Track E Implementation Science, Health Systems and Economics

Due to Lesotho's high adult HIV prevalence (23%), considerable resources have been allocated to the HIV/AIDS response, while resources for non‐communicable diseases have lagged. Since November 2011, the Elizabeth Glaser Pediatric AIDS Foundation (EGPAF) has supported Lesotho Ministry of Health to roll out Family Health Days (FHDs), an innovative strategy to increase community access to integrated health services, with a focus on hard‐to‐reach areas where immunization coverage, HIV service uptake, and screening and treatment for chronic diseases are low.

E3 -Effect of HIV funding/programming on sexual and reproductive health (SRH) services WEPDE0101 Managing change in performance-based project funding: a case study of PPFN Rapid Emergency Scale Up Plan to increase HCT/FP/STIs integrated services Background: While policy and implementation support for SRHR/HIV integration is increasing, significant questions and uncertainties remain about what such programming means in practice. This is particularly the case in concentrated HIV epidemics, where little is still known about what integration should look like for key populations, including sex workers, men who have sex with men (MSM), transgenders, injection drug users (IDUs) and people living with HIV (PLHIV). While integration is a desirable goal in the long-run particularly for clinical services, joining programs and systems that are not ready could compromise quality and access for these groups that already face difficulty in obtaining appropriate services for both HIV and SRH needs. Methods: India HIV/AIDS Alliance undertook a global review of over 160 resources available on the websites of selected national and international organisations, including NGOs, technical support agencies and UN agencies. The resources included case studies, mappings, toolkits, policy briefings and program reports. Results: The review identified the most common challenges in designing and implementing SRH/HIV integrated programs for key populations. These included stigma and discrimination, low levels of demand, lack of rights-based approaches, low attention to gender inequality, low understanding of key populations' specific SRH needs, lack of capacity and sensitivity among service providers, lack of strong referral systems and inadequate resources for additional interventions. The review highlighted key steps that organisations can take to successfully integrate SRHR and HIV in their responses for key populations. Conclusion: While SRH/HIV integrated programs present an important opportunity to respond to the unmet needs of key populations, integration that is premature, overly rapid or too large-scale risks compromising rather than enhancing key populations' access to high quality HIV and SRH services. Good practice principles, including gender equality, human rights-based approaches, meaningful involvement of communities, for work with key populations are particularly critical in effective HIV/SRHR integration.
E8 -Task-shifting MOPDE0103 Implementation of VMMC efficiency elements in four sub-Saharan countries: service delivery methods and provider attitudes with VMMC providers and the in-charge officer, and compilation of service statistics.
. Surgical method . Task shifting (allowing non-physicians to perform VMMC) . Task-sharing (allowing non-physicians to conduct aspects of VMMC) . Rotation among multiple bays in the operating theater . Bundling of supplies and tools . Use of electrocautery instead of ligating sutures Results: The results are useful for monitoring service delivery in each country and conducting cross country comparisons. The data shows considerable variation by country on 5 of the 6 elements, the exception being nearly universal use of the forceps-guided method (see Table 1). The study revealed stark differences on task-shifting, both in terms of actual practice and provider attitudes toward it. For example, in Zimbabwe, while no providers report using task-shifting, 86.5% say they would implement the practice given the choice. Countries also differed in the use of multiple beds per provider (rotation) during VMMC, with100% of providers in Zimbabwe reported using multiple beds, compared to 38% percent in Kenya, reflecting a different service delivery model. Provider reported use of electrocautery ranged from 5.4% in Tanzania to 88.6% in South Africa. Conclusion: The decision to adopt these elements is generally based on national policies. However, the review of these practices across four countries demonstrates alternative methods of service delivery, and is intended to incite potential changes to these policies in the name of increasing efficiency. . Surgical method (e.g., forceps-guided) . Task shifting (allowing non-physicians to perform VMMC) . Task-sharing (allowing non-physicians to conduct aspects of VMMC) . Rotation among multiple bays in the operating theater . Bundling of supplies and tools . Use of electrocautery instead of ligating sutures Methods: Data collection took place at 14Á30 VMMC sites per country (73 sites total) from April-December 2011. It included observation of the clinical facilities, observation of VMMC procedures, interviews with VMMC providers and the in-charge officer, and compilation of service statistics. A total of 357 providers reported on their attitudes towards their jobs and on the pervasiveness of burnout in the field of VMMC. Results: The results are useful for monitoring service delivery in each country and conducting cross country comparisons. The data shows varied levels of burnout and job satisfaction among providers across countries (see Table 1 and Table 2). The various survey instruments allow for an analysis of which program, site and provider level factors may contribute to increased  burnout. This paper presents results on: type of site (mobile/ static/ outreach), site volume, provider cadre, education and additional training, time in field, full/part-time status, and variety of tasks performed. For example, preliminary analyses from Zimbabwe indicated that physicians experience higher levels of burnout than their nurse colleagues. Conclusion: The review of these practices across the four countries will shed light on alternative methods of staffing and service delivery that contribute to lower levels of burnout among VMMC providers. The results could inform policies and program planning for VMMC as well as other clinical HIV services.

MOPDE0106
A comparative analysis of two high-volume male medical circumcision (MMC) operational models with similar service delivery outcomes in different settings within Gauteng and KwaZulu-Natal provinces in South Africa: urban Centre for HIV/AIDS Prevention Studies (CHAPS) versus rural-SACTWU Worker Health Program (SWHP) N. Soboil 1 , J. Cockburn 1 , D. Rech 2 and D. Taljaard 2 1 SACTWU, Worker Health Program, Cape Town, South Africa. 2 CHAPS, Johannesburg, South Africa Presenting author email: nikki@swtzn.co.za Background: In 2010 South Africa initiated rapid scale up of MMC services nationally. In support of the service delivery, CHAPS operates a ''specialized fixed'' VMMC programme in a large urban setting in Johannesburg, Gauteng. SWHP operates a ''roving'' VMMC programme in the rural district of Uthukela, KwaZulu-Natal. Both programmes have achieved similar uptake of MMC services using different operational models, with each achieving the highest number of MMCs nationally within their applicable settings.
Methods: A comparative analysis was conducted of CHAPS' ''fixed site'' and SWHP's ''roving team'' over a one year period to analyse unique attributes of each, namely: social mobilisation; staff and capital resources; extent of partnerships with district Departments of Health (DOH), service quality, and cost comparisons between models.
Results: Social mobilisation activities were similar but customised to address local customs and preferences. There was variation in staffing and capital expenditure due to CHAPS' initial capital outlay for a fixed site and its permanent team of 27 versus SWHP's skeleton team of 12 with no investment in facilities. Both programmes have strong partnerships with their district DOHs, although SWHP benefited from more support in terms of interim staffing and provision of surgical consumables. This contributed to variances in expenditure and cost per MMC by programme. Adverse events and service quality were comparable, but roving services required greater time and resources in tracking client follow up than fixed services. Table 1.
Conclusion: This comparative analysis shows models need to be customised to address the requirements of each target population group and their geographic settings. In urban settings, fixed sites, drawing on large dense surrounding populations, maintain good daily numbers and follow up rates. In rural settings, the roving model, with significant government support, is economical and effective in reaching targets and covering communities spanning large areas. Background: Evidence is mixed on the benefit of medical male circumcision (MMC) for women. Questions remain whether there is direct benefit in reducing HIV acquisition among women; whether men put women at risk due to early resumption of sex; whether mass rollout of MMC leads to risk compensation in ways that compromise women's ability to negotiate condom use; and whether MMC increases patriarchal behaviors such as gender-based violence. We present data from various studies to address these questions and make the case for women's shareholding position in MMC. Methods: We reviewed over 30 published and ongoing studies on MMC for evidence on early resumption of sex, risk compensation, benefits of MMC to women, and women's views of and reactions to MMC rollout in different African countries.
Results: About 10 studies have returned the verdict of no risk compensation among circumcised men. In addition, there are few but scientifically solid evidence emerging showing up to 47% direct benefit of MMC in reducing HIV transmission to women, particularly when sex is not resumed early. However, four studies show consistent evidence of early resumption of sex before wound healing, ranging from 24-42% overall, 63-76% among married/ cohabiting men, and 12-48% among HIV-positive men. About 15 studies indicate that women support MMC though misconceptions are still common, with up to 77% unaware of the need for sexual abstinence. Some believe that scale-up of MMC translates into reduced risk hence less need for safe sex.
Conclusion: With the current scale up of MMC for HIV prevention in 14 African countries, there is no evidence of risk compensation; however, early resumption of sex and women's low risk perception because their partners are circumcised may undermine the benefits of MMC to women. MMC programs should target women with correct and complete information and advocate for their involvement in partners' decisions on circumcision.
E15 -Managing HIV supply chain challenges with limited resources THPDE0101 Lessons from the rapid response of Namibia's supply chain when antiretroviral treatment guidelines changed to tenofovir-based first-line regimen MoHSS enlisted the support of the Supply Chain Management System (SCMS) to analyse the supply chain implications of the changes and provide support for a smooth transition. Methods: Using MS Excel (r) spreadsheets, SCMS modelled implications of increasing the CD4 threshold for ART initiation and changing to tenofovir-based first line regimens. Backed by Namibian Government funding commitment, CMS accelerated internal procurement processes to avail tenofovir-based formulations, reviewed stock control parameters to manage reduced demand for stavudine and expanded ARV storage capacity to cope with the increased volumes of ARVs. Monitoring systems were intensified including quarterly tracking of trends in consumption and treatment regimens. Results: Within 8 months, CMS had managed to accumulate adequate quantities of tenofovir formulations, clearing the way for the national roll-out of the new ART guidelines starting September 2010. As a result, over 34% of the approximately 90,000 adults on ART were taking tenofovir-based regimens one year later. In addition, a potential wastage of about US$ 950,000 worth of stavudine formulations was prevented and stock outs of tenofovir formulations avoided.
Interventions to optimise existing storage space at CMS created an additional 328 pallet locations, a 74% increase in racked storage capacity in five rooms for ARVs and general pharmaceuticals. Conclusion: A rapid response was achievable due to the Namibian government's commitment and the agility of the supply chain. Timely availability of information through robust monitoring systems established to capture changing consumption patterns informed procurement planning and minimised the risks associated with change in ART guidelines. Background: Crisis followed the October 2010 Presidential elections in Cote d'Ivoire. International sanctions were applied, the banking system collapsed, security deteriorated and internal displacement prevented access to healthcare. SCMS, a PEPFAR program administered by USAID, adapted the national HIV/AIDS supply chain strategies to respond to changing conditions and to avoid treatment interruption Methods: From the outset, SCMS established a crisis-committee whose daily calls between the field and Washington focused on information sharing with the Ministry of Health (MOH), Mission, USG team, Implementing Partners (IPs) and international donors. SCMS jointly reviewed supply plans and stock-status reports with international donors and IPs and adjusted planned orders to meet urgent national demands. Under threat of air, land and sea border closings, SCMS re-routed all deliveries though its' Regional Distribution Center (RDC) in Accra, Ghana. The RDCs, in turn, pulsed commodities into the country as demand required and security permitted. Meanwhile, in country, the MOH provided patients with two months treatment instead of the usual one. The distribution mechanisms, however, were disrupted and SCMS together with IPs created an interim distribution plan in which IPs picked up their products and distributed them to sites. Results: Deliveries continued throughout the crisis; no health facility experienced a stockout of critical products and treatment interrup-tion was averted. Donor collaboration ensured planned deliveries met patient needs. The RDC in neighboring Ghana provided flexibility to maximize in-country stock levels, arrange last-minute flights and make urgent deliveries. On-the-ground collaboration with the MOH and IPs ensured medicines were not only available at service delivery points, but in the patients' hand.
Conclusion: By creatively leveraging in-country knowledge, regional resources and procurement volumes of supplies, ARVs and other critical products reached patients throughout the crisis.

THPDE0103
Use of simple, low cost innovations to improve availability of HIV-related commodities at public hospitals in Uganda Background: Currently, HIV/AIDS care support from donors is shifting from direct service delivery to systems strengthening of countryowned programs. However, frequent stock-outs of HIV supplies occur, as, in many cases, the systems for supply of HIV-and non-HIVrelated commodities remain separate, and health facilities do not order for HIV commodities when required. The Strengthening Uganda's Systems for Treating AIDS Nationally (SUSTAIN) Project worked closely with and supported 17 public hospitals in Uganda to harmonize ordering for commodities from the governments' National Medical Stores to improve availability and service delivery. Methods: SUSTAIN utilized four low-cost interventions to strengthen supply chain management systems: identification of focal persons responsible for submission of commodity orders to the national stores; training of the order focal persons in ART logistics management followed by onsite coaching and mentoring; telephone text message reminders to the focal persons one week prior to order deadlines; and development of an order checklist to harmonize submission of HIV-and non-HIV-related commodities. Results: From January to October 2011, average ordering rates for all commodities improved from 46% to 71%. Improvements were noted in order submission rates for HIV-related commodities: from 29% to 81% for ARVs, 24% to 69% for PMTCT commodities, and 22% to 63% for HIV test kits. 54% of facilities harmonized submission of orders for all commodities, an improvement from 12% in January 2011. Improved submission of orders resulted in increased availability of HIV-related commodities and quality of care. Conclusion: Improving availability of commodities and supplies available through the national system can be achieved by supporting hospitals to utilize low cost, simple innovations to improve ordering rates.

THPDE0105
Overcoming challenges in supply chain management amidst rapid scale-up of antiretroviral services Background: Fragile health commodity supply chain systems have long been a factor inhibiting universal access to care and treatment in ART programs. Weaknesses in the various components of the logistics cycle contribute to non-dependable and irregular supply of quality products at service delivery points which causes poor healthseeking behaviour, treatment interruption and poor adherence and therefore treatment failure and resistance. Methods: In June 2010, STAR-EC, a five year USAID-funded project implemented by JSI, initiated training and mentorship of health workers in pediatric ART, logistics management of opportunistic infection medicines and stores management to build capacity for quantification. District focal persons were facilitated to support the sites to submit orders to the national suppliers. Storage was improved through the provision of medicine cabinets and pallets. Destruction of expired commodities was done in high priority health centres to generate space for increasing quantities of products. STAR-EC later supported the collection of these supplies from the facilities to the district stores for destruction by MoH. Data management was improved through the printing and distribution of logistics management information tools that encouraged the use of more optimal formulations. During periods of stock outs, buffer commodities were provided and collaboration with other partners mitigated non-availability at facilities. Consistent communication with national stores to determine availability of supplies and support to facilities to make and retrieve emergency orders was provided.
Results: The ordering rate for ARVs increased from 46% to 96% and was maintained as shown in Figure 1. Consequently, the availability of products improved which led to increase in number of clients enrolled onto ART as shown in the Figure 2.   Background: For a small NGO supporting a community health worker program in rural villages, ensuring accurate timely data collection is difficult. Paper based systems are time-consuming to collect and often provide incomplete data. A simple, cost-effective monitoring system using mobile phones can speed up data collection and provide better data. This project aimed to use electronic reporting to better capture the quality of support (social, family, personal) provided by community health workers.
Methods: The project adapted paper-based monthly reporting tools already in use for implementation on the mobile phone platform using open access Frontline software. 38 CHWs (21% of the total number) were selected from a diverse set of locations. Staff built buy in from participating CHWs for the new technology and trained them on the system. Data was collected through SMS to a central database. CHWs received an initial sum of phone credit and were reimbursed based on the number of messages received, regardless of accuracy. To assess the comparative benefit of the system, SIC measured the collection and accuracy of CHWs using paper surveys.
Results: One month after initiation, participating CHWs had sent monthly patient information by text for 37.2% of their clients with 9.8% of messages having data errors. For paper, at one month, the percentage of monthly patient reports collected was 6%. At three months, a higher percentage of paper reports (78.7%) had been collected than electronic (45.0%). Paper surveys were two and a half times as likely to have missing information. Conclusion: The mobile phone system has a clear advantage in timeliness of data. Over time, however, consistency with mobile phone reporting declined because of difficulties with reimbursements and challenges with system maintenance Next steps are adjusting the reimbursement system and evaluating whether the system is better suited for use in ongoing reporting, annual evaluation or not at all.

WEAE0304
Mobile telemedicine for improved community-level clinical decision making, referrals and medical information transmission and storage: a pilot study in Nairobi, Kenya Background: Shortages of health professionals in low-income countries make task shifting to community health workers (CHWs) crucial to AIDS care and treatment. Multiple studies of CHW utilization reveal poor linkages to the formal health sector, lack of clinical decision making support in the field and weak integration of information gathered at the community level with the national health information system. Methods: A pilot study was performed at a public HIV clinic in Nairobi that provided CHWs with a mobile telephone system (ClinipakMobile) loaded with surveys that support appropriate clinical decision making and link to an electronic patient information database. Each CHW underwent a one-day training on use of ClinipakMobile and used ClinipakMobile at home visits with clients living with HIV. Variables assessed were: ability of the CHW to use ClinipakMobile based on standardized role-play assessments, CHW satisfaction surveys, ability of the server to store clinical data, and patient-based outcomes assessed by an end-of-study chart review of CHW referrals. Results: Seventeen CHWs participated. The CHWs quickly learned and retained skills in using ClinipakMobile across three time points (week 0, 3 and 6) and expressed great satisfaction. The database maintained function and integrity with only one down-time period of 24 hours. CHW patient interviews using ClinipakMobile recorded multiple ''red flag'' answers for antiretroviral non-adherence (21%), side effects (30%) and opportunistic infection symptoms (33%) resulting in 30% of patients receiving ClinipakMobile-recommended referrals to the health facility. The chart review revealed these problems were infrequently noted by clinicians at follow-up visits. Conclusion: ClinipakMobile is an easy-to-use mhealth tool to improve CHW functioning to care for PLWHA. Busy clinicians may not note important clinical ''red flag'' that CHWs record at home visits. In the future, patient information gathered on ClinipakMobile should be downloaded from the server daily and given directly to clinicians to improve care during clinic visits.

FRAE0104
An automated platform for delivery of CD4 results to SMSenabled antenatal clinic printers, Botswana Background: CD4 determination and is a key step in the prevention of mother-to-child transmission cascade. However, timely access to centrally-processed CD4 results remains a challenge for antenatal clinics in resource-limited settings. Methods: As part of an interventional study to improve antenatal access to HAART, we developed an automated platform to wirelessly communicate CD4 results to portable SMS-enabled thermal printers (iBacsTel) in referral antenatal clinics. The system, authored in Python over a MySQL database using Django, directly collects results from flow cytometer (FACSCalibur) output files and integrates demographics and validation status from electronic records. The system then sends validated results via a SMS gateway (Kannel) to printers in referring antenatal clinics. Clinic receipt is confirmed centrally via SMS. A monitoring platform (see Figure 1) permits quick central assessment of clinic printer status, laboratory delays, deviations from expected testing patterns (e.g. due to supply outages, staffing shortages), and identification of women with low CD4 counts. We deployed the system in 16 clinics around Gaborone, Botswana (2 clinics added per month) between August 2011 and February 2012. Results: Antenatal CD4 results for 367 women have been transmitted and confirmed, including 92 women qualifying for HAART (CD4 B250 cells/mL). Pre-intervention delivery, largely via hand delivery of printed results, was estimated at 21 days. After an initial adaptation phase, the mean number of days from phlebotomy to confirmed clinic result receipt improved to 3.0 days (95% CI 2.1Á3.9 days, Figure 2). Improvements in both laboratory and results transmission times have resulted from real-time identification of laboratory delays, software modifications, and preemptive management of printer airtime periods. Clinicians report satisfaction, noting substantial improvements in reliability and timeliness. Conclusion: Direct transmission of CD4 results to antenatal practitioners is feasible and reduces time to result receipt. The system permits central monitoring of clinic and lab performance and prioritization of pregnant women with low CD4.

TULBE02
Implementation of a wireless GPRS-based monitoring system for point-of-care CD4 testing at rural primary health facilities in Mozambique   Results: Data from the twelve POC CD4 devices representing a total of 336 testing days and 3,109 CD4 tests (with 620 quality control runs) were aggregated and reported by the interface. The system identified 376 error events, including 19 quality control failures (Levey-Jennings !2 standard deviations), 4 failures to run daily controls before testing, 45 operator test errors, 6 possible device malfunction events, and 21 instances of attempted use of expired reagents. These events prompted follow-up corrective action by supervisors more quickly than with previous paper-based reporting methods.
Conclusion: This is the first description of a remote wireless monitoring system to track performance of POC diagnostic tests for HIV in resource-limited settings. This GPRS-based system can improve the management of POC diagnostics by capturing real-time events that might otherwise go unnoticed. This enables detailed management of the quality and performance of POC CD4 testing in remote areas and rapid response to resolve problems and provide

Questions Responses
Why were you given a phone? (50 of the 70 women who were given a phone were contactable by the donated phone)

Track E Implementation Science, Health Systems and Economics
Conclusion: The distribution of cell phones to families of HIV-positive children identified with the EID program in Haiti has helped our teams monitor the progress of children. Approximately 20% of PCR positive children identified are lost to follow up within the first year and contact by phone can help. Phones are available for approximately 20USD in Haiti; the program is not prohibitively expensive. 30% of the phones were non-functional/stolen within 21 months of the project. The use of mobile phones, combined with an active community tracing program has benefitted the Haiti EID program.

MOPDE0203
Risk factors for late-stage HIV disease presentation at initial HIV diagnosis in Durban, South Africa Background: Despite expanded access to HIV testing, most newlydiagnosed South Africans present with severe immunosuppression. We sought to determine the risk factors for presentation with latestage HIV disease and the perceived barriers to presenting earlier for care. Methods: We enrolled and surveyed adults prior to HIV testing at four outpatient clinics in urban and peri-urban areas of Durban from August 2010 to November 2011. Late-stage HIV disease was defined as a CD4 count B100 cells/mm 3 . We used multivariate logistic regression models to determine the effects of sex, emotional health, social support, distance from the clinic, employment, perceived barriers to receiving healthcare, and foregoing healthcare to use money for food, clothing, or housing (''competing needs to healthcare'') on presentation for care with late-stage HIV disease. Results: Among 3,669 adults screened, 830 (22.6%) were enrolled, newly-diagnosed HIV-infected, and obtained a CD4 result. Among those, 279 (33.6%) presented with late-stage HIV disease. In multivariate analyses, participants who lived 05 kilometers away (OR 2.8, 95% CI 1.7Á4.7), had reported competing needs to healthcare (OR 1.7, 95% CI 1.2Á2.4), were male (OR 1.7, 95% CI 1.2Á2.3), worked outside the home (OR 1.5, 95% CI 1.1Á2.1), perceived health service delivery barriers (OR 1.5, 95% CI 1.1Á2.1), and had poor emotional health (OR 1.4, 95% CI 1.0Á1.9) had higher odds of late-stage HIV disease presentation. Conclusion: In Durban, South Africa, the strongest independent risk factors for presentation with late-stage HIV disease were living further from the clinic, being male, and having competing needs to healthcare. Self-reported barriers related to personal illness, costs of care, and poor perceived service delivery were also significantly associated with late-stage disease presentation. Future studies should examine whether use of mobile units and financial assistance may reduce presentation with late-stage HIV disease in resourcepoor settings.

THPDE0301
Towards universal awareness of HIV status: a systematic review on uptake of home-based HIV testing in sub-Saharan Africa Background: Improving access to HIV testing is recognized as a key priority in scaling up HIV treatment and prevention services. Homebased testing (HBT) as an approach to delivering wide-scale HIV testing is explored in this study. Methods: A systematic review and random effects meta-analysis of published studies reporting on uptake of HBT in sub-Saharan Africa since 2000 were conducted to assess the proportion of individuals accepting HBT and receiving their test result. Three electronic databases were searched. Results: Our initial search yielded 1199 articles, 114 were reviewed as full-text articles and 19 publications involving 21 studies (N 0524,787 offered HBT) were included for final review. The studies came from 5 countries: Uganda, Malawi, Kenya, South Africa, Zambia.

Track E Implementation Science, Health Systems and Economics
The proportion of people who accepted HBT (N0474,516) ranged from 58.1% to 100%, with a pooled proportion of 83.8% (95% CI: 80.9Á86.6%) ( Figure 1). Heterogeneity was high (t 2 0.13). Sixteen studies reported on the number of people who received the result of HBT (N 0432,835). The proportion of individuals receiving their results out of all those offered testing ranged from 24.9% to 99.7% with a pooled proportion of 77.4% (95% CI: 74.0Á80.7%), (t 2 0.12) ( Figure 2). HIV prevalence ranged from 2.9%Á36.5%. New diagnosis of HIV following HBT ranged from 40Á79% of those testing positive. Forty-eight percent of those offered testing were men and they were just as likely to accept HBT as women (pooled odds ratio 0.84 (95% CI: 0.56Á1.26)(t 2 0.33). The proportion of individuals previously tested for HIV among those offered a test ranged from 5Á66%. No evidence was reported of negative consequences of HBT. Conclusion: Home based voluntary counselling and testing has the potential to dramatically increase awareness of HIV status in previously undiagnosed men and women in sub-Saharan Africa. HBT is a gateway to accessing care early and the benefits for individual and public health, both for treatment and prevention, make it an invaluable tool in the fight against HIV.

THPDE0302
Thinking outside the docs: expanding access to HIV testing services through the delivery of HIV testing at the Department of Motor Vehicles in Washington, D.C. Presenting author email: afulwood@fmcsinc.org Background: Washington DC has an estimated HIV prevalence rate of 3.2% and only 50% may be aware of their infection. Using large volume public service venues may promote HIV testing and directly increase access to HIV testing, thereby increasing the number of residents who know their HIV status. Family and Medical Counseling Service, Inc. (FMCS) implemented a novel program to provide HIV testing at the Department of Motor Vehicles (DMV) which provides driver's license and automobile tag services to over 150,000 residents annually. Methods: Dedicated project staff discuss the importance of routine HIV testing and offer the test to everyone awaiting DMV services. Rapid HIV testing is conducted in a private office inside the DMV, and all who test positive are immediately referred to care and support services. We present data describing the HIV testing outcomes from the program. We also asked a sample of 406 persons who didn't test about their reason for refusal. Results: From October 2010 to February 2012, 108,188 individuals were offered an HIV test, 8,152 (8%) accepted, 6,788 (83%) were tested, and 34 (0.5%) were positive. The most common reasons for refusal were ''already had a test in the past six months'' (138/406, 30%) and ''do not feel I'm at risk'' (76/406, 17%). Twenty-three percent (n 01,566) of those tested had never been tested before; the majority of whom were aged 13Á24. Conclusion: Conducting HIV testing in high volume non-clinical settings, such as the DMV, is a feasible strategy to engage individuals in HIV counseling and testing services, including those who have never tested before. Expansion of this program model to similar public service sites may be necessary to increase access to HIV testing services, encourage routine screening and increase the percentage of individuals in the general population who know their HIV status.

THPDE0304
The national HIV counselling and testing campaign and treatment expansion in South Africa: a return on investments in combination prevention T. Mbengashe 1 , Z. Nevhutalu 2 , M. Chipimo 3 , T. Chidarikire 1 and L. Diseko 1 1 National Department of Health, HIV Cluster, Pretoria, South Africa. 2 South Africa National AIDS Council, Pretoria, South Africa. 3

UNAIDS, Pretoria Country Office, Pretoria, South Africa
Background: South Africa has an estimated 5.6 million people living with HIV, representing a quarter of the disease burden in sub-Sahara Africa and one fifth of the global disease burden. The National HCT Campaign was launched in April 2010 and ended in June 2011, coordinated by the National Nerve Centre. Methods: The Campaign was launched at national, provincial and district levels and targeted everyone between the ages of 12Á60 years, not exclusive of younger or older persons. Confidentiality and informed consent was required to test and every site was linked to a referral facility providing ART, care and support. HCT was done at health facilities, work places and community outreach sites. Results: South Africa committed a significant budget of US$3.5 billion to support the HCT Campaign and Treatment Expansion. By June 2011 a record 13,269,746 HIV tests were conducted and 2,155,312 (16%) people tested positive of whom 48% had CD4 counts above 350. Over 400,000 patients were initiated on ART, of whom 57,000 were pregnant women. National Laboratory DNA PCR results indicated that mother to child transmission rates at 8 weeks reduced to 5% compared to 11% in 2009. Over 8 million Track E Implementation Science, Health Systems and Economics people were screened for TB and 185 million male condoms and 524,000 female condoms were distributed. 237 males were medically circumcised, exceeding the Campaign target of 100,000. The procurement system achieved a 53% ARV price reduction through competitive tenders and 3,686 (80%) health facilities were capacitated to deliver ART through task shifting and training of 10,542 nurses. Conclusion: South Africa has succeeded in implementing a massive HCT campaign linked to a comprehensive HIV prevention model that optimizes treatment for prevention and reduction of new HIV infections. The Campaign created a shift from testing for treatment to testing for prevention and promoted combination prevention to turn the AIDS epidemic tide for South Africa. Background: Recent estimates indicate that 40% of the HIV infections in the Netherlands are undiagnosed. Earlier diagnosis of HIV will improve individual health outcomes and reduce further spread. Self tests for HIV have become available, which allow individuals to test at home without involvement of health care professionals. Although self tests may help to increase HIV test uptake, there are concerns about test quality, counseling and medical follow up. The Amsterdam Public Health Service has initiated the HivTest@Home project to develop and evaluate a service that provides reliable HIV self tests using oral fluid in combination with an Internet counseling strategy for individuals at high risk for HIV, especially MSM and migrants from HIV endemic countries.
Methods: A website and logistic infrastructure will be developed that provides an intake with personal advice about HIV self testing, and enable individuals to purchase the HIV self test we provide online. With the test, they will receive a code to access a pre-test trajectory including step-by-step instructions, counseling, and low-threshold contact options with health care professionals. For those who test positive, a follow-up procedure will be developed to motivate them to access regular health care for confirmation testing and referral to an HIV clinic. Results: The service will be launched in 2013, accompanied by a media campaign targeting high risk groups. We aim to distribute 2,000 tests within a 12-month period, assuming an HIV prevalence of newly diagnosed individuals of 2.5-5.0%. Using data collected online from participants, web statistics, interviews with participants, and clinical follow-up data, we will evaluate the feasibility and acceptability of the service and its effectiveness in identifying undiagnosed HIV infections. Conclusion: If feasible and effective this new method can create another reliable low-threshold testing option for different risk populations -an option hardly needed to break the cycle of lowtesting numbers.
E26 -Co-designing and implementing programmes at national and regional scale TULBE03 HIV treatment as prevention: driving health system development and improved health outcomes in British Columbia

Track E Implementation Science, Health Systems and Economics
Ministry of Health, Victoria, Canada Presenting author email: ciro.panessa@gov.bc.ca Background: Jurisdictions use a mix of horizontal and vertical approaches to achieve health system reform. Horizontal approaches like British Columbia's Innovation and Change Agenda aim to strengthen the overall health system, while vertical strategies, supported by specialized provincial structures, address specific health issues. Both approaches have drawbacks, so public health leaders such as Jaime Sepú lveda and Julio Frenk have described a third way, ''the diagonal approach'': explicit intervention priorities, with cost effectiveness and improved reach and engagement embedded as foundational principles, can generate positive disease-specific outcomes for individuals and populations, while dealing with generic health system issues such as human resource development and quality assurance. Methods: BC is using an explicit HIV Treatment as Prevention (TasP) pilot, Seek and Treat to Optimally Prevent HIV (STOP HIV), as a diagonal approach to drive innovation and strategic change in BC's health system. Emphasis has been placed on several outcomes: improved screening; increased diagnoses due to improved screening; and increased linkage to care/treatment. Several discrete initiatives have been deployed to enhance health system capacity to reach (''Seek'') and sustainably engage (''Treat'') vulnerable groups in an overall program of care. Results: STOP HIV is driving health system change. Latest data is confirming that British Columbians' in the pilot areas are being better engaged leading to increased HIV testing, diagnoses, treatment and reduced AIDS mortality*see Figures [1Á3]. Conclusion: STOP HIV is a real-world implementation of HIV TasP, aimed at reducing HIV/AIDS-related morbidity and mortality, as well as new HIV diagnoses, and as a result containing associated health system costs. The initiative has become a driver of diagonal health care reform*and by engaging the most vulnerable and hard to reach, contributes to reducing health inequalities.  Track E Implementation Science, Health Systems and Economics WEPDE0105 'Less is more': assessing the effectiveness of a HIV prevention program for long distance trucker drivers in India after a re-designing in implementation structure V.R. Tirumalasetti 1 , B. Mahapatra 2 , S. Juneja 1 and I. Singh 1 1 TCI Foundation, HIV/AIDS Communicatios, Delhi, India. 2 Population Council, New Delhi, India Presenting author email: vasudha.tcif@tcil.com Background: A large scale HIV prevention programs among long distance trucker drivers was started in 2003 across 36 Transhipment locations (TSLs) alongside the national highways. Program was re-designed in the year 2006 by reducing number intervention locations to 17 high performing TSLs. The current research assesses the effectiveness of program after the re-design process. Methods: Program monitoring data over seven years and two rounds of cross-sectional behavioral survey conducted in January 2008 (N 01402) and July 2009 (N 01407) were used. Indicators on program outreach and service utilization were examined. Multivariate logistic regression models were used to assess if the increase from first survey to next was significant. Results: Coverage of program increased from 43535 in 2004 to 311667 in 2010, though there was a decline in number of TSLs. Along with this, there was a significant increase in number of truckers utilizing clinical services. Analysis from survey data showed that there was a seven-fold increase in clinic visit in last 12 months from 2008 to 2009 (21% vs. 63%, P B 0.001). Significant level of increase was also observed in percent of truckers who watched street plays, participation in health exhibitions. Furthermore, increase from round-1 to round-4 was significant in the following indicators: received condom (13% to 22%, P B 0.001), oneone counseling (15% to 21%, PB 0.01). Treatment seeking for sexually transmitted infection (STI)-related symptoms from STI clinics also increased six times during this time (16% vs. 50%, P B0.001). Conclusion: The re-designing of the intervention has increased the program coverage and service utilization. Implementing truckers program in limited number of high-impact locations after proper planning can help in saturating the coverage and optimum utilization of the available resources. Background: 'Test-and-treat' programs combining expanded HIV diagnosis, linkage to care, and ART delivery are under consideration, but real-world experience is lacking. During 2011, in rural southwestern Uganda, implementation of ART eligibility to CD4B 350/uL began. Concomitantly, we (A) conducted a community-wide HIV testing/linkage-to-care campaign, and (B) offered ART to adults with CD4]350 via a research study (EARLI: NCT01479634). One year later, we conducted a second health campaign and examined the population distribution of HIV RNA levels. Methods: During weeklong campaigns in May 2011 and 2012, all Kakyerere Parish residents were offered HIV testing (Determine, Inverness) in multi-disease diagnosis and linkage ''health fairs''. In HIV' individuals, HIV RNA levels were measured by a validated fingerprick blood collection method and RT-PCR (Abbott). We assessed population HIV RNA levels by computing (1) the proportion of persons with an undetectable VL, (2) the median VL, and (3) the mean log(VL) among HIV' persons. Results: After community mobilization, 4,343 and 4,872 persons attended the 2011 and 2012 campaigns, respectively. We estimated 69% and 71% community participation based on census data from 2011 and 2012, respectively. Adult HIV prevalence ( ]18yrs.) was 7.8% in 2011 (179/2,282 adults) and 9.4% in 2012 (210/2271 adults). Prevalence was 18.6% on the final day of the health fair located at the parish trading center, and 8.2% across prior days (p B 0.001). A substantially higher proportion of HIV' individuals had an undetectable HIV RNA level in 2012 vs. 2011 (55% vs. 37%), and both median VL and mean log(VL) were lower in 2012 (see table). Background: HIV and TB treatment is provided free of charge in the public sector in South Africa. However, patients may lose time and incur expenses due to utilizing these services. We measure time and financial costs of utilizing HIV and TB services, and examine in how far they lead to financial distress (patients' self-report of either borrowing money or selling assets to finance healthcare). Methods: We randomly selected patients in a two-stage-sampling scheme and collected data in patient exit interviews: 300 patients on antiretroviral treatment (ART), 200 enrolled in a pre-ART programme, and 300 receiving TB treatment in rural KwaZulu-Natal, South Africa. We assess factors associated with financial distress in multiple regressions, controlling for sex, age and employment.

May 2011 Health
Results: Most patients utilizing healthcare were women: 79% (in pre-ART), 62% (ART), 53% (TB). The average times patients spent utilizing care at the last clinic visit was: 3.5 (pre-ART), 2.8 (ART), and 1.1 hours (TB). The average total costs of utilizing healthcare during the last visit were: 15 (pre-ART), 25 (ART), and 20 South African Rand (ZAR) (TB). Transport was the single largest cost component in all three patient categories. 39% (ART), 31% (pre-ART), and 41% (TB) of patients reported financial distress due to healthcare utilization. For each additional hour spent utilizing healthcare, the odds of financial distress increased by 3% (pre-ART), 21% (ART) and 64% (TB). For every additional 10 ZAR spent on utilizing healthcare, the odds of financial distress increased by 25% (pre-ART), 9% (ART), and 6% (TB). Conclusion: In this poor, rural community, very high proportions of patients utilizing HIV or TB services report financial distress due to healthcare utilization. Frequent, free transport to HIV and TB clinics would likely substantially reduce the time and financial burdens of healthcare, as well as the prevalence of financial distress.

MOPDE0205
Economic spillover effects of ART on rural South African households Background: ART rollout may have important indirect effects on households and communities. We assess household spillover effects of ART on economic outcomes, in a region of rural South Africa with high HIV prevalence. Methods: Longitudinal socioeconomic data from a large population surveillance system (n !100,000) were linked with clinical records from the government ART program serving the area (n 06964). We first describe the extent of household-level exposure to ART among residents in the community. We then employ panel data techniques to evaluate the impact of ART on household assets, food insecurity, and perceived financial status; household composition; and the labor market participation and school enrollment of other household Track E Implementation Science, Health Systems and Economics members. We assess trends in these economic outcomes, relative to the dates when household members accessed the ART program and/ or initiated ART. As a point of contrast, we examine similar trends relative to dates of death due to HIV in the era before the public sector ART rollout. Results: By 2010, two-fifths of area residents lived with someone who had sought care in the government ART program. Prior to widespread availability of ART, households in the region experiencing an HIV-related death faced large economic costs. In contrast, we find little evidence that households of ART initiators experience significant economic hardship due to HIV illness. Conclusion: South Africa?s public sector ART rollout has provided substantial social protection for households against the economic costs formerly associated with HIV illness and death. Through indirect exposure to ART via household living arrangements, large populations now enjoy the economic benefits of ART.

E34 -Cost-effectiveness of optimizing diagnostics and monitoring tools TUAE0101
Cost-effectiveness of lateral-flow urine LAM for TB diagnosis in HIV-positive South African adults  Table 1. We adopted a lifetime time horizon with 3% discounting and performed both sensitivity analysis and probabilistic uncertainty analysis.
Results: For every 1000 patients tested, adding lateral-flow urine LAM to microscopy generated 80 incremental appropriate TB treatments and averted 58 DALYs, at a cost of $1400 (95% uncertainty range: $716Á$5773) per DALY averted ( Table 2). The majority of incremental costs reflected TB treatment. Incremental cost-effectiveness was most sensitive to assay specificity, cost of TB treatment, life expectancy after TB cure, and cohort TB prevalence ( Figure 1 and Figure 2). The probability of acceptability was 99.1% at a per-DALY willingness-to-pay threshold equal to per-capita gross domestic product ($11,000) and 85.3% at a threshold of $2000 ( Figure 3).      Track E Implementation Science, Health Systems and Economics $15,370/year for the regimen); the cohort is 84% male, mean CD4 317/ml. From the US health system perspective (2010 USD), we report incremental cost-effectiveness ratios (ICERs, $/QALY) compared against a $100,000/QALY threshold. We also project the potential annual savings for those initiating ART (incident diagnoses, Â2,500) and those eligible to switch to the generic-based regimen (prevalent cases, Â147,000) Results: Compared to No ART, generic-based ART has an ICER of $21,600/QALY. Compared to generic-based ART, brand-name ART increases lifetime costs by $43,900, and per-person survival gains by 0.36 quality-adjusted life years (due to increased efficacy and reduced resistance from emtricitabine compared to lamivudine), for an ICER of $121,300/QALY (Table). ICERs are sensitive to decreases in generic-based ART efficacy and to generic drug costs; most plausible combinations lead to ICERs !$100,000/QALY (Figure). Estimated annual savings if all eligible US incident/prevalent patients switched to the generic-based regimen are $1.01B.
Conclusion: Compared to a slightly-less effective generic-based regimen, the cost-effectiveness of the guideline-recommended brand-name regimen exceeds the $100,000/QALY threshold. Switching to generic-based regimens would yield substantial savings for programs that fund HIV treatment.
E42 -Effects of public-private partnerships, and working with faith-based groups and traditional healers on HIV outcomes Background: In many developing countries, reliance on private providers for healthcare, even among the poor, is high. However, less is known about health seeking behavior specific to HIV/AIDS care. As countries increasingly view the private health sector as an integral part of the overall health system, it is critical to understand the current and potential role of private providers in addressing HIV/ AIDS needs. Methods: We analyze 20 Demographic and Health Surveys from 18 countries conducted between 2005 and 2011 in which information on source of HIV test is available.
Results: Use of private providers for HIV testing varies greatly across countries (see Figure 1). Overall use of the private sector for HIV testing is strongly correlated with use of the private sector for other health services but not strongly correlated with per capita GDP, uptake of HIV testing, or overall HIV prevalence. At the household level, use of the private sector is highly correlated with household wealth in nearly all countries (see Figure 2) but not strongly associated with level of education, marital status or location (urban vs. rural) in most countries. Background: Supply chains are essential to securing availability of life-saving commodities for patients. Unfortunately, logistics training is often absent from the professional education of staff operating the supply chain. In-service training, while traditionally used to teach supply chain skills, takes personnel away from their work, is expensive to implement and often relies heavily on external technical assistance. USAID is investing in pre-service training (PST) as a costeffective and sustainable alternative to build a competent health workforce in supply chain management (SCM). Methods: The Supply Chain Management System (SCMS), a project under PEPFAR administered by USAID, works with local universities to incorporate SCM courses into pharmacy and nursing programs and build the capacity of university staff to plan, teach and manage those courses. Course content is based on local context and existing supply chain systems and prepares health workers for day-to-day supply chain tasks they will encounter post-graduation.

Figure 2. Among Women
Results: As a result of SCMS PST initiatives, the Namibian School of Pharmacy now includes pharmaceutical management and procurement courses in its curriculum. In Zimbabwe, with support from the USAID j DELIVER PROJECT, three pharmacy programs offer SCM courses, with nursing programs to follow suit.    Background: In order to improve access to antiretroviral therapy (ART), Lesotho was one of the first countries in southern Africa to adopt task-shifting from hospitals to nurse-led health centers (HC). We compared outcomes in patients who initiated ART at HC with those who started at hospitals in two catchment areas in rural Lesotho.
Methods: Patients aged ]16 years who started ART in one of two hospitals or one of their 12 affiliated HCs between 2008 and 2011 were included. Baseline characteristics were compared using Mann-Whitney and Pearson's chi-square tests. No follow-up was defined as no recorded follow-up after starting ART, lost to follow-up (LTFU) as not returning to the facility for ]180 days after the last visit. Crude retention in care was estimated using Kaplan-Meier methods. Mortality, no follow-up and LTFU in patients who started ART in hospitals and HCs were compared using logistic and cause-specific Cox regression models. Odds ratios and hazard ratios, adjusted for gender, age, CD4 count, WHO stage at start of ART and type of firstline ART regimen are reported.
Results: A total of 3,733 patients, including 2,036 (54.5%) who started ART at HC and 1?697 (45.5%) at hospital level, were analyzed. Patients who started ART at hospitals had more advanced disease than those in the HC-group (Table 1). Overall, 46 patients (2.3%) in the HC and 58 (3.4%) in the hospital group had no follow-up visit. Over 5,504 person-years, 414 patients died (11.1%) and 577 (15.5%) were lost to follow-up. Crude, three-year retention was higher at HC (67.8% (95% CI: 64.6Á70.7) versus 61.9% (58.8Á64.9)). However, in Track E Implementation Science, Health Systems and Economics adjusted analyses, mortality and LTFU were similar in both groups (

Track E Implementation Science, Health Systems and Economics
Background: As essential services for PMTCT are increasingly decentralized to antenatal care (ANC) sites, the consequences of shifting services from dedicated HIV care and treatment (C&T) clinics remain incompletely explored. We compared service delivery at ANC and C&T clinics in Kinshasa, DRC, a low HIV prevalence, resourcedeprived setting. Methods: In 10/2010, an enhanced standard of care was introduced at 44 ANC sites: personnel were retrained to implement the 2010 WHO PMTCT guidelines including Option A and co-located postdelivery care, and were provided with new individual-level tracking tools and supportive supervision. Women were encouraged to enroll at either of two affiliated C&T sites for continued PMTCT and HIV care but could opt to receive AZT-based prophylaxis at ANC sites when it became available alongside CD4 testing in 2011. Antiretroviral therapy was available only at C&T sites. Conclusion: Individual-level tracking of mothers and infants was feasible in Kinshasa and revealed that PMTCT services were delivered less effectively at sites historically focused on ANC rather than HIV C&T. While decentralization increased care access, its potential to further reduce vertical transmission cannot be fully realized without sustained training and supervisory support to ensure optimal quality of service delivery throughout the entire PMTCT cascade.

THAE0104
Preventing mother-to-child HIV transmission through community-based approach in Nepal

Track E Implementation Science, Health Systems and Economics
Conclusion: Despite similar clinical and immunologic outcomes, children in the outreach group were less likely to achieve virologic suppression, potentially due to lower adherence. Continued adherence counseling is critical for the success of decentralized care.
E 5 6 -S c a l i n g -u p m e t h o d s a n d experiences for HIV programmes WEAE0401 Sustaining quality while scaling-up adolescent ART: findings from Zimbabwe?s largest adolescent cohort Background: As Anti Retroviral Treatment (ART) programmes improve survival in early life the number of adolescents in need of ART will continue to increase. Programmes will need to scale up services which address the particular needs of adolescents. Here we assess the experience of such scale up in a large urban adolescent cohort.
Methods: Between 2004 and 2010 9,360 adults and 1,776 adolescents commenced antiretroviral therapy (ART) at Mpilo OI ART clinic, Bulawayo. A package of specific activities for adolescents was implemented, to provide comprehensive care including active defaulter tracing, a comprehensive, dedicated counselling programme and psycho-social activities operating both inside and outside the clinic, such as a youth club. Adolescents were engaged in decisions regarding their care. In this retrospective cohort analysis adolescents were defined as those ] 10 and B19 years old at ART initiation date. Cox's proportional hazards model was used to calculate hazards and the log rank test to assess significance. Results: Between 2004 and 2010 a six-fold increase in adolescent ART initiations occurred ( Figure 1). 12 month adolescent loss to follow increased in the first 3 years of the programme, peaking at 7%, subsequently falling to below 5% by the end of the study period. There was no significant difference between adult and adolescent hazard of death. (HR 00.92, p00.3793) Loss to follow up was significantly higher in adults than adolescents. (HR 01.92, pB0.0005) (Figure 2). Conclusion: These results contrast other research, which generally show adolescent outcomes to be worse than those in adults.
Outcomes in expanding programmes have also been shown to suffer elsewhere. As more HIV positive children survive into adolescence, ART programmes must scale up comprehensive services for this group; we show that good results are feasible with dedicated clinical and psychosocial resources in resource poor settings.

THAE0105
Nothing for us without us: community led approaches towards successful implementation of HIV prevention programs: experiences from southern India   Track E Implementation Science, Health Systems and Economics mandate is to provide technical assistance to the prevention programs. Results: Factors that contributed significantly to improved performance were that these projects are components of broader, holistic agenda that puts women at its centre, addressing women's needs, including, but not limited to, HIV prevention. Programs generate evidence-based plans from program data. Sitebased peer plans are used, which plan for individual FSW and tailor services according to individual profiles, have improved TI performance. This has helped Peer Educators to spend quality time in the field, while also improving their productivity. Supportive hand holding by Technical Support Unit has helped in diagnosing problems and its resolution. Conclusion: Embedding HIV prevention within broader programs that respond to women's needs, and in which FSWs play a deciding role, increases likelihood of success of the programs. As NACP-III nears its end and the nation prepares for NACP-IV, these findings indicate promising direction for future.
E57 -Integration of HIV services rather than stand-alone services, and services which bond the community care with the health facility services WEPDE0206 HIV prevention through pharmacies network in Ukraine: improving access to services for injecting drug users and commercial sex workers G. Naduta ICF International HIV/AIDS Alliance in Ukraine, Policy and Partnership, Kyiv, Ukraine Presenting author email: naduta@aidsalliance.org.ua Background: The prevalence of HIV infection in Ukraine is 1.33% (adult). In Europe it is one of the highest rates. The highest concentration of-among IDUs. Pharmacy interventions started in Ukraine in 2007. The clients of pharmacy interventions are mainly belong to IDUs, including female IDUs, and commercial sex workers (CSWs). Currently 143 pharmacies are implementing pharmacy-based prevention projects in cooperation with 25 NGOs in 13 out of 27 regions of Ukraine. Methods: Organization of the project activities: an NGO and a pharmacy sign a cooperation agreement. NGO: transfers consumables to the pharmacy, trains pharmacy workers, informs the target group about the project, pays for the pharmacy services within the project and reports to the donor agency funding the project (ICF ''International HIV/AIDS Alliance in Ukraine'', funded by the Global Fund). Pharmacy: provides free of charge services to the program clients, provides the program client cards to the new clients (IDUs and FSWs), and keeps record of the clients and distributed products. The basic package at the pharmacies includes: syringe distribution/ exchange (3Á5), alcohol swabs, condoms, information products, delivery of the harm reduction program client cards to the new clients. Counseling on safe behavior, referral of the clients to NGOs to receive other services (testing for HIV, STIs, counseling by specialists) are also provided. Results: Achievements for 2011: General year coverage: 27 435 clients, who made more than 387 000 visits to the pharmacies.11 714 new clients (IDUs, CSWs) became the participants and were issued à Participant card of the pharmacy programme. Syringes exchange organized in 32 pharmacies (out of 143)permission for such exchange obtained from the local controlling institutions.
Conclusion: Approximately 20.5 % of the over-all number of new clients (57 143) was involved into the HIV prevention programs of Alliance in Ukraine through the pharmacies in 2011.
E58 -Role of community organizations in linking people to HIV services and strengthening the health system WEAE0204 Community-based adherence support associated with improved virological suppression in adults receiving antiretroviral treatment: five-year outcomes from a multicentre cohort study in South Africa Background: Sub-Saharan African antiretroviral treatment (ART) programmes have a severe professional staff shortage resulting in inadequate support for patients which contributes to increasing virologic failure. However, community-based adherence-support (CBAS) workers now form a significant work-force. This study evaluated the effect of CBAS on virological outcomes at 57 nongovernmental organization-supported public ART facilities in four South African provinces. Methods: CBAS workers provide education and psychosocial support for ART patients through home visits to assess and address adherence challenges. A multicentre cohort study was performed, including adults starting ART between January 2004ÁSeptember 2010 at clinics where patients were eligible to receive CBAS. Prospectively collected routine electronic data were analysed. Outcome measures were virological suppression (VS) (viral load B 400 copies/ml) at six-monthly intervals until 5 years of ART. Multivariable generalised estimating equations were used to compare VS between patients who received and not receiving CBAS, using an intention-totreat approach. Extreme-case sensitivity analyses were performed to estimate bias due to missing viral load results. Results: 66,953 patients were included: 19,668 (29.4%) received CBAS and 47,285 (70.6%) did not. At baseline, median age was 34.8 years; 45,844 (68.5%) were female; median CD4 cell count was 125 cells/ml (IQR: 65Á175). After six months of ART, VS was 76.6% (95% CI: 75.8%-77.5%) and 72.0% (CI: 71.3%-72.5%) in CBAS patients and non-CBAS patients, respectively (PB 0.0001). In multivariable analyses after one and five years of ART, CBAS was associated with improved VS, adjusted odds ratio (aOR) 1  Background: In the DRC, traditionally stigmatized groups such as MSM are also those most at risk of HIV infection. In the first activity of its kind in Kinshasa, Projet Integré de VIH/SIDA au Congo (ProVIC), a USAID/PEPFAR-funded HIV/AIDS project, partnered with local nongovernmental organization Progrè s Santé Sans Prix (PSSP) to assess the receptiveness to and feasibility of providing voluntary nighttime mobile HIV counseling and testing (HCT), or ''moonlight'' testing, to MSM using rapid finger-prick technology. Providing quality services to MSM and addressing their specific needs, including combating stigma, are important to reversing the HIV/AIDS epidemic in the DRC. Methods: MSM were identified through risk mapping conducted at 15 sites in Kinshasa in 2010. Teams worked with groups of MSM peer educators from both the community and PSSP to provide HCT services using the rapid finger-prick method. Four mobile HCT teams, each including two laboratory technicians and a supervisor, conducted two moonlight HCT clinics per week. Attendance at the mobile HCT was carefully monitored, and observations on the strategy's successes were solicited from program staff.
Results: MSM attendance was initially low, with only 20 MSM accessing HCT services per night; over the course of the pilot, however, this number grew to an average of 80 clients per night. A seroprevalence of 30 percent was found among MSM ages 20 to 30. Involving MSM peer counselors in this intervention played an important role in the feasibility and effectiveness of this pilot. Conclusion: Rapid moonlight HIV testing for MSM led to increased service utilization because services were more easily accessible; clients were pleased to receive prompt, quality services without the perception of negative judgment. Staff reported that involving MSM peer counselors in this intervention plays an important role in these outcomes. Nighttime testing for MSM under ProVIC has since been expanded to two additional provinces in 2012.

Track E Implementation Science, Health Systems and Economics
Background: Post-electoral unrest in Ivory Coast led to major population displacement. Looting, lack of personnel, shortage of drugs, difficult physical accessibility and insecurity challenged access to HIV services. Given the individual and public health impact of antiretroviral therapy (ART) interruption, addressing continuity of treatment was defined as a priority since up to 14,500 Ivorians under treatment could be affected. Methods: As such, UNHCR ensured that all partners integrate HIV and AIDS related Identification, Communication and Case Management in the overall relief response. Out reach and registration of refugees and internally displaced persons were used as opportunities to link up those in need of ART. Messages delivered during gatherings encouraged PLWH to contact focal points for care and treatment. HIV Joint assessment missions highlighted lack of ART-related inputs conflict affected areas. Results: In Ghana and Liberia, among 258 refugees living with HIV registered, 142 were receiving ART. 60% of PLWH receiving ART in western Ivory Coast and 83% in Abidjan had interrupted their treatment for more than 3 weeks at time of identification. Language barriers (refugees), fear for stigmatisation and lack of ART services in refugees and IDP hosting areas were main reasons for treatment interruption. People in need of ART were referred to closest treatment centres. Cross-border arrangements were made between Liberia and Ivory Coast to overcome a difference in second line therapy. UNHCR procured portable CD4 machines in conflict-affected areas to improve treatment quality. Conclusion: Early identification of cases and communication through different strategies was essential to limit treatment disruption. Crossborder coordination and contingency planning for drug procurement and distribution enables continuity of treatment. The importance of continuity of HIV-related business continuity including treatment need to be further emphasised to patients, as well as planners and relief assistance managers. Media should be cautious when talking about HIV status of a displaced population.

THPDE0203
The challenge of maintaining continuum of care and support to PLHIV in health facilities located in military conflict zones in Ivory Coast Background: Post-election crisis in 2010 in Ivory Coast caused militaries conflicts in several cities, disturbing health centres activities. NGO Aconda tried to support provision of care to PLWHA. To describe strategies implemented by Aconda to ensure continuum of care and support to PLWHA in health centers located in conflict zones. Methods: It is a prospective study which describes activities of Aconda in health centres located at conflict area during war in Ivory Coast. Data has been collected from monthly reports. Strategies were based on reduction of HIV activities package to an absolute minimum: supplying ARV to health centres; providing ART to patients in follow; continuing PMTCT and providing ARV prophylaxis to pregnant women. Communities based Organizations made active research of patients lost to follow and referred them to health facilities for care. Pharmacists gave systematically at least 3 monthly ARV provision to patients coming to visits. Counselling and testing Activities, biology and ART initiation were stopped. Results: From January to June 2011, 05 health regions: Abidjan, Duékoué, Guiglo, Bloléquin and Toulépleu were located conflict zones. We noted 45/68 centres were functional: 40 in Abidjan, 3 in Guiglo and 2 in Duékoué. We note that 11 centres have been looted; 7 completely and 4 partially. Because of non accessibility for drug's vehicle, 32 sites had failures in delivering ART. In 28 sites, less than 50% of medical staff was present. In 12 centres where caregivers were absent, the provision of ART to patients has been ensured by counsellors and data managers. In the 3 months, 17,471 PLWHA (8.7% of children) received ART and 475 pregnant women, the ART prophylaxis. Conclusion: Militaries conflicts cause a dislocation of health systems. Aconda managed to maintain care and support to PLWHA in health centres in war area, by reducing to absolute minimum services. Results: To date, the programme has covered total 46 PLHA families having average 4 members per family living at rural areas. A total of 143 PLHA has been supported, out of which, 43% were male (n061), 33% (n047) were female and the rest were children. After affected by HIV, the house hold income dropped, on average, by 33%, whereas household expenditure increased by 110%, which was mainly for medical purposes. Through the programme, the following services were provided: vocational training to all covered adult male and female (76%, n0108); job assistance (n 017), 42% (n 07) of them were female; microcredit loans to start small scale business (n 07, all were widowed female; health loans (n 022), 68% (n 012) of them were female and 13% were children (n03). Lump sum educational assistance grant (45 US Dollar per year per children) was provided to 35 HIV affected children (24% of the total covered PLHA). Moreover, job place advocacy and sensitization on HIV/AIDS were also done in different organizations (n 012), Compared to the baseline, the average monthly income in each house hold increased 5.7 times higher after 1 year of coverage (mean9SD in US Dollar, 35911 vs 201921, p B0.005).

Conclusion:
The success of this small scale programme demonstrates the importance of economic coverage for restoration of PLHA livelihood. Long term, large scale comprehensive programme is therefore needed further.
E63 -Economic evaluation of prevention, t re a t m e n t , c a re a n d mi t i ga t i o n programmes, adherence schemes, sustainability of programmes and financial strengthening initiatives

MOAE0201
Cost and efficiency analysis of the Avahan HIV prevention programme for high risk groups in India Background: The India AIDS Initiative (Avahan) is involved in scale up of HIV prevention interventions among high-risk groups (HRGs). The few existing studies limit their cost analysis to the implementing non-governmental organisations (NGOs). This is the first study to analyse the costs of HIV prevention at scale, over four years, including systems costs. These systems costs are particularly important for those replicating scale-up of HIV prevention in different settings. We present here the final results of a four year costing effort.
Methods: Financial and economic costs were collected from 64 districts, 137 NGOs and supporting costs at the state and costs of programme management by the Bill and Melinda Gates Foundation over four years, presented in US$ 2008 (3% discount rate). The intervention package included outreach, STI services, condoms, capacity building, community mobilisation, advocacy and enabling environment. Results: Unit costs per person reached, per contact, per estimated population ranged US$ 232 (44Á840), US$ 68 (13Á242) and US$ 175 (95Á443) respectively. Key factors driving cost variation include scale. Although costs fell with scale, they also increased over time with inclusion of new districts and activities as the programme developed (including high risk men who have sex with men, advocacy, and community mobilisation). The total economic costs (2004Á2008) were US$ 97,966,216 of which approximately 35% was spent at NGO level, 30% at the State level and 35% at the national level which were mainly incurred for capacity building and programme management. Conclusion: The rapid scaling up HIV prevention requires significant investment in expertise enhancement and programme administration. These are not captured in routine cost analysis, which may misinform resource planning. At scale the unit cost of services provision falls, however actual cost reduction may take longer as programmes evolve, and activities are added. Policy makers should take into account these additional costs while planning public health interventions.

MOAE0204
Increasing investment in syringe exchange is cost-saving HIV prevention: modeling hypothetical syringe coverage levels in the United States  Methods: A costing exercise was conducted for 20 OVC programs in Ethiopia and 19 in Botswana. The data collection process involved conducting interviews with program managers and finance officers. This study followed an economic approach, where the value of all resources deemed necessary to deliver services to OVC were taken into account.
Results: Most of the organizations in both countries offered between 4 and 8 services to OVC. The most common services offered to OVC were education, psychosocial support and food/nutrition. The cost of reaching an OVC differed significantly across countries, with a unit cost of $80 in Ethiopia and $946 in Botswana. Even for specific Figure 1. Track E Implementation Science, Health Systems and Economics services, the cost differed significantly both within and across countries. Conclusion: This analysis provides important comparative information about the range of costs by service area and the overall cost per child reached. The analysis shows that unit costs decline significantly as programs are scaled-up (diminishing marginal returns), but then increase as OVC programs grow bigger (diseconomies of scale). Differences in unit costs appear to be largely attributable to the type and intensity of services provided, as well as the overall difference in the cost of living between the two countries. The analysis could not, however, make any conclusions about the effectiveness of the interventions provided. It is recommended that future OVC research focus on measuring both costs and effectiveness.

FRAE0103
Economic evaluation of the national program to prevent mother-to-child transmission of HIV in Ghana  Track E Implementation Science, Health Systems and Economics thought to influence unit cost, including facility level and facility ownership. Results: We found an average cost of US$56 for providing HTC services, US$739 for seropositive pregnant women on ARV prophylaxis, and US$1,054 for sero-positive pregnant women on ARV treatment. The cost for providing PMTCT services to an HIV-exposed infant ranged from US$29 to US$69 (Figure 1). Direct costs accounted for the majority of total cost, comprised primarily of costs for antiretroviral drugs, laboratory testing, and staff time; suggesting that the unit cost of delivering PMTCT services will not vary significantly in terms of changes in client load (Figure 2). Conclusion: We found higher staff costs for delivering the same services at higher-level facilities, where more specialized staff deliver services, suggesting task shifting could achieve cost efficiencies. These data can be used to better inform resource allocation decisions as PMTCT programs are scaled up in Ghana and other countries.

MOAE0205
Transition from stavudine to tenofovir and zidovudine for first-line treatment of HIV/AIDS in low-and middle-income countries Conclusion: Transition from stavudine has been slow, though some countries have exhibited rapid transition. Price remains a critical concern in transition from stavudine to ZDV or TDF regimens. E65 -Economic strengthening initiatives (inc luding mic ro-finance, health insurance, etc.)

WEAE0402
Self-esteem, self-efficacy and hope among vulnerable adolescents affected by HIV participating in communitybased savings and lending groups in rural Nyanga district, Zimbabwe Background: Approximately 25% of Zimbabwe's children are estimated to be orphaned, with the majority orphaned due to AIDSrelated illnesses. HIV-affected households often face severe financial constraints making them unable to meet their basic needs. Financial stress can negatively impact overall mental health, life satisfaction, and wellbeing in numerous ways. Perceived control over one's economic situation and sense of self can mediate financial distress. Methods: Over a one year period, 2000 adolescents participated in community-based savings and lending groups (SLG), and 50% also received life skills education (LSE). A quantitative survey was administered to a randomly selected sub-sample of the project population matched with a sample of adolescents from a nonintervention control ward. Surveys were administered in Shona and back translated into English. Data were entered into SPSS and analyzed using univariate and bivariate measures. Results: 160 adolescents participated in the survey. Average age was 15-years-old with an average household size of seven members. Adolescent SLG group members (n 0139) reported statistically significant higher scores on self-efficacy (p B.001), self-esteem (p B.01), and hope (p B.01) than adolescents from the control group (n 021). There were no statistically significant differences between adolescents who received SLG only compared to those who received SLG plus LSE, although most of the random sample received both interventions. While 75% of adolescents reported that caregivers decided how general household income was spent, 45% of adolescents reported caregivers decided how their SLG funds would be spent; 33%made independent decisions regarding their funds. Hope scale scores positively correlated with increased decision making power over funds. Conclusion: Adolescents affected by HIV face many challenges including poverty and unknown futures. Economic empowerment may provide an option for responding to both immediate financial concerns and also building mental health resilience and outcomes for the future.
E66 -Consequences of test-and-treat programmes on investments in human capital and on household finances/ wealth THAE0102 Improved employment and children's education outcomes in households of HIV-positive adults with high CD4 counts: evidence from a community-wide health campaign in Uganda Background: Despite growing evidence that socio-economic outcomes among HIV-infected adults show improvement after antiretroviral therapy (ART) initiation, little is known about the variation in these outcomes among a population that also includes individuals with high CD4 counts and those not enrolled in care. We examined associations between CD4 count and socio-economic outcomes among adults participating in a community-wide health campaign in a rural Ugandan parish. Methods: A one-week community health campaign offering diagnostic and treatment services for HIV and other infectious and noncommunicable diseases was conducted in May 2011. Data on campaign participants? employment were collected, and a detailed household socio-economic survey was conducted among a random subset of participants. Multivariable regression was used to assess relationships between CD4 count and employment and educational outcomes. Results: 2,323 adults (74% of the community) participated in the campaign. 179 adults (7.8%) tested HIV-positive and 46% were newly diagnosed. HIV-positive adult participants with CD4 !550 and 350Á 550 worked 4.8 and 5.3 more days during the past month than those with CD4B200 (p B0.05). No differences in work patterns were found between participants with CD4 200-350 and B200. The association was similar among those on ART and not on ART. Children's school enrollment was also associated with adults' CD4 counts. Children in households of adults with CD4 !350 had 20% higher school enrollment rates than children in households of adults with CD4 B200 (p B 0.05). Finally, socio-economic outcomes of HIVparticipants with high CD4 counts resembled those of HIV-negative participants. Conclusion: Outcomes of HIV-positive adults with high CD4 counts are not only better than those of adults with low CD4 counts, they also resemble those of HIV-negative adults. Early initiation of ART could generate economic benefits by preventing a decline in employment and education outcomes and maintaining them at levels seen among HIV-negative peers.  Figure 1); gains were larger for women (9.5 years) than for men (6.6 years). Changes in the distribution of lifespans reveal a substantial shift towards older ages (Figure 2). Applying standard estimates of the value of a statistical life-year (1-3 times per capita GDP), ART roll-out in this community has led to lifetime gains of $26,000 to $77,000 per capita. This is 2-6 times larger than the per capita cost of providing lifelong ART to every person who contracts HIV in this community, which we estimate at $13,400, based on 2012 cost figures from the President?s Emergency Plan for AIDS Relief. Conclusion: ART roll-out has resulted in large increases in adult life expectancy in rural South Africa. The economic value of these health gains far outweighs the costs of treatment. Future research is needed to understand how people perceive these changes in longevity, and how these changes affect attitudes, such as fatalism,  Track E Implementation Science, Health Systems and Economics and behaviors, such as healthcare-seeking, risk-taking, and educational choices.

MOPDE0206
The impact of antiretroviral therapy on the social, economic and working conditions of patients with HIV in Malawi Background: Antiretroviral therapy (ART) for treatment of HIV/AIDS patients in less developed countries (LDCs) produces an impact on individual health and indirectly impacts socio-economic status. We investigated the impact of an ART program on work productivity and household income as these are key parameters for cost-benefit analyses of HIV public programs and policies. Methods: Health status of HIV-infected patients on ART was determined at baseline and subsequent visits through assessment of Body Mass Index (BMI), hemoglobin, viral load and CD4 cell subsets. Socio-economic status was assessed exploring productivity (hours worked), income, and other economic variables through patient interviews. Using a prospective design we followed a cohort of patients from the DREAM program Malawi with the following inclusion criteria: HIV positivity, age ]15, ART initiation within 2 months of enrollment. All subjects had at least 8 months of follow-up post-ART initiation. Health, income and productivity parameters were evaluated through paired t-test. Results: 165 subjects were followed from 1/2008 to 3/2009. The overall health status of subjects improved significantly based on clinical and virologic parameters (Table). A positive overall impact on productivity and income was noted with hours worked in the last week increasing by 25%, hours worked in the last month increasing by 31%, income generated in the last week increasing by 85%, income in the last month increased by 80% (Table). For patients who were unemployed at baseline (n037) mean income was too low ( B4$ per month) to be relevant for the analysis. With the exclusion of unemployed patients at baseline, the increase in last week/last month hours worked was '35% and '43% respectively, and last week/last month income '93% and '89% respectively ('400$ per year). Conclusion: ART therapy has a strong impact on patients productivity and income, and provides an economic incentive towards sustainability of ART delivery programs in LDC.

WEPDE0201
Does condom use affect the earnings of commercial sex workers? New evidence from a survey of female sex workers in India S. Mondal and I. Gupta Institute of Economic Growth, Delhi, India Presenting author email: kumar.swadhin@gmail.com Background: Unsafe sex is the major cause of transmission of HIV in developing countries like India. Condom use is, therefore, seen as the most effective instrument against the transmission of HIV. However, use of condoms in commercial sex is still nowhere universal. Part of the reason could lie in the evidence that sex workers lose income when they enforce the use of condoms among their clients. This paper re-examines how condom use affect the price for commercial sex. Methods: A survey of 5498 female sex workers from four high HIV prevalence states of India was analyzed using a simultaneous equations model using three-stage least squares method, to identify the causality between condom use and price for commercial sex. Results: The data indicated that 77 percent of the sex workers are consistent condom user. The analysis shows that there both price of sex and condom use influences each other. Also, results indicate that all types of sex workers (Road, Brothel, Hotel and Bar) except brothel-based workers received 35 percent higher earnings when they have protected sex than unprotected sex. Brothel based sex workers earned 24 percent less when they enforced the use of condom. The other exogenous factors that affect the price are negotiating power, level of education, type of sex workers (other sources of income, number of client visit per day, contract under pimp etc.). Hotel and bar based sex worker getting comparatively higher price. Conclusion: The market for sex work has changed in favour of sex workers and condom use generally no longer has a negative premium for most of the sex workers. It is, therefore, possible to step up the intensity of behavioral intervention for enhancing E68 -Investment frameworks and models in a period and in settings of decreasing financial resources.

MOAE0203
Company-level ART provision to employees is cost saving: a modelled cost-benefit analysis of the impact of HIV and antiretroviral treatment in a mining workforce in South Africa Background: HIV impacts heavily on rates of absenteeism, labour force turnover and, ultimately, cost of operations of companies in sub-Saharan Africa. Employers increasingly supply HIV testing and antiretroviral treatment (ART) programmes at the workplace. The full economic impact of ART provision on a mining workforce has not yet been analysed. Methods: We developed a dynamic health-state transition model to analyse the economic impact of HIV and the cost benefit of ART provision in a mining company in South Africa between 2003 and 2022. The model was fitted to the workplace using information on the size, composition, turnover, HIV prevalence and CD4' counts of the workforce from company records. Bottom-up analyses of economic costs at the company supplied data on inpatient and outpatient resource utilisation and the costs of absenteeism and replacing a sick worker. Costs are analysed from the company perspective and presented undiscounted and discounted at 5%. Results: As a result of lower mortality and morbidity in the employees covered by ART, survival in employment of HIV-positive employees grows by 7% as ART coverage increases from 21% to 80% of eligible HIV-positive employees by 2022. The associated reduction in absenteeism and benefit payments more than offsets the additional cost of the ART programme, leading to a 9% decrease in the total and annual cost of HIV to the company and a 15% decrease in the mean cost per HIV-positive employee by 2012 (Table 1). 48% and 37% of cost savings are due to reductions in benefits and absenteeism, respectively (

Track E Implementation Science, Health Systems and Economics
Methods: For 20 countries with generalized HIV epidemics, current national strategic plans and progress reports were reviewed to assess the national response in terms of: planning, measured by the inclusion of youth-specific strategies within national AIDS plans; and implementation, or the extent to which prevention activities reach the intended audience.
Results: All 20 countries include youth-specific strategies in their current national AIDS plan, and school-based HIV prevention was the youth prevention strategy most often included. Governments of all 20 countries report that school-based HIV education is reaching the majority of people in need, and included in primary, secondary, and teacher training curricula. The proportion of schools providing lifeskills based HIV education varies from 2% to 100%. Programmes for out-of-school youth, behaviour change communication, and condom promotion were commonly included in national strategies, however, their content, quality and coverage were generally not reported. In UNGASS country progress reports, few countries disaggregate by age and sex the UNGASS indicators relevant to young people, or report comparable statistics over time.
Conclusion: In 20 high-prevalence countries, HIV prevention among young people is considered a priority in national plans, and the most widely implemented intervention for youth is school-based prevention. Monitoring youth-focused programmes should be improved to assess coverage, quality and delivery through comparable data over time. At a minimum, reporting UNGASS indicators by age and sex will improve the usefulness of national progress reports in tracking prevention efforts for youth. As a priority, systems are needed to report planning and programmes to promote condoms and HIV testing among young people, given the scarcity of data at the national level.

MOAE0302
Sustainability of HIV programs: lessons learned from sustainability analyses in four countries Track E Implementation Science, Health Systems and Economics different interventions, and between target populations to be reached by a given intervention. Prioritization is more challenging where no cost-effective information exists, e.g., when dealing with non-clinical interventions. In such cases, the relevance of the intervention to the specific country context and the capacity to scale-up and sustain were assessed; 2) Efficiency, which includes, among others, partial or full integration of HIV services with the wider health and social services; and 3) Resource mobilization, from traditional sources such as governments, as well as innovative resources such as a levy tax on airlines.
Conclusion: With the decline in donor HIV funding, national HIV Programs will have to reassess the components of their respective programs in the context of sustainability. Countries will benefit from integrating sustainability plans into their strategies, operational plans and funding proposals.

WEAE0105
Affordability of HIV/AIDS treatment in developing countries: an analysis of ARV drug price determinants Conclusion: Procurement policies should pay attention not only to higher prices practiced by innovator firms for drugs recommended in first-line regimens, but also to drugs with patents arriving to expiration whose prices are higher than drugs with patents expiring later. Where patent protection is concerned, high prices may impair the delivery of higher quality treatment in developing countries.

TULBE06
Scenario-based cost projections for PEPFAR resource requirements for the ART program in Ethiopia from FY2011-FY2015 Background: Providing HIV treatment consumes substantial program resources. Forecasting future resource requirements can help ensure uninterrupted service delivery as treatment programs grow. We undertook cost analyses and projections for PEPFAR resources required to continue supporting the national ART program in Ethiopia under various scenarios from FY2011 through FY2015.

Track E Implementation Science, Health Systems and Economics
Methods: We collected data on actual expenditures devoted to supporting treatment programs in FY 2010, and used these data to calculate unit costs for various services. Data about patient and program characteristics were collected from performance reports, indicator targets, results of cohort analyses, and known program mark-ups and prices. We used the PEPFAR ART Costing Model 2010 to calculate resource requirments under scenarios describing different rates of new patient enrollment. The base case scenario assumed continued program growth at its current pace, adding 4,000 new ART patient slots every month from an initial patient volume of 207,458 ART patients. We assumed that Pre-ART patient volume continues to grow proportional to ART patient volume. This base case scenario was compared to 1) a slower (50% reduction) in patient uptake from base case scenario, 2) linear scale-up to universal access targets, 3) rapid scale-up to universal access targets.
Results: Faster rate of scale-up requires markedly increased resources for expanding capacity to accommodate a large number of patients at the initial phase, but in subsequent years, annual costs drop sharply and drop below the slower rate of scale-up for universal access. However, the total 5 year cost was higher than in all other scenarios. ARV costs represent a small fraction of total PEPFAR costs, as these costs are largely covered by the national government through Global Fund and other donors. Conclusion: Programs need to weigh the higher costs of rapid scaleup, and the improved health outcomes and economic gains of universal access in making planning decisions.

MOAE0303
Spending on ART by provinces in South Africa: trends, cost drivers, (in)efficiencies and sustainability Background: In 2003, the South African government committed to providing ART free to HIV-positive patients attending public health care facilities. National Treasury increased the budget allocations to cope with the rapid rate of ART roll-out, and has committed to increasing its allocations by 22% on average over the Medium Term Expenditure Framework (MTEF) period (2010/11 to 2012/13). Initially, provinces struggled with lack of capacity, stock-outs and even underspending. In recent years, some of these challenges have been reduced as procurement and supply systems have been improved and human resource capacity has been expanded. However, examining the different provincial ART expenditure in relation to respective numbers of ART patients and HIV prevalence shows interesting variations, highlighting possible areas for improved efficiency in some provinces.

Methods:
The total public spending that was labelled as''ART'' by the South African provinces was identified for the years 2007/08, 2008/ 09 and 2009/10, and broken down into cost components. This was limited by the level of disaggregation available in the public expenditure records. Simple unit costs were calculated using the total provincial ART spending per annum, divided by the number of public ART patients in that province in each year.
Results: The rough estimates of unit costs over the three years show increasing efficiencies in delivery, with variations in the key cost drivers per province. Comparing the provincial spending with estimated numbers of patients in need of treatment indicates that allocations and spending appear to be along lines of capacity to deliver, rather than of need. As spending on ART in SA dramatically increased, spending on prevention interventions decreased nominally and proportionally, raising concerns of sustainability. The attached figures present some key findings.

Conclusion:
The study concludes with recommendations for improved efficiency of ART spending and delivery, and sug-    Track E Implementation Science, Health Systems and Economics  (Table 1). Estimated cost to the national program is 0.68USD / patient/year for scale up from 0.42million currently on ART to 0.8million patients expected to receive ART by 2017 ( Figure 1). Cost estimates for individual components ie; IVR alone and SMS alone for 0.8million patients are 0.64USD/patient/year and 0.16USD/patient/ year respectively (Table 2). Conclusion: From our study, the Indian program would incur an overall cost of 0.086% of its current 5year budget to provide such adherence support. The costing results of this study are relevant for policy makers when considering interventions to enhance adherence support within the national program.
E75 -Use of operations research to support the response to HIV prevention and treatment

MOPDE0202
Factors associated with non-utilisation of antenatal care services by women delivering with an unknown HIV status at four poly-clinics in Chitungwiza, Zimbabwe  Background: Evidence has emerged linking the lack of readily available accurate information of the patent status of key ARVs with high prices in some countries. Despite their similar socio-economic circumstances, it shows that countries often pay significantly different prices for the same ARVs. Greater cost savings can be achieved if organizations and officials involved in the national procurement have more transparent information regarding the patent status of key ARVs.
Methods: UNDP has co-sponsored the development of a methodology written by Barbara Milani and Cecilia Oh that can be used by officials involved in procuring ARVs and other essential medicines in low and middle income countries to determine their patent status and ascertain whether and when generics can be freely imported into their country. Results: The methodology was preliminarily tested in Canada, China, India, and South Africa in 2006 and proved to be effective in determining the patent status of ARVs. An updated and expanded version of the methodology is being finalized and will be launched before the AIDS Conference in July 2012. Conclusion: In light of the decrease in AIDS funding and growing need for countries to lower treatment cost, the methodology can be used by officials and organizations involved in procurement to ensure that generic ARVs are imported legally, and by low and middle income countries to facilitate the use of other flexibilities, where necessary, if valid patents for key ARVs are hampering the procurement of more affordable equivalents of comparable quality.

WEAE0103
Understanding voluntary licensing: an analysis of current practices and key provisions in antiretroviral voluntary licenses Background: Voluntary licensing has gained increased attention as a potential policy measure to improve access to generic versions of widely-patented antiretroviral (ARV) medicines in developing countries, particularly in relation to the Medicines Patent Pool, an international voluntary licensing mechanism. This study offers a description of ''baseline'' voluntary licensing practices before UNI-TAID's decision to establish the Pool in 2009, analyzes the evolution of licensing practices since then, and highlights the types of license provisions with the most important implications for public health.

Methods:
We compiled data on all publicly-announced voluntary licenses between major ARV patent-holders and generic firms, and analyzed various provisions covering: geographic scope for production and sales; degree of competition enabled for end products and active pharmaceutical ingredients; royalty rates; freedom to coformulate into fixed-dose-combinations and pediatric formulations; technology transfer; regulatory data; and various other provisions with public health relevance. Results: There is wide variation in voluntary licensing practices, with geographic scope ranging from one to 112 countries; number of licensees ranging from one to unlimited; royalty rates ranging from 0% to 5% of the generic price; freedom to co-formulate ranging from none to unlimited; and access to technology transfer and regulatory data ranging from minimal to extensive. Other terms and conditions may have important public health impact, but the lack of transparency on the text of licenses limits the scope of the analysis. Conclusion: Voluntary licenses offer one potential route to improving access to low-cost, generic ARVs in a predictable, sustainable manner. However, attention must be paid to license provisions that can limit access to generic medicines, restrict competition, create onerous burdens, or otherwise reduce the public health benefits of licensing. Increased transparency in licensing practices, critical analyses of license provisions, and improved license terms are necessary to maximize the potential public health benefit of this policy tool.

WEAE0104
Panorama of the pharmaceutical patenting and sanitary registration of ARVs drugs in Brazil: implications to access and to health industrial complex P. Villardi Associação Brasileira Interdisciplinar de AIDS, Rio de Janeiro, Brazil Presenting author email: pedro@abiaids.org.br Background: In Brazil the number of patents and patent applications has increased, making it difficult to place generic drugs in the market. At the same time, it is fundamental that the pharmaceutical companies-either innovators or generic-register their drugs, formulations or combinations in the National Agency Sanitary Surveillance-ANVISA (acronym in Portuguese) so that the general population to access is their sanitary registration.
Methods: This study was conducted between April 2010 and April 2011 and it was funded by the Brazilian Ministry of Health. The research resulted in two methodological outcomes: 1) an ARV patent and sanitary registration database; 2) research methodology aimed at overcoming one key difficulty faced in Brazil in this area of investigation, which is that patents are not automatically linked to their corresponding final products. The methodology was designed for Brazil, but it may be adapted to other countries. Regarding to sanitary registration of ARVs, ANVISA provides an online database of granted and not granted sanitary registrations. Results: Patents and registration for 23 single-drug medicines and 8 combinations were reviewed. The study identified 96 patents or patent applications related to these drugs and combinations, which account for more than 3 patents/drug. One study result is that the granting of a secondary patent delays the entry of generic medication in the market. In the case of darunavir, the first patent will expire in 2016, but if the last patent application is granted (for Tibotec), the patent applicant, will benefit from a monopoly until 2025.

Conclusion:
The results of the research led to the following policy recommendations: 1: Wide application of the Bolar Exception and rigid standards for patent examinations; 2: Incentives for pre and post-grant oppositions; 3: A possible solution for the non-transparent Background: Access to HIV programmes for refugees and internally displaced persons (IDPs) is a human rights issue and a public health priority for affected and host populations. Primary source of funding for HIV programmes for many countries is the Global Fund. This article analyses the current HIV National Strategic Plans (NSPs) and Global Fund approved proposals from rounds 1Á8 for countries in Africa hosting populations with refugees and/or IDPs to document their inclusion. Methods: The review was limited to countries in Africa as they constitute the highest caseload of refugees and IDPs affected by HIV.
Only countries with refugee and/or IDP populations of 010,000 persons were included. NSPs were retrieved from primary and secondary sources while approved Global Fund proposals were obtained from the organisation's website.
Results: Majority of countries did not mention IDPs (57%) compared with 48% for refugees in their HIV NSPs. A minority (21Á29%) of HIV NSPs referenced and included activities for refugees and IDPs. The majority of countries with 010,000 refugees and IDPs did not include these groups in their approved HIV proposals (61%Á83%). Conclusion: Besides legal obligations, Governments have a public health imperative to include these groups in NSPs and funding proposals. Governments may wish to add a component for refugees that is additional to the needs for their own citizens. The inclusion of forcibly displaced persons in funding proposals may have positive direct effects for host populations as international and UN agencies often have strong logistical capabilities that could benefit both populations. Given recent developments in the Global Fund and HIV funding in general, the HIV funding situation for forcibly displaced persons may worsen. It is essential for their inclusion to occur if we are to reach the stated goal of universal access and the Millennium Development Goals. is the largest donor in Southern Africa, spending over US $6 billion since 2004 and supporting ARV treatment for 1.5 million people. However, since the introduction of PEPFAR conflicting narratives have appeared regularly. These conflicting narratives often tend to differently describe and interpret the content, intent, impact and outcome of PEPFAR policies, particularly with regards to prostitution.
Methods: This analysis utilized a competing narrative approach to identify recurrent and conflicting narratives relating to PEPFAR and its implementation among organizations and populations in southern Africa. A competing narrative methodology allows the identification and comparison of differing interpretations of phenomena. Here narratives found in published, ethnographic and interview sources from sex work and public policy communities in Southern Africa and elsewhere were analyzed and similarities and differences in these communities' interpretations of the PEPFAR policy compared. Results: The presentation and discussion of PEPFAR's anti-prostitution clause vary widely, and have varied over time. Sex worker communities and organizations have spoken of PEPFAR's "dark ripple effect"-the incremental phase-out of sex worker-accessed services, the increasing isolation of sex workers, and fears about disseminating and sharing information about sex work programs and funding. In contrast, public policy communities have been slow to acknowledge, publicly, the policy's suggested failures, or to provide clarifying guidance or response to proposed shortcomings. Conclusion: Among the two often-opposing communities, conflicting narratives about PEPFAR and its implementation were found. Evident were narratives of PEPFAR's suggested harms and the lack of response to it. This analysis builds on earlier work to suggest that fostering improved dialogue between sex work and US funding communities would help to better broad understandings of PEPFAR's impacts, while potentially translating to improved HIV service access for sex workers.
E89 -Financial sustainability of the response to HIV and AIDS

MOAE0304
Ensuring the financial sustainability of the national AIDS response in a low-middle income country with the growing HIV epidemic: is it feasible in the next few years in Ukraine? Background: Ukraine is a lower-middle income country with the most severe and growing HIV-epidemic in Europe (1.1% estimated adult prevalence), concentrated among people who inject drugs (PWID), sex workers (SWs) and men who have sex with men (MSM). Ukraine's financial needs for the adequate HIV response are dramatically increasing while it continues to highly depend upon external funding. The major challenge is to sustain financing of the HIV response while the global financial crisis prompts uncertainty of the level and scope of future donor support, especially for middle-income countries. We aimed to examine the financial sustainability of Ukraine's HIV response, and make recommendations for its mid-and long-term financing strategies.

Track E Implementation Science, Health Systems and Economics
Methods: We analyzed trends in the annual expenditures using the National AIDS Spending Assessments; financing needs for 2011Á2015 using the 2011Á2013 National Operational Plan and the gap analysis for 2014Á2015; future domestic and external funding allocations; policy environment, and implications of all those on future financial sustainability.
Results: The HIV expenditures dramatically increased since 2004 (USD3.3M) with external and non-governmental resources comprising 42Á48% of the USD 65.4Á72.8M in 2009Á2010. Despite the increasing future funding allocations from all sources, the financial gap remains substantial estimated at USD 138M in 2015. All HIV care and prevention services remain free through public and community-level settings regardless of the funding source.
Conclusion: While domestic investments for treatment increase, HIV prevention especially for key populations is almost fully funded externally. Allocation of a larger proportion of governmental resources is critical, along with policy initiatives aimed at reduction of costs per client through efficient service delivery modalities, including HIV service integration into primary and other care. While donor-funding commitments will be still required for the next few years, establishment of alternative modalities like financing HIV from a national health insurance, introducing cost-sharing arrangements, and facilitating private-public partnerships will be needed.  (2009Á14) has an increased emphasis on promoting the sustainability of the HIV/AIDS response. Such sustainability is related to greater country ownership, local capacity, and strengthened health systems. Investments promoting sustainability can compete for resources with the rapid delivery of services.
There is need to measure the costs associated with an increased emphasis on sustainability.
Methods: The USAID HPI Costing Task Order evaluated the cost of PEPFAR support to service delivery and sustainability for the PMTCT program in Kenya. Data were analyzed from two implementing partners: FHI360 and the Elizabeth Glaser Pediatric AIDS Foundation. Data sources included implementation and financial records as well as interviews with partner staff. From this data, unit costs of PEPFAR support per mother receiving ARV-based prophylaxis and per infection averted were calculated for the study period, 2005Á2010.
Results: The average unit cost across EGPAF and FHI360 of PEPFAR support (excluding commodity costs) per HIV-positive mother provided antiretroviral-based prophylaxis declined by 52% from $567 to $271. PEPFAR support for commodities and supply chain added $48 to the latter cost in 2010. The unit cost of PEPFAR support per averted mother-to-child infection declined by 66% from $7,117 to $2,440, excluding commodity costs. The sustainability-related proportion of the unit cost of support to PMTCT increased from 12 to 33% over the study period. Investment into sustainability has grown in proportion and complexity while overall unit costs of support have declined.
Conclusion: Investing in sustainability did not inhibit the expansion of PMTCT services. Such investments build long-term viability and enhance the prospect of a transition from reliance on donor support. We do not see a trade-off between scaling-up services and investing in sustainability. Rather, investment in sustainability is critical to service delivery scale-up and an effective, affordable use of PEPFAR resources.
E92 -Global recession and funding cutbacks: unleashing new challenges for sustainability

WEAE0106
The 'middle-income' curse: should global aid and treatment access decisions be based on national economic criteria?

K. Bhardwaj Independent Lawyer (HIV, Health and Human Rights), New Delhi, India
Background: Of late, donors adjusting their funding policies and companies and agencies working on treatment access initiatives have included World Bank set economic criteria in their decision making. What has been the impact of these new parameters in HIV funding and treatment access in so-called "middle income" countries?
Methods: The analysis looks at 1) Changes in Global Fund funding criteria; 2) Changes in price discounts by MNCs on ARVs and 3) The coverage of the Medicines Patent Pool licence announced in 2011. The situations of India, China, Argentina, Botswana and Namibia are examined in particular.
Results: The study finds that in all three cases, economic criteria are now playing a key role in determining the level and extent of assistance in relation to funding and treatment access. While the Global Fund funding criteria attempts to balance the economic status of a developing country with its disease burden, in the other two situations, the use of economic criteria appears motivated only by private interests in profits.
Conclusion: The use of GDP to determine access to HIV funding and treatment has gained acceptance in part because of the impression that the "middle income" status of a country indicates the ability of the country or its people to fund HIV programmes or HIV treatment. The GDP does not however reflect actual situations in country or the fact that several HIV programmes with marginalized communities are only in existence because of neutral, evidence based funding from the Global Fund. Such an approach undermines the very basis of Universal Access and is likely to be counterproductive both in the short term (with countries hesitating from adopting evidence of treatment as prevention into policies) and in the long term (as HIV programmes in these countries shrink or are shaped by national priorities and prejudices including in relation to marginalized groups).

THPDE0204
Patient outcomes in Lubumbashi, Democratic Republic of Congo after disruption in HIV care and treatment due to decreased Global Fund appropriations