HIV transmission (ϕ) within HIV sero-discordant couples (SDCs) in sub-Saharan Africa (SSA) is critical to understanding HIV epidemiology and implementation of intervention programs. Previously, we estimated ϕ and its variation and found that it varies across countries (
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      ). We also found that ϕ is within the range, but overall higher, than that estimated in setting-specific longitudinal studies following cohorts of SDCs in Africa (
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      ).
      A key limitation in earlier estimates is that the modeling approach did not factor the temporal variation in HIV prevalence, sexual risk-behavior heterogeneity, and scale-up of antiretroviral therapy (ART). Although the spatial dimension does not appear to play a critical role in sero-discordancy dynamics (
      • Cuadros D.F.
      • Abu-Raddad L.J.
      Geographical Patterns of HIV Sero-Discordancy in High HIV Prevalence Countries in Sub-Saharan Africa.
      ), historical and future patterns of sero-discordancy were found dependent on the temporal changes in HIV epidemics (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ).
      We present here new estimates for ϕ in 23 SSA countries using an elaborate modeling approach that accommodates for epidemic temporal variation, sexual risk-behavior heterogeneity, and ART scale-up. The new estimates also utilize all rounds of the nationally representative household surveys, the Demographic and Health Surveys (DHS) (
      • MEASURE DHS
      Demographic and health surveys. ICF Macro.
      ), for each country in SSA up to 2015.

      Methods

      A pair-based model was used to estimate the mean ϕ over the duration of the modeled epidemics. Briefly, the model consists of a set of coupled nonlinear differential equations stratifying the population according to stable-couple status, risk-behavior group, HIV status, and ART-treatment status (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ). The model accommodated for epidemic temporal variation, in its different phases, and for the risk of acquiring the infection from other non-stable partners (i.e. through extramarital sex). Ten risk groups, a sexual-risk mixing matrix, and temporal changes in risk behavior were also incorporated to account for heterogeneity and declines in risk of HIV exposure (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ,
      • Awad S.F.
      • Abu-Raddad L.J.
      Could there have been substantial declines in sexual risk behavior across sub-Saharan Africa in the mid-1990s?.
      ). The country-specific ART coverage was set as reported (
      • UNAIDS
      AIDSinfo. Coverage of people recieving ART.
      ), and ART reduced the overall HIV infectiousness (by 96% (
      • Cohen M.S.
      • Chen Y.Q.
      • McCauley M.
      • Gamble T.
      • Hosseinipour M.C.
      • Kumarasamy N.
      • et al.
      Prevention of HIV-1 infection with early antiretroviral therapy.
      )) and slowed disease progression (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ).
      Countries were included based on availability of DHS sero-prevalence surveys to apply the model to these data (
      • MEASURE DHS
      Demographic and health surveys. ICF Macro.
      ). The model was parameterized using natural history and epidemiology data from SSA (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ). We estimated ϕ by fitting the country-specific 1990-2015 HIV prevalence and sero-discordancy measures (
      • MEASURE DHS
      Demographic and health surveys. ICF Macro.
      ,
      • Chemaitelly H.
      • Cremin I.
      • Shelton J.
      • Hallett T.B.
      • Abu-Raddad L.J.
      Distinct HIV discordancy patterns by epidemic size in stable sexual partnerships in sub-Saharan Africa.
      ). All measures were weighted equally in the fitting. Multivariate uncertainty analysis was conducted for the model structural parameters, and the country-specific ϕ mean and 95% uncertainty interval (UI) were estimated using a log-normal fit. Association between the country-specific HIV prevalence and estimated ϕ was assessed using linear regression.
      Further details about this model and its parameterization can be found elsewhere (
      • Awad S.F.
      • Chemaitelly H.
      • Abu-Raddad L.J.
      Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
      ).

      Results

      Table S1 shows the derived DHS HIV prevalence and sero-discordancy measures that were fitted using the model. Estimated country-specific ϕ (in absence of ART) ranged between 4.2% (95% UI: 1.9%-6.3%) and 47.4% (95% UI: 37.2%-69.0%) per person-year (ppy), with a median of 12.4% ppy (Figure 1). The median was 9.4% ppy in low-HIV-prevalence countries (HIV prevalence ≤5%) and 25.2% ppy in high-HIV-prevalence countries (HIV prevalence >5%).
      Figure 1
      Figure 1Mean and 95% uncertainty interval for the country-specific annual risk of HIV transmission (ϕ) from the infected to the uninfected partner in a stable HIV sero-discordant couple across sub-Saharan Africa. The blue bars indicate the predictions of
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      , while the red bars indicate the predictions of the present study. Countries are shown by sub-region (
      • UNICEF
      Subregions and regions of Africa. UNICEF classifications for The State of Africa’s Children 2008 based on United Nations regional groupings.
      ) and in order of increasing 2015 HIV prevalence. HIV prevalence in 2015 is based on UNAIDS estimates (
      • UNAIDS
      Epidemiological data, HIV estimates 1990-2015.
      ).
      Figure 2 illustrates the association between the estimated ϕ and HIV prevalence in 2015, with Burundi and Rwanda excluded as outliers. ϕ was strongly correlated with HIV prevalence—Pearson correlation coefficient of 0.92 with HIV prevalence explaining 85% of the variation. ϕ increased by 1.31% (95% confidence interval: 1.00%-1.55%) ppy for every 1% increase in HIV prevalence (p-value < 0.001). The correlation was also assessed with HIV prevalence in earlier years, to capture the correlation at different epidemic phases. The correlation coefficient increased steadily with year, and was strong for all years starting from the mid-1990s. For example, the correlation coefficient was 0.88 with HIV prevalence in 2000, thereby explaining 77% of the variation.
      Figure 2
      Figure 2The variation in the mean annual risk of HIV transmission (ϕ) from the infected to the uninfected partner in a stable HIV sero-discordant couple with respect to HIV prevalence in the population. Country-specific estimates marked in red are excluded from the analysis as outliers. HIV prevalence in 2015 is based on UNAIDS estimates (
      • UNAIDS
      Epidemiological data, HIV estimates 1990-2015.
      ).

      Discussion

      We presented new estimates of HIV infectiousness within SDCs in 23 SSA countries. Despite the new analytical approach and updated input data, ϕ estimates were similar to those estimated earlier (Figure 1; p-value for difference in overall mean = 0.71) (
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      ). The estimated ϕ values indicated also large variability in infectiousness across countries. The contrasting HIV epidemic trajectories in SSA are possibly a consequence of the variability in ϕ—HIV prevalence in different epidemic phases was strongly associated with HIV infectiousness (Figure 2). This highlights how ϕ is high and needs to be addressed in high HIV prevalence countries, such as in Swaziland with a ϕ of 47% ppy and HIV prevalence of 29%. As more quality population-based data become available over the coming years, it will be useful to update these estimates based on more comprehensive data from more survey rounds.
      Our estimates for ϕ are roughly within the range of estimates from longitudinal studies, of 1.2%-22.0% ppy (
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      )—but higher on average. The lower observed ϕ in longitutal studies may be explained by known biases, such as higher levels of sero-status disclosure and intense counseling, in the studied cohorts versus the population at large, as well as selection bias of more “resistant” and “surviving” couples in the recruitment of SDCs (
      • Chemaitelly H.
      • Awad S.F.
      • Abu-Raddad L.J.
      The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
      ).
      The observed variability in ϕ across countries is supported by a body of evidence indicating a role for biological cofactors in HIV transmission such as male circumcision (
      • Weiss H.A.
      • Halperin D.
      • Bailey R.C.
      • Hayes R.J.
      • Schmid G.
      • Hankins C.A.
      Male circumcision for HIV prevention: from evidence to action.
      ), co-infections (whether sexually transmitted or not) (
      • Korenromp E.L.
      • de Vlas S.J.
      • Nagelkerke N.J.
      • Habbema J.D.
      Estimating the magnitude of STD cofactor effects on HIV transmission: how well can it be done.
      ,
      • Abu-Raddad L.J.
      • Patnaik P.
      • Kublin J.G.
      Dual infection with HIV and malaria fuels the spread of both diseases in sub-Saharan Africa.
      ,
      • Abu-Raddad L.J.
      • Barnabas R.V.
      • Janes H.
      • Weiss H.A.
      • Kublin J.G.
      • Longini Jr., I.M.
      • et al.
      Have the explosive HIV epidemics in sub-Saharan Africa been driven by higher community viral load?.
      ), variation in the susceptibility to HIV (
      • Nagelkerke N.
      • de Vlas S.J.
      • Jha P.
      • Luo M.
      • Plummer F.A.
      • Kaul R.
      Heterogeneity in host HIV susceptibility as a potential contributor to recent HIV prevalence declines in Africa.
      ,
      • Kaul R.
      • Cohen C.R.
      • Chege D.
      • Yi T.J.
      • Tharao W.
      • McKinnon L.R.
      • et al.
      Biological Factors that May Contribute to Regional and Racial Disparities in HIV Prevalence.
      ), virus sub-type (
      • Novitsky V.
      • Ndung’u T.
      • Wang R.
      • Bussmann H.
      • Chonco F.
      • Makhema J.
      • et al.
      Extended high viremics: a substantial fraction of individuals maintain high plasma viral RNA levels after acute HIV-1 subtype C infection.
      ), and hormonal contraception use (
      • Heffron R.
      • Donnell D.
      • Rees H.
      • Celum C.
      • Mugo N.
      • Were E.
      • et al.
      Use of hormonal contraceptives and risk of HIV-1 transmission: a prospective cohort study.
      ). Behavioral factors may also play a role such as variations in uptake of condom use (
      • Hughes J.P.
      • Baeten J.M.
      • Lingappa J.R.
      • Magaret A.S.
      • Wald A.
      • de Bruyn G.
      • et al.
      Partners in Prevention HSVHIVTST. Determinants of per-coital-act HIV-1 infectivity among African HIV-1-serodiscordant couples.
      ) and coital frequency (
      • Brown M.S.
      Coitus, the proximate determinant of conception: inter-country variance in sub-Saharan Africa.
      ). While our study estimated the overall mean ϕ across epidemic evolution, it would be of interest in future studies to investigate how these multiple cofactors could have affected ϕ’s variation across countries and over time.
      Our study has limitations. We assumed a specific level of sexual mixing between risk groups based on earlier work (
      • Abu-Raddad L.J.
      • Longini Jr., I.M.
      No HIV stage is dominant in driving the HIV epidemic in sub-Saharan Africa.
      ), but this assumption had no impact on the estimated ϕ, as demonstrated in a sensitivity analysis (Figure S1). We assumed that ϕ was constant during the observed time frame (1980-2030), but prevalence of cofactors affecting ϕ, such as male circumcision, may have changed over time—the estimated ϕ is simply an overall average over the modeled epidemic duration. Burundi and Rwanda were excluded from the regression analysis because of very high ϕ relative to HIV prevalence—possibly because of too small and non-representative SDC datasets.
      In summary, HIV infectiousness within sexual partnerships appears to vary by country probably due to a combination of biological and behavioral cofactors. This heterogeneity may explain, partly, the differences in the scales of SSA epidemics. This cumulative and consistent evidence for heterogeneity in infectiousness may underline a hallmark of the epidemiology of HIV—the extensive geographical variability in transmission patterns in comparison to other infections.

      Contributors

      SFA co-designed and programmed the model, conducted the modeling analyses, and wrote the first draft of the article. HC managed the Demographic and Health Surveys databases and conducted the statistical analyses on these databases. LJA conceived and led the design of the study and model, analyses, and drafting of the article. All authors contributed to the interpretation of the results and writing of the article.

      Competing interests

      We declare that we have no conflict of interest to disclose.

      Funding

      This publication was made possible by NPRP grant number 6-681-3-173 from the Qatar National Research Fund (a member of Qatar Foundation). The findings achieved herein are solely the responsibility of the authors.

      Acknowledgements

      The authors are grateful for infrastructure support provided by the Biostatistics, Epidemiology, and Biomathematics Research Core at Weill Cornell Medicine-Qatar. The authors are also thankful for Measure Demographic and Health Surveys (Measure DHS) for putting these national surveys in the service of science, and for the United States Agency for International Development and other donors supporting these initiatives.

      References

        • Abu-Raddad L.J.
        • Longini Jr., I.M.
        No HIV stage is dominant in driving the HIV epidemic in sub-Saharan Africa.
        AIDS. 2008; 22: 1055-1061
        • Abu-Raddad L.J.
        • Patnaik P.
        • Kublin J.G.
        Dual infection with HIV and malaria fuels the spread of both diseases in sub-Saharan Africa.
        Science. 2006; 314: 1603-1606
        • Abu-Raddad L.J.
        • Barnabas R.V.
        • Janes H.
        • Weiss H.A.
        • Kublin J.G.
        • Longini Jr., I.M.
        • et al.
        Have the explosive HIV epidemics in sub-Saharan Africa been driven by higher community viral load?.
        AIDS. 2013; 27: 981-989
        • Awad S.F.
        • Abu-Raddad L.J.
        Could there have been substantial declines in sexual risk behavior across sub-Saharan Africa in the mid-1990s?.
        Epidemics. 2014; 8: 9-17
        • Awad S.F.
        • Chemaitelly H.
        • Abu-Raddad L.J.
        Temporal evolution of HIV sero-discordancy patterns among stable couples in sub-Saharan Africa.
        2017 (Under Review)
        • Brown M.S.
        Coitus, the proximate determinant of conception: inter-country variance in sub-Saharan Africa.
        J Biosoc Sci. 2000; 32: 145-159
        • Chemaitelly H.
        • Cremin I.
        • Shelton J.
        • Hallett T.B.
        • Abu-Raddad L.J.
        Distinct HIV discordancy patterns by epidemic size in stable sexual partnerships in sub-Saharan Africa.
        Sex Transm Infect. 2012; 88: 51-57
        • Chemaitelly H.
        • Awad S.F.
        • Abu-Raddad L.J.
        The risk of HIV transmission within HIV-1 sero-discordant couples appears to vary across sub-Saharan Africa.
        Epidemics. 2014; 6: 1-9
        • Cohen M.S.
        • Chen Y.Q.
        • McCauley M.
        • Gamble T.
        • Hosseinipour M.C.
        • Kumarasamy N.
        • et al.
        Prevention of HIV-1 infection with early antiretroviral therapy.
        N Engl J Med. 2011; 365: 493-505
        • Cuadros D.F.
        • Abu-Raddad L.J.
        Geographical Patterns of HIV Sero-Discordancy in High HIV Prevalence Countries in Sub-Saharan Africa.
        Int J Environ Res Public Health. 2016; : 13
        • Heffron R.
        • Donnell D.
        • Rees H.
        • Celum C.
        • Mugo N.
        • Were E.
        • et al.
        Use of hormonal contraceptives and risk of HIV-1 transmission: a prospective cohort study.
        Lancet Infect Dis. 2012; 12: 19-26
        • Hughes J.P.
        • Baeten J.M.
        • Lingappa J.R.
        • Magaret A.S.
        • Wald A.
        • de Bruyn G.
        • et al.
        Partners in Prevention HSVHIVTST. Determinants of per-coital-act HIV-1 infectivity among African HIV-1-serodiscordant couples.
        J Infect Dis. 2012; 205: 358-365
        • Kaul R.
        • Cohen C.R.
        • Chege D.
        • Yi T.J.
        • Tharao W.
        • McKinnon L.R.
        • et al.
        Biological Factors that May Contribute to Regional and Racial Disparities in HIV Prevalence.
        Am J Reproduct Immunol. 2011; 65: 317-324
        • Korenromp E.L.
        • de Vlas S.J.
        • Nagelkerke N.J.
        • Habbema J.D.
        Estimating the magnitude of STD cofactor effects on HIV transmission: how well can it be done.
        Sex Transm Dis. 2001; 28: 613-621
        • MEASURE DHS
        Demographic and health surveys. ICF Macro.
        2017 (Available: http://www.measuredhs.com/)
        • Nagelkerke N.
        • de Vlas S.J.
        • Jha P.
        • Luo M.
        • Plummer F.A.
        • Kaul R.
        Heterogeneity in host HIV susceptibility as a potential contributor to recent HIV prevalence declines in Africa.
        AIDS. 2009; 23: 125-130
        • Novitsky V.
        • Ndung’u T.
        • Wang R.
        • Bussmann H.
        • Chonco F.
        • Makhema J.
        • et al.
        Extended high viremics: a substantial fraction of individuals maintain high plasma viral RNA levels after acute HIV-1 subtype C infection.
        AIDS. 2011; 25: 1515-1522
        • UNAIDS
        Epidemiological data, HIV estimates 1990-2015.
        2015 (Available: http://www.unaids.org/en/dataanalysis/datatools/aidsinfo)
        • UNAIDS
        AIDSinfo. Coverage of people recieving ART.
        2015 (Available at: http://aidsinfo.unaids.org/. [Accessed April, 2016])
        • UNICEF
        Subregions and regions of Africa. UNICEF classifications for The State of Africa’s Children 2008 based on United Nations regional groupings.
        2008 (Available at: https://www.unicef.org/wcaro/WCARO_SOAC08_Fig011.pdf)
        • Weiss H.A.
        • Halperin D.
        • Bailey R.C.
        • Hayes R.J.
        • Schmid G.
        • Hankins C.A.
        Male circumcision for HIV prevention: from evidence to action.
        AIDS. 2008; 22: 567-574