Human Immunodeficiency Virus Incidence Among Women at High-Risk of Human Immunodeficiency Virus Infection Attending a Dedicated Clinic in Kampala, Uganda: 2008-2017

HIV incidence among women engaging in sex work in sub-Saharan Africa: a systematic review and meta-analysis

BACKGROUND

Women who engage in sex work in sub-Saharan Africa have a high risk of acquiring HIV infection. HIV incidence has declined among all women in sub-Saharan Africa, but trends among women who engage in sex work are poorly characterised. We synthesised data on HIV incidence among women who engage in sex work in sub-Saharan Africa and compared these with the total female population to understand relative incidence and trends over time.

METHODS

We searched MEDLINE, Embase, Global Health, and Google Scholar from Jan 1, 1990, to Feb 28, 2024, and grey literature for studies that reported empirical estimates of HIV incidence among women who engage in sex work in any sub-Saharan Africa country. We calculated incidence rate ratios (IRRs) compared with total female population incidence estimates matched for age, district, and year, did a meta-analysis of IRRs, and used a continuous mixed-effects model to estimate changes in IRR over time.

FINDINGS

From 32 studies done between 1985 and 2020, 2194 new HIV infections were observed among women who engage in sex work over 51 490 person-years. Median HIV incidence was 4·3 per 100 person years (IQR 2·8-7·0 per 100 person-years). Incidence among women who engage in sex work was eight times higher than matched total population women (IRR 7·8 [95% CI 5·1-11·8]), with larger relative difference in western and central Africa (19·9 [9·6-41·0]) than in eastern and southern Africa (4·9 [3·4-7·1]). There was no evidence that IRRs changed over time (IRR per 5 years: 0·9 [0·7-1·2]).

INTERPRETATION

Across sub-Saharan Africa, HIV incidence among women who engage in sex work remains disproportionately high compared with the total female population. However, constant relative incidence over time indicates HIV incidence among women who engage in sex work has declined at a similar rate. Location-specific data for women who engage in sex work incidence are sparse, but improved surveillance and standardisation of incidence measurement approaches could fill these gaps. Sustained and enhanced HIV prevention for women who engage in sex work is crucial to address continuing inequalities and ensure declines in new HIV infections.

FUNDING

Bill & Melinda Gates Foundation, UK Research and Innovation, National Institutes of Health.

TRANSLATION

For the French translation of the abstract see Supplementary Materials section.

HIV incidence among women engaging in sex work in sub-Saharan Africa: a systematic review and meta-analysis

Introduction

HIV incidence among women in sub-Saharan Africa (SSA) has declined steadily, but it is unknown whether new infections among women who engage in sex work (WESW) have declined at a similar rate. We synthesised estimates of HIV incidence among WESW in SSA and compared these to the wider female population to understand levels and trends in incidence over time.

Methods

We searched Medline, Embase, Global Health, Popline, Web of Science, and Google Scholar from January 1990 to October 2022, and grey literature for estimates of HIV incidence among WESW in SSA. We included studies reporting empirical estimates in any SSA country. We calculated incidence rate ratios (IRR) compared to age-district-year matched total female population incidence estimates. We conducted a meta-analysis of IRRs and used a continuous mixed-effects model to estimate changes in IRR over time.

Results

From 32 studies between 1985 and 2020, 2,194 new HIV infections were observed in WESW over 51,000 person-years (py). Median HIV incidence was 4.3/100py (IQR 2.8-7.0/100py), declining from a median of 5.96/100py between 1985 and 1995 to a median of 3.2/100py between 2010 and 2020. Incidence among WESW was nine times higher than in matched total population women (RR 8.6, 95%CI: 5.7-12.9), and greater in Western and Central Africa (RR 22.4, 95%CI: 11.3-44.3) than in Eastern and Southern Africa (RR 5.3, 95%CI: 3.7-7.6). Annual changes in log IRRs were minimal (-0.1% 95%CI: -6.9 to +6.8%).

Conclusions

Across SSA, HIV incidence among WESW remains disproportionately high compared to the total female population but showed similar rates of decline between 1990 and 2020. Improved surveillance and standardisation of approaches to obtain empirical estimates of sex worker incidence would enable a clearer understanding of whether we are on track to meet global targets for this population and better support data-driven HIV prevention programming.

Thirty years of change in HIV incidence among adults in the Kyamulibwa General Population Cohort in rural southwest Uganda, 1989-2021

OBJECTIVES

To document the changes in HIV incidence over thirty years in Kalungu district, Uganda.

METHODS

Since 1989, residents aged ≥15 years old have been tested for HIV, and data were collected on HIV risk factors annually and later, biennially in the Kyamulibwa open cohort. In the 2019-2021 survey, people living with HIV self-reported on knowledge of their HIV status, antiretroviral therapy (ART) use, and their most recent viral load data were obtained from health facilities. The HIV seroconversion dates were randomly imputed between the last negative and first positive test dates using a uniform distribution.

RESULTS

Among 20,959 residents who were HIV-negative, 669 seroconverted within 176,659 person-years. Data showed a downward trend in age-adjusted HIV incidence over 30 years (P <0.001) even though HIV prevalence steadily increased with ART availability from 2004. Comparing 1990-1992 and 1996-1998, HIV incidence declined by 43% (0.79 to 0.45/100 person-years, P = 0.002). Between 1999 and 2011, the incidence remained stable at 0.49/100 person-years (95% confidence interval: 0.41-0.58) in men but slowly increased in women (average age-adjusted hazard ratio = 1.13 per 3 years, 95% confidence interval: 1.03-1.24; trend P-value = 0.02). After 2011, however, the incidence trends reversed and continued to decline in men and women and in all age groups.

CONCLUSION

Facilitating HIV testing and timely ART initiation, and supporting ART adherence must be emphasized alongside sustainable prevention measures.

Ongoing HIV transmission following a large outbreak among people who inject drugs in Athens, Greece (2014-20)

BACKGROUND AND AIMS

The human immunodeficiency virus (HIV) outbreak among people who inject drugs (PWID) in Athens, Greece in 2011-13 was the largest recent epidemic in Europe and North America. We aimed to assess trends in HIV prevalence, drug use and access to prevention among PWID in Athens to estimate HIV incidence and identify risk factors and to explore HIV-1 dispersal using molecular methods during 2014-20.

METHODS

Two community-based HIV/hepatitis C programmes on PWID were implemented in 2012-13 (n = 3320) and 2018-20 (n = 1635) through consecutive respondent-driven sampling (RDS) rounds. PWID were uniquely identified among rounds/programmes. We obtained RDS-weighted HIV prevalence estimates per round for 2018-20 and compared them to 2012-13. We assessed changes in HIV status, behaviours and access to prevention in PWID participating in both periods. We estimated HIV incidence in a cohort of seronegative PWID as the number of HIV seroconversions/100 person-years during 2014-20 and used Cox regression to identify associated risk factors. Molecular sequencing and phylogenetic analysis were performed in HIV seroconverters.

RESULTS

HIV prevalence per round ranged between 12.0 and 16.2% in 2012-13 and 10.7 and 11.3% in 2018-20 with overlapping 95% confidence intervals (95% CI). Among PWID participating in both programmes, HIV prevalence (95% CI) increased from 14.2% (11.7-17.1%) in 2012-13 to 22.0% (19.0-25.3%) in 2018-20 (P < 0.001). There was a deterioration in socio-economic characteristics such as homelessness [from 16.2% (95% CI = 13.5-19.2%) to 25.6% (22.3-29.0%)], a shift in cocaine use [16.6% (13.9-19.6%) versus 28.1% (24.7-31.7%], reduced access to free syringes [51.8% (48.0-55.7%) versus 44.5% (40.7-48.3%)] and a decrease in daily injecting [36.2% (32.6-39.9%) versus 28.5% (25.2-32.1%)]. HIV incidence (95% CI) in 2014-20 was 1.94 (1.50-2.52) new cases/100 person-years and younger age, lower educational level, larger injection network and daily injecting were risk factors. Almost 9% of HIV seroconversions occurred within a newly expanding phylogenetic cluster.

CONCLUSIONS

In Athens, Greece, compared with the period 2012-13, in the period 2018-20 there was a deterioration in socio-economic conditions among people who inject drugs, an increase in the use of cocaine, reduced access to needle and syringe programmes and stable low levels of human immunodeficiency virus testing. Ongoing human immunodeficiency virus transmission was documented during 2014-20 in existing as well as new transmission clusters.

Phylogenetic Networks and Parameters Inferred from HIV Nucleotide Sequences of High-Risk and General Population Groups in Uganda: Implications for Epidemic Control

Phylogenetic inference is useful in characterising HIV transmission networks and assessing where prevention is likely to have the greatest impact. However, estimating parameters that influence the network structure is still scarce, but important in evaluating determinants of HIV spread. We analyzed 2017 HIV pol sequences (728 Lake Victoria fisherfolk communities (FFCs), 592 female sex workers (FSWs) and 697 general population (GP)) to identify transmission networks on Maximum Likelihood (ML) phylogenetic trees and refined them using time-resolved phylogenies. Network generative models were fitted to the observed degree distributions and network parameters, and corrected Akaike Information Criteria and Bayesian Information Criteria values were estimated. 347 (17.2%) HIV sequences were linked on ML trees (maximum genetic distance ≤4.5%, ≥95% bootstrap support) and, of these, 303 (86.7%) that consisted of pure A1 (n = 168) and D (n = 135) subtypes were analyzed in BEAST v1.8.4. The majority of networks (at least 40%) were found at a time depth of ≤5 years. The waring and yule models fitted best networks of FFCs and FSWs respectively while the negative binomial model fitted best networks in the GP. The network structure in the HIV-hyperendemic FFCs is likely to be scale-free and shaped by preferential attachment, in contrast to the GP. The findings support the targeting of interventions for FFCs in a timely manner for effective epidemic control. Interventions ought to be tailored according to the dynamics of the HIV epidemic in the target population and understanding the network structure is critical in ensuring the success of HIV prevention programs.

Use of propensity score matching to create counterfactual group to assess potential HIV prevention interventions

The design of HIV prevention trials in the context of effective HIV preventive methods is a challenge. Alternate designs, including using non-randomised 'observational control arms' have been proposed. We used HIV simulated vaccine efficacy trials (SiVETs) to show pitfalls that may arise from using such observational controls and suggest how to conduct the analysis in the face of the pitfalls. Two SiVETs were nested within previously established observational cohorts of fisherfolk (FF) and female sex workers (FSW) in Uganda. SiVET participants received a licensed Hepatitis B vaccine in a schedule (0, 1 and 6 months) similar to that for a possible HIV vaccine efficacy trial. All participants received HIV counselling and testing every quarter for one year to assess HIV incidence rate ratio (IRR) between SiVET and non-SiVET (observational data). Propensity scores, conditional on baseline characteristics were calculated for SiVET participation and matched between SiVET and non-SiVET in the period before and during the SiVET study. We compared IRR before and after propensity score matching (PSM). In total, 3989 participants were enrolled into observational cohorts prior to SiVET, (1575 FF prior to Jul 2012 and 2414 FSW prior to Aug 2014). SiVET enrolled 572 participants (Jul 2012 to Apr 2014 in FF and Aug 2014 to Apr 2017 in FSW), with 953 non-SiVET participants observed in the SiVET concurrent period and 2928 from the pre-SiVET period (before Jul 2012 in FF or before Apr 2014 in FSW). Imbalances in baseline characteristics were observed between SiVET and non-SiVET participants in both periods before PSM. Similarly, HIV incidence was lower in SiVET than non-SiVET; SiVET-concurrent period, IRR = 0.59, 95% CI 0.31-0.68, p = 0.033 and pre-SiVET period, IRR = 0.77, 95% CI 0.43-1.29, p = 0.161. After PSM, participants baseline characteristics were comparable and there were minimal differences in HIV incidence between SiVET and non-SiVET participants. The process of screening for eligibility for efficacy trial selects participants with baseline characteristics different from the source population, confounding any observed differences in HIV incidence. Propensity score matching can be a useful tool to adjust the imbalance in the measured participants' baseline characteristics creating a counterfactual group to estimate the effect of interventions on HIV incidence.

The Molecular Epidemiology and Transmission Dynamics of HIV Type 1 in a General Population Cohort in Uganda

The General Population Cohort (GPC) in south-western Uganda has a low HIV-1 incidence rate (<1%). However, new infections continue to emerge. In this research, 3796 HIV-1 pol sequences (GPC: n = 1418, non-GPC sites: n = 1223, Central Uganda: n = 1010 and Eastern Uganda: n = 145) generated between 2003–2015 were analysed using phylogenetic methods with demographic data to understand HIV-1 transmission in this cohort and inform the epidemic response. HIV-1 subtype A1 was the most prevalent strain in the GPC area (GPC and non-GPC sites) (39.8%), central (45.9%) and eastern (52.4%) Uganda. However, in the GPC alone, subtype D was the predominant subtype (39.1%). Of the 524 transmission clusters identified by Cluster Picker, all large clusters (≥5 individuals, n = 8) involved individuals from the GPC. In a multivariate analysis, clustering was strongly associated with being female (adjusted Odds Ratio, aOR = 1.28; 95% CI, 1.06–1.54), being >25 years (aOR = 1.52; 95% CI, 1.16–2.0) and being a resident in the GPC (aOR = 6.90; 95% CI, 5.22–9.21). Phylogeographic analysis showed significant viral dissemination (Bayes Factor test, BF > 3) from the GPC without significant viral introductions (BF < 3) into the GPC. The findings suggest localized HIV-1 transmission in the GPC. Intensifying geographically focused combination interventions in the GPC would contribute towards controlling HIV-1 infections.

Comparison of HIV Risk Behaviors Between Clinical Trials and Observational Cohorts in Uganda

Many key populations have high-risk behaviors for HIV infection making them suitable for HIV vaccine efficacy trials. However, these behaviors may change when participants enroll into a trial. We used HIV simulated vaccine efficacy trials (SiVETs) nested within observational cohorts of fisherfolks and female sex workers in Uganda to evaluate this difference. We screened observational cohort participants for enrolment into SiVETs, until 572 were enrolled. Those not enrolled (n = 953) continued participation in the observational cohorts. We determined risk behaviors at baseline and at 1 year, assigned a numeric score to each behavior and defined composite score as the sum of reported behaviors. We compared changes in scores over 12 months. Both observational cohorts and SiVETs saw a significant decrease in score but greatest in the SiVETs. Investigators recruiting for trials from these populations should consider the likely effect of reduction in risk behaviors on incident HIV infection and trial statistical power.

High Levels of Acquired HIV Drug Resistance Following Virological Nonsuppression in HIV-Infected Women from a High-Risk Cohort in Uganda

HIV drug resistance (HIVDR) is of increasing health concern, especially among key populations. We investigated the prevalence of virological suppression (VS), prevalence and correlates of HIVDR in HIV-infected women, enrolled in a high-risk cohort. We enrolled 267 women initiated on first-line antiretroviral therapy (ART) between 2015 and 2018. Participants' plasma samples were analyzed for HIV RNA viral load (VL) and genotypic resistance testing was performed on those with VL nonsuppression (defined as VL ≥1,000 copies/mL). We used the Stanford HIVDR database-algorithm to assess HIVDR mutations and logistic regression to assess risk factors for VL nonsuppression and HIVDR. We observed an overall VS prevalence of 76.0% (203/267) and detected respective acquired drug resistance prevalence to non-nucleoside reverse transcriptase inhibitors (NNRTIs) and nucleoside reverse transcriptase inhibitors (NRTIs) of 81.3% [confidence interval (CI) 67.4-91.1] and 45.8% (CI 31.4-60.8) among the 48 successfully genotyped VL nonsuppressors. NNRTI mutations were observed in 81.3% (39/48) of the genotyped participants and 45.8% (22/48) had both NRTI and NNRTI mutations. The mutation K103N was detected in 62.5% (30/48) of participants, 41.7% (20/48) had M184V/I, 14.6% had K65R, and 12.5% (6/48) had thymidine analog mutations (TAMs). None of the analyzed potential risk factors, including age and duration on ART, was significantly correlated with VL nonsuppression or HIVDR. Although high levels of NNRTI mutations support the transition to dolutegravir, the presence of NRTI mutations, especially TAMs, may compromise dolutegravir-based regimens or other second-line ART options. The moderate VS prevalence and high HIVDR prevalence therefore call for timely ART switching and intensive adherence counseling.

Missed Study Visits and Subsequent HIV Incidence Among Women in a Predominantly Sex Worker Cohort Attending a Dedicated Clinic Service in Kampala, Uganda

BACKGROUND

There is limited evidence on the relationship between sustained exposure of female sex workers (FSWs) to targeted HIV programmes and HIV incidence. We investigate the relationship between the number of missed study visits (MSVs) within each episode of 2 consecutively attended visits (MSVs) and subsequent HIV risk in a predominantly FSW cohort.

METHODS

Women at high risk of HIV are invited to attend an ongoing dedicated clinic offering a combination HIV prevention intervention in Kampala, Uganda. Study visits are scheduled once every 3 months. The analysis included HIV-seronegative women with ≥1 follow-up visit from enrollment (between April 2008 and May 2017) to August 2017. Cox regression models were fitted adjusted for characteristics on sociodemographic, reproductive, behavioral, and sexually transmitted infections (through clinical examination and serological testing for syphilis).

FINDINGS

Among 2206 participants, HIV incidence was 3.1/100 (170/5540) person-years [95% confidence interval (CI): 2.6 to 3.5]. Incidence increased from 2.6/100 person-years (95% CI: 2.1 to 3.2) in episodes without a MSV to 3.0/100 (95% CI: 2.2 to 4.1) for 1-2 MSVs and 4.3/100 (95% CI: 3.3 to 5.6) for ≥3 MSVs. Relative to episodes without a MSV, the hazard ratios (adjusted for confounding variables) were 1.40 (95% CI: 0.93 to 2.12) for 1-2 MSVs and 2.00 (95% CI: 1.35 to 2.95) for ≥3 MSVs (P-trend = 0.001).

CONCLUSION

Missing study visits was associated with increased subsequent HIV risk. Although several factors may underlie this association, the finding suggests effectiveness of targeted combination HIV prevention. But exposure to targeted interventions needs to be monitored, facilitated, and sustained in FSWs.

Phylogenetic and Demographic Characterization of Directed HIV-1 Transmission Using Deep Sequences from High-Risk and General Population Cohorts/Groups in Uganda

Across sub-Saharan Africa, key populations with elevated HIV-1 incidence and/or prevalence have been identified, but their contribution to disease spread remains unclear. We performed viral deep-sequence phylogenetic analyses to quantify transmission dynamics between the general population (GP), fisherfolk communities (FF), and women at high risk of infection and their clients (WHR) in central and southwestern Uganda. Between August 2014 and August 2017, 6185 HIV-1 positive individuals were enrolled in 3 GP and 10 FF communities, 3 WHR enrollment sites. A total of 2531 antiretroviral therapy (ART) naïve participants with plasma viral load >1000 copies/mL were deep-sequenced. One hundred and twenty-three transmission networks were reconstructed, including 105 phylogenetically highly supported source–recipient pairs. Only one pair involved a WHR and male participant, suggesting that improved population sampling is needed to assess empirically the role of WHR to the transmission dynamics. More transmissions were observed from the GP communities to FF communities than vice versa, with an estimated flow ratio of 1.56 (95% CrI 0.68–3.72), indicating that fishing communities on Lake Victoria are not a net source of transmission flow to neighboring communities further inland. Men contributed disproportionally to HIV-1 transmission flow regardless of age, suggesting that prevention efforts need to better aid men to engage with and stay in care.

Comparison of retention in observational cohorts and nested simulated HIV vaccine efficacy trials in the key populations in Uganda

Background

Outcomes in observational studies may not best estimate those expected in the HIV vaccine efficacy trials. We compared retention in Simulated HIV Vaccine Efficacy Trials (SiVETs) and observational cohorts drawn from two key populations in Uganda.

Methods

Two SiVETs were nested within two observational cohorts, one in Fisherfolk (FF) and another one in Female Sex Workers (FSW). Adult participants in each observational cohort were screened for enrolment into SiVETs. Those screened-out or not screened continued participation in the observational (non-SiVET) cohorts. SiVET participants were administered a licensed hepatitis B vaccine in a schedule that mimicked an actual HIV vaccine efficacy trial. Both cohorts were followed for 12 months and retention was assessed through dropout, defined as lost to follow up, being uncontactable, refusal to continue or missing the last study clinic visit. Dropout rates were compared using Poisson models giving rate ratios and 95% confidence intervals (95%CI).

Results

Out of 1525 participants (565 FF and 960 FSW), 572 (38%) were enrolled into SiVETs (282-FF and 290-FSW), and 953 (62%) remained in the non-SiVET cohorts. Overall, 326 (101 SiVET, 225 non-SiVET) dropped out in 1260 Person Years of Observation (PYO), a dropout rate of 25.9 /100 PYO (95%CI: 23.2–28.8); fewer dropped out in the SiVET cohorts (18.4, 95% CI: 15.1–22.4) than in the non-SiVET cohorts (31.6, 95% CI: 27.8–36.1), rate ratio (RR) =0.58, 95% CI: 0.46–0.73. In all cohorts, the dropout was more marked in FSW than in FF population. Duration lived in community was associated with dropout in both SiVETs and religion in both non-SiVET cohorts.

Conclusion

The rate of dropout was lower in SiVET compared to non-SiVET cohort. Though the difference in dropout between SiVET and non-SiVET was generally similar, the actual dropout rates were higher in the FSW population. Conduct of SiVETs in these key populations could mean that designing HIV Vaccine Efficacy Trials will benefit from lower dropout rate shown in SiVET than non-SiVET observational cohort.