© International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Internet: www.worldbank.org; Telephone: 202 473 1000 This work is a product of the staff of The World Bank with external contributions. Note that The World Bank does not necessarily own each component of the content included in this work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Executive Directors of The World Bank or other partner institutions or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 4.0 International (CC BY 4.0) unported license https://creativecommons.org/licenses/by/4.0/. Under the Creative Commons Attribution license, you are free to copy, distribute and adapt this work, including for commercial purposes, under the following conditions: Attribution – Please cite the work as follows: Improving Allocative Efficiency of the HIV Response in Kenya: A County-level Analysis Using the Optima HIV Model. 2020. Washington DC: World Bank. License: Creative Commons Attribution CC BY 4.0 Translations – If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in its translation. All queries on rights and licenses should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington DC, 20433, USA; fax: 202-522-2625; email: pubrights@worldbank.org. IMPROVING ALLOCATIVE EFFICIENCY OF THE HIV RESPONSE IN KENYA: A County-level Analysis Using the Optima HIV Model Benard Lukoba, Joseph Simiyu and Wendy Chege (National AIDS Control Council) Sherrie Kelly and Mark Minnery (Burnet Institute) Lonjezo Sithole and Zara Shubber (World Bank) July 2020 This page is for collation purposes. TABLE OF CONTENTS ACKNOWLEDGEMENTS ..................................................................................................................... ix ABBREVIATIONS ................................................................................................................................. xi EXECUTIVE SUMMARY .................................................................................................................... xiii Key recommendations for HIV investment in Kenya ...................................................... xiv SECTION 1 INTRODUCTION ............................................................................................................ 1 Kenya’s HIV epidemic and response ....................................................................... 1 Rationale for this study ........................................................................................... 3 SECTION 2 OBJECTIVES ................................................................................................................... 5 Objective 1: Optimization within counties ........................................................................ 5 Objective 2: Impact of past HIV spending ......................................................................... 5 SECTION 3 METHODS ..................................................................................................................... 7 Model choice ........................................................................................................... 7 Study design ............................................................................................................ 7 Data sources ............................................................................................................ 7 Populations modeled .............................................................................................. 7 HIV programs modeled ........................................................................................... 8 Model overview ...................................................................................................... 8 Optimization............................................................................................................ 8 Model choice ........................................................................................................... 9 Study design ............................................................................................................ 9 Data sources ............................................................................................................ 9 Populations modeled ............................................................................................ 10 HIV programs modeled ......................................................................................... 10 Model overview .................................................................................................... 10 Optimization.......................................................................................................... 11 SECTION 4 STUDY LIMITATIONS .................................................................................................. 13 SECTION 5 RESULTS ...................................................................................................................... 15 SECTION 6 RECOMMENDATIONS FOR HIV INVESTMENT IN KENYA......................................... 27 SECTION 7 CONCLUSION .............................................................................................................. 29 REFERENCES ..................................................................................................................................... 30 APPENDICES ..................................................................................................................................... 31 Appendix 1: Optima HIV model ....................................................................................... 31 Appendix 2: Key model parameters ................................................................................ 34 Appendix 3: Example model calibration for Nairobi county ........................................... 39 Appendix 4: HIV program unit costs ................................................................................ 42 Appendix 5: Example cost coverage curves for Nairobi county...................................... 44 Appendix 6: Supporting data for results and additional results ..................................... 47 v TABLE OF CONTENTS BOXES 5.1 Illustrative analysis: To estimate the optimized 100% annual resource allocations across counties for 2019 to 2030 to minimize new HIV infections and HIV-related deaths by 2030 ........................................... 21 TABLES A2.1 Latest reported key data and estimates used to inform the Optima HIV county models for Kenya ............................................................................... 34 A2.2 Key parameters used to inform the Optima HIV county models for Kenya ........ 37 A4.1 HIV program unit costs at the county and national level ..................................... 42 A4.2 HIV program unit costs for care and treatment by county .................................. 42 A4.3 HIV program unit costs for HIV testing services by county .................................. 43 A6.1 100% annual optimized allocation within counties represented nationally for 2019 to 2030, and percent of the national budget ....................... 47 A6.2 Changing levels of HIV budgets optimized within counties for 2019 to 2030, represented nationally ........................................................................... 47 A6.3 Percentage of budget allocated by HIV program for changing budgets optimized within counties represented nationally for 2019 to 2030 ......................................................................................................... 48 A6.4 Percentage of total budget allocated to HIV prevention and treatment with changing budgets levels optimized within counties represented nationally for 2019 to 2030 ............................................... 48 A6.5 HIV program coverage levels at 100% HIV budget optimized within counties represented nationally for 2019 to 2030.................................... 48 A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 .............. 54 A6.7 Percentage of total annual HIV budget to optimally reallocate within counties, 2019 to 2030 .............................................................................. 59 A6.8 100% HIV budget optimization across counties, 2019 to 2030 ........................... 61 A6.9 Percentage of total annual HIV budget to optimally reallocate across counties, 2019 to 2030 .............................................................................. 66 A6.10 100% annual HIV budget optimization within counties, represented nationally 2019‒25 .......................................................................... 68 MAP 1.1 County HIV prevalence of adults aged 15–64 years ............................................... 2 vi TABLE OF CONTENTS FIGURES 1.1 County HIV prevalence of adults aged 15–64 years ............................................... 2 5.1 Annual HIV budget optimization within counties to 2030 represented at the national level .............................................................................................. 16 5.2 100% annual HIV budget optimizations within counties, 2019 to 2030 .............. 17 5.3 Impact of optimizing the 100% annual HIV budgets within counties from 2019 to 2030 on new HIV infections represented nationally ....... 18 5.4 Impact of optimizing the 100% annual HIV budgets within counties from 2019 to 2030 on HIV-related deaths represented nationally ..................... 18 5.5 Changing levels of HIV budgets optimized within counties for 2019 to 2030 represented nationally ............................................................................ 20 5.6 Impact of optimizing changing levels of HIV budgets within counties from 2019 to 2030 on new HIV infections represented nationally...................... 20 5.7 Impact of optimizing changing levels of HIV budgets within counties from 2019 to 2030 on HIV-related deaths represented nationally ..................... 21 5.8 100% annual HIV budget optimizations for targeted HIV services across counties, 2019 to 2030 .............................................................................. 23 5.9 Estimated new HIV infections with 100% HIV budget optimized across counties from 2019 to 2030 represented nationally ................................ 24 5.10 Estimated HIV-related deaths with 100% HIV budget optimized across counties from 2019 to 2030 represented nationally ................................ 24 5.11 2015–19 HIV program spending removed to estimate the impact this spending had on new HIV infections and HIV-related deaths over this period ............................................................................................................. 25 A1.1 HIV health states compartments and transmission-related interactions across the care cascade represented in Optima HIV ....................................................... 31 A1.2 Risk-based population mixing patterns represented in Optima HIV ................... 32 A6.1 50% annual HIV budget optimized within counties ............................................. 49 A6.2 90% annual HIV budget optimized within counties ............................................. 50 A6.3 110% annual HIV budget optimized within counties ........................................... 51 A6.4 150% annual HIV budget optimized within counties ........................................... 52 A6.5 200% annual HIV budget optimized within counties ........................................... 53 vii This page is for collation purposes ACKNOWLEDGEMENTS This study is the result of a collaboration of various institutions and individuals who all made essential contributions to the work presented in this report. Contributors within each organization are listed in alphabetical order. The core study, analysis, and report-writing team comprised of Benard Lukoba, Joseph Simiyu and Wendy Chege (National AIDS Control Council); Sherrie Kelly and Mark Minnery (Burnet Institute); and Lonjezo Sithole and Zara Shubber (World Bank). Substantial strategic and technical inputs were also provided by Dr. Nduku Kilonzo, Regina Ombam, Dennis Kamren, John Kamigwi, Emmy Chesire, Joshua Gitonga, Geoffrey Obonyo, and Caro Ngare, (National AIDS Control Council); Dr Joyce Wamicwe, Irene Mukui, and Helgar Musyoki, (National AIDS and STIs Control Programme); Costing Experts; the National TWG; Azfar Hussain and Anna Roberts (Burnet Institute); Marelize Görgens and Nejma Cheikh (World Bank). The team would also like to thank all the County AIDS Coordinators, County HIV/AIDS Coordinators (CACS), County AIDS and STI coordinators (CASCOs) and the Strategic Information Technical Working Group who provided critical data, insights and support, as well as other stakeholders and colleagues, including colleagues from the National AIDS Control Council, Kenya National Treasury, the Joint United Nations Programme on HIV/AIDS (UNAIDS), National AIDS and STIs Control Programme (NASCOP) and President's Emergency Plan for AIDS Relief (PEPFAR). Cover and design: Theo Hawkins (World Bank) ix This page is for collation purposes. ABBREVIATIONS AIDS Acquired Immune Deficiency Syndrome ART Antiretroviral therapy CD4 Cluster of differentiation 4 FSW Female sex worker HTS HIV testing services KASF Kenya AIDS Strategic Framework KENPHIA Kenya Population-based HIV Impact Assessment KNBS Kenya Bureau of Statistics MSM Men who have sex with men NACC National AIDS Control Council NASCOP National AIDS and STIs Control Programme PWID People who inject drugs SBCC Social behavior change communication STI Sexually transmitted infections UNAIDS Joint United Nations Programme on HIV/AIDS VL Viral load VMMC Voluntary medical male circumcision xi This page is for collation purposes EXECUTIVE SUMMARY D espite significant progress in Kenya’s national HIV/AIDS response, Kenya’s HIV epidemic remains the fifth largest in the world in terms of the number of people living with HIV, which was estimated to be 1.4 million in 2018 (1). HIV continues to be a leading cause of adult morbidity and mortality (2). There is significant There is significant heterogeneity in HIV risk by population heterogeneity in HIV risk by and across Kenya’s 47 counties. Kenya exhibits a range of population and across Kenya’s subnational HIV epidemic patterns including generalized, 47 counties. Kenya exhibits a concentrated, and mixed with almost a third (30%) of new range of subnational HIV HIV infections estimated to have occurred among key epidemic patterns including populations such as female sex workers (FSWs) (3). generalized, concentrated, and Estimates for HIV prevalence across counties range from mixed with almost a third (30%) <0.1% in Garissa to 19.6% in Homa Bay, with a national HIV of new HIV infections estimated prevalence estimate of 6.6% among females and 3.1% to have occurred among key among males aged 15–64 years and 0.7% among children populations such as female sex 0–14 years in 2018 (1). workers. Although the Kenya Government has increased spending on the HIV response in the last decade, around 70% of the national HIV response is still reliant on external donor financing (4). With uncertainty around availability of future international HIV/AIDS funding, the sustainability of Kenya’s HIV/AIDS response is at risk. For Kenya to sustain its national HIV response, the best decisions and delivery choices must be made to help ensure that the response is as efficient and effective as possible. This report outlines findings from an allocative efficiency modeling analysis for Kenya’s HIV epidemic and response that was conducted by the National AIDS Control Council (NACC), the National AIDS and STIs Control Programme (NASCOP), the World Bank, and the Optima Consortium for Decision Science, in consultation with other stakeholders at the national and county levels. This study was conducted to inform the 2020/21–2024/25 Kenya AIDS Strategic Framework (KASFII). We estimated the optimized resource allocations within counties, whereby the total HIV budget for each county was kept the same, and across counties, where resources could be shifted between counties. The time horizon for this analysis was from 2019 to 2030 with the objective to minimize new HIV infections and HIV-related deaths by 2030 to align with the Kenya’s Vision 2030. Estimated outcomes for infections and deaths were also presented in the Appendix for 2025 to align with the new 2020/21–2024/25 KASFII. xiii TABLE OF CONTENTS KEY RECOMMENDATIONS FOR HIV INVESTMENT IN KENYA ► Reallocate the latest reported HIV budget prioritising the scale up of care and treatment as well as cost-effective preventative programmes to minimize the number of new HIV infections and HIV-related deaths. To minimize new HIV infections and HIV-related deaths by 2030, recommendations to optimize resources within counties includes prioritizing scale-up of care and treatment, HIV prevention and testing programs targeting females sex workers (FSW), HIV prevention services for the general population (condoms and SBCC), and HIV prevention and testing programs targeting people who inject drugs (PWID) by 2030. This could lead to 50,000 more new HIV infections (almost 10% more) and 40,000 more HIV-related deaths (almost 15% more) being averted. ► The current HIV budget should be maintained at minimum to avoid reversing the gains made in the HIV response. Decreasing the latest reported budget by 50% is estimated to potentially result in 84,000 more new HIV infections (5% more) and 74,000 more HIV-related deaths (6% more) over the 2019 to 2030 period when compared to the status quo. To continue progress in reducing the HIV epidemic at least maintaining the HIV budget is therefore recommended. Effective budget increases are The current HIV budget possible through, for example, implementation should be maintained at efficiency gains (not explored by this analysis), such minimum to avoid reversing as using more optimal service delivery modalities, the gains made in the HIV reduced cost of antiviral regimens, and reduced response. spending on non-targeted programs, among others. ► Additional interventions and innovations to further reduce service delivery costs and increase effectiveness will be required if Kenya is to reach the 2030 target to end AIDS as a public health threat. Even with a doubling of budget for the HIV response optimized within all counties, the 2030 HIV incidence reductions targets are unlikely to be met, meaning there are diminishing marginal returns with the current available ‘toolbox’ of interventions. In countries with large existing disease burdens such as Kenya, reducing HIV incidence to such low levels will need personalized and pre-emptive HIV prevention strategies. xiv SECTION 1 INTRODUCTION K enya’s 2014/15–2018/19 AIDS Strategic Framework (KASF)I marked a milestone in the country’s HIV response in the wake of a new constitution that devolved political and economic power to its 47 newly created counties. The KASFI recognized and emphasized the importance of focusing on effective evidence-based interventions and prioritising investments for improved health outcomes for all. This included an investment case approach with emphasis on geographical, population and intervention prioritisation, and feasibility and sustainability for impact. As Kenya approached the end of the 2014/15–2018/19 KASFI period and started preparations for the new 2020/21–2024/25 KASFII, a county-level HIV allocative efficiency analysis was conducted to review and re-evaluate approaches to ensure that future investments and response were efficient and effective. The aim of any allocative Allocative efficiency analyses allow the most cost-effective efficiency analysis is to guide resource allocation, within a defined resource envelope, to be the financing of the right estimated. The aim of any allocative efficiency analysis is to interventions, for the right people, in the right places, to guide the financing of the right interventions, for the right maximize health outcomes. people, in the right places, to maximize health outcomes. Kenya’s HIV epidemic and response While there has been a gradual decline in the number of annual new adult HIV infections to 36,000, there were still an estimated at 1.4 million people living with HIV in Kenya in 2018 (1), with population and geographical heterogeneity across the country. Four Western counties, Homa Bay, Kisumu, Migori, Siaya, are described as hyperendemic with HIV prevalence estimates higher than 13%. Furthermore, almost half (43%) of all new HIV infections were estimated to have occurred in only five counties, namely Nairobi, Homa Bay, Kisumu, Siaya, and Migori. Female sex workers in Nairobi had the highest respondent driven sampling estimate for HIV prevalence of any other group (29.3% [95% CI: 24.6%–34.9%], N=593, November 2010–January 2011), followed by people who inject drugs in Nairobi (18.7% [12.2%–26.7%], N=263, January–March 2011), and men who have sex with men in Nairobi (18.2% [13.1%–23.6%], N=563, July–October 2010) (3). 1 TABLE OF CONTENTS Map 1.1 County HIV prevalence of adults aged 15–64 years Source: KENPHIA 2018 Preliminary Report, 2020. Kenya’s HIV response is widely praised as a success story with significant achievements in reducing new HIV infections and HIV-related deaths over the last decade. The response at the national and county levels is guided by the KASFI. The country reports reaching the first and second 90-90-901 targets, with 79.5% (95% CI: 77.0%–82.0%) of adults aged 15–64 living with HIV knowing their HIV status, 96.0% (95% CI: 94.7%–97.3%) of those adults who know their status were on ART, and 90.6% 95% CI: 88.5%–92.7%) of those adults on ART having achieved viral suppression. The number of people on HIV treatment has also 1 By 2020, 90% of all people living with HIV will know their HIV status, 90% of all people with diagnosed HIV infection will receive sustained antiretroviral therapy and 90% of all people receiving antiretroviral therapy will have viral suppression. 2 METHODS significantly increased, almost doubling over the last five years, from approximately 600,000 in 2013 to 1.12 million in 2018 (5). This translates to an increase in ART coverage from 43% to 75% over this period. Across the counties, however, there is significant disparity in ART program performance. ART coverage among adults living with HIV at the county level ranges from 23% to 99% (table A2.1). Eleven counties latest reported ART coverage among adults 15 years and older above 80%, 25 counties coverage between 50% and 79%, nine 32% to 49%, one county an ART coverage of 23% (Tana River), and one county with coverage not reported (Mombasa). There is similar county-wide heterogeneity in prevention of mother-to-child transmission (PMTCT) coverage, with 12 counties achieving greater than 80% coverage, 21 counties achieving 50%–79% coverage, 11 counties achieving 24%–49% coverage, and three counties having less than 23% coverage of PMTCT (5). The incidence of mother-to-child transmission of HIV at 18 months in Kenya stood at 11.5% in 2017, down from 14.0% in 2014 (5). There has also been progress in scale- up of non-ART prevention efforts. For example, the annual number of voluntary medical male circumcisions (VMMC) conducted in 2017 (230,854) surpassed the 200,000 circumcisions annual targets set for the 2014/15–2018/19 KASFI mid- (2017) and end- term (2019). The proportion of men circumcised is now greater than 95%. Similar achievements have been reported in the coverage of HIV programmes targeting key populations, with over 80% of FSWs, MSM, and PWID being reached with combination prevention programmes in 2017. Rationale for this study Since 70% of Kenya’s HIV response is externally funded and donor funding is declining (7), the future of HIV funding for Kenya is at risk. Furthermore, since Kenya graduated to the World Bank’s lower-middle income country status in 2015, it will be less eligible for donor funding in this round of Global Fund disbursement. For Kenya to sustain its national HIV response, in addition to the 2014/15–2018/19 KASFI For Kenya to sustain its national commitment to increase domestic HIV financing to 50% of HIV response, in addition to the the total program budget, sound decisions and delivery 2014/15-2018/19 KASFI choices must be made to help ensure that the HIV response commitment to increase is as efficient and effective as possible. Kenya’s diversity in domestic HIV financing to 50% HIV epidemics, ranging from generalized to concentrated of the total program budget, and mixed, with high heterogeneity in the contribution of sound decisions and delivery key populations to county-level HIV transmission also choices must be made to help necessitates within-county resource allocations and policy ensure that the HIV response is choices that are tailored to the local epidemic profile in each as efficient and effective as county. To help inform the next iteration of the KASF possible. (KASFII), this modelling study was conducted by the National AIDS Control Council (NACC) and the National AIDS and STIs Control Programme (NASCOP), in collaboration with the World Bank, the Optima Consortium for Decision Science, and other stakeholders. 3 This page is for collation purposes. SECTION 2 OBJECTIVES To improve the HIV response, resources can be optimally reallocated within counties in Kenya. To this end, the following study objectives were examined: Objective 1: Optimization within counties  To estimate the optimized annual resource allocations within counties for 2019 to 2030 to minimize new HIV infections and HIV-related deaths by 2030 for varying budget levels 2, and to present estimated outcomes for infections and deaths by 2025 to align with the end of the new 2020/21–2024/25 KASFII.  To examine future HIV epidemic trends at the national and county levels to better understand the HIV epidemic and thereby help guide strategic planning through to 2030 to align with Kenya’s Vision 2030. Objective 2: Impact of past HIV spending  To estimate the epidemiological impact of the 2014/15–2018/19 KASFI spending on new HIV infections and HIV-related deaths averted compared with no spending over this period, as well as to evaluate and learn from the successes in the HIV response in Kenya. 2 Varying budget levels of 50%, 90%, 100%, 110%, 150%, and 200% were examined. Kenya’s HIV response is widely praised as a success story with significant achievements in reducing new HIV infections and HIV-related deaths over the last decade. Khadija Rama, standing with a group of people, is the founder 5 of Pepo La Tumaini Jangwani, HIV/AIDS Community Rehabilitation Program, Orphanage & Clinic. offers hope, support and care for orphan and vulnerable children living with HIV/AIDS in Nairobi, Kenya, Africa. Photo: Joseph Sohm/Dreamstime.com This page is for collation purposes. SECTION 3 METHODS Model choice This analysis was conducted using the Optima HIV model version 2.9.2. Optima HIV is an epidemiological model of HIV transmission overlaid with an economic analysis compartment that contains a resource optimization algorithm. Study design A separate Optima HIV model was created for each of the 47 counties in Kenya. County models used in this analysis were developed by the country team (officers within NACC and NASCOP), the World Bank, and the Optima Consortium for Decision Science, in consultation with other stakeholders. Data sources Data and estimates used to inform the county models were retrieved from national, county- level, and stakeholder reports, HIV program data sources, other publications, and from expert opinion (table A1).3 Populations modeled General population and key population groups were included in the Optima HIV model for each of the 47 counties in Kenya. General population groups include females aged 0–14 years, males 0–14, females 15–24, males 15–24, females 25–49, males 25–49, females 50 years and older (females 50+), and males 50 years and older. Key populations include female sex workers (FSW), clients of female sex workers (clients), men who have sex with men (MSM), and people who inject drugs (PWID). 3 Reports included but were not limited to the Behavioural Assessment of Key Populations in Kenya Polling Booth Survey (2017); County Profile reports (2014, 2016, 2018); Demographic and Health Survey (2014); District Health Information System 2 and other program data; DREAMS Overview (2016-2017); HIV Financing County Profiles (2018); Kenya AIDS Indicator Survey (2012); Kenya AIDS Response Program reports; Kenya AIDS Strategic Framework (2014/15-2018/19) and Mid-Term Review Report (2018); Kenya HIV Estimates report (2015 and 2017); Kenya Integrated Biological and Behavioral Surveillance reports; Kenya Most At Risk Populations Size Estimate Consensus report (2013); Kenya National Bureau of Statistics County Projections (2017); Key populations reports; National AIDS Spending Assessment reports (2006-2007 and 2009-2011); PEPFAR Country Operational Plan Strategic Direction Summary (2017); UNAIDS AIDSinfo and Key Population Atlas; and United Nations World Population Prospects (WPP, 2017) estimates and projections. 7 TABLE OF CONTENTS HIV programs modeled Seven HIV programs were included in this analysis, including: (1) care and treatment (comprised of antiretroviral therapy (ART) and prevention of mother-to-child therapy (PMTCT)); (2) HIV prevention services (condoms and social behavior change communication (SBCC)); (3) HIV testing services (biomedical services only); (4) HIV prevention and testing programs targeting female sex workers (FSW); (5) HIV prevention and testing programs targeting men who have sex with men (MSM); (6) HIV prevention and testing programs targeting people who inject drugs (PWID) including needle-syringe programs opiate substitution therapy (OST); and (7) voluntary medical male circumcision (VMMC). Pre-exposure prophylaxis (PrEP) programs were not modelled. Only targeted HIV programs were considered in this analysis. Non-targeted programs, representing programs whose direct impact on the epidemic is not readily determinate, were excluded. These non- targeted programs include environmental HIV programs, human resources, management, monitoring and evaluation, non-disaggregated prevention programs, orphans and vulnerable children, other HIV care, social protection. All costs are reported in United States Dollars (USD) with no discounting applied. Model overview Optima HIV has a population-based dynamic, compartmental model. It has a disease transmission module that is calibrated to demographic (population size, and birth and background death rates) and epidemiological (HIV, ulcerative STIs, and tuberculosis prevalence) estimates. It is informed by sexual and injecting behavioural values and mixing patterns, as well as programmatic data for testing and treatment across the care cascade (time for linkage to care, percentage lost-to-follow-up, and treatment failure rate). The model assumes parameter values for relative disease-related transmissibility and disease progression specified for acute infection and CD4 health states. People on suppressive ART are assumed to have a 92% reduction in HIV transmission compared with people not on ART (table A2.2). People not on suppressive ART and not in early or late stage infection were assumed to have a relative HIV transmission rate of 50%. Different death rates by health state (acute infection or CD4 stage), ART status (on suppressive or non-suppressive ART), and tuberculosis cofactor were applied. HIV program cost and coverage data are used to generate cost functions for each HIV program. Cost functions represent the relationship between cost and coverage and coverage and outcomes for each program. Different values for changes in Optimization analyses were transmissibility are applied for condom use, circumcision, and conducted to estimate the treatment types (table A2.2). Additional Optima HIV model most cost-effective investment details are provided in Kerr et al. (8). across a combination of HIV programs to minimize new Optimization HIV infections and HIV- related deaths. Optimization analyses were conducted to estimate the most cost-effective investment across a combination of HIV programs to minimize new HIV infections and HIV-related deaths. The optimization was conducted using the adaptive stochastic descent (ASD) algorithm (9). Resources were optimized either within or across counties. For optimizations within counties, the total HIV 8 METHODS budget for each county is maintained and resources optimized within each county budget. In contrast, for optimizations across counties, resources could be shifted between counties and optimized for cost-effectiveness. For the optimization within counties, constraints were applied within the optimizations to ensure those on treatment remained on treatment unless lost by natural attrition. Thus, in the optimized allocation, budgets for antiretroviral therapy (ART) and prevention of mother-to-child transmission (PMTCT) could not be reduced to ensure at least the same number of people were maintained on treatment. Optimization across counties is an illustrative analysis only and no constraints were applied. Prioritization for given programs within the optimization is Prioritization for given defined for this study as a proportional scale-up of allocation programs within the of any magnitude from the latest reported level. Finally, the optimization is defined for this model algorithm aimed to estimate a theoretical optimal study as a proportional scale- distribution of resources and emphasis of different HIV up of allocation of any programmatic responses which minimizes both new HIV magnitude from the latest infections and HIV-related deaths given the local epidemic reported level. parameters and data, cost of delivering services, subject to the constraints as defined. Appendices 2‒5 show key model parameters, model calibration and cost curve figures for Nairobi county as an example, and unit costs used in this study. Model choice This analysis was conducted using the Optima HIV model version 2.9.2. Optima HIV is an epidemiological model of HIV transmission overlaid with an economic analysis compartment that contains a resource optimization algorithm. Study design A separate Optima HIV model was created for each of the 47 counties in Kenya. County models used in this analysis were developed by the country team (officers within NACC and NASCOP), the World Bank, and the Optima Consortium for Decision Science, in consultation with other stakeholders. Data sources Data and estimates used to inform the county models were retrieved from national, county- level, and stakeholder reports, HIV program data sources, other publications, and from expert opinion (table A2.1).4 4 Reports included but were not limited to the Behavioural Assessment of Key Populations in Kenya Polling Booth Survey (2017); County Profile reports (2014, 2016, 2018); Demographic and Health Survey (2014); District Health Information System 2 and other program data; DREAMS Overview (2016-2017); HIV Financing County Profiles (2018); Kenya AIDS Indicator Survey (2012); Kenya AIDS Response Program reports; Kenya AIDS Strategic Framework (2014/15-2018/19) and Mid-Term Review Report (2018); Kenya HIV Estimates report (2015 and 2017); Kenya Integrated Biological and Behavioral Surveillance reports; Kenya Most At Risk Populations Size Estimate Consensus report (2013); Kenya National Bureau of Statistics County Projections (2017); Key populations reports; National AIDS Spending Assessment reports (2006-2007 and 2009-2011); PEPFAR Country Operational Plan Strategic Direction Summary (2017); UNAIDS AIDSinfo and Key Population Atlas; and United Nations World Population Prospects (WPP, 2017) estimates and projections. 9 TABLE OF CONTENTS Populations modeled General population and key population groups were included in the Optima HIV model for each of the 47 counties in Kenya. General population groups include females aged 0–14 years, males 0–14, females 15–24, males 15–24, females 25–49, males 25–49, females 50 years and older (females 50+), and males 50 years and older. Key populations include female sex workers (FSW), clients of female sex workers (clients), men who have sex with men (MSM), and people who inject drugs (PWID). HIV programs modeled Seven HIV programs were included in this analysis, including: (1) care and treatment (comprised of antiretroviral therapy (ART) and prevention of mother-to-child therapy (PMTCT)); (2) HIV prevention services (condoms and social behavior change communication (SBCC)); (3) HIV testing services (biomedical services only); (4) HIV prevention and testing programs targeting female sex workers Only targeted HIV programs (FSW); (5) HIV prevention and testing programs targeting men were considered in this who have sex with men (MSM); (6) HIV prevention and testing analysis. programs targeting people who inject drugs (PWID) including needle-syringe programs opiate substitution therapy (OST); and (7) voluntary medical male circumcision (VMMC). Pre-exposure prophylaxis (PrEP) programs were not modelled. Only targeted HIV programs were considered in this analysis. Non-targeted programs, representing programs whose direct impact on the epidemic is not readily determinate, were excluded. These non-targeted programs include environmental HIV programs, human resources, management, monitoring and evaluation, non-disaggregated prevention programs, orphans and vulnerable children, other HIV care, social protection. All costs are reported in United States Dollars (USD) with no discounting applied. Model overview Optima HIV has a population-based dynamic, compartmental model. It has a disease transmission module that is calibrated to demographic (population size, and birth and background death rates) and epidemiological (HIV, ulcerative STIs, and tuberculosis prevalence) estimates. It is informed by sexual and injecting behavioural values and mixing patterns, as well as programmatic data for testing and treatment across the care cascade (time for linkage to care, percentage lost-to-follow-up, and treatment failure rate). The model assumes parameter values for relative disease-related transmissibility and disease progression specified for acute infection and CD4 health states. People on suppressive ART are assumed to have a 92% reduction in HIV transmission compared with people not on ART (table A2.2). People not on suppressive ART and not in early or late stage infection were assumed to have a relative HIV transmission rate of 50%. Different death rates by health state (acute infection or CD4 stage), ART status (on suppressive or non-suppressive ART), and tuberculosis cofactor were applied. HIV program cost and coverage data are used to generate cost functions for each HIV program. Cost functions represent the relationship between cost and coverage and coverage and outcomes for each program. Different values for changes in transmissibility are applied 10 METHODS for condom use, circumcision, and treatment types (table A2.2). Additional Optima HIV model details are provided in Kerr et al. (8). Optimization Optimization analyses were conducted to estimate the most cost-effective investment across a combination of HIV programs to minimize new HIV infections and HIV-related deaths. The optimization For the optimization within was conducted using the adaptive stochastic descent counties, constraints were applied (ASD) algorithm (9). Resources were optimized either within the optimizations to within or across counties. For optimizations within ensure those on treatment counties, the total HIV budget for each county is remained on treatment unless lost maintained and resources optimized within each county by natural attrition. budget. In contrast, for optimizations across counties, resources could be shifted between counties and optimized for cost-effectiveness. For the optimization within counties, constraints were applied within the optimizations to ensure those on treatment remained on treatment unless lost by natural attrition. Thus, in the optimized allocation, budgets for antiretroviral therapy (ART) and prevention of mother-to- child transmission (PMTCT) could not be reduced to ensure at least the same number of people were maintained on treatment. Optimization across counties is an illustrative analysis only and no constraints were applied. Prioritization for given programs within the optimization is defined for this study as a proportional scale-up of allocation of any magnitude from the latest reported level. Finally, the model algorithm aimed to estimate a theoretical optimal distribution of resources and emphasis of different HIV programmatic responses which minimizes both new HIV infections and HIV-related deaths given the local epidemic parameters and data, cost of delivering services, subject to the constraints as defined. Appendices 2–5 show key model parameters, model calibration and cost curve figures for Nairobi county as an example, and unit costs used in this study. After scale-up of treatment for all diagnosed people living with HIV, the next priority to improve outcomes across the cascade of care and avert infections and deaths is to increase diagnoses by scaling-up testing. 11 Kenyan women stand in line to get health checkup for HIV/AIDS at the Pepo La Tumaini Jangwani, HIV/AIDS Community Rehabilitation. Photo: Joseph Sohm/Dreamstime.com This page is for collation purposes. SECTION 4 STUDY LIMITATIONS A s with any modelling study, there are limitations with this analysis that should be considered when interpreting results and recommendations. One of the main limitations for this study was around the availability of data at the county-level. Limitations in data availability and reliability can lead to uncertainty around projected results. The model optimization algorithm accounts for inherent uncertainty, but it may not be possible to account for all aspects of uncertainty because of poor quality or insufficient data. The following data were available at the county-level and were used to inform the respective county models or to guide calibration of the county models: demographic data (population size, birth rate, background death rate) and treatment (antiretroviral therapy (ART), prevention of mother-to-child transmission (PMTCT), and opiate substitution therapy (OST; by proportion of PWID)), and annual numbers of HIV tests, HIV diagnoses, HIV infections, HIV-related deaths, ART initiations, and those virally suppressed, as well as the percent of people living with HIV who know their status. The availability of detailed county-level costing data, which are needed to generate cost functions, that is the relationship between spending and coverage and coverage and outcomes, was limited. County-level estimates of expenditure were only readily available for two program The availability of detailed county-level costing data, categories (1) overall HIV prevention and (2) care and which are needed to generate treatment (6). Care and treatment county expenditure cost functions, that is the estimates were used directly for this modeled intervention. relationship between spending Care and treatment unit costs were calculated using care and coverage and coverage and treatment estimates of expenditures and numbers of and outcomes, was limited. people on antiretroviral therapy for each county (table A4.2). Since it cost some counties more to put an average person living with HIV on treatment for one year, for various operational and programmatic reasons, unit costs for care and treatment differed between counties with an average unit cost of US$663, ranging from US$85 in Kiambu to US$3,677 in Wajir. Overall HIV prevention expenditure estimates for each county needed to be disaggregated into separate HIV prevention programs that were modeled. In consultation with the country team, county expenditures for HIV prevention programs, including HIV testing, condoms and SBCC, VMMC, HIV testing and prevention targeting FSW, HIV testing and prevention targeting MSM, and HIV testing and prevention targeting PWID, were derived by triangulating county expenditure estimates for overall prevention for 2013/2014, 2014/2015, and 2015/2016; county coverage values for 13 TABLE OF CONTENTS HIV testing for 2015 and for VMMC for 2017; nationally informed unit costs for each HIV prevention program (tables A4.1 and A4.2); and the latest reported proportion spent on each HIV prevention program of the national prevention budget. Another key indicator to inform this model exercise is HIV prevalence. HIV prevalence estimates at the county-level were not available for all populations. The following HIV prevalence estimates were used to calculate missing estimates: overall HIV prevalence by county for 2018 (figure 1), HIV prevalence for adults 15–49 years of age for 2017 (overall, male, and female) (6), estimated numbers of people living with HIV by county for children aged 0–14, youth 15–24, and adults 15 years or older for 2015 (6), and HIV prevalence values for female sex workers for 2010 (3) and 2012 (10). Estimates of HIV prevalence from the Spectrum model were not generated at the county level, only at the regional level, so could not be used to compare trends for county model calibrations. County-level values for HIV testing, breastfeeding rates, sexual and injecting partnership and behaviour, average time to be linked to care, percentage of people in care who are lost to follow-up, average number of viral load monitoring tests conducted per person per year, and treatment failure rate were not readily available. As such, national values were used as proxy values for each county model. Pre-exposure prophylaxis data by country was not available at the county-level and was therefore not considered in this analysis. The sum of all county annual HIV diagnoses values was high given the testing rates and HIV incidence estimates. This is likely due to double counting of diagnoses (e.g., repeat testing in antenatal care) and the study team chose to put less weight on diagnoses values during the calibration process. The full results from the Kenya Population-based HIV Impact Assessment (KENPHIA) were not available when this study was conducted and could therefore not be used to inform the study. The preliminary KENPHIA report has since been made available and was used to validate estimates generated in this analysis. During the workshop held in Nairobi, Kenya in September 2019, the study team worked closely with the country team to validate the derived values used to inform county models. County model calibrations and county costing values were shared widely with county program teams to further ascertain the plausibility of the derived values. Finally, these findings are only modeled projections and have not been confirmed in a practical setting in Kenya. The country models used in this study have been calibrated to reflect county-level epidemiological estimates provided by the country team, but validation of results suggesting optimized reallocations that will lead to reductions in infections and deaths in real-world practice has not been done. Shifting resources following evidence from this study will not always be feasible and may not necessarily be politically favorable but should be considered for greater impact. 14 SECTION 5 RESULTS OBJECTIVE 1: To estimate the optimized annual resource allocations within counties for 2019 to 2030 to minimize new HIV infections and HIV-related deaths by 2030 for varying budget levels5, estimate outcomes for infections and deaths by 2025 to align with the end of the new 2019/20–2024/25 KASFII, and examine future HIV epidemic trends at the national and county levels to better understand the HIV epidemic to guide strategic planning through to 2030 to align with Kenya’s Vision 2030. Recommendations to optimize annual allocations within counties to 2030 to minimize new HIV infections and HIV-related deaths by 2030 using the 47 county projects, represented at the national- level, include prioritizing scale-up of care and treatment, HIV prevention and testing programs targeting females sex workers (FSW), and HIV prevention and testing programs targeting people who inject drugs (PWID) (figure 5.1; tables A6.1, A6.5, While there has been a gradual A6.7). These recommendations align with country decline in the number of annual strategic plans to treat more people diagnosed with HIV, new adult HIV infections to as well as to increase treatment coverage of people living 36,000, there were still an with HIV from an estimated 75% in 2018 (5) to potentially estimated at 1.4 million people achieving 79% by 2030. It is also estimated that female living with HIV in Kenya in 2018 sex workers and their clients accounted for almost 20% of (1), with population and new HIV infections in 2018; therefore, prioritizing HIV geographical heterogeneity across prevention and testing programs targeting female sex the country. workers will be important to minimize HIV transmission among sex workers, clients, and their other partners. See appendix table A6.6 for budget values to support figure 5.2 and table A6.7 for changes in allocation, as well as figures A6.1‒A6.5 for optimization allocations within counties for other budget levels. For all counties, it is recommended to scale-up care and treatment over the period between 2019 and 2030 to minimize infections and deaths by 2030, except for Kakamega, Kilifi, and Nairobi counties where it is recommended to maintain care and treatment budgets at latest reported levels. This aligns with the national targets of increasing treatment coverage. Furthermore, since it was estimated that almost one fifth of all new HIV infections in 2018 were among female sex workers and their clients, it is recommended to scale-up HIV testing and prevention targeting FSW in all counties 1 to 80-times higher than current coverage across counties, with the highest increase 5 Varying budget levels of 50%, 90%, 100%, 110%, 150%, and 200% were examined. 15 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK recommended in Nakuru county and an overall recommended 3-times increase in funding for this program at the national level. It is important to be mindful that county-level latest reported budgets for this program, and other prevention programs, were derived from county-level 2015/16 spending on prevention from the 2018 County Profile reports and national unit costs for the respective programs. Figure 5.1 Annual HIV budget optimization within counties to 2030 represented at the national level $600 $500 HIV testing and prevention for PWID Spending in Millions USD HIV testing and prevention for MSM $400 HIV testing and prevention for FSW $300 VMMC HIV prevention services (condoms and SBCC) $200 HIV testing services (biomedical only) $100 Care and treatment $0 100% latest 100% optimized reported Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; SBCC = social behaviour change communication; VMMC = voluntary male medical circumcision; USD = United States dollar. If annual county budgets were optimized within counties from 2019 to minimize infections and deaths by 2030, it is estimated that an additional 29,000 new HIV infections (6% more) could be averted by 2025, and 50,000 more could be averted (almost 10% more) by 2030 compared with the latest reported allocation being maintained over this period (figure 5.3). This is however not enough to reach the 2030 end AIDS target of a 90% reduction in HIV incidence by 2030 from 2010 levels. 16 Figure 5.2 100% annual HIV budget optimizations within counties, 2019 to 2030 $90 $80 $70 $60 Bars per county Latest reported (top) $50 Optimized (bottom) Millions USD $40 $30 $20 $10 $0 Tana River Turkana Machakos Muranga Kirinyaga Kwale Siaya Meru Busia Kitui Kisii Bomet Wajir Vihiga Nyeri Nyandarua Nyamira Nakuru Nairobi Laikipia Kisumu Kericho Homa Bay Baringo Narok Embu Marsabit Tharaka Nithi Nandi Migori Taita Taveta Samburu Mombasa Kiambu Kajiado Isiolo Garissa Uasin Gishu Trans Nzoia Mandera Makueni Kilifi Kakamega Lamu Elgeyo Marakwet Bungoma West Pokot Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; VMMC = voluntary male medical circumcision. RESULTS 17 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Figure 5.3 Impact of optimizing the 100% annual HIV budgets within counties from 2019 to 2030 on new HIV infections represented nationally 120,000 New HIV infections 100,000 Latest reported 80,000 Optimized 60,000 40,000 20,000 90% reduction target from 2010 level 0 2010 2015 2020 2025 2030 Source: Authors from Optima data. If annual county budgets were optimized within counties from 2019 to minimize infections and deaths by 2030, it is estimated that an additional 24,000 HIV-related deaths (14% more) could be averted by 2025, and 40,000 more could be averted (almost 15% more) compared with the latest reported allocation being maintained over this period (figure 5.4). Figure 5.4 Impact of optimizing the 100% annual HIV budgets within counties from 2019 to 2030 on HIV- related deaths represented nationally 80,000 HIV-related deaths 70,000 60,000 50,000 40,000 Latest reported Optimized 30,000 20,000 10,000 0 2010 2015 2020 2025 2030 Source: Authors from Optima data. Optimization within counties under varying budget levels from 2019 to 2030 to minimize infections and deaths shows that if more budget were to become available, more should be spent on prevention services including VMMC and particularly HIV testing (figure 5.5). However, with increasing budget, overall, there are diminishing returns on investment (figures 5.6 and 5.7). The ratio of new HIV 18 RESULTS infections averted through optimization of resources within all counties to the percent of increased budget is 14.8 for the 110% optimized budget but drops to 10.5 for the 150% optimization budget and to 7.2 for the 200% optimized budget. This ratio is an indicator of the amount of return on investment should additional HIV funding be made available versus the amount of infections or deaths that could be averted. It is important to note that even with a doubling of budget that is optimized, the reduction in new HIV infections would not be enough to reach the 2030 target of a 90% reduction in new infections by 2030 from 2010 level. Diminishing returns with increased HIV budget optimized within all counties is also observed for HIV-related deaths with a ratio of 7.3 for the 110% optimized budget, which drops to 2.2 at 150% optimized budget, but rises slightly to 2.7 at 200% optimized budget. One possible reason that may help explain this trend is that at 200% optimized budget, it could be possible to achieve the target of 90% of those aware of their HIV status receiving treatment by 2030, thus driving down deaths even further. Recommendations for prioritizing resource allocation change as budget amounts vary. For example, at 200% budget it is recommended to allocate 9% of the national budget on prevention, versus 4% at 100% budget, and 2% at 50% budget (figures 5.2, 5.5, A4.1, and A6.2). Inversely, if less funding were available, it is recommended to make care and treatment an even higher priority, whereby, at 50% budget 98% of the budget should be allocated for treatment, compared with 96% at 100% optimized budget, and 91% at 200% optimized budget. It is important to note that programs with an indirect impact on the HIV epidemic, such as program management and strategic information, are not included in the optimized budget. At all budget levels examined in this analysis, it is recommended to allocate 90% or more of the budget towards treatment. This holds true for all counties, except for Machakos county where at 100% budget it is recommended to allocate 80% of the budget to treatment under optimized allocation, up from the approximate 75% of the budget allocated for treatment in the latest reported allocation. This is the third lowest proportion allocated to care and treatment in the latest reported budget of all counties: only Kiambu, 60%, and Turkana, 62%, allocated less for treatment in the latest reported allocation. However, of these three counties, Machakos was estimated to have the lowest budget contribution for testing services of its overall HIV budget, 22%, compared with the other two counties (Turkana 26% and Kiambu 34%). Under the optimized allocation, it is recommended for Machakos to retain a higher proportion of its total budget for testing, 18%, than the other two counties. This suggests Machakos has a greater need to prioritize HIV testing, while still scaling up treatment. Nevertheless, for Machakos, while treatment coverage in 2018 was already relatively high at almost 90%, even with 80% of the budget allocated for treatment (with a recommended 5% increase in spending on care and treatment over this period) it is estimated that by 2030 treatment coverage would still increase to almost 95%. More specifically, under optimized allocation based on percent of the total budget, with increasing budget it is recommended to decrease the proportional allocation for care and treatment and HIV testing and prevention programs for FSW at 150% budget and above, and VMMC, while increasing allocation for HIV testing services (biomedical only), HIV prevention services (condoms and SBCC), and HIV testing and prevention programs for MSM. Recommended allocations for HIV testing and prevention programs for PWID remain relatively stable regardless of budget level. If total county budgets were to be decreased, it is recommended to prioritize care and treatment even more. For example, at 50% total budget, VMMC, HIV testing and prevention for FSW, and HIV testing and 19 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK prevention for PWID should get a bigger proportion of the reduced budget, but HIV testing services (biomedical only), HIV prevention services (condoms and SBCC), and HIV testing and prevention for MSM should get less priority. Figure 5.5 Changing levels of HIV budgets optimized within counties for 2019 to 2030 represented nationally $1,200 HIV testing and prevention for PWID $1,100 HIV testing and prevention for MSM HIV testing and prevention for FSW $1,000 VMMC HIV prevention services (condoms and SBCC) $900 HIV testing services (biomedical only) Care and treatment $800 Millions USD $700 $600 $500 $400 $300 $200 100% latest 50% 90% 100% 110% 150% 200% reported optimized optimized optimized optimized optimized optimized Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; SBCC = social behaviour change communication; VMMC = Voluntary male medical circumcision. Figure 5.6 Impact of optimizing changing levels of HIV budgets within counties from 2019 to 2030 on new HIV infections represented nationally 120,000 100,000 80,000 New HIV infections 60,000 50% optimized 90% optimized 40,000 100% latest reported 100% optimized 110% optimized 20,000 150% optimized 200% optimized 90% reduction target from 2010 level 0 2010 2015 2020 2025 2030 Source: Authors from Optima data. 20 RESULTS Figure 5.7 Impact of optimizing changing levels of HIV budgets within counties from 2019 to 2030 on HIV- related deaths represented nationally 80,000 HIV-related deaths 70,000 60,000 50,000 40,000 50% optimized 30,000 90% optimized 100% latest reported 20,000 100% optimized 110% optimized 10,000 150% optimized 200% optimized 0 2010 2015 2020 2025 2030 Source: Authors from Optima data . The optimized allocation to minimize new HIV infections and HIV-related deaths by 2025 to align with the end of the new 2019/20–2024/25 KASFII are shown in table A6.10. Box 5.1 Illustrative analysis: To estimate the optimized 100% annual resource allocations across counties for 2019 to 2030 to minimize new HIV infections and HIV-related deaths by 2030 An illustrative analysis whereby resource allocations are optimized across counties without constraints shows how resources could be shifted towards certain counties and then how these adjusted county budgets could be invested towards the most cost-effective HIV programs. Findings for 100% budget optimization across counties show that among the seven counties with the highest estimated number of new HIV infections in 2018, it is recommended to shift the largest amount of funds for any county to six out of seven of these high-burden counties (figure 5.8; tables A8.8 and A6.9). The seven counties with the highest estimated number of new HIV infections, in descending order, are Homa Bay, Siaya, Kisumu, Nairobi, Migori, Kiambu, and Nakuru, accounted for an estimated 25% of all new HIV infections in Kenya in this year. Optimal redistribution of funds to these six counties would account for over 50% of the total budget for the national HIV program in Kenya. Of these top seven counties, Nairobi was the only county for which it is recommended to allocate less funding under optimized allocation. Since Nairobi county has the highest estimate of people living with HIV (PLHIV) of any county in Kenya, approximately 191,000 in 2018 or 12% of the total 1.49M PLHIV nationally, it follows that the biggest county HIV budget, $83M in 2018 or 14% of the national budget, was allocated to this county. However, based on model recommendations, more new HIV infections could be averted nationally over the 2019 to 2030 period if some of the HIV budget for Nairobi county were shifted to other counties and thereafter all county budgets optimally allocated. Box continued… 21 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Box 5.1 Illustrative analysis: To estimate the optimized 100% annual resource allocations across counties for 2019 to 2030 to minimize new HIV infections and HIV-related deaths by 2030 (continued) Prioritizing the most cost-effective HIV programs as part of this illustrative optimization includes scaling-up care and treatment in over half of the counties, as well as prioritizing VMMC, HIV prevention and testing programs targeting FSW, and HIV prevention and testing programs targeting PWID. Besides, some funding should be maintained for HIV testing services mainly targeting the general population and for HIV prevention services, including condoms and SBCC. This suggests that in the context of Kenya’s HIV response, there is potential for leveraging additional impact if resources were permitted to shift across not only programs, but also geographical space. From this illustrative analysis of resource optimization across counties, whereby the treatment program of a given county could be defunded, shifts in resources across counties according to burden and cost-effectiveness criteria could lead to 160,000 more new HIV infections (over 20% more) and 73,000 more HIV-related deaths (almost 30% more) being prevented by 2030. For counties where it is recommended to increase the total budget from 2019 to 2030, it is also recommended to scale-up care and treatment (figure 5.6, tables A6.8 and A6.9). For Kirinyaga, Kitui, Machakos, and Nyeri counties where a decreased total budget is recommended, it is still recommended to scale-up care and treatment. If annual county budgets were optimized across counties from 2019 to 2030 to minimize infections and deaths by 2030, then an additional 160,000 new HIV infections could be averted (over 20% more) compared with the latest reported allocation being maintained over this period (figure 5.3) 22 Figure 5.8 100% annual HIV budget optimizations for targeted HIV services across counties, 2019 to 2030 (sorted in ascending order by optimized allocation)* $90 $80 $70 $60 Millions USD $50 Bars per county $40 Latest reported (top) Optimized (bottom) $30 $20 $10 $0 Tharaka Nithi Homa Bay West Pokot Bomet Kwale Machakos Nyeri Nyamira Laikipia Turkana Busia Meru Kitui Kisii Kisumu Siaya Nyandarua Vihiga Nairobi Wajir Narok Nakuru Tana River Samburu Baringo Embu Kirinyaga Kericho Mandera Garissa Marsabit Isiolo Elgeyo Marakwet Kajiado Nandi Muranga Trans Nzoia Makueni Migori Kilifi Uasin Gishu Kiambu Mombasa Bungoma Kakamega Lamu Taita Taveta Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. Note: * = This is an illustrative analysis only and no constraints were applied to the model. FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; VMMC = Voluntary male medical circumcision. RESULTS 23 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Figure 5.9 Estimated new HIV infections with 100% HIV budget optimized across counties from 2019 to 2030 represented nationally New HIV infections 120,000 100,000 Latest reported 80,000 Optimized 60,000 40,000 20,000 0 2010 2015 2020 2025 2030 Source: Authors from Optima data. Figure 5.10 Estimated HIV-related deaths with 100% HIV budget optimized across counties from 2019 to 2030 represented nationally HIV-related deaths 80,000 60,000 40,000 Latest reported Optimized 20,000 0 2010 2015 2020 2025 2030 Source: Authors from Optima data. OBJECTIVE 2: To estimate the epidemiological impact of the 2014/15–2018/19 KASFI spending on new HIV infections and HIV-related deaths averted compared with no spending over this period, to evaluate and learn from the successes in the HIV response in Kenya. At the national level, it is estimated that no spending on the HIV program during the 2014/15– 2018/19 KASFI period compared with what was spent over this 5-year period could have resulted 24 RESULTS in 500,000 more new HIV infections (almost 150% more) and 600,000 more HIV-related deaths (over 300% more) (figure 5.11). This suggests that investments made by the HIV program over the 2014/15–2018–19 KASFI period was effective in preventing at least half a million new HIV infections and over half a million HIV-related deaths in Kenya. Figure 5.11 2015–19 HIV program spending removed to estimate the impact this spending had on new HIV infections and HIV-related deaths over this period New HIV infections HIV-related deaths 230,000 200,000 210,000 180,000 190,000 160,000 170,000 140,000 150,000 120,000 100,000 130,000 80,000 110,000 60,000 90,000 40,000 70,000 20,000 50,000 0 2010 2012 2014 2016 2018 2010 2012 2014 2016 2018 Zero spending 2015-2019 Zero spending 2015-2019 Latest reported spending Latest reported spending Source: Authors from Optima data. 25 This page is for collation puposes SECTION 6 RECOMMENDATIONS FOR HIV INVESTMENT IN KENYA ► Reallocate the latest reported HIV budget prioritising the scale up of care and treatment as well as cost-effective preventative programmes to minimize the number of new HIV infections and HIV-related deaths. To minimize new HIV infections and HIV- related deaths by 2030, recommendations to optimize resources within counties includes prioritizing scale-up of care and treatment, HIV prevention and testing programs targeting females sex workers (FSW), HIV prevention services for the general population (condoms and SBCC), and HIV prevention and testing programs targeting people who inject drugs (PWID) by 2030. This could lead to 50,000 more new HIV infections (almost 10% more) and 40,000 more HIV-related deaths (almost 15% more) being averted. While ART remains the most cost-effective program for preventing HIV-related deaths, it is also highly effective at reducing incidence by providing a While ART remains the most preventative mechanism for those on treatment to limit new cost-effective program for HIV infections. After scale-up of treatment for all diagnosed preventing HIV-related deaths, people living with HIV, the next priority to improve outcomes it is also highly effective at across the cascade of care and avert infections and deaths is reducing incidence by to increase diagnoses by scaling-up testing. This is achieved providing a preventative by increasing the coverage of HIV prevention programs mechanism for those on targeting key populations that include a testing component treatment to limit new HIV as well as the HIV testing services program. Those newly infections. diagnosed can then be put on treatment. It is important to note that outcomes shown for 2025 (end of the new 2020/21–2024/25 KASFII period) and for the 10-year horizon (2030 to match the Kenya Vision 2030) used for this analysis were too short for the impact of certain interventions, such as VMMC, to be realized and whose effectiveness may therefore be undervalued in this study (11). Optimized allocation recommendations will need to be tempered by logistical realities. As such, there may be economic, programmatic, and even ethical reasons why certain HIV services would not be defunded completely in any county. ► The current HIV budget should be maintained at minimum to avoid reversing the gains made in the HIV response. Decreasing the latest reported budget by 50% is estimated to potentially result in 84,000 more new HIV infections (5% more) and 74,000 more HIV-related deaths (6% more) over the 2019 to 2030 period when compared to the status quo. To continue progress in reducing the HIV epidemic at least maintaining the HIV budget is therefore recommended. Effective budget increases are possible through, for example, implementation efficiency gains not explored by this analysis, such as using more optimal service delivery modalities, reduced cost of antiviral regimens, and reduced spending on non- targeted programs, among others. 27 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK ► Additional interventions and innovations to further reduce service delivery costs and increase effectiveness will be required if Kenya is to reach the 2030 target to end AIDS as a public health threat. Even with a doubling of budget for the HIV response optimized within all counties, the 2030 HIV incidence reductions targets are unlikely to be met, meaning there are diminishing marginal returns with the current available ‘toolbox’ of interventions. In countries with large existing disease burdens such as Kenya, reducing HIV incidence to such low levels will need personalized and pre-emptive HIV prevention strategies. 28 SECTION 7 CONCLUSION 1. SIGNIFICANT PROGRESS HAS BEEN MADE IN THE HIV RESPONSE IN KENYA. It is estimated that HIV spending over the 2015–2019 KASFI period led to the prevention of half a million new HIV infections and HIV-related deaths. However, there are still an estimated 1.4 million people living with HIV in the country, the fifth highest national burden worldwide. The country has achieved remarkable reductions in new HIV infections and HIV-related deaths over the last decade. Driving these improvements has been the progress towards the 95-95-95 targets. There have been great achievements in male circumcision and for HIV service coverage among key populations. However, there is significant disparity in ART program performance across the counties and double the proportion of women are estimated to be living with HIV than men. 2. EVEN GREATER IMPACT COULD BE ACHIEVED THROUGH ADDITIONAL ALLOCATIVE EFFICIENCY OF THE LATEST REPORTED HIV BUDGET. National HIV incidence is declining, but so is external donor funding for HIV, which the country currently relies on to fund 70% of its HIV response. For the same amount of total funding, more infections and deaths could be averted if the highest impact programs are prioritized. It is imperative that funding for the HIV response should be maintained and this will likely require implementation efficiency gains to be made. Should more resources become available, HIV prevention interventions should be further prioritized. While financial resources may not be easily redistributed, nor reallocation be politically favorable, opportunities to allocate HIV financial resources more efficiently should be explored to achieve the biggest gain with the available resources. 3. THE PRIMARY BENEFIT OF OPTIMIZATION TO IMPROVE ALLOCATIVE AND IMPLEMENTATION EFFICIENCY LIES IN CREATING AN OBJECTIVE PLATFORM TO MAKE EVIDENCE-INFORMED RESOURCE DECISIONS. This is with the caveat that modelling relies on strong assumptions of data quality and the impact of targeted and non-targeted programs. Deploying the recommendations provided in this report should consider the costs and benefits of using optimization as a basis for resource allocation. 29 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK REFERENCES 1 Kenya Population-based HIV Impact Assessment (KENPHIA) 2018 Preliminary Report Nairobi: National AIDS and STI Control Programme (NASCOP) and the Ministry of Health, Kenya; 2020. 2. Global Burden of Disease Factsheet, Kenya Seattle, Washington: Institute for Health Metrics Evaluation; 2020 [Available from: http://www.healthdata.org/kenya]. 3. 2010–2011 Integrated Biological and Behavioural Surveillance Survey among Key Populations in Nairobi and Kisumu, Kenya. Nairobi: Ministry of Health and Sanitation; 2014. 4. Kenya National AIDS Spending Assessment 2017. Nairobi, Kenya: Ministry of Health, National AIDS Control Council; 2018. 5. Kenya AIDS Response Progress Report 2018. Nairobi, Kenya: National AIDS Control Council, Ministry of Health; 2019. 6. Kenya HIV County Profiles 2018. Nairobi, Kenya: Ministry of Health, National AIDS Control Council; 2019. 7. Kates J, Wexler A, Foundation KF, Lief E, UNAIDS. Donor Government Funding for HIV in Low- and Middle- Income Countries in 2018. San Francisco, California: Henry J Kaiser Family Foundation; 2019. 8. Kerr CC, Stuart RM, Gray RT, Shattock AJ, Fraser-Hurt N, Benedikt C, et al. Optima: A Model for HIV Epidemic Analysis, Program Prioritization, and Resource Optimization. Journal of Acquired Immune Deficiency Syndromes (1999). 2015;69(3):365-76. 9. Kerr CC, Dura-Bernal S, Smolinski TG, Chadderdon GL, Wilson DP. Optimization by Adaptive Stochastic Descent. Plos One. 2018;13(3):e0192944. 10. National AIDS and STI Control Programme (NASCOP), Kenya. Kenya AIDS Indicator Survey 2012: Final Report. Nairobi: NASCOP; 2014. 11. Shattock AJ, Kerr CC, Stuart RM, Masaki E, Fraser N, Benedikt C, et al. In the interests of time: improving HIV allocative efficiency modelling via optimal time-varying allocations. J Int AIDS Soc. 2016;19(1):20627. 30 APPENDICES APPENDIX 1: OPTIMA HIV MODEL This Appendix provides a brief technical overview of Optima. A more detailed summary of the model and methods is provided elsewhere. Optima is based on a dynamic, population-based HIV model. Figure A1.1 shows the disease progression implemented in the model. Optima tracks the entire population of people living with HIV (PLHIV) across stages of CD4 count, including key aspects of the antiretroviral therapy (ART) service delivery cascade. Figure A1.2 provides a summary of the populations and mixing patterns used in the Optima HIV. Figure A1.1 HIV health states compartments and transmission-related interactions across the care cascade represented in Optima HIV (8) Source: Optima consortium. 31 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Schematic diagram of the health state structure of the model. Each compartment represents a single population group with the specified health state while each arrow represents the movement of numbers of individuals between health states. All compartments except for susceptible represent people living with HIV. Death includes all causes of death. Figure A1.2 Risk-based population mixing patterns represented in Optima HIV (8) Source: Optima consortium The model uses a linked system of ordinary differential equations to track the movement of PLHIV between HIV health states; the full set of equations can be accessed via the Optima supplementary index provided in the overall population is partitioned in 2 ways: by population group and by HIV health state. Individuals are assigned to a given population group based on their dominant risk. 6 HIV infections occur through the interaction between different populations by regular, casual, or commercial (including transactional) sexual partnerships, through sharing of injecting equipment or through mother-to-child transmission. The force-of-infection is the rate at which uninfected individuals become infected, and it depends on the number and type of risk events to which individuals are exposed in a given period (either within their population groups or through interaction with other population groups) and the infection probability of each event. Mathematically, the force of-infection has the general form: where λ is the force-of-infection, β is the transmission probability of each event, and n is the effective number of at-risk events (i.e., n gives the average number of interaction events with HIV-infected people where HIV transmission may occur). 6 However, to capture important cross-modal types of transmission, relevant behavioral parameters can be set to non-zero values (e.g., males who inject drugs may engage in commercial sex; some MSM may have female sexual partners). 32 APPENDICES There is one force-of-infection term for each type of interaction [e.g., casual sexual relationships between male sex workers and female sex workers (FSW)]; the force-of-infection for a given population will be the sum of all interaction types. In addition to the force-of-infection rate, which is the number of individuals who become infected with HIV per year, there are 7 other ways individuals may change health states. The change in the number of people in each compartment is determined by the sum over the relevant rates described above, multiplied by the population size of the compartments on which they act. 33 34 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK APPENDIX 2: KEY MODEL PARAMETERS Table A2.1 Latest reported key data and estimates used to inform the Optima HIV county models for Kenya HIV prevalence M15–49 HIV prevalence F15–49 HIV prevalence Clients HIV prevalence PWID HIV prevalence MSM HIV prevalence FSW 0–14 ART coverage 15+ ART coverage PMTCT coverage PLHIV all ages 0–14 on ART 15+ on ART 0–14 LHIV 15+ LHIV County Kenya 4.5% 5.2% 1,388,169 105,213 1,493,382 1,035,618 75% 86,325 82% 76% 11.0% 17.2% 5.3% 6.1% Baringo 1.1% 1.9% 5,397 477 5,874 3,450 64% 364 76% 72% 4.0% 4.4% 1.4% 1.6% Bomet 1.6% 2.7% 9,761 863 10,624 9,186 94% 850 99% 91% 5.9% 6.5% 1.9% 2.2% Bungoma 2.4% 3.9% 27,648 2,396 30,044 22,375 81% 2,104 88% 76% 7.1% 7.2% 2.7% 3.0% Busia 5.7% 9.4% 35,527 3,078 38,606 33,654 95% 2,415 78% 67% 17.1% 17.2% 6.4% 7.2% Elgeyo Marakwet 1.4% 2.3% 4,400 389 4,789 2,540 58% 232 60% 76% 4.9% 5.3% 1.6% 1.9% Embu 1.6% 3.8% 9,866 855 10,721 7,846 80% 747 87% 70% 7.9% 6.3% 1.9% 2.2% Garissa 0.3% 1.4% 2,356 532 2,888 1,296 55% 71 13% 16% 2.9% 1.2% 0.4% 0.4% Homa Bay 19.1% 22.1% 128,199 10,722 138,921 100,677 79% 9,655 90% 90% 47.7% 41.8% 22.6% 26.0% Isiolo 1.9% 4.3% 2,889 250 3,139 1,972 68% 248 99% 44% 9.1% 6.9% 2.2% 2.5% Kajiado 3.3% 5.5% 22,850 2,020 24,869 11,658 51% 845 42% 77% 11.6% 12.5% 3.9% 4.4% Kakamega 3.4% 5.6% 48,752 4,224 52,976 40,525 83% 3,977 94% 66% 10.1% 10.3% 3.8% 4.3% Kericho 2.4% 4.1% 16,111 1,424 17,535 16,045 100% 1,462 NA 63% 8.6% 9.1% 2.9% 3.3% Kiambu 2.1% 5.9% 56,622 2,394 59,016 34,417 61% 1,972 82% NA 13.5% 8.8% 2.5% 3.0% Kilifi 2.3% 5.4% 30,597 2,422 33,019 21,224 69% 2,233 92% 69% 11.3% 8.5% 2.7% 3.1% Kirinyaga 1.7% 4.6% 13,893 588 14,481 9,074 65% 625 NA 83% 8.9% 5.3% 1.9% 2.1% Kisii 4.0% 4.7% 34,950 2,923 37,874 27,571 79% 2,337 80% 55% 9.3% 14.0% 4.7% 5.3% Table continued… Table A2.1 Latest reported key data and estimates used to inform the Optima HIV county models for Kenya (continued) HIV prevalence M15–49 HIV prevalence F15–49 HIV prevalence Clients HIV prevalence PWID HIV prevalence MSM HIV prevalence FSW 0–14 ART coverage 15+ ART coverage PMTCT coverage PLHIV all ages 0–14 on ART 15+ on ART 0–14 LHIV 15+ LHIV County Kisumu 15.0% 17.4% 112,862 9,439 122,301 101,527 90% 8,225 87% 87% 36.9% 57.2% 17.6% 20.3% Kitui 2.7% 6.1% 26,375 2,286 28,661 17,257 65% 2,003 88% 53% 11.7% 8.7% 3.0% 3.4% Kwale 2.3% 5.4% 17,877 1,415 19,292 7,286 41% 785 55% 49% 13.1% 10.4% 2.8% 3.3% Laikipia 2.3% 3.8% 8,530 754 9,284 5,386 63% 457 61% 68% 8.1% 8.8% 2.7% 3.1% Lamu 1.8% 4.3% 2,445 194 2,638 954 39% 99 51% 77% 8.8% 6.7% 2.1% 2.4% Machakos 2.2% 5.1% 27,695 2,400 30,095 22,712 82% 2,148 90% 81% 10.7% 8.5% 2.6% 3.0% Makueni 2.5% 5.7% 22,621 1,960 24,581 15,841 70% 1,719 88% 55% 12.1% 9.4% 2.9% 3.4% Mandera 0.1% 0.3% 805 182 987 445 55% 39 21% 5% 1.5% 0.8% 0.1% 0.2% Marsabit 0.8% 1.8% 2,372 206 2,577 1,155 49% 160 78% 28% 3.7% 2.9% 0.9% 1.1% Meru 1.4% 3.3% 22,090 1,914 24,005 17,283 78% 1,649 86% 58% 6.9% 5.3% 1.7% 1.9% Migori 12.2% 14.2% 79,146 6,619 85,765 65,820 83% 6,175 93% 93% 28.1% 42.5% 14.1% 16.0% Mombasa 2.5% 5.9% 38,548 3,051 41,599 41,748 NA 2,630 86% 100% 15.7% 13.1% 3.2% 3.8% Murang'a 2.2% 6.2% 29,144 1,232 30,376 12,922 44% 935 76% 59% 11.9% 7.3% 2.5% 2.9% Nairobi 4.7% 7.5% 182,856 8,137 190,993 140,724 77% 7,611 94% 90% 14.5% 15.6% 5.4% 6.1% Nakuru 2.9% 4.8% 45,549 4,026 49,575 37,619 83% 2,963 74% 81% 10.2% 11.0% 3.4% 3.9% Nandi 1.7% 2.9% 11,712 1,035 12,748 7,681 66% 667 64% 78% 6.1% 6.6% 2.0% 2.3% Narok 2.3% 3.9% 16,810 1,486 18,296 7,512 45% 887 60% 54% 7.8% 8.3% 2.7% 3.1% Nyamira 3.9% 4.5% 17,537 1,467 19,004 13,439 77% 1,357 93% 53% 10.8% 17.5% 4.7% 5.6% Nyandarua 1.9% 5.2% 15,355 649 16,005 5,944 39% 539 83% 83% 9.3% 5.6% 2.1% 2.3% Nyeri 1.9% 5.5% 20,559 869 21,428 12,643 61% 732 84% 81% 10.1% 6.1% 2.2% 2.5% Table continued… APPENDICES 35 36 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Table A2.1 Latest reported key data and estimates used to inform the Optima HIV county models for Kenya (continued) HIV prevalence M15–49 HIV prevalence F15–49 HIV prevalence Clients HIV prevalence PWID HIV prevalence MSM HIV prevalence FSW 0–14 ART coverage 15+ ART coverage PMTCT coverage PLHIV all ages 0–14 on ART 15+ on ART 0–14 LHIV 15+ LHIV County Samburu 1.5% 2.5% 2,820 249 3,069 1,197 42% 232 93% 32% 5.4% 5.7% 1.8% 2.1% Siaya 19.4% 22.4% 113,605 9,501 123,107 80,123 71% 7,462 79% 73% 46.7% 72.0% 22.7% 26.0% Taita Taveta 2.5% 5.8% 9,462 749 10,211 4,710 50% 352 47% 37% 14.0% 11.2% 3.0% 3.6% Tana River 0.8% 1.8% 2,071 164 2,235 657 32% 74 45% 24% 4.3% 3.5% 0.9% 1.1% Tharaka Nithi 1.9% 4.4% 7,779 674 8,453 6,022 77% 507 75% 39% 9.3% 7.2% 2.2% 2.6% Trans Nzoia 3.7% 6.1% 26,610 2,352 28,962 12,510 47% 1,144 49% 44% 12.9% 13.8% 4.3% 4.9% Turkana 2.7% 4.5% 21,343 1,887 23,230 4,945 23% 713 38% 48% 9.8% 10.6% 3.2% 3.7% Uasin Gishu 3.3% 5.5% 29,640 2,620 32,260 29,557 100% 2,067 79% 71% 11.7% 12.6% 3.9% 4.5% Vihiga 4.0% 6.7% 18,346 1,590 19,935 13,131 72% 1,370 86% 49% 12.0% 12.1% 4.5% 5.1% Wajir 0.03% 0.2% 262 59 321 194 74% 9 15% 0% 0.9% 0.5% 0.1% 0.1% West Pokot 1.3% 2.2% 5,524 488 6,012 3,164 57% 478 98% 45% 4.3% 4.3% 1.5% 1.7% Sources: 2018 County Profile reports, other than for HIV prevalence estimates derived from 2010 IBBS values. Note: ART = antiretroviral therapy; F = female; FSW = female sex worker; LHIV = living with HIV; M = male; MSM = men who have sex with men; PWID = people living with HIV. APPENDICES Table A2.2 Key parameters used to inform the Optima HIV county models for Kenya Interaction-related transmissibility (% per act) Insertive penile-vaginal intercourse 0.04% Receptive penile-vaginal intercourse 0.08% Insertive penile-anal intercourse 0.09% Receptive penile-anal intercourse 1.38% Intravenous injection 0.80% Mother-to-child (breastfeeding) 36.70% Mother-to-child (non-breastfeeding) 20.50% Relative disease-related transmissibility Acute infection 5.60 CD4 (>500) 1.00 CD4 (500) to CD4 (350–500) 1.00 CD4 (200–350) 1.00 CD4 (50–200) 3.49 CD4 (<50) 7.17 Disease progression (average years to move) Acute to CD4 (>500) 0.3 CD4 (500) to CD4 (350–500) 1.11 CD4 (350–500) to CD4 (200–350) 3.10 CD4 (200–350) to CD4 (50–200) 3.90 CD4 (50–200) to CD4 (<50) 1.9 Changes in transmissibility (%) Condom use 95% Circumcision 58% Diagnosis behavior change 0% STI cofactor increase 265% Opiate substitution therapy 54% PMTCT 90% Unsuppressive ART 50% Suppressive ART 92% Disutility weights Untreated HIV, acute 0.15 Untreated HIV, CD4 (>500) 0.01 Untreated HIV, CD4 (350–500) 0.02 Untreated HIV, CD4 (200–350) 0.07 Untreated HIV, CD4 (50–200) 0.27 Untreated HIV, CD4 (<50) 0.55 Treated HIV 0.05 Treatment recovery due to suppressive ART (average years to move) CD4 (350–500) to CD4 (>500) 2.20 CD4 (200–350) to CD4 (350–500) 1.42 CD4 (50–200) to CD4 (200–350) 2.14 CD4 (<50) to CD4 (50–200) 0.66 Time after initiating ART to achieve viral suppression (years) 0.20 Number of VL tests recommended per person per year 1.00 Table A2.2 continued… 37 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Table A2.2 Key parameters used to inform the Optima HIV county models for Kenya (continued) CD4 change due to non-suppressive ART (%/year) CD4 (500) to CD4 (350–500) 3% CD4 (350–500) to CD4 (>500) 15% CD4 (350–500) to CD4 (200–350) 10% CD4 (200–350) to CD4 (350–500) 5% CD4 (200–350) to CD4 (50–200) 16% CD4 (50–200) to CD4 (200–350) 12% CD4 (50–200) to CD4 (<50) 9% CD4 (<50) to CD4 (50–200) 11% Death rate (% mortality per year) Acute infection 0% CD4 (>500) 0% CD4 (350–500) 1% CD4 (200–350) 1% CD4 (50–200) 8% CD4 (<50) 43% Relative death rate on suppressive ART 30% Relative death rate on non-suppressive ART 70% Tuberculosis cofactor 217% Source: Authors from Optima data. Note: ART = antiretroviral therapy 38 APPENDICES APPENDIX 3: EXAMPLE MODEL CALIBRATION FOR NAIROBI COUNTY 39 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 40 APPENDICES Source: Authors from Optima data. 41 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK APPENDIX 4: HIV PROGRAM UNIT COSTS Table A4.1 HIV program unit costs at the county and national level Unit costs, 2018 HIV programs, county level Low bound High bound HIV prevention services (condoms and SBCC) $3.00 $4.00 HIV testing and prevention programs for FSW $13.00 $15.00 HIV testing and prevention programs for MSM $8.16 $16.32 HIV testing and prevention programs for PWID $13.00 $20.00 Voluntary medical male circumcision (VMMC) $33.91 $42.40 HIV programs, national level Unit costs, 2018 Antiretroviral therapy (ART) $273.27 Prevention of mother-to-child transmission (PMTCT) $638.62 HIV testing services (biomedical only) $6.50 HIV prevention services (condoms and SBCC) $3.22 HIV testing and prevention programs for FSW $14.39 HIV testing and prevention programs for MSM $8.16 HIV testing and prevention programs for PWID $13.33 Voluntary medical male circumcision (VMMC) $33.91 Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PMTCT = prevention of mother-to-child transmission; PWID = people living with HIV; SBCC = social behaviour change communication; VMMC = Voluntary male medical circumcision. Table A4.2 HIV program unit costs for care and treatment by county Care and treatment Care and treatment County unit cost, 2018 County unit cost, 2018 Baringo $881.25 Marsabit $1,426.03 Bomet $370.20 Meru $714.07 Bungoma $235.02 Migori $394.69 Busia $303.09 Mombasa $347.06 Elgeyo Marakwet $1,580.66 Muranga $245.41 Embu $1,058.41 Nairobi $516.18 Garissa $2,982.66 Nakuru $653.51 Homa Bay $324.85 Nandi $389.29 Isiolo $1,296.15 Narok $769.98 Kajiado $981.17 Nyamira $676.83 Kakamega $497.08 Nyandarua $676.25 Kericho $440.06 Nyeri $484.17 Kiambu $85.37 Samburu $1,726.41 Kilifi $653.93 Siaya $547.88 Kilifi $646.24 Taita Taveta $975.21 Kirinyaga $672.35 Tana River $1,410.73 Kisii $414.48 Tharaka Nithi $3,166.24 Kisumu $694.92 Trans Nzoia $465.65 Table A4.2 continued… 42 APPENDICES Table A4.2 HIV program unit costs for care and treatment by county (continued) Care and treatment unit Care and treatment County cost, 2018 County unit cost, 2018 Kitui $681.12 Turkana $364.25 Kwale $448.63 Uasin Gishu $384.90 Laikipia $1,205.46 Vihiga $598.99 Lamu $123.70 Wajir $9,383.46 Machakos $742.17 West Pokot $905.79 Makueni $3,677.23 Median $663.14 Mandera $881.25 Minimum $85.37 Maximum $9,383.46 Source: Authors from Optima data. Table A4.3 HIV program unit costs for HIV testing services by county HIV testing services HIV testing services Unit costs, 2018 Unit costs, 2018 Low High Low High County bound bound County bound bound Baringo $6.00 $8.00 Mandera $6.00 $7.00 Bomet $6.00 $8.00 Marsabit $6.00 $7.00 Bungoma $6.00 $7.00 Meru $6.00 $7.00 Busia $2.00 $2.16 Migori $0.75 $2.00 Elgeyo Marakwet $6.00 $7.00 Mombasa $2.60 $4.00 Embu $6.00 $7.00 Muranga $2.75 $3.50 Garissa $6.00 $7.00 Nairobi $6.00 $7.00 Homa Bay $2.00 $3.50 Nakuru $5.00 $6.60 Isiolo $6.00 $7.00 Nandi $6.00 $6.30 Kajiado $5.70 $7.00 Narok $5.50 $7.00 Kakamega $6.00 $7.00 Nyamira $4.00 $6.00 Kericho $6.00 $7.00 Nyandarua $5.50 $7.00 Kiambu $5.20 $6.00 Nyeri $5.20 $7.00 Kilifi $5.70 $7.00 Samburu $5.70 $7.00 Kilifi $5.70 $7.00 Siaya $0.50 $1.70 Kirinyaga $5.10 $5.90 Taita Taveta $6.00 $7.00 Kisii $4.00 $5.30 Tana River $5.50 $7.00 Kisumu $1.50 $3.00 Tharaka Nithi $5.00 $7.40 Kitui $5.00 $7.00 Trans Nzoia $5.50 $7.00 Kwale $6.00 $7.00 Turkana $5.50 $7.00 Laikipia $5.70 $7.00 Uasin Gishu $5.00 $7.00 Lamu $6.00 $7.00 Vihiga $5.50 $7.00 Machakos $2.25 $3.50 Wajir $6.00 $7.00 Makueni $6.00 $8.00 West Pokot $6.00 $8.00 Median $5.70 $7.00 Minimum, maximum $0.50 $8.00 Source: Authors from Optima data. 43 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK APPENDIX 5: EXAMPLE COST COVERAGE CURVES FOR NAIROBI COUNTY HIV testing services (biomedical only) HIV prevention services (condoms and SBCC) 44 APPENDICES 45 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK Source: Authors from Optima data. 46 APPENDIX 6: SUPPORTING DATA FOR RESULTS AND ADDITIONAL RESULTS Table A6.1 100% annual optimized allocation within counties represented nationally for 2019 to 2030, and percent of the national budget Latest reported 100% budget optimized within % National budget HIV programs national budget counties, represented nationally Latest reported budget 100% budget optimized Care and treatment $538,574,507 $576,905,275 90% 96% HIV testing services (biomedical only) $45,147,988 $15,467,543 8% 3% VMMC $8,105,135 $3,353,402 1.4% 0.6% HIV testing and prevention for FSW $760,474 $2,436,503 0.1% 0.4% HIV prevention services (condoms and SBCC) $6,013,639 $399,169 1.01% 0.07% HIV testing and prevention for PWID $21,615 $68,805 0.004% 0.011% HIV testing and prevention for MSM $28,043 $20,704 0.005% 0.003% Total $598,651,401 $598,651,401 100% 100% Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; SBCC = social behaviour change communication ; VMMC = Voluntary male medical circumcision. Table A6.2 Changing levels of HIV budgets optimized within counties for 2019 to 2030, represented nationally 100% 50% 90% 100% 110% 150% 200% HIV programs latest reported optimized optimized optimized optimized optimized optimized Care and treatment $538,574,507 $293,132,793 $528,210,917 $576,905,275 $628,109,660 $825,658,933 $1,086,351,556 HIV testing services (biomedical only) $45,147,988 $692,481 $5,077,022 $15,467,543 $23,149,900 $53,570,225 $73,304,590 HIV prevention services (condoms and SBCC) $6,013,639 $0 $87,497 $399,169 $1,222,746 $11,384,574 $28,641,997 VMMC $8,105,135 $3,134,817 $3,085,941 $3,353,402 $3,478,321 $4,539,827 $5,777,675 HIV testing and prevention for FSW $760,474 $2,302,012 $2,260,502 $2,436,503 $2,444,255 $2,654,661 $3,008,416 HIV testing and prevention for MSM $28,043 $6,680 $9,626 $20,704 $35,719 $81,799 $117,114 HIV testing and prevention for PWID $21,615 $56,918 $54,756 $68,805 $75,940 $87,083 $101,455 Total $598,651,401 $299,325,701 $538,786,261 $598,651,401 $658,516,541 $897,977,102 $1,197,302,802 Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; SBCC = social behaviour change communication ; VMMC = Voluntary male medical APPENDICES circumcision. 47 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 48 Table A6.3 Percentage of budget allocated by HIV program for changing budgets optimized within counties represented nationally for 2019 to 2030 100% latest 50% 90% 100% 110% 150% 200% HIV programs reported optimized optimized optimized optimized optimized optimized Care and treatment 90% 98% 98% 96% 95% 92% 91% HIV testing services (biomedical only) 8% 0.2% 0.9% 3% 4% 6% 6% HIV prevention services (condoms and SBCC) 1% 0% 0.02% 0.07% 0.2% 1% 2% VMMC 1% 1% 0.6% 0.6% 0.5% 0.5% 0.5% HIV testing and prevention for FSW 0.1% 0.8% 0.4% 0.4% 0.4% 0.3% 0.3% HIV testing and prevention for MSM 0.005% 0.002% 0.002% 0.003% 0.005% 0.009% 0.010% HIV testing and prevention for PWID 0.004% 0.019% 0.010% 0.011% 0.012% 0.010% 0.008% Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people living with HIV; SBCC = social behaviour change communication ; VMMC = Voluntary male medical circumcision. Table A6.4 Percentage of total budget allocated to HIV prevention and treatment with changing budgets levels optimized within counties represented nationally for 2019 to 2030 100% latest 50% 90% 100% 110% 150% 200% HIV program categories reported optimized optimized optimized optimized optimized optimized Total prevention 10.0% 2.1% 2.0% 3.6% 4.6% 8.1% 9.3% Total treatment 90.0% 97.9% 98.0% 96.4% 95.4% 91.9% 90.7% Source: Authors. Table A6.5 HIV program coverage levels at 100% HIV budget optimized within counties represented nationally for 2019 to 2030 HIV programs 100% latest reported 100% optimized Care and treatment (ART and PMTCT) 600,000 640,000 HIV testing services (biomedical only) 6,400,000 2,200,000 HIV prevention services (condoms and SBCC) 1,700,000 110,000 VMMC 210,000 87,000 HIV testing and prevention for FSW 54,000 170,000 HIV testing and prevention for MSM 2,300 1,700 HIV testing and prevention for PWID 1,300 4,200 Source: Authors from Optima data. Note: ART = antiretroviral therapy; FSW = female sex worker; MSM = men who have sex with men; PMTCT = prevention of mother-to-child transmission; PWID = people living with HIV; SBCC = social behaviour change communication ; VMMC = Voluntary male medical circumcision. Figure A6.1 50% annual HIV budget optimized within counties $45 $40 $35 Bars per county Latest reported (top) Optimized (bottom) $30 $25 Millions USD $20 $15 $10 $5 $0 Nakuru Marsabit Vihiga Nyeri Nyandarua Nyamira Kwale Bomet Wajir Turkana Kitui Kisii Siaya Nairobi Meru Laikipia Busia Machakos Kericho Homa Bay Baringo Narok Kisumu Embu Tharaka Nithi Tana River Nandi Samburu Mombasa Migori Kirinyaga Isiolo Garissa Muranga Mandera Kakamega Uasin Gishu Trans Nzoia Makueni Lamu Kajiado Elgeyo Marakwet Taita Taveta Kilifi Kiambu Bungoma West Pokot Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. APPENDICES Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, VMMC = voluntary male medical circumcision. 49 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 50 Figure A6.2 90% annual HIV budget optimized within counties $80 $70 Bars per county $60 Latest reported (top) Optimized (bottom) $50 Millions USD $40 $30 $20 $10 $0 Kirinyaga Kwale Nyamira Kisumu Siaya Meru Busia Bomet Wajir Vihiga Turkana Nyandarua Laikipia Kitui Kisii Nyeri Nakuru Nairobi Machakos Kericho Homa Bay Baringo Narok Embu Migori Marsabit Tharaka Nithi Tana River Nandi Garissa Samburu Mombasa Isiolo Trans Nzoia Muranga Mandera Makueni Lamu Kilifi Kiambu Kakamega Kajiado Uasin Gishu Elgeyo Marakwet Taita Taveta Bungoma West Pokot Care and treatment HVI testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, VMMC = voluntary male medical circumcision. Figure A6.3 110% annual HIV budget optimized within counties $100 $90 $80 Bars per county $70 Latest reported (top) Optimized (bottom) $60 Millions USD $50 $40 $30 $20 $10 $0 Uasin Gishu Kwale Turkana Siaya Nyeri Meru Kisii Busia Wajir Vihiga Nyandarua Nyamira Nairobi Machakos Kitui Bomet Nakuru Laikipia Kisumu Tana River Narok Kirinyaga Kericho Homa Bay Baringo Embu Trans Nzoia Nandi Muranga Mombasa Migori Marsabit Garissa Tharaka Nithi Elgeyo Marakwet Samburu Makueni Kilifi Mandera Kiambu Isiolo Kakamega Kajiado Lamu Taita Taveta Bungoma West Pokot Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs APPENDICES Source: Authors. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, VMMC = voluntary male medical circumcision. 51 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 52 Figure A6.4 150% annual HIV budget optimized within counties $140 $120 $100 $80 Millions USD $60 $40 $20 $0 Vihiga Nyandarua Nyamira Kwale Bomet Wajir Kisii Turkana Siaya Nyeri Meru Kitui Nairobi Laikipia Busia Narok Machakos Embu Baringo Nakuru Kisumu Kericho Homa Bay Tharaka Nithi Tana River Nandi Samburu Mombasa Migori Muranga Marsabit Kirinyaga Isiolo Trans Nzoia Mandera Makueni Kakamega Garissa Uasin Gishu Lamu Kajiado Elgeyo Marakwet Kilifi Kiambu Bungoma West Pokot Taita Taveta Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, VMMC = voluntary male medical circumcision. Figure A6.5 200% annual HIV budget optimized within counties $180 $160 $140 $120 Millions USD $100 $80 $60 $40 $20 $0 Siaya Kwale Kisumu Meru Bomet Wajir Vihiga Turkana Nyeri Nyandarua Nyamira Nakuru Laikipia Kitui Kisii Nairobi Busia Machakos Kericho Homa Bay Baringo Narok Embu Migori Marsabit Tharaka Nithi Tana River Nandi Garissa Samburu Mombasa Kirinyaga Isiolo Trans Nzoia Muranga Mandera Makueni Kilifi Kiambu Kakamega Uasin Gishu Lamu Kajiado Elgeyo Marakwet Taita Taveta Bungoma West Pokot Care and treatment HIV testing services HIV prevention services VMMC FSW programs MSM programs PWID programs Source: Authors from Optima data. APPENDICES Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, VMMC = voluntary male medical circumcision. 53 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 54 Table A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 HIV prevention HIV testing services Care and services (condoms and FSW MSM PWID Total County Scenario treatment (biomedical only) SBCC) VMMC programs programs programs budget Baringo Latest reported $3,250,922 $487,240 $41,345 $0 $2,170 $0 $0 $3,781,677 Optimized $3,771,903 $0 $0 $0 $9,774 $0 $0 $3,781,677 Bomet Latest reported $3,795,302 $732,895 $138,331 $43,642 $3,173 $0 $0 $4,713,343 Optimized $4,702,819 $0 $0 $0 $10,524 $0 $0 $4,713,343 Bungoma Latest reported $5,107,718 $1,293,019 $80,951 $0 $38,316 $900 $755 $6,521,659 Optimized $5,845,066 $609,337 $0 $0 $65,342 $220 $1,694 $6,521,659 Busia Latest reported $9,664,843 $516,681 $142,195 $456,395 $28,791 $676 $567 $10,810,148 Optimized $10,135,237 $310,560 $0 $291,834 $70,326 $244 $1,947 $10,810,148 Elgeyo Latest reported $4,342,077 $353,945 $47,141 $0 $1,507 $0 $0 $4,744,670 Marakwet Optimized $4,736,337 $0 $0 $0 $8,333 $0 $0 $4,744,670 Embu Latest reported $9,183,850 $541,821 $127,319 $0 $5,192 $61 $234 $9,858,477 Optimized $9,836,136 $0 $0 $0 $21,176 $0 $1,165 $9,858,477 Garissa Latest reported $3,746,225 $129,987 $17,968 $0 $6,996 $600 $414 $3,902,190 Optimized $3,875,435 $0 $0 $0 $25,511 $0 $1,243 $3,902,190 Homa Bay Latest reported $34,111,042 $1,925,676 $192,814 $1,513,878 $5,368 $1,952 $151 $37,750,881 Optimized $37,149,352 $0 $0 $572,005 $28,455 $0 $1,069 $37,750,881 Isiolo Latest reported $2,264,373 $78,072 $17,388 $102 $1,962 $42 $20 $2,361,959 Optimized $2,351,989 $0 $0 $102 $9,737 $0 $131 $2,361,959 Kajiado Latest reported $10,600,549 $849,895 $22,411 $407 $24,029 $62 $25 $11,497,378 Optimized $11,445,180 $0 $0 $407 $51,739 $0 $52 $11,497,378 Kakamega Latest reported $19,811,955 $1,856,134 $187,211 $0 $11,470 $1,220 $1,024 $21,869,014 Optimized $19,811,955 $2,009,095 $0 $0 $46,731 $806 $427 $21,869,014 Table continued… Table A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 (continued) HIV prevention HIV testing services Care and services (condoms and FSW MSM PWID Total County Scenario treatment (biomedical only) SBCC) VMMC programs programs programs budget Kericho Latest reported $6,551,128 $1,132,027 $107,612 $226,960 $5,948 $178 $60 $8,023,913 Optimized $7,809,966 $0 $0 $191,798 $21,946 $0 $204 $8,023,913 Kiambu Latest reported $3,213,795 $1,831,363 $221,987 $0 $51,954 $1,220 $1,024 $5,321,343 Optimized $5,124,804 $115,277 $0 $0 $75,217 $858 $5,189 $5,321,343 Kilifi Latest reported $13,772,970 $999,980 $97,744 $1,356 $30,956 $727 $610 $14,904,343 Optimized $13,772,970 $996,804 $90,126 $1,356 $40,115 $2,971 $0 $14,904,343 Kirinyaga Latest reported $6,573,568 $740,917 $67,620 $0 $3,933 $0 $422 $7,386,460 Optimized $7,361,983 $7,409 $0 $0 $17,068 $0 $0 $7,386,460 Kisii Latest reported $19,800,751 $1,704,517 $113,022 $0 $45,026 $1,058 $887 $21,665,261 Optimized $20,003,096 $1,550,384 $0 $0 $108,785 $2,532 $464 $21,665,261 Kisumu Latest reported $42,147,243 $2,337,739 $269,869 $1,797,027 $21,665 $2,446 $2,052 $46,578,041 Optimized $45,572,660 $379,160 $0 $540,418 $77,305 $1,620 $6,878 $46,578,041 Kitui Latest reported $12,767,821 $1,353,235 $2,941 $0 $4,014 $951 $0 $14,128,962 Optimized $13,802,865 $304,779 $0 $0 $20,034 $1,285 $0 $14,128,962 Kwale Latest reported $6,380,058 $363,766 $71,870 $610 $19,050 $447 $375 $6,836,176 Optimized $6,796,896 $0 $0 $610 $38,669 $0 $0 $6,836,176 Laikipia Latest reported $3,697,176 $494,462 $58,346 $610 $19,483 $67 $384 $4,270,528 Optimized $4,238,983 $0 $0 $610 $30,415 $0 $520 $4,270,528 Lamu Latest reported $1,694,873 $79,606 $17,388 $0 $2,372 $178 $149 $1,794,566 Optimized $1,783,401 $0 $0 $0 $11,165 $0 $0 $1,794,566 Machakos Latest reported $3,020,338 $879,099 $146,446 $0 $37,450 $1,307 $429 $4,085,069 Optimized $3,262,914 $750,664 $0 $0 $68,318 $2,815 $358 $4,085,069 Makueni Latest reported $13,301,166 $966,526 $107,469 $0 $27,059 $98 $0 $14,402,318 Optimized $13,660,541 $379,793 $309,043 $0 $52,680 $261 $0 $14,402,318 Table continued… APPENDICES 55 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 56 Table A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 (continued) HIV prevention HIV testing services Care and services (condoms and FSW MSM PWID Total County Scenario treatment (biomedical only) SBCC) VMMC programs programs programs budget Mandera Latest reported $1,941,577 $83,330 $25,502 $0 $7,577 $0 $0 $2,057,986 Optimized $2,042,013 $0 $0 $0 $15,973 $0 $0 $2,057,986 Marsabit Latest reported $1,908,023 $74,367 $15,649 $237 $869 $0 $0 $1,999,145 Optimized $1,995,023 $0 $0 $237 $3,885 $0 $0 $1,999,145 Meru Latest reported $12,639,113 $1,095,231 $93,887 $0 $40,048 $941 $789 $13,870,009 Optimized $13,676,015 $113,616 $0 $0 $80,378 $0 $0 $13,870,009 Migori Latest reported $26,994,656 $1,345,552 $159,776 $973,624 $12,265 $1,413 $1,186 $29,488,472 Optimized $28,198,654 $803,180 $0 $435,878 $43,286 $2,014 $5,460 $29,488,472 Mombasa Latest reported $15,023,512 $729,957 $347,976 $21,092 $25,760 $605 $508 $16,149,410 Optimized $15,906,507 $32,202 $0 $21,162 $165,917 $1,681 $21,940 $16,149,410 Muranga Latest reported $3,550,369 $507,725 $67,234 $0 $2,212 $412 $346 $4,128,298 Optimized $4,120,646 $0 $0 $0 $7,652 $0 $0 $4,128,298 Nairobi Latest reported $76,142,281 $5,358,997 $1,546,306 $516,992 $162,572 $3,818 $3,203 $83,734,169 Optimized $76,142,281 $6,906,679 $0 $0 $663,807 $3,221 $18,182 $83,734,169 Nakuru Latest reported $24,106,103 $2,905,864 $328,425 $238,625 $1,078 $2,471 $288 $27,582,854 Optimized $27,252,562 $0 $0 $243,922 $86,369 $0 $0 $27,582,854 Nandi Latest reported $3,952,806 $901,914 $52,917 $173,144 $3,084 $0 $0 $5,083,865 Optimized $4,892,897 $0 $0 $177,750 $13,218 $0 $0 $5,083,865 Narok Latest reported $6,478,647 $867,204 $51,264 $2,340 $3,284 $0 $0 $7,402,739 Optimized $7,382,472 $0 $0 $2,348 $17,919 $0 $0 $7,402,739 Nyamira Latest reported $9,043,146 $1,404,761 $50,234 $0 $4,586 $824 $0 $10,503,551 Optimized $10,481,225 $0 $0 $0 $22,326 $0 $0 $10,503,551 Nyandarua Latest reported $5,184,140 $584,987 $37,094 $0 $4,967 $0 $486 $5,811,674 Optimized $5,793,982 $0 $0 $0 $17,692 $0 $0 $5,811,674 Table continued… Table A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 (continued) HIV prevention HIV testing services Care and services (condoms and FSW MSM PWID Total County Scenario treatment (biomedical only) SBCC) VMMC programs programs programs budget Nyeri Latest reported $7,976,176 $1,011,784 $117,079 $0 $5,202 $0 $580 $9,110,821 Optimized $8,890,712 $198,605 $0 $0 $21,443 $0 $61 $9,110,821 Samburu Latest reported $1,836,898 $223,412 $19,320 $0 $847 $0 $0 $2,080,477 Optimized $2,075,555 $0 $0 $0 $4,922 $0 $0 $2,080,477 Siaya Latest reported $44,934,550 $1,154,392 $120,750 $1,735,955 $11,570 $1,637 $123 $47,958,977 Optimized $47,386,078 $0 $0 $507,568 $64,465 $0 $866 $47,958,977 Taita Latest reported $5,363,680 $221,761 $21,059 $0 $11,906 $0 $0 $5,618,406 Taveta Optimized $5,573,523 $0 $0 $0 $44,883 $0 $0 $5,618,406 Tana River Latest reported $1,096,139 $216,795 $123,648 $0 $903 $0 $0 $1,437,485 Optimized $1,432,790 $0 $0 $0 $4,695 $0 $0 $1,437,485 Tharaka Latest reported $19,257,062 $416,611 $26,275 $0 $5,219 $0 $341 $19,705,508 Nithi Optimized $19,679,213 $0 $0 $0 $25,911 $0 $383 $19,705,508 Trans Latest reported $5,783,377 $873,327 $121,890 $0 $4,630 $0 $0 $6,783,224 Nzoia Optimized $6,759,712 $0 $0 $0 $23,512 $0 $0 $6,783,224 Turkana Latest reported $2,606,202 $1,091,818 $109,544 $375,994 $14,584 $1,083 $909 $4,200,134 Optimized $3,804,389 $0 $0 $338,799 $56,946 $0 $0 $4,200,134 Uasin Latest reported $10,427,658 $1,382,875 $223,814 $0 $13,818 $222 $71 $12,048,458 Gishu Optimized $11,990,235 $0 $0 $0 $58,048 $175 $0 $12,048,458 Vihiga Latest reported $8,193,543 $633,159 $27,628 $0 $18,184 $427 $3,203 $8,876,144 Optimized $8,824,047 $0 $0 $0 $51,778 $0 $319 $8,876,144 Wajir Latest reported $2,036,212 $77,272 $14,490 $0 $1,693 $0 $0 $2,129,667 Optimized $2,119,658 $0 $0 $0 $9,756 $0 $252 $2,129,667 Table continued… APPENDICES 57 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 58 Table A6.6 100% annual HIV budget optimizations within counties, 2019 to 2030 (continued) HIV testing HIV prevention services services Care and (biomedical (condoms and FSW MSM PWID Total County Scenario treatment only) SBCC) VMMC programs programs programs budget West Latest reported $3,298,894 $336,291 $14,490 $26,145 $5,732 $0 $0 $3,681,552 Pokot Optimized $3,632,599 $0 $0 $26,597 $22,356 $0 $0 $3,681,552 Total Latest reported $538,574,507 $45,147,988 $6,013,639 $8,105,135 $760,474 $28,043 $21,615 $598,651,401 national Optimized $576,905,275 $15,467,543 $399,169 $3,353,402 $2,436,503 $20,704 $68,805 $598,651,401 90.0% 7.5% 1.0% 1.4% 0.1% 0.005% 0.004% 100.0% 96.4% 2.6% 0.1% 0.6% 0.4% 0.003% 0.011% 100.0% Source: Authors from Optima data. Note: Latest reported program budgets for each county were derived using the county budgets for treatment and for overall prevention; FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, SBCC = social behavior change communication; VMMC = voluntary male medical circumcision. Table A6.7 Percentage of total annual HIV budget to optimally reallocate within counties, 2019 to 2030 Percentage of total HIV budget recommended to allocate to each program (percentage difference of total budget between the optimized and latest reported allocation) by county HIV prevention HIV testing and Care and HIV testing services (condoms HIV testing and prevention for County treatment (biomedical) and SBCC) VMMC prevention for FSW PWID Total budget Baringo 99.7% (+13.8%) Not prioritized Not prioritized Not prioritized 0.3% (+0.2%) Not prioritized $3,781,677 Bomet 99.8% (+19.3%) Not prioritized Not prioritized Not prioritized 0.2% (+0.2%) Not prioritized $4,713,343 Bungoma 89.6% (+11.3%) 9.3% (-10.5%) Not prioritized Not prioritized 1.0% (+0.4%) 0.03% (+0.01%) $6,521,659 Busia 93.8% (+4.4%) 2.9% (-1.9%) Not prioritized 2.7% (-1.5%) 0.7% (+0.4%) 0.02% (+0.01%) $10,810,148 Elgeyo Marakwet 99.8% (+8.3%) Not prioritized Not prioritized Not prioritized 0.2% (+0.1%) Not prioritized $4,744,670 Embu 99.8% (+6.6%) Not prioritized Not prioritized Not prioritized 0.2% (+0.2%) 0.01% (+0.01%) $9,858,477 Garissa 99.3% (+3.3%) Not prioritized Not prioritized Not prioritized 0.7% (+0.5%) 0.03% (+0.02%) $3,902,190 Homa Bay 98.4% (+8.0%) Not prioritized Not prioritized 1.5% (-2.5%) 0.1% (+0.1%) 0.003% (+0.002%) $37,750,881 Isiolo 99.6% (+3.7%) Not prioritized Not prioritized 0.004% (+0.0%) 0.4% (+0.3%) 0.006% (+0.005%) $2,361,959 Kajiado 99.5% (+7.3%) Not prioritized Not prioritized 0.004% (+0.0%) 0.5% (+0.2%) Not prioritized $11,497,378 Kakamega 90.6% (+0.0%) 9.2% (+0.7%) Not prioritized Not prioritized 0.2% (+0.2%) 0.002% (-0.003%) $21,869,014 Kericho 97.3% (+15.7%) Not prioritized Not prioritized 2.4% (-0.4%) 0.3% (+0.2%) 0.003% (+0.002%) $8,023,913 Kiambu 96.3% (+35.9%) 2.2% (-32.2%) Not prioritized Not prioritized 1.4% (+0.4%) 0.1% (+0.1%) $5,321,343 Kilifi 92.4% (+0.0%) 6.7% (+0.0%) 0.6% (-0.001%) 0.01% (+0.0%) 0.2% (+0.0%) Not prioritized $14,904,343 Kirinyaga 99.7% (+10.7%) 0.1% (-9.9%) Not prioritized Not prioritized 0.2% (+0.1%) Not prioritized $7,386,460 Kisii 92.3% (+0.9%) 7.2% (-0.7%) Not prioritized Not prioritized 0.5% (+0.2%) 0.002% (-0.002%) $21,665,261 Kisumu 97.8% (+7.4%) 0.8% (-4.2%) Not prioritized 1.2% (-2.7%) 0.1% (+0.1%) 0.02% (+0.01%) $46,578,041 Kitui 97.7% (+7.3%) 2.2% (-7.4%) Not prioritized Not prioritized 0.1% (+0.1%) Not prioritized $14,128,962 Kwale 99.4% (+6.1%) Not prioritized Not prioritized 0.01% (+0.0%) 0.5% (+0.2%) Not prioritized $6,836,176 Laikipia 99.3% (+12.7%) Not prioritized Not prioritized 0.01% (+0.0%) 0.7% (+0.2%) 0.012% (+0.003%) $4,270,528 Lamu 99.4% (+4.9%) Not prioritized Not prioritized Not prioritized 0.6% (+0.4%) Not prioritized $1,794,566 Machakos 79.9% (+5.9%) 18.4% (-3.1%) Not prioritized Not prioritized 1.6% (+0.7%) 0.009% (-0.002%) $4,085,069 Makueni 94.8% (+2.5%) 2.6% (-4.1%) 2.1% (+0.01%) Not prioritized 0.3% (+0.1%) Not prioritized $14,402,318 Mandera 99.2% (+4.9%) Not prioritized Not prioritized Not prioritized 0.7% (+0.4%) Not prioritized $2,057,986 Marsabit 99.8% (+4.4%) Not prioritized Not prioritized 0.01% (+0.0%) 0.1% (+0.1%) Not prioritized $1,999,145 APPENDICES Table continued… 59 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 60 Table A6.7 Percentage of total annual HIV budget to optimally reallocate within counties, 2019 to 2030 (continued) Percentage of total HIV budget recommended to allocate to each program (percentage difference of total budget between the optimized and latest reported allocation) by county HIV prevention HIV testing and Care and HIV testing services (condoms HIV testing and prevention for County treatment (biomedical) and SBCC) VMMC prevention for FSW PWID Total budget Meru 98.6% (+7.5%) 0.8% (-7.1%) Not prioritized Not prioritized 0.5% (+0.2%) Not prioritized $13,870,009 Migori 95.6% (+4.1%) 2.7% (-1.8%) Not prioritized 1.5% (-1.8%) 0.1% (+0.1%) 0.02% (+0.01%) $29,488,472 Mombasa 98.5% (+5.5%) 0.2% (-4.3%) Not prioritized 0.1% (+0.0%) 1.0% (+0.8%) 0.1% (+0.1%) $16,149,410 Muranga 99.8% (+13.8%) Not prioritized Not prioritized Not prioritized 0.1% (+0.1%) Not prioritized $4,128,298 Nairobi 90.9% (+0.0%) 8.2% (+1.8%) Not prioritized Not prioritized 0.7% (+0.5%) 0.02% (+0.02%) $83,734,169 Nakuru 98.8% (+11.4%) Not prioritized Not prioritized 0.9% (+0.02%) 0.3% (+0.3%) Not prioritized $27,582,854 Nandi 96.2% (+18.5%) Not prioritized Not prioritized 3.5% (+0.1%) 0.2% (+0.1%) Not prioritized $5,083,865 Narok 99.7% (+12.2%) Not prioritized Not prioritized 0.03% (+0.0%) 0.2% (+0.1%) Not prioritized $7,402,739 Nyamira 99.8% (+13.7%) Not prioritized Not prioritized Not prioritized 0.2% (+0.1%) Not prioritized $10,503,551 Nyandarua 99.7% (+10.5%) Not prioritized Not prioritized Not prioritized 0.3% (+0.2%) Not prioritized $5,811,674 Nyeri 97.6% (+10.0%) 2.2% (-8.9%) Not prioritized Not prioritized 0.2% (+0.1%) 0.001% (-0.006%) $9,110,821 Samburu 99.8% (+11.5%) Not prioritized Not prioritized Not prioritized 0.2% (+0.1%) Not prioritized $2,080,477 Siaya 98.8% (+5.1%) Not prioritized Not prioritized 1.6% (-2.6%) 0.1% (+0.1%) 0.002% (+0.002%) $47,958,977 Taita Taveta 99.2% (+3.7%) Not prioritized Not prioritized Not prioritized 0.7% (+0.5%) Not prioritized $5,618,406 Tana River 99.7% (+23.4%) Not prioritized Not prioritized Not prioritized 0.3% (+0.2%) Not prioritized $1,437,485 Tharaka Nithi 99.9% (+2.1%) Not prioritized Not prioritized Not prioritized 0.1% (+0.1%) 0.002% (+0.0%) $19,705,508 Trans Nzoia 99.7% (+14.4%) Not prioritized Not prioritized Not prioritized 0.3% (+0.2%) Not prioritized $6,783,224 Turkana 90.6% (+28.5%) Not prioritized Not prioritized 8.1% (-0.9%) 1.3% (+1.0%) Not prioritized $4,200,134 Uasin Gishu 99.5% (+13.0%) Not prioritized Not prioritized Not prioritized 0.4% (+0.3%) Not prioritized $12,048,458 Vihiga 99.4% (+7.1%) Not prioritized Not prioritized Not prioritized 0.5% (+0.3%) 0.004% (-0.032%) $8,876,144 Wajir 99.5% (+3.9%) Not prioritized Not prioritized Not prioritized 0.4% (+0.3%) 0.01% (+0.01%) $2,129,667 West Pokot 98.7% (+9.1%) Not prioritized Not prioritized 0.7% (+0.01%) 0.6% (+0.4%) Not prioritized $3,681,552 National 96.4% (+6.4%) 2.6% (-5.0%) 0.1% (-0.01%) 0.6% (-0.8%) 0.4% (+0.2%) 0.01% (+0.01%) $598,651,401 Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, SBCC = social behavior change communication; VMMC = voluntary male medical circumcision. Table A6.8 100% HIV budget optimization across counties, 2019 to 2030 HIV testing HIV prevention services services Care and (biomedical (condoms and FSW MSM PWID County Scenario treatment only) SBCC) VMMC programs programs programs Total budget Baringo Latest reported $3,250,922 $487,240 $41,345 $0 $2,170 $0 $0 $3,781,677 Optimized $6,260,411 $0 $0 $262,581 $8,383 $0 $45 $6,531,420 Bomet Latest reported $3,795,302 $732,895 $138,331 $43,642 $3,173 $0 $0 $4,713,343 Optimized $6,069,575 $440,809 $0 $178,980 $11,791 $0 $0 $6,701,155 Bungoma Latest reported $5,107,718 $1,293,019 $80,951 $0 $38,316 $900 $755 $6,521,659 Optimized $5,989,094 $725,922 $0 $225,324 $83,608 $511 $2,148 $7,026,607 Busia Latest reported $9,664,843 $516,681 $142,195 $456,395 $28,791 $676 $567 $10,810,148 Optimized $11,988,897 $400,578 $0 $230,156 $60,613 $118 $1,818 $12,682,180 Elgeyo Latest reported $4,342,077 $353,945 $47,141 $0 $1,507 $0 $0 $4,744,670 Marakwet Optimized $2,631 $0 $0 $224,341 $8,426 $0 $67 $235,465 Embu Latest reported $9,183,850 $541,821 $127,319 $0 $5,192 $61 $234 $9,858,477 Optimized $59,314 $0 $0 $251,678 $26,386 $0 $1,560 $338,938 Garissa Latest reported $3,746,225 $129,987 $17,968 $0 $6,996 $600 $414 $3,902,190 Optimized $4,155 $0 $0 $11,425 $21,673 $0 $779 $38,032 Homa Bay Latest reported $34,111,042 $1,925,676 $192,814 $1,513,878 $5,368 $1,952 $151 $37,750,881 Optimized $67,200,229 $1,603,507 $1,256,646 $623,501 $30,736 $0 $1,451 $70,716,070 Isiolo Latest reported $2,264,373 $78,072 $17,388 $102 $1,962 $42 $20 $2,361,959 Optimized $30,053 $0 $0 $77,291 $10,405 $0 $145 $117,894 Kajiado Latest reported $10,600,549 $849,895 $22,411 $407 $24,029 $62 $25 $11,497,378 Optimized $147,327 $0 $0 $415,141 $58,239 $0 $143 $620,850 Kakamega Latest reported $19,811,955 $1,856,134 $187,211 $0 $11,470 $1,220 $1,024 $21,869,014 Optimized $17,402,318 $470,869 $0 $423,987 $42,120 $0 $226 $18,339,520 Kericho Latest reported $6,551,128 $1,132,027 $107,612 $226,960 $5,948 $178 $60 $8,023,913 Optimized $8,186,760 $7,250 $0 $221,183 $25,395 $0 $313 $8,440,901 APPENDICES Table continued… 61 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 62 Table A6.8 100% HIV budget optimization across counties, 2019 to 2030 (continued) HIV testing HIV prevention services services Care and (biomedical (condoms and FSW MSM PWID County Scenario treatment only) SBCC) VMMC programs programs programs Total budget Kiambu Latest reported $3,213,795 $1,831,363 $221,987 $0 $51,954 $1,220 $1,024 $5,321,343 Optimized $5,201,407 $713,448 $0 $267,961 $84,883 $1,121 $6,113 $6,274,933 Kilifi Latest reported $13,772,970 $999,980 $97,744 $1,356 $30,956 $727 $610 $14,904,343 Optimized $0 $1,115,462 $215,903 $212,916 $75,722 $2,459 $0 $1,622,462 Kirinyaga Latest reported $6,573,568 $740,917 $67,620 $0 $3,933 $0 $422 $7,386,460 Optimized $6,968,951 $0 $0 $153,899 $15,208 $0 $0 $7,138,058 Kisii Latest reported $19,800,751 $1,704,517 $113,022 $0 $45,026 $1,058 $887 $21,665,261 Optimized $18,277,744 $0 $0 $444,887 $98,288 $0 $164 $18,821,083 Kisumu Latest reported $42,147,243 $2,337,739 $269,869 $1,797,027 $21,665 $2,446 $2,052 $46,578,041 Optimized $50,347,599 $422,994 $0 $593,490 $83,225 $1,815 $7,411 $51,456,534 Kitui Latest reported $12,767,821 $1,353,235 $2,941 $0 $4,014 $951 $0 $14,128,962 Optimized $13,532,829 $57,447 $0 $389,895 $19,319 $529 $0 $14,000,019 Kwale Latest reported $6,380,058 $363,766 $71,870 $610 $19,050 $447 $375 $6,836,176 Optimized $8,713,512 $0 $0 $327,288 $40,739 $0 $655 $9,082,194 Laikipia Latest reported $3,697,176 $494,462 $58,346 $610 $19,483 $67 $384 $4,270,528 Optimized $4,768,627 $0 $0 $86,905 $32,668 $0 $623 $4,888,823 Lamu Latest reported $1,694,873 $79,606 $17,388 $0 $2,372 $178 $149 $1,794,566 Optimized $12,447 $0 $0 $59,747 $11,909 $0 $0 $84,103 Machakos Latest reported $3,020,338 $879,099 $146,446 $0 $37,450 $1,307 $429 $4,085,069 Optimized $3,226,608 $366,641 $0 $0 $53,922 $0 $0 $3,647,171 Makueni Latest reported $13,301,166 $966,526 $107,469 $0 $27,059 $98 $0 $14,402,318 Optimized $13,240,511 $32,277 $0 $320,646 $47,096 $113 $0 $13,640,643 Mandera Latest reported $1,941,577 $83,330 $25,502 $0 $7,577 $0 $0 $2,057,986 Optimized $0 $0 $0 $9,346 $14,158 $0 $54 $23,558 Table continued… Table A6.8 100% HIV budget optimization across counties, 2019 to 2030 (continued) HIV HIV testing prevention services services Care and (biomedical (condoms and FSW MSM PWID County Scenario treatment only) SBCC) VMMC programs programs programs Total budget Marsabit Latest reported $1,908,023 $74,367 $15,649 $237 $869 $0 $0 $1,999,145 Optimized $33,091 $0 $0 $17,391 $3,133 $0 $0 $53,615 Meru Latest reported $12,639,113 $1,095,231 $93,887 $0 $40,048 $941 $789 $13,870,009 Optimized $13,496,324 $0 $0 $359,047 $72,033 $0 $0 $13,927,404 Migori Latest reported $26,994,656 $1,345,552 $159,776 $973,624 $12,265 $1,413 $1,186 $29,488,472 Optimized $28,536,795 $817,892 $0 $419,498 $40,627 $367 $5,282 $29,820,461 Mombasa Latest reported $15,023,512 $729,957 $347,976 $21,092 $25,760 $605 $508 $16,149,410 Optimized $15,912,406 $40,586 $0 $156,690 $145,970 $1,212 $19,847 $16,276,711 Muranga Latest reported $3,550,369 $507,725 $67,234 $0 $2,212 $412 $346 $4,128,298 Optimized $5,717,609 $341,867 $0 $316,044 $10,500 $0 $0 $6,386,020 Nairobi Latest reported $76,142,281 $5,358,997 $1,546,306 $516,992 $162,572 $3,818 $3,203 $83,734,169 Optimized $68,133,947 $0 $0 $0 $433,782 $0 $8,062 $68,575,791 Nakuru Latest reported $24,106,103 $2,905,864 $328,425 $238,625 $1,078 $2,471 $288 $27,582,854 Optimized $31,460,454 $0 $0 $657,450 $101,094 $0 $0 $32,218,998 Nandi Latest reported $3,952,806 $901,914 $52,917 $173,144 $3,084 $0 $0 $5,083,865 Optimized $4,952,168 $58,422 $0 $263,181 $13,639 $0 $0 $5,287,410 Narok Latest reported $6,478,647 $867,204 $51,264 $2,340 $3,284 $0 $0 $7,402,739 Optimized $17,123,972 $699,468 $0 $498,546 $17,533 $0 $0 $18,339,519 Nyamira Latest reported $9,043,146 $1,404,761 $50,234 $0 $4,586 $824 $0 $10,503,551 Optimized $10,756,907 $0 $0 $237,982 $22,892 $0 $0 $11,017,781 Nyandarua Latest reported $5,184,140 $584,987 $37,094 $0 $4,967 $0 $486 $5,811,674 Optimized $6,744,659 $500,778 $0 $166,890 $13,935 $0 $0 $7,426,262 Table continued… APPENDICES 63 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 64 Table A6.8 100% HIV budget optimization across counties, 2019 to 2030 (continued) HIV HIV testing prevention services services Care and (biomedical (condoms and FSW MSM PWID County Scenario treatment only) SBCC) VMMC programs programs programs Total budget Nyeri Latest reported $7,976,176 $1,011,784 $117,079 $0 $5,202 $0 $580 $9,110,821 Optimized $8,713,280 $58,117 $0 $151,899 $17,551 $0 $0 $8,940,847 Samburu Latest reported $1,836,898 $223,412 $19,320 $0 $847 $0 $0 $2,080,477 Optimized $118,154 $0 $0 $133,721 $4,831 $0 $0 $256,706 Siaya Latest reported $44,934,550 $1,154,392 $120,750 $1,735,955 $11,570 $1,637 $123 $47,958,977 Optimized $65,598,685 $681,641 $0 $496,058 $63,909 $0 $1,129 $66,841,422 Taita Taveta Latest reported $5,363,680 $221,761 $21,059 $0 $11,906 $0 $0 $5,618,406 Optimized $185,968 $0 $0 $97,183 $38,564 $0 $0 $321,715 Tana River Latest reported $1,096,139 $216,795 $123,648 $0 $903 $0 $0 $1,437,485 Optimized $0 $0 $0 $44,852 $3,789 $0 $0 $48,641 Tharaka Latest reported $19,257,062 $416,611 $26,275 $0 $5,219 $0 $341 $19,705,508 Nithi Optimized $0 $0 $0 $167,132 $26,158 $0 $544 $193,834 Trans Nzoia Latest reported $5,783,377 $873,327 $121,890 $0 $4,630 $0 $0 $6,783,224 Optimized $10,908,768 $1,482,172 $0 $465,405 $25,449 $0 $0 $12,881,794 Turkana Latest reported $2,606,202 $1,091,818 $109,544 $375,994 $14,584 $1,083 $909 $4,200,134 Optimized $6,478,517 $179,826 $0 $552,965 $74,950 $0 $0 $7,286,258 Uasin Gishu Latest reported $10,427,658 $1,382,875 $223,814 $0 $13,818 $222 $71 $12,048,458 Optimized $13,715,164 $507,233 $0 $464,015 $60,670 $0 $0 $14,747,082 Vihiga Latest reported $8,193,543 $633,159 $27,628 $0 $18,184 $427 $3,203 $8,876,144 Optimized $9,936,208 $0 $0 $200,031 $51,630 $0 $486 $10,188,355 Table continued… Table A6.8 100% HIV budget optimization across counties, 2019 to 2030 (continued) HIV HIV testing prevention services services Care and (biomedical (condoms and FSW MSM PWID County Scenario treatment only) SBCC) VMMC programs programs programs Total budget Wajir Latest reported $2,036,212 $77,272 $14,490 $0 $1,693 $0 $0 $2,129,667 Optimized $0 $0 $0 $7,135 $7,359 $0 $32 $14,526 West Pokot Latest reported $3,298,894 $336,291 $14,490 $26,145 $5,732 $0 $0 $3,681,552 Optimized $4,895,633 $0 $0 $214,959 $21,054 $0 $0 $5,131,646 Total, Latest reported $538,574,507 $45,147,988 $6,013,639 $8,105,135 $760,474 $28,043 $21,615 $598,651,401 national Optimized $571,049,708 $11,725,206 $1,472,549 $12,100,642 $2,235,964 $8,245 $59,097 $598,651,411 % of total latest 90.0% 7.5% 1.0% 1.4% 0.1% 0.005% 0.004% 100.0% reported budget % of total 95.4% 2.0% 0.2% 2.0% 0.4% 0.001% 0.010% 100.0% optimized budget Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, SBCC = social behavior change communication; VMMC = voluntary male medical circumcision. APPENDICES 65 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 66 Table A6.9 Percentage of total annual HIV budget to optimally reallocate across counties, 2019 to 2030 Percentage of total HIV budget recommended to allocate to each program (percentage difference of total budget between the optimized and latest reported allocation) by county Optimized % HIV prevention HIV testing and HIV testing and budget increase Care and HIV testing services (condoms prevention for HIV testing and prevention for County or decrease treatment (biomedical) and SBCC) VMMC FSW prevention for MSM PWID Baringo 73% 95.8% (+9.8%) Not prioritized Not prioritized 4.0% (+4.0%) 0.1% (+0.1%) 0% (+0%) 0.0007% (+0.0007%) Bomet 42% 90.5% (+10%) 6.5% (-8.9%) 6.5% (-8.9%) 2.6% (+1.7%) 0.2% (+0.1%) 0% (+0%) 0% (+0%) Bungoma 8% 85.2% (+6.9%) 10.3% (-9.4%) 10.3% (-9.4%) 3.2% (+3.2%) 1.2% (+0.6%) 0.007% (-0.007%) 0.03% (+0.02%) Busia 17% 94.5% (+5.1%) 3.1% (-1.6%) 3.1% (-1.6%) 1.8% (-2.4%) 0.5% (+0.2%) 0.001% (-0.005%) 0.01% (+0.01%) Elgeyo -95% 1.1% (-90.3%) Not prioritized Not prioritized 95.2% (+95.2%) 3.6% (+3.5%) 0% (+0%) 0.03% (+0.03%) Marakwet Embu -97% 17.4% (-75.6%) Not prioritized Not prioritized 74.2% (+74.2%) 7.8% (+7.7%) 0% (-0.0006%) 0.5% (+0.5%) Garissa -99% 10.9% (-85.0%) Not prioritized Not prioritized 30.0% (+30.0%) 57.0% (+56.8%) 0% (-0.02%) 2.0% (+2.0%) Homa Bay 87% 95.0% (+4.6%) 2.2% (-2.8%) 2.2% (-2.8%) 0.8% (-3.1%) 0% (+0%) 0% (-0.005%) 0.002% (-0.003%) Isiolo -95% 25.4% (-70.3%) Not prioritized Not prioritized 65.5% (+65.5%) 8.8% (+8.7%) 0% (-0.002%) 0.1% (+0.1%) Kajiado -95% 23.7% (-68.4%) Not prioritized Not prioritized 66.8% (+66.8%) 9.4% (+9.2%) 0% (-0.001%) 0.02% (+0.02%) Kakamega -16% 94.8% (+4.2%) 2.5% (-5.9%) 2.5% (-5.9%) 2.3% (+2.3%) 0.2% (+0.2%) 0% (-0.01%) 0.001% (-0.004%) Kericho 5% 96.9% (+15.3%) 0% (-14%) Not prioritized 2.6% (-0.2%) 0.3% (+0.2%) 0% (-0.002%) 0.004% (+0.002%) Kiambu 18% 82.8% (+22.4%) 11.3% (-23%) 11.3% (-23%) 4.2% (+4.2%) 1.4% (+0.4%) 0.02% (-0.01%) 0.1% (+0.1%) Kilifi -89% 0% (-92.4%) 68.7% (+62%) 68.7% (+62%) 13.1% (+13.1%) 4.7% (+4.5%) 0.2% (+0.1%) 0% (-0.005%) Kirinyaga -3% 97.6% (+8.6%) Not prioritized Not prioritized 2.1% (+2.1%) 0.2% (+0.2%) 0% (+0%) 0% (+0%) Kisii -13% 97.1% (+5.7%) Not prioritized Not prioritized 2.3% (+2.3%) 0.5% (+0.3%) 0% (-0.005%) 0.001% (-0.004%) Kisumu 10% 97.8% (+7.3%) 0.8% (-4.1%) 0.8% (-4.1%) 1.1% (-2.7%) 0.2% (+0.1%) 0.004% (-0.002%) 0.01% (+0.01%) Kitui -1% 96.6% (+6.2%) 0.4% (-9.1%) 0.4% (-9.1%) 2.7% (+2.7%) 0.1% (+0.1%) 0.004% (-0.003%) 0% (-0.007%) Kwale 33% 95.9% (+2.6%) Not prioritized Not prioritized 3.6% (+3.5%) 0.4% (+0.2%) 0% (-0.01%) 0.007% (+0.001%) Laikipia 14% 97.5% (+10.9%) Not prioritized Not prioritized 1.7% (+1.7%) 0.7% (+0.2%) 0% (-0.002%) 0.01% (+0.01%) Lamu -95% 14.7% (-79.6%) Not prioritized Not prioritized 71.0% (+71.0%) 14.2% (+14.0%) 0% (-0.01%) 0% (-0.01%) Machakos -11% 88.4% (+14.5%) 10% (-11.4%) 10.0% (-11.4%) Not prioritized 1.5% (+0.6%) 0% (-0.03%) 0% (-0.03%) Makueni -5% 97.0% (+4.7%) 0.2% (-6.4%) 0.2% (-6.4%) 2.3% (+2.3%) 0.3% (+0.2%) 0.0008% (+0.0001%) 0% (-0.0007%) Table continued… TableA6.9 Percentage of total annual HIV budget to optimally reallocate across counties, 2019 to 2030 Percentage of total HIV budget recommended to allocate to each program (percentage difference of total budget between the optimized and latest reported allocation) by county Optimized % HIV prevention HIV testing and HIV testing and budget increase Care and HIV testing services (condoms prevention for HIV testing and prevention for County or decrease treatment (biomedical) and SBCC) VMMC FSW prevention for MSM PWID Mandera -99% 0% (-94.3%) Not prioritized Not prioritized 39.6% (+39.6%) 60.1% (+59.7%) 0% (+0%) 0.2% (+0.2%) Marsabit -97% 61.7% (-33.7%) Not prioritized Not prioritized 32.4% (+32.4%) 5.8% (+5.8%) 0% (+0%) 0% (+0%) Meru 0.4% 96.9% (+5.7%) Not prioritized Not prioritized 2.5% (+2.5%) 0.5% (+0.2%) 0% (-0.007%) 0% (-0.007%) Migori 1% 95.6% (+4.1%) 2.7% (-1.8%) 2.7% (-1.8%) 1.4% (-1.8%) 0.1% (+0.1%) 0.001% (-0.004%) 0.02% (+0.01%) Mombasa 1% 97.7% (+4.7%) 0.2% (-4.2%) 0.2% (-4.2%) 0.9% (+0.8%) 0.9% (+0.7%) 0.007% (+0.004%) 0.1% (+0.1%) Muranga 55% 89.5% (+3.5%) 5.3% (-6.9%) 5.3% (-6.9%) 4.9% (+4.9%) 0.2% (+0.1%) 0% (-0.01%) 0% (-0.01%) Nairobi -18% 99.3% (+8.4%) Not prioritized Not prioritized Not prioritized 0.6% (+0.4%) 0% (-0.005%) 0.01% (+0.01%) Nakuru 17% 97.6% (+10.2%) Not prioritized Not prioritized 2% (+1.1%) 0.3% (+0.3%) 0% (-0.01%) 0% (-0.009%) Nandi 4% 93.6% (+15.9%) 1.1% (-16.6%) 1.1% (-16.6%) 4.9% (+1.5%) 0.3% (+0.2%) 0% (+0%) 0% (+0%) Narok 148% 93.3% (+5.8%) 3.8% (-7.9%) 3.8% (-7.9%) 2.7% (+2.6%) 0.1% (+0.1%) 0% (+0%) 0% (+0%) Nyamira 5% 97.6% (+11.5%) Not prioritized Not prioritized 2.1% (+2.1%) 0.2% (+0.2%) 0% (-0.008%) 0% (-0.008%) Nyandarua 28% 90.8% (+1.6%) 6.7% (-3.3%) 6.7% (-3.3%) 2.2% (+2.2%) 0.2% (+0.1%) 0% (+0%) 0% (+0%) Nyeri -2% 97.4% (+9.9%) 0.6% (-10.4%) 0.6% (-10.4%) 1.6% (+1.6%) 0.2% (+0.1%) 0% (+0%) 0% (+0%) Samburu -88% 46% (-42.2%) Not prioritized Not prioritized 52.0% (+52.0%) 1.9% (+1.8%) 0% (+0%) 0% (+0%) Siaya 39% 98.1% (+4.4%) 1% (-1.3%) 1.0% (-1.3%) 0.7% (-2.8%) 0.1% (+0.1%) 0% (-0.003%) 0.002% (-0.002%) Taita Taveta -94% 57.8% (-37.6%) Not prioritized Not prioritized 30.2% (+30.2%) 12.0% (+11.8%) 0% (+0%) 0% (+0%) Tana River -97% 0% (-76.2%) Not prioritized Not prioritized 92.2% (+92.2%) 7.8% (+7.7%) 0% (+0%) 0% (+0%) Tharaka -99% 0% (-97.7%) Not prioritized Not prioritized 86.2% (+86.2%) 13.5% (+13.5%) 0% (+0%) 0.3% (+0.3%) Nithi Trans Nzoia 90% 84.6% (-0.5%) 11.5% (-1.3%) 11.5% (-1.3%) 3.6% (+3.6%) 0.2% (+0.1%) 0% (+0%) 0% (+0%) Turkana 73% 88.9% (+26.8%) 2.4% (-23.5%) 2.4% (-23.5%) 7.5% (-1.3%) 1.0% (+0.7%) 0% (-0.03%) 0% (-0.03%) Uasin Gishu 22% 93.0% (+6.4%) 3.4% (-8%) 3.4% (-8.0%) 3.1% (+3.1%) 0.4% (+0.3%) 0% (-0.002%) 0% (-0.002%) Vihiga 15% 97.5% (+5.2%) Not prioritized Not prioritized 1.9% (+1.9%) 0.5% (+0.3%) 0% (-0.005%) 0.005% (+0%) Wajir -99% 0% (-95.6%) Not prioritized Not prioritized 49.1% (+49.1%) 50.7% (+50.6%) 0% (+0%) 0.2% (+0.2%) West Pokot 39% 95.4% (+5.7%) Not prioritized Not prioritized 4.1% (+3.4%) 0.4% (+0.3%) 0% (+0%) 0% (+0%) Source: Authors from Optima data. Note: FSW = female sex worker; MSM = men who have sex with men; PWID = people who inject drugs, SBCC = social behavior change communication; VMMC = voluntary male medical APPENDICES circumcision. 67 MODELLING ANALYSIS TO SUPPORT THE NEXT ITERATION OF THE KENYA AIDS STRATEGIC FRAMEWORK 68 Table A6.10 100% annual HIV budget optimization within counties, represented nationally 2019‒25 Baseline Optimized Baseline Optimized HIV programs 100% budget 100% budget % total budget % total budget Care and treatment $538,574,507 $576,551,518 90.0% 96.3% HIV testing services (HTS) (biomedical only) $45,147,988 $17,401,809 7.5% 2.9% HIV prevention services (condoms and SBCC) $6,013,639 $296,755 1.00% 0.05% VMMC $8,105,135 $2,035,885 1.4% 0.3% HIV prevention for FSW $760,474 $2,313,651 0.1% 0.4% HIV prevention for MSM $28,043 $2,456 0.0047% 0.0004% HIV prevention for PWID $21,615 $49,327 0.004% 0.008% Total $598,651,401 $598,651,401 100% 100% 69