Independent Evaluation of Phase 1 of the Affordable Medicines Facility malaria (AMFm) KENYA. Endline Outlet Survey and Context Description.

Transcription

1 Independent Evaluation of Phase 1 of the Affordable Medicines Facility malaria (AMFm) KENYA Endline Outlet Survey and Context Description July 2012 Division of Malaria Control Andrew Nyandigisi Dorothy Naisiae Agnetta Mbithi Pharmacy and Poisons Board Stephen Kimatu Contributors: African Population and PSI/Kenya ACTwatch Group Independent Health Research Centre Evaluation Team Kanyiva Nuindi Edna Ogada Marilyn Wamukoya Mbogo Bunyi Frederick Wekesah Blessing Mberu Yohannes Kinfu Kathryn O Connell Hellen Gatakaa Stephen Poyer Illah Evance Julius Ngigi Mitsuru Toda Meghan Bruce Tanya Shewchuk Catherine Goodman Sarah Tougher Yazoume Ye Andrea Mann Barbara Willey Ruilin Ren Fred Arnold Kara Hanson Copyright 2012 Population Services International (PSI). All rights reserved

2 Suggested citation: ACTwatch Group, African Population and Health Research Centre (APHRC), PSI/Kenya, Kenya Division of Malaria Control, Kenya Pharmacy and Poison s Board, and the Independent Evaluation Team. (2012) Endline Outlet Survey Report for the Independent Evaluation of Phase 1 of the Affordable Medicines Facility malaria (AMFm): Kenya. Nairobi, Kenya: ACTwatch/APHRC/ PSI/Kenya. i P age Copyright 2012 Population Services International (PSI). All rights reserved

3 Table of Contents Table of Contents... ii List of Tables... iv List of Figures... vii Definitions... VIII List of Abbreviations... X Executive Summary... XII Overview of the independent evaluation process... XII Endline outlet survey methods... XIII Key findings from the outlet survey... XV Key findings on AMFm implementation: process and key contextual factors... XXIII 1. Background Overview of the AMFm phase Overview of the AMFm Phase 1 Independent Evaluation (IE) Country background context Overview of the country Description of health care system Epidemiology of malaria Antimalarial Policies and Regulatory Environment Malaria control strategy Malaria financing Methods Outlet survey Outlet survey indicators Sampling Approach Data collection Data processing Data analysis Results Outlet survey Characteristics of the sample Availability of antimalarial drugs Antimalarials in stock Antimalarials in stock by type Stockouts of quality assured ACTs Population coverage of outlets with quality assured ACTs Pricing of antimalarials (Affordability) Cost to patients of antimalarials Gross percentage markup between purchase price and retail selling price Availability and cost to patients of diagnostic tests (RDT/microscopy) Quality assured ACTs market share Provider knowledge of first line antimalarial treatment and ACT regimen AMFm logo ii P age

4 4. AMFm implementation: process and key contextual factors Introduction Methods Findings AMFm implementation process Implementation of AMFm supporting interventions Summary of findings Quality of data collected Availability of quality assured ACTs Pricing/affordability of quality assured ACTs Market share of quality assured ACTs References Acknowledgements Appendices Questionnaire ACTs classified as quality assured Final sample Survey team Description of outlet types visited for this survey Sampling weights Assumptions for calculating Adult Equivalent Treatment Doses (AETDs) iii P age

5 List of Tables Table : Primary indicators Table : Outlets enumerated by location, drugs stocked and final interview status [Kenya], Table : Outlets enumerated [Kenya], Table : Outlets with antimalarials in stock* [Kenya], 2011 ** Table : Number of products audited [Kenya], Table : Outlets with at least one staff member who completed secondary school or primary school [Kenya] Table : Outlets with a staff member with a health related qualification [Kenya], Table : Outlets with antimalarials in stock in [Kenya], Table : Outlets with non artemisinin therapy in stock Kenya, Table a: Outlets with artemisinin monotherapy in stock (ALL DOSAGE FORMS ) Kenya, Table b: Outlets with ORAL artemisinin monotherapy in stock Kenya, Table : Outlets with non quality assured ACTs in stock Kenya, Table c: Public health facility outlets with quality assured ACTs among ALL PUBLIC HEALTH FACILITIES in Kenya, Table : Outlets with stock outs of quality assured ACTs [Kenya], Table : of the population living in censused locations with outlets with qualityassured ACTs in stock at the time of survey [Kenya], Table : Cost to patients of non artemisinin therapy, in US dollars Kenya, Table : Cost to patients of artemisinin monotherapy, in US dollars Kenya, Table : Cost to patients of non quality assured ACTs, in US dollars Kenya, Table : Cost to patients of quality assured ACTs, in US dollars Kenya, Table a : Cost to patients of the most popular antimalarial in terms of national private forprofit outlet sales volumes (SP) for ALL DOSAGE TYPES, in US dollars Kenya, Table b: Cost to patients of the most popular antimalarial in terms of national private forprofit outlet sales volumes (SP) for TABLETS, in US dollars Kenya, Table c: Cost to patients of the most popular antimalarial in terms of national outlet sales volumes for all outlet types (SP) for ALL DOSAGE TYPES, in US dollars Kenya, Table d: Cost to patients of the most popular antimalarial in terms of national outlet sales volumes for all outlet types (SP) for TABLETS, in US dollars Kenya, Table : Gross percentage markup between purchase price and retail selling price of nonartemisinin therapy [Kenya], Table : Gross percentage markup between purchase price and retail selling price of artemisinin monotherapy [Kenya], Table : Gross percentage markup between purchase price and retail selling price of nonquality assured ACTs [Kenya], iv P age

6 Table a: Gross percentage markup between purchase price and retail selling price of qualityassured ACTs [Kenya], Table b: Gross percentage markup between purchase price and retail selling price of qualityassured ACTs, by presence of the AMFm logo, [Kenya], Table : Median total gross markup between first line buyer price and retail selling price of quality assured ACTs bearing the AMFm logo, in 2010 US dollars, [Kenya], Table : Availability of any diagnostic test for malaria, [Kenya], Table : Availability of malaria microscopy, [Kenya], Table : Cost to patients of malaria microscopy in 2010 US dollars [Kenya], Table : Availability of rapid diagnostic tests for malaria, [Kenya], Table : Cost to patients of rapid diagnostic tests (RDTs) for malaria in 2010 US dollars [Kenya], Table : breakdown of antimalarial sales volumes by antimalarial type, Kenya, Table : Market share of quality assured ACTs (QAACTs), for all QAACTs, QAACTs with the AMFm logo and QAACTs without the AMFm logo, Kenya, Table : breakdown of antimalarial sales volumes by outlet type, Kenya, Table : Provider knowledge of first line antimalarial treatment, [Kenya], Table : Provider knowledge of dosing regimen for quality assured ACTs (QAACTs) for an adult. [Kenya], Table : Provider knowledge of dosing regimen for quality assured ACTs (QAACTs) for a child, [Kenya], Table : Reasons for not stocking quality assured ACTs (QAACTs) by private providers, [Kenya], Table : Provider recognition of AMFm logo, [Kenya], Table : Provider knowledge of the AMFm logo, [Kenya], Table : Sources from which providers have seen or heard of the AMFm logo, [Kenya], Table : of antimalarials bearing the AMFm logo, [Kenya], Table : Provider knowledge of the AMFm programme, [Kenya], Table : Sources from which providers have seen or heard of AMFm, [Kenya], Table : Provider stating that there is a maximum/recommended retail price (RRP) for antimalarials with the AMFm logo, [Kenya], Table : Provider stating the correct maximum/recommended retail price (RRP) for antimalarials with the AMFm logo, [Kenya], Table : Providers who have received training on antimalarials with the AMFm logo, [Kenya], Table : Summary of key factors likely to have supported or hindered achievement of AMFm goals in Kenya v P age

7 Table : List of Quality Assured ACTs for availability, price and market share indicators Table : List of clusters/sub districts sampled and their population, [Kenya], Table : List of APHRC staff members involved in the survey, [Kenya ], Table : AETD Calculation details by antimalarial type vi P age

8 List of Figures Figure 1. Availability of antimalarials among all outlets, by outlet type... XVI Figure 2. Outlet types stocking antimalarials...xvii Figure 3. Availability of antimalarials, among outlets stocking at least one antimalarial, by outlet type... XVIII Figure 4. Proportion of outlets with microscopic blood testing facilities and rapid diagnostic tests XIX Figure 5. Median price of antimalarial treatment per AETD in the private sector, by outlet type... XX Figure 6. Market share of AETDs sold/distributed in the past week (7 days) within outlet types... XXI Figure 7. Market share of AETD sold/distributed in the past week (7 days) across outlet types...xxii Figure 8. Provider knowledge of recommended first line treatment and dosing regimens... XXIII Figure : AMFm Phase 1 Results Framework... 3 Figure : The Independent Evaluation Impact model... 4 Figure : The Independent Evaluation Design... 4 Figure : Location of Kenya... 7 Figure : Public health delivery system, Kenya... 8 Figure : Distribution of Health Facilities by type and Controlling Agency, Figure : Malaria transmission risk map [Transmission zones], [2009], [Kenya] Figure : Survey flow diagram, [Kenya], Figure : Timeline of key events related to AMFm implementation process and context in Kenya vii P age

9 Definitions Adult Equivalent Treatment Dose (AETD) Antimalarial Artemisinin based Combination Therapy (ACT) Artemisinin monotherapy Booster Sample Censused location Child Equivalent Treatment Dose (CETD) Combination therapy Dosing/treatment regimen Enumerated Outlets First line treatment Monotherapy Non artemisinin therapy Outlet Pediatric formulation An AETD is the number of milligrams (mg) of an antimalarial drug needed to treat a 60 kg adult. Any medicine recognized by the WHO for the treatment of malaria. Medicines used solely for the prevention of malaria were excluded from analysis of key indicators in this report. An antimalarial that combines artemisinin or one of its derivatives with an antimalarial or antimalarials of a different class. Refer to Combination Therapy (below). An antimalarial medicine that has a single active compound, where this active compound is artemisinin or one of its derivatives. A booster sample is an extra sample of units (or in this case outlets) of a type not adequately represented in the main survey, but which are of special interest. In this survey, we have included a booster sample of public health facilities and part one pharmacies in the entire division that includes the selected location, consisting of all of the public health facilities and part one pharmacies in the district that are not in the selected location. A location where field teams conducted a full census of all outlets with the potential to sell antimalarials. A CETD is the number of milligrams (mg) of an antimalarial drug needed to treat a 10 kg child. The use of two or more classes of antimalarial drugs/molecules in the treatment of malaria that have independent modes of action. The posology or timing and number of doses of an antimalarial used to treat malaria. This schedule often varies by patient weight. Outlets that were visited by a member of one of the field teams, and at a minimum basic descriptive information was collected (sections C1 C9 of the outlet survey questionnaire). The government recommended treatment for uncomplicated malaria. Kenya s first line treatment is Artemether Lumefantrine. An antimalarial medicine that has a single mode of action. This may be a medicine with a single active compound or a synergistic combination of two compounds with related mechanisms of action. An antimalarial medicine that does not contain artemisinin or any of its derivatives. Any point of sale or provision of a commodity to an individual. Outlets are not restricted to stationary points of sale and may include mobile units or individuals. Refer to Appendix 8.5 for a description of the outlet types visited for this survey. Antimalarial drug packaged specifically for children. VIII P age

10 Quality assured Artemisinin Based Combination Therapies (QAACTs) QAACTs are ACTs that comply with the Global Fund to Fight AIDS, Tuberculosis and Malaria s Quality Assurance Policy. For the purpose of the Independent Evaluation, a QAACT is any ACT which appeared on the Global Fund's indicative list of antimalarials meeting the Global Fund's quality assurance policy prior to baseline or endline data collection (see ), or which previously had C status in an earlier Global Fund quality assurance policy and was used in a program supplying subsidized ACTs. Rapid Diagnostic Test (RDT) for malaria Screened Screening criteria Sub district (SD) At endline, QAACTs were defined as any ACT which appeared on the Global Fund s indicative list of antimalarials meeting its quality assurance policy as at September 2011, or which previously had C status in an earlier Global Fund quality assurance policy and was used in a program supplying subsidized ACTs. Malaria rapid diagnostic tests, sometimes called "dipsticks" or malaria rapid diagnostic devices, assist in the diagnosis of malaria by providing evidence of the presence of malaria parasites in human blood. RDTs do not require laboratory equipment, and can be performed and interpreted by non clinical staff. An outlet that was administered the screening questions (S1 to S4) of the outlet survey questionnaire (see Screening criteria). The set of requirements that must be satisfied before the full questionnaire is administered. In this survey, an outlet met the screening criteria if (1) they had antimalarials in stock at the time of the survey visit, or (2) they report having stocked them in the past three months. The primary sampling unit, or cluster, for the outlet survey. It is an administrative unit determined by the Ministry of Health (MOH) that host a population size of approximately 10,000 to 15,000 inhabitants. These units frequently are defined by geographical, health, or political boundaries, and are based around wards. In Kenya they were defined as locations, and this served as the primary sampling unit. IX P age

11 List of Abbreviations ACT AETD AL AM AMT AMFm APHRC API ASAQ CETD CFW CHW CI CIA CIERPA CRDH DCs DfID DHS DNDi DOMC FBO GDP GPS HAI HDI IE IEC IMF IPT IQR IRS ITN JICA KCM KDHS KEMSA kg KI KNMS KNBS KSH Artemisinin based Combination Therapy Adult Equivalent Treatment Dose Artemether lumefantrine Antimalarial Artemisinin monotherapy Affordable Medicines Facility malaria African Population and Health Research Centre Active Pharmaceutical Ingredient Artesunate and Amodiaquine Child Equivalent Treatment Dose Community and Family Wellness Community Health Worker Confidence Interval Central Intelligence Agency Centre International d'études et de Recherches sur les Populations Africaines Centre de Recherche pour le Développement Humain Data Contributors Department for International Development Demographic and Health Surveys Drugs for Neglected Diseases initiative Division of Malaria Control Faith based Organization Gross domestic product Global Positioning System Health Action International Human Development Index Independent Evaluation/Evaluator Information, Education and Communication International Monetary Fund Intermittent Preventive Treatment Interquartile Range Indoor Residual Spraying Insecticide Treated Net Japan International Cooperation Agency Kenya Country Mechanism Kenya Demographic and Health Survey Kenya Medical Suppliers Agency Kilogram Key Informant Kenya National Medical Services Kenya National Bureau of Statistics Kenya Shillings X P age

12 LLIN LSHTM MCH MEDS Mg MICC MICS MIS MOH MOMS NAFDAC nat NGO NMCP NOC nqaact OTC PHF PMI POM POP PSI PPB PPS PV QAACT RDT RRP SD SI SP SOP TWG UN UNDP US USD USP WFB WHO/AFRO WHO Long lasting Insecticidal Net London School of Hygiene and Tropical Medicine Maternal and Child Health Mission for Essential Drugs and Supplies Milligrams Malaria Interagency Coordinating Committee Multiple Indicator Cluster Survey Malaria Indicator Survey Ministry of Health Ministry of Medical Services National Agency for Food and Drug Administration and Control Non artemisinin Therapy Non governmental organization National Malaria Control Program National Oversight Committee Non Quality assured Artemisinin based Combination Therapy Over the Counter Public Health Facility President s Malaria Initiative Prescription only Medicine Part one Pharmacy Population Services International Poisons and Pharmacy Board Probability proportional to size Pharmacovigilance Quality assured Artemisinin based Combination Therapy Rapid Diagnostic Test Recommended Retail Price Sub district Supporting Interventions Sulfadoxine pyrimethamine Standard Operating Procedures Technical Working Group United Nations United Nations Development Program United States United States Dollar US Pharmacopeia World Fact Book World Health Organization/Africa region World Health Organization XI P age

13 Executive Summary Overview of the independent evaluation process The independent evaluation is part of a multi faceted monitoring and evaluation framework developed for Phase 1 of the Affordable Medicines Facility malaria (AMFm). It is intended to assess whether, and to what extent, AMFm Phase 1 achieves its objectives. The findings of the independent evaluation will be summarized in a report to be considered by the Global Fund Board at the end of Phase 1. The four main objectives of AMFm are: (i) to increase ACT affordability, (ii) to increase ACT availability, (iii) to increase ACT use, including among vulnerable groups, and (iv) to crowd out other oral antimalarials by gaining market share. Through a competitive bid, the Global Fund contracted ICF Macro and the London School of Hygiene and Tropical Medicine (LSHTM) to conduct the IE. The IE was carried out in all of the currently operational Phase 1 pilots (Ghana, Kenya, Madagascar, Niger, Nigeria, Tanzania (mainland and Zanzibar), and Uganda.). In addition, the Global Fund contracted with Data Contributors (DCs) that were responsible for in country fieldwork, data analysis and country reports. These institutions are Population Services International (PSI), Drugs for Neglected Diseases initiative (DNDi), and Centre de Recherche pour le Développement Humain (CRDH). PSI's ACTwatch Project 1 was responsible for the work in Kenya (which was subcontracted to the African Population and Health Research Centre), Madagascar, Nigeria, Uganda, Tanzania mainland (which was subcontracted to the Ifakara Health Institute) and Zanzibar. The ACTwatch Project also contributed evidence for the baseline work in Nigeria and Madagascar, which was conducted prior to the IE surveys. DNDi subcontracted with the Research and Development Unit, Komfo Anokye Teaching Hospital, Kumasi, to undertake the work in Ghana. CRDH subcontracted with the Centre International d'etudes et de Recherches sur les Populations Africaines (CIERPA) to undertake the work in Niger. The IE is based on a non experimental design with a pre and post test intervention assessment in which each participating country is treated independently as a case study. In addition to measuring the changes in key indicators pre and post intervention, the evaluation includes an assessment of the implementation process and a comprehensive documentation of the context both to inform assessments about causality and to aid in generalizability to other contexts. The current report is based on the endline assessment in Kenya conducted by ACTwatch and the African Population and Health Research Centre. The results of the baseline survey can be found in the Kenya baseline report (PSI/Kenya, the ACTwatch Group and the Independent Evaluation Team, 2011), and for all pilots in the Multi Country Baseline Report (Independent Evaluation Team, 2011). Analysis of changes between baseline and endline outlet surveys will be presented in the Multi Country Endline Report (forthcoming), together with the data the IE team has compiled from national household surveys. In 1 ACTwatch is a five year multi country research project designed to provide a comprehensive picture of the anti malarial market to inform national and international anti malarial drug policy decision making (Shewchuk et al., 2011). The project is being conducted in seven malaria endemic countries: Benin, Cambodia, the Democratic Republic of Congo, Madagascar, Nigeria, Uganda and Zambia from 2008 to ACTwatch measures which antimalarials are available, where they are available and at what price and who they are used by. These indicators are measured over time and across countries through three study components: outlet surveys, supply chain studies and household surveys. Nationally representative outlet surveys examine the market share of different antimalarials passing through public facilities and private retail outlets (O Connell et al., 2011). Supply chain research provides a picture of the supply chain serving drug outlets, and measures mark ups at each supply chain level. On the demand side, nationally representative household surveys capture treatment seeking patterns and use of anti malarial drugs, as well as respondent knowledge of antimalarials (Littrell et al., 2011). For further information see. XII P age

14 addition country case studies on context/process were conducted by the IE, and these case studies are summarized in the present report. Endline outlet survey methods Endline outlet surveys were conducted in all pilots between September and December The endline surveys in Madagascar, Nigeria and Uganda were undertaken through the ACTwatch project using methods that are compatible with the IE surveys. The IE used a sampling approached based on that used in ACTwatch. (O Connell et al., 2011) A cluster sampling approach was used because there were no reliable lists of all outlets stocking antimalarials. Clusters were locations, with an average of 10,000 to 15,000 inhabitants. In Kenya, 57 locations were selected with probability proportional to size (PPS) a sampling technique in which the probability that a particular location is selected is proportional to its population size. The sample size was powered to detect a change of 20% percentage points in availability of quality assured ACTs between baseline and endline in rural and urban areas. The inclusion criteria for this study were outlets that stocked an antimalarial at the time of survey or had stocked antimalarials in the previous three months. An outlet is defined as any point of sale or provision of commodities for individuals. Outlets included in the survey were: 1) public health facilities (national referral hospitals, provincial general hospitals, health centres or sub health centres, and dispensaries); 2) private not for profit health facilities (hospitals, clinics, and dispensaries operated by non governmental, community based, or religious organizations); 3) community health workers (community health workers, traditional birth attendants and home based care); 4) private for profit health facilities (private hospitals, clinics and dispensaries, and midwives); 5) registered pharmacies (registered pharmacies licensed by the Pharmacy and Poisons Board); 6) drug stores/chemists (small businesses that are not registered with the Pharmacy and Poisons Board); 7) general retailers (supermarkets, duka, kiosks, market stalls, and petrol stations); and 8) itinerant drug vendors (hawkers, which are itinerant salesperson without a physical location). Refer to the Appendix for definitions of each type of outlet included in the analysis. A structured questionnaire was developed, based on the ACTwatch questionnaire. Fieldworkers recorded the outlets basic details and then asked a screening question about the availability of antimalarials to decide whether to proceed with the full interview or not. The questionnaire was administered to a senior person at the outlet to collect data on outlet identification, outlet characteristics, provider knowledge, antimalarials and rapid diagnostic tests (RDTs) stocked, and stockouts of quality assured ACTs. They recorded information on audit sheets on all antimalarials and RDT products stocked in terms of their price and volume sold in the past week. Data quality control tools used in the field were based on those developed by ACTwatch. All data were double entered and verified by APHRC using an Access database developed by the ACTwatch project. To ensure a high level of data quality, APHRC undertook data cleaning using a detailed guideline provided by the IE team, and guidance from ACTwatch. For the analysis, the Independent Evaluators provided a tabulation plan for all tables presented in this report, and analysis do files in Stata, which produced all the required indicators and XIII P age

15 automatically generated the tables. APHRC, in collaboration with ACTwatch, adapted these analysis files to the country setting and ran the analysis using STATA version 11, recording results in a log file. XIV P age

16 Key findings from the outlet survey Data were collected between 7th of November 2011 and the 7th of January A total of 13,376 outlets were approached. Of these, 1990 outlets were not screened for various reasons: 46 providers refused to be interviewed; 1430 outlets were closed down permanently; 408 outlets were not open at the time of the survey visit; in 86 outlets, providers were not available for interview at the time of survey visit; 20 providers were unable to be interviewed for other reasons. Overall, 11,368 outlets agreed to participate in the ACTwatch outlet survey and were screened. Of these, 2,112 outlets met our screening criteria; however, interviews could not be conducted for 24 outlets. Of the 2,088 interviews conducted, 232 reported having stocked antimalarials at any point in the three months prior to the interview and 1,856 outlets stocked antimalarials at the time of the interview. XV P age

17 AVAILABILITY OF ANY ANTIMALARIAL: Stocking rates of any antimalarial varied by outlet type. In the public/not for profit sector, 97% of public health facilities and 94% of private not for profit facilities had at least one antimalarial in stock on the day of interview. No community health workers stocked antimalarials on the day of survey. In the private for profit sector more than 90% of private for profit facilities, pharmacies and drugs stores stocked antimalarials. This is in contrast to 4.4% of general retailers (Figure 1). Figure 1. Availability of antimalarials among all outlets, by outlet type XVI P age

18 OUTLET TYPES STOCKING ANTIMALARIALS: Figure 2 shows the relative distribution of all outlets that had at least one antimalarial in stock. Drug stores were the most common type of outlet stocking antimalarials (78%), followed by private health facilities (7%) and public health facilities (5%). Figure 2. Outlet types stocking antimalarials XVII P age

19 AVAILABILITY OF DIFFERENT CLASSES OF ANTIMALARIALS: Among facilities that stocked antimalarials at any time in the three months preceding the survey, overall QAACT availability in 2011 was 66%. This was lower in the private for profit sector (60%) than in the public not for profit sector (97%). However, there was some variation within the private for profit sector. QAACT availability was higher in pharmacies (92%), drug stores (87%), and private for profit facilities (74%), than general retailers (24%). For private for profit sector, QAACT availability was higher in urban areas than in rural areas (71% versus 55%). In the public and private not for profit outlets, availability of QAACTs with and without the AMFm logo was very similar, but in the private for profit sector, availability of QAACTs with the logo was 59% compared with only 5% for QAACTs without the logo (data not shown). Figure 3. Availability of antimalarials, among outlets stocking at least one antimalarial, by outlet type XVIII P age

20 AVAILABILITY OF DIAGNOSTIC BLOOD TESTING: Of outlets stocking antimalarials in the last three months, 79.4% of private not for profit facilities, 56% of public health facilities and 46% of private for profit health facilities offered any test services. Microscopic blood testing was more widely available than RDTs. Availability of RDT services was less than 20% across all outlet types. Figure 4. Proportion of outlets with microscopic blood testing facilities and rapid diagnostic tests XIX P age

21 PRICE OF ANTIMALARIALS: In the public and private not for profit outlets, the median QAACT price was USD0.00, reflecting the policy of free ACT provision. Pooling all sectors, the median price was USD 0.46 (data not shown). The median price of QAACTs in the private for profit sector was USD 0.58, although still somewhat higher than the RRP of USD Prices remained slightly higher in urban areas (USD 0.61 versus USD 0.46 in rural areas) (data not shown). The median price for a QAACT with the AMFm logo was USD 0.46 per AETD overall (exactly equal to the RRP) (data not shown). In the private for profit sector the price of a QAACT with the AMFm logo was USD The median price of a QAACT with the AMFm logo among private for profit outlets is exactly equal to the median price of the most popular antimalarial which is not a QAACT (SP) in private for profit outlets, whether this is measured in tablet form or among all dosage types. The price of non quality assured ACTs in the private for profit sector was USD Figure 5. Median price of antimalarial treatment per AETD in the private sector, by outlet type XX P age

22 VOLUMES OF ANTIMALARIALS SOLD/DISTRIBUTED: Market share of QAACTs was 57% of all antimalarials sold/distributed in the week preceding the survey, with nats accounting for 38% of the overall market share. Overall market share of oral AMTs was negligible (0.05%). Market share of QAACTs was highest in the private for profit sector (60%), as compared with the private not for profit outlets (44.4%) and public health facility outlets (35%). In the private not for profit health facilities, 37% of the markets share were non QAACTs. QAACTs with the AMFm logo accounted for 88% of all QAACTS dispensed overall and 97% of all QAACTs dispensed in the private for profit sector (data not shown). Figure 6. Market share of AETDs sold/distributed in the past week (7 days) within outlet types XXI P age

23 The private for profit sector was responsible for 62% of all antimalarials sold or distributed in 2011, accounting for 89% and 52% in urban and rural areas, respectively. Figure 7. Market share of AETD sold/distributed in the past week (7 days) across outlet types XXII P age

24 PROVIDER KNOWLEDGE: Overall, 71% of providers were able to correctly state AL as the recommended first line treatment for uncomplicated malaria in Kenya. Knowledge was higher among providers at public/not for profit outlets, compared to the private sector (96% vs. 66% respectively). Providers were more likely to be able to correctly state the dosing regimen of AL for a child versus an adult (data not shown). Figure 8. Provider knowledge of recommended first line treatment and dosing regimens Key findings on AMFm implementation: process and key contextual factors AMFm implementation: Seven private sector FLBs registered and established relationships with manufacturers, of which six had placed orders by the end of The FLB for the public sector was the Kenya Medical Supplies Agency (KEMSA). The first orders for copaid QAACTs were placed in July 2010 by a private for profit FLB and delivered in August A total of 15 months elapsed between the date the first drugs arrived in Kenya and the midpoint of endline outlet survey fieldwork. Supporting interventions mainly started in February 2011, giving nine months of effective SI implementation. The supporting interventions included a communication campaign, training of private sector health workers, pharmacovigilance activities and a recommended retail price set at USD 0.46 for all pack sizes. A total of 28.4 million copaid QAACT treatments were delivered between July 2010 and December 2011 (0.9 treatments per person at risk of malaria), half of which were delivered to the private for profit sector. The application of the Global Fund s demand levers in XXIII P age

25 Kenya resulted in only 56% of treatments requested by FLBs in the second half of 2011 being approved. Context: A predicted malaria epidemic led to an emergency response, although the epidemic did not arise. Mass distribution of LLINs took place. There was depreciation of the Kenya shilling. Political support for AMFm was high. ACTs did not have over the counter status. XXIV P age

26 1. Background 1.1 Overview of the AMFm phase 1 The success of malaria control efforts depends on a high level of coverage and use of effective antimalarials, such as artemisinin based combination therapies (ACTs). Although these antimalarials have been procured in large amounts by countries, evidence suggests that ACT use still remains far below target levels. Reasons suggested for the low uptake of ACTs include interruptions in public sector supply; limited availability outside major urban centers; the high prices of the drugs, particularly in the private sector; lack of provider adherence to new recommendations; and patient self treatment with other more common and cheaper antimalarials. Lowering the cost of ACTs to the end user through a subsidy mechanism could be an effective way to increase their uptake. In response to these issues, the Affordable Medicines Facility malaria (AMFm) hosted by The Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM, or the Global Fund) was set up. As described by Adeyi and Atun, AMFm is a financing mechanism designed to incorporate three elements: (1) price reductions through negotiations with manufacturers of ACTs; (2) a buyer subsidy, via a co payment at the top of the global supply chain by AMFm on behalf of eligible buyers from the public, private for profit and private not for profit sectors; and (3) support for interventions to promote appropriate use of ACTs. Examples of these supporting interventions include training providers and outreach to communities to promote ACT utilization. AMFm is being tested in a first phase that includes 9 pilots in 8 countries: Cambodia, Ghana, Kenya, Madagascar, Niger, Nigeria, Tanzania (mainland and Zanzibar) and Uganda. It is expected that in the last quarter of 2012, the Global Fund Board will make a decision regarding the future of the AMFm on the basis of evidence gathered during Phase 1 regarding progress toward achieving its four stated objectives: (i) increased ACT affordability, (ii) increased ACT availability, (iii) increased ACT use, including among vulnerable groups, and (iv) crowding out oral artemisinin monotherapies, chloroquine and sulfadoxine pyrimethamine (SP) by gaining market share. The AMFm Phase 1 Independent Evaluation (IE) has been commissioned to address the need for evidence on which to base the Global Fund Board decision. 1.2 Overview of the AMFm Phase 1 Independent Evaluation (IE) Through a competitive bid, the Global Fund contracted ICF Macro and the London School of Hygiene and Tropical Medicine (LSHTM) to conduct the IE. The IE was carried out in all of the currently operational Phase 1 pilots (Ghana, Kenya, Madagascar, Niger, Nigeria, Tanzania (mainland and Zanzibar) and Uganda. 2 In addition, the Global Fund contracted Data Contributors (DCs) to be responsible for in country fieldwork, data analysis and country reports. These institutions are Population Services International (PSI), Drugs for Neglected Diseases initiative (DNDi), and Centre de Recherche pour le Développement Humain (CRDH). PSI was responsible for the work in Kenya, Madagascar, Nigeria, Uganda, Tanzania mainland (which was subcontracted to the Ifakara Health Institute) and Zanzibar. PSI's ACTwatch Project ( has contributed evidence for 2 In March 2011, the AMFm Ad Hoc Committee decided to remove Cambodia from the evaluation due to the lack of an eligible ACT for subsidy. 1 P age

27 the baseline work in Nigeria and Madagascar, which was conducted prior to the IE surveys. DNDi subcontracted with the Research and Development Unit, Komfo Anokye Teaching Hospital, Kumasi, to undertake the work in Ghana. CRDH subcontracted with the Centre International d'etudes et de Recherches sur les Populations Africaines (CIERPA) to undertake the work in Niger. The purpose of the IE is to assess how AMFm has evolved in each pilot, and estimate changes between the baseline and endline surveys in the values of key measures (availability, price, market share and use of quality assured ACTs) to inform decisions regarding the future of AMFm beyond Phase 1. The IE is based on the AMFm (Phase 1) Monitoring and Evaluation (M&E) Results Framework, with a focus on Outputs and Outcomes (Figure 1.2.1). 2 P age

28 Figure : AMFm Phase 1 Results Framework ACTs: Artemisinin based combination therapies; IEC: Information Education and Communication; GF: Global Fund, AMTs: Antimalarial Treatment; SP: Sulfadoxine pyrimethamine, CQ: Chloroquine; TA: Technical Assistance, SIVs: Supporting Interventions Source: Global Fund, AMFm Phase 1 Monitoring and Evaluation Framework, 2009 The IE is therefore designed to answer four questions related to the availability, affordability, market share and use of ACTs. These questions are formulated as follows: 1. Has the AMFm mechanism helped increase the availability of quality assured ACTs to patients across public, private for profit and not for profit sectors, in rural/urban areas? 2. Has the AMFm mechanism helped to reduce the cost of quality assured ACTs to patients at public, private for profit and not for profit outlets in rural/urban areas to a price comparable to the price of most popular antimalarials? 3. Has the AMFm mechanism helped increase use of quality assured ACTs, including among vulnerable groups, such as poor people, rural residents and children? 4. Has the AMFm mechanism helped increase the market share of quality assured ACTs relative to all antimalarial treatments in the public, private for profit and not for profit sectors in rural/urban areas? To answer these questions, building on the AMFm results framework, the IE impact model (Figure 1.2.2) foresees that subsidizing ACTs, accompanied by effective supporting interventions, will lead to a decrease in the ACT price. It is therefore anticipated in the model that if ACT price decreases, more outlets will be willing to stock the product and thereby increase availability. The increase in availability, and the substantial decrease in price, will potentially lead to an increase in use. 3 P age

29 Figure : The Independent Evaluation Impact model Price reductions through negotiations with manufacturers, a subsidy in the form of a buyer co-payment, and supporting interventions ACT Price Decreased ACT Availability Increased ACT Access and Use Increased Malaria Burden Decreased Inputs, Process Outputs Outcomes Impact While an evaluation based on a quasi experimental design would have provided stronger evidence to attribute any change in primary outcomes to the intervention, it is challenging to execute such a study design for an evaluation of a complex public health program such as the AMFm, which is implemented on a national scale with multiple players. The IE therefore uses a pre and posttest/intervention design (Figure 1.2.3) in which each participating country is treated independently as a case study. As the literature suggests for the evaluation of such a complex intervention, in addition to measuring the changes in key indicators pre and post intervention, the evaluation includes an assessment of the implementation process to determine whether any lack of impact reflects implementation failure or genuine ineffectiveness. It also includes comprehensive documentation of context, both to inform assessments about causality and to aid in generalizability to other contexts. Figure : The Independent Evaluation Design The evaluation, therefore, includes two major components: (1) a pre and post intervention study of key outcomes through outlet surveys and use of secondary household survey data, and (2) documentation of key features of the context at baseline and endpoint, and the implementation process in each pilot. The descriptions of context and implementation process provide the information needed to interpret the changes in outcomes over the implementation period, and to judge whether any observed changes are likely due to AMFm. 4 P age

30 The evaluation is based on primary data collected from outlet surveys conducted at baseline and endline (for questions related to availability, affordability and market share of ACTs); secondary data from national household surveys (for question related to use of ACT), such as Demographic and Health Surveys (DHS), Malaria Indicator Surveys (MIS), Multiple Indicator Cluster Surveys (MICS) and ACTwatch household surveys; in depth interviews with key stakeholders involved in the drug supply chain in the country; and review of documents such as reports from AMFm operations research, malaria treatment guidelines, pharmacy regulations, country level reports from MOH and donor partners, including national malaria control strategy documents, results from national surveys, and any other documents relevant to the context data described above. For each country, relevant indicators will be computed for the baseline and endpoint from the outlet surveys. For secondary data from existing national household surveys, appropriate indicators will be extracted from existing reports. To assess change, the IE will calculate the percentage point change or the percent change (whichever is relevant for each indicator) between the baseline and the endpoint. Contextual information will then be processed to help in the interpretation of these results. Pilot specific case studies will be produced, making use of the qualitative and quantitative approaches described above, to document and describe how the AMFm has evolved in each country. The evaluation will distinguish two parts: (i) the upstream part, with emphasis on the business model of the AMFm as a financing platform; and (ii) the downstream part, with emphasis on service delivery to increase access to and use of ACTs, including by poor people. In the case studies, findings from nationally representative outlet surveys will be compared before and after the introduction of the AMFm, taking into account relevant contextual information and results from operational research that become available to help learn how and why the new model unfolds in a variety of contexts, while drawing lessons that can help future operations. While this section gives an overview of the IE to provide the reader with the relevant context, this report presents the country process, context and results of the endline outlet survey for Kenya. This is Step 3 of a four step process. Step 1 included the baseline outlet survey and country specific baseline report. Step 2 integrated these results into a Multi country Baseline Report produced by the Independent Evaluation Team. Findings from this endline outlet survey will be used to inform Step 4, the development of the full AMFm Phase 1 Independent Evaluation report, which will include results from all operational phase 1 pilots, to be submitted to the Global Fund. 5 P age

31 1.3 Country background context Overview of the country The Republic of Kenya is a country in East Africa, which lies astride the equator and is bordered by Ethiopia, Sudan, Uganda, Tanzania and Somalia (Figure 1.3.1). It covers an area of 582,646 square kilometers, of which about 80% of the land area is arid and semi arid and only 20% is arable (National Malaria Strategy a, 2009). It is divided into eight provinces, which are further subdivided into districts, divisions, locations, and sub locations. According to the 2009 census, the population of Kenya is 38.6 million, representing a population increase of 35.0% over the past decade (World Bank, 2010). Of the population, 42.9% is aged below 15 years, and 67.7% live in rural areas (Ministry of State for Planning, 2009). Kenya is home to seven primary ethnic groups, in additional to a significant population of non Kenyan Africans and people of Arab, European, and Asian descent. English and Kiswahili are the official languages of Kenya, although there are a number of indigenous languages. The Kenya National Bureau of Statistics 2010) reported that 84.9% of women and 90.8% of men aged 15 and above were literate. There is, however, great disparity between the provinces with 96% of women in Nairobi being literate compared to 21% of the women in North Eastern province. Geographically, the country can be divided into two regions, the lowlands including the coastal and Lake Basin lowlands, and highlands extending on both sides of the Great Rift Valley. Rainfall and temperatures are defined by altitude and closeness to bodies of water. The coast experiences a tropical climate with high rainfall and warm temperatures year round. The rest of the country experiences four seasons, with a dry period from January to March, the long rainy season from March to May, followed by a long dry spell from May to October, and then the short rains between October and December (Central Intelligence Agency (CIA), 2011). Kenya is a low income country with a Gross Domestic Product (GDP) per capita of US $ per capita (International Monetary Fund, 2011), and an economy that is predominantly agrarian. The Human Development Index (HDI) is 0.470, ranking the country 128th out of 169 countries. Poverty is a significant issue in both rural and urban areas, with overall poverty standing at 49% and 39% in rural and urban Kenya respectively. Kenya experiences strong regional disparities in poverty as well as high rates of inequality of income distribution, particularly in urban areas. 6 P age

32 Figure : Location of Kenya Source: Source: CIA World Fact book Kenya Description of health care system Kenya s health care system consists of a network of more than 6,700 health facilities countrywide, with the public sector accounting for half of facilities (MOPHS, 2010). In 2002, the total number of health personnel was 59,000, representing a density of 189 health personnel per 100,000 people. The public sector accounts for 60% of total health personnel, of whom about 70% are concentrated in hospitals (Wamai et al., 2009). Kenya has two line ministries responsible for health care delivery the Ministry of Medical Services (MOMS) and the Ministry of Public Health and Sanitation (MOPHS) that together have the mandate to implement the National Health Sector Strategic Plan. 7 P age

33 Health Service Delivery Structure The public delivery system follows a pyramidal structure as shown in Figure Figure : Public health delivery system, Kenya Source: Ministry of Health, 2006 MOH Norms and Standards First level care consists of a community unit that is governed by community health committees, and is targeted to include 50 CHWs per unit. Levels 2 and 3 administer preventative care, with a basic level of curative services on a sub district level through dispensaries, health centers, and maternity/nursing homes. Levels 4 6 are comprised of primary, secondary, and tertiary hospitals including two national hospitals Moi Referral and Teaching Hospital in Eldoret and Kenyatta National Hospital in Nairobi (Kenya HIV/MCH Service Provision Assessment Survey, 2004). Health facilities in Kenya are generally controlled by one of four bodies: 1) Ministry of Health 2) Other public institutions 3) Faith based organizations (FBO) and other nongovernmental organizations (NGO) 4) Private organizations The distribution of health facilities by type and controlling agency is presented in figure P age

34 Figure : Distribution of Health Facilities by type and Controlling Agency, 2010 Controlling Agency Ministry of Health Hospitals Health Centers Dispensaries Maternity& Nursing Homes Clinics TOTAL , ,170 Other Public Institutions Faith Based and NGO Private ,756 2,263 TOTAL , ,787 6,769 Source: Health Management Information System List of Health Facilities in Kenya, September 2010 The private for profit health sector, which is comprised of private hospitals, clinics, dispensaries and drug outlets, accounts for about 33.4% of health facilities in the country (MoH, 2010). Health Financing Policies By 2004, the Kenyan government replaced a cost sharing policy with the following for the public sector: Level 2 and 3 facilities are not to charge user fees for curative services (including medicines) other than KES 10 and KES 20 at levels 2 and 3, respectively, for registration. Once registered, the patient receives a client card, and treatment is then free for children under 5. For adults, they pay a minimal amount as part of a cost sharing initiative (facility improvement fund). Children under 5 years are entitled to free medical care in all public and faith based health facilities. There is a waiver system in place for older patients who cannot afford treatment (MOPHS, 2006). Publicly procured medicines for priority health programs, such as contraceptives and medicines for malaria, HIV/AIDS and TB are provided for free in all public and faith based health facilities. For other conditions in adults and children over 5 years at public facilities in levels 2 3, and 4 6, cost sharing is applied. Expenditure tracking surveys have, however, shown that the implementation of this policy is not strictly adhered to and patients are still paying for services that should be free or higher fees than the recommended (MoH, 2008). Faith based organizations tend to charge for the cost of treatment for most conditions, except where waivers and subsidies exist. Health services and medicines in the private for profit health sector are provided in these facilities on a full cost recovery basis. A study on access to essential 9 P age

35 medicines conducted in 2009 showed that the prices of medicines in private for profit facilities were 3.29 times the international reference price (WHO/Health Action International [HAI] 2010). Pharmaceutical sector structure The pharmaceutical sector in Kenya consists of three main segments, namely manufacturers, distributors and retailers, which are all regulated by the Pharmacy and Poison s Board (PPB): Manufacturers There were 45 registered manufacturers for pharmaceutical products in Kenya in 2009 (PPB, 2009). These include local manufacturing companies, multi national corporations, as well as subsidiaries or joint ventures. The industry compounds and packages medicines, repackages formulated drugs and processes bulk drugs into doses, using predominantly imported active ingredients and excipients. In 2005, it was reported that Kenya was the largest producer of pharmaceuticals in the Common Market for Eastern and Southern Africa region, supplying about 50% of the region s market (PPB, 2005). Distributors The three main distributors of pharmaceuticals are: o The Kenya Medical Suppliers Agency (KEMSA) is a parastatal under the Ministry of Medical Services. Its main mandate is to procure, warehouse and distribute medical commodities to public and faith based health facilities. The KEMSA distribution system consists of a kit based push system, and an inventory based ordering (pull) system. o Mission for Essential Drugs and Supplies (MEDS) is a Christian not for profit organization, which undertakes procurement and supply of essential drugs and medical supplies to faith based facilities, some public facilities and other not for profit healthcare providers. MEDS operates a pull system, where the client places orders. o Private wholesalers are pharmaceutical wholesalers registered with the PPB. In 2008, there were 212 private wholesalers. Only registered pharmacists can obtain a Wholesale Dealer s License. Registered Outlets and Drug Stores there are two types of registered pharmaceutical retailers in Kenya (deemed pharmacies in the report). These differ from unregistered outlets, e.g., drug stores. o Pharmacist Premises is a pharmacy that is superintended by a registered pharmacist. The PPB website records 405 pharmacist premises that were registered in 2010 (PPB, 2010). o Pharmaceutical Technologist Premises is a pharmacy that is superintended by a Pharmaceutical Technologist, who has worked under the supervision of another qualified superintendent (pharmacist or pharmaceutical technologist) for a period of not less than six years from the time of obtaining the diploma in pharmacy (PPB, 2006). The PPB records include 547 pharmaceutical technologist premises registered in o Private clinics/dispensaries are smaller than hospitals, some of which are not registered with the Ministry of Health. They provide consultations and examinations, as well as prescription and over the counter medicines. Dispensaries are run by registered nurses or clinical officers. o Drug stores: Small businesses that are not registered with the PPB, but that sell various classes of prescription and over the counter medicine at commercial prices, and may 10 P age

36 also sell other goods like cosmetics. They may be staffed by chemists, pharmacists, pharmaceutical technologists, pharmacy assistants and health practitioners. There is no restriction on the type of drugs that are sold in pharmaceutical technologist premises and both types of retail outlets are considered pharmacies. Treatment Seeking Behavior Malaria is the leading cause of outpatient morbidity in Kenya, representing a third of health care utilization (Macro, 2004). Studies on malaria treatment seeking behavior in the general population have shown that people seek treatment in both the formal and informal sector, although the first source of treatment is often the informal sector (Abuya et al., 2007, Chuma et al., 2009). However, about a quarter of Kenyans do not seek medical treatment when they are ill (MoH, 2003). For those that do seek treatment outside the home, studies suggest the majority self treat using over the counter drugs, followed by a visit to a trained health worker or a traditional healer (DOMC, 2010). The proportion of children under five with fever who took an antimalarial drug is low at 23%, with the percentage of those receiving correct treatment taking an ACT on the same day or the next day after the onset of fever at only 4% (DHS, 2010). 11 P age

37 1.3.3 Epidemiology of malaria Malaria is the leading cause of morbidity and mortality in Kenya, with close to 70% (24 million) of the population living in transmission areas (DHS, 2010). Kenya has four malaria epidemiological zones which are largely defined by altitude, rainfall patterns and temperature (MoH, 2010). These areas include: 1. Endemic areas around Lake Victoria in Western Kenya and in the Coastal regions. Transmission in these areas is intense throughout the year, but peaks from June to August and again in late November following the rainy seasons. 2. Seasonal transmission areas. These are arid and semi arid areas of the northern and south eastern parts of the country, which experience short periods of intense transmission during the rainy season. 3. Epidemic prone areas of Western highlands of Kenya. Malaria transmission here is seasonal with considerable year to year variation. Case fatality rates during an epidemic can be up to ten times greater than those experienced in regions where malaria occurs regularly. 4. Low risk malaria areas. This zone covers the central highlands of Kenya including Nairobi, where temperatures are too low to allow completion of the sporogenic cycle of the malaria parasite in the vector. Plasmodium falciparum is the predominant species of malaria in Kenya, accounting for 98.2% of parasites. According to Kenya s National Malaria Strategy , the expansion of coverage of parasite and vector control interventions has resulted in a decline in the malaria disease burden, its severity and transmission patterns. Various sources suggest that there have been important reductions in morbidity and mortality in different parts of the country including malaria sentinel site data (Zhou et al., 2011), hospital admissions data and studies on transmission intensity (Okiro et al., 2010; Okiro et al., 2007). An updated map on malaria risk and endemicity produced in 2009 (Figure 1.3.4) provides prevalence estimates whereby the darker the shade of red, the higher the prevalence and where blue represents 0% prevalence. 12 P age

38 Figure : Malaria transmission risk map [Transmission zones], [2009], [Kenya] Source: Kenya Division of Malaria Control Antimalarial Policies and Regulatory Environment The MOMS and MOPHS released the third edition of the National Guidelines for the Diagnosis, Treatment and Prevention of Malaria in Kenya in May 2010, which provides guidelines for the management of uncomplicated and severe malaria, and malaria in pregnancy. In 2004, the Division of Malaria Control (DoMC) adopted artemether lumefantrine combination therapy (AL) as the first line treatment for uncomplicated malaria. The second line treatment is dihydroartemisin piperaquine, which is not a WHO pre qualified drug, limiting the availability of funding for procurement. Parenteral quinine or parenteral artemisinins are recommended for the treatment of severe malaria, with intramuscular quinine, or in its absence rectal artesunate or intramuscular artemether, used as pre referral treatment. The 2010 guidelines recommend the use of parasitological diagnosis, either in the form of microscopy or a rapid diagnostic test, for all suspected cases of malaria. Under current policy, medicines to treat malaria should be free at all public and faith based facilities. The PPB is the drug regulatory authority established under the Pharmacy and Poisons Act, Chapter 244 of the Laws of Kenya. The Board regulates the practice of pharmacy and the manufacture and trade in drugs and poisons. It is one of the statutory institutions responsible for quality control in the Ministry of Medical Services. ACTs, including AL, are categorized as prescription only medications (Part I poisons), limiting their distribution to government facilities, private clinics and registered pharmacies. Discussions are ongoing as to whether ACTs should be re categorized as Part II poisons, in which case they would be available over the counter and could be distributed by community health workers. The debate has focused on balancing increased access with the risk of increased inappropriate usage (Key Informant 13 P age

39 Interviews). In September 2009, the PPB issued a communiqué that amodiaquine and artemisinin monotherapies were prohibited and that the usage indication for SP must be updated to reflect the current treatment guidelines. Despite these efforts, the drugs still remain on the market (Key Informant Interviews). Small scale subsidy pilots were also in place in Kenya during the time of data collection. In 2007, an initiative was launched to make AL available through private shops in targeted rural areas. Nine Community and Family Wellness (CFW) shops, located in three districts participated in the intervention (five in Embu, three in Kirinyaga, and one in Mbeere). As part of this initiative, AL was distributed to the CFWs from the government central medical store and administered free of charge to patients with uncomplicated malaria after confirmation by an RDT. Patients were to pay $0.65 for a consultation and RDT before receiving AL for confirmed malaria (Sabot et al, 2009). In 2008, nine sub locations in Western Kenya were selected as part of an intervention, where retail outlets were provided with subsidized packs of pediatric ACTs (branded as Tibamal, a pretested name derived from the Kiswahili words Tiba ya Malaria, meaning malaria cure). The outlets were instructed to sell the packs at a retail price of 0.25 USD (20 KSH), and this price was printed on the drug packaging. In addition, supporting interventions included training of retail outlet staff and community awareness activities were conducted (Kangwana et al., 2011) Malaria control strategy The key malaria control interventions in Kenya are outlined in the strategy document (KNMS, 2009). The focus is on case management of malaria, management of malaria in pregnancy, vector control, and epidemic preparedness and response. Case management of malaria The aim is to have 80% of all self managed fever cases receiving prompt and effective treatment and 100 per cent of all fever cases who present to health facilities receiving parasitological diagnosis and effective treatment by Case management activities include: updating drug policy, training health workers, improving use of diagnostics (microscopy/rdt), and ensuring equitable access to ACTs. Malaria treatment services are provided at all levels of service provision in the public health system. Levels 2 and 3 are equipped to manage uncomplicated malaria. Severe malaria is treated as a medical emergency and patients are referred to higher levels of care. Pre referral treatment is, however, given at the lower levels. It is planned that by 2013, home management of malaria (level 1, community) will be implemented in malaria endemic districts through CHWs. By the end of the data collection period however, CHWs were not permitted to provide ACTs or RDTs. Access to diagnostic testing services is being expanded through the provision of RDTs to facilities that do not have microscopy. It is expected that with increased access to RDTs, provided through a Global Fund Round 10 Grant, that rational use of ACTs will improve thereby increasing access by decreasing demand (Key Informant Interviews). 14 P age

40 Results of a nationally representative biannual survey to monitor outpatient malaria case management by the Division of Malaria Control (DOMC) conducted in April 2010, showed that 81.6% of the facilities were reporting at least one of the four age bands of AL packs in stock, and all four packs were available in only 41.1% of facilities. Between May and July 2010, approximately 8,300,000 adults received an ACT under the appropriate treatment guidelines for malaria representing about 60% of the expected targets under Kenya s Global Fund Round 4 indicators (Global Fund, 2011). Management of malaria and anemia in pregnancy Intermittent preventive treatment (IPT) is recommended for pregnant women living in malaria endemic areas and is to be provided for free in public facilities. The proportion of women who received the recommended two doses of sulfadoxine pyrimethamine (SP) (second and third trimesters) increased from 5% in 2003 to 14% in , while 36% of women received at least one dose of SP (DHS, 2010). The National Malaria Control Strategy aims for universal coverage of Long Lasting Insecticidal Nets (LLINs) by Studies conducted in , found that 49% of pregnant women were sleeping under a LLIN compared with only 4% in 2003 (DHS, 2010). Integrated vector management Kenya implements two main methods for vector control, the distribution of LLINs and indoor residual spraying (IRS). In order to attain universal net coverage, the government utilizes three distribution channels: 1) Mass distribution of LLINs in endemic and epidemic prone areas every 3 years to all households, as well as distribution in schools and health facilities 2) Routine distribution to pregnant women and children under one year through ANC, child health clinics, and Community Health Workers 3) Social marketing of LLINs The KDHS reported that 55.7% of households were found to own at least one Insecticide treat Net (ITN). Net usage was found to be lower, with only 46% of children sleeping under an ITN the night before the survey. The Division of Malaria Control (DOMC) has focused indoor residual spraying (IRS) efforts on 16 epidemic prone districts covering approximately 6.5 million people in 1.3 million homesteads, with a target of annually spraying 80% of the households in these districts 6 8 weeks before the onset of heavy rains (May August). 15 P age

41 Epidemic preparedness and response The National Malaria Control Strategy aims to ensure that all malaria epidemic prone districts have the capacity to detect, and are prepared to respond, to malaria epidemics annually. Epidemic preparedness and response efforts have to date focused on IRS Malaria financing Funding for malaria control activities in Kenya comes from the government, and external contributions from the Global Fund, USAID/PMI, other bilateral agencies including DFID and JICA, United Nations (UN) agencies including UNICEF, and other local and international NGOs. The Kenyan Government contributed less than one percent of the financing for malaria activities in 2008 and 2009 (WHO, 2010) and this mainly supports health infrastructure, funding for some health commodities including medicines, and salaries of health workers engaged directly as part of the overall service delivery mandate of the health sector. The Global Fund traditionally, the Global Fund has been the largest contributor towards malaria financing in Kenya. By 2010, the Global Fund had disbursed US$146 million to malaria control efforts in Kenya for the provision of prompt and effective treatment of malaria, epidemic detection and control, use of insecticide treated bed nets for vector control and development of human capacity through training. The DOMC is currently spending from Phase II of the Round 4 Grant, which was to end in January 2011 but has received a six month no cost extension. The total amount of the Round 4 Grant was US$162.2 million, with US$76.1 million disbursed under Phase I and US$86 million under Phase II. Of the Round 4 Phase II grant, US $ 18,394,872 had been earmarked for procurement of ACTs. With the AMFm, the same quantity of drugs would now cost US $1,601,699, resulting in a saving of over US$16 million. Money saved though AMFm procurement is planned to be spent on supporting interventions including: a public awareness and education campaign; training, monitoring and supervision of health workers, strengthening pharmacovigilance; and improvement of the regulatory environment of ACTs (Global Fund, 2011). USAID/President s Malaria Initiative (PMI) Full PMI implementation in Kenya began in 2008, with a budget of US$19.7 million for 2008/2009. The proposed budget for 2009/2010 was US$40 million to support procurement and distribution of ITNs, improved case management, IRS and behavioral change activities. By 2010, PMI had distributed close to 15 million ACTs with the objective to fill gaps while continuing to strengthen the supply chain and logistics systems to ensure a steady supply of these essential medications (PMI, 2010). Department for International Development (DFID) has been contributing to malaria control activities in Kenya since DFID funding to Kenya is channeled through the WHO and PSI. The proposed WHO managed malaria funding between 2010 and 2013 is approximately US$ 8.1 million (US$ 2.7 million per year) with additional support of US$ 16.2 million per year for three years to continue support to PSI s distribution of LLINs. 16 P age

42 Key informant interviews suggest that funding for malaria control projects in Kenya has been weakened by the global economic crisis, including a depreciation of the Kenya shilling, and a reduction in DFID funding. Key informants suggest that this did not, however, affect the availability of ACTs in the public sector, since the Global Fund, which is the main funder for purchase of ACTs, operates off shore accounts and makes payments to suppliers in USD, although WHO Kenya s planned stationing of officers in regional bases in Kenya to provide support to the Malaria Control Coordinators will no longer take place. Implementation process of AMFm in Kenya The AMFm grant amendment was signed on 14 July 2010 and an official launch activity was held in Nairobi on 17 August. The first disbursement by the Global Fund was made in August 2010 (via a Harley s Limited order in August). First line buyers commenced placing orders in July 2010 and by the end of 2010, they had placed orders for just under 7 million treatment units. By November 23, 2010 (the end of baseline data collection), nine first line buyer agreements had been signed (7 from the private for profit sector, 2 from the public sector). Of these, 635,000 treatment units arrived in Kenya either before or during the data collection period for the baseline survey. The first medicines were delivered on August 10, 2010, however no information is available about the date the first medicines cleared customs, but there is evidence of co paid ACTs being available on August 17, 2010 (reported in the Daily Nation newspaper). Further promotional activities took place in early October, with a national newspaper supplement on 1 October and posters and leaflets distributed to pharmacies. Further supporting interventions such as national level marketing campaigns did not commence by the end of the baseline survey. The fact that some national launch activities had taken place and that a small quantity of co paid ACTs was available in the country before the start of baseline data collection implies that some of the indicators may capture AMFm implementation. 17 P age

43 2. Methods 2.1 Outlet survey Outlet survey indicators The following table shows the primary indicators to be measured through the outlet survey, and presented in this report. Results are presented by urban and rural locations and nationally. They will also be presented by outlet type (though there may not be sufficient power to detect statistical differences between outlet types). Table : Primary indicators Availability indicators 1.1 Proportion of censused outlets that have any antimalarials in stock at the time of survey visit in rural and urban areas 1.2 Proportion of outlets that have non artemisinin monotherapy or non artemisinin combination therapy in stock among outlets with any antimalarials in stock at the time of survey visit in rural and urban areas 1.3 Proportion of outlets that have artemisinin monotherapy in stock among outlets with any antimalarials in stock at the time of survey visit in rural and urban areas 1.4 Proportion of outlets that have non quality assured ACTs in stock among outlets with any antimalarials in stock at the time of survey visit in rural and urban areas 1.5 Proportion of outlets that have quality assured ACTs in stock at the time of survey visit among outlets with any antimalarials in stock at the time of survey visit in rural and urban areas 1.6 Proportion of outlets with antimalarials in stock at the time of survey visit that have been out of stock of quality assured ACTs for at least 1 day in the last 7 days in rural and urban areas 1.7 Proportion of the population living in a sub district where there is at least one outlet that had a qualityassured ACT in stock at the time of the survey visit in rural and urban areas Pricing indicators 2.1 Median cost to patients of one adult equivalent treatment dose (AETD) of quality assured ACTs in rural and urban areas 2.2 Median cost to patients of one AETD of non quality assured ACTs in rural and urban areas 2.3 Median cost to patients of one AETD of artemisinin monotherapy in rural and urban areas 2.4 Median cost to patients of one AETD of non artemisinin monotherapy or non artemisinin combination therapy in rural and urban areas 2.5 Median percentage markup between retail purchase and selling price of quality assured ACTs in rural and urban areas 2.6 Median total markup from first line buyer purchase price to retail selling price for quality assured ACTs Market share indicators 4.1 Total volume of quality assured ACTs sold or distributed in the last week, as a proportion of the total volume of all antimalarials sold or distributed in the last week in rural and urban areas Sampling Approach The IE team developed the sampling strategy for all the AMFm Phase 1 Independent Evaluation outlet surveys. The sampling approach was based on that used in ACTwatch outlet surveys (including outlets in the public, private for profit and not for profit sectors). The target sampling units were all types of outlets that have the potential to sell or provide antimalarials in Kenya. A probability sample of 23 locations out of 238 was selected from urban 18 P age

44 domains and 34 locations out of 2190 were selected from rural domains, giving a total of 57 locations. The outlets were classified into two main categories: Category 1: public health facilities (this included national referral hospitals, provincial general hospitals, health centers or sub health centers, and dispensaries), and pharmacies. Category 2: other antimalarial drug sellers, including private, NGO, and mission based hospitals, clinics, dispensaries, and supermarkets, market stalls, petro station, convenience stores, community health workers, and itinerant drug vendors. Sampling procedures were used to select outlets within each category, as described below. Sample size determination The sample size calculations have been based on Indicator 1.5 Availability of QAACTs (the proportion of outlets that sell QAACTs as a share of the number of outlets that have stocks of any kind of anti malarials at the time of the survey). This indicator has been chosen because of concerns that QAACTS will be too rare at baseline to provide a feasible sample size for the price/affordability indicators. After some calculations based on ACTwatch data, it was determined that the average number of non free QAACTs per sub district is very small, especially in the rural domain. This would require a very large sample size (in terms of the number of sample locations) and such a large sample size is not feasible within the budget constraints to power the calculations on the detection of the median price change of the non free QAACTs. Therefore, we have focused the sample size calculations on the ability to detect a 20 percentage point change increase in QAACT availability. The required sample size is calculated in 3 steps: 1. The required number of outlets with anti malarials in stock on the day of the survey 2. The number of outlets that must be enumerated to arrive at this number of antimalarial selling outlets ( gross sample size ) 3. The number of sub districts 3 that must be visited to arrive at this number of outlets (operational sample size). 1. Required number of outlets The sample size required for the endline survey is a function of the sample size achieved in the baseline survey calculated based on a similar formula as in the baseline survey: n Deff Z 1 P(1 P)(1 1/ ) Z ( P P ) P (1 P ) / P (1 P ) where: n = desired sample size for the endline survey P 1 = the value of the indicator measured at baseline survey P 2 = the expected value of the indicator at the endpoint survey P = (P 1 +P 2 )/2 λ=the ratio of the baseline survey sample size over the endline survey sample size which can only be solved in a recursive calculation A sub district in this study is defined as a geographical area or enumeration unit with approximately 10,000 to 15,000 inhabitants. In the case of Kenya, this corresponds to a location. 19 P age

45 Z 1 α = the standard normal 1 α quintile corresponding to an α (type I) error with a one sided test; replace α by α/2 if a two sided test is desired Z 1 β = the standard normal 1 β quintile corresponding to the power of the test Deff = the design effect for cluster sampling estimated from the baseline survey The required sample size has been calculated on the basis of the following assumed values of the key parameters: P 1 = the value of the key outcome indicator measured at the baseline survey P 2 = the expected value of the indicator at the endline survey; a 20 percentage point difference will be desired (assuming that a 20% point increase is the minimum necessary to justify the importance in public health policy terms) P = (P 1 +P 2 )/2 P 2 P 1 = the change in availability from baseline to follow up is 20 percentage points (the minimum level of change felt to be needed to justify continuing the intervention on public health grounds) Z 1 α = 1.64 corresponding to an α (type I) error of 5% with a one sided test Z 1 β = 0.84 corresponding to a power of test at 80% (or a type II error of 20%) The sample size for the Kenya endline survey is calculated for the urban and rural domain separately because of the required detectable targeted changes by domain. Design effect is estimated at Deff = 14 for the urban domain, and Deff=12 for the rural domain which are much higher than the assumed values in the baseline survey. By applying the above formula with the requested parameters, this gives a number of 1024 outlets in the urban domain, 900 outlets in the rural domain, which is the number of outlets having any kind of anti malarial stocks at the time of the survey. 2. Number of outlets that must be enumerated The estimated gross sample size (number of outlets enumerated) needed for the QAACT availability indicator is determined by the following formula for urban and rural domains separately: N n / P am where P am is the proportion of outlets having anti malarial stocks at the time of the survey among all outlets enumerated. In this equation, the assumptions are as follows: N = desired sample size of all outlets for monitoring availability indicators, n is the number of outlets with anti malarial stocks at the time of the survey. P am is the proportion of outlets having anti malarials in stock at the time of the survey among all outlets enumerated estimated from the baseline survey for each of the urban and rural domains. The baseline survey data indicate that 11% of the outlets in urban and rural areas, on average, have anti malarials in stock at the time of the survey. By applying these percentages to the above formula, a total number of 9300 outlets in the urban domain and 8300 outlets in the rural domain must be interviewed in order to detect a 20 percentage point increase in QAACT availability, for urban and rural domains separately. 20 P age

46 3. Number of sub districts To convert the gross sample size into the number of sub districts/locations, we apply the estimated average number of outlets per sub district/location (n outlet ). The number of outlets needed to reach the required number of outlets with anti malarials in stock may be different for urban and rural areas depending on the average number of outlets per sub district/location (n outlet ) and the percentage of outlets with anti malarials (P am ) in urban and rural areas separately. The Kenya baseline survey results show that there are, on average, 408 outlets interviewed per urban sub district and 243 outlets interviewed per rural sub district. By applying these estimated parameters for Kenya, the ultimate number of sub districts/locations required to reach the estimated number of outlets would be 23 in the urban domain and 34 in the rural domain, giving a total of 57 sub districts required in the country. Selection procedure of the sub districts Multi staged sampling methodology was conducted, with a booster sample used at the division level. The sampling frame used was the list of the 2,428 locations in 8 provinces of Kenya, using information from the 1999 National Census. The urban rural definition of the locations was based on its percentage of urban/peri urban population: a cut off of 65% or more urban was used to define urban locations; all the rest were defined as rural. With the 65% cut off, there were 238 locations that were defined as urban, which coincided with the government provided number of urban locations. The sample was stratified into urban and rural areas. Implicit stratification was achieved by separating the urban and rural domains according to malaria endemicity, so each domain was stratified into 4 strata reflecting the 4 levels of endemicity in Kenya. Implicit stratification by Province, District and Division was also achieved by sorting the sampling frame within each explicit stratum before the sample selection. At the first stage, locations were selected proportional to the population. A probability sample of 23 locations out of 238 locations was selected from urban domains and 34 locations out of 2190 were selected from rural domains, giving a total of 57 locations. For locations with population sizes above 40,000, these were segmented into smaller segments of approximately inhabitants. One segment was then randomly selected as the AMFm cluster. For the booster sample, division level listings of public health facilities were obtained from the 2010 Kenya Master Health Facility List. For each randomly selected location, all public health facilities located in the corresponding division of the selected location were included. Exceptions for this approach are made for localities with population sizes over 40,000. In this case, the selected cluster (segment) served as the sub district and the location served as the booster sample area. A similar approach was used for the booster sample of registered pharmacies, using the 2010 PPB listing of registered Pharmacies in Kenya. 21 P age

47 Within each location, a census of all outlets that meet the inclusion criteria was conducted Data collection The IE team developed the generic questionnaire in consultation with the DCs, the Global Fund and other key stakeholders. The questionnaire is based on the ACTwatch outlet survey questionnaire, and wherever possible the questions have been kept the same to permit comparability with data collected in ACTwatch surveys in Nigeria and Madagascar. The IE team made several adaptations to the ACTwatch tool to ensure that the IE indicators were included and other requests from key stakeholders were met (e.g., the addition of questions on stockouts of quality assured ACTs, training courses attended, and knowledge on proper dosing of quality assured ACTs). To ensure high quality data across countries, the IE team used Standard Operating Procedures (SOPs) developed for the ACTwatch project. The SOPs outline each element of data collection and management, e.g., questionnaire translation, questionnaire pretesting, fieldworker training, and double data entry. The ACTwatch training materials were adapted and modified at the country level as required. Preparatory phase Questionnaire modules were used in the outlet survey: a screening questionnaire for all outlets, and for eligible outlets: a provider questionnaire, an antimalarial audit and a RDT audit. The screening questionnaire was used to identify outlets that were eligible for the audit and provider interviews. The provider questionnaire collected information on outlet demographics (e.g., health qualifications of staff, number of staff that prescribe or dispense medicines), provider knowledge of the first line treatment, recognition of the AMFm logo, perceptions of the most effective antimalarial medicine, and stock outs of QAACTs. The antimalarial audit questionnaire collected data for each antimalarial stocked, including information on brand name, generic name and strengths, package type and size, recall of volumes sold over the week before the survey, recall of last purchase price and selling price. The RDT audit questionnaire collected data on each RDT stocked, including information on brand name, recall of volumes sold over the week before the survey, recall of last purchase price and selling price (to adults and children under five). Separate questions were also asked on microscopic testing availability, recall of price and number administered over the week before the survey. The Independent Evaluator (IE) provided the generic questionnaire in English. The generic IE questionnaire was modified to cater for country specific responses and adapted to suit the context of Kenya (e.g., names of administrative boundaries, types of anti malarial outlets, titles of health worker cadres, first line anti malarial treatment, and local currency). Before the survey, the questionnaire was pilot tested in Nairobi, Kenya and translated into Swahili. The study received ethical clearance from the Kenya National Health Ethics & Research Committee on 16 September P age

48 Standardised ACTwatch training materials, modified by the Independent Evaluator and adapted to the national setting, were administered by APHRC and ACTwatch central staff. The fieldwork training was conducted between 21September and 1 October 2011 in Nairobi. The training of field interviewers ended on 29 September, and two days of additional training was provided for Field Supervisors and Quality Controllers between 30 September and 1 October, A field exercise was conducted on 28 of September. Ninety seven trainees attended the first day of training, and a total of 86 were finally selected for the field work. Field worker training sessions covered completing the questionnaire, informed consent, conducting the census, and identifying outlet types. Interviewers were also trained on how to identify antimalarial medicines, including the differences between ACTs and non ACTs, trade names and generics, packaged and loose tablets, and the various formulations. One day of field practice was also conducted within Nairobi. The supervisor and quality controller training sessions covered roles and responsibilities, coding of questionnaires, error checks for questionnaire validity, field monitoring and reporting and back checking of questionnaires. Fieldwork Data collection started on 7 October and was completed on 10 December Field work was conducted by a team of 86 field staff who were organized into six teams assigned to specific geographic areas (provinces). Overall, the field team included 12 supervisors, 12 quality controllers, 6 regional coordinators and 56 interviewers. Each province had two sub teams, each including a supervisor, quality controller, and up to five interviewers. Official lists of pharmacies operating in all locations were obtained from the PPB, and official lists of the public health facilities were obtained from the 2010 Kenya Master Health Facility List. During the data collection, a more updated Kenya Master Health Facility Listing (2011) became available and was used for areas where data collection had not yet commenced. These outlets were listed according to their location in the locations and used to identify pharmacies and public health facilities in advance of data collection. During data collection, survey teams consulted with location and district level officials, who reviewed the health facility listings and maps. Any new registered pharmacies and public health facilities were updated in the original list. Officials and community elders also helped to define location boundaries and to identify other outlets with the potential to sell drugs. A snowball technique was also used: outlets included in the survey were asked to identify other outlets stocking or with the potential to stock medicine in the locations. In rural areas, interviewers first covered the main trading center and then asked the outlet owners/providers for the locations of other outlets found in their village. In urban areas, interviewers were allocated different streets, and outlets were approached in a logical manner. In both urban and rural areas, key features, such as roads, were used as markers to avoid double coverage. For each outlet that was identified during the census, the outlet type and location were noted, along with its longitude and latitude coordinates (obtained via hand held GPS units). The fieldworker then proceeded to identify the senior staff member currently present at the outlet, and screening questions were administered. For outlets that were eligible, the interviewer then read the 23 P age

49 information sheet to the senior staff person and obtained witnessed oral consent. Consenting providers were then administered the interview. A maximum of three call backs were made to outlets that were either closed at the time of interview, or where the interview was interrupted. The quality controller on each team conducted 5% of a random sample of questionnaires (both screeners and complete audits). Special attention was given to refusals, or questionnaires with substantial missing data and/or non response. Once questionnaires had been reviewed by the supervisor or quality controller, the questionnaires were then sent to APHRC headquarters in Nairobi. The questionnaires were reviewed again by APHRC staff in Nairobi, and coding of drugs and RDTs was also conducted. Any questionnaires with inconsistencies were sent immediately back to the field within the same week. Both supervisors and quality controllers accompanied interviewers during field collection, and conducted spot checks. Regional coordinators, the division of malaria control, the division of pharmacy and poisons board, and ACTwatch Central teams also provided additional supervision, ensuring that a census had been adequately completed. These staff also helped to review the questionnaires. Questionnaires from the field were sent and received via certified courier services. Field reports were ed to the field coordinator (based in Nairobi) on a daily basis, who provided monitoring updates to the ACTwatch Project. The field coordinator also compiled all monitoring figures and reviewed all questionnaires for coding issues Data processing An Access database was developed by the ACTwatch project. Data were double entered by APHRC from 24 October 2011 and completed 7 January All data are backed up on a central server at ACTwatch central and off site via back up using an external hard drive. 24 P age

50 2.1.5 Data analysis Data analysis process Detailed data cleaning guidelines giving step by step instructions on how to clean each section of the data using range and consistency checks were utilized during the analysis process. Commands executed for data cleaning were documented using a syntax file, and the results of running these commands using a log file. A standardized tabulation plan was used for all tables presented in this report, which were produced using standard analysis do files in Stata. Analysis was run using Stata version11, recording results in a log file. Accounting for the survey design in data analysis We accounted for three aspects of the sampling design during the analysis: Sampling weights: Sample weights were calculated for the outlet survey data to allow for 1) difference in sampling probabilities due to variation in the size of strata, 2) the oversampling for the booster sample, and 3) the sampling strategy, which involves a census of outlets in locations of varying size, selected using probability proportional to size (PPS) sampling. Weights were based on sampling probabilities and were calculated by the IE after data cleaning was complete. Appendix 8.6 provides a detailed description of the calculations performed and weights used. Clustering: As the sample was clustered at the level of the division for the booster sample and the location for other outlets, the calculation of the standard errors takes the clustering into account because outlets in a given cluster are likely to be more similar to each other than to outlets in other clusters. (The standard errors did not take into account clustering of products within outlets because a complete list of all relevant products in each outlet was obtained and no sampling was performed). Stratification: As locations were sampled separately in each stratum, it was necessary to adjust for this in the calculation of standard errors. To account for these design features in the tabulations, we used the STATA commands for analyzing complex survey data ( svy commands) to weight the data and calculate confidence intervals which account for clustering and stratification. We declared the primary sampling unit (division), the weight variable (wt), the strata and the finite population correction (fpc) equaling the sampling fraction for each stratum (the number of sampled locations in a stratum divided by the total number of locations in the stratum, or 0.5 if the sampling fraction was greater than 50 percent). 4 This was specified as: svyset division [pweight=wt ], strata(strata) fpc(fpc) We calculated a proportion and its 95 percent confidence interval (CI) as: 4 For simplicity we used the division as the primary sampling unit for both booster sample and main sample outlets, as it is rare for there to be more than one main sample location from the same stratum in a division. However, in Kenya, all the main clusters came from 57 unique divisions. 25 P age

51 svy: proportion VariableName Classification of antimalarials For the purpose of analysis, antimalarials were split into three categories, in line with the Independent Evaluation indicators, which require information on non artemisinin therapy (nat), artemisinin monotherapy (AMT) and artemisinin based combination therapy (ACT). AMT were further classified into oral and non oral AMT, as while non oral AMT are recommended for treatment of severe malaria, the removal of oral AMT from the market is a key policy goal. ACTs were further sub divided into quality assured ACTs (QAACTs) and non quality assured ACTs. QAACTs are ACTs that comply with the Global Fund to Fight AIDS, Tuberculosis and Malaria s Quality Assurance Policy. For the purpose of the Independent Evaluation, a QAACT is any ACT which appeared on the Global Fund's indicative list of antimalarials meeting the Global Fund's quality assurance policy prior to baseline or endline data collection (see or which previously had C status in an earlier Global Fund quality assurance policy and was used in a program supplying subsidized ACTs. At endline, QAACTs were defined as any ACT which appeared on the Global Fund s indicative list of antimalarials meeting its quality assurance policy as at September 2011, or which previously had C status in an earlier Global Fund quality assurance policy and was used in a program supplying subsidized ACTs. AA list of all ACTs qualifying as QAACTs at the time of the endline survey is included in Appendix 8.2. Calculation of antimalarial volumes, prices and markups Antimalarial volume and price data are reported in terms of adult equivalent treatment doses (AETDs) using an AETD calculator developed by ACTwatch with some modifications. An AETD is defined as the number of milligrams (mg) of an antimalarial drug needed to treat a 60 kg adult (refer to Appendix 8.7 for details). The number of mg/kg used to calculate one AETD was defined as what was recommended for a particular drug in the treatment guidelines for uncomplicated malaria in areas of low drug resistance issued by WHO (as of 5 April 2011). Where WHO treatment guidelines did not exist, such as for Halofantrine (Halfan), or Dihydroartemisinin, AETDs were based on the product manufacturer s treatment guidelines. In the case of ACTs, which have two or more active antimalarial ingredients packaged together (either co formulated or co blistered), the strength of the artemisinin based component was used as the basis for the AETD calculations. Information collected on the medicine strength and unit size, as listed on the product packaging, was then used to calculate the number of AETDs contained in each unit. Market share was calculated by dividing the number of AETDs of a particular antimalarial category sold by the total number of AETDs of all antimalarials sold. In cases where outlets stocked antimalarials, but some or all sales volumes were missing, we did not impute for missing values. Price data were collected in local currencies and adjusted to 2010 prices in order to facilitate comparisons to baseline estimates which were adjusted to 2010 prices for all pilots. Prices were adjusted using the ratio of the average national consumer price index for 2011 to the national average consumer price index for 2010 (source International Monetary Fund (IMF), International Financial Statistics). These 2010 prices were then converted to their USD equivalent using the average interbank rate for 2010 (USD = KES, source *******). Price data are reported using 26 P age

52 median and inter quartile range, which are appropriate for describing distributions likely to be skewed. Retail gross percentage markups were calculated for each product as the difference between selling price and purchase price, divided by purchase price. In cases where an outlet received an antimalarial for free from its supplier and distributes the product for free, the retail markup was set to 0%. In cases where an outlet received an antimalarial for free from its supplier, but does not distribute the product for free, the retail markup was set to missing. The tables that present markup data indicate the number of observations set to missing for this reason. 3. Results Outlet survey 3.1 Characteristics of the sample Throughout the document, reference will be made to the letters in this flow diagram (A to E) as a reminder of which subset of outlets a given table is referring to. 27 P age

53 Figure : Survey flow diagram, [Kenya], 2011 A B Outlets enumerated* [13,376] Outlets screened [11,386] C Outlets which met screening criteria: 1=[1,871] or 2= [241] D Outlets interviewed** [2,088] E Outlets with antimalarials in stock on day of visit [1,856] Outlets not screened [1,990] Outlets which did not meet Screening criteria [9,274] Outlets not interviewed [24] Outlets with not antimalarials in stock on day of visit*** [232] Interview interrupted : [0] Eligible respondent not available/time not convenient for interview : [86] Outlet not open at the time : [408] Outlet closed permanently : [1,430] Other : [20] Refused : [46] Interview interrupted : [1] Eligible respondent not available/time not convenient for interview : [10] Outlet not open at the time : [0] Outlet closed permanently [0] Other : [8] Refused : [5] 1: Antimalarials in stock on day of visit ; 2: No antimalarials in stock on day of visit, but antimalarials in stock in previous 3 months *Enumerated means were visited and filled in at a minimum basic descriptive information (sections C1 C9 of questionnaire) **Interviewed means that final interview status was completed or partially completed but had stock in previous 3 months ***Outlets with no antimalarials in stock on day of visit but had stock in previous 3 months Table : Outlets enumerated by location, drugs stocked and final interview status [Kenya], 2011 Final interview status Urban n Rural n Total n Number of outlets enumerated (Flow Diagram Reference A) Number of outlets stocking drugs at the time of the survey visit Number of outlets meeting the screening criteria* (Flow Diagram Reference C) Number of outlets stocking antimalarials at the time of the survey visit (Flow P age

54 Diagram Reference E) Number of outlets without antimalarials in stock at the time of the survey visit, but who had antimalarials in stock at some time in the 3 months previous to the survey Final interview status Outlet Not Screened Interview interrupted Eligible respondent not available Outlet not open at the time Outlet closed permanently Refused Other Outlet did not meet screening criteria Outlet met screening criteria, but not interviewed (total) Interview interrupted Eligible respondent not available Outlet not open at the time Refused Other** Completed interview Partially completed interview Interview interrupted Eligible respondent not available Outlet not open at the time Refused Other Response rate (%)*** % % % Proportion of outlets enumerated that were screened Proportion of outlets meeting screening criteria that were interviewed * The number of outlets meeting the screening criteria is defined as the sum of the number of outlets stocking antimalarials at the time of the survey and the number of outlets without antimalarials in stock at the time of the survey, but who had antimalarials in stock at some time in the 3 months previous to the survey **The 8 classified as other did not have antimalarials in stock on the day of survey visit but stocked them at some time in the 3 months prior to the survey. These were not interviewed and classified as not meeting screening criteria at data collection. *** Response rate was calculated as outlets where final interview status was Completed interview or Partially completed interview as a percentage of all outlets meeting the screening criteria (i.e. flow diagram reference D divided by C). Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 29 P age

55 Table : Outlets enumerated [Kenya], 2011 Number of outlets enumerated, by location and type of outlet* Urban Rural Total Type of outlet Censused Locations Booster sample Total Censused Locations Booster sample Total Censused Locations Booster sample Total Public health facility Community health worker Private not for profit health facility Private for profit outlet Private for profit health facility Pharmacy Drug Store General retailer 6, ,467 4, ,779 11, ,246 Itinerant drug vendor Total 7, ,370 5, ,142 12, ,512 Total 7, ,646 5, ,727 12, ,373 * Flow diagram reference A. For three of the enumerated outlets, the outlet type was not indicated and this could not be ascertained. Therefore, there is a difference in N between table and Flow Diagram Reference A. Of these three outlets missing outlet type information, two outlets (1 urban and 1 rural) did not meet screening criteria while the other was a completed interview in an urban location Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 30 P age

56 Table : Outlets with antimalarials in stock* [Kenya], 2011 ** Number of outlets with antimalarials in stock at the time of the survey where an interview was conducted, by location and type of outlet Urban Rural Total Type of outlet Censused Locations Booster sample Total Censused Locations Booster sample Total Censused Locations Booster sample Public health facility Community health worker Total Private not for profit health facility Private for profit outlet Private for profit health facility Pharmacy Drug Store General retailer Itinerant drug vendor Total , ,378 Total , , ,855 * Flow diagram reference E. An interview was conducted if final interview status for an outlet was Completed interview or Partially completed. An outlet type for one of the completed interviews in an urban location was not indicated and could not be determined hence the difference between the figure on this table and diagram reference E ** These numbers form the denominator for all subsequent tables, unless specified otherwise. Any variation in the stated denominator in subsequent tables is due to missing data on specific variables. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 31 P age

57 Table : Number of products audited [Kenya], 2011 Number of products audited by outlet type, product type, and location Urban Rural Total Number of products audited Number of products audited Number of products audited Quality assured ACTs Public health facility 397 1,102 1,499 Community health worker Private not for profit health facility Private for profit outlet 1, ,048 Total 2,050 1,619 3,669 Non quality assured ACTs Public health facility Community health worker Private not for profit health facility Private for profit outlet 1, ,106 Total 1, ,153 Artemisinin monotherapy Public health facility Community health worker Private not for profit health facility Private for profit outlet Total Non Artemisinin therapy Public health facility ,041 Community health worker Private not for profit health facility Private for profit outlet 1, ,111 Total 1,671 1,590 3,261 All antimalarials Public health facility 725 1,880 2,605 Community health worker Private not for profit health facility Private for profit outlet 5,166 1,511 6,677 Total 5,992 3,552 9,544 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 32 P age

58 Table : Outlets with at least one staff member who completed secondary school or primary school [Kenya] 2011 Outlets with at least one staff member who completed secondary school or primary school* (n) as a percentage of all outlets with any antimalarials in stock at the time of the survey visit (N), by location and type of outlet. Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) At least one staff member completed primary school Public health facility N Community health worker Private not for profit health facility Private for profit outlet Private for profit health facility Pharmacy Drug Store General retailer 96.8 (90.3, 99.0) (85.7, 95.3) (87.9, 95.6) 377 Itinerant drug vendor Total 99.3 (97.7, 99.8) (93.0, 98.0) (95.1, 98.4) 1375 Total 99.3 (97.8, 99.8) (94.5, 98.3) (95.9, 98.6) 1852 At least one staff member completed secondary school Public health facility Community health worker Private not for profit health facility Private for profit outlet Private for profit health facility Pharmacy Drug Store (95.5, 99.9) (97.4, 99.9) 474 General retailer 69.6 (57.4, 79.5) (49.8, 68.4) (53.2, 68.8) 378 Itinerant drug vendor Total 93.3 (87.1, 96.6) (73.9, 86.9) (79.7, 89.2) 1376 Total 93.7 (87.9, 96.8) (78.9, 89.3) (82.8, 90.8) 1853 * The two groups are not mutually exclusive. Providers noted as having completed primary school include those who have completed secondary school and those who have not completed secondary school but who have completed primary school. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 33 P age

59 Table : Outlets with a staff member with a health related qualification [Kenya], 2011 Outlets with at least one staff member with a health related qualification* (n) as a percentage of all outlets with any antimalarials in stock at the time of the survey visit (N), by location and type of outlet. Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 99.5 (96.5, 99.9) (88.5, 99.7) (89.7, 99.7) 431 Community health worker Private not for profit health facility 97.7 (84.4, 99.7) (79.1, 99.4) (83.9, 99.3) 46 Private for profit outlet Private for profit health facility 95.8 (90.1, 98.3) (86.8, 99.8) (92.6, 99.1) 266 Pharmacy Drug Store 97.1 (93.1, 98.8) (82.0, 95.0) (88.3, 95.9) 474 General retailer 1.6 (0.4, 5.8) (1.1, 7.9) (1.1, 6.5) 372 Itinerant drug vendor Total 76.1 (67.0, 83.3) (39.1, 65.0) (50.3, 68.5) 1370 Total 77.4 (68.9, 84.2) (50.4, 70.3) (57.6, 72.6) 1847 * A health related qualification was defined as pharmacy, nurse or medical doctor related training. Pharmacy related training includes pharmacy studied to a certificate or diploma level; Nurse related training includes studying nursing to a certificate level (nurse aid) and diploma level; Medical doctor training includes clinical officers who studied medicine to a diploma level and fully qualified physicians. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 34 P age

60 3.2 Availability of antimalarial drugs Antimalarials in stock Table : Outlets with antimalarials in stock in [Kenya], 2011 Indicator 1.1 Outlets that had any antimalarials in stock at the time of the survey visit* (n) as a percentage of all outlets where screening questions were completed** (N), by location and type of outlet Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 95.0 (90.7, 97.4) (91.2, 99.0) (92.0, 98.8) 445 Community health worker Private not for profit health facility 80.4 (58.6, 92.2) (82.4, 99.7) (83.5, 97.7) 50 Total Public/Not for profit 26.1 (10.3, 51.9) (12.7, 51.3) (13.9, 48.1) 844 Private for profit outlet Private for profit health facility 90.2 (86.1, 93.2) (78.5, 96.0) (83.7, 94.4) 301 Pharmacy 98.7 (95.2, 99.7) (97.3, 99.8) 263 Drug Store 95.4 (90.9, 97.7) (91.1, 98.0) (92.7, 97.4) 503 General retailer 2.7 (1.7, 4.2) (3.3, 7.8) (3.1, 6.2) 9424 Itinerant drug vendor Total 11.2 (9.8, 12.7) (8.6, 12.6) (9.3, 12.2) Total 11.6 (10.2, 13.1) (10.1, 14.0) (10.4, 13.3) * Flow diagram reference E ** Flow diagram reference B. Screening questions asked whether outlets had any medicines in stock that day, or any antimalarials in stock that day, and if not whether they had had any medicines, or any antimalarials, in stock in the previous 3 months. For three of the enumerated outlets, the outlet type was not indicated and this could not be ascertained. Therefore, there is a difference in N between table and Flow Diagram Reference B. Of these three outlets missing outlet type information, two outlets (1 urban and 1 rural) did not meet screening criteria while the other was a completed interview in an urban location Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 35 P age

61 3.2.2 Antimalarials in stock by type Table : Outlets with non artemisinin therapy in stock Kenya, 2011 Indicator 1.2 Outlets that had non artemisinin monotherapy or non artemisinin combination therapy in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 85.8 (75.4, 92.3) (88.2, 98.0) (88.6, 97.3) 431 Community health worker Private not for profit health facility 67.3 (27.7, 91.7) (84.2, 99.7) (76.0, 97.6) 46 Total Public/Not for profit 77.2 (52.8, 91.1) (90.4, 98.1) (88.6, 96.6) 477 Private for profit outlet Private for profit health facility 61.2 (54.4, 67.6) (63.3, 84.6) (63.2, 76.5) 265 Pharmacy 73.9 (66.8, 79.9) (89.6, 99.3) (77.7, 90.3) 256 Drug Store 70.0 (59.6, 78.7) (77.0, 90.5) (72.6, 84.1) 473 General retailer 86.5 (72.0, 94.1) (70.1, 89.3) (73.0, 89.0) 378 Itinerant drug vendor Total 71.6 (64.4, 77.8) (76.7, 85.9) (74.7, 82.0) 1372 Total 71.9 (65.3, 77.7) (79.8, 88.0) (77.3, 84.0) 1849 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 36 P age

62 Table a: Outlets with artemisinin monotherapy in stock (ALL DOSAGE FORMS ) Kenya, 2011 Indicator 1.3 Outlets that had artemisinin monotherapy in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 13.3 (7.2, 23.4) (0.6, 5.1) (1.3, 5.3) 431 Community health worker Private not for profit health facility 43.5 (23.8, 65.4) (5.2, 32.0) (10.2, 34.6) 46 Total Public/Not for profit 27.4 (17.4, 40.4) (2.1, 9.2) (4.2, 11.2) 477 Private for profit outlets Private for profit health facility 28.6 (20.9, 37.8) (9.5, 23.5) (15.2, 26.2) (28.0, 31.6 (19.1, 47.3) 218 Pharmacy 85.8) (30.5, 61.4) 255 Drug Store 11.4 (6.1, 20.4) (3.3, 15.5) (5.5, 14.2) 473 General retailer Itinerant drug vendor Total 14.3 (9.6, 20.7) (4.5, 9.7) (7.0, 11.5) 1369 Total 15.0 (10.3, 21.4) (4.4, 8.8) (6.9, 10.9) 1846 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 37 P age

63 Table b: Outlets with ORAL artemisinin monotherapy in stock Kenya, 2011 Outlets that had oral artemisinin monotherapy in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility Community health worker Private not for profit health facility Total Public/Not for profit Private for profit outlets Private for profit health facility 1.1 (0.1, 8.5) (0.1, 2.9) 263 Pharmacy (0.4, 16.7) (0.2, 8.2) 254 Drug Store General retailer Itinerant drug vendor Total 0.3 (<0.1, 2.3) (<0.1, 0.4) (<0.1, 0.6) 1,367 Total 0.3 (<0.1, 2.2) 1044 <0.1 (<0.1, 0.3) (<0.1, 0.5) 1,844 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 38 P age

64 Table : Outlets with non quality assured ACTs in stock Kenya, 2011 Indicator 1.4 Outlets that had non quality assured ACTs in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 11.5 (5.6, 22.2) (2.7, 12.0) (3.3, 11.5) 431 Community health worker Private not for profit health facility 24.2 (9.5, 49.2) (1.3, 19.7) (3.9, 19.9) 46 Total Public/Not for profit 17.4 (9.1, 30.9) (3.0, 10.7) (4.3, 11.2) 477 Private for profit outlet Private for profit health facility 47.8 (36.9, 59.0) (16.9, 43.8) (27.2, 45.0) 264 Pharmacy 92.2 (86.7, 95.5) (70.4, 92.8) (81.0, 93.4) 256 Drug Store 54.0 (43.9, 63.8) (20.6, 41.6) (31.5, 48.4) 473 General retailer (0.6, 5.3) (0.5, 4.3) 377 Itinerant drug vendor Total 42.3 (34.4, 50.5) (12.8, 24.4) (21.0, 30.6) 1370 Total 40.8 (33.4, 48.7) (11.2, 21.3) (18.7, 27.3) 1847 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 39 P age

65 Table a: Outlets with quality assured ACTs in stock Kenya, 2011 Indicator 1.5 Outlets that had quality assured ACTs in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 92.0 (80.9, 96.9) (93.6, 98.9) (93.7, 98.6) 431 Community health worker Private not for profit health facility 93.9 (73.2, 98.9) (94.5, 99.7) 46 Total Public/Not for Profit 92.9 (84.2, 97.0) (95.3, 99.1) (95.1, 98.6) 477 Private for profit outlets Private for profit health facility 75.5 (64.7, 83.7) (65.3, 79.5) (67.8, 79.2) 265 Pharmacy 93.3 (90.6, 95.2) (76.3, 95.5) (86.2, 94.7) 256 Drug Store 91.9 (87.1, 95.0) (76.1, 89.8) (82.3, 90.9) 474 General retailer 14.6 (6.7, 29.1) (14.7, 41.8) (14.6, 36.9) 378 Itinerant drug vendor Total 70.6 (62.4, 77.7) (44.3, 65.9) (52.1, 67.7) 1373 Total 71.9 (63.9, 78.7) (54.4, 71.6) (59.0, 72.0) 1850 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 40 P age

66 Table b: Outlets with quality assured ACTs with and without the AMFm logo in stock Kenya, 2011 Indicator 1.5 Outlets that had quality assured ACTs with and without the AMFm logo in stock (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Stocked QAACTs with the AMFm logo Public health facility 69.6 (57.3, 79.7) (55.0, 79.1) (56.3, 78.4) 431 Community health worker Private not for profit health facility 55.8 (23.1, 84.1) (36.9, 78.8) (39.6, 75.7) 46 Total Public/Not for Profit 63.1 (43.7, 79.1) (53.2, 77.3) (54.3, 75.9) 477 Private for profit outlets Private for profit health facility 72.5 (63.4, 80.1) (63.9, 78.4) (66.2, 77.1) 265 Pharmacy 92.8 (89.4, 95.2) (76.3, 95.5) (85.8, 94.5) 256 Drug Store 88.9 (83.7, 92.6) (73.6, 88.0) (79.7, 88.7) 474 General retailer 14.6 (6.7, 29.1) (14.5, 41.1) (14.4, 36.3) 378 Itinerant drug vendor Total 68.4 (60.5, 75.5) (43.3, 64.8) (50.8, 66.1) 1373 Total 68.1 (60.3, 75.1) (47.0, 65.5) (52.7, 66.5) 1850 Stocked QAACTs without the AMFm logo Public health facility 55.6 (43.3, 67.3) (49.9, 77.8) (50.6, 76.2) 431 Community health worker Private not for profit health facility 46.0 (16.6, 78.4) (40.1, 80.4) (40.2, 75.7) 46 Total Public/Not for profit 51.1 (32.5, 69.4) (50.4, 76.5) (50.7, 74.0) 477 Private for profit outlets Private for profit health facility 8.0 (4.7, 13.5) (5.6, 17.1) (6.1, 13.9) 263 Pharmacy 10.6 (7.0, 15.6) (6.7, 20.8) (7.9, 15.8) 254 Drug Store 10.2 (7.4, 13.9) (2.1, 11.2) (4.5, 10.8) 473 General retailer (0.1, 1.8) (0.1, 1.5) 377 Itinerant drug vendor Total 7.4 (5.7, 9.5) (2.5, 6.4) (3.8, 6.8) 1367 Total 9.9 (8.6, 11.4) (12.1, 19.6) (11.5, 16.7) 1844 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 41 P age

67 Table c: Public health facility outlets with quality assured ACTs among ALL PUBLIC HEALTH FACILITIES in Kenya, 2011 Public health facilities that had quality assured ACTs in stock (n) as a percentage of ALL PUBLIC HEALTH FACILITIES screened (N), location Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 87.4 (77.6, 93.3) (89.4, 97.2) (89.5, 96.6) 445 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 42 P age

68 3.2.3 Stockouts of quality assured ACTs Table : Outlets with stock outs of quality assured ACTs [Kenya], 2011 Indicator 1.6 Outlets that were out of stock of all quality assured ACTs for at least 1 day in the last 7 days (n) as a percentage of outlets with any quality assured ACTs in stock at the time of the survey visit or in the 4 weeks preceding the survey visit (N), by location and type of outlet* Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 2.2 (0.9, 5.3) (1.4, 9.7) (1.4, 9.0) 412 Community health worker Private not for profit health facility 2.4 (0.3, 17.7) (0.6, 31.4) (0.7, 24.0) 42 Private for profit outlet Private for profit health facility 7.1 (2.6, 18.2) (1.3, 13.7) (2.4, 11.6) 228 Pharmacy 2.7 (1.2, 5.9) (0.6, 3.5) 241 Drug Store 5.7 (2.9, 11.0) (4.6, 16.2) (4.6, 12.1) 447 General retailer 1.5 (0.3, 6.6) (1.0, 9.0) (1.0, 7.7) 139 Itinerant drug vendor Total 5.6 (3.1, 10.1) (3.7, 9.3) (4.1, 8.3) 1055 Total 5.4 (3.0, 9.5) (3.7, 7.8) (4.0, 7.3) 1510 *This indicator measures stockouts of quality assured ACTs among outlets that have recently stocked these products. The denominator may include outlets which had no antimalarials in stock on the day of the survey but which had stocked them in the previous 3 months. A stockout is defined as being out of stock of all quality assured ACTs for at least 1 day in the last seven days. Outlets that have recently stocked QAACTs are defined as outlets with any QAACTs in stock at the time of the survey visit or in the 4 weeks preceding the survey visit. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 43 P age

69 3.2.4 Population coverage of outlets with quality assured ACTs Table : of the population living in censused locations with outlets with quality assured ACTs in stock at the time of survey [Kenya], 2011 Indicator 1.7: Population living in a censused locations where there was at least one of a given type of outlet with a qualityassured ACT in stock at the time of the survey visit (n) as a percentage of the total population living in all the censused locations (N), by location. Urban Rural Total (95% CI) N (95% CI) N (95% CI) N At least one public health facility stocking quality assured ACTs At least one private not for profit health facility stocking quality assured ACTs At least one private for profit outlet stocking quality assured ACTs At least one community health worker stocking quality assured ACTs 63.4 (38.5, 82.7) (76.9, 99.5) (79.0, 95.0) (23.4, 67.5) 47.7 (30.8, 65.1) 47.0 (32.9, 61.5) 97.9 (85.4, 99.7) 87.5 (70.0, 95.5) 89.7 (75.9, 96.0) At least one outlet of any type stocking quality assured ACTs Source of population data: Central Bureau of Statistics, Kenya, 1999 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 44 P age

70 3.3 Pricing of antimalarials (Affordability) Cost to patients of antimalarials Table : Cost to patients of non artemisinin therapy, in US dollars Kenya, 2011 Indicator 2.4: Median cost* to patients of one adult equivalent treatment dose (AETD)** of non artemisinin monotherapy or nonartemisinin combination therapy, by location, type of outlet and dosage form. Urban Rural Total Number of products Number of products Number of products Median cost Median cost Median cost Type of outlet [IQR] [IQR] [IQR] All dosage forms Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 1036 Community health workers Private not for profit health facility 2.07 [0.12, 9.74] [0.00, 7.31] [0.00, 7.31] 104 Private for profit outlet 1.73 [0.69, 9.79] [0.46, 7.31] [0.52, 7.31] 2045 Total 1.55 [0.58, 9.21] [0.23, 2.42] [0.35, 3.45] 3185 Tablets Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 655 Community health workers Private not for profit health facility 0.17 [0.00, 1.45] [0.00, 1.15] [0.00, 1.15] 57 Total Public/Not for profit 0.00 [0.00, 0.17] [0.00, 0.00] [0.00, 0.00] 712 Private for profit outlet 0.86 [0.52, 1.73] [0.35, 1.04] [0.40, 1.15] 1332 Total 0.86 [0.46, 1.73] [0.17, 0.86] [0.29, 1.04] 2044 Oral liquids Public health facility 7.31 [5.84, 14.61] [n/a] [0.00, 7.31] 13 Community health workers Private not for profit health facility Private for profit outlet [4.87, 12.14] [6.82, 24.35] [6.82, 24.35] [2.42, 20.24] [2.42, 20.24] [2.76, 17.05] [2.76, 17.05] [2.76, 19.48] [2.76, 19.48] 524 Total Injectables Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 368 Community health workers Private not for profit health facility [6.09, 17.05] [0.00, 12.18] [0.00, 12.18] 31 Private for profit outlet [6.09, 24.35] [7.31, 12.18] [7.31, 12.18] 218 Total 9.74 [4.87, 19.48] [0.00, 8.52] [0.00, 9.74] 617 Other Public health facility Community health workers Private not for profit health facility Private for profit outlet Total * 1 USD = KES; Source: Consumer Price Index; CPI 2010 = & CPI 2011 = ; Source: International Finance Statistics IMF) ** An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 45 P age

71 Table : Cost to patients of artemisinin monotherapy, in US dollars Kenya, 2011 Indicator 2.3 Median cost * to patients of one adult equivalent treatment dose (AETD)** of artemisinin monotherapy, by location, type of outlet and dosage form. Urban Rural Total Number of products Number of products Number of products Median cost Median cost Median cost Type of outlet [IQR] [IQR] [IQR] All dosage forms Public health facility 8.29 [0.00, 13.81] [0.00, 0.00] [0.00, 6.91] 32 Community health workers Private not for profit health [9.21, 27.63] [11.05, 16.58] [11.05, 22.10] 17 facility Private for profit outlet [11.05, 27.63] [13.26, 16.58] [11.60, 22.10] 382 Total [11.05, 27.63] [11.60, 16.58] [11.05, 22.10] 431 All oral dosage forms Public health facility Community health workers Private not for profit health facility Private for profit outlet 3.45 [3.45, 10.36] [n/a] [3.45, 10.36] 4 Total 3.45 [3.45, 10.36] [n/a] [3.45, 10.36] 4 Tablets Public health facility Community health workers Private not for profit health facility Total Public/Not for profit Private for profit outlet 3.45 [3.45, 3.45] [n/a] [3.45, 3.45] 3 Total 3.45 [3.45, 3.45] [n/a] [3.45, 3.45] 3 Oral liquids Public health facility Community health workers Private not for profit health facility Private for profit outlet [n/a] [n/a] 1 Total [n/a] [n/a] 1 Injectables Public health facility 8.29 [0.00, 13.81] [0.00, 0.00] [0.00, 6.91] 32 Community health workers Private not for profit health [9.21, 27.63] [11.05, 16.58] [11.05, 22.10] 17 facility Private for profit outlet [11.05, 27.63] [13.42, 16.58] [12.09, 22.56] 378 Total [11.05, 27.63] [11.60, 16.58] [11.05, 22.10] 427 Other Public health facility Community health workers Private not for profit health facility Private for profit outlet Total * 1 USD = KES; Source: Consumer Price Index; CPI 2010 = & CPI 2011 = ; Source: International Finance Statistics IMF) ** An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 46 P age

72 Table : Cost to patients of non quality assured ACTs, in US dollars Kenya, 2011 Indicator 2.2: Median cost* to patients of one adult equivalent treatment dose (AETD)** of non quality assured ACTs by location, type of outlet and dosage form. Urban Rural Total Number of products Number of products Number of products Median cost Median cost Median cost Type of outlet [IQR] [IQR] [IQR] All dosage forms Public health facility 2.30 [0.46, 6.14] [0.00, 0.00] [0.00, 0.00] 31 Community health workers Private not for profit health 6.14 [5.76, 10.36] [0.00, 0.00] [0.00, 7.67] 14 facility Private for profit outlet 7.52 [4.03, 11.94] [4.03, 11.67] [4.03, 11.67] 2044 Total 7.37 [4.03, 11.94] [3.63, 11.05] [4.03, 11.67] 2089 Tablets Public health facility 2.76 [0.00, 5.18] [0.00, 0.00] [0.00, 0.00] 16 Community health workers Private not for profit health 5.76 [3.45, 11.94] [0.00, 0.00] [0.00, 3.45] 6 facility Total Public/Not for profit 5.76 [2.88, 7.74] [0.00, 0.00] [0.00, 0.00] 22 Private for profit outlet 4.60 [3.45, 6.91] [2.88, 5.76] [3.45, 6.16] 1306 Total 4.60 [3.45, 6.91] [2.30, 5.41] [3.45, 5.99] 1328 Oral liquids Public health facility 2.30 [1.15, 10.28] [n/a] [0.00, 1.15] 13 Community health workers Private not for profit health 7.67 [6.14, 10.36] [6.14, 10.36] 8 facility Private for profit outlet [9.21, 15.96] [10.74, 17.19] [9.21, 16.58] 717 Total [9.21, 15.96] [10.44, 17.19] [9.21, 16.58] 738 Injectables Public health facility Community health workers Private not for profit health facility Private for profit outlet Total Other*** Public health facility 0.00 [n/a] [n/a] [0.00, 0.00] 2 Community health workers Private not for profit health facility Private for profit outlet [13.81, 26.71] [12.89, 13.81] [13.81, 21.41] 21 Total [13.81, 26.71] [0.00, 13.81] [12.89, 18.42] 23 * 1 USD = KES; Source: Consumer Price Index; CPI 2010 = & CPI 2011 = ; Source: International Finance Statistics IMF) ** An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial *** The other dosage form includes suppositories and granules. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 47 P age

73 Table : Cost to patients of quality assured ACTs, in US dollars Kenya, 2011 Indicator 2.1: Median cost* to patients of one adult equivalent treatment dose (AETD)** of quality assured ACTs, by presence of the AMFm logo, location and type of outlet Urban Rural Total Type of outlet Median cost [IQR] Number of products Median cost [IQR] Number of products Median cost [IQR] Number of products Adult equivalent treatment dose (AETD)* All QAACTs Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 1498 Community health worker Private not for profit health facility 0.46 [0.00, 0.92] [0.00, 0.77] [0.00, 0.77] 122 Total Public/Not for Profit 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 1620 Private for profit outlet 0.61 [0.46, 1.15] [0.46, 0.92] [0.46, 0.92] 1984 Total 0.58 [0.46, 1.15] [0.00, 0.46] [0.00, 0.69] 3604 QAACTs with the AMFm logo Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 704 Community health worker Private not for profit health facility 0.46 [0.00, 1.15] [0.00, 0.69] [0.00, 0.69] 48 Total Public/Not for Profit 0.00 [0.00, 0.46] [0.00, 0.00] [0.00, 0.00] 752 Private for profit outlet 0.61 [0.46, 1.15] [0.46, 0.81] [0.46, 0.92] 1887 Total 0.58 [0.46, 1.15] [0.00, 0.58] [0.46, 0.81] 2639 QAACTs without the AMFm logo Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 792 Community health worker Private not for profit health facility 0.00 [0.00, 0.69] [0.00, 1.15] [0.00, 0.92] 74 Total Public/Not for Profit 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 866 Private for profit outlet 0.92 [0.58, 2.30] [0.46, 3.45] [0.46, 2.30] 97 Total 0.46 [0.00, 0.92] [0.00, 0.00] [0.00, 0.00] 963 Pediatric formulation Pack for a two year old child (10kg)*** All QAACTs Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 185 Community health worker Private not for profit health facility 0.00 [0.00, 0.46] [0.00, 0.00] [0.00, 0.46] 15 Total Public/Not for Profit 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 200 Private for profit outlet 0.46 [0.35, 0.52] [0.32, 0.46] [0.35, 0.46] 317 Total 0.46 [0.35, 0.46] [0.00, 0.35] [0.00, 0.46] 517 QAACTs with the AMFm logo Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 97 Community health worker Private not for profit health facility 0.46 [0.46, 0.46] [0.00, 0.58] [0.00, 0.58] 6 Total Public/Not for Profit 0.00 [0.00, 1.61] [0.00, 0.00] [0.00, 0.00] 103 Private for profit outlet 0.46 [0.35, 0.46] [0.32, 0.46] [0.35, 0.46] 304 Total 0.46 [0.35, 0.46] [0.00, 0.46] [0.00, 0.46] 407 QAACTs without the AMFm logo Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 88 Community health worker Private not for profit health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 9 Total Public/Not for Profit 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 97 Private for profit outlet 0.58 [0.46, 0.58] [0.35, 0.46] [0.46, 0.58] 13 Total 0.46 [0.00, 0.58] [0.00, 0.00] [0.00, 0.00] 110 * 1 USD = KES; Source: Consumer Price Index; CPI 2010 = & CPI 2011 = ; Source: International Finance Statistics IMF) ** An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial. ***Pediatric formulations (PFs) are packages intended for children. In the calculation of median cost we include only packages whose age (weight) range includes a 2 year old (10kg) child. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 48 P age

74 Table a : Cost to patients of the most popular antimalarial in terms of national private for profit outlet sales volumes (SP) for ALL DOSAGE TYPES, in US dollars Kenya, 2011 Indicator 2.4: Median cost to patients of one adult equivalent treatment dose (AETD)* of the most popular antimalarial in terms of national private for profit sales volumes in Kenya that is not a quality assured ACT, for ALL DOSAGE TYPES (SP), by location and type of outlet Urban Rural Total Type of outlet Median cost [IQR] Number of products Median cost [IQR] Number of products Median cost [IQR] Number of products 0.69 [0.46, 1.38] [0.35, 0.69] [0.35, 0.92] 981 Private for profit outlet * An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) Table b: Cost to patients of the most popular antimalarial in terms of national private for profit outlet sales volumes (SP) for TABLETS, in US dollars Kenya, 2011 Indicator 2.4: Median cost to patients of one adult equivalent treatment dose (AETD)* of the most popular antimalarial in terms of national private for profit sales volumes in Kenya that is not a quality assured ACT, for TABLETS (SP), by location and type of outlet Urban Rural Total Number of products Number of products Median cost Median cost Median cost Type of outlet [IQR] [IQR] [IQR] 0.52 [0.35, 0.69 [0.46, 1.38] [0.35, 0.69] 364 Private for profit outlet 0.86] 973 * An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) Number of products Table c: Cost to patients of the most popular antimalarial in terms of national outlet sales volumes for all outlet types (SP) for ALL DOSAGE TYPES, in US dollars Kenya, 2011 Indicator 2.4: Median cost to patients of one adult equivalent treatment dose (AETD)* of the most popular antimalarial in terms of national sales volumes for all outlet types in Kenya that is not a quality assured ACT, for ALL DOSAGE TYPES (SP), by location and type of outlet Urban Rural Total Type of outlet Median cost [IQR] Number of products Median cost [IQR] Number of products Median cost [IQR] Number of products 0.69 [0.46, 1.38] [0.35, 0.60] [0.35, 0.86] 1307 All outlets * An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) Table d: Cost to patients of the most popular antimalarial in terms of national outlet sales volumes for all outlet types (SP) for TABLETS, in US dollars Kenya, 2011 Indicator 2.4: Median cost to patients of one adult equivalent treatment dose (AETD)* of the most popular antimalarial in terms of national sales volumes for all outlet types in Kenya that is not a quality assured ACT, for TABLETS (SP), by location and type of outlet Urban Rural Total Type of outlet Median cost [IQR] Number of products Median cost [IQR] Number of products Median cost [IQR] Number of products 0.69 [0.46, 1.38] [0.35, 0.58] [0.35, 0.78] 1299 All outlets * An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 49 P age

75 3.3.2 Gross percentage markup between purchase price and retail selling price Table : Gross percentage markup between purchase price and retail selling price of non artemisinin therapy [Kenya], 2011 Median percentage markup between purchase price and retail selling price of non artemisinin monotherapy or non artemisinin combination therapy by location and type of outlet Type of outlet Urban Rural Total Number Median markup [IQR] Number of products Median markup [IQR] of product s Median markup [IQR] Number of products Public health facility 0.0 [0.0, 0.0] [0.0, 0.0] [0.0, 0.0] 969 Community health worke Private not for profit health facility 40.0 [0.0, 71.4] [0.0, 76.5] [0.0, 75.8] 84 Private for profit outlet Private for profit health 66.7 [33.3, 62.2 [33.3, 150.0] [33.3, 100.0] 91 facility 100.0] 259 Pharmacy 42.9 [30.0, 66.7] [29.4, 66.7] [30.0, 66.7] 417 Drug Store 57.9 [31.6, 100.0] [30.0, 75.0] [31.6, 76.5] 697 General retailer 33.3 [20.0, 78.6] [20.0, 66.7] [20.0, 66.7] 223 Itinerant drug vendor Total 53.8 [30.2, 100.0] [25.0, 81.8] [28.6, 87.5] 1596 Total 50.0 [25.0, 100.0] [0.0, 66.7] [0.0, 75.0] 2649 Note: 44 markups were treated as missing, because the purchase price was zero and the selling price was non zero. Refer to section 0 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 50 P age

76 Table : Gross percentage markup between purchase price and retail selling price of artemisinin monotherapy [Kenya], 2011 Median percentage markup between purchase price and retail selling price of artemisinin monotherapy by location and type of outlet Type of outlet Median markup [IQR] Urban Rural Total Number of products Median markup [IQR] Number of products Median markup [IQR] Public health facility 9.1 [0.0, 28.6] [0.0, 0.0] [0.0, 0.0] 22 Number of products Community health worker Private not for profit health facility 42.9 [31.9, 50.9] [33.3, 207.7] [31.9, 207.7] 14 Private for profit outlet Private for profit health 36.4 [25.0, 66.7] facility [30.8, 50.0] [25.0, 60.0] 84 Pharmacy 33.3 [25.0, 50.0] [15.4, 36.4] [19.8, 38.5] 117 Drug Store 33.3 [25.0, 40.0] [10.0, 12.0] [12.0, 35.0] 122 General retailer Itinerant drug vendor Total 33.3 [25.0, 50.0] [12.0, 50.0] [20.0, 50.0] 323 Total 33.3 [25.0, 50.0] [12.0, 50.0] [18.8, 50.0] 359 Note: 8 markups were treated as missing, because the purchase price was zero and the selling price was non zero. Refer to section 0 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 51 P age

77 Table : Gross percentage markup between purchase price and retail selling price of non quality assured ACTs [Kenya], 2011 Median percentage markup between purchase price and retail selling price of non quality assured ACTs by location and type of outlet Type of outlet Median markup [IQR] Urban Rural Total Number of products Median markup [IQR] Number of products Median markup [IQR] Public health facility 9.1 [0.0, 25.0] [0.0, 0.0] [0.0, 0.0] 22 Number of products Community health worker Private not for profit health facility [11.1, 25.0] [0.0, 0.0] [0.0, 17.6] 14 Private for profit outlet Private for profit health facility 29.6 [17.6, 36.4] [25.0, 66.7] [21.6, 50.0] 180 Pharmacy 32.4 [24.5, 37.5] [18.4, 36.4] [21.7, 36.4] 828 Drug Store 31.0 [20.0, 38.9] [19.0, 48.9] [20.0, 40.0] 655 General retailer [14.3, 42.9] [14.3, 42.9] 3 Itinerant drug vendor Total 31.0 [20.0, 38.5] [20.7, 48.9] [20.5, 42.9] 1666 Total 30.6 [20.0, 37.6] [19.0, 48.9] [20.0, 41.5] 1702 Note: 6 markups were treated as missing, because the purchase price was zero and the selling price was non zero. Refer to section 0 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 52 P age

78 Table a: Gross percentage markup between purchase price and retail selling price of quality assured ACTs [Kenya], 2011 Indicator 2.5: Median percentage markup between purchase price and retail selling price of quality assured ACTs by location and type of outlet Type of outlet Median markup [IQR] Urban Rural Total Number of products Median markup [IQR] Number of products Median markup [IQR] Number of products Public health facility 0.0 [0.0, 0.0] [0.0, 0.0] [0.0, 0.0] 1441 Community health worker Private not for profit health facility [0.0, 14.3] [0.0, 25.0] [0.0, 18.4] 98 Private for profit outlet Private for profit health 60.0 [33.3, 42.9 [33.3, 50.0 [33.3, facility 100.0] 100.0] 100.0] 267 Pharmacy 42.9 [33.3, 66.7] [33.3, 66.7] [33.3, 66.7] 563 Drug Store 48.1 [33.3, 77.8] [33.3, 73.9] [33.3, 73.9] 769 General retailer 33.3 [14.3, 53.8] [14.3, 33.3] [14.3, 33.3] 90 Itinerant drug vendor Total 50.0 [33.3, 81.8] [33.3, 73.9] [33.3, 73.9] 1689 Total 42.9 [33.3, 72.4] [0.0, 42.9] [0.0, 60.0] 3228 Note: 56 markups were treated as missing, because the purchase price was zero and the selling price was non zero. Refer to section 0 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 53 P age

79 Table b: Gross percentage markup between purchase price and retail selling price of quality assured ACTs, by presence of the AMFm logo, [Kenya], 2011 Indicator 2.5: Median percentage markup between purchase price and retail selling price of quality assured ACTs by presence of the AMFm logo by location and type of outlet Type of outlet Median markup [IQR] Urban Rural Total Number of products Median markup [IQR] Number of products Median markup [IQR] Number of products QAACTs with the AMFm logo Public health facility 0.0 [0.0, 0.0] [0.0, 0.0] [0.0, 0.0] 665 Community health worker Private not for profit health facility 13.6 [ 100.0, 33.3] [0.0, 57.9] [0.0, 50.0] 40 Private for profit outlet Private for profit health 60.0 [33.3, 60.0 [33.3, 100.0] [33.3, 100.0] 60 facility 100.0] 251 Pharmacy 42.9 [33.3, 66.7] [33.3, 66.7] [33.3, 66.7] 543 Drug Store 48.1 [33.3, 78.6] [33.3, 73.9] [33.3, 73.9] 738 General retailer 33.3 [14.3, 53.8] [14.3, 33.3] [14.3, 33.3] 88 Itinerant drug vendor Total 50.0 [33.3, 81.8] [33.3, 73.9] [33.3, 75.0] 1620 Total 42.9 [33.3, 78.6] [0.0, 60.0] [0.0, 66.7] 2325 QAACTs without the AMFm logo Public health facility 0.0 [0.0, 0.0] [0.0, 0.0] [0.0, 0.0] 775 Community health worker Private not for profit health facility 0.0 [0.0, 0.0] [0.0, 0.0] [0.0, 0.0] 58 Private for profit outlet Private for profit health facility 28.6 [11.1, 100.0] [0.0, 39.5] [0.0, 39.5] 16 Pharmacy 32.6 [22.2, 38.9] [25.0, 150.0] [25.0, 47.7] 20 Drug Store 42.9 [33.3, 71.4] [33.3, 66.7] [33.3, 66.7] 31 General retailer [33.3, 50.0] [33.3, 50.0] 2 Itinerant drug vendor Total 40.0 [14.3, 71.4] [33.3, 50.0] [15.4, 66.7] 69 Total 0.0 [0.0, 33.3] [0.0, 0.0] [0.0, 0.0] 902 Note: 56 markups were treated as missing, because the purchase price was zero and the selling price was non zero. Refer to section 0 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 54 P age

80 Table : Median total gross markup between first line buyer price and retail selling price of quality assured ACTs bearing the AMFm logo, in 2010 US dollars, [Kenya], 2011 Indicator 2.6: Median total gross markup* between first line buyer price** and retail selling price per adult equivalent treatment dose (AETD)*** of quality assured ACTs bearing the AMFm logo, by location and type of outlet Urban Rural Total Number Number of of products products Number of products Median markup Median markup Median markup Type of outlet [IQR] [IQR] [IQR] Public health facility 0.05 [ 0.11, [ 0.11, 0.04] [ 0.11, 0.04] 621 Community health worker Private not for profit health facility [ 0.11, 0.99] [ 0.05, 0.53] [ 0.05, 0.62] 46 Private for profit outlet Private for profit health facility 0.58 [0.35, 1.13] [0.35, 0.99] [0.35, 1.04] 295 Pharmacy 0.52 [0.39, 1.11] [0.35, 1.25] [0.35, 1.11] 639 Drug Store 0.52 [0.35, 1.09] [0.30, 0.53] [0.35, 0.89] 839 General retailer 0.35 [0.30, 0.63] [0.30, 0.35] [0.30, 0.35] 108 Itinerant drug vendor Total 0.52 [0.35, 1.09] [0.30, 0.66] [0.30, 0.89] 1881 Total 0.50 [0.35, 1.08] [ 0.03, 0.46] [0.30, 0.73] 2548 *Median total gross markup is the median of the difference between the retail selling price and the mean first line buyer price for each QAACT. **First Line Buyer (FLB) price data were provided by The Global Fund. *** An AETD is the number of milligrams (mg) of a given drug that is required to treat a 60 kg adult. AETDs were calculated for every audited antimalarial Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 55 P age

81 3.3.3 Availability and cost to patients of diagnostic tests (RDT/microscopy) Any diagnostic test Table : Availability of any diagnostic test for malaria, [Kenya], 2011 Outlets where any diagnostic tests (microscopy or RDT) were available* (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit** (N), by location and type of outlet. Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 69.9 (58.5, 79.3) (45.7, 63.6) (47.6, 64.0) 431 Community health worker Private not for profit health facility 91.5 (62.9, 98.5) (61.1, 90.4) (66.5, 91.0) 46 Total Public/Not for profit 80.0 (67.6, 88.4) (52.2, 68.0) (55.1, 69.2) 477 Private for profit outlet Private for profit health facility 66.1 (54.9, 75.7) (23.2, 47.2) (37.1, 54.8) 263 Pharmacy 16.9 (9.9, 27.2) (8.6, 38.3) (11.1, 28.1) 255 Drug Store 13.8 (9.9, 18.7) (2.6, 12.3) (6.0, 13.3) 471 General retailer Itinerant drug vendor Total 24.3 (20.1, 29.0) (5.7, 14.4) (10.9, 17.4) 1364 Total 27.5 (23.1, 32.4) (15.5, 22.7) (18.5, 24.3) 1841 * Malaria microscopic testing is considered to be available if the respondent reported that the service is available in the outlet on the day of the survey visit ** Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 56 P age

82 Malaria microscopy Table : Availability of malaria microscopy, [Kenya], 2011 Outlets where malaria microscopic tests were available* (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit** (N), by location and type of outlet Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 58.4 (45.9, 69.8) (32.2, 49.8) (34.1, 50.4) 431 Community health worker Private not for profit health facility 80.0 (51.9, 93.7) (61.1, 90.4) (64.6, 89.1) 46 Total Public/Not for profit 68.5 (55.2, 79.3) (39.8, 59.1) (42.7, 60.1) 477 Private for profit outlet Private for profit health facility 64.1 (54.4, 72.7) (17.0, 38.4) (31.2, 49.6) 263 Pharmacy 8.3 (4.4, 15.2) (7.0, 34.3) (6.4, 22.2) 255 Drug Store 11.0 (8.5, 14.2) (2.0, 10.8) (4.8, 10.8) 471 General retailer Itinerant drug vendor Total 22.1 (18.2, 26.5) (4.6, 11.1) (9.4, 14.7) 1364 Total 24.7 (20.5, 29.5) (12.4, 18.4) (15.4, 20.6) 1841 * Malaria microscopic testing is considered to be available if the respondent reported that the service is available in the outlet on the day of the survey visit ** Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 57 P age

83 Table : Cost to patients of malaria microscopy in 2010 US dollars [Kenya], 2011 Median cost to patients of one malaria diagnostic test with microscopy, by outlet type and patient age Urban Rural Total Type of outlet Median cost [IQR] Number of outlets reporting price of malaria microscopy Median cost [IQR] Number of outlets reporting price of malaria microscopy Median cost [IQR] Number of outlets reporting price of malaria microscopy Adults Public health facility 0.58 [0.35, 0.58] [0.35, 0.58] [0.35, 0.58] 220 Community health worker Private not for profit healt facility 0.58 [0.58, 0.92] [0.46, 0.58] [0.46, 0.58] 36 Private for profit outlet Private for profit health facility 1.15 [0.58, 1.15] [0.58, 1.15] [0.58, 1.15] 149 Pharmacy 1.15 [0.58, 1.73] [0.58, 1.15] [0.58, 1.15] 26 Drug Store 1.15 [0.58, 1.15] [0.58, 0.58] [0.58, 1.15] 29 General retailer Itinerant drug vendor Total 1.15 [0.58, 1.15] [0.58, 0.92] [0.58, 1.15] 204 Total 0.92 [0.58, 1.15] [0.46, 0.58] [0.46, 1.04] 460 Children Public health facility 0.23 [0.00, 0.58] [0.00, 0.58] [0.00, 0.58] 221 Community health worker Private not for profit healt facility 0.58 [0.58, 0.92] [0.35, 0.58] [0.35, 0.58] 36 Private for profit outlet Private for profit health facility 0.81 [0.58, 1.15] [0.46, 0.58] [0.58, 1.15] 152 Pharmacy 1.15 [0.58, 1.15] [0.58, 1.15] [0.58, 1.15] 25 Drug Store 0.92 [0.58, 1.15] [0.58, 0.58] [0.58, 1.15] 31 General retailer Itinerant drug vendor Total 0.92 [0.58, 1.15] [0.58, 0.58] [0.58, 1.15] 208 Total 0.58 [0.58, 1.15] [0.23, 0.58] [0.35, 0.58] 465 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 58 P age

84 Diagnostic test with rapid diagnostic tests Table : Availability of rapid diagnostic tests for malaria, [Kenya], 2011 Outlets where rapid diagnostic tests were available (n) as a percentage of outlets with any antimalarials in stock at the time of the survey visit* (N), by location and type of outlet. Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 17.6 (8.6, 32.7) (9.2, 37.6) (9.7, 35.7) 431 Community health worker Private not for profit health facility 32.1 (15.1, 55.7) (0.6, 20.9) (4.3, 19.6) 46 Total Public/Not for profit 24.4 (15.0, 37.1) (7.8, 30.6) (9.3, 29.3) 477 Private for profit outlet Private for profit health facility 7.2 (2.8, 17.4) (3.5, 16.7) (4.2, 13.6) 264 Pharmacy 10.3 (5.3, 19.2) (0.8, 10.5) (3.8, 11.8) 255 Drug Store 5.7 (2.5, 12.5) (0.1, 8.0) (1.3, 6.7) 474 General retailer Itinerant drug vendor Total 5.0 (3.0, 8.4) (0.8, 4.8) (1.8, 4.8) 1371 Total 6.1 (4.4, 8.5) (2.6, 8.1) (3.4, 7.4) 1848 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 59 P age

85 Table : Cost to patients of rapid diagnostic tests (RDTs) for malaria in 2010 US dollars [Kenya], 2011 Median cost to patients of one rapid diagnostic test for malaria, by location, outlet type and patient age Urban Rural Total Type of outlet Median cost [IQR] Number of RDT products Median cost [IQR] Number of RDT products Median cost [IQR] Adults Number of RDT products Public health facility 0.00 [0.00, 0.00] [0.00, 0.35] [0.00, 0.35] 68 Community health worker Private not for profit health facility 0.58 [0.58, 1.15] [0.35, 0.35] [0.35, 0.58] 7 Private for profit outlet Private for profit health facility 1.15 [0.75, 1.15] [0.35, 1.15] [0.58, 1.15] 22 Pharmacy 1.73 [1.15, 1.73] [1.73, 1.73] [1.15, 1.73] 29 Drug Store 1.15 [1.15, 1.73] [0.58, 0.58] [0.58, 1.15] 15 General retailer Itinerant drug vendor Total 1.15 [1.15, 1.73] [0.35, 1.15] [0.58, 1.15] 66 Total 1.15 [0.58, 1.15] [0.00, 0.58] [0.00, 1.15] 141 Children Public health facility 0.00 [0.00, 0.00] [0.00, 0.00] [0.00, 0.00] 67 Community health worker Private not for profit healt facility 0.58 [0.58, 1.15] [0.35, 0.35] [0.35, 0.58] 7 Private for profit outlet Private for profit health facility 1.15 [0.75, 1.15] [0.23, 1.15] [0.58, 1.15] 22 Pharmacy 1.38 [1.15, 1.73] [1.15, 1.15] [1.15, 1.73] 29 Drug Store 1.15 [0.58, 1.15] [0.58, 0.58] [0.58, 1.15] 15 General retailer Itinerant drug vendor Total 1.15 [0.75, 1.15] [0.23, 1.15] [0.58, 1.15] 66 Total 1.15 [0.58, 1.15] [0.00, 0.58] [0.00, 1.15] 140 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 60 P age

86 3.3 Quality assured ACTs market share Table : breakdown of antimalarial sales volumes by antimalarial type, Kenya, 2011 Indicator 4.1: Total number of AETDs of each antimalarial type sold or distributed in the week preceding the survey visit (n), as a percentage all antimalarial AETDs sold or distributed in the week preceding the survey visit by outlets with any antimalarials in stock at the time of the survey visit (N), by location and type of outlet Urban Rural Total N N N Public health facility Quality assured ACTs Non quality assured ACTs Oral artemisinin monotherapy Non oral artemisinin monotherapy Non artemisinin therapy Total 100 2, , ,726.3 Community health worker Quality assured ACTs Non quality assured ACTs Oral artemisinin monotherapy Non oral artemisinin monotherapy Non artemisinin therapy Total Private not for profit health facility Quality assured ACTs Non quality assured ACTs Oral artemisinin monotherapy Non oral artemisinin monotherapy Non artemisinin therapy Total Private for profit outlet Quality assured ACTs Non quality assured ACTs Oral artemisinin monotherapy Non oral artemisinin monotherapy Non artemisinin therapy Total , , ,761.4 All outlets Quality assured ACTs Non quality assured ACTs Oral artemisinin monotherapy Non oral artemisinin monotherapy Non artemisinin therapy Total , , ,464.1 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 61 P age

87 Table : Market share of quality assured ACTs (QAACTs), for all QAACTs, QAACTs with the AMFm logo and QAACTs without the AMFm logo, Kenya, 2011 Indicator 4.1: Total number of AETDs of QAACTs sold or distributed in the week preceding the survey visit (n), as a percentage all antimalarial AETDs sold or distributed in the week preceding the survey visit by outlets with any antimalarials in stock at the time of the survey visit (N), for all QAACTs, QAACTs with the AMFm logo and QAACTs without the AMFm logo, by location and type of outlet Urban Rural Total N N N Public health facility All QAACTs QAACTs with logo QAACTs without logo Total number of AEDTS sold 2, , ,726.3 Community health worker All QAACTs QAACTs with logo QAACTs without logo Total number of AEDTS sold Private not for profit health facility All QAACTs QAACTs with logo QAACTs without logo Total number of AEDTS sold Private for profit outlet All QAACTs QAACTs with logo QAACTs without logo Total number of AEDTS sold 22, , ,761.4 All outlets All QAACTs QAACTs with logo QAACTs without logo Total number of AEDTS sold 25, , ,464.1 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 62 P age

88 Table : breakdown of antimalarial sales volumes by outlet type, Kenya, 2011 Total number of AETDs sold or distributed in the week preceding the survey visit by each outlet type (n), as a percentage of all antimalarial AETDs sold or distributed in the week preceding the survey visit by all outlets with any antimalarials in stock at the time of the survey visit (N), by location and antimalarial type Urban Rural Total N N N Quality assured ACTs Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total , , ,151.7 Non quality assured ACTs Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total 100 3, ,073.6 Oral artemisinin therapies Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total Non oral artemisinin therapies Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total Non artemisinin therapies Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total 100 5, , ,714.8 All antimalarials Public health facilities Community health workers Private not for profit health facilities Private for profit outlets Total , , ,464.1 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 63 P age

89 3.5 Provider knowledge of first line antimalarial treatment and ACT regimen Table : Provider knowledge of first line antimalarial treatment, [Kenya], 2011 Providers able to correctly identify the antimalarial for first line treatment (n) as a percentage of outlets with antimalarials in stock at the time of the survey visit* (N), by location and type of outlet Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 94.8 (86.4, 98.2) (94.8, 99.1) (95.0, 98.9) 431 Community health worker Private not for profit health facility 97.7 (84.4, 99.7) (70.9, 95.5) (75.6, 96.2) 46 Total Public/Not for Profit 96.2 (90.2, 98.6) (91.1, 97.9) (91.7, 97.8) 477 Private for profit outlet Private for profit health facility 91.8 (85.9, 95.3) (74.0, 89.5) (80.3, 90.7) 264 Pharmacy 92.0 (86.5, 95.4) (80.2, 95.1) (85.8, 94.4) 257 Drug Store 88.0 (85.1, 90.5) (79.8, 90.8) (83.2, 90.0) 473 General retailer 26.0 (13.8, 43.6) (20.9, 49.9) (21.6, 45.7) 376 Itinerant drug vendor Total 75.5 (67.2, 82.2) (50.3, 72.1) (57.9, 73.4) 1370 Total 76.6 (68.5, 83.2) (58.6, 76.5) (63.4, 76.8) 1847 * Flow diagram reference E Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 64 P age

90 Table : Provider knowledge of dosing regimen for quality assured ACTs (QAACTs) for an adult. [Kenya], 2011 Providers able to correctly describe the dosing regimen for quality assured ACTs for an adult (n) as a percentage of the number of outlets with QAACTs in stock at the time of the survey visit (N), by location and type of outlet* Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 94.8 (90.6, 97.2) (85.2, 93.8) (86.0, 93.9) 411 Community health worker Private not for profit health facility (67.6, 99.1) (72.8, 99.3) 43 Total Public/Not for profit 97.3 (94.8, 98.6) (85.9, 94.6) (87.1, 94.9) 454 Private for profit outlet Private for profit health facility 86.9 (81.2, 91.0) (87.5, 99.2) (87.3, 96.3) 199 Pharmacy 92.4 (87.3, 95.6) (91.7, 98.0) 232 Drug Store 92.3 (81.8, 96.9) (88.8, 98.4) (89.1, 97.0) 422 General retailer 74.2 (52.4, 88.3) (61.5, 85.5) (63.1, 84.4) 96 Itinerant drug vendor Total 90.0 (83.4, 94.1) (84.8, 95.7) (86.6, 94.2) 949 Total 90.5 (84.6, 94.3) (86.3, 95.0) (87.5, 94.0) 1403 * Correctly describe implies that the respondent correctly stated the number of tablets that should be taken at a time, the number of times the medicine should be taken per day and the duration of the dose in number of days for a 60kg adult for a specific product which they selected from the QAACTs that they stocked. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 65 P age

91 Table : Provider knowledge of dosing regimen for quality assured ACTs (QAACTs) for a child, [Kenya], 2011 Providers able to correctly describe the dosing regimen for quality assured ACT for a child (n) as a percentage of the number of outlets with QAACTs in stock at the time of the survey visit (N), by location and type of outlet* Type of outlet Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Public health facility 88.1 (80.4, 93.0) (81.2, 93.2) (81.9, 92.8) 412 Community health worker Private not for profit health facility 94.6 (77.5, 98.9) (54.1, 94.0) (61.2, 94.3) 43 Total Public/Not for profit 91.2 (84.0, 95.4) (79.4, 91.8) (80.7, 91.8) 455 Private for profit outlet Private for profit health facility 59.4 (45.1, 72.3) (54.5, 71.7) (54.3, 69.1) 198 Pharmacy 71.1 (60.7, 79.6) (65.2, 86.0) (66.3, 80.4) 235 Drug Store 70.0 (62.0, 76.9) (48.9, 70.6) (56.7, 71.3) 425 General retailer 40.4 (18.9, 66.3) (28.1, 56.7) (29.2, 55.0) 97 Itinerant drug vendor Total 65.8 (57.6, 73.1) (51.0, 64.1) (55.4, 65.7) 955 Community health worker Total 67.7 (60.4, 74.2) (59.8, 72.0) (61.8, 71.1) 1410 * Correctly describe implies that the respondent correctly stated the number of tablets that should be taken at a time, the number of times the medicine should be taken per day and the duration of the dose in number of days for child under 2 years (10kg) for a specific product which they selected from the QAACTs that they stocked. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 66 P age

92 Table : Reasons for not stocking quality assured ACTs (QAACTs) by private providers, [Kenya], 2011 Private for profit providers stating a specific reason for why they were not stocking QAACTs (n) as a percentage of all private forprofit outlets* not stocking QAACTs at the time of the survey visit** (N), by location Reason Urban Rural Total (95% CI) N (95% CI) N (95% CI) N Too expensive 12.4 (9.4, 16.1) (7.5, 14.7) (8.4, 14.1) 615 Not profitable 3.7 (1.9, 7.0) 3.2 (1.9, 5.6) 3.3 (2.2, 5.1) The outlet is not allowed to sell them 9.3 (6.0, 14.0) 11.4 (8.8, 14.7) 11.0 (8.8, 13.6) They have too many side effects 1.6 (0.4, 6.2) 1.5 (0.4, 5.7) 1.5 (0.5, 4.4) They do not work well 4.5 (2.1, 9.2) 5.6 (1.9, 15.3) 5.3 (2.2, 12.4) They are not available/my suppliers do 27.5 (20.0, 25.3 (20.0, 31.3) 28.1 (18.8, 39.7) not have it in stock 36.5) 15.9 (11.9, 18.8 (15.3, 22.9) 15.1 (10.4, 21.5) My customers do not ask for them 20.8) 26.3 (21.5, 22.4 (17.5, 28.2) 27.3 (21.5, 34.1) I don t know about these drugs 31.7) I am temporarily out of stock 18.0 (12.5, 25.2) 12.0 (7.2, 19.3) 13.2 (9.0, 18.9) Other Reason: Other 1.8 (0.6, 5.5) 0.7 (0.1, 3.4) 0.9 (0.3, 2.6) Other Reason: Not responsible for stock (0.4, 3.5) 0.9 (0.3, 2.8) Other Reason: Customers get free from PHF 0.1 (0.0, 0.9) 0.5 (0.1, 2.3) 0.4 (0.1, 1.7) Other Reason: New outlet, not yet fully stocked 1.1 (0.3, 3.6) 0.8 (0.2, 3.3) 0.9 (0.3, 2.5) Other Reason: Not malaria season/low malaria prevalence (0.2, 3.0) 0.6 (0.2, 2.4) Other Reason: Chemist/Clinic nearby, people go there 4.2 (2.0, 8.3) 2.9 (1.3, 6.1) 3.2 (1.8, 5.6) Other Reason: Don't have skills to prescribe 1.7 (0.7, 4.1) 1.0 (0.3, 3.5) 1.1 (0.5, 2.8) Other Reason: No capital/no market 3.3 (1.3, 8.0) 2.0 (0.8, 4.9) 2.2 (1.1, 4.5) Other Reason: Don't stock such drugs 0.2 (0.0, 1.6) 0.3 (0.0, 2.1) 0.3 (0.1, 1.5) * This indicator excludes responses from public health facilities and CHW. ** Note that a provider could give more than one response to this question. may add to more than 100 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 67 P age

93 3.6 AMFm logo Table : Provider recognition of AMFm logo, [Kenya], 2011 Providers able to recognize the AMFm logo* (n) as a percentage of the number of outlets with antimalarials in stock at the time of the survey visit** (N), by location and type of outlet Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 84.4 (72.6, 91.7) (68.0, 88.0) (69.4, 87.7) 430 Community health worker Private not for profit health facility 73.0 (36.5, 92.7) (42.5, 76.3) (47.1, 76.6) 46 Private for profit outlet Private for profit health facility 91.0 (86.7, 94.1) (81.1, 94.1) (84.9, 93.4) 264 Pharmacy 98.3 (93.9, 99.6) (82.8, 99.8) (93.2, 99.5) 257 Drug Store 97.9 (95.0, 99.1) (89.8, 98.2) (93.4, 98.2) 474 General retailer 45.3 (36.6, 54.4) (38.2, 61.7) (39.5, 58.8) 377 Itinerant drug vendor Total 84.5 (77.7, 89.5) (62.6, 82.0) (69.4, 83.0) 1372 Total 84.2 (77.7, 89.1) (64.9, 81.1) (70.2, 82.2) 1848 * All respondents were shown a visual aid depicting the AMFm logo. And were asked whether they have seen the symbol before. A provider is able to recognize the AMFm logo if they answer that they have seen the symbol before. ** Flow diagram reference E. Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 68 P age

94 Table : Provider knowledge of the AMFm logo, [Kenya], 2011 Providers stating a specific meaning of the AMFm logo (n) as a percentage of outlets that recognized the AMFm logo (N), by location* Urban Rural Total Meaning of AMFm logo (95% CI) N (95% CI) N (95% CI) N Effective/quality antimalarial 28.2 (22.8, 34.2) (22.5, 35.7) (23.9, 33.6) 1539 Affordable antimalarial 19.1 (13.6, 26.3) 14.0 (7.7, 24.2) 15.6 (10.7, 22.1) An antimalarial in high demand 2.0 (1.1, 3.7) 4.3 (2.4, 7.6) 3.6 (2.2, 5.9) Effective/quality medicine 4.9 (2.8, 8.6) 6.8 (4.6, 9.9) 6.2 (4.5, 8.6) Affordable medicine 6.5 (4.7, 8.8) 3.2 (1.1, 8.4) 4.2 (2.5, 6.9) A medicine in high demand 1.1 (0.5, 2.1) 0.8 (0.3, 2.5) 0.9 (0.4, 1.9) It means nothing 0.3 (0.1, 1.1) 0.6 (0.1, 2.5) 0.5 (0.2, 1.7) Artemisinin Combined Therapy (ACT) 28.3 (24.7, 32.2) 23.1 (19.1, 27.6) 24.7 (21.7, 28.0) Recommended treament 8.8 (6.4, 11.8) 5.0 (3.1, 8.0) 6.2 (4.6, 8.3) Subsidized medicine 7.1 (5.6, 9.1) 3.5 (0.9, 12.8) 4.7 (2.3, 9.0) I don t know what it means 20.9 (16.9, 25.6) 27.0 (22.1, 32.6) 25.1 (21.6, 29.1) Other meaning: Other 2.9 (1.4, 5.7) 3.4 (2.1, 5.6) 3.3 (2.2, 4.8) Other meaning: Combination of two or more drugs 0.1 (0.0, 0.7) 0.1 (0.0, 0.4) 0.1 (0.0, 0.3) Other meaning: Herbal 3.6 (2.1, 6.0) 7.7 (4.8, 12.2) 6.4 (4.3, 9.5) medicine/environment Other meaning: Antimalarial 4.7 (2.9, 7.5) 6.6 (4.1, 10.7) 6.0 (4.1, 8.9) Other meaning: Medicine 0.5 (0.2, 1.3) 1.2 (0.5, 3.0) 1.0 (0.4, 2.2) Other meaning: Government 0.1 (0.0, 0.3) 1.2 (0.3, 4.8) 0.9 (0.2, 3.2) Other meaning: Logos/Symbols/Trade marks 1.9 (1.0, 3.6) 2.0 (0.9, 4.2) 2.0 (1.1, 3.4) * Note that providers could give more than one response to this question. may add to more than 100 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 69 P age

95 Table : Sources from which providers have seen or heard of the AMFm logo, [Kenya], 2011 Providers stating a specific source where they have seen or heard of the AMFm logo (n) as a percentage of providers that recognized the AMFm logo (N), by location* Urban Rural Total Source (95% CI) N (95% CI) N (95% CI) N On malaria medicine packaging 56.7 (49.6, 63.5) (42.3, 59.8) (46.5, 58.9) 1539 On medicine packaging 27.3 (19.5, 36.7) 23.1 (18.7, 28.1) 24.4 (20.6, 28.6) On posters 13.7 (9.9, 18.6) 14.6 (8.4, 24.3) 14.3 (9.6, 20.8) On billboards 2.3 (1.3, 3.8) 1.9 (0.9, 4.1) 2.0 (1.2, 3.4) On TV/radio 70.8 (65.3, 75.8) 65.6 (58.7, 71.9) 67.2 (62.2, 71.8) On a prescription 2.4 (1.4, 3.8) 1.4 (0.6, 3.2) 1.7 (1.0, 2.9) In newspapers/magazines 6.4 (4.7, 8.7) 6.6 (4.2, 10.3) 6.6 (4.8, 9.0) In pharmacies/ drug shops 16.2 (12.7, 20.4) 21.4 (15.2, 29.3) 19.8 (15.2, 25.4) In private clinics 3.7 (2.2, 6.0) 6.9 (3.7, 12.6) 5.9 (3.5, 9.9) In public health facilities 11.2 (7.8, 15.9) 16.7 (11.4, 23.9) 15.0 (10.9, 20.3) In training 6.2 (4.3, 8.9) 5.5 (3.5, 8.4) 5.7 (4.2, 7.7) From a supplier 6.2 (3.8, 9.9) 1.7 (0.9, 3.2) 3.1 (2.0, 4.8) From a public event 1.9 (1.0, 3.7) 3.8 (1.7, 8.2) 3.2 (1.6, 6.2) From a local leader (0.0, 2.0) 0.2 (0.0, 1.3) From a friend/family member 0.4 (0.1, 1.5) 0.6 (0.2, 1.7) 0.5 (0.2, 1.3) On the internet 0.9 (0.3, 2.6) 0.9 (0.2, 4.9) 0.9 (0.3, 3.0) Brochure/pamphlets 1.9 (1.1, 3.3) 1.9 (1.0, 3.4) 1.9 (1.2, 2.9) Medical/pharmacology books 1.5 (0.7, 3.3) 1.1 (0.4, 2.7) 1.2 (0.6, 2.3) Don t Know 0.1 (0.1, 0.3) 1.1 (0.2, 5.0) 0.8 (0.2, 3.3) Other source: Other 1.0 (0.4, 2.2) 0.9 (0.3, 2.5) 0.9 (0.4, 1.9) Other source: Government 0.4 (0.1, 1.8) 0.1 (0.0, 1.0) 0.2 (0.1, 0.7) Other source: Customer 0.4 (0.1, 1.3) 0.7 (0.2, 2.1) 0.6 (0.2, 1.5) Total * Note that providers could give more than one response to this question. may add to more than 100 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 70 P age

96 Table : of antimalarials bearing the AMFm logo, [Kenya], 2011 Antimalarials bearing the AMFm logo (n) as a percentage of all antimalarials audited (N), by location and antimalarial type Urban Rural Total Type of antimalarial (95% CI) N (95% CI) N (95% CI) N Quality assured ACTs Public health facility 47.3 (36.3, 58.6) (32.7, 62.5) (33.7, 61.4) 1497 Community health worker Private not for profit health facility 45.8 (16.1, 78.8) (20.6, 57.0) (23.4, 55.9) 122 Private for profit outlet 94.2 (92.2, 95.7) (91.3, 96.2) (92.5, 95.5) 2048 Total 89.4 (85.8, 92.1) (61.4, 77.0) (69.7, 80.8) 3667 All other antimalarials Public health facility 0.6 (0.2, 2.3) (0.5, 2.7) (0.5, 2.5) 1102 Community health worker Private not for profit health facility Private for profit outlet 2.0 (1.1, 3.7) (0.6, 3.3) (1.0, 2.8) 4618 Total 1.9 (1.0, 3.5) (0.6, 2.8) (0.9, 2.5) 5860 Total 32.4 (29.2, 35.7) (24.6, 32.0) (26.9, 32.3) 9527 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 71 P age

97 Table : Provider knowledge of the AMFm programme, [Kenya], 2011 Providers who have heard of a programme that reduces the prices of antimalarial medicines known as ACTs (n) as a percentage of outlets with antimalarials in stock at the time of the survey visit (N), by location and type of outlet Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 70.3 (60.1, 78.8) (55.6, 77.1) (56.9, 76.6) 431 Community health worker Private not for profit health facility 75.6 (49.9, 90.6) (30.3, 65.3) (37.6, 67.9) 46 Private for profit outlet Private for profit health facility 70.8 (60.0, 79.7) (46.5, 79.3) (55.3, 76.7) 263 Pharmacy 80.0 (71.3, 86.5) (59.6, 90.6) (69.9, 86.7) 255 Drug Store 70.9 (64.9, 76.2) (56.8, 76.3) (62.1, 74.7) 473 General retailer 36.5 (24.3, 50.8) (30.6, 50.9) (31.4, 48.6) 378 Itinerant drug vendor Total 63.7 (56.2, 70.6) (46.4, 62.5) (51.4, 63.2) 1369 Total 64.2 (56.8, 71.0) (49.0, 63.0) (53.0, 63.7) 1846 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 72 P age

98 Table : Sources from which providers have seen or heard of AMFm, [Kenya], 2011 Providers stating a specific source where they have seen or heard of a programme that reduces the prices of antimalarial medicines known as ACTs (n) as a percentage of providers that have heard of a programme that reduces the prices of antimalarial medicines known as ACTs (N), by location* Urban Rural Total Source (95% CI) N (95% CI) N (95% CI) N On malaria medicine packaging 15.3 (10.8, 21.3) (5.6, 21.0) (8.1, 18.6) 1237 On medicine packaging 8.9 (4.9, 15.6) 7.4 (4.2, 13.0) 7.9 (5.2, 11.8) On posters 9.3 (6.6, 12.9) 6.8 (3.6, 12.5) 7.6 (5.1, 11.1) On billboards 1.2 (0.5, 2.7) 1.0 (0.4, 2.6) 1.0 (0.5, 2.1) On TV/radio 92.3 (89.4, 94.4) 91.5 (86.9, 94.6) 91.7 (88.6, 94.0) On a prescription 1.0 (0.4, 2.4) 2.3 (0.9, 5.6) 1.9 (0.8, 4.1) In newspapers/magazines 14.4 (11.3, 18.2) 11.2 (6.8, 18.0) 12.2 (8.9, 16.5) In pharmacies/ drug shops 5.6 (3.2, 9.6) 7.3 (4.7, 11.3) 6.8 (4.7, 9.6) In private clinics 0.9 (0.4, 2.1) 4.7 (1.9, 10.9) 3.5 (1.5, 8.0) In public health facilities 6.3 (3.6, 10.9) 11.4 (6.8, 18.4) 9.8 (6.3, 14.9) In training 8.3 (5.5, 12.4) 9.0 (6.3, 12.6) 8.8 (6.7, 11.4) From a supplier 8.1 (6.0, 10.9) 1.4 (0.6, 3.0) 3.4 (2.3, 5.1) From a public event 2.6 (1.3, 5.1) 3.6 (1.9, 6.9) 3.3 (2.0, 5.6) From a local leader (0.1, 1.9) 0.3 (0.1, 1.3) From a friend/family member 2.2 (1.2, 3.9) 0.2 (0.1, 0.6) 0.8 (0.5, 1.4) SMS messages 0.3 (0.1, 2.1) (0.0, 0.6) On the internet 0.3 (0.1, 0.8) 0.6 (0.2, 2.0) 0.5 (0.2, 1.4) Don t Know (0.1, 1.4) 0.3 (0.1, 0.9) Other source: Other 1.4 (0.6, 3.2) 1.1 (0.3, 4.1) 1.2 (0.5, 2.9) Other source: Government 1.5 (0.6, 3.4) 0.6 (0.1, 2.6) 0.9 (0.4, 2.0) Other source: Customer 0.6 (0.2, 2.1) 0.3 (0.1, 1.3) 0.4 (0.2, 1.0) * Note that providers could give more than one response to this question. may add to more than 100 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 73 P age

99 Table : Provider stating that there is a maximum/recommended retail price (RRP) for antimalarials with the AMFm logo, [Kenya], 2011 Providers stated that there is a RRP for antimalarials with the AMFm logo (n) as a percentage of outlets with antimalarials in stock at the time of the survey visit (N), by location and type of outlet Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 71.3 (59.1, 81.1) (67.8, 87.3) (68.4, 86.1) 431 Community health worker Private not for profit health facility 87.7 (60.3, 97.1) (36.8, 79.3) (45.8, 80.8) 46 Private for profit outlet Private for profit health facility 81.4 (74.4, 86.8) (70.2, 86.2) (74.2, 84.9) 264 Pharmacy 98.1 (95.1, 99.3) (70.2, 98.5) (86.2, 98.6) 255 Drug Store 93.8 (90.9, 95.9) (83.6, 97.0) (88.4, 96.1) 474 General retailer 41.7 (31.2, 53.0) (30.9, 55.9) (32.7, 53.4) 378 Itinerant drug vendor Total 79.3 (73.1, 84.4) (55.8, 76.8) (63.0, 77.8) 1371 Total 79.3 (73.2, 84.3) (59.1, 76.7) (64.7, 77.6) 1848 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 74 P age

100 Table : Provider stating the correct maximum/recommended retail price (RRP) for antimalarials with the AMFm logo, [Kenya], 2011 Providers stated the correct RRP for antimalarials with the AMFm logo (n) as a percentage of providers who responded that there was a RRP for antimalarials with the AMFm logo (N), by location and type of outlet Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 91.8 (83.8, 96.0) (86.7, 97.8) (87.5, 97.5) 343 Community health worker Private not for profit health facility 79.4 (48.6, 94.0) (42.5, 93.4) (51.9, 91.4) 32 Private for profit outlet Private for profit health facility 85.7 (78.2, 90.9) (80.2, 99.0) (86.0, 95.2) 220 Pharmacy 98.7 (97.2, 99.4) (86.4, 97.9) (92.6, 98.6) 248 Drug Store 96.8 (91.0, 98.9) (95.4, 99.6) (95.2, 99.0) 445 General retailer 87.3 (75.3, 93.9) (87.6, 98.1) (87.9, 96.7) 180 Itinerant drug vendor Total 92.9 (90.0, 95.0) (93.6, 98.2) (93.4, 96.7) 1093 Total 92.4 (89.6, 94.6) (93.0, 97.1) (92.7, 95.9) 1468 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 75 P age

101 Table : Providers who have received training on antimalarials with the AMFm logo, [Kenya], 2011 Outlet where at least one staff member has received training on antimalarials with the AMFm logo (n) as a percentage of outlets with antimalarials in stock at the time of the survey visit (N), by location and type of outlet Urban Rural Total Type of outlet (95% CI) N (95% CI) N (95% CI) N Public health facility 20.0 (14.2, 27.5) (7.3, 18.1) (8.1, 18.1) 431 Community health worker Private not for profit health facility 19.0 (6.3, 45.2) (9.1, 37.9) (10.3, 34.3) 46 Private for profit outlet Private for profit health facility 17.9 (11.0, 27.7) (14.4, 39.7) (14.6, 32.6) 264 Pharmacy 21.1 (14.6, 29.5) (17.7, 36.8) (17.9, 30.3) 255 Drug Store 19.9 (15.0, 26.0) (6.8, 19.0) (11.1, 19.9) 473 General retailer (1.0, 9.1) (0.8, 7.3) 378 Itinerant drug vendor Total 15.1 (11.2, 20.0) (7.1, 15.7) (9.2, 15.5) 1370 Total 15.3 (11.6, 20.0) (8.3, 14.8) (10.0, 15.1) 1847 Source: AMFm Phase 1 Independent Evaluation Outlet Survey (Endline) 76 P age

102 4. AMFm implementation: process and key contextual factors 4.1 Introduction In order to document the implementation process of AMFm (supply of copaid ACTs and supporting interventions) and contextual factors that may influence the effectiveness of AMFm, the Independent Evaluation team collected data in each AMFm pilot, using key informant interviews (KII), a structured questionnaire on supporting interventions, and document review. These data were collected to facilitate an assessment of (1) whether any improvement observed in AMFm indicators between baseline and endline is likely to be due to AMFm and (2) whether a lack of improvement in indicators can be reasonably attributed to a failure of AMFm. The methods and findings for this activity in Kenya are summarised below. 4.2 Methods Three types of key informants were included in interviews: those centrally involved in AMFm implementation, antimalarial importers, and other stakeholders who were knowledgeable about the AMFm process or other key contextual factors. A total of 28 interviews were conducted in November and December Oral consent was obtained for all interviews, and participants were assured of confidentiality and given the option of whether they wanted their interviews to be recorded. Notes were taken during all interviews. Interviewers used a semi structured interview guide that covered AMFm governance, registration of first line buyers, ordering and distribution of copaid drugs, supporting interventions (e.g., communications, training, regulation and recommended retail prices), diagnostics, and key contextual events (e.g., weather anomalies, economic and political factors, changes in other malaria control activities and changes in the health system more broadly). Using the agreed template, the information from each interview was then broken into the appropriate reporting categories, and findings across interviews were synthesized. In addition, a form for quantifying supporting interventions was sent to the relevant authorities for completion. Finally, key documents were reviewed such as policy documents, briefing documents and reports prepared by CHAI, Global Fund grant documents, preliminary findings of research studies, data on first line buyer orders, and communication materials. 4.3 Findings AMFm implementation process Governance structure for AMFm There are five main governance structures some old and some new which are pertinent to the governance of AMFm in Kenya. These are: 77 P age

103 1. the Kenya Country Mechanism (KCM) or National Oversight Committee (NOC), previously the Global Fund Country Coordinating Mechanism (CCM) 2. the Malaria Interagency Coordinating Committee (MICC) 3. the Global Fund Technical Working Group (TWG) 4. the AMFm TWG or Task Force 5. Stakeholder Forums The latter two were established as a direct result of the need for planning and coordination around AMFm; the first three have been in existence for far longer. The KCM/NOC focuses on all Global Fund coordination issues, for instance the selection of principal recipients (PRs) and management of the process for the selection of sub recipients (SRs) in coordination with the PRs and the programs. The MICC is the highest decision making body for malaria in Kenya. It meets quarterly and receives updates on all areas of malaria control to facilitate key decisions for which the necessary technical discussions and deliberations have already taken place within the respective TWGs. The Global Fund TWG oversees and provides technical assistance in the implementation of Global Fund malariarelated activities. The AMFm TWG was formed out of the initial group of stakeholders who took part in the AMFm proposal writing for Kenya with ad hoc representation from other stakeholders on an as needed basis. The core members are the Division of Malaria Control (DOMC) Case Management TWG, M&E TWG and Advocacy, Communication and Social Mobilization TWG; Clinton Health Access Initiative (CHAI); Pharmacy & Poisons Board (PPB) and PSI/Kenya. The TWG spearheads the planning, coordination and monitoring of all AMFm related activities and reports back to the Case Management TWG. The meetings are said to have been very useful in tracking progress, identifying constraints and responding flexibly to novel issues. The main challenge of the AMFm TWG has been the slow progress of key AMFm activities, such as private sector training courses and IEC/BCC activities, which led to some fatigue on discussing the same issues of delays over and over, leading to a slowing down of the dialogue. In the lead up to the grant signing, the TWG met weekly, then after grant signature, monthly. The Stakeholder Forums were seen as crucial for advocacy and sensitization for AMFm, especially with the private sector, and were included in the grant proposal as an avenue for engaging with stakeholders in much the same way as Global Fund grants have annual or semi annual stakeholder workshops. There are three key phases to the Stakeholder Forums; the pre grant phase, the early period of grant writing and post grant signature. The pre grant phase involved advocacy for AMFm, both within and outside government. The grant writing phase involved engaging with those who would be directly affected by AMFm professionally or economically, such as the Pharmaceutical Society of Kenya, large pharmaceutical distributors, local manufacturers and other health professionals. At least four stakeholder meetings were held in the run up to AMFm in 2009 and P age

104 The pre grant Stakeholder Forums are viewed as largely successful in fulfilling their stated objectives. By the time the copaid medicines came to the country, AMFm was said to be well understood. Indeed, many expressed the opinion that we would have been lost without the [stakeholder] forums. Some private sector stakeholders were very enthusiastic and even offered to help kick start the IEC/BCC campaigns with their own funds given the delays in procuring IEC/BCC services. A perceived challenge has been poor attendance at the forums, and getting the various players to put all their cards on the table. There were initial misgivings about AMFm because of local manufacturers and importers fearing loss of business, and lack of understanding of the mechanism s objectives and operation. The latter has been largely overcome, although the issue of how local manufacturers take part in such procurement processes still remains salient (ALMA et al. 2011). Public sector AMFm copaid ACT supply mechanism The first line buyer (FLB) for the public sector was the Kenya Medical Supplies Agency (KEMSA). The initial determination of quantities of AL to be ordered was undertaken by the DOMC and partners through a formal quantification workshop, but the quantities determined (31 million doses, including a buffer stock of 13 million) had to be aligned with available resources (approximately USD 14 million), so only 16.6 million doses were included in the tender. The tender process took from October 2010 to May Contracts were signed with Ajanta Limited for the two older age packs in April 2011 and with Novartis Pharma AG for Coartem Dispersible for the two pediatric age packs in May Both orders were split into two equal call downs, with an initial call down of 50% requested immediately, although in practice they were received in more batches in dribs and drabs. The first consignment arrived on 27 June , and by December 2011, 8.5 million doses for the public sector had been delivered in Kenya, but Ajanta Limited had not delivered their full order by December These delays were perceived to be partly due to a global shortage of the artemesia active pharmaceutical ingredient (API). An additional emergency procurement was made through the President s Malaria Initiative (PMI) for 4 million doses, of which 3 million arrived between July and December Public sector facility stockouts were not reported in 2011, except for a four week period in July and August (MOPHS/DOMC et al. 2010, 2011a, 2011b). It is acknowledged that despite the strides made by KEMSA in building up efficient procurement and supply management systems, government procurement processes for ACTs were still very lengthy. It took nine months between the tender closure and the first consignment of the copaid ACTs being delivered in Kenya, which is fairly typical of medicine procurement processes. What was unusual was that six months after the first consignment of copaid ACTs was delivered, Ajanta Limited had not been paid for their consignment. Indeed, by December 2011 no manufacturer had received the government share of the price of copaid ACTs delivered to the public sector. This appeared to reflect administrative bottlenecks in terms of invoicing and invoice processing, involving the DOMC, KEMSA and the PR (Ministry of Finance). Novartis Pharma AG had to make special representations to their risk management section in order to avoid further orders to KEMSA being barred, and KEMSA had to 79 P age

105 make representations to Ajanta not to stop further deliveries. The delay in payment may have further compounded public sector delivery delays due to a shortage of API. Private sector Seven private sector FLBs were registered and established relationships with manufacturers, three with Novartis Pharma AG, and the others with four other manufacturers. Six were reported to be active in ordering copaid ACTs with one having dropped out much earlier in the process. There were no major issues reported in the registration of FLBs, the determination of order quantities or customs clearance, although at first there was some confusion as to whether customs levies should be charged on the full or subsidized price of copaid ACTs. The DOMC and CHAI, through the relevant Stakeholder Forums and the AMFm TWG, were instrumental in clarifying the process and providing assistance. The first private sector copaid ACTs were received in Kenya in August For the 18 month period from July 2010 (soon after grant signing) to December 2011, a total of 12.8 million treatment doses were received in Kenya by private sector FLBs. There was no discernible pattern in the total monthly receipts, except from April August 2011 when the drug supply was on an upward trend, peaking in August 2011 and declining thereafter. In sharp contrast to the issues identified above for the public sector, the speed and efficiency of the private sector surprised everybody. The first consignment of copaid ACTs was in Kenya before the ink dried on the AMFm grant signature. In fact, the speed of the private sector practically dictated the pace of some activities which were scheduled for much later, such as the national launch in August 2010, so that members of the public could tie the new copaid ACTs with the government s efforts at increasing access to quality assured ACTs and to ensure adherence to the recommended retail price (RRP). Some FLBs that were new to the ACT market were very innovative and daring in their ordering and marketing, and quickly made substantial orders. However, orders from Novartis were slower to arrive, with the first consignment not being delivered until January 2011, for a number of reasons: Novartis had to register a new packaging for the private sector Coartem Novartis made a strategic decision to work with very enthusiastic FLBs rather than their traditional agents for premium Coartem, and the process of identifying and establishing these new partners took time Some traditional FLBs for Coartem were initially skeptical about AMFm and therefore very conservative in their ordering Novartis faced supply constraints due to the need to meet orders in other countries, meaning that delivery times were longer than expected Several FLBs reported that from August 2011, the demand shaping levers applied by the Global Fund to align orders more with the burden of disease led to a restriction in the number of packs for the two older age groups which could be ordered, and a slowing down in the order process. 80 P age

106 Private sector FLBs reported that quantities imported soon disappeared off the shelves as there was a huge, unanticipated demand. Many ramped up their sales teams to increase sales in rural areas by hiring additional telemarketers and travelling regional representatives Implementation of AMFm supporting interventions The first disbursement from the Global Fund to cover supporting interventions (SI) was for USD 7.4 million, which was received in August Requests to the Global Fund for further disbursements were made in December 2010 and July 2011, but these were both declined. The December 2010 request was declined primarily because there were still substantial funds remaining from the first disbursement, and the July 2011 request was declined primarily due to accounting issues. It is possible that the decline of further disbursement requests led to some slowing down of SI implementation in the latter half of 2011, although there were also other causes of SI delays, as described below. Communication The most important supporting intervention for AMFm in scale, scope and financial expenditure was the IEC/BCC campaign, which was allocated a total of USD 5,681,487 up to December The campaign was tendered in stages by KEMSA, with various components won by Access Leo Burnett, ReelForge, 29 radio stations, and 4 TV stations. The main IEC/BCC activities that were planned were the following: 1. National launch 2. Developing and airing five radio messages 3. Developing and airing four TV messages 4. Developing, printing and disseminating 400,000 posters 5. Holding community meetings in 558 locations 6. Facilitating 186 road shows. The messaging in the print, electronic and broadcast media was graduated, with earlier messages focusing on what AMFm is, including the price of the ACTm medicines and their superiority to other non recommended medicines, and recognition of outlets where ACTm medicines were sold. Later, messages focused on the need to get tested before treatment, the need to use ACTm when positive for malaria and the need to adhere to the full treatment dose even when symptoms of the illness subside. With time, prevention messages such as the use of long lasting insecticidal nets (LLINs) were also incorporated. The main target areas were Nyanza, Western and Coast provinces, although radio and TV spots had a national reach. Stakeholders agreed that this was the biggest malaria IEC/BCC campaign they had ever seen, and that it was very successful in its stated objectives of: o demand creation for AMFm copaid ACTs o creating awareness about the RRP of Ksh 40 (USD 0.46) o the need for diagnostics before dispensing ACTs 81 P age

107 On the other hand, the sheer scale of the IEC/BCC campaign it accounted for more than 80% of the entire SI budget led to risk aversion in procuring the services, leading to further delays in a system which was already fraught with procurement challenges. This was compounded by the fact that it was the first time KEMSA had procured IEC/BCC services, as opposed to commodities, and as a result there was a lot of back and forth; a lot of learning for all concerned. The national launch was held on 26 August in Nairobi, officiated by the Minister of Public Health and Sanitation, and was widely reported in the media. A newspaper supplement was distributed through the local daily papers in September Other activities were substantially delayed by the procurement challenges. Due to the delays in IEC/BCC procurement, PSI/Kenya and CHAI co funded a stop gap measure between December 2010 and January 2011, involving sold here posters for retail outlets, and radio and TV spots. Over 70% of the IEC/BCC budget was for radio messaging, which began in February 2011, with a total of 17,560 radio spots estimated to have been aired between March and December Posters were also printed, although there were some problems with their distribution. The road shows took place by December 2011, albeit with some logistic difficulties, but the community meetings had not yet begun. Recommended retail price The Recommended Retail Price (RRP) of copaid ACTs as communicated to members of the public through the IEC/BCC messages was Ksh 40 (USD 0.46) for all pack sizes. The price was developed and set through the Stakeholder Forums with the private sector. FLBs were of the opinion that the RRP should not be displayed on the packaging to allow for free competition and hopefully further price reductions. The Pharmacy and Poisons Board (PPB) has no legal mandate to enforce prices for pharmaceuticals. The only enforcement strategy, therefore, is customer awareness and empowerment through public messages. Between August and December 2010, initial media reports showed high prices of the copaid ACTs circulating in the market, reflecting limited supply at that time. However, once the full (Global Fundfunded) media campaign had kicked in, prices stabilized at or close to the RRP. Health Action International (HAI) price monitoring surveys (HAI/Africa 2011a, 2011b) indicated a median price of approximately Ksh 40 in private sector outlets. The DOMC s own inspection visit in Nyanza province involving 240 retail pharmacies in April 2011 showed a similar outcome. The issue of price stigma that the copaid medicines are viewed as poor quality or suspect because they are very inexpensive has come up in previous assessments (Appleford 2011) and is one that is still of concern to the DOMC and stakeholders. Training Training, as with other SIs, was outsourced via an open national tender, won by four agencies: Sema, Lisa, MEDS Consultancy and Maseno University. The target was to train 5,890 private sector health workers, with 4,520 to be trained by mid grant (July 31, 2011). However, training did not begin until October 2011, reflecting delays due to procurement challenges and the need to supply a revised 82 P age

108 training plan to the Global Fund. By December 2011, only 733 private sector health workers had been trained in Western and Nyanza provinces. The DOMC judged that doctors and pharmacists had been underrepresented because these were three day residential training courses and perhaps the strategy of using their professional associations to reach out to them was not effective. The private sector FLBs have also been instrumental in sensitizations/trainings. They have trained their own distributors and reached out to healthcare professionals through their regular meetings by sponsoring continuous education meetings or sponsoring the DOMC case management officers to give talks at annual professional gatherings, such as those for clinical officers and pharmacists. It is estimated that approximately 1,000 healthcare professionals have been reached through such strategies between January and December Pharmacovigilance Pharmacovigilance (PV) and post market surveillance activities under AMFm have benefited from ongoing work under other funding streams and technical assistance, notably from USAID through United States Pharmacopeia (USP), Management Sciences for Health and WHO. For antimalarials and ACTs, quality has been the main focus of these PV and regulatory activities. Five sentinel sites (Nairobi, Mombasa, Kakamega, Eldoret and Kisumu) have been established and supplied with a minilab, which can perform qualitative and semi quantitative tests (USP et al. 2007). To date, the DOMC has conducted two rounds of quality testing the first in 2009 under USP funding and the second between January and February 2011 using AMFm SI funds. An inspection visit was also conducted in Nyanza province in April 2011 involving 240 chemists. A routine inspection by the PPB in mid November 2011 also resulted in a crackdown on unlicensed outlets in that province while AMFm IE endline outlet survey data collection was ongoing. Only one PV supervisory visit has been conducted using AMFm resources, reflecting the delayed disbursement of funds. ACTs are still prescription only medicines (POM) in Kenya. However, in practical terms, the POM status of ACTs has not been an impediment to access to copaid ACTs because of the disconnect between de jure regulation and real life medicine regulation in Kenya (Amin et al. 2007). Many POM medicines are available over the counter (OTC) and regulatory infringement such as the presence of unregistered pharmacies and even unregistered products has been documented. Other AMFm supporting interventions Two other supporting interventions planned under AMFm were improving ACT access through community health workers (CHWs) to help reach poor and vulnerable populations, and operational research. The DOMC proposed to piggyback on Kenya s overall Community Strategy for health by improving access to ACTs through existing community health units in Western and Nyanza provinces, where the burden of malaria is highest. The strategy was to procure and distribute ACTs to CHWs, train 80 community health extension workers and 2,000 CHWs on malaria case management in the same provinces and strengthen supervision. Some 1.18 million copaid AL treatment doses were procured and distributed in Western and Nyanza between June and December The training and supervision activities had not taken place by December 2011, due to delays in procurement for training, the design and production of training materials. 83 P age

109 No operational research planned under AMFm and funded by the Global Fund had taken place by December Implementation of non AMFm interventions In addition to AMFm, two other key malaria interventions were implemented during 2011: ACT and RDT availability increased in parts of North Eastern, Coast and Rift Valley Provinces following increased commodity supply to avert a predicted epidemic in the last quarter of In August 2011, the Kenya meteorological department issued a red alert of possible torrential rains in parts of Coast Province, North Eastern and Rift Valley for the period October December In response, the DOMC anticipated a sharp increase in malaria cases and developed a response plan in collaboration with the UK Department for International Development, the US President s Malaria Initiative, Mentor Initiative and UNICEF. Activities took place between August and December Approximately 400,000 RDTs were sent to public facilities in the epidemic prone areas identified. Also, the usual distribution cycle of antimalarials from KEMSA was hastened so that the areas did not run out of stock during the anticipated epidemic and surveillance and monitoring activities were enhanced. In addition, some 20,000 ampoules of artesunate injections were sent for severe cases. An IEC/BCC strategy was also drawn up for the emergency campaign involving demand creation for proper diagnosis and effective treatment of malaria. Advertisements were placed on radio, especially on vernacular stations in the affected areas. A number of radio talk shows and call in sessions were also done. Affected districts were assisted in overall planning and coordination and given money in case the epidemic did occur. As it turned out, the anticipated very heavy rains did not come to pass. However, the response is likely to have increased QAACT and RDT availability in the public sector, as well as QAACT market share. Approximately 6 million LLINs were distributed. From March 2011, the DOMC undertook a rolling campaign aimed at distributing approximately 11 million LLINs in the whole of Western, Nyanza, and Coast Provinces and selected districts in Rift Valley and Central Provinces in line with the epidemiology of malaria in Kenya. The LLINs were procured by a combination of Global Fund, PMI, World Bank and World Vision funds, with the bulk of funding coming from the Global Fund Round 4 Phase 2 malaria grant. The objective of the campaign was to enable Kenya to attain universal coverage of LLINs, i.e., one LLIN for every two persons at risk of malaria. By December 2011, approximately 6 million LLINs had been distributed to all targeted provinces except Coast. The impact on AMFm indicators is not clear, although it is likely to have reduced malaria incidence and therefore demand for ACTs. Although Kenya s official policy is that all suspected cases of malaria should be subject to a blood test for confirmation, in practice, availability of diagnostic tests in the public and private sectors has remained very limited. In addition to the RDTs distributed to the epidemic prone areas (see above), RDTs have been rolled out to six districts participating in a pilot of the SMS for Life project. 84 P age

110 Finally, the international ban on artemisinin monotherapies in 2006 and subsequent reissuing of government circulars in 2006 and 2008 in Kenya, which ban the sale, manufacturer and importation of artemisinin monotherapies have reduced the supply of monotherapies. The ban was reinforced by an agreement between AMFm FLB and the Global Fund to not to sell these medicines, which is likely to have provided a supportive environment for AMFm. Key events and context Two key contextual factors are domestic inflation and the high levels of political support for AMFm. The AMFm price indicator is likely to have been affected by the loss of value of the Kenya shilling against the main international currencies, high inflation and fuel shortages, which all led to price increases for essential commodities. The Kenya shilling had been stable against the United States dollar until February 2011, at approximately Ksh to the USD, when it started fluctuating, hitting rock bottom in October 2011, at to the dollar. It has subsequently regained its value from mid October 2011 onwards after the Central Bank of Kenya intervened. Understandably, the cost of fuel and transportation went up during this time. Monthly inflation was estimated at 3.21% in October 2010; by September 2011, it was at 17.32%. Similarly, the cost of a liter of premium petrol went from Ksh 95.0 in October 2010 to in August 2011 (Parliamentary Service Commission 2011). One might expect these price shocks to have a knock on effect on ACT prices because of increased import costs and increased cost of inland transportation. However, this was believed to be unlikely to have been significant for copaid ACTs because of the low cost paid by FLBs and the margin available with the Ksh 40 RRP being sufficient to absorb some increases in distribution costs, especially for pediatric doses. The fact that AMFm enjoyed high level political support from the Minister of Public Health and Sanitation has been instrumental in advocacy and pushing the process forward. The emphasis on the importance of reducing childhood mortality was therefore seen to override domestic concerns such as loss of market share by local manufacturers. Conclusion Table summarizes key factors likely to have supported or hindered achievement of AMFm goals in Kenya and Figure presents a timeline of all key events related to AMFm implementation and context. The key findings of the case study can be summarized as follows: There was enthusiastic uptake of the opportunity to order copaid ACTs by private sector FLBs, and no major issues in supply to the private sector were reported until the last quarter of 2011 when the application of demand shaping levers by the Global Fund began to slow down and restrict some orders. However, given that downstream supply does not quickly dry up even with stockouts at the central level, this is unlikely to have substantially affected QAACT availability at the time of the endline outlet survey, though it may have affected availability during collection of the remote areas study data. By contrast, the public sector drug supply has faced significant delays due to procurement 85 P age

111 and delivery challenges. A large scale IEC/BCC campaign has been implemented that is perceived to have been successful in creating demand and raising awareness about the RRP. The campaign was delayed by several months, but stop gap activities were put in place by other stakeholders in the mean time. Training activities for private sector health workers were also heavily delayed, meaning that only 733 had been trained by December A response to a predicted malaria epidemic in late 2011 may have served to increase QAACT availability and market share in some provinces. AMFm has received high level political support, and private sector stakeholders have also been very supportive. Table : Summary of key factors likely to have supported or hindered achievement of AMFm goals in Kenya Factors which are likely to have supported achievement of AMFm goals Effective sensitization and mobilization of private sector FLBs Fast pace of the private sector in terms of ordering, processing and distribution systems Increased distribution of ACTs to epidemic areas as part of epidemic preparedness and response RRP which was well publicized. Large scale IEC/BCC campaign Ban on monotherapies, and FLBs undertaking with the Global Fund not to sell artemisinin monotherapies Lack of enforcement of POM status of ACTs Factors which are likely to have hindered achievement of AMFm goals Delays in public sector procurement process for ACTs Delays in delivery of some public and private orders Inadequate supplies in the private sector, initially, which may have pushed up prices Delays in procurement of IEC/BCC and training interventions Rationing of orders through demand levers may have affected QAACT availability in P age

112 Figure : Timeline of key events related to AMFm implementation process and context in Kenya Activity AMFm grants and orders Application to the Global Fund for AMFm AMFm Grant signature First disbursement request First disbursement under AMFm grant Arrival of first consignment of copaid ACTs Harleys Limited AMFm supporting interventions AMFm National Launch IEC/BCC activities Public sector tender award to Ajanta and first Ajanta lld Inspection visit to 240 chemists in Nyanza Public sector tender award to Novartis and first Novartis First lld consignment of public sector copaid ACTs delivered by Aj End Global t Fund Round 4 Phase II (host grant for AMFm) New Global Fund order management system begins to affect K Forecast d of torrential rains in parts of Kenya Second calldown of public sector ACTs Lowest recorded value of the Kenya shilling to the dollar Signing of Round 10 malaria grant Training of 732 private health workers in Western and N P i Crackdown on unregistered pharmacies in Nyanza province Non AMFm interventions LLIN campaign Receipt of PMI emergency procurement of ACTs for public t Research activities IE baseline outlet survey data collection IE endline outlet survey data collection IE country case study Jul Jun Jul Aug Sep Oct NovDec Jan FebMarAprMayJun Jul Aug Sep Oct NovDec 87 P age

113 5. Summary of findings 5.1 Quality of data collected During data collection quality control (QC) persons travelled with the field teams and visited a sample of the outlets visited by the data collectors. The QCs went to 5%of outlets which met the screening criteria and 5% which did not. They asked basic background questions to the outlet attendants, and audited a selection of drugs present. Data were reviewed with supervisors on a daily basis. Completed questionnaires were also reviewed by the supervisors. Any issues were fed back to field teams, which minimised problems that arose in the field. In addition, regular communication with ACTWatch Central ensured that problems could be resolved quickly when issues arose in the field. No major issues appeared during data cleaning or analysis. 5.2 Availability of quality assured ACTs Among facilities that stocked antimalarials at any time in the three months preceding the survey, overall QAACT availability in 2011 was 66%. This was lower in the private for profit sector (60%) than in the public not for profit sector (97%). However, there was some variation within the private forprofit sector. QAACT availability was higher in pharmacies (92%), drug stores (87%), and private forprofit facilities (74%), than general retailers (24%). For private for profit sector, QAACT availability was higher in urban areas than in rural areas (71% versus 55%). In the public and private not for profit outlets, availability of QAACTs with and without the AMFm logo was very similar, but in the private for profit sector, availability of QAACTs with the logo was 59% compared with only 5% for QAACTs without the logo. Availability of nats was 81% overall and 93% in the public not for profit sector and 77% in private for profit sector. 5.3 Pricing/affordability of quality assured ACTs In the public and private not for profit outlets, the median QAACT price was USD0.00, reflecting the policy of free ACT provision. Pooling all sectors, the median price was USD The median price of QAACTs in the private for profit sector was USD 0.58, although still somewhat higher than the RRP of USD Prices remained slightly higher in urban areas (USD 0.61 versus USD 0.46 in rural areas). The median price for a QAACT with the AMFm logo was USD 0.46 per AETD overall (exactly equal to the RRP). In the private for profit sector the price of a QAACT with the AMFm logo was USD The median price among private for profit outlets is exactly equal to the median price of the most 88 P age

114 popular antimalarial which is not a QAACT (SP) in private for profit outlets, whether this is measured in tablet form or among all dosage types. The price of non quality assured ACTs in the private forprofit sector was USD The gross percentage markup on QAACTs, measured among private for profit outlets, was 48%. The gross percentage markup in private for profit outlets was higher on QAACTs with the logo than on QAACTs without the logo (50% vs. 40%), but because of the large differences in the price of these products, the absolute markup on copaid QAACTs was much lower. As a comparator, the gross percentage markup on non artemisinin therapy (nat) in the private for profit sector was 50%, and very similar to that of QAACTS. The median total markup from first line buyer price to retail price in private for profit outlets was low, at only USD Market share of quality assured ACTs Market share of QAACTs was 57% of all antimalarials sold/distributed in the week preceding the survey, with nats accounting for 38% of the overall market share. Overall market share of oral AMTs was negligible (0.05%). Market share of QAACTs was highest in the private for profit sector (60%), as compared with the private not for profit outlets (44.4%) and public health facility outlets (35%). In the private not for profit health facilities, 37% of the markets share were non QAACTs. QAACTs with the AMFm logo accounted for 88% of all QAACTS dispensed overall and 97% of all QAACTs dispensed in the private for profit sector. The private for profit sector was responsible for 62% of all antimalarials sold or distributed at baseline, accounting for 89% and 52% in urban and rural areas, respectively. 89 P age

115 6. References Abuya T, Mutemi W, Karisa B, et al., (2007). Use of over the counter malaria medicines in children and adults in three districts in Kenya: implications for private medicine retailer interventions, Malaria Journal 6:57. Adeyi, O. and Atun, R. (2010). "Universal access to malaria medicines: innovation in financing and delivery," Lancet 376 (9755): Amin AA, Walley T,Kokwaro GO,Winstanley, PA and Snow RW(2007). Reconciling national treatment policies and drug regulation in Kenya. Health Policy and Planning 22: Appleford G. (2011). Affordable Medicines Facility malaria. Kenya Stakeholder consultation. Nairobi, September. Arrow, K., Panosian, C.B. and Gellband H. (2004). Saving Lives, Buying Time: Economics of malaria drugs in an age of resistance, National Academies Press, Washington, DC. Chuya A, Abuya T, Memusi D, et al., (2009). Reviewing the literature on access to prompt and effective malaria treatment in Kenya: implications for meeting the Abuja targets, Malaria Journal 8:243. Craig, P., Dieppe P. et al. (2008). Developing and evaluating complex interventions: new guidance, Medical Research Council, < Central Intelligence Agency, (2011) World fact Book, < world factbook/geos/bn.html> de Savigny, D. and Adam T., eds. (2009). Systems thinking for health systems strengthening, Alliance for Health Policy and Systems Research and WHO, Geneva. DOMC et al. (2011a). Monitoring out patient malaria case management under the 2010 diagnostic and treatment policy in Kenya progress during Division of Malaria Control (DOMC), Ministry of Public Health and Sanitation, KEMRI/Wellcome Trust Research Programme and Management Sciences for Health. DOMC et al. (2011b) Monitoring out patient malaria case management under the 2010 diagnostic and treatment policy in Kenya progress during the first 18 months. Division of Malaria Control (DOMC), Ministry of Public Health and Sanitation, KEMRI/Wellcome Trust Research Programme and Management Sciences for Health. DOMC et al. (2010). Monitoring out patient malaria case management under the 2010 diagnostic and treatment policy in Kenya baseline results. Division of Malaria Control (DOMC), Ministry of Public Health and Sanitation, KEMRI/Wellcome Trust Research Programme and Management Sciences for Health. 90 P age

116 HAI/Africa (2011a). Retail prices of ACTs copaid by the AMFm and other antimalarial medicines in Ghana, Kenya, Nigeria and Tanzania. A report of price tracking surveys. June 2011 HAI/Africa. (2011b) Retail prices of ACTs copaid by the AMFm and other antimalarial medicines in Ghana, Kenya, Madagascar, Nigeria, Tanzania and Uganda. A report of price tracking surveys. September/October Littrell M, Gatakaa H, Phok S, Allen H, Yeung S, Chuor CM, Dysoley L, Socheat D, Spiers A, White C, Shewchuk T, Chavasse D, O'Connell KA: Case management of malaria fever in Cambodia: results from national anti malarial outlet and household surveys. Malaria Journal 2011, 10:328. Ministry of Medical Services, (2008). Ministry of Medical Services Strategic Plan, , Nairobi. O'Connell KA, Gatakaa H, Poyer S, Njogu J, Evance I, Munroe E, Solomon T, Goodman C, Hanson K, Zinsou C, Akulayi L, Raharinjatovo J, Arogundade E, Buyungo P, Mpasela F, Adjibabi CB, Agbango JA, Ramarosandratana BF, Coker B, Rubahika D, Hamainza B, Chapman S, Shewchuk T, and Chavasse D (2011). Got ACTs? Availability, price, market share and provider knowledge of anti malarial medicines in public and private sector outlets in six malaria endemic countries. Malaria Journal 10:326. Parliamentary Service Commission (2011). Parliamentary Budget Office: Inflation and exchange rates in Kenya, why we must act decisively. Nairobi, October. United States Pharmacopeia Drug Quality and Information Program and collaborators(2007). Ensuring the quality of medicines in resource limited countries: An operational guide. Rockville, Md.: The United States Pharmacopeial Convention. Sabot, O.J., et al., (2009). "Piloting the global subsidy: The impact of subsidized artemisinin based combination therapies distributed through private drug shops in rural Tanzania," PLoS ONE 4(9): e6857.sabot O.J., Yeung, S., Pagnoni, F., Gordon, M., Petty, N., Schmits, K. and Talisuna A., (2008). Distribution of Artemisinin Based Combination Therapies through Private Sector Channels Lessons from Four Country Case Studies. Prepared for the Consultative Forum on AMFm the Affordable Medicines Facility malaria September 27 28, 2008 Washington, DC. Shewchuk T, O'Connell KA, Goodman C, Hanson, K, Chapman S, and Chavasse D(2011). The ACTwatch project: methods to describe anti malarial markets in seven countries. Malaria Journal 10:325. Shretta, R., Amin, A., Tetteh, G., Wahito, J., Nyandigisi, A., Juma, E., (2009). Increasing Community Access of ACTs in Kenya by Changing the Regulatory Status from POM to OTC: A Guidance document, Submitted to the U.S. Agency for International Development by the Strengthening Pharmaceutical Systems (SPS) Program, Management Sciences for Health, Arlington, VA.. 91 P age

117 Snow R.W., Okiro, E.A., Noor, A.M., Munguti, K., Tetteh, G., Juma, E., (2009). The coverage and impact of malaria intervention in Kenya , Division of Malaria Control, Ministry of Public Health and Sanitation. United Nations, (2010) United Nations Demographic Year Book, Department of Economic and Social Affairs, New York. United Nations, (2011). World Statistics Pocket Book 2010.Department of Economic and Social Affairs, Series V, No. 35, New York. United Nations Population Division, (2010). World Population Prospects: The 2010 Revision, accessed on 11 March 2011, World Bank, (2010). World databank, < World Health Organization, (2010). World Health Statistics 2010, France. World Health Organization (2009). Systems thinking for health systems strengthening, Geneva. World Health Organization, (2010a). World Malaria Report 2010, WHO Press: Geneva. 92 P age

118 7. Acknowledgements This report presents the results of the Kenya Endline Outlet Survey. It is a comprehensive, nationally representative outlet survey designed to evaluate Phase 1 of the Affordable Medicines Facility for malaria. This report would not have been possible without the efforts of a large number of people who assisted with the data analysis, data processing, and preparation of the report, as well as those who worked tirelessly to collect the endline survey data and analyze the results. We particularly appreciate the efforts of the Division of Malaria Control and the Pharmacy and Poisons Board for providing overall support for the survey. We would also like to thank APHRC for conducting the survey, to PSI/Kenya for providing project management support. We express our appreciation to the Independent Evaluator, ICF International and LSHTM for their assistance during the research process, and to the Global Fund for their support. We would also like to express our thanks to the field teams and individuals involved in the survey. Their names are presented in the Appendix. Finally, we would like to thank the thousands of providers who took time to complete the interview. Without them, it would not be possible to provide these results. 93 P age

119 8. Appendices 8.1 Questionnaire 94 P age

120 95 P age

121 96 P age

122 97 P age

123 98 P age

124 99 P age

125 100 P age

126 101 P age

127 102 P age

128 103 P age

129 104 P age

130 105 P age

131 106 P age

132 107 P age