BackgroundAntiretroviral Treatment (ART) significantly reduces HIV transmission. We conducted a cost-effectiveness analysis of the impact of expanded ART in South Africa.MethodsWe model a best case scenario of 90% annual HIV testing coverage in adults 15–49 years old and four ART eligibility scenarios: CD4 count <200 cells/mm3 (current practice), CD4 count <350, CD4 count <500, all CD4 levels. 2011–2050 outcomes include deaths, disability adjusted life years (DALYs), HIV infections, cost, and cost per DALY averted. Service and ART costs reflect South African data and international generic prices. ART reduces transmission by 92%. We conducted sensitivity analyses.ResultsExpanding ART to CD4 count <350 cells/mm3 prevents an estimated 265,000 (17%) and 1.3 million (15%) new HIV infections over 5 and 40 years, respectively. Cumulative deaths decline 15%, from 12.5 to 10.6 million; DALYs by 14% from 109 to 93 million over 40 years. Costs drop $504 million over 5 years and $3.9 billion over 40 years with breakeven by 2013. Compared with the current scenario, expanding to <500 prevents an additional 585,000 and 3 million new HIV infections over 5 and 40 years, respectively. Expanding to all CD4 levels decreases HIV infections by 3.3 million (45%) and costs by $10 billion over 40 years, with breakeven by 2023. By 2050, using higher ART and monitoring costs, all CD4 levels saves $0.6 billion versus current; other ART scenarios cost $9–194 per DALY averted. If ART reduces transmission by 99%, savings from all CD4 levels reach $17.5 billion. Sensitivity analyses suggest that poor retention and predominant acute phase transmission reduce DALYs averted by 26% and savings by 7%.ConclusionIncreasing the provision of ART to <350 cells/mm3 may significantly reduce costs while reducing the HIV burden. Feasibility including HIV testing and ART uptake, retention, and adherence should be evaluated.

BackgroundIntegrated disease prevention in low resource settings can increase coverage, equity and efficiency in controlling high burden infectious diseases. A public-private partnership with the Ministry of Health, CDC, Vestergaard Frandsen and CHF International implemented a one-week integrated multi-disease prevention campaign.MethodResidents of Lurambi, Western Kenya were eligible for participation. The aim was to offer services to at least 80% of those aged 15–49. 31 temporary sites in strategically dispersed locations offered: HIV counseling and testing, 60 male condoms, an insecticide-treated bednet, a household water filter for women or an individual filter for men, and for those testing positive, a 3-month supply of cotrimoxazole and referral for follow-up care and treatment.FindingsOver 7 days, 47,311 people attended the campaign with a 96% uptake of the multi-disease preventive package. Of these, 99.7% were tested for HIV (87% in the target 15–49 age group); 80% had previously never tested. 4% of those tested were positive, 61% were women (5% of women and 3% of men), 6% had median CD4 counts of 541 cell/µL (IQR; 356, 754). 386 certified counselors attended to an average 17 participants per day, consistent with recommended national figures for mass campaigns. Among women, HIV infection varied by age, and was more likely with an ended marriage (e.g. widowed vs. never married, OR.3.91; 95% CI. 2.87–5.34), and lack of occupation. In men, quantitatively stronger relationships were found (e.g. widowed vs. never married, OR.7.0; 95% CI. 3.5–13.9). Always using condoms with a non-steady partner was more common among HIV-infected women participants who knew their status compared to those who did not (OR.5.4 95% CI. 2.3–12.8).ConclusionThrough integrated campaigns it is feasible to efficiently cover large proportions of eligible adults in rural underserved communities with multiple disease preventive services simultaneously achieving various national and international health development goals.

BackgroundDelivery of community-based prevention services for HIV, malaria, and diarrhea is a major priority and challenge in rural Africa. Integrated delivery campaigns may offer a mechanism to achieve high coverage and efficiency.MethodsWe quantified the resources and costs to implement a large-scale integrated prevention campaign in Lurambi Division, Western Province, Kenya that reached 47,133 individuals (and 83% of eligible adults) in 7 days. The campaign provided HIV testing, condoms, and prevention education materials; a long-lasting insecticide-treated bed net; and a water filter. Data were obtained primarily from logistical and expenditure data maintained by implementing partners. We estimated the projected cost of a Scaled-Up Replication (SUR), assuming reliance on local managers, potential efficiencies of scale, and other adjustments.ResultsThe cost per person served was $41.66 for the initial campaign and was projected at $31.98 for the SUR. The SUR cost included 67% for commodities (mainly water filters and bed nets) and 20% for personnel. The SUR projected unit cost per person served, by disease, was $6.27 for malaria (nets and training), $15.80 for diarrhea (filters and training), and $9.91 for HIV (test kits, counseling, condoms, and CD4 testing at each site).ConclusionsA large-scale, rapidly implemented, integrated health campaign provided services to 80% of a rural Kenyan population with relatively low cost. Scaling up this design may provide similar services to larger populations at lower cost per person.

No abstract

Caching popular contents in advance is an important technique to achieve low latency and reduce the backhaul costs in future wireless communications. Considering a network with base stations distributed as a Poisson point process, optimal content placement caching probabilities are obtained to maximize the average success probability (ASP) for a known content popularity (CP) profile, which in practice is time-varying and unknown in advance. In this paper, we first propose two online prediction (OP) methods for forecasting CP viz., popularity prediction model (PPM) and Grassmannian prediction model (GPM), where the unconstrained coefficients for linear prediction are obtained by solving constrained non-negative least squares. To reduce the higher computational complexity per online round, two online learning (OL) approaches viz., weighted-follow-theleader and weighted-follow-the-regularized-leader are proposed, inspired by the OP models. In OP, ASP difference (i.e, the gap between the ASP achieved by prediction and that by known content popularity) is bounded, while in OL, sub-linear MSE regret and linear ASP regret bounds are obtained. With MovieLens dataset, simulations verify that OP methods are better for MSE and ASP difference minimization, while the OL approaches perform well for the minimization of the MSE and ASP regrets.