Geographical Associations of HIV Prevalence in Female Sex Workers From Nairobi, Kenya (2014-2017).

BACKGROUND: Kenya's HIV epidemic is heterogeneously distributed. Although HIV incidence in Kenya has shown signs of recent decline, focused interventions are still needed for female sex workers (FSWs). Geospatially informed approaches have been advocated for targeted HIV prevention. We quantified heterogeneity in HIV burden in Nairobi-based FSWs by place of origin within Kenya and hotspots and residence within Nairobi. METHODS: Data were collected as part of enrolment in the Sex Workers Outreach Program in Nairobi between 2014 and 2017. Prevalence ratios were used to quantify the risk of HIV by high-prevalence counties using modified Poisson regression analyses. Crude and fully adjusted models were fitted to the data. In heterogeneity analyses, hotspots and residences were aggregated to the Nairobi constituency level (n = 17). Inequality in the geographic distribution of HIV prevalence was measured using the Gini coefficient. RESULTS: A total of 11,899 FSWs were included. Overall HIV prevalence was 16%. FSWs originating from a high-prevalence country were at 2-fold increased risk of living with HIV in adjusted analysis (prevalence ratio 1.95; 95% CI: 1.76 to 2.17). HIV prevalence was also highly heterogeneous by hotspot, ranging from 7% to 52% by hotspot (Gini coefficient: 0.37; 95% CI: 0.23 to 0.50). By contrast, the constituency of residence had a Gini coefficient of 0.08 (95% CI: 0.06 to 0.10), suggesting minimal heterogeneity by residence. CONCLUSION: HIV prevalence in FSWs is heterogeneous by place of work within Nairobi and by county of birth within Kenya. As HIV incidence declines and financial commitments flatline, tailoring interventions to FSWs at highest HIV risk becomes increasingly important.

Declines in HIV prevalence in female sex workers accessing an HIV treatment and prevention programme in Nairobi, Kenya over a 10-year period.

OBJECTIVES: Empirical time trends in HIV prevalence in female sex workers (FSWs) are helpful to understand the evolving HIV epidemic, and to monitor the scale-up, coverage, and impact of ongoing HIV prevention and treatment programmes. DESIGN: Serial HIV prevalence study. METHODS: We analyzed time trends in HIV prevalence in FSWs accessing services at seven Sex Worker Outreach Programme (SWOP) clinics in Nairobi from 2008 to 2017 (N = 33 560). The Mantel--Haenszel test for trend and independent samples Kruskal--Wallis test were used to analyze categorical and continuous variables, respectively. Multivariable binomial regression was used to estimate prevalence ratios/year, adjusting for several covariates. RESULTS: HIV prevalence decreased over time in all age groups. This was particularly evident among FSWs less than 25 years of age; HIV was 17.5% in 2008-2009, decreasing to 12.2% in 2010-2011, 8.3% in 2012-2013, 7.3% in 2014-2015, and 4.8% in 2016-2017 (P < 0.0001). Over time, FSWs reported increased condom use, particularly with regular partners, more frequent prior HIV testing, and were less likely to report a history of vaginal discharge (P < 0.0001). In adjusted analyses compared with 2008, HIV prevalence decreased in 2011 (aPR 0.64; 95% CI: 0.46-0.90), 2012 (aPR 0.58; 95% CI: 0.41-0.81), 2013 (aPR 0.53; 95% CI: 0.38-0.73), 2014 (aPR 0.48; 95% CI: 0.34-0.67), 2015 (aPR 0.50; 95% CI: 0.35-0.70), 2016 (aPR 0.40; 95% CI: 0.28-0.57), and 2017 (aPR 0.33; 95% CI: 0.22-0.50). CONCLUSION: HIV prevalence has decreased among FSW accessing SWOP in Nairobi, Kenya. This decline is consistent with the scale-up of HIV prevention and treatment efforts, both in FSWs and in the general population.

Single-lead f-wave extraction using diffusion geometry.

OBJECTIVE: A novel single-lead f-wave extraction algorithm based on the modern diffusion geometry data analysis framework is proposed. APPROACH: The algorithm is essentially an averaged beat subtraction algorithm, where the ventricular activity template is estimated by combining a newly designed metric, the 'diffusion distance', and the non-local Euclidean median based on the non-linear manifold setup. We coined the algorithm [Formula: see text]. MAIN RESULTS: Two simulation schemes are considered, and the new algorithm [Formula: see text] outperforms traditional algorithms, including the average beat subtraction, principal component analysis, and adaptive singular value cancellation, in different evaluation metrics with statistical significance. SIGNIFICANCE: The clinical potential is shown in the real Holter signal, and we introduce a new score to evaluate the performance of the algorithm.