CASE STUDY: IDENTIFYING BEST PRACTICES IN HIV SERVICE DELIVERY

Transcription

1 CASE STUDY: IDENTIFYING BEST PRACTICES IN HIV SERVICE DELIVERY

2 Table of Contents background 1 perpetual hiv counseling and testing (phct) at ampath 5 hiv counseling at ampath 5 fltr a model to Integrate hiv and ncd treatment and prevention 6 phct program costing 12 cost categories 14 costing limitations 18 benefits of the phct program 19 conclusion 24 1 tanzania: the rch platform approach

3 introduction AMPATH - Kenya The Academic Model Providing Access to Healthcare (AMPATH) is a collaborative partnership between Moi University School of Medicine and Moi Teaching and Referral Hospital in Eldoret, Kenya and a consortium of academic institutions in the United States led by Indiana University School of Medicine. It provides HIV care and treatment for clients in eight counties in western Kenya, a catchment population of about 3.5 million people. A core component of the partnership is AMPATH s Perpetual Home-based Counseling and Testing (PHCT) program, through which PHCT counselors are assigned to a designated location for two years or more and tasked with going door-to-door to counsel and test for HIV and link HIV-positive individuals to care and treatment services. The program trains counselors to use smartphones to capture individual and household information that is uploaded to the AMPATH Medical Record Systems (AMRS). As a result, all participating individuals within a community are enrolled and registered into a database, a process that greatly improves the quality and scope of care. The PHCT program s Find, Link, Treat and Retain (FLTR) approach provides community members with faster, more effective access to HIV care services through the support of the dedicated counselors and the database.

4 BACKGROUND Country Context and Situation Analysis Kenya POPULATION: 46 M Kenya covers some 580,000 square kilometers (225,000 square miles) and has an estimated population of about 46 million that has been growing at an annual rate of around 2.1 percent. It is a relatively young country, as more than 42 percent of residents are aged 14 or younger and just 2.6 percent are older than 65. Nearly 80 percent of Kenya s population lives in rural areas. Kenya s Health System TERTIARY HOSPITALS SECONDARY HOSPITALS PRIMARY HOSPITALS HEALTH CENTERS, MATERNITIES, NURSING HOMES DISPENSARIES/CLINICS > INTERFACE < COMMUNITY: VILLAGE/HOUSEHOLDS/FAMILIES/INDIVIDUALS COMMUNITY: VILLAGE/HOUSEHOLDS/FAMILIES/INDIVIDUALS According to recent data, the fertility rate is at 3.57 percent (births per woman) and life expectancy at birth for females is 65 years and 62 years for males. The country s infant mortality rate is 41 per 1,000 live births and the under-five mortality rate is 71 per 1,000 live births. 1 The maternal mortality ratio is 488 per 100,000 births. The national Ministry of Health (MoH) oversees the Kenyan health system and has responsibility for most financing and setting guidelines and standards regarding treatment and care. A recent decentralization initiative means that 47 county governments are now in charge of implementation. Nongovernmental organizations (NGOs), including faith-based organizations (FBOs), and private health facilities also play major roles. In 2008, the government operated 48 percent of the country s health facilities, with the NGO and private for-profit sectors accounting for 15 percent and 34 percent of all facilities, respectively. 2 The Kenya Essential Package for Health (KEPH) represents the integration of all health programs into a single package that focuses its interventions towards the improvement of health at different phases (cohorts) of the human development cycle. Kenya s health care system is structured in a step-wise manner so that complicated cases are referred to a higher level (see Figure 1 above). Level 1, the community level, is the foundation of service delivery priorities. Village health committees (VHCs) are organized in each community; through them, households and individuals are able to participate and contribute to the identification of their own health priorities and those of their village. The idea is that real ownership and commitment can be expected when communities are allowed to define their own priorities and services are subsequently provided to support those priorities. Higher up the chain, levels 2-3 promote KEPH activities related to promotive and preventive care, while levels 4-6 support curative and rehabilitative-related activities. Gaps in the system are filled by private and NGO-run units. Of the more than 9,700 facilities providing care in Kenya, FBOs operate 25 percent of the non-public facilities at levels 2, 3, and 4 (clinics, dispensaries, health centers, other hospitals, and primary hospitals). A 2009 memorandum of understanding (MoU) between the Kenyan government and faith-based and other NGO providers includes the engaged NGOs in the public sector s joint annual planning and monitoring and evaluation processes. In July 2013, the National Hospital Insurance Fund (NHIF) released the results of a study supporting the creation of a Household Insurance Subsidy Program (HISP). Among its projected benefits, the HISP is expected to enable provision of a comprehensive in- and out-patient health care service package for the very poor. Through this subsidy program, several thousand households across the country will have access to free health care services, which should result in earlier health care-seeking practices. The implementation plan for putting the HISP into action is currently being discussed and drafted by the MoH along with a number of other stakeholders. 1 See 2 Annual Health Sector Statistics Report (2008). Division of Health Management Information Systems, Kenya Ministry of Health. 1 AMPATH - Kenya

5 INCREASE IN PLHIV POPULATION 1.4M M 2013 HIV Epidemic in Kenya HIGHER HIV PREVALENCE IN WOMEN AND GIRLS 57% 1.1 MILLION CHILDREN ORPHANED BY AIDS 62,000 HIV-RELATED DEATHS IN 2012 KENYA IS EXPERIENCING A MIXED AND GEOGRAPHICALLY HETEROGENEOUS HIV EPIDEMIC, which includes both a generalized epidemic among the mainstream population and a concentrated one among certain key populations. HIV prevalence among adults aged 15 to 64 years decreased nationally from 7.2 percent, as measured in Kenya AIDS Indicator Survey (KAIS) 2007, to 5.6 percent in Prevalence estimates by county underscore the geographical variability of the HIV burden across the country, with rates ranging from an estimated high of 27.1 percent in Homa Bay County to below 0.2 percent in Wajir County. Ten counties have an estimated prevalence higher than the national average, while HIV prevalence in seven counties is estimated to be less than 2 percent. 3 Evidence shows a stabilizing epidemic among the general population and an elevated epidemic among the key populations. Various studies, including a 2010 mode of transmission study have revealed high HIV prevalence among a number of key affected groups, including sex workers, people who inject drugs, men who have sex with men (MSM), truck drivers and cross-border mobile populations. For example, HIV prevalence among MSM in Kenya is almost three times that among the general population, at 18.2 percent. 4 Also of note is that HIV prevalence for women was 6.9 percent compared with 4.2 percent of men, a finding that reflects the greater vulnerability of women to HIV transmission in Kenya. The number of people living with HIV is estimated to have increased from about 1.4 million in 2009 to 1.6 million in , with women and girls comprising about 57 percent of those individuals. An estimated 1.1 million children have been orphaned by AIDS and in 2012 nearly 62,000 people died from HIV-related illnesses. 6 Policy and Programmatic Response In 1999, the Kenyan government declared HIV/AIDS a national disaster, and established the National AIDS Control Council (NACC) under the Office of the President to coordinate a multi-sectoral national response. A five-year strategic plan was issued in The establishment of NACC and the National AIDS & STI Control Programme (NASCOP) in the Ministry of Health provided the national institutional framework for coordination of HIV activities. NACC has developed formal policies and guidelines to support program planning and implementation in such areas as antiretroviral therapy (ART), voluntary counseling and testing, sexual and reproductive health services for young people, and HIV education in primary school settings. 3 NASCOP (2012). Kenya AIDS Indicator Survey NASCOP (2012). Kenya AIDS Indicator Survey UNAIDS (2013). Global Report Available at 6 USAID (2013). Kenya HIV/AIDS Fact Sheet. See AMPATH - Kenya 2

6 HIV Treatment in Kenya Kenya set a target of having 1 million people on ART by the end of That ambitious goal was not met, but results have nevertheless been impressive. ART coverage has been scaled up rapidly over the last 10 years, from approximately 6,000 people on ART in 2003 to about 656,000 in 2013, an estimated 78 percent of eligible adults. Some of the steps taken to scale up ART include increased allocation of funding for purchasing antiretroviral drugs, development of ART delivery sites, deployment and training of health care workers, treatment literacy campaigns, efforts to reduce HIV-related stigma, increasing utilization of HIV testing and counseling, continuous update of the HIV treatment protocols in line with World Health Organization (WHO) recommendations, improved HIV commodity quantification and forecasting, and strengthened services for tuberculosis/hiv co infection. HIV counseling and testing has been a major focus of the response in Kenya. Multiple strategies have been adopted, including provider-initiated testing and counseling; outreach testing and counseling events; home-based counseling and testing; and integration of HIV testing and counseling with antenatal care (ANC) services and those treating sexually transmitted infections (STIs) and focused on sexual and reproductive health. The Kenya Demographic Health Survey (KDHS) and KAIS findings show an increase in the number of people tested and receiving their results annually in the last five years. A longstanding concern, however, is that women are far more likely to be tested than men. Despite an increase in access to HIV treatment for children, the overall coverage for children remains extremely low. According to NASCOP, the average time to link a baby to care in Kenya is three years a loss of valuable time during which many of these children die. Kenya scaled up early infant diagnosis (EID) to reach 45 percent of infants annually by Yet this level still remains low and there is need for enhanced scale-up strategies to be put in place. 78% NOW RECEIVING TREATMENT 45% EARLY INFANT DIAGNOSIS Prevention of Mother-to-Child Transmission of HIV (PMTCT) NASCOP has taken several actions to expand and strengthen PMTCT interventions in the country. In 2000, a National Technical Working Group (TWG) on PMTCT was formed. The TWG, co-chaired by NASCOP and the Division of Reproductive Health within the MoH, coordinates implementation and provides technical support to the National PMTCT Programme, launched in Between 2009 and 2011, infant infection rates were reduced by 50 percent. However, the 2013 UNAIDS Progress Report found that PMTCT coverage in Kenya fell by 20 percent in , a decline attributed to disruptions in health services due to a series of strikes by doctors and nurses demanding an increase in health care funding. In 2013, the government of Kenya adopted the WHO-recommended Option B+ regimen, a step that replaced the administration of the drug AZT (option A) to HIV-positive expectant mothers with standard triplecombination ART regimens regardless of CD4 count. 4,500 PMTCT SITES 5X INCREASE FROM MILLION PREGNANT WOMEN COUNSELED AND TESTED FOR HIV 3.5X INCREASE FROM 2005 Some 4,500 PMTCT sites were operating in 2012, an increase from 926 in The number of pregnant women counseled and tested for HIV over that period rose sharply as a result, from around 318,000 to 1.2 million in Kenya s Progress towards Elimination of Mother-To-Child Transmission (EMTCT) of HIV (2013). Mudany MA, Sirengo M, Nganga LW, Gichangi A. Sex Transm Infect 2013;89:A373-A374 doi: /sextrans AMPATH - Kenya

7 Funding for HIV and AIDS in Kenya 70% OF FUNDING COMES FROM EXTERNAL SOURCES 17% OF FUNDING FROM GOVERNMENT OF KENYA 13% OF FUNDING FROM PRIVATE KENYAN SOURCES HIV/AIDS services continue to be largely financed by development partners. From 2009 to 2013, the total spending on the HIV response increased from Kshs63 billion ($705 million) to Kshs72 billion ($806 million) respectively. 8 External funding sources contributed over 70 percent of total spending on HIV/AIDS, with the Kenyan government providing about 17 percent. Private or household sources accounted for the remaining 13 percent of funding. The amount of funding provided from the national public budget increased from KES 8 billion ($90 million) to KES 13 billion ($146 million) during this period. In recognition of the fact that development partner resources for HIV/AIDS were to be reduced starting in , the Kenyan government organized a steering committee to explore alternative Health Sector Challenges 1.69 HEALTHCARE WORKERS PER 1,000 PEOPLE modalities of financing, especially focusing on mobilization of domestic resources. Three innovative approaches were identified, including: increasing central government allocation to the health sector to meet the 2001 Abuja Declaration, under which African governments pledged to allocate 15 percent of national budgets to health each year; introducing a special tax on air travel passengers; and covering ART services through the National Hospital Insurance Fund. These measures will be further articulated as the 47 county governments begin the process of budget preparations and allocations. In 2010, the newly adopted constitution mandated the decentralization of funding and service provision and created 47 County governments that were established after the 2013 elections. Among the objectives of the devolution of government are to promote social and economic development and the provision of proximate, easily accessible services throughout Kenya, according to Kenya s 2010 Constitution. 20% FILL RATES IN RURAL DISPENSARIES % FILL RATES IN DISTRICT HOSPITALS While there is an absolute shortage of health care workers, the provider-population ratio of 1.69/1000 (for all cadres of providers) is relatively high compared with other countries in the region. The most pressing problem is the drastically unequal distribution of workers between urban/rural areas, across regions, and by level of care. Rural dispensaries have 20 percent fill rates of their nursing positions, while district hospitals have 120 percent fill rates. Approximately 25 percent of the health budget for the entire public sector is taken up by the two referral hospitals. 8 Kenya AIDS Response Progress Report 2014, National AIDS Control Council, March 2014 AMPATH - Kenya 4

8 Perpetual HIV Counseling and Testing (PHCT) at AMPATH AMPATH delivers primary health care services in western Kenya in partnership with the MoH, including care and treatment for more than 158,000 people living with HIV in over 500 clinical facilities in both urban and rural environments. AMPATH reaches over 200,000 people per year through its home-based testing and counseling services. Also, it supports PMTCT services across 100 sites in western Kenya. On a monthly basis HIV testing is provided for 99 percent of all mothers who attend antenatal care clinics in AMPATH s catchment area (approximately 3,000 expectant mothers each month). Pregnant women who are diagnosed with HIV are referred to one of AMPATH s HIV treatment clinics and over 90 percent of women are successfully referred into care. AMPATH also provides care and treatment for cancer, cardiovascular and pulmonary diseases, diabetes and mental illness. Health care services are further supported through provision of food security, micro-financing and agricultural programs (supporting over 14,000 people living with HIV), legal services, work with orphans and vulnerable children and infrastructure development. AMPATH s electronic health information systems including the AMPATH medical record system (AMRS), its Open Medical Records System (OpenMRS) platform, and electronic clinical decision support systems have enabled decentralization of services and promote task-shifting among health care and community-based providers, leading to high-quality, cost-effective care delivery. HIV Counseling and Testing at AMPATH In 2007, 85 percent of residents in AMPATH s catchment area of then 1.8 million people still did not know their HIV status. In an effort to increase knowledge about HIV status among people in the region, AMPATH began a HIV homebased counseling and testing (HBCT) pilot project in which community health workers sensitized communities and 5 AMPATH - Kenya

9 counselors went door-to-door offering HIV education, counseling, and testing to residents in their houses. The pilot HBCT program began in Kosirai Division in the North Rift Valley with a predominantly rural population of approximately 36,000 people across 9,000 households. AMPATH trained and deployed local residents known as community mobilizers who introduced the concept of HBCT to residents. The counselors used handheld devices equipped with a global positioning system (GPS) to collect data from their visits that could be directly uploaded into a database system. During the pilot, 95 percent of the total eligible persons, almost 20,000 people, accepted counseling about HIV. Of those, 96 percent received HIV tests. All who tested positive were referred to AMPATH clinics for care, and those who tested negative received counseling on how to avoid HIV infection. Based on the success of the pilot, AMPATH expanded the home-based testing and counseling program to other parts of the catchment area. A 2010 study published in the Journal of the International AIDS Society 9 examined the impact of HBCT on the clinical status of newly enrolled patients at AMPATH s treatment centers. The study analyzed results from 19,552 adults newly enrolled in care between August 2008 and April 2010, comparing their point of entry, including 946 from HBCT, 8,073 from provider-initiated testing and counseling (PITC), 272 from tuberculosis clinics, and 10,261 from voluntary counseling and testing (VCT) sites. The highest proportion of HIV-positive pregnant women enrolled in care had been tested through the HBCT program (11% of pregnant women linked to care). The lowest percentage had been tested at VCT sites and at tuberculosis clinics. Compared with the other points of entry, HBCT provided the highest percentage (24 percent) of individuals in discordant couples with effective linkage to care. Patients who tested HIV positive through the HBCT program had the highest median CD4 cell count at enrollment (median 323 cells/mm 3 ), followed by VCT and then PITC. In addition, 86 percent of patients with WHO stage I/II disease came to care through HBCT, compared with 62 percent in VCT, 54 percent in PITC, and 21 percent in the tuberculosis clinic. Overall, the study concluded that HBCT was effective in gaining patient acceptance for testing and linking patients to care earlier in the course of HIV disease. FLTR A Model to Integrate HIV and NCD Treatment and Prevention Although AMPATH s initial focus was providing HIV treatment and prevention services, the program has expanded to address other communicable and non-communicable diseases (NCDs). There is a high burden of NCDs in Kenya, but in general the health system has not been well structured to address them. Seeing HIV as a chronic disease, AMPATH determined that it could expand its services to include other chronic illnesses, providing earlier diagnosis, treatment and prevention and improve overall engagement in health care. AMPATH developed a model called FLTR (pronounced flitter ): case Finding, Linkage, Retention and Treatment. The case finding is done by incorporating blood pressure and random blood sugar tests into the existing HIV counseling and testing program. Treatment systems utilize clinics initially established to provide HIV care, but whose capacity has been broadened to provide care for other chronic diseases. The use of lower-level health facilities for confirmation of diagnosis, triage and care for early disease was deemed most appropriate to address the high burden of NCDs. The need for an efficient referral system, critical for successful programs addressing NCDs, led to the placement of centers of excellence 9 Kimaiyo, S et al (2010). Know Your epidemic: Scaling-up Home-based HIV Counseling and Testing in Western Kenya. Journal of the International AIDS Society (2010). AMPATH - Kenya 6

10 (COEs) at the referral hospital in Eldoret and mini-coes at district-level facilities. Multiple stakeholders were involved in the program design, with government being a major partner for sustainability. The key innovations of the FLTR model include: Demand Creation: Actively seeking patients out rather than waiting for them to come into care, leading to earlier diagnosis and treatment. Task-shifting: The standard policy has only physicians treating NCDs. In the FLTR model, nurses and clinical officers are also key treatment and care providers. Data collection: Moving paper to electronic medical records provides a significant opportunity for improved quality control and feedback. Improved technology: Use of hand-held devices provides an opportunity for real time decision support and eases work load. FIGURE 1 LAYERED PROGRAMS WITHIN AMPATH: POTENTIAL FOR INTEGRATION COMMUNITY-BASED STAFF AND THEIR PRIMARY ROLE FACILITY-BASED HIV CARE DECREASED INCIDENCE HIV > 50% MOTHER TO CHILD TRANSMISSION <5% IN CATCHMENT DECREASE MATERNAL, INFANT AND UNDER 5 MORTALITY > 50% LEVEL 1-6 CARE FOR CANCER, HYPERTENSION, LUNG DISEASE AND MENTAL DISORDERS TRADITIONAL AMPATH PHCT/FLTR (FIND, LINK, TREAT, RETAIN) PRIMARY HEALTH CARE NCD OVC / COMMUNITY HEALTH WORKER (CHW) PERPETUAL HOME-BASED COUNSELING AND TESTING (PHCT) PHCT, CHW AND COMMUNITY HEALTH EXTENSION WORKER (CHEW) PHCT, CHW AND CHEW Building on the success of the HBCT program, AMPATH developed a new model called Perpetual Home-based Counseling and Testing (PHCT) in furtherance of the objective to address other chronic health care needs through its HIV testing services. Rather than travel from town to town, PHCT counselors are assigned to their own designated location indefinitely. PHCT counselors counsel and test for HIV, screen for tuberculosis, diabetes and elevated blood pressure, treat children for intestinal worms, provide information on safer sex practices, and distribute condoms. They take responsibility for linking to care all those who might need it, including those who test positive for HIV. Household data are recorded on a handheld device, which is then synched with the AMPATH electronic medical record system. By having the counselor remain within the specific location for an extended period of time, the PHCT approach seeks to establish ongoing relationships with community members, which, in turn, builds trust and 7 AMPATH - Kenya

11 improves engagement in health services. Through this approach, AMPATH is working to improve reliability of HIV prevalence data, monitor the impact of its interventions, and reduce the loss to follow-up. FIGURE 2 ELECTRONIC MEDICAL RECORD BASED ON HANDHELD ANDROID PHONES COMMUNITY HEALTH WORKERS (CHWS) MAKE HOME VISITS OR CLIENT VISITS LOCAL DISPENSARY CHW SCANS PATIENTS MEDICAL ID CARD WEB-BASED NETWORK SERVER EVENTUAL CONNECTIVITY TO ALL AREAS WITHIN OUR CATCHMENT AREA CHW PERFORMS BASIC ASSESSMENT BASED ON DECISION SUPPORT IN THE PHONE BASED EMR DATA ENTERED DIRECTLY INTO THE PHONE DATA FROM PREVIOUS VISITS AVAILABLE FOR DECISION SUPPORT In areas where AMPATH s door-to-door testing has already occurred, the PHCT counselor systematically offers retesting in his or her assigned catchment area to individuals who previously tested negative at a rate of one third of the total population within the catchment area each year. In the settings where door-to-door testing has not yet occurred (approximately 1.4 million persons within the overall catchment area), PHCT counselors will test the entire population within their designated areas over a period of two years. The ultimate goal of this effort is to demonstrate for Kenya and beyond that this approach to HIV testing and counseling can decrease the incidence of HIV infections by more than 50 percent, as well as reduce vertical transmission to less than 3 percent in a large population. In its 2013 progress report, AMPATH reported that over 350,000 clients received HIV testing in that year. A total of 127,064 people (58,475 males and 68,589 females) were tested for HIV during the year through PHCT and mobile VCT approaches. An added advantage to the PHCT approach is that other services were provided alongside HIV testing and counseling, including tuberculosis screening, deworming of children aged 2-5 years, and direct linkage to care. Counselors also provided demonstrations of how and why to use condoms, and collectively distributed a total of 975,215 male condoms and 224 female condoms. Provider-initiated counseling and testing uptake in 2013 improved from 156,314 tested in 2012 to 199,766 (of whom: AMPATH - Kenya 8

12 males, 89,618 and females, 110,148). The number of people newly diagnosed with HIV dropped from 9,701 to 9,167 in the year. Also noted in the AMPATH 2013 progress report was the drop in prevalence among those tested, from 5.6 percent in 2012 to 4.6 percent in Another optimistic finding was that more couples turned up together for testing: 6,232 couples in 2013 compared with 5,706 in In addition, 893 health care workers tested for HIV in 2013, compared with 627 in An In-Depth Look at the PHCT Program in Two Catchment Areas The Pangaea/CHAI project team visited the AMPATH and Moi University Teaching Hospital in Eldoret in May The site visit included meetings with representatives from the Ministry of Health and AMPATH and hospital staff based in Eldoret as well as field visits to meet with HIV counselors in the nearby Kosirai catchment area. In addition to the interviews, the team reviewed routine PHCT monitoring and evaluation data collected by AMPATH and commodity data to better understand resource needs for the program. The site visit provided an opportunity to see the PHCT program in action, speak with personnel and volunteers about their work, and work with AMPATH staff to develop costing estimates for the program case study. The cost analysis focused specifically on two of the PHCT program s twelve catchment areas Kosirai, which is in Nandi County, close to the city of Eldoret, and Bunyala, which lies in the western county of Busia, on the border with Uganda. (Although Bunyala was included in the costing analysis, the team did not make a site visit specifically to the catchment area.) The results of this work are described below. Overview Of Patient Population Bunyala and Kosirai have similar population sizes, each with around 68,000 individuals living in the catchment area. Kosirai is close to AMPATH s base in Eldoret and AMPATH has a longstanding relationship with communities throughout the Kosirai catchment area. AMPATH has been providing health care services in the area for decades, and started its original home-based HIV testing program in the area in 2007, before launching the PHCT program in Access to HIV services is strong in the area, and overall prevalence is low at less than 1 percent. Bunyala, by contrast, has an adult HIV prevalence of 8.3 percent and the PHCT program was only started in The catchment area is located nearly 200 kilometers (125 miles) west of Eldoret and sits on the shores of Lake Victoria near the border with Uganda. Through 2013, all AMPATH PHCT catchment areas were on a three-year testing cycle, where the program aimed to reach every household in the catchment area once in a three-year period. However, the PHCT program is currently transitioning to a more targeted approach, where high prevalence catchment areas will be reached every year. (Although this costing analysis only captured 2013 data, the cost implications of these changes need to be considered as well as how costs are likely to change over time under this new approach.) 9 AMPATH - Kenya

13 TABLE 1 Summary of Population and HIV-Related Information, Kosirai and Bunyala PATIENT NUMBERS KOSIRAI BUNYALA Total Catchment Population 68,000 67,669 Total Catchment Targeted in ,667 22,556 # of Individuals Reached % Reached of Total 2013 Target # of Total Households Reached Average # of Individuals Per Household # of Individuals Eligible % Eligible of Total Reached # Tested or Previously Known HIV+ % Tested or Known Status of Reached # Tested % Tested of Total Eligible # HIV+ (including Previously Known) Positivity Rate # Total HIV+ and Linked % of HIV+ Patients Linked # Newly Tested HIV+ and Linked % of Newly Tested HIV+ Patients Linked Improving Linkage to Care 3,223 14% ,010 62% 2,006 62% 1,981 99% 32 2% 18 56% 7 100% 11,640 52% 2, ,258 62% 7,769 67% 7,102 98% % % 44 36% AMPATH has been engaged in HIV home-based counseling and testing since Testing through this mechanism has identified a large number of HIV-positive individuals, many of whom may not have learned their HIV status until they developed symptoms of AIDS. These are people who do not necessarily have a health concern that prompted them to go to a facility to either ask to get tested or have a healthcare provider offer HIV testing as part of routine care. Often they are reluctant to engage in care following their test result. In an interview, an AMPATH staff person framed the problem clearly, People here really struggle. The caregiving burden is high for most adults. The cost of living over the past few years in Kenya has increased substantially while wages have not. But we collectively expect all these people, who for the moment look and feel fine, to leave their work and come to an HIV clinic within 30 days, 6 months, even a year; to come to terms with their diagnosis and that of family members in many cases; and disclose to friends and family. Often when they go to a facility the queues are long, the clinicians aren t particularly friendly or even often helpful, and they ask a lot of rude or embarrassing questions. To me it seems like a big ask. From my perspective and that of AMPATH s, it s a call to action. It s a call for culturally appropriate patient-centered care. It s a call for HIV care providers to make care easily accessible, affordable, and effective, to enable people living with HIV particularly if they are stable either on or off treatment to be in care. To address these challenges, AMPATH is working to improve opportunities for on-going client engagement through expanded community-based service delivery. These include: GISE groups - Group Integrated Saving Enterprise (GISE) groups, serve as cooperative saving and lending groups, as well as HIV support groups. The GISE aspect acts as a major incentive for people to attend. There they receive information and support, including encouragement to go be assessed at an HIV clinic. AMPATH will soon provide ART to stable patients through these groups, a form of community-based low risk express care. Stable patients not on ART can also be seen in these groups. Care Navigators - Care Navigators are people living with HIV who have a record of good adherence to treatment and care. The Care Navigator s role is to receive people at the clinic who have tested positive through either PITC or PHCT. The Care Navigators provides additional counseling, gives them a tour of the facility and an orientation to HIV disease, care, and treatment. They act as role models, treatment educators, and adherence counselors. Peer Navigators - AMPATH will also test a new cadre called Peer Navigators, starting with a pilot in Eldoret of HIV-positive street youth, in which HIV prevalence is very high 15% among girls and about 6% overall. The AMPATH prevention team will assign a PHCT counselor who will specifically target street youth and street families. Street youth who either are known to be positive or who newly test positive will be invited to link up with a navigator who is aged between18-30, HIVpositive and on treatment, essentially clean and sober, who has been connected to the streets. The navigators will work to help other positive street youth engage with and adhere to care and treatment. AMPATH - Kenya 10

14 2013 Bunyala Patient Numbers In 2013, the PHCT program reached 11,640 individuals in Bunyala (nearly 20 percent of the total population), and tested 7,221 for HIV. (Although some of the individuals who were reached refused testing services, it is also important to note that children under 13 who are reached are only eligible for testing services if their mother is HIV-positive, missing or deceased.) Overall, 788 individuals were found to be HIV-positive. This included a group that already knew their HIV status but were not linked to care. Of these 788 people, PHCT program successfully linked 87 percent into care. However, only 36 percent of those testing HIV positive for the first time were successfully linked to care Kosirai Patient Numbers In 2013, the PHCT program reached 3,223 individuals (about 5 percent of the total population) and tested 1,983 individuals for HIV. Of the 32 individuals who were found to be HIV-positive, the program successfully linked 56 percent into care. Out of these 32 individuals, a number were already aware of their status and linked to care, contributing to the seemingly low rate of linkage to care. FIGURE 3 Kosirai Patient Targets Under Optimal Personnel Efficiency 30,000 25,000 20,000 15,000 10,000 5,000 0 PHCT PATIENTS REACHED NUMBER TARGETED BY PHTC EACH YEAR (22,667 PPL/6,869 HOUSEHOLDS) NUMBER REACHED WITH CURRENT COUNSELORS AT OPTIMAL EFFICIENCY (12,375 PPL/3,760 HOUSEHOLDS) NUMBER CURRENTLY REACHED (3,223 PPL/967 HOUSEHOLDS) The low prevalence in Kosirai significantly increases the costs of HIV service delivery in the catchment area. However, utilization rates of the program s counselors also appear to have an impact on costs in Kosirai. The PHCT program aims to have counselors reach 2-3 households each day. In Kosirai, there were five counselors assigned to the program, and they cumulatively reached 967 total households in This indicates that each counselor reached an average of 193 households during the costing year, or less than one per day (assuming 250 working days in a year). While not all working days can be expected to be spent on active outreach, the target of 2-3 households per day suggests that higher utilization could be achieved. (In contrast, seven counselors were deployed in Bunyala in 2013, who cumulatively reached 2,778 households. This translates to an average of 397 households per counselor per year, which corresponds to an average of 1.6 households per counselor per day.) In reaching fewer households, the Kosirai program is ultimately reaching and testing fewer individuals, so the costs 11 AMPATH - Kenya

15 of the program are spread across a smaller denominator. Preliminary analysis suggests that tripling the average number of households reached in Kosirai, to an average of 2.4 per day, would reduce costs per person tested by over 50 percent. However, even with more optimal efficiency, it appears that additional resources may be needed to hire a greater number of counselors to meet Kosirai s annual target of reaching everyone in the catchment area once every three years (see inset). PHCT program costing: Analysis of two catchment areas over a one-year period To better understand the resource implications of the PHCT program, the study team reviewed costs associated with the program in Bunyala and Kosirai. These catchment areas were chosen for their differences rather than their similarities. As mentioned above, Bunyala has an adult HIV prevalence of 8.3 percent and the PHCT program in the area focuses solely on HIV counseling and testing. Kosirai, on the other hand, has low adult HIV prevalence, at less than 1 percent, and in 2012, diversified its PHCT service delivery package to include testing for certain NCDs in addition to HIV testing. For both catchment areas, HIV-related costs were collected over a 12-month period (January through December 2013) across five main cost categories: laboratory costs, including tests, reagents, lab equipment and clinical supplies; personnel costs, including staff salaries and compensation for work-related travel; other costs, which encompasses non-lab commodities (e.g., condoms) and general program running costs; programming costs, which includes software development/maintenance; and demand generation, which includes community outreach to maintain and improve acceptance of the PHCT program. In Kosirai, also collected were costs associated with the NCDs being assessed and addressed through the program. This primarily included relevant laboratory tests and equipment and the incremental costs of additional personnel time necessary to execute testing and education regarding NCDs on top of the program s core HIV component. Findings from the analysis indicate that AMPATH s home-based HIV testing program does not require significant resources per person tested. At a basic level, the cost of the program ranged from $5.46 per person tested in Bunyala to $15.41 per person tested in Kosirai. Costs in Kosirai are higher because fewer people were tested. The primary drivers of cost were personnel, accounting for 70 percent to 86 percent of overall costs, and to a lesser extent, lab commodities and equipment. Other costs were largely driven by condom provision as part of the PHCT program s community outreach and non-lab equipment. In both catchment areas, infrastructure and running costs of the program were very limited, as were software development and demand generation costs. When considering cost per HIV-positive person identified, however, the cost differences between the two catchment areas is much larger: just $50 in Bunyala, but $955 in Kosirai, where, given low prevalence, many more people have to be tested to identify one HIV-positive individual. Cost per HIV-positive person includes both those who newly test positive and those who already know their status and are not yet in care. In all catchment areas, the program has encountered a significant number of HIV-positive individuals who are not in care, and this population accounts for the majority of the program s successful linkages. AMPATH - Kenya 12

16 TABLE 2 High-Level Cost Breakdown by Catchment Area BUNYALA (hiv prevalence 8.3%) Program-Level Costs $5.46 PP TESTED $1.21 LABS $3.84 PERSONNEL Cost PP Tested $.12 $.11 BUNYALA - COST PER PERSON TESTED $.18 OTHER Cost PP HIV+ (Tested or ID) Cost PP HIV+/ Linked Primary Cost Categories Personnel $3.84 $35.23 $40.47 Labs $1.21 $11.06 $12.70 Other Costs $0.18 $1.66 $1.91 Supplementary Cost Categories Programming Costs $0.11 $1.00 $1.15 Demand Generation $0.12 $1.10 $1.27 Total Cost $5.46 $50.05 $57.50 FIGURE 4 DEMAND GENERATION PROGRAMMING COSTS kosirai (hiv prevalence < 1%) Program-Level Costs $15.41 PP TESTED $1.52 LABS $13.23 PERSONNEL Cost PP Tested KOSIRAI - COST PER PERSON TESTED $.01 $.24 Cost PP HIV+ (Tested or ID) $.42 OTHER Cost PP HIV+/ Linked Primary Cost Categories Personnel $13.23 $ $1, Labs $1.52 $93.90 $ Other Costs $0.42 $26.09 $46.38 Supplementary Cost Categories Programming Costs $0.24 $14.74 $26.21 Demand Generation $0.01 $0.46 $0.82 Total Cost $15.41 $ $1, DEMAND GENERATION PROGRAMMING COSTS AMPATH captures data on all patients who are reached by the PHCT program and successfully linked into care (as defined by registration at an ART facility). This allowed for calculation of the cost per HIV-positive person (those newly tested and those with previously known status) linked to care. These outcome-adjusted costs are important because they begin to show resource intensity required to link HIV-positive individuals into the treatment and care cascade, and demonstrate the variations in costs across areas with substantially different HIV prevalence. While both catchment areas door-to-door counselors were able to test more than 60 percent of those individuals reached and over 98 percent of those eligible the low prevalence in the Kosirai catchment area means only 32 HIV-positive people were identified in 2013 (as opposed to 788 in Bunyala). As mentioned earlier, this led to the large difference in cost per person who tested positive for HIV. In addition, a greater proportion of people who were identified as HIV-positive were subsequently registered for treatment in Bunyala (87 percent) than in Kosirai (56 percent). This likely reflects the fact that in Kosirai the HIV program is more established and a greater proportion of people living with HIV are already linked to care. The lower linkage rates in this case therefore should not be read as poor performance, but rather as a confirmation of the greater incremen- 13 AMPATH - Kenya

17 tal effort required to identify and link HIV-positive people over time. This effort translates to a much higher cost per person who tests positive and is linked: $58 in Bunyala compared with $1,698 in Kosirai. The full costs and patient numbers for each catchment area are described below in depth. At a high level, these findings confirm and help quantify that HIV prevalence and testing yield rates that may significantly affect the resources required to launch and maintain a home-based testing program. The resource intensiveness and cost-effectiveness of different testing strategies are key points of consideration as HIV programs contemplate the inclusion of home-based HIV testing in their suite of interventions. COSTING METHODOLOGY Through the analysis conducted for this report, costs were captured relating to the home-based HIV testing program, including all services necessary to confirm and communicate test results to individuals and then link HIV-positive patients into care. Costs of demand generation necessary for the program to operate at the community level are also included. Data were collected for a 12-month period (January- December 2013) to capture the annual costs of maintaining the program; also considered, however, were costs that were critical to starting up the program in A top-down costing approach was primarily used for this analysis; it consisted of comprehensively collecting all the resources used for the PHCT program in 2013 and then allocating a portion of them to the PHCT program and then to the specific catchment areas for the analysis (i.e., Kosirai and Bunyala). This total cost was then divided by the total number of people tested to arrive at a unit cost cost per person tested. Outcome-adjusted unit costs were also calculated to show the cost per person testing positive for HIV and cost per HIV-positive person linked to care. It is important to note that this case represents the costs of a home-based testing and linkage program in the context of a well-established, partner-led health care system with strong resource support and an extensive electronic medical record (EMR) system. The cost structure, especially starting costs, may be different for programs that do not have the same level of infrastructure. cost categories Personnel Personnel are the primary driver of costs under the PHCT program, accounting for 70 percent of total costs in Bunyala ($3.84 per person tested) and 86 percent of total costs in Kosirai ($13.23 per person tested). This includes direct personnel costs for staff who provide counseling or clinical care to patients including counselors, their immediate supervisors (who are responsible for EID and ELISA sample collection), and lab technicians. It also includes indirect costs for program staff such as data managers and shared administrative support. Direct personnel made up between 74 and 78 percent of overall personnel costs in the catchment areas in Program-level supervision costs In addition to the personnel costs around service delivery, there are a number of Eldoret-based AMPATH staff who support overall supervision of the program. They include program administrators, program managers and techni- AMPATH - Kenya 14

18 cal advisors. For the analysis, the costs of compensation for program-level supervision staff were captured as well as other key costs that support the role of these personnel such as travel costs for supervisory visits and running costs related to the PHCT administrative office in Eldoret. In 2013, program-level supervision costs totaled $0.63 per person tested in Bunyala and $1.95 per person tested in Kosirai. These costs are not included in the main cost breakdown, as supervision support is highly dependent on health system capacity and may not be necessary for all programs that are considering implementation of this model. It is also important to note that program-level supervision is only one part of the total above facility costs that go into a program. This case study has not looked at the full landscape of such costs that support AMPATH s PHCT program. Labs Individuals reached by the HIV component of the PHCT program have access to five types of lab tests. All consenting individuals are tested in the home with a rapid diagnostic kit (except for HIV-exposed infants, who provide a dried blood spot [DBS] sample for EID testing.) When individuals test positive in the home, they are provided with confirmatory testing through a rapid diagnostic test (in this case UniGold HIV 1/2). ELISA testing is done if there is a discrepancy in the results. Individuals were also screened for tuberculosis, and counselors are trained to take sputum samples for microscopy testing at the nearest tuberculosis diagnostic center. Lastly, HIV-positive women are offered pregnancy tests. Lab costs totaled $1.21 per person tested in Bunyala and $1.52 per person tested in Kosirai. However, lab costs between the two catchment areas differ dramatically when looking at outcome-adjusted unit costs (see Figure 5). Costs were FIGURE 5 Breakdown of Lab Costs by Catchment Area BUNYALA (hiv prevalence 8.3%) Program-Level Costs Cost PP Tested Cost PP HIV+ (tested or ID) Cost PP HIV+ / Linked $200 KOSIRAI BUNYALA $ Tests $1.01 $9.22 $10.59 Equipment $0.07 $0.66 $0.76 Clinic Supplies $0.13 $1.17 $1.35 Labs $1.21 $11.06 $12.70 kosirai (hiv prevalence <1%) Program-Level Costs Cost PP Tested Cost PP HIV+ (tested or ID) Cost PP HIV+ / Linked Tests $1.11 $68.64 $ Equipment $0.12 $7.20 $12.80 Clinic Supplies $0.29 $18.06 $32.11 Labs $1.52 $93.90 $ TOTAL LAB COSTS (USD) $150 $100 $50 0 $ % $1.52 COST PER PERSON TESTED $11.06 $93.90 COST PER PERSON HIV+ $12.70 COST PER PERSON HIV+/LINKED 15 AMPATH - Kenya

19 primarily driven by commodities, including rapid test kits, and EID and ELISA tests although the overall cost category also includes lab equipment and clinical supplies (e.g., gloves for counselors, medical disposal bags, etc.). Sample transportation costs are not included in the overall lab cost calculations. Some sample transportation costs are accounted for within personnel calculations as counselors and supervisors are given a travel stipend to cover transport to different communities and to sample collection sites. Additional sample transportation costs are covered through existing AMPATH sample transportation infrastructure, which supports programs much broader than the PHCT program. Other Costs In addition to the main lab and personnel cost categories, data were collected on both commodity and non-commodity other costs that were essential to the HIV testing program. In 2013, these other costs totaled $0.18 per person tested in Bunyala and $0.42 per person tested in Kosirai. On the commodity side, the cost of condoms provided through the PHCT program was included. In 2013, counselors disseminated 15,000 condoms in Bunyala (~2 per person tested or identified as HIV-positive) and 10,000 condoms in Kosirai (~5 per person tested or identified as living with HIV). Given the high volume, condom provision accounted for nearly half of overall other costs in Bunyala and Kosirai, although AMPATH did note that some of the total volume may have been supplied at no cost by the MoH. When other programs are looking to PHCT as a model for implementation in other settings, it could be important to consider whether or not these condom costs are essential to the success of the program. Given the reach of many HIV prevention programs, condom dissemination may be ongoing within communities already. Condom provision therefore may not need to be factored into a community-based testing program, or potentially it could be done at a more limited level. Without condom provision, total other costs are reduced to $0.10-$0.23 per person tested. AMPATH - Kenya 16

20 Also calculated were the annual non-commodity other costs, including expenses related to non-lab equipment, running costs (e.g., electricity, fuel, etc.), and relevant buildings. Equipment is a particularly important consideration for the PHCT program, given the heavy use of technology used to track patient data through smartphones and link information collected to AMPATH s overall EMR system. However, when amortized and spread over the full beneficiary population, non-lab equipment costs were very limited on a per patient basis. Programming Costs In addition to the equipment costs related to the handheld-device-based data input system (discussed above), the costs of programming the software used by the program to track outcomes across each catchment area were included in the analysis. Overall, these programming costs were very low at $0.11 per person tested in Bunyala and $0.24 per person tested in Kosirai. AMPATH used open-source software for the PHCT program, which limited overall costs. The overall programming costs also included the development costs for the data-capture forms used by counselors to input information that was then uploaded or transferred to the broader EMR database. Also captured and included in overall programming costs were costs necessary for synchronization of data throughout the year. Demand Generation A key component of AMPATH s work in communities is sensitization and demand generation work. As the PHCT program was getting started, AMPATH worked with community leaders to discuss the program and determine the best way to work with communities. However, there is also an ongoing demand generation component to ensure acceptance of PHCT services in each catchment area. AMPATH personnel work with local leaders to identify and overcome challenges around the understanding and uptake of PHCT services. The costs of various demand generation meetings/consultations within the costing year were captured, although these were negligible parts of overall costs. (In Bunyala, demand generation costs were $0.12 per person tested and in Kosirai such costs were only $0.01 per person tested.) Testing For Non-Communicable Diseases (NCDs) The Kosirai catchment area had implemented an NCDs service delivery component to the broader PHCT program prior to Thus although the HIV testing program was the focus of this analysis, the costs associated with the NCDs component were collected in Kosirai as well in order to demonstrate the potential costs associated with building upon a community-based HIV testing program to reach communities for diverse services. In Kosirai, patients who are reached by the PHCT program have access to all the HIV services described in this report. In addition, counselors provide education about diabetes, hypertension and heart disease and discuss lifestyle changes that could reduce risk for many NCDs; furthermore, they offer blood pressure testing and glucose monitoring to assess risk for hypertension and diabetes. 17 AMPATH - Kenya