Prevalence of curable STIs and bacterial vaginosis during pregnancy in sub-Saharan Africa: a systematic review and meta-analysis.

OBJECTIVE: STIs remain a global public health problem with a high burden among pregnant women. STIs in pregnant women may lead to various adverse pregnancy outcomes. In most sub-Saharan African countries, syndromic management is used for screening and treatment of STIs. We aimed to update and summarise pooled prevalence of curable STIs and bacterial vaginosis (BV) among pregnant women in sub-Saharan Africa. METHODS: Electronic databases and reference lists of relevant published and unpublished studies were searched from March 2015 to October 2020. Studies were included if they estimated prevalence of Chlamydia trachomatis (CT), Trichomonas vaginalis (TV), Neisseria gonorrhoeae (NG), Treponema pallidum (syphilis), Mycoplasma genitalium (MG) and BV among pregnant women in sub-Saharan Africa. Meta-analyses were performed with observed prevalences corrected for diagnostic errors to estimate the pooled prevalence of diagnosed infections by region. RESULTS: A total of 48 studies met the inclusion criteria, providing 85-point prevalence estimates for curable STIs and BV. Pooled prevalence estimates (with 95% CI and number of women tested) were as follows: MG: 13.5% (4.0-27.2, n=1076); CT: 10.8% (6.9-15.5, n=6700); TV: 13.8% (10.0-18.0, n=9264); NG: 3.3% (2.1-4.7, n=6019); syphilis: 2.9% (2.0-4.0, n=95 308) and BV: 36.6% (27.1-46.6, n=5042). By region, BV was the most prevalent and ranged from 28.5% (24.5-32.8, n=1030) in Eastern Africa to 52.4% (33.5-70.9, n=2305) in Southern Africa; NG had the lowest prevalence, ranging from 1.4% (95% CI 0.1 to 3.1, n=367) in Central Africa to 4.4% (95% CI 2.6 to 6.4, n=4042) in Southern Africa. CONCLUSION: The prevalence of curable STIs and BV in sub-Saharan Africa is substantial in pregnant women but most prevalent in Southern Africa where HIV prevalence is highest. It is crucial to integrate screening of curable STIs into antenatal care programmes that have previously focused on diagnosis and treatment of syphilis and HIV.

Prevalence, incidence and associated risk factors of STIs during pregnancy in South Africa.

OBJECTIVE: STIs during pregnancy increase adverse pregnancy and birth outcomes and may increase HIV risk. STI syndromic management is standard of care in South Africa. Our study evaluated the prevalence and incidence of STIs in pregnant women and the associated risk factors. METHODS: We combined data from two prospective observational studies of pregnant women enrolled while attending their first antenatal clinic (ANC) visit in Tshwane District and Cape Town. Women ≥18 years were tested at first ANC visit and at their first postpartum visit for Chlamydia trachomatis, Neisseria gonorrhoeae and Trichomonas vaginalis using Xpert assays (Cepheid, USA). We evaluated the prevalence and incidence of STI and the associated risk factors using multivariable regression models. RESULTS: We enrolled 669 pregnant women, 64% (n=427) from Tshwane District and 36% (n=242) from Cape Town; 80% (n=534) were women living with HIV (WLHIV) and 20% (n=135) without HIV. At enrolment, 37% (n=250) were diagnosed with at least one STI, of which 76% (n=190) were asymptomatic. STI prevalence was 40% (n=213) in WLHIV and 27% (n=37) in women without HIV (p=0.01). Baseline STI infection was associated with younger age (OR=0.95 per year, 95% CI 0.92 to 0.98), higher gestational age (adjusted OR (aOR)=1.03 per week, 95% CI 1.00 to 1.05), single relationship status (aOR=1.53, 95% CI 1.09 to 2.15) and HIV status (aOR=1.86, 95% CI 1.17 to 2.95). Of 419 participants with no STI at baseline, 21 had an incident STI during follow-up, with a mean follow-up time of 140 days. The incidence rate of STI during pregnancy and early post partum was 15 infections per 100 women-years (95% CI 9 to 23). Younger age was associated with STI incidence. CONCLUSION: Our study shows high prevalence and incidence of STIs in pregnancy, especially in WLHIV, demonstrating the need for STI screening in ANC to prevent adverse pregnancy and birth outcomes. Most STI cases were asymptomatic and would have gone untreated with syndromic management. Aetiological STI screening is urgently needed to reduce the burden of STIs in pregnancy.

Are associations between HIV and human papillomavirus transmission due to behavioural confounding or biological effects?

OBJECTIVES: Cohort studies have shown significant increased risk of HIV acquisition following human papillomavirus (HPV) detection and increased risk of new HPV detection in individuals with HIV infection, after adjusting for behavioural risk factors. This study uses an individual-based model to assess whether confounding sexual behaviour factors and network level effects can explain these associations between HIV and HPV infection status, without biological interactions. METHODS: The model simulates infection with 13 oncogenic HPV types and HIV. It allows for different relationship types, with heterogeneity in probabilities of concurrency and rates of partner change. No effect of prevalent HPV infection on HIV acquisition is assumed and vice versa. The model is calibrated to South African HIV and type-specific HPV prevalence data using a Bayesian approach. The model is used to simulate cohorts with quarterly HIV and HPV testing from 2000 to 2002. These simulated data are analysed using proportional hazard models. RESULTS: The mean of the unadjusted HRs of HIV acquisition following detection of an oncogenic HPV type calculated for each simulated cohort is 2.6 (95% CI 2.2 to 3.1). The mean of the unadjusted HRs for the effect of HIV on newly detected HPV is 2.5 (95% CI 2.2 to 2.8). Simulated associations between HIV and HPV infection status are similar to corresponding empirical estimates. In sensitivity analyses in which HIV and HPV were assumed to increase each other's transmission risk, simulated associations were stronger but not inconsistent with empirical estimates. CONCLUSIONS: Although we cannot rule out the possibility that associations between HIV and HPV transmission may be due in part to biological interactions, these results suggest that observed associations could be explained entirely by residual confounding by behavioural factors and network-level effects that observational studies cannot account for.

Estimated mortality of adult HIV-infected patients starting treatment with combination antiretroviral therapy.

OBJECTIVE: To provide estimates of mortality among HIV-infected patients starting combination antiretroviral therapy. METHODS: We report on the death rates from 122 925 adult HIV-infected patients aged 15 years or older from East, Southern and West Africa, Asia Pacific and Latin America. We use two methods to adjust for biases in mortality estimation resulting from loss from follow-up, based on double-sampling methods applied to patient outreach (Kenya) and linkage with vital registries (South Africa), and apply these to mortality estimates in the other three regions. Age, gender and CD4 count at the initiation of therapy were the factors considered as predictors of mortality at 6, 12, 24 and >24 months after the start of treatment. RESULTS: Patient mortality was high during the first 6 months after therapy for all patient subgroups and exceeded 40 per 100 patient years among patients who started treatment at low CD4 count. This trend was seen regardless of region, demographic or disease-related risk factor. Mortality was under-reported by up to or exceeding 100% when comparing estimates obtained from passive monitoring of patient vital status. CONCLUSIONS: Despite advances in antiretroviral treatment coverage many patients start treatment at very low CD4 counts and experience significant mortality during the first 6 months after treatment initiation. Active patient tracing and linkage with vital registries are critical in adjusting estimates of mortality, particularly in low- and middle-income settings.