The Private Sector and HIV/AIDS in Africa: Taking Stock of Six Years of Applied Research

Background: Until recently, little was known about the costs of the HIV/AIDS epidemic to businesses in Africa and business responses to the epidemic. This paper synthesizes the results of a set of studies conducted between 1999 and 2006 and draws conclusions about the role of the private sector in Africa’s response to AIDS. Methods: Detailed human resource, financial, and medical data were collected from 14 large private and parastatal companies in South Africa, Uganda, Kenya, Zambia, and Ethiopia. Surveys of small and medium-sized enterprises (SMEs) were conducted in South Africa, Kenya, and Zambia. Large companies’ responses or potential responses to the epidemic were investigated in South Africa, Uganda, Kenya, Zambia, and Rwanda. Results: Among the large companies, estimated workforce HIV prevalence ranged from 5%¬37%. The average cost per employee lost to AIDS varied from 0.5-5.6 times the average annual compensation of the employee affected. Labor cost increases as a result of AIDS were estimated at anywhere from 0.6%-10.8% but exceeded 3% at only 2 of 14 companies. Treatment of eligible employees with ART at a cost of $360/patient/year was shown to have positive financial returns for most but not all companies. Uptake of employer-provided testing and treatment services varied widely. Among SMEs, HIV prevalence in the workforce was estimated at 10%-26%. SME managers consistently reported low AIDS-related employee attrition, little concern about the impacts of AIDS on their companies, and relatively little interest in taking action, and fewer than half had ever discussed AIDS with their senior staff. AIDS was estimated to increase the average operating costs of small tourism companies in Zambia by less than 1%; labor cost increases in other sectors were probably smaller. Conclusions: Although there was wide variation among the firms studied, clear patterns emerged that will permit some prediction of impacts and responses in the future.

How Much Does It Cost to Provide Antiretroviral Therapy for HIV/AIDS in Africa?

Background: Many African countries are rapidly expanding HIV/AIDS treatment programs. Empirical information on the cost of delivering antiretroviral therapy (ART) for HIV/AIDS is needed for program planning and budgeting. Methods: We searched published and gray sources for estimates of the cost of providing ART in service delivery (non-research) settings in sub-Saharan Africa. Estimates were included if they were based on primary local data for input prices. Results: 17 eligible cost estimates were found. Of these, 10 were from South Africa. The cost per patient per year ranged from $396 to $2,761. It averaged approximately $850/patient/year in countries outside South Africa and $1,700/patient/year in South Africa. The most recent estimates for South Africa averaged $1,200/patient/year. Specific cost items included in the average cost per patient per year varied, making comparison across studies problematic. All estimates included the cost of antiretroviral drugs and laboratory tests, but many excluded the cost of inpatient care, treatment of opportunistic infections, and/or clinic infrastructure. Antiretroviral drugs comprised an average of one third of the cost of treatment in South Africa and one half to three quarters of the cost in other countries. Conclusions: There is very little empirical information available about the cost of providing antiretroviral therapy in non-research settings in Africa. Methods for estimating costs are inconsistent, and many estimates combine data drawn from disparate sources. Cost analysis should become a routine part of operational research on the treatment rollout in Africa.

Efficient End-Host Architecture for High Performance Communication Using User-level Sandboxing

Current low-level networking abstractions on modern operating systems are commonly implemented in the kernel to provide sufficient performance for general purpose applications. However, it is desirable for high performance applications to have more control over the networking subsystem to support optimizations for their specific needs. One approach is to allow networking services to be implemented at user-level. Unfortunately, this typically incurs costs due to scheduling overheads and unnecessary data copying via the kernel. In this paper, we describe a method to implement efficient application-specific network service extensions at user-level, that removes the cost of scheduling and provides protected access to lower-level system abstractions. We present a networking implementation that, with minor modifications to the Linux kernel, passes data between "sandboxed" extensions and the Ethernet device without copying or processing in the kernel. Using this mechanism, we put a customizable networking stack into a user-level sandbox and show how it can be used to efficiently process and forward data via proxies, or intermediate hosts, in the communication path of high performance data streams. Unlike other user-level networking implementations, our method makes no special hardware requirements to avoid unnecessary data copies. Results show that we achieve a substantial increase in throughput over comparable user-space methods using our networking stack implementation.

On the Interaction Between an Operating System and Web Server

This paper examines how and why web server performance changes as the workload at the server varies. We measure the performance of a PC acting as a standalone web server, running Apache on top of Linux. We use two important tools to understand what aspects of software architecture and implementation determine performance at the server. The first is a tool that we developed, called WebMonitor, which measures activity and resource consumption, both in the operating system and in the web server. The second is the kernel profiling facility distributed as part of Linux. We vary the workload at the server along two important dimensions: the number of clients concurrently accessing the server, and the size of the documents stored on the server. Our results quantify and show how more clients and larger files stress the web server and operating system in different and surprising ways. Our results also show the importance of fixed costs (i.e., opening and closing TCP connections, and updating the server log) in determining web server performance.

Unicast Routing: Cost-Performance Tradeoffs

The objective of unicast routing is to find a path from a source to a destination. Conventional routing has been used mainly to provide connectivity. It lacks the ability to provide any kind of service guarantees and smart usage of network resources. Improving performance is possible by being aware of both traffic characteristics and current available resources. This paper surveys a range of routing solutions, which can be categorized depending on the degree of the awareness of the algorithm: (1) QoS/Constraint-based routing solutions are aware of traffic requirements of individual connection requests; (2) Traffic-aware routing solutions assume knowledge of the location of communicating ingress-egress pairs and possibly the traffic demands among them; (3) Routing solutions that are both QoS-aware as (1) and traffic-aware as (2); (4) Best-effort solutions are oblivious to both traffic and QoS requirements, but are adaptive only to current resource availability. The best performance can be achieved by having all possible knowledge so that while finding a path for an individual flow, one can make a smart choice among feasible paths to increase the chances of supporting future requests. However, this usually comes at the cost of increased complexity and decreased scalability. In this paper, we discuss such cost-performance tradeoffs by surveying proposed heuristic solutions and hybrid approaches.

On the Cost of Supporting Multihoming and Mobility

As the Internet has evolved and grown, an increasing number of nodes (hosts or autonomous systems) have become multihomed, i.e., a node is connected to more than one network. Mobility can be viewed as a special case of multihoming—as a node moves, it unsubscribes from one network and subscribes to another, which is akin to one interface becoming inactive and another active. The current Internet architecture has been facing significant challenges in effectively dealing with multihoming (and consequently mobility). The Recursive INternet Architecture (RINA) [1] was recently proposed as a clean-slate solution to the current problems of the Internet. In this paper, we perform an average-case cost analysis to compare the multihoming / mobility support of RINA, against that of other approaches such as LISP and MobileIP. We also validate our analysis using trace-driven simulation.