Scale-up of assisted partner services in Kenya: assessing linkage to care, integration and costs

Abstract

Despite marked progress in achieving universal 95-95-95 targets, gaps still exist, especially in improving individual awareness of HIV status. According to the 2018 Kenya Population-based HIV Impact Assessment (KenPHIA) report, approximately 79.5% of individuals were aware of their status, 96.0% were on antiretroviral therapy, and 90.6% were virally suppressed. Men were less likely to be aware of their HIV status compared to women (72.6% vs 82.7%) necessitating HTS strategies to effectively target this ‘hard-to-reach’ group. HIV assisted partner services (aPS), or healthcare provider supported notification of sex partners to newly diagnosed HIV-positive individuals, have been used to bridge this gap in HIV testing, and have been shown to be safe, effective, and cost-effective. aPS was scaled up within the national HIV testing services (HTS) program in Kenya in 2016 after World Health Organization (WHO) recommended the intervention. Our objective was to assess linkage to care, integration, and costs of scaling up aPS within the national HTS program in Kenya. In the first study, we used data from nine facilities randomized to receive immediate aPS in a cluster-randomized trial conducted in Kenya. We estimated linkage to care - defined as HIV clinic registration - and ART initiation separately for index clients and their sex partners. We found that only two-thirds of newly diagnosed HIV-positive sex partners, and known HIV-positive sex partners not enrolled in care at study enrolment, linked to care after receiving aPS. However, once linked to care, ART initiation was high (>85%) regardless of whether the participant was an index client, newly-diagnosed or known HIV-positive sex partner not previously linked to care. We recommend that HIV aPS programs optimize HIV care for these individuals, especially those who are younger and single. In the second study, we used an integrated conceptual framework to assess the extent of aPS integration, institutionalization, and sustainability in routine HTS programs. This study was conducted within the aPS scale-up project – an implementation science study to implement and evaluate the effectiveness of aPS when integrated within routine HTS, and assess implementation outcomes including implementation fidelity, acceptability, demand, and costs. We conducted semi-structured key informant in-depth interviews with aPS stakeholders at national, county, facility and community levels, and found that aPS was well integrated into the national HTS program within two years of scale-up. Funding limitations, human resource constraints, and low community awareness were noted as major barriers to service provision and long-term sustainability. To overcome these barriers, we recommend increased resource allocation for aPS (funding, human resources) and community health volunteer-facilitated community-level awareness. In the third study, using a payer perspective, we estimated the cost of integrating aPS into routine HTS within the aPS scale-up project in Kisumu and Homa Bay counties. We conducted microcosting, analyzing costs by start-up (August 2018), and recurrent costs one-year after aPS implementation (Kisumu: August 2019, Homa Bay: January 2020), and conducted time-and-motion observations. The average weighted incremental cost of integrating aPS into the existing HTS program was $7,485.97 per facility per year, with recurrent costs accounting for approximately 90% of costs. Average unit costs per male sex partner (MSP) traced, tested, testing HIV-positive, and on antiretroviral therapy were $34.54, $42.50, $108.71 and $152.28, respectively, and varied by county and facility type, with larger volume facilities, especially county and sub-county hospitals, having higher total incremental costs and lower average unit costs. The largest cost drivers were personnel (49%) and transport (13%). We found significant cost variations across facilities offering aPS with high volume facilities having low average unit costs per MSP. We recommend facility prioritization to improve efficiency in resource allocation, especially healthcare personnel, potentially reducing the time and cost spent on delivering aPS. This dissertation contributes to the growing implementation science literature on aPS and highlights the need to prioritize resources as funding support towards HIV programs declines. As aPS is scaled-up, especially in resource-limited settings, policymakers and implementers will need to regularly review program data to identify sub-groups of PLWH requiring additional support before linking to HIV care and treatment services, and address communication gaps on aPS. Future research on cost-efficient strategies optimizing healthcare worker allocation during aPS is also critically important.