Kenya: 2020 Supply Chain Modeling . Forecasting Demand from 2020-2024

Publication date: 2010

Kenya: 2020 Supply Chain Modeling Forecasting Demand from 2020–2024 DECEMBER 2010 This publication was produced for review by the U.S. Agency for International Development. It was prepared by the USAID | DELIVER PROJECT, Task Order 1. Kenya: 2020 Supply Chain Modeling Forecasting Demand Over 2020–2024 The authors' views expressed in this publication do not necessarily reflect the views of the U.S. Agency for International Development or the United States Government. USAID | DELIVER PROJECT, Task Order 1 The USAID | DELIVER PROJECT, Task Order 1, is funded by the U.S. Agency for International Development under contract no. GPO-I-01-06-00007-00, beginning September 29, 2006. Task Order 1 is implemented by John Snow, Inc., in collaboration with PATH; Crown Agents Consultancy, Inc.; Abt Associates; Fuel Logistics Group (Pty) Ltd.; UPS Supply Chain Solutions; The Manoff Group; and 3i Infotech. The project improves essential health commodity supply chains by strengthening logistics management information systems, streamlining distribution systems, identifying financial resources for procurement and supply chain operations, and enhancing forecasting and procurement planning. The project also encourages policymakers and donors to support logistics as a critical factor in the overall success of their health care mandates. Recommended Citation Tuin, Astrid, Kelly L. Ganamet, and Donald A. Hicks. 2010. Kenya: 2020 Supply Chain Modeling. Forecasting Demand from 2020-2024. Arlington, Va.: USAID | DELIVER PROJECT, Task Order 1. Abstract In 2010, LLamasoft, with technical assistance from the USAID | DELIVER PROJECT, Task Order 1, developed a modeling framework to forecast public health supply chain needs and enable policymakers to strengthen the logistics situation. Here, the model was applied to understand and analyze the current and future state (2020–2024) supply chain requirements for procuring and distributing essential medical commodities in Kenya. The developed methodology in this report can be employed in any country for any future time frame. Cover photo: Current network structure in Kenya. USAID | DELIVER PROJECT John Snow, Inc. 1616 Fort Myer Drive, 11th Floor Arlington, VA 22209 USA Phone: 703-528-7474 Fax: 703-528-7480 Email: askdeliver@jsi.com Internet: deliver.jsi.com                                                         Contents Acronyms. v Acknowledgments . vii Executive Summary . ix Background. 1 Overview . 1 Project Objectives . 1 Modeling Framework. 1 Future State Scenario . 3 Methodology . 5 Overview . 5 Data Collection . 5 Baseline and Future Projection Assumptions. 6 Quantification Analysis . 8 Key Findings . 9 Baseline Analysis . 9 Future State Analysis. 10 Discussion . 23 Conclusion. 25 Appendices A. Kenya Population Data 2010 & 2020. 27 B. Prevalence Rate Sources and Description . 31 C. Kenya Prevalence Rates . 33 D. Number of HIV, Tuberculosis, and Malaria Cases in 2020 . 35 E. Resources Used for Material Requirements Model. 39 F. Material Requirements Model: Treatments . 41 G. Assumptions Material Requirements Model. 45 H. Treatment Rates by Condition . 47 I. Vehicle Characteristics KEMSA Owned Fleet .49 J. Outputs Generated by Models and Scenarios . 51 iii                                       Figures 1. Visual Overview of the Interdependent Model Framework, Which Utilizes a Country’s “Health State” to Determine Supply Chain Recommendations by Taking into Account Observed or Expected Trends . 3 2. Current Network Structure in Kenya . 9 3. Supply Chain Metrics for 2010 and 2020–2024. 11 4. Supply Chain Metrics for 2010 and 2020–2024. 12 5. Supply Chain Metrics for 2010 and 2020–2024. 13 6. Supply Chain Metrics for 2010 and 2020–2024. 14 7. Results of Greenfield Analysis . 15 8. Supply Chain Financials for Baseline and Adding Regional Warehouse Scenarios .16 9. Supply Chain Service Metrics for Baseline and Adding Regional Warehouse Scenarios .17 10. Supply Chain Metrics . 18 11. Future State Supply Chain Network . 19 12. Results of Greenfield Analysis . 20 13. Supply Chain Financials Comparing Multiple Scenarios: Expansion of Central Warehouse (3PL and the Central Warehouse) versus Assessing Multiple Variables (KEMSA-owned Fleet, Regional Warehouses, and Shipment Schedule) . 21 Tables 1. Baseline Supply Chain Cost Comparison . 10 2. Baseline Supply Chain Metrics . 10 3. Commodities Resulting in an Unmet Need of 75 Percent or Greater .12 4. Comparison of Transportation Cost Components (U.S. Dollars) .18 5. Supply Chain and Service Metrics . 21 6. Summary of Total Supply Chain and Procurement Budgets (U.S. Dollars) .24 iv Acronyms 3PL third-party logistics provider KEMSA Kenya Medical Supplies Agency MDG Millennium Development Goals PHeNOM Public Health Network Optimization and Modeling tool WHO World Health Organization v vi Acknowledgments The team would like to thank the staff at John Snow, Inc., for all their guidance and support. This study would not have been possible without them all. The team would also like to thank the staff of the Kenya Medical Supplies Agency for providing us with much of the data used in this project. The U.S. Agency for International Development contracts funded the technical assistance, in-country projects, and research that enabled the authors to produce the experience and lessons learned here. We are deeply grateful to the team of professionals in the Commodity Security and Logistics Division in the Office of Population, Health, and Nutrition, and the Office of HIV and AIDS who provided the support for the production of this publication. vii viii Executive Summary In the developing world, logistical shortcomings can have a serious impact on the quality of human health care. As such, the USAID | DELIVER PROJECT has partnered with LLamasoft, Inc., to develop a reusable modeling framework to forecast developing countries’ public health supply chain needs for future time periods. The model was designed to be robust and general so that it can be applied to any country for any future time frame and provide policymakers with key data to guide effective design of their supply chain networks. Here, the framework was applied to Kenya for the years 2020 through 2024 with the goal of demonstrating how the modeling approach can be utilized to help policymakers accurately visualize and understand the most likely and possible situations facing them in 10 years. The project objectives were accomplished by modeling the relationships between key public health variables including population, disease prevalence, and economic conditions, along with the resulting health supply material requirements. The modeling framework consisted of three separate but interlinked models: 1) a health model for predicting the location and quantity of treatable health conditions of interest in the future, 2) a material requirements model for translating the project health conditions into delivery needs for the supply system, and 3) a supply chain model that generates metrics of interest by modeling the pull and flow of generated material requirements through a defined supply chain network. Together, the interlinked framework, which represents the three models of the system (health, material requirements, and supply chain), and a data gathering tool allow for the model to be built and applied quickly and flexibly for other scenarios, time frames, and countries. For example, a change made in the material requirements model simulating a potential future state scenario automatically cascades through to the supply chain model and is taken into account in the supply chain optimization. The data gathering template guides a modeler to extract and use the appropriate data. Substantial legwork for populating the health, material requirements, and supply chain models with accurate data has been completed as part of this project; this will reduce the time investment needed to generate these models for future projects that use the same framework. This study found that if Kenya’s network structure does not change to accommodate the growing population, the proportion of need met by the Kenya Medical Supplies Agency will decrease from 35 percent in 2010 to 28.5 percent in 2024. Multiple future state scenarios were assessed to determine in what ways the network could be restructured in a cost-effective manner to increase service levels, and the study found that the most cost-effective implementation would be to increase the number of replenishments each year. The main conclusion is that if nothing is done to address the increased demand, the majority of the public sector will not be serviced, resulting in loss of lives. As such, it is essential that stakeholders understand the importance of investing in the supply chain network and the timing of delivery. ix x Background Overview Public health supply chains deliver essential medical commodities to underserved communities in the developing world. As such, logistical shortcomings can have a serious impact on the quality of human health care. In some situations, it can literally mean the difference between life and death. By strengthening the existing supply chain systems, the availability of essential commodities to health care providers and consumers is greatly increased, resulting in improved health for communities in underserved areas. Today, with increased health needs due to the world’s growing population and changing disease burdens, it is imperative that public health systems are optimized to ensure cost- effective and reliable supply chains to meet those demands. The USAID | DELIVER PROJECT, in collaboration with LLamasoft, Inc., has undertaken the 2020 Supply Chain Modeling project as a means to develop a reusable framework to forecast developing countries’ public health supply chain needs for future time periods. Here, the framework has been applied to Kenya for the years 2020 through 2024 to enable policymakers to strengthen the logistics situation. The 2020 model is designed so that it can be applied quickly to any country for any future time frame, helping policymakers to accurately visualize and understand the most likely and possible situations facing them, and to make informed decisions about how to design effective supply chains to meet those demands. Project Objectives The goal of this project is to predict future supply chain needs and performance metrics over a five- year period (2020–2024) in order to inform Kenyan policymakers and improve their long-range strategic planning processes. The three main objectives are— 1. To develop a general and reusable methodology for creating three interlinked models: a health model, a material requirements model, and a supply chain model. 2. To apply the health, material requirements, and supply chain models to understand and analyze the current and future state (2020–2024) supply chain requirements for procuring and distributing essential medical commodities in Kenya. 3. To determine the most robust supply chain network in Kenya by applying multiple future state scenarios to the modeling framework. Modeling Framework The approach for this study employed a modeling framework that consisted of three separate but interlinked models: 1) a health model for predicting the location and quantity of treatable health conditions of interest in the future, 2) a material requirements model for translating the health conditions into delivery needs for the supply system, and 3) a supply chain model that generates metrics of interest by modeling the pull and flow of generated material requirements through a 1 defined supply chain network. Forecasting of essential health commodities was accomplished by analyzing the relationships between the three interdependent models. Health Model In line with the United Nation’s Millennium Development Goals (MDGs) 4, 5, and 6 (child health, maternal health, and combat HIV and AIDS, respectively), the emphasis for the conditions included in the health model was on reducing child mortality, improving maternal health, and combating HIV and AIDS, malaria, and other serious diseases that have the greatest long-term detrimental effects on lifetime human potential contribution. Conditions modeled for reducing child mortality included measles, vitamin A deficiency, diarrhea, worms, and respiratory infections. Sexually transmitted infection prevalence, maternal mortality, and pregnancy and birth rates were modeled in accordance with MDG 5 to improve maternal health. HIV and AIDS, malaria, tuberculosis, and leprosy were included for MDG 6 as these are very serious diseases that can result in severe illness or death. In addition to conditions within the MDGs, selected lifestyle diseases, including cardiovascular diseases, hypertension, diabetes, and asthma, were included in the health model. It is expected that their prevalence rates will significantly increase in developing countries over the coming years. Furthermore, the World Health Organization (WHO) Global Burden of Disease work was reviewed, and the Pareto principle, also known as the 80-20 rule, was applied to this data. Greater emphasis was placed on keeping diseases in the model that represent the 20 percent of diseases that cause 80 percent of disability-adjusted life years and mortalities for the MDG region of sub-Saharan Africa. Prevalence rates were obtained for each of the conditions described previously and broken down by age group when possible. Additionally, geocoded population data for Kenya was acquired at the district level. As the main components of the health model, the prevalence rates and population data were used to derive the number of people who need treatment for these conditions and diseases, and how the treatment should be distributed across Kenya. Material Requirements Model The materials for the material requirements model focus on key pharmaceuticals needed to treat a person who suffers from one of the in-scope conditions. In addition to medicines required to treat conditions, diagnostic and preventive commodities like HIV and malaria test kits, bed nets, vaccine packages, family planning commodities, and mother-child health-related items were included. Another component of the material requirements model is treatment rates for each disease; not everyone who suffers from a disease receives treatment, which needs to be taken into account when determining the overall material needs for a country. An additional bundle of essential medicines was also determined to be required to cover other disease and conditions not identified specifically. For each included health commodity, the price, quantity, weight, and volume were used as characteristic definitions. Supply Chain Model The supply chain model was defined by the following parameters: procurement and supply chain financials (warehouse operating, administrative, labor, and transportation), supply chain configuration (warehouse location, available warehouse space and capacity, and transportation capacity), and service requirements (percentage of need satisfied, and service distance and hours). The network was modeled at an aggregated district level based on demand-driven flows of the 2 material needs that are determined through the material requirements model. Overall supply chain budget can be modeled both as a constraint and as a variable to be tested. With this approach, observed trends for any of the three models of the framework can be easily incorporated to create different scenarios and compare the overall effects on the supply chain model outputs. The framework is illustrated in figure 1. Figure 1. Visual Overview of the Interdependent Model Framework, Which Utilizes a Country’s “Health State” to Determine Supply Chain Recommendations by Taking into Account Observed or Expected Trends Future State Scenario In this study, two main questions were considered for future state analyses. First, if the network structure does not change from its current operating state, how will this affect the service levels in 2020–2024? Additionally, what budget will be necessary for procurement and supply chain operations? Second, if the network was to change to meet the needs of the Kenyan population, what are ways to restructure the network in a cost-effective manner that increases service levels? Multiple options were considered, and those researched included 1) increased replenishment rate, 2) expansion of the central warehouse, 3) additional tier in the supply chain (i.e., regional warehouses), 4) Kenya Medical Supplies Agency (KEMSA)-owned fleet rather than using third-party logistics providers (3PLs), and 5) assessing multiple variables simultaneously. 3 4 Methodology Overview The project team first collected the required data from available sources to use as inputs for the health, material requirements, and supply chain models, which together makeup the modeling framework. The team then determined the necessary assumptions for both the baseline and future state scenarios. Next, the Public Health Network Optimization and Modeling (PHeNOM) tool, a rapid supply chain optimization and analysis tool developed by LLamasoft, Inc., was employed to determine the key supply chain, service, and financial metrics for the baseline and future state supply chain scenarios. Finally, the results were analyzed and summarized. Data Collection The following data was collected for Kenya to be used as inputs for the health model, material requirements model, and supply chain model. Health Model Census data for 1999 and projections for 2000 to 2010 available from the Kenya National Bureau of Statistics were used to predict the expected Kenyan population for the years 2020–2024. The census data is available down to the district level (Kenya had 69 districts in 1999) and is split by gender and age. The population numbers for 2010 and 2020 by district can be found in Appendix A. The prevalence rates for the in-scope conditions were taken from sources such as the Kenyan Ministry of Public Health and Sanitation, WHO and U.S. President’s Emergency Plan for AIDS Relief reports, peer-reviewed publications, and conference abstracts. An overview of the resources used and a description of what the rate/number represents for each condition are listed in Appendix B; the actual numbers for the prevalence rates by region can be found in Appendix C. The number of cases for a certain condition was then calculated from the prevalence rates and projected population data. Appendix D is an example of these results, listing the number of cases by district in 2020 for HIV, tuberculosis, and malaria. Material Requirements Model The material requirements model is a compilation of all the information needed to derive a country’s health supply needs from its health model. This entails gathering data on what treatment a person with an in-scope condition should receive, as required by age and gender. Appendices E, F, G, and H summarize the material requirements model. Appendix E contains the resources used for the model, Appendix F lists the treatments for each condition, the assumptions made for the material requirements are gathered by condition in Appendix G, and Appendix H lists the assumed treatment rates by condition for 2010 and 2020–2024. 5 The total material needs of a country were then calculated based on the health model and the treatments from the material requirements model. The treatment rates were applied to predict the total forecasted demand. Supply Chain Model The Kenya public health supply chain is operated by KEMSA. From previous LLamasoft in-country work in Kenya and KEMSA data, it is known that there is currently one central warehouse where supplies are imported, which has been at capacity since 2009. In addition to the central warehouse, there are eight smaller warehouses (depots) in the capital cities of the regions (provinces); however, these are mainly used to store larger pieces of equipment. All transportation occurs through 3PLs as KEMSA does not own its own vehicle fleet. In order to build the Kenya supply chain model, the following data elements were gathered. First, latitude and longitude values for the district capitals were obtained and used for distance calculations and to distribute medical supplies based on population (using the population data gathered in the health model). Next, the volume and weight for one package of each commodity in the material requirements model was determined as well as the commodity cost. When available, this was taken from the historical shipment data; for those not available, the International Drug Price Indicator Guide (Management Sciences for Health 2008), Logistics Fact Sheet: ARV Drugs (DELIVER 2008), or Logistics Fact Sheet: HIV Test Kits (DELIVER 2008) were used. Historical shipment data for the year 2009-2010 was provided by KEMSA, detailing the amount and value of commodities shipped out of the central warehouse by date. This data was used to 1) create an additional commodity in the supply chain model that represents the other essential medicines which are not yet specifically modeled as part of the disease- and program-specific commodities identified in the material requirements model, 2) determine what fraction of commodities go down to the lowest level of the supply chain (modeled as the district capitals in the supply chain model) and which commodities stay at a higher echelon (modeled as the region capitals in the supply chain model), and 3) calculate the 2009 average fill rate of orders, which is between 60 percent and 70 percent. Additionally, an itemized budget for the logistics operations in 2009-2010 was obtained from KEMSA and included administrative, transportation, warehousing, order management, inventory, and procurement costs. The KEMSA budget was used to 1) derive 3PL transportation costs, on a per cubic meter per kilometer basis; 2) calculate the cost to lease a square foot of warehouse space; 3) determine the average handling rate with which items are received and shipped by an employee at the central warehouse; and 4) calculate the per unit administrative cost associated with ordering commodities. The historical shipment and itemized budget were used to calibrate and validate the model that was built for 2010, which functions as the baseline to compare the results for the 2020– 2024 future state models. Baseline and Future Projection Assumptions General Assumptions The following assumptions were made across the modeling framework:  The exchange rate for Kenyan shilling to U.S. dollar was assumed to be 81.8:1, which was the rate as of June 4, 2010. 6  No inflation is assumed so that the monetary values reported in the model reflect constant prices. Health Model Assumptions The following assumptions were made across the health model framework:  Population growth rate was assumed equal for all districts.  Current condition prevalence rates were assumed for 2020–2024. Material Requirements Model Assumptions The following assumptions were made across the material requirements model framework:  The assumptions made for treatments by condition are summarized in Appendix G.  The treatments (number of tablets/vials of medicine per person per year) are assumed to be the same between the baseline and 2020–2024, with the exception of the assumption that a new vaccine package that is bulkier and more expensive will be used for 2020–2024.  Based on The Business of Health in Africa (International Finance Corporation 2008), 50 percent of health supply needs in sub-Saharan countries is fulfilled by the private sector and the other 50 percent by the public sector. Due to KEMSA’s fill rate of 60 percent to 70 percent (calculated from historical shipment data), it is assumed that KEMSA currently fulfills 35 percent of the country’s total need. Supply Chain Model Assumptions The following assumptions were made across the supply chain model framework:  The derived per square foot warehouse leasing cost is used for the central warehouse and any new warehouses considered in the 2020–2024 scenarios.  The derived administrative cost per commodity unit is applied at the central warehouse in the baseline and future state scenarios; however, it is not applied at any regional warehouses considered in future state scenarios.  The derived 3PL transportation rate is used for both the baseline and future state scenarios for which 3PLs are considered.  The derived handling rate is used in the future state scenarios at the central warehouse and any new warehouses that are considered.  Warehouse employee annual wages are estimated at US$9,700 based on wages from WHO- Choosing Interventions that are Cost-Effective: 10 percent of logistics supply manager wage plus 90 percent of logistics storekeeper wage.  Vehicle driver annual wages are estimated at US$7,700 based on wages from WHO-Choosing Interventions that are Cost-Effective: 10 percent of transportation manager wage plus 90 percent of driver wages.  For future state scenarios where a KEMSA-owned vehicle fleet is considered, the vehicle characteristics are summarized in Appendix I. 7 Quantification Analysis First, using PHeNOM, a public health–oriented supply chain modeling and optimization tool developed by LLamasoft, Inc., the modeling framework was used to generate a baseline set of metrics given current conditions and the most likely expected trends based on current trajectories. The baseline was used to analyze how the system is currently operating. It is important as a first step to validate that the model is able to reproduce current conditions so that the future state scenarios can be run with confidence. It was necessary to consider donated commodities in the baseline model to calculate the current state costs. The model included 69 district facilities and 130 total products. Next, the modeling framework was reapplied under a scenario analysis approach, with key variables adjusted and relationships tested, in order to forecast a range of possible future state supply chain metrics utilizing a variety of possible conditions. For each scenario, there were three main types of output: financials, service metrics, and supply chain metrics. The financial metrics include supply chain costs (transportation, vehicle, facility operating, warehousing, and administrative), procurement costs, safety stock investment, and supply chain cost rate (percentage of supply chain cost compared with value of delivered goods). The supply chain cost rate begins to examine the effectiveness of a supply chain; in other words, how the supply chain costs compare to the value of goods that is flowing through this supply chain. However, it is only useful if there is a baseline value or data from multiple scenarios to compare. The service metrics include met and unmet need, warehouse in-stock fill rates, service time and distance to the customers, delivered value (value of goods delivered countrywide), and delivered value per person (delivered value divided by the country’s population). Lastly, supply chain metrics contain warehouse data (type, number, location, size, number of employees, and cost), vehicle data (number, number of drivers, and cost), and safety stock levels needed to reach in-stock fill rates. The output metrics from the future state scenarios were analyzed and consolidated to provide actionable information for policymakers, who can in turn make decisions and plans that encourage one set of metric outcomes and potentially avoid other less desirable sets. A description of the outputs the framework generates and what is included in each output is detailed in Appendix J. 8 Key Findings Baseline Analysis The first step in any modeling exercise is to provide a benchmark for the software and verify that the tool can reproduce the known data. Once the software is able to meet this requirement, the user can have confidence in the tool’s ability to predict future state scenarios. PHeNOM was validated for the Kenyan 2020 project by building a model using known 2009-2010 data and comparing the model outputs to 2009-2010 actual metrics (provided by KEMSA). The supply chain warehouse space was calibrated to its actual value and as previously mentioned, it was assumed that this public health network serves 35 percent of the population. Figure 2 depicts the current network structure in Kenya. The central warehouse is located in Nairobi, which delivers products directly out to the districts (these lanes are represented by purple lines). As illustrated by the greater number of district warehouses, the majority of the Kenyan population is focused in the south/southwest of the country. Figure 2. Current Network Structure in Kenya Nairobi As shown in table 1, the supply chain costs predicted by PHeNOM range from within 0.0 percent to 2.9 percent of the actual cost, and the total cost is within 1.7 percent. The numbers indicate that the assumptions used in the model are valid and PHeNOM is capable of reproducing the 2009 current state in Kenya. This analysis also demonstrates that the largest cost currently facing KEMSA is transportation at 46.3 percent of the current budget. This is nearly double the second largest cost (facility at 25.0 percent). The calculated supply chain metrics for the baseline model are provided in table 2. The delivered value is much higher than the total procurement cost because HIV, malaria, and tuberculosis medicine was donated by various organizations. Procurement cost captures only the cost of commodities actually paid for by KEMSA; delivered value also includes the value of the donated goods. The delivered value per person and supply chain cost rate will be compared to future 9 state scenarios. The cost to procure the medicine and diagnostic and preventive commodities in 2009-2010 was quite high at U.S.$72 million. Table 1. Baseline Supply Chain Cost Comparison PHeNOM Output Percent of Total Budget (%) KEMSA Data Percent Difference Predicted versus Actual (%) Total transportation cost US$2,624,702 46.3 US$2,704,126 2.9 Total facility cost US$1,414,500 25.0 US$1,414,463 0.0 Total administrative cost US$269,306 4.7 US$272,406 1.1 Total warehousing cost $1,359,146 24.0 $1,375,891 1.2 Total supply chain cost $5,667,654 - $5,766,886 1.7 Table 2. Baseline Supply Chain Metrics Total safety stock investment Total procurement cost US$12,170,104.40 US$72,771,526.44 Delivered value US$287,137,385.66 Delivered value per person US$7.17 Supply chain cost rate (%) 1.97 Future State Analysis The following future state scenarios explored two key questions. First, if the network structure does not change from its current operating state but the population continues to grow over time, how will this affect the service levels? Second, how can the supply chain network be restructured to account for additional demand and increased service levels? Multiple options were considered and included: 1) increased replenishment rate, 2) expansion of the central warehouse, 3) additional tier in the supply chain (i.e., regional warehouses), 4) KEMSA-owned fleet rather than using 3PLs, and 5) assessing multiple variables simultaneously. Network Structure Remains the Same The first future state scenario examined how population growth in 2020–2024 would affect service level if the supply chain network structure remained unchanged. The Kenyan population was predicted for the years 2020–2024 using projections from historical data; the population is expected to increase from 19.1 percent to 25.1 percent, from 40.0 million in 2010 to 49.5 million in 2020, 50.5 million in 2021, 51.5 million in 2022, 52.4 million in 2023, and 53.4 million in 2024. Figure 3 shows the proportion of the population’s need that will be served by KEMSA in 2020–2024 if the network remains the same. As illustrated in the graph, the need satisfied by KEMSA will decrease from 35 percent in 2010 to 28.5 percent in 2024. Hence, only 28.5 percent of the Kenyan population will 10 receive life-saving medicine and diagnostic and preventive commodities through the public sector. The 71.5 percent (38.2 million people) not served by KEMSA in 2024 will have to receive service from the private sector or will not receive health care at all. Figure 3. Supply Chain Metrics for 2010 and 2020–2024 Figure 4A shows the delivered value per person for 2020–2024 compared to the baseline state in 2010. As expected, the delivered value per person decreases from US$7.17 in 2010 to US$6.82 in 2020 and continues to decrease over the next four years. Although demand is increasing, the throughput of products remains the same. Figure 4B demonstrates the delivered value for 2010 and 2020–2024. This graph is a bit misleading in the respect that it shows the delivered value increasing over time. In reality, this is an artifact of the modeling software; the tool optimizes on revenue so it chooses commodities with the highest revenue but lowest cost to deliver. However, it does demonstrate a real-world situation. If the network does not change to accommodate the increased demand, there will not be enough space to house all of the necessary commodities nor will there be enough budget for procurement of the goods or supply chain costs to push the goods through the network. As such, it will be necessary for KEMSA to choose which commodities to procure, and a decision will be made that will neglect a certain group or groups. This point is illustrated in table 3, which shows the commodities for which KEMSA meets less than 25 percent of the need in this scenario (for some commodities, the need met by KEMSA decreases even more than the aggregated need met by KEMSA of 28.5 percent). Hence, in this situation, a patient that required aminophylline as an asthma treatment would not receive it. In this analysis, the PHeNOM tool has dealt with the issue of supply chain infrastructure that cannot serve all demand by choosing to distribute products that have the highest value. In reality, this situation shows that, without investment to increase supply chain capacity, government officials will face increasingly difficult choices about which populations to serve and which conditions to treat. 11 Figure 4. Supply Chain Metrics for 2010 and 2020–2024 A B Table 3. Commodities Resulting in an Unmet Need of 75 Percent or Greater Commodity Condition Proportion of Need Served by the Kenya Medical Supplies Agency Aminophylline 25 mg/mL, 10 mL Asthma 0 Determine HIV rapid test kits, 100 tests HIV and AIDS 0 Stavudine 30mg/lamivudine/nevirapine (30 mg/150 mg/200 mg), 60 tabs/bottle HIV and AIDS 4 Salbutamol 2 mg/5 mL, 100 mL, syrup Respiratory infection/asthma 11 Zidovudine 150 mg, 100 caps HIV and AIDS 14 Cotrimoxazole 240 mg/5 mL, 100 mL HIV and AIDS/respiratory infection/diarrhea 14 Other essential medicines Other 14 The total procurement and supply chain costs for 2024 were calculated by PHeNOM and found to be U.S.$359.54 million and U.S.$5.76 million, respectively. To compare back to the current 2009- 2010 costs (U.S.$72.77 million for total procurement and U.S.$5.67 million for supply chain costs), the supply chain cost remains relatively the same (this is expected as the network has not changed and remains at maximum capacity) while the procurement cost increases by U.S.$286.77 million. The procurement cost is higher because it is assumed that commodities will not be donated as they were in 2009 and that KEMSA must pay for all commodities. This demonstrates that even if more supplies are procured or donated, they will not be delivered without investment in the network structure. If the need is not being met now and the proper investment is not made in supply chain costs, it will not be possible to fulfill an even greater need (which will occur due to population growth). It is important to not only push supplies into the country but also to invest in the supply chain network to handle the increased demand flow. 12 Restructuring the Supply Chain The following future state scenarios (2020–2024) examined the effect of restructuring the current network structure to account for the additional demand flowing through the system. In the scenarios, fulfilling 35 percent, 50 percent, or 100 percent of the countrywide need was utilized. Increased Replenishment Rate In the current situation, KEMSA receives a shipment of goods once per year and stores the commodities in the central warehouse in Nairobi. Here, the effect of increased replenishments, two or four yearly shipments, in 2020–2024 was examined. When fulfilling 35 percent or 50 percent of the demand, only two shipments are necessary per year. For a requirement of 100 percent demand, if two shipments are utilized, only 63 percent to 66 percent (2020–2024) of the need will be satisfied. Hence, 33 percent to 37 percent of the Kenyan population will not receive essential life-saving medicine and diagnostic and preventive commodities. However, 100 percent of the demand can be fulfilled if four shipments are utilized. Figures 5A and 5B demonstrate the delivered value and supply chain cost as compared to the baseline (2010). It was determined from the baseline model (2010) that the total logistics cost is currently US$5,667,654. As such, there is a need for a 17 percent, 38 percent, and 42 percent investment into the supply chain logistics cost to meet 35 percent, 50 percent, and 100 percent of demand in 2024, respectively. Data is provided for 100 percent demand and two shipments in figure 5, but this does not provide the full picture as four shipments would be necessary to actually fulfill 100 percent demand. This scenario demonstrates the importance of investing in the supply chain and the timing of delivery. Figure 5. Supply Chain Metrics for 2010 and 2020–2024 A B Central Warehouse Expansion In this scenario, the effect of allowing the central warehouse to expand and accommodate 35 percent, 50 percent, and 100 percent of the 2020–2024 predicted demand was examined. All other aspects of the network remained as is. The current size of the central warehouse is 204,894 square feet; however, it is already at capacity. Figure 6A shows the increase in the central warehouse that is necessary in order to meet 35 percent, 50 percent, and 100 percent of the need of the Kenyan population. For 35 percent, it must be expanded 22 percent (average over 2020–2024); for 50 percent, an expansion of 45 percent is necessary (average over 2020–2024); to meet 100 percent of the need, 75 percent growth is required (average over 2020–2024). From this data, it is apparent that 13 in order to meet the increased need, the central warehouse needs to expand significantly. PHeNOM also predicted the number of warehouse employees that would be required to ensure the warehouse is running efficiently at the increased capacity, as provided in figure 6B. Figures 6C and 6D show the delivered value and supply chain cost as compared to the baseline (2010). As in the previous scenario, it is apparent from the increased supply chain cost that it is necessary to invest in the supply chain network in order to fulfill the future state demand. Figure 6. Supply Chain Metrics for 2010 and 2020–2024 Number of Employees in Central Warehouse 2020–2024 35% 153-165 50% 219-236 100% 437-472 A B C D Addition of Regional Warehouses The next scenario examined the benefits of adding regional warehouses to the network. In this “decentralized” supply chain structure, the goods are first delivered to the central warehouse in Nairobi and then flow through regional warehouses rather than directly to the districts. This creates an intermediary holding facility, reducing the amount of supply stored at the central warehouse. From the regional warehouses, all shipments are routed to the districts. There can be multiple benefits to adding regional warehouses. First, there is the ability to pool products from central to regional warehouses on larger trucks rather than utilize more frequent shipments on smaller trucks, which can lower the transportation cost. Second, as the regional warehouses are, for the most part, closer to the districts than the central warehouse, a decentralized network would move stock closer to the population, reducing the service distance and hence, delivery times. A “greenfield” analysis was conducted to determine the optimal location for one, two, or three regional warehouse(s) based on the overall demand in the network. The analysis determined that Nakuru would be the ideal location if only a single warehouse were to be added (figure 7A), Kisumu and Mwingi if two warehouses were utilized (figure 7B), and Kisumu, Nyeri, and Bamburi if three 14 Nakuru Nairobi Nairobi Mwingi Kisumu Kisumu Bamburi Nyeri Nairobi A B C warehouses were added (figure 7C). However, the feasibility of the locations would be dependent on land availability and economic and political considerations, which are beyond the scope of this study. An interesting finding is that the location of all the regional warehouses is predicted to be concentrated in the south of the country. This is not too surprising, however, because the majority of the Kenyan population is focused in the south/southwest of the country. Figure 7. Results of Greenfield Analysis Central warehouse is shown in yellow, predicted regional warehouses in red, and district customers in blue. The green lines indicate the flow of demand from the central warehouse to the regional warehouses to the districts. A) Location of a single regional warehouse. B) Location of two regional warehouses. C) Location of three regional warehouses. Figures 8A and 8B provide the supply chain financials for the baseline (no regional warehouses) compared to future state scenarios with one, two, or three regional warehouses being utilized in the network. For this scenario, only fulfilling 50 percent of the countrywide need was examined. As shown in figure 8A, the total logistics cost almost doubles between the baseline (no regional warehouse) and when a single regional warehouse is utilized. However, there is marginal cost difference between adding a single regional warehouse compared to two or three regional warehouses. In figure 8B, the percentage for each component of the logistics cost is given for the baseline versus three regional warehouses being utilized; the largest cost shifts from transportation in the baseline to facility costs when regional warehouses are added. 15 Figure 8. Supply Chain Financials for Baseline and Adding Regional Warehouse Scenarios A B To determine if the large investment necessary for the addition of regional warehouses is worthwhile, the resulting supply chain service metrics were examined. The average service distance and hours for the baseline compared to the three regional warehouse scenarios is provided in figures 9A and 9B, and the maximum service distance and hours is provided in figures 9C and 9D. As demonstrated by this data, the service levels do not improve significantly with the addition of intermediate facilities. Hence, the marginal service improvements do not justify the large investment necessary for the supply chain logistics. LLamasoft’s experience in other countries, in addition to these results, suggests that country size is a good indicator of whether an additional warehouse layer would be beneficial to the supply chain network. Based on this analysis, Kenya is not large enough to justify the investments needed for additional intermediate facilities. 16 Figure 9. Supply Chain Service Metrics for Baseline and Adding Regional Warehouse Scenarios A B C D Kenya Medical Supplies Agency–Owned Fleet In the current supply chain network, the largest cost facing KEMSA is transportation between the central warehouse and districts. To date, a 3PL is being utilized to deliver the goods. Here, it was assessed whether it is beneficial for KEMSA to own a vehicle fleet. For this scenario, only fulfilling 50 percent of the countrywide need was examined. Figure 10A shows the transportation cost required for 3PLs and a KEMSA-owned fleet over 2020– 2024. For the KEMSA-owned fleet, two depreciation rates were examined: 10 percent and 25 percent. In the best operating conditions, the vehicle fleet is well maintained, drivers are highly trained, and roads have been improved. However, in reality this is rarely the case in sub-Saharan Africa, and a higher depreciation rate such as 25 percent is more applicable. In this study, both 10 percent and 25 percent were examined to determine the cost effects of depreciation over time. The motivation was to assess the savings that could result from better maintenance of the vehicles, in addition to providing information on the cost of KEMSA owning its own vehicle fleet. As shown in figure 10A, there is a large upfront investment to own a vehicle fleet, but it becomes cost-effective over time. If the vehicles are well maintained and a 10 percent depreciation rate can be assumed, it is actually cheaper over time for KEMSA to own its own fleet rather than outsourcing to a 3PL. PHeNOM also calculated that it would be necessary for KEMSA to purchase 99 to 107 vehicles and employ 94 to 102 drivers. As demonstrated by figure 10B, the total supply chain costs summed over five years, maintenance of the vehicle fleet is extremely important. If a 25 percent 17 depreciation rate is assumed, there is a US$3.7 million difference over five years from the 10 percent depreciation rate data, and the cost savings versus 3PL transportation becomes negligible. Figure 10. Supply Chain Metrics A B A breakdown of the transportation cost is provided in table 4. In 2020, for the KEMSA-owned fleet, the largest component is the vehicle purchase (56 percent of the cost for 10 percent depreciation; 52 percent of the cost for 25 percent depreciation); fuel cost is second (24 percent of the cost for 10 percent depreciation; 22 percent of the cost for 25 percent depreciation). However, in 2024, fuel cost becomes most expensive as the vehicle cost significantly drops after the initial investment was made (53 percent of the cost for 10 percent depreciation; 44 percent of the cost for 25 percent depreciation). So, while there is a large cost associated with owning a vehicle fleet in the first year to purchase the vehicles, this becomes a cost-effective option for KEMSA as long as the vehicles are properly maintained over time. If the vehicles are not properly maintained, then the currently employed 3PL option is relatively the same cost. Experience of publicly owned transport fleets suggests that maintenance is typically a weakness in the public sector. Table 4. Comparison of Transportation Cost Components (U.S. Dollars) 2020 Third-Party Logistics Provider (3PL) Own Fleet, 10 Percent Depreciation Own Fleet, 25 Percent Depreciation 3PL $4,619,254 Vehicle purchase $4,708,559 $4,708,478 Vehicle fixed cost $936,755 $1,643,011 Vehicle fuel cost $1,962,488 $1,962,454 Vehicle driver wages $726,498 $726,486 Total transportation cost $4,619,254 $8,334,300 $9,040,429 18 2020 Third-Party Logistics Provider (3PL) Own Fleet, 10 Percent Depreciation Own Fleet, 25 Percent Depreciation 2024 3PL $4,988,436 Vehicle purchase $97,379 $97,291 Vehicle fixed cost $1,011,623 $1,774,324 Vehicle fuel cost $2,119,335 $2,119,298 Vehicle driver wages $784,562 $784,548 Total transportation cost $4,988,436 $4,012,899 $4,775,461 Assessing Multiple Variables In the final scenario, three variables were examined in combination: addition of regional warehouses, multiple replenishment schedule (six shipments per year), and a KEMSA-owned fleet (small-, medium-, and large-sized trucks; 10 percent depreciation rate of the vehicles). For this scenario, only fulfilling 100 percent of the countrywide need was examined. Figure 11 shows a graphical representation of the future state network. PHeNOM was allowed to choose between two delivery methods: 1) central warehouse directly to districts (on medium-sized trucks) and 2) central warehouse to regional warehouse (on large-sized trucks) to districts (on small-sized trucks). The different transportation lanes are colored according to the size truck being utilized. Figure 11. Future State Supply Chain Network WH - warehouse. A greenfield analysis was again conducted to determine the optimal locations for regional warehouses based on the overall demand in the network. In this scenario, eight regional warehouses were assessed but only seven were utilized by PHeNOM (Nyeri, Nakuru, Kisumu, Mumias, Eliye Springs, and Wajir), as shown in figure 12A. As previously mentioned, the feasibility of the locations would be dependent on land availability and economic and political considerations, which are beyond the scope of this study. Warehouses were added in the north part of the country, unlike the 19 previous scenario. The size of each warehouse is provided in figure 12B. As expected, the largest warehouse is the central warehouse in Nairobi for both 2020 and 2024, as all goods must first flow through here. The next largest warehouse is predicted to be Mumias for both 2020 and 2024; Mumias is centered in an area in the southwest heavily populated by district customers. The Nyeri regional warehouse was only opened in 2024, when an increased amount of demand flows through the system. Figure 12. Results of Greenfield Analysis B Eliye Springs Mumias Kisumu Wajir Nakuru Nyeri Bamburi Nairobi A Size (sq ft) 2020 2024 Nairobi central warehouse 718,967 776,429 Bamburi depot 46,692 50,424 Nyeri depot 0 6,513 Kisumu depot 49,382 53,329 Wajir depot 6,209 6,706 Eliye Springs depot 9,238 9,976 Mumias depot 78,442 84,711 Nakuru depot 41,031 44,310 A) Locations of the predicted regional warehouses. Warehouses are shown in blue and district customers in green. B) Size of warehouses for 2020 and 2024. Figures 13A and 13B provide the supply chain financials for the scenario “expanding the central warehouse” (3PL and the central warehouse) compared to “assessing multiple variables” (KEMSA- owned fleet, regional warehouses, and shipment schedule). For this scenario, fulfilling 100 percent of the countrywide need was examined. As shown in figure 13A, the total logistics cost is significantly cheaper for the assessing multiple variables scenario than only expanding the central warehouse. This suggests that multiple changes to the network may be necessary to reduce the supply chain costs. In figure 13B, the percentage for each component of the logistics cost is given for the scenario “expanding the central warehouse” (3PL and the central warehouse) compared to “assessing multiple variables” (KEMSA-owned fleet, regional warehouses, and shipment schedule). The majority of the cost when only expanding the central warehouse is from transportation. However, for the assessing multiple variables scenario, both transportation and warehousing costs contribute the most to the overall supply chain costs. The addition of regional warehouses drives up the warehouse cost. 20 Figure 13. Supply Chain Financials Comparing Multiple Scenarios: Expansion of Central Warehouse (3PL and the Central Warehouse) versus Assessing Multiple Variables (KEMSA-owned Fleet, Regional Warehouses, and Shipment Schedule) A B Supply chain metrics and service metrics are compared for two scenarios in table 5: assessing multiple variables and KEMSA-owned fleet (central warehouse only). In the assessing multiple variables scenario, small-, medium-, and large-sized trucks were utilized, and the number of trucks and drivers necessary to fulfill 100 percent of the demand varies between 33 to 37, 102 to 109, and 22 to 24 for trucks, respectively, and 150 to 162 drivers. The most interesting finding is that the addition of regional warehouses improves the service time. The average service distance decreases from 310.7 kilometers to 243.6 kilometers, and the average service hours (this includes driving time only, not loading/unloading of the vehicle) from 6.2 hours to 4.9 hours. Table 5. Supply Chain and Service Metrics 100 Percent, 2020–2024 Central and Regional Warehouses and Kenya Medical Supplies Agency–Owned Fleet Only Central Warehouse and Kenya Medical Supplies Agency–Owned Fleet Number of small trucks 33–37 0 Number of medium trucks 102–109 198–214 Number of large trucks 22–24 0 Number of vehicle drivers 150–162 189–204 Maximum service distance (kilometers) 1,006.2 1,006.2 Average service distance (kilometers) 243.6 310.7 Maximum service hours 20.1 20.1 21 100 Percent, 2020–2024 Central and Regional Warehouses and Kenya Medical Supplies Agency–Owned Fleet Only Central Warehouse and Kenya Medical Supplies Agency–Owned Fleet Average service hours 4.9 6.2 22 Discussion The focus of this project was to build a general modeling framework that can be used to forecast the public health supply chains for any country for any future time frame. Here, it was applied to Kenya for the years 2020 through 2024 to demonstrate how the modeling approach can be utilized to help policymakers accurately visualize and understand the most likely and possible situations facing them in 10 years. However, it can be applied to any country to help make informed decisions about how to design a cost-effective supply chain that fulfills the necessary future demands. Multiple future state scenarios were conducted to demonstrate the various ways a supply chain network could be restructured to account for additional demand and increased service levels. However, this is by no means a comprehensive list of the different variables that can be assessed. For example, the effect of varying prevalence rates, the implementation of new vaccines/treatments, or adding another import point in Mombasa could also be examined. Table 6 shows a summary of total supply chain and procurement budgets for all scenarios across the various levels of demand fulfillment (35 percent, 50 percent, and 100 percent). In general, the most cost-effective implementation would be to increase the number of replenishments each year. It would, however, require coordination between multiple stakeholders to accomplish a replenishment schedule. In most cases, other scenarios are very similar in their total supply chain budget. Table 6 also demonstrates that to improve service levels, it is necessary to double (50 percent) or quadruple (100 percent) the supply chain budget from the baseline cost. 23 Table 6. Summary of Total Supply Chain and Procurement Budgets (U.S. Dollars) Percent (%) Total Supply Chain Budget (2024) Total Procurement Budget (2024) Baseline (2010) $5,667,654 $73 million 35 Central warehouse and increased replenishments (two shipments) $6,825,746 $375 million 35 Central warehouse and third-party logistics provider (3PL) $7,209,749 35 Central warehouse and Kenya Medical Supplies Agency (KEMSA)-owned fleet (10% depreciation) $6,502,979 35 Central warehouse KEMSA-owned fleet (25% depreciation) $7,036,754 50 Central warehouse and increased replenishments (two shipments) $9,144,845 $533 million 50 Central warehouse and 3PL $10,299,637 50 Central warehouse and KEMSA-owned fleet (10% depreciation) $9,323,892 50 Central warehouse and KEMSA-owned fleet (25% depreciation) $10,086,427 50 Central and two regional warehouses, and 3PL $22,191,007 100 Central warehouse and increased replenishments (four shipments) $16,875,195 $1,067 million 100 Central warehouse and 3PL $20,599,278 100 Central warehouse and KEMSA-owned fleet (10% depreciation) $18,647,790 100 Central warehouse and KEMSA-owned fleet (25% depreciation) $20,172,859 100 Central and regional warehouses, and KEMSA-owned fleet (shipments every two months) $14,265,173 24 Conclusion The modeling framework was developed and tested by applying it to analyze Kenya and its current and future supply chain states. Having done the groundwork and overcoming the challenges along the way, a robust framework now exists that can be easily reused and applied to other countries and scenarios. The interlinked modeling framework, which represents the health, material requirements, and supply chain models, and a data gathering tool allow for future models to be built and applied quickly and flexibly. The substantial legwork performed in order to shape the current software structure and populate it with accurate data will significantly reduce the time and effort needed to create models for future projects using this robust framework. The developed data gathering template will serve as a guideline for a modeler to extract and use the appropriate data. Based on the analysis work conducted, the following results should be highlighted. First, if the Kenyan population continues to grow at a steady pace and the supply chain network remains as is, proportion of need met by KEMSA will decrease from 35 percent in 2010 to 28.5 percent in 2024. Additionally, it is important to not only push supplies into the country but to also invest in the supply chain network to handle the increased flow. The forecasted supply chain and procurement budget in 2024 is U.S.$5.76 million and U.S.$359.54 million, respectively. Multiple cost-effective ways to restructure the supply chain network were examined in order to meet the needs of the growing Kenyan population over the years 2020–2024. In the first scenario, it was found that two replenishments per year would be necessary to meet 35 percent and 50 percent of the total demand, while four replenishments would be needed for 100 percent. In the second scenario, expansion of the central warehouse, the central warehouse must be expanded 22 percent (average over 2020–2024) to meet 35 percent of the future demand, an expansion of 45 percent is necessary (average over 2020–2024) for 50 percent, and 75 percent growth is required (average over 2020–2024) to meet 100 percent of the need. Additionally, a decentralization scenario was conducted where regional warehouses are utilized. This scenario demonstrated that a large investment would be necessary and only result in marginal service improvements. Although there would be shorter service distance and time to districts, Kenya as a country is small enough that the impact on costs and overall network efficiency would be minimal. Additionally, it was assessed whether it would be cost-effective for KEMSA to utilize an in-house vehicle fleet. The analysis found that although there would be a large upfront investment to own a vehicle fleet, it becomes cost-effective over time. Furthermore, if the vehicles are well maintained and a 10 percent depreciation rate can be assumed, it is actually cheaper over time for KEMSA to own its own fleet rather than outsourcing to a 3PL. However, if a 25 percent depreciation rate is assumed, there is a US$3.5 million difference over five years from the 10 percent depreciation rate data, and utilizing a 3PL is relatively the same cost. This study demonstrates the importance of vehicle maintenance and how much of a factor it can play into supply chain transportation costs. In the last scenario, multiple variables were assessed together. Here, decreased overall supply chain costs were found; for fulfilling 100 percent of the demand, it was the cheapest option. This suggests that multiple changes to the network may be necessary to significantly reduce the supply chain costs and increase service levels. The majority of the cost when only expanding the central warehouse is 25 from transportation. However, for the assessing multiple variables scenario, both transportation and warehousing costs together contribute the most to the overall supply chain costs. In general, the most cost-effective implementation would be to increase the number of replenishments each year. The power of PHeNOM is the ability to predict and visualize the future. Changes to strengthen the logistics situation are required, but they can be manageable if policymakers understand the most likely and possible situations facing them. If nothing is done to address the increased demand due to a growing Kenyan population, the public sector will not be serviced, which ultimately can result in loss of lives. The key message from this study is that stakeholders must understand the importance of investing in supply chain and the timing of delivery. Future opportunities with PHeNOM building on the framework developed for this project exist and include applying the framework to multiple other countries to compare and identify regional trends. Additionally, information could be obtained, such as the typical benchmark numbers for the supply chain cost rate and other service metrics. Furthermore, additional scenarios could be conducted and include (among others) the following: 1) other network structures to consider; 2) the effect of higher fuel prices; 3) the effect of changing demographics due to urbanization, and economic/political changes and events; 4) the effect of changing prevalence rates due to vaccine introduction, disease outbreaks, and disease eradication; and 5) the effect of feedback loops such as lower maternal morbidity leads to higher birth rates and therefore larger population on the supply chain needs of a country. Lastly, PHeNOM could be employed to investigate the relation of supply chain service metrics to health outcomes. For example, how does the number of patients that are treated/untreated translate into quality-adjusted life years/disability-adjusted life years? 26 Appendix A Kenya Population Data 2010 & 2020 Name Province District Admin Level 2010 2020 KENYA Country 40,046,566 49,458,956 KENYA_CENTRAL CENTRAL Province 5,198,935 6,420,870 KENYA_CENTRAL_KIAMBU CENTRAL KIAMBU District 1,038,638 1,282,757 KENYA_CENTRAL_KIRINYAGA CENTRAL KIRINYAGA District 638,117 788,095 KENYA_CENTRAL_MARAGUA CENTRAL MARAGUA District 541,610 668,910 KENYA_CENTRAL_MURANG'A CENTRAL MURANG'A District 486,232 600,513 KENYA_CENTRAL_NYANDARUA CENTRAL NYANDARUA District 669,947 827,404 KENYA_CENTRAL_NYERI CENTRAL NYERI District 922,972 1,139,910 KENYA_CENTRAL_THIKA CENTRAL THIKA District 901,419 1,113,280 KENYA_COAST COAST Province 3,472,226 4,288,326 KENYA_COAST_KILIFI COAST KILIFI District 759,848 938,442 KENYA_COAST_KWALE COAST KWALE District 692,598 855,386 KENYA_COAST_LAMU COAST LAMU District 101,474 125,322 KENYA_COAST_MALINDI COAST MALINDI District 393,045 485,426 KENYA_COAST_MOMBASA COAST MOMBASA District 928,371 1,146,567 KENYA_COAST_TAITA COAST TAITA District 344,349 425,289 KENYA_COAST_TANA RIVER COAST TANA RIVER District 252,541 311,895 KENYA_EASTERN EASTERN Province 6,465,973 7,985,711 KENYA_EASTERN_EMBU EASTERN EMBU District 388,360 479,639 KENYA_EASTERN_ISIOLO EASTERN ISIOLO District 140,803 173,897 KENYA_EASTERN_KITUI EASTERN KITUI District 719,531 888,645 KENYA_EASTERN_MACHAKOS EASTERN MACHAKOS District 1,265,678 1,563,161 KENYA_EASTERN_MAKUENI EASTERN MAKUENI District 1,077,075 1,330,230 KENYA_EASTERN_MARSABIT EASTERN MARSABIT District 169,583 209,435 KENYA_EASTERN_MBEERE EASTERN MBEERE District 238,650 294,749 KENYA_EASTERN_MERU CENTRAL EASTERN MERU CENTRAL District 696,442 860,128 KENYA_EASTERN_MERU NORTH EASTERN MERU NORTH District 843,257 1,041,449 KENYA_EASTERN_MOYALE EASTERN MOYALE District 74,655 92,205 KENYA_EASTERN_MWINGI EASTERN MWINGI District 424,143 523,830 27 Name Province District Admin Level 2010 2020 KENYA_EASTERN_NITHI (MERU SOUTH) EASTERN NITHI (MERU SOUTH) District 286,814 354,222 KENYA_EASTERN_THARAKA EASTERN THARAKA District 140,982 174,121 KENYA_NAIROBI AREA NAIROBI AREA PROVINCE Province 2,991,987 3,695,213 KENYA_NAIROBI AREA_NAIROBI AREA NAIROBI AREA NAIROBI AREA District 2,991,987 3,695,213 KENYA_NORTH-EASTERN NORTH-EASTERN Province 1,343,153 1,658,841 KENYA_NORTH-EASTERN_GARISSA NORTH-EASTERN GARISSA District 547,946 676,727 KENYA_NORTH-EASTERN_MANDERA NORTH-EASTERN MANDERA District 349,520 431,677 KENYA_NORTH-EASTERN_WAJIR NORTH-EASTERN WAJIR District 445,687 550,438 KENYA_NYANZA NYANZA Province 6,131,514 7,572,642 KENYA_NYANZA_BONDO NYANZA BONDO District 333,337 411,685 KENYA_NYANZA_GUCHA (SOUTH KISII) NYANZA GUCHA (SOUTH KISII) District 643,475 794,714 KENYA_NYANZA_HOMA BAY NYANZA HOMA BAY District 402,798 497,472 KENYA_NYANZA_KISII CENTRAL NYANZA KISII CENTRAL District 686,536 847,894 KENYA_NYANZA_KISUMU NYANZA KISUMU District 704,086 869,578 KENYA_NYANZA_KURIA NYANZA KURIA District 212,035 261,864 KENYA_NYANZA_MIGORI NYANZA MIGORI District 718,799 887,747 KENYA_NYANZA_NYAMIRA NYANZA NYAMIRA District 695,352 858,783 KENYA_NYANZA_NYANDO NYANZA NYANDO District 418,702 517,114 KENYA_NYANZA_RACHUONYO NYANZA RACHUONYO District 428,749 529,516 KENYA_NYANZA_SIAYA NYANZA SIAYA District 670,335 827,892 KENYA_NYANZA_SUBA NYANZA SUBA District 217,310 268,383 KENYA_RIFT VALLEY RIFT VALLEY Province 9,753,920 12,046,445 KENYA_RIFT VALLEY_BARINGO RIFT VALLEY BARINGO District 369,913 456,858 KENYA_RIFT VALLEY_BOMET RIFT VALLEY BOMET District 534,379 659,976 KENYA_RIFT VALLEY_BURET RIFT VALLEY BURET District 442,370 546,346 KENYA_RIFT VALLEY_EAST MARAKWET RIFT VALLEY EAST MARAKWET District 196,316 242,457 KENYA_RIFT VALLEY_KAJIADO RIFT VALLEY KAJIADO District 566,857 700,086 KENYA_RIFT VALLEY_KEIYO RIFT VALLEY KEIYO District 200,831 248,036 KENYA_RIFT VALLEY_KERICHO RIFT VALLEY KERICHO District 654,017 807,733 KENYA_RIFT VALLEY_KOIBATEK RIFT VALLEY KOIBATEK District 192,880 238,211 KENYA_RIFT VALLEY_LAIKIPIA RIFT VALLEY LAIKIPIA District 449,771 555,486 KENYA_RIFT VALLEY_NAKURU RIFT VALLEY NAKURU District 1,657,107 2,046,589 KENYA_RIFT VALLEY_NANDI RIFT VALLEY NANDI District 807,939 997,836 KENYA_RIFT VALLEY_NAROK RIFT VALLEY NAROK District 510,591 630,594 KENYA_RIFT VALLEY_SAMBURU RIFT VALLEY SAMBURU District 200,390 247,486 KENYA_RIFT VALLEY_TRANS MARA RIFT VALLEY TRANS MARA District 238,141 294,124 KENYA_RIFT VALLEY_TRANS-NZOIA RIFT VALLEY TRANS-NZOIA District 803,629 992,503 KENYA_RIFT VALLEY_TURKANA RIFT VALLEY TURKANA District 629,397 777,334 KENYA_RIFT VALLEY_UASIN GISHU RIFT VALLEY UASIN GISHU District 869,300 1,073,611 28 Name Province District Admin Level 2010 2020 KENYA_RIFT VALLEY_WEST POKOT RIFT VALLEY WEST POKOT District 430,092 531,179 KENYA_WESTERN WESTERN Province 4,688,858 5,790,908 KENYA_WESTERN_BUNGOMA WESTERN BUNGOMA District 1,223,582 1,511,165 KENYA_WESTERN_BUSIA WESTERN BUSIA District 517,372 638,972 KENYA_WESTERN_ELGON WESTERN ELGON District 665,790 822,280 KENYA_WESTERN_KAKAMEGA WESTERN KAKAMEGA District 842,382 1,040,367 KENYA_WESTERN_LUGARI WESTERN LUGARI District 301,420 372,271 KENYA_WESTERN_SHIROTSA/MUMIAS WESTERN SHIROTSA/MUMIAS District 188,509 232,814 KENYA_WESTERN_TESO WESTERN TESO District 253,359 312,910 KENYA_WESTERN_VIHIGA WESTERN VIHIGA District 696,444 860,130 29 30 Appendix B Prevalence Rate Sources and Description Condition Source Date Description HIV HIV Spatial Data Repository, PEPFAR HIV Spatial Data webpage 2003 % prevalence of HIV, Men ages 15­ 45 & % prevalence of HIV, Women ages 15-45 Tuberculosis (a) Estimated epidemiological burden of TB, all forms WHO, TB Data webpage 2008 Prevalence rate per 100,000 Number Births World Health Statistics, 2009 WHO Report 2002 Live births in 2002 per 1,000 population Measles Kenya Reported Cases of Measles WHO Data Statistics and Graphs 2008 Total Cases reported Malaria Reported Malaria cases per 100,00 WHO spreadsheet 1990­ 2006 % Prevalence, average of 1990­ 2006 data Leprosy Division of Leprosy, Tuberculosis, and Lung Disease – Kenya Ministry of Public Health and Sanitation 2006 # cases in 2006 (only for 5 provinces given) Childhood Diarrheal “Mortality Country Factsheet 2006” WHO 2006 Assuming 25% for children and 10% for adults Worms “Worms: Education and Health Externalities in Kenya” National Bureau of Economic Research 2002 Adding 5% to worldwide rate of 25% Respiratory Infections “Community Understanding of pneumonia in Kenya” African Health Sciences, Vol 8 2008 Using Pneumonia as a proxy ­ Children 13% (Upper limit of Range); Adults: 10% (guess) Syphilis “The epidemiology of gonorrhea, chlamydial infection and syphilis in four African cities.” NLM Gateway 2001 4% Overall STI “Female condom introduction and STI prevalence” NLM Gateway 2000 17% Diabetes “Global prevalence of Diabetes: Estimates for the year 200 and projections for 2030” Diabetes Care (Volume 27) 2004 3% (worldwide estimate) HTN SECTION 1.02 “PREVALENCE, DETECTION, MANAGEMENT, AND CONTROL OF HYPERTENSION IN ASHANTI, WEST AFRICA” SECTION 1.03 AMERICAN HEART ASSOCIATION 2004 30% (estimate from Ghana) Asthma “Global Burden and management of Asthma in developing countries” International Union Against Tuberculosis and Lung Disease 10% (estimate from neighboring countries) 31 32 Appendix C Kenya Prevalence Rates Region (Province) Condition Central Coast Eastern Nairobi Area North- Eastern Nyanza Rift Valley Western HIV Men 2.00% 4.80% 1.50% 7.80% 0.00% 11.60% 3.60% 3.80% HIV Women 7.60% 6.60% 6.10% 11.90% 0.00% 18.30% 6.90% 5.80% Malaria 13.15% 13.15% 13.15% 13.15% 13.15% 13.15% 13.15% 13.15% Tuberculosis 180 180 180 180 180 180 180 180 Measles 1282 1282 1282 1282 1282 1282 1282 1282 Number of Births 37.5 37.5 37.5 37.5 37.5 37.5 37.5 37.5 Resp Inf - Child 7% 7% 7% 7% 7% 7% 7% 7% Resp Inf - Adult 10% 10% 10% 10% 10% 10% 10% 10% Diarrheal - Child 25% 25% 25% 25% 25% 25% 25% 25% Diarrheal - Adult 10% 10% 10% 10% 10% 10% 10% 10% Syphilis 4% 4% 4% 4% 4% 4% 4% 4% Other STI 23% 23% 23% 23% 23% 23% 23% 23% Leprosy 0 37 37 37 0 37 0 37 Vitamin A Def 100% Children 0-4 100% Children 0-4 100% Children 0-4 100% Children 0-4 100% Children 0-4 100% Children 0-4 100% Children 0-4 100% Children 0-4 Diabetes 3% 3% 3% 3% 3% 3% 3% 3% Hypertension 30% 30% 30% 30% 30% 30% 30% 30% Asthma 10% 10% 10% 10% 10% 10% 10% 10% Ascariasis Richuriasis Worms 25% 25% 25% 25% 25% 25% 25% 25% 33 34 Appendix D Number of HIV, Tuberculosis, and Malaria Cases in 2020 Country_Region_District Tuberculosis HIV Malaria KENYA 89,026 3,265,614 3,561,573 KENYA_CENTRAL 11,558 309,076 0 KENYA_CENTRAL_KIAMBU 2,309 61,162 0 KENYA_CENTRAL_KIRINYAGA 1,419 37,671 0 KENYA_CENTRAL_MARAGUA 1,204 32,657 0 KENYA_CENTRAL_MURANG'A 1,081 29,661 0 KENYA_CENTRAL_NYANDARUA 1,489 40,010 0 KENYA_CENTRAL_NYERI 2,052 55,141 0 KENYA_CENTRAL_THIKA 2,004 52,773 0 KENYA_COAST 7,719 243,527 127,881 KENYA_COAST_KILIFI 1,689 53,868 28,153 KENYA_COAST_KWALE 1,540 48,912 42,769 KENYA_COAST_LAMU 226 7,089 627 KENYA_COAST_MALINDI 874 27,664 14,563 KENYA_COAST_MOMBASA 2,064 64,053 34,397 KENYA_COAST_TAITA 766 24,195 4,253 KENYA_COAST_TANA RIVER 561 17,745 3,119 KENYA_EASTERN 14,374 307,382 30,346 KENYA_EASTERN_EMBU 863 18,264 0 KENYA_EASTERN_ISIOLO 313 6,497 869 KENYA_EASTERN_KITUI 1,600 34,874 4,443 KENYA_EASTERN_MACHAKOS 2,814 59,957 1,563 KENYA_EASTERN_MAKUENI 2,394 51,450 1,330 KENYA_EASTERN_MARSABIT 377 7,905 6,283 KENYA_EASTERN_MBEERE 531 11,445 1,474 KENYA_EASTERN_MERU CENTRAL 1,548 32,503 4,301 KENYA_EASTERN_MERU NORTH 1,875 39,966 5,207 35 Country_Region_District Tuberculosis HIV Malaria KENYA_EASTERN_MOYALE 166 3,501 2,766 KENYA_EASTERN_MWINGI 943 20,689 524 KENYA_EASTERN_NITHI (MERU SOUTH) 638 13,567 1,063 KENYA_EASTERN_THARAKA 313 6,763 522 KENYA_NAIROBI AREA 6,651 355,359 0 KENYA_NAIROBI AREA_NAIROBI AREA 6,651 355,359 0 KENYA_NORTH-EASTERN 2,986 0 9,519 KENYA_NORTH-EASTERN_GARISSA 1,218 0 6,767 KENYA_NORTH-EASTERN_MANDERA 777 0 0 KENYA_NORTH-EASTERN_WAJIR 991 0 2,752 KENYA_NYANZA 13,631 1,140,889 1,382,578 KENYA_NYANZA_BONDO 741 62,180 164,674 KENYA_NYANZA_GUCHA (SOUTH KISII) 1,430 119,614 3,974 KENYA_NYANZA_HOMA BAY 895 75,183 174,115 KENYA_NYANZA_KISII CENTRAL 1,526 127,775 4,239 KENYA_NYANZA_KISUMU 1,565 129,956 260,873 KENYA_NYANZA_KURIA 471 39,293 2,619 KENYA_NYANZA_MIGORI 1,598 133,661 8,877 KENYA_NYANZA_NYAMIRA 1,546 129,148 8,588 KENYA_NYANZA_NYANDO 931 77,588 129,278 KENYA_NYANZA_RACHUONYO 953 80,040 158,855 KENYA_NYANZA_SIAYA 1,490 126,083 372,552 KENYA_NYANZA_SUBA 483 40,368 93,934 KENYA_RIFT VALLEY 21,684 629,428 231,707 KENYA_RIFT VALLEY_BARINGO 822 24,061 1,371 KENYA_RIFT VALLEY_BOMET 1,188 34,850 2,640 KENYA_RIFT VALLEY_BURET 983 28,270 2,732 KENYA_RIFT VALLEY_EAST MARAKWET 436 12,766 1,212 KENYA_RIFT VALLEY_KAJIADO 1,260 36,363 2,800 KENYA_RIFT VALLEY_KEIYO 446 13,012 1,240 KENYA_RIFT VALLEY_KERICHO 1,454 41,975 4,039 KENYA_RIFT VALLEY_KOIBATEK 429 12,448 476 KENYA_RIFT VALLEY_LAIKIPIA 1,000 29,018 0 KENYA_RIFT VALLEY_NAKURU 3,684 106,575 0 KENYA_RIFT VALLEY_NANDI 1,796 52,125 8,981 KENYA_RIFT VALLEY_NAROK 1,135 32,912 2,522 KENYA_RIFT VALLEY_SAMBURU 445 13,108 990 36 Country_Region_District Tuberculosis HIV Malaria KENYA_RIFT VALLEY_TRANS MARA 529 15,474 1,471 KENYA_RIFT VALLEY_TRANS-NZOIA 1,787 51,977 29,775 KENYA_RIFT VALLEY_TURKANA 1,399 40,769 155,467 KENYA_RIFT VALLEY_UASIN GISHU 1,933 55,778 5,368 KENYA_RIFT VALLEY_WEST POKOT 956 27,947 10,624 KENYA_WESTERN 10,424 279,954 1,779,541 KENYA_WESTERN_BUNGOMA 2,720 72,813 453,349 KENYA_WESTERN_BUSIA 1,150 31,030 287,537 KENYA_WESTERN_ELGON 1,480 39,816 246,684 KENYA_WESTERN_KAKAMEGA 1,873 50,249 312,110 KENYA_WESTERN_LUGARI 670 17,931 11,168 KENYA_WESTERN_SHIROTSA/MUMIAS 419 11,177 69,844 KENYA_WESTERN_TESO 563 15,091 140,809 KENYA_WESTERN_VIHIGA 1,548 41,848 258,039 37 38 Appendix E Resources Used for Material Requirements Model Condition Resource Tuberculosis For determining treatments: Treatment of Tuberculosis Guidelines, fourth Edition, WHO (http://whqlibdoc.who.int/publications/2010/9789241547833_eng.pdf) For determining percentage of patients requiring various treatments: Data collected from TB control programs and estimates generated by WHO (http://www.who.int/tb/country/data/download/en/index.html) Pharmaceutical product details: from STOP TB Product list (http://www.stoptb.org/gdf/drugsupply/drugs_available.asp) STIs Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) HIV/AIDS MSFs: Untangling the Web of Antiretroviral Drug Prices: http://utw.msfaccess.org/ UNGASS 2010, country Report: http://data.unaids.org/pub/Report/2010/kenya_2010_country_progress_report_en.pdf Epidemiological Fact Sheet on HIV/AIDS 2008: http://apps.who.int/globalatlas/predefinedReports/EFS2008/full/EFS2008_KE.pdf PowerPoint presentation by WHO 2008 survey on ARV use: http://www.who.int/hiv/amds/who_survey_arv_use_2008_market_renaudthery.pdf “NEW WHO recommendations on ART regimen: Preliminary assumptions on future use of 1st and 2nd line regimen” PowerPoint (Rapid d4T example was used (2012)): http://www.who.int/hiv/amds/who_new_ar_recom_assump_future_renaudthery.pdf WHO Antiretroviral therapy of HIV infection in infants and children: http://www.who.int/hiv/pub/guidelines/paediatric020907.pdf WHO survey: ARV Use in 2008 and market trends in low and middle income countries: http://www.who.int/hiv/amds/who_survey_arv_use_2008_market_renaudthery.pdf Finalization of WHO ART Treatment Guidelines for Children: http://www.who.int/hiv/amds/who_paediatric_art_guidelines_crowley.pdf "Methodology and Assumptions used to estimate the Cost of Scaling Up Selected Child Health Interventions", WHO, March 2005. WHO survey: ARV Use in 2008 and market trends in low and middle income countries” http://www.who.int/hiv/amds/who_survey_arv_use_2008_market_renaudthery.pdf USAID | DELIVER PROJECT, Task Order 1. 2008. Logistics Fact Sheets: ARV Drugs. Arlington, Va.: USAID | DELIVER, PROJECT, Task Order 1. http://deliver.jsi.com/dlvr_content/resources/allpubs/factsheets/LogiFactShee_ARV_Comp.pdf USAID | DELIVER PROJECT, Task Order 1. 2008. Logistics Fact Sheets: HIV Test Kits. Arlington, Va.: USAID | DELIVER PROJECT, Task Order 1. http://deliver.jsi.com/dlvr_content/resources/allpubs/factsheets/LogiFactShee_HIVT_Comp.pdf 39 Condition Resource Diarrhea Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) Internal experts Malaria Guidelines for the Treatment of Malaria, Second edition, WHO, 2010. Leprosy National NTLP guidelines (http://www.nltp.co.ke/docs/National_NLTP_Guideline.pdf) Weekly epidemiological record (http://www.who.int/wer/2009/wer8433.pdf) Worms Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) Respiratory Infections Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) Internal experts "Methodology and Assumptions used to estimate the Cost of Scaling Up Selected Child Health Interventions", WHO, March 2005. Family Planning Profiles for Family Planning and Reproductive Health Programs, 2nd Edition by the Futures Group (http://www.policyproject.com/pubs/generalreport/Profiles116FP2ed.pdf) United Nations, Department of Economic and Social Affairs, Population Division: World Population Prospects DEMOBASE extract. 2007. Vitamins "Methodology and Assumptions used to estimate the Cost of Scaling Up Selected Child Health Interventions", WHO, March 2005. “Constraints to Scaling Up Health Related MDGs: Costing and Financial Gap Analysis:, WHO, September 2009. Diabetes Mellitus “Constraints to Scaling Up Health Related MDGs: Costing and Financial Gap Analysis:, WHO, September 2009. Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) Maternal Health “Constraints to Scaling Up Health Related MDGs: Costing and Financial Gap Analysis:, WHO, September 2009. Hypertension & Cardiovascular Disease “Constraints to Scaling Up Health Related MDGs: Costing and Financial Gap Analysis:, WHO, September 2009. Asthma Clinical Guidelines for Diagnosis and treatment of common Conditions in Kenya, 2002. (http://apps.who.int/medicinedocs/documents/s16427e/s16427e.pdf) Vaccines “Landscape Analysis: Trends in Vaccine Availability and Novel Vaccine Delivery technologies: 2008-2025”, OPTIMIZE, July 2008. Other Resources International Finance Corporation. 2008. Business of Health in Africa. Washington, D.C.: International Finance Corporation. Management Sciences for Health. 2008. International Drug Price Indicator Guide. Arlington, Va.: Management Sciences for Health. 40 Appendix F Material Requirements Model: Treatments Condition Type Treatment Tuberculosis 1st line 2HRZE/4HR; likelihood of isoniazide resistance is low in African countries re-treatment 2HRZES/1HRZE/5HRE MDR-TB Z, Km or Cm, Ofx, Eto, Cs STIs Genital Discharge Norfloxacillin, doxycycline, metronidazole or clotrimazole pesaries Genital ulcer Cotrimoxazole or ceftriaxone Syphilis Benzathine or Procaine Penicillin Injection HIV/AIDS - Adult 1st line d4T/3TC NVP or EFV AZT/3TC NVP or EFV TDF/XTC with NVP or EFV 2nd line LPV/r ATV IDV/r or SQV/r with TDF/XTC AZT/3TC AZT/XTC/TDF ABC/ddI d4T/3TC Note XTC=3TC or FTC HIV/AIDS - Pediatric 1st line d4T/3TC/NVP AZT/3TC/NVP AZT/3TC/EFV d4T/3TC/EFV ABC/3TC/LPV/r 2nd line: ABC/d4T/LPV/r ABC/ddI/LPV/r d4T/3TC/ddI d4T/3TC/LPV/r 41 Condition Type Treatment 3TC/ddI/LPV/r AZT/3TC/LPV/r HIV/AIDS - PMTCT Infants: Breastfeeding: daily NVP from birth until 6 weeks of life Non-breastfeeding: daily AZT and NVP from birth until 6 weeks of age. Mothers (who do not qualify for ART) - option A Ante partum AZT QD, starting at week 14 sd-NVP at onset of labor and delivery HIV/AIDS - PMTCT Mothers (who do not qualify for ART) - option A AZT/3TC during labor and delivery AZT/3TC for 7 days post-partum Mothers (who do not qualify for ART) - option B AZT+ 3TC+ LPV/r AZT+ 3TC+ ABC AZT+ 3TC+ EFV TDF+ XTC+ EFV Note All positive pediatrics and exposed infants until testing confirms negative results: co-trimoxazole All qualifying TB/HIV patients: co-trimoxazole RTKs: serial testing Diarrhea Majority on ORS and Normal saline (if needed) with zinc for pediatrics Antibiotics used: cotrimoxazole, metronidazole, erythromycin and nalidixic acid Malaria Uncomplicated 97% on coartem and 3% on SP, second-line: quinine and doxycycline Severe Artesunate injection and coartem Pregnancy Quinine, clindamycin, coartem and artesunate injection IPTp: SP RDTs ITNs Leprosy Paucibacillary Dapsone and rifampicin Multi-bacillary Dapsone, rifampicin and clofazimine Worms Acariasos Mebendazole or albendazole Trichuriasis Mebendazole or albendazole Hookworm Mebendazole or albendazole Respiratory Infections Pediatrics Outpatient: amoxicillin, Co-trimoxazole benzylpen, procain pen and/or chloramphenicol Inpatient: chloramphenicol, Benzylpenicillin,and/or gentamicin Discharged on: co-trimoxazole, amoxicillin, nebulized salbutamol and/or tablet Adult Outpatient: amoxicillin, erythromycin, benzylpen, procain pen. Inpatient: crystalline penicillin, gentamicin, chloramphenical, and/or erythromycin 42 Condition Type Treatment discharge on: Amoxicillin or erythromycin Family Planning Oral Contraceptives Injectables Family Planning IUD Acceptors Condom (male) Vaginals Vitamins Measles and general pediatric supplementation: Vitamin A supplemented to all children at age 3, 9 and 15 months and every 6 months thereafter to age 5. Diabetes Mellitus Metformin and glibenclamide (glyburide) with 5% requiring insulin Maternal Health Eclampsia and severe pre-eclampsia: hydralazine and magnesium Ante-partem haemorrhage: ergometrine, iron and oxytocin Postpartum haemorrhage, manual removal of placenta and active management of the third stage of labor: oxytocin Hypertensio n and cardivascular disease Cardiovascular disease: hydrochlorothiazide, enalapril, simvastatins, atenolol and calcium channel blocker Congestive heart failure: digoxin, furosemide, enalapril and potassium Asthma Salbuterol and beconase; aminophylline injection for severe asthma Nebulizer treatment every 6 hours for total of 3 days Oral treatment for a total of 5 days Vaccines Basic package: BCG, DPT Hep B, oral polio (OPV), measles and tetanus. Possible future package: BCG or TB, DPT HepB Hib, Measles, seasonal Flu, Cholera, CMV, ETEC, HepA, HepE, HPV, Malaria, MCV-TT, MMR, PCV, Rotavirus RSV, Typhoid, YF, OPV. 43 44 Appendix G Assumptions Material Requirements Model Condition Assumption Tuberculosis Co-trimoxazole therapy is given for all HIV positive patients. 45% of TB patients tested for HIV were positive; increased this to 60% based on comments from advisors and other country examples. 2.02% of TB patients had MDR TB Included all syringes and water for injection that would be required for the injections STIs Syndromic approach used for genital discharge and genital ulcer according to the national guidelines. Treatment for syphilis covers all forms and averages treatment durations. HIV/AIDS - Adults Most patients would be on WHO recommended treatments (for 1st-line) “NEW WHO recommendations on ART regimen: Preliminary assumptions on future use of 1st and 2nd line regimen” PowerPoint (Rapid transition from d4T example was used (2012)): http://www.who.int/hiv/amds/who_new_ar_recom_assump_future_renaudthery.pdf Where possible chosen a fixed-dose combination (FDC) assuming there will be more of a trend towards this. Assumed 80% on NVP and 20% on EFV based on 2006 survey of ART use: http://www.who.int/hiv/amds/ReportDemandForecastforARV2007-2008.pdf HIV/AIDS - Pediatric Taken out all d4T regimens and substitute with AZT. Combine all regimens for one total list of all 1st and 2nd line regimens. Overestimated needs for the 5-14 year old group, at times,to adult doses. Daily zinc (10mg) to prevent diarrhea in HIV positive pediatrics. HIV/AIDS - PMTCT Pregnant women that qualify for ART are covered under adult ART. Two options for PMTCT given in new WHO guidelines: Breastfeeding was assumed to be at 50% for option A and 50% for option B. Duration was assumed to be for 12 months (12 months from WHO document: HIV and infant feeding Revised Principles and Recommendations Rapid Advice, November 2009) There are four options for triple therapy in option B of PMTCT-mothers treatment. Choose 25% for each. HIV/AIDS - Cotrimoxazole Half the children between 0-4 would be taking syrup and the rest on tablets. Children 5-14 would be 25% on child dose and the rest on 400+80. TB/HIV adults should be placed on daily Cotrimoxazole preventative therapy. Diarrhea Average weight of 12 kg for children 0-4years old Average weight of 25kg for children 5-8 years old 45 Condition Assumption Average weight of 35kg for children 8-14 years old Average weight of 60 kg for adults Assumed 20% of children between 5-14 will be on solutions and the other 80% on tablets/capsules For higher weight pediatric patients, did not go over the adult recommended dosage. For pediatric patients, rounded a treatment up to the nearest full bottle, since remaining medication cannot be stored once reconstituted. Malaria Coartem was used as a placeholder for the packaging of any ACT Leprosy Assume 50% un-supervised multi-bacillary leprosy treatments No treatments required for children under 15 years old Worms Within each age group the prevalence or incidence rate of disease is constant. Worms Mebendazole and Albendazole treatments are both recommended. Assume 50% of 0-4 year olds are 0-2 years old Assume all 0-4 year olds are taking solutions Assume 20% of 5-15 year olds are taking solutions. Respiratory Infections 30% of men would need hospitalization 5% of women would need hospitalization 50% of children would need hospitalization 3 days average hospital stay for adults and children, then discharged with oral meds for 10 days for a total of 13 days Of those who are not hospitalized, 50% receive an injection for the first dose of antibiotics and oral meds for a total of 10 days 5% of adults have penicillin allergy Assumed 20% of children between 5-14 will be on solutions and the other 80% on tablets/capsules Patients 5 years and older can take tablets/capsules Used average weight of 12 kg for children 0-4 years old Used average weight of 30 kg for children 5-14 years old Family Planning Did not account for requirements needed for female and male sterilization as these medications will be covered in the general bundle of essential medicines. Diabetes Mellitus 10ml vial of insulin is most likely used, with one vial per month as an average. Three injections per day for a 100/month Maternal Health All general essential medicines like IV solutions, lignocaine, diazepam and antibiotics have been taken out and will be assumed to be included as part of the essential medicines bundle. Asthma Epinephrine and prednisone would be included as part of the essential medicines bundle Average weight of 12 kg for children 0-4 years old Average weight of 25 kg for children 5-8 years old Average weight of 35 kg for children 8-14 years old Loading dose of Aminophylline=6mg/kg 46 Appendix H Treatment Rates by Condition Disease Treatment Rate 2010 Treatment Rate 2020­ 2024 HIV/AIDS (first line/second line) 12.15% 49% / 1% Malaria 34% 34% Tuberculosis 85% 85% Vaccines 85% 85% Vitamin A 15% 85% Diarrhea 20% 20% Leprosy 20% 85% Worms 35% 85% Respiratory Infections 35% 35% Syphilis 10% 85% Other STI 10% 85% Measles 20% 20% Diabetes 5% 5% Hypertension & cardiovascular disease 10% 10% Asthma 5% 5% Family Planning 39% 100% Maternal Health 52% 100% Other Essential Meds bundle 100% 100% 47 48 Appendix I Vehicle Characteristics KEMSA Owned Fleet Vehicle Type 1 Ton vehicle 3.5 Ton vehicle 7 Ton vehicle Notes Example Toyota Tundra /Tacoma Nissan UD 40 /Mitsubishi Canter Nissan UD 70 /Isuzu Van Maximum payload (kg) 1,085 3,150 6,300 1 ton = 900 kg Average weight per trip (kg) 543 1,575 5,670 50 or 90% of max payload Purchase cost (USD) $33,000.00 $47,500.00 $68,500.00 US Numbers from JSI Fixed Cost (10% Depreciation) $4,800.00 $9,450.00 $10,450.00 Maintenance / repair + 10% depreciation Fixed Cost (25% Depreciation) $9,750.00 $16,575.00 $20,725.00 Maintenance / repair + 25% depreciation Maintenance /Repair $1,500.00 $4,700.00 $3,600.00 Based on Malawi historical data Depreciation (10%) $3,300.00 $4,750.00 $6,850.00 10% of purchase cost Depreciation (25%) $8,250.00 $11,875.00 $17,125.00 25% of purchase cost Per Distance Cost (fuel) $0.32 $0.40 $0.49 Based on Malawi historical data Average speed (km/hr.) 40 50 60 Estimates 49 50 Appendix J Outputs Generated by Models and Scenarios Output Cost / Metric Name Description Total Transportation Cost For 3PL: total cost incurred for transportation based on per cubic meter per kilometer rate For KEMSA owned fleet: total cost incurred for transportation. This includes purchase of new vehicles, fixed costs for vehicles (maintenance, repair, depreciation), fuel costs (per kilometer), and labor costs for vehicle drivers Total Facility Cost This is the cost to operate the warehouse(s), which includes rent and electricity. Total Administrative Cost The total cost to order commodities, it is a per commodity unit cost that is applied. Total Warehousing Cost Total Supply Chain Cost The cost of labor for warehouse employees shipping and receiving commodities Sum of transportation, facility, administrative, and warehousing costs Total Safety Stock Investment The total investment needed to buy commodities to keep stocks are the recommended safety stock level, based on the fill-rate service requirement Total Procurement Cost The total investment needed to buy the needed commodities (price of commodities * # commodities) Delivered Value The total value of goods flowing through the supply chain (equals total procurement cost unless commodities are donated) Delivered Value per Person Delivered Value / country’s population Supply Chain Cost Rate (%) Total Supply Chain Cost / Delivered Value (in %) 51 52 For more information, please visit deliver.jsi.com. USAID | DELIVER PROJECT John Snow, Inc. 1616 Fort Myer Drive, 11th Floor Arlington, VA 22209 USA Phone: 703-528-7474 Fax: 703-528-7480 Email: askdeliver@jsi.com Internet: deliver.jsi.com

View the publication