ABSTRACT Background: MEASURE Evaluation is conducting an impact evaluation of the Organized Network of Services for Everyone’s (ONSE) Health project, in Malawi. ONSE aims to reduce maternal, newborn, and child morbidity and mortality. Aims: The primary goal of the impact evaluation is to estimate the extent to which the ONSE project has impacted health outcomes. Methods: The impact evaluation uses a quasi-experimental approach in three ONSE and three non-ONSE districts. The end line survey will use a difference-in-differences (DID) approach to estimate the causal impact of the ONSE project on changes in health and facility outcomes. Baseline data were collected from April to July 2017 from 7,929 households and 139 health facilities. Results and Conclusions: Skilled antenatal care (ANC) attendance and skilled birth attendance were almost universal. One-half of pregnant women received the recommended four or more ANC visits during their pregnancy. Knowledge of key maternal and newborn danger signs was very low. The availability of services for family planning, ANC, and basic obstetric care was very high. Readiness to provide services was more varied. Obstetrics was the area with the lowest general readiness of all service types. Assisted vaginal delivery and removal of retained products were the two signal functions of basic emergency obstetric and newborn care (BEmONC) provided by the lowest percentage of health facilities and hospitals. The end line survey will provide follow-up data on these indicators and will measure change over the project period. Malawi ONSE Impact Evaluation: Baseline Report v Malawi ONSE Impact Evaluation: Baseline Report 7 ACKNOWLEDGMENTS MEASURE Evaluation would like to thank Veronica Chirwa, Collins Kwizombe, and Lilly Banda of the United States Agency for International Development (USAID)/Malawi for their support of the impact evaluation of the Organized Network of Services for Everyone’s (ONSE) Health project. We are especially grateful to the staff of our local research partner, the Centre for Social Research (CSR) in Zomba: Blessings Chinsinga, Peter Mvula, Jacob Mazalale, and Massy Chiocha. We are indebted to our data collection team, especially the 18 supervisors from CSR: Fiskani Msutu, Diana Namalueso, Elina Mkandawire, Leah Mziya, Prisca Ngoma, Tobias Maunde, Madalitso Banda, Tendai Lipipa, Butao Masautso, Linda Kalulu, Sandra Kuntembwe, Juliet Magombo, Getson Uladi, Alfred Grey, Baina Kassim, Mphatso Mwandali, Mphatso Chiphwanya, and Madalitso Mwasima. Last, our sincere thanks to the gracious people of Malawi who participated in the interviews. We thank MEASURE Evaluation’s knowledge management team for editorial and production services. Cover: Young mother in Malawi. Photo: © 2008 Lisa Basalla, courtesy of Photoshare 8 Malawi ONSE Impact Evaluation: Baseline Report CONTENTS Figures ......................................................................................................................................................................... 10 Tables........................................................................................................................................................................... 11 Abbreviations.............................................................................................................................................................. 13 Executive Summary................................................................................................................................................... 15 Introduction........................................................................................................................................................... 15 Background............................................................................................................................................................ 15 Impact Evaluation Research Questions and Methods.................................................................................... 15 Results..................................................................................................................................................................... 17 Conclusions............................................................................................................................................................ 19 Next Steps.............................................................................................................................................................. 19 Introduction................................................................................................................................................................ 21 Background............................................................................................................................................................ 21 The Malawi ONSE Project ................................................................................................................................. 22 The ONSE Impact Evaluation........................................................................................................................... 25 Methods....................................................................................................................................................................... 26 Selection of Study Districts................................................................................................................................. 26 Quantitative Household Survey.......................................................................................................................... 26 Health Facility Assessment.................................................................................................................................. 28 Strengths and Limitations.................................................................................................................................... 29 Results.......................................................................................................................................................................... 30 Household Population and Housing Characteristics........................................................................................... 30 Household Population......................................................................................................................................... 30 Housing Characteristics....................................................................................................................................... 38 Water and Sanitation Characteristics................................................................................................................. 40 Assistance Provided to Households.................................................................................................................. 42 Characteristics of Women of Reproductive Age .................................................................................................. 44 Characteristics of WRA ....................................................................................................................................... 44 Exposure to Mass Media ..................................................................................................................................... 49 Exposure to Moyo Ndi Mpamba, Usamalireni!..................................................................................................... 52 Satisfaction with Health Services ....................................................................................................................... 52 Family Planning.......................................................................................................................................................... 56 Current Use of Family Planning......................................................................................................................... 56 Beliefs about Family Planning ............................................................................................................................ 61 Maternal Health.......................................................................................................................................................... 65 Antenatal Care....................................................................................................................................................... 65 Assistance During Delivery................................................................................................................................. 68 Postnatal Care........................................................................................................................................................ 69 Women’s Knowledge about Pregnancy and Childbirth ................................................................................. 73 Malawi ONSE Impact Evaluation: Baseline Report 9 Child Health................................................................................................................................................................ 78 Background Characteristics of Children under Three .................................................................................... 78 Breastfeeding ......................................................................................................................................................... 80 Prevalence and Treatment of Fever, Acute Respiratory Infection, and Diarrhea...................................... 84 Women’s Knowledge about Common Infectious Diseases in Children..................................................... 89 Service Availability and Readiness Assessment..................................................................................................... 94 Characteristics of Health Facilities..................................................................................................................... 94 Basic Client Services and Amenities.................................................................................................................. 95 Availability of Priority Medicines for Mothers and Children ........................................................................ 97 FP Services........................................................................................................................................................... 100 Maternal Health Services................................................................................................................................... 102 Child Health......................................................................................................................................................... 112 Malaria .................................................................................................................................................................. 117 Conclusion ................................................................................................................................................................121 Primary Outcomes.............................................................................................................................................. 121 Secondary Outcomes ......................................................................................................................................... 123 Comparability of Project and Comparison Domains................................................................................... 123 Exposure to Other Interventions..................................................................................................................... 124 Implications for the Impact Evaluation.......................................................................................................... 124 Next Steps............................................................................................................................................................ 125 References.................................................................................................................................................................126 Appendix A. Additional Tables.............................................................................................................................128 Appendix B. Key Indicator Data Synthesis.........................................................................................................173 Appendix C. Methods Supplement.......................................................................................................................185 Appendix D. Data Collection Tools.....................................................................................................................274 Appendix E. Survey Implementation ...................................................................................................................293 Appendix F. Evaluation Scope of Work..............................................................................................................295 Appendix G. Team Members and Disclosures of Conflicts of Interest.........................................................308 10 Malawi ONSE Impact Evaluation: Baseline Report FIGURES Figure 1. Malawi ONSE theory of change....................................................................................................................22 Figure 2. Percentage of WRA who knew the four recommended components of birth preparedness planning......................................................................................................................................................74 Figure 3. Percentage of WRA who identified serious warning/danger signs during pregnancy .........................75 Figure 4. Percentage of WRA who identified serious warning/danger signs during childbirth ..........................76 Figure 5. Percentage of WRA who identified serious warning/danger signs of newborn complications..........77 Figure 6. Percentage of WRA who knew the signs and symptoms of malaria .......................................................89 Figure 7. Percentage of WRA who knew the signs and symptoms of pneumonia................................................91 Figure 8. Percentage of WRA who knew the causes of pneumonia.........................................................................92 Figure 9. Percentage of WRA who knew the signs and symptoms of diarrhea......................................................92 Figure 10. Percentage of WRA who knew the causes of diarrhea ............................................................................93 Malawi ONSE Impact Evaluation: Baseline Report 11 TABLES Table 1. ONSE interventions, by district............................................................................................................... 24 Table 2. Household population by age and sex ................................................................................................... 31 Table 3. Household by type, size, and select characteristics............................................................................... 32 Table 4. Sex, age, and education of household heads.......................................................................................... 34 Table 5. Sex, age, and education of household heads by wealth quintile.......................................................... 35 Table 6. Household possessions.............................................................................................................................. 37 Table 7. Housing characteristics.............................................................................................................................. 39 Table 8. Water and sanitation ................................................................................................................................. 40 Table 9. Health-related assistance/support provided to households................................................................ 43 Table 10. Characteristics of WRAs......................................................................................................................... 45 Table 11. Education of WRA by age, wealth, and district .................................................................................. 47 Table 12. Percentage of WRA who were exposed to specific media on a weekly basis ....... 50 Table 13. Exposure to Moyo ndi mpamba, Usamalireni!.................................................................................... 52 Table 14. Satisfaction with health services among WRA.................................................................................... 53 Table 15. Current use of contraception among WRA......................................................................................... 56 Table 16. Contraceptive use among married or sexually active women who were not pregnant and who did not desire more children.................................................................................................................... 57 Table 17. Informed choice among current users of modern methods ages 15 to 49 who started the last episode of contraceptive use in the three years before the survey..................................................................... 59 Table 18. Beliefs about FP, percentage of WRA who believed statements were completely true ............... 62 Table 19. Women’s ANC in the past three years.................................................................................................. 66 Table 20. Use of SP/Fansidar during last pregnancy in the past three years................................................... 67 Table 21. Women’s delivery care in the past three years..................................................................................... 68 Table 22. Women’s PNC in the past two years among those who delivered at a health facility .................. 70 Table 23. Table 23. Newborn PNC in the past two years among those who were delivered at a health facility........................................................................................................................................................ 71 Table 24. Women’s PNC in the past two years among those who did not deliver at a health facility ........ 72 Table 25. Newborn PNC in the past two years among those who were not delivered at a health facility........................................................................................................................................................ 73 Table 26. Characteristics of children under three years of age........................................................................... 78 Table 27. Breastfeeding observation and counseling among last births in facilities in the past two years.................................................................................................................................................. 80 Table 28. Breastfeeding observation and counseling among last births outside of health facilities in the past two years....................................................................................................................................................... 81 Table 29. Percentage of children who were still being breastfed among last births in the past three years.............................................................................................................................................................................. 83 Table 30. Prevalence and treatment of fever in children under three in the past two weeks........................ 85 Table 31. Prevalence and treatment of ARI in children under three in the past two weeks......................... 87 12 Malawi ONSE Impact Evaluation: Baseline Report Table 32. Prevalence and treatment of diarrhea in children under three in the past two weeks .................. 88 Table 33. Percentage of WRA who correctly identified mosquitos as the cause of malaria.......................... 90 Table 34. Characteristics of sampled health facilities........................................................................................... 94 Table 35. Availability of basic client services......................................................................................................... 95 Table 36. Availability of basic amenities for client services................................................................................ 96 Table 37. Percentage of facilities with priority medicines for mothers............................................................. 98 Table 38. Percentage of facilities with priority medicines for children ............................................................. 99 Table 39. Availability of FP methods among facilities that provided FP services........................................101 Table 40. Availability of guidelines, trained staff, and equipment for FP services........................................102 Table 41. Availability of maternal health services...............................................................................................104 Table 42. Signal functions of BEmONC and CEmONC performed in the past 12 months.....................105 Table 43. Availability of guidelines, trained staff, and equipment for delivery services...............................106 Table 44. Percentage of facilities offering ANC that had indicated medicines .............................................109 Table 45. Availability of items for infection control during provision of ANC............................................110 Table 46. Availability of malaria services in facilities offering ANC ...............................................................112 Table 47. Availability of child health services .....................................................................................................113 Table 48. Availability of guidelines, trained staff, and equipment for child curative services.....................115 Table 49. Availability of malaria services, and guidelines, trained staff, and diagnostic capacity ...............118 Table 50. Availability of malaria medicines .........................................................................................................120 Malawi ONSE Impact Evaluation: Baseline Report 13 ABBREVIATIONS ACT artemisinin combination therapy ANC antenatal care ANC 4+ four or more ANC visits during pregnancy ARI acute respiratory infection BEmONC CDCS basic emergency obstetric and newborn care Country Development Cooperation Strategy CEmONC CHAM comprehensive emergency obstetric and newborn care Christian Health Association of Malawi CSR Centre for Social Research DHIS 2 District Health Information System Version 2 DHMT District Health Management Team DHS Demographic and Health Survey DID difference-in-differences EA enumeration area FHP family health package FP family planning HSS health systems strengthening IE impact evaluation IMCI integrated management of childhood illnesses IMPAC integrated management of pregnancy and childbirth IPT intermittent prevention treatment IPTp intermittent prevention treatment during pregnancy ITN insecticide treated net IUD intrauterine device MCPR modern contraceptive prevalence rate MDG Millennium Development Goal MES MDG Endline Survey MICS Multiple Indicator Cluster Survey MIS Malaria Indicator Survey MNCH maternal, neonatal, and child health ONSE Organized Network of Services for Everyone ORS oral rehydration salts PNC postnatal care RDT rapid diagnostic test RH reproductive health RMNCH reproductive, maternal, newborn, and child health SARA Service Availability and Readiness Assessment SBCC strategic behavior change communication SPA Service Provision Assessment SSDI Support for Service Delivery Integration 14 Malawi ONSE Impact Evaluation: Baseline Report USAID United States Agency for International Development WASH water, sanitation, and hygiene WHO World Health Organization WRA women of reproductive age Malawi ONSE Impact Evaluation: Baseline Report 15 EXECUTIVE SUMMARY Introduction Led by the University of North Carolina at Chapel Hill, MEASURE Evaluation is conducting an impact evaluation of the Malawi Organized Network of Services for Everyone’s (ONSE) Health project. ONSE is implemented by Management Sciences for Health with support from the United States Agency for International Development (USAID) Malawi Mission. ONSE aims to reduce maternal, newborn, and child morbidity and mortality by improving access to, the quality of, and demand for priority health services. Of interest to the impact evaluation is ONSE’s “smart approach” to capacity building and problem solving. “Smart” capacity building involves the co-development of capacity building plans with District Health Management Teams (DHMTs) to identify district-specific capacity building needs related to the provision of health services. An example of “smart” problem solving is the use of locally collected data by DHMTs to identify health priorities and develop solutions to problems identified. Background Malawi is one of the few countries in sub-Saharan Africa to achieve Millennium Development Goal (MDG) 4 for child survival, which aimed to reduce the under-five mortality rate by two-thirds between 1990 and 2015. Nevertheless, neonatal, infant, and under-five mortality rates remain high. The major causes of infant and child death in Malawi are pneumonia, malaria, diarrhea, HIV/AIDS, and malnutrition. Malaria is endemic in Malawi and accounts for more than 40 percent of all hospitalizations for children under five (National Statistics Office [Malawi] and ICF, 2017.). There are several other factors that contribute to high under-five mortality rates. Only one-half of all Malawian children ages 12 to 23 months have received all age-appropriate vaccinations, and only one in 10 children ages 24 to 35 months have received all age-appropriate vaccinations. Nearly four in 10 (37 percent) children under five are stunted, an indication of chronic undernutrition. Water and sanitation-related diseases are also a leading cause of death for children under five (National Statistics Office [Malawi] and ICF, 2017). Although nine in 10 births in Malawi occur in health facilities, maternal mortality in Malawi remains high at 439 deaths per 100,000 live births, and one-half of women do not receive a postnatal checkup within 41 days of delivery. The main causes of maternal mortality are postpartum hemorrhage, eclampsia, and sepsis. Malawi has made progress in family planning/reproductive health (FP/RH), with about six in 10 married women ages 15 to 49 using a modern method of FP (National Statistics Office [Malawi] and ICF, 2017). Impact Evaluation Research Questions and Methods The ONSE impact evaluation aims to answer three primary research questions: • What is the impact of the ONSE project on changes in health and facility outcomes compared with changes in these outcomes in districts that did not receive the ONSE project? 16 Malawi ONSE Impact Evaluation: Baseline Report • How was “smart” capacity building and problem solving operationalized in each district? • What is the impact of ONSE’s community engagement and mobilization activities in communities where this intervention was implemented compared with communities that did not receive this intervention? The evaluation employs a quasi-experimental design in which pre- and post-differences in the outcomes of interest will be compared between project and comparison districts to measure the impact of ONSE (Table ES1). Project districts were purposively selected. They are Machinga, Nkhotakota, and Salima. Comparison districts were selected based on recent estimates of key health indicators. They are Mzimba, Ntchisi, and Nsanje. Data collection methods consisted of a baseline (2017) and end line (2021) quantitative household survey, a baseline and end line health facility assessment, implementation process monitoring (2018–2020), and an end line qualitative study. In collaboration with a local research partner, the Centre for Social Research (CSR), MEASURE Evaluation collected baseline data from April 28 to June 30, 2017 on the intended health and facility outcomes of the ONSE project. This report presents the findings of the household and facility baseline surveys. A difference in differences (DID) analysis is planned for the impact evaluation at the time of the end line survey. Results of the evaluation will be used to inform future health programming in Malawi and will generate evidence on approaches that tailor project activities at the district level based on varying needs. Table ES1. Primary and secondary outcomes of interest Primary outcome areas Source: Household survey Secondary outcome areas Source: Facility survey • Antenatal care • Maternal health • Postnatal care (PNC) • FP • Care seeking for children under three • Patient satisfaction • Women’s knowledge of key maternal and newborn warning/danger signs • Women’s knowledge of symptoms and causes of key childhood illnesses • Beliefs about FP • Exposure to behavior change messaging • General service availability • Availability and readiness to provide services for: ✓ FP ✓ Antenatal care (ANC) ✓ Basic obstetric and newborn care ✓ Comprehensive obstetric care ✓ Preventative and curative child health services ✓ Malaria services Malawi ONSE Impact Evaluation: Baseline Report 17 Results The household survey reached 7,929 households and 7,542 women of reproductive age (WRA) in three project and three comparison districts. A summary of key health indicators for ANC, birth attendance, PNC, and contraceptive use is provided in Table ES2. The receipt of skilled ANC and skilled birth attendance were almost universal in both the project and comparison domains. However, only about one-third of women received ANC in their first trimester and only one-half of pregnant women receive the recommended four or more ANC visits during their pregnancy. The rate of postnatal checks was about two-thirds, with almost all of these women receiving PNC within two days of childbirth. The modern contraceptive prevalence rate was approximately 55 percent among WRA who are married or living with a man, and was approximately 46 percent among all WRA. The household survey also asked women about key danger signs for specific maternal and child health issues to gauge their knowledge of priority topics for the Government of Malawi and in general. Knowledge was very low about most warning and danger signs during pregnancy, childbirth, and for newborn complications. Few women knew what types of issues to include in a birth plan, with the least frequently considered topic being who would care for the other children in the household during childbirth. Severe bleeding during pregnancy and childbirth were the only warning/danger signs reported by more than one-half of women surveyed. Only approximately one-third of women knew that difficulty breathing and high fever were serious danger signs for newborns. The Government of Malawi’s National Health Communications Strategy also calls for improving knowledge of key symptoms and causes of diarrhea, pneumonia, and malaria. Knowledge of symptoms and the cause of malaria were the most well-known of these three infectious diseases. Women reported that breathing problems (approximately 60 percent) and not dressing warmly enough (70 percent) were symptoms and causes of pneumonia, with almost no knowledge of other symptoms or causes. Approximately 40 percent of women in both domains knew that loose and watery stools for more than three days were a symptom of diarrhea. Knowledge of the causes of diarrhea was higher than that for pneumonia. Between 35 and 40 percent of the women in both domains reported lack of safe drinking water and food contamination as causes of diarrhea, and about one-fifth of women reported eating rotten food and not washing hands after defecation as other causes. 18 Malawi ONSE Impact Evaluation: Baseline Report Table ES2. Summary of health outcomes Health outcome Project Comparison Total ANC Skilled ANC 98.6 96.9 97.7 ANC visits in the first trimester 30.8 35.2 32.9 Four or more ANC visits during pregnancy 51.8 55.8 53.7 Number of live births in the past three years 1,801 1,598 3,399 Birth attendance Skilled birth attendance 93.5 94.7 94.1 Number of births in the past three years 1,801 1,598 3,399 PNC Women receiving postnatal health checks 63.7 65.1 64.3 Women receiving postnatal health check within two days of birth 61.1 59.3 60.3 Number of live births in the past two years 1,333 1,147 2,480 Contraceptive use Modern contraceptive prevalence rate among WRA who are married or living with a man 55.5 53.7 54.6 Number of WRA who are married or living with a man 2,522 2,178 5,240 Modern contraceptive prevalence rate among all women 45.6 45.8 45.7 Number of WRA who are married, living with a man, or unmarried and sexually active 3,582 3,576 7,158 Total number of WRA 3,776 3,766 7,542 Malawi ONSE Impact Evaluation: Baseline Report 19 The health facility survey was administered to all public and Christian Health Association of Malawi (CHAM) facilities in the project and comparison domains. The survey revealed high availability of services for FP, ANC, and basic obstetric care (more than 90 percent in almost all facilities). Readiness to provide services was more varied. The readiness to provide services indicators were low across most dimensions of the measure, including staffing and guidelines, medicines and commodities, diagnostics, and equipment. Staffing and guidelines were lacking in all areas, although FP readiness on this dimension reached more than 30 percent in the comparison domain. The availability of medicines and commodities varied by service type but was also low. Malaria and ANC were the two areas with the greatest availability of medicines and commodities, with 32 percent and 24 percent of the facilities, respectively, having all required items on the day of the visit. Although some facilities had the required resources to meet staffing and guidelines requirements for basic obstetric and newborn care, very few met other requirements for basic obstetric and newborn care or comprehensive care. Obstetrics was the area with the lowest general readiness of all service types. Conclusions The Malawi ONSE impact evaluation seeks to test the hypothesis that the interventions implemented by ONSE will improve health outcomes for women, newborns, and children in the project domain compared to the comparison domain. The household survey conducted in 2017 as part of the Malawi ONSE impact evaluation establishes baseline indicators for household and women’s background characteristics, primary outcomes, and exposure to project or similar interventions in both the project and comparison domains. The health facility survey, conducted at the same time as the household survey, establishes baseline estimates for secondary outcomes related to the availability of health services and facility readiness to provide specific services in both the project and comparison domains. The baseline survey reveals important information about the project and comparison domains that will be considered during the end line impact analysis. First, about one-third of households in both domains reported receiving support from a related intervention in the 12 months leading up to the survey. Malaria was the most commonly reported area of support, followed by water, sanitation, and hygiene (WASH). More information will be sought regarding these projects and the potential of their activities to influence the ONSE project’s outcomes. With this information, an appropriate strategy will be developed to control for any contamination resulting from outside interventions during the ONSE project. Second, the surveys reveal important similarities and differences in households and facilities in the evaluation sample. Facility service availability and readiness were generally similar in project and comparison domains. Methodological techniques will be applied at end line, where necessary, to account for these differences. Next Steps End line data collection is planned for 2021. The same households will be interviewed at that time to evaluate the impact of ONSE on the health outcomes of interest in the project domain. The DID approach will be 20 Malawi ONSE Impact Evaluation: Baseline Report used to compare pre- and post-intervention differences in outcomes between the project and comparison domains. Qualitative analysis will aim to describe and understand differences in how respondents in ONSE’s project communities were exposed to its strategic behavior change communication (SBCC) campaign. Ongoing implementation process monitoring will occur through the time of the end line survey. The monitoring will focus on how the “smart” approach was operationalized in the project domain and will seek to identify the pathways through which this approach affects project beneficiaries. Implementation process monitoring will also provide information about exposure to other activities that may affect the outcomes of the impact evaluation. Malawi ONSE Impact Evaluation: Baseline Report 21 INTRODUCTION Led by the University of North Carolina at Chapel Hill, MEASURE Evaluation is conducting an impact evaluation of the ONSE project. ONSE is implemented Management Sciences for Health1 with support from the USAID Malawi Mission. ONSE aims to reduce maternal, newborn, and child morbidity and mortality by improving access to, the quality of, and demand for priority health services. The ONSE impact evaluation employs a quasi-experimental design in which outcomes of interest are compared between project and comparison districts over time to measure the impact of ONSE. Data collection methods include a baseline (2017) and end line (2021) quantitative household survey, a baseline and end line health facility assessment, implementation process monitoring (2018–2020), and an end line qualitative study. In collaboration with the local research partner, CSR, MEASURE Evaluation collected baseline data from April 28 to June 30, 2017 on the intended health and facility outcomes of ONSE. This report presents the findings of the household and facility baseline surveys. Background Malawi is one of the few countries in sub-Saharan Africa to achieve MDG 4 for child survival, which aims to reduce the under-five mortality rate by two-thirds between 1990 and 2015. Nevertheless, neonatal, infant, and under-five mortality rates remain high at 27, 42, and 63 per 1,000 live births, respectively (National Statistics Office [Malawi] and ICF, 2017). The major causes of infant and child death in Malawi are pneumonia, malaria, diarrhea, HIV/AIDS, and malnutrition. Nine in 10 births in Malawi occur in a health facility. The country has experienced significant gains in skilled assistance during delivery, which increased from 55 percent in 1992 to 90 percent in 2015‒16. Nevertheless, maternal mortality in Malawi remains high, at 439 deaths per 100,000 live births. One-half of women do not receive a postnatal checkup within 41 days of delivery (National Statistics Office [Malawi] and ICF, 2017). The main causes of maternal mortality in Malawi are postpartum hemorrhage, eclampsia, and sepsis. Malawi has also made progress in FP/RH. About six in 10 (58 percent) married women ages 15 to 49 use a modern method of FP. However, nearly one in five married women have an unmet need for FP, defined as the proportion of married women who want to delay or stop childbearing but are not using FP (National Statistics Office [Malawi] and ICF, 2017). Only one-half of all Malawian children ages 12 to 23 months have received all age-appropriate vaccinations, and only one in 10 children ages 24 to 35 months have received all age-appropriate vaccinations. Nearly four in 10 (37 percent) children under five in Malawi are stunted, an indication of chronic undernutrition (National Statistics Office [Malawi] and ICF, 2017). 1 Award number AID-612-C-17-00001, with award dates November 15, 2016 to November 15, 2021. 22 Malawi ONSE Impact Evaluation: Baseline Report Malaria is endemic in Malawi, accounting for more than 40 percent of all hospitalizations of children under five. Use of insecticide treated nets to prevent malaria increased from 39 percent in 2010 to 43 percent in 2015‒16 among children under five, and from 35 percent to 44 percent among pregnant women in the same period. In the two weeks before the 2015‒16 Demographic and Heath Survey (DHS), 29 percent of children under five had fever, the primary symptom of malaria. Although treatment/advice was sought for two-thirds of these children, only one-half had blood taken for testing. Water and sanitation-related diseases are also a leading cause of death for children under five. Approximately 85 percent of rural households in Malawi have access to an improved source of drinking water, whereas only one-half of all households use improved toilet facilities. According to the 2015‒16 Malawi DHS, although a place for washing hands was observed in 83 percent of households, soap and water were observed in only 11 percent of handwashing locations. Another 26 percent of handwashing locations had just water. The Malawi ONSE Project The continued improvement in health outcomes and reductions in neonatal, infant, child, maternal morbidity, and mortality in Malawi will largely depend on improving access to and the quality of essential services and the supporting health system. The Malawi ONSE project is USAID/Malawi’s flagship health program and focuses on these improvements. Working in 16 districts, ONSE will support more than 400 health facilities that provide essential health care services to over one-half of Malawi’s population. The ONSE project theory of change is depicted in Figure 1. Figure 1. Malawi ONSE theory of change Malawi ONSE Impact Evaluation: Baseline Report 23 The goal of ONSE is to improve health outcomes, including maternal, newborn, and child survival and well￾being in Malawi. The objectives of the project’s targeted intermediate outcomes are: • Increase access to priority health services for maternal, newborn, and child health (MNCH); FP/RH; malaria; and WASH. • Improve the quality of priority health services. • Strengthen the performance of health systems. • Increase demand for priority health services. Key interventions, or inputs, are: • Health systems strengthening (HSS): HSS will be implemented in all 16 ONSE districts to improve the management and supervision of human resources for health, governance, policy implementation, and use of data for decision making at the district level. • A family health package (FHP): The FHP is ONSE’s service delivery component and will be implemented in 11 of the 16 project districts. The FHP focuses on improving access to and the quality of MNCH, FP/RH, and WASH services. • Malaria services: Malaria services will be integrated with the FHP and will be provided in 10 of the 16 project districts based on need. Table 1 presents the ONSE interventions, by district. 24 Malawi ONSE Impact Evaluation: Baseline Report Table 1. ONSE interventions, by district District Intervention HSS FHP Malaria Balaka • • • Lilongwe • • • Machinga • • • Salima • • • Nkhotakota • • • Mulanje • • Kasungu • • Dowa • • Chitipa • • Karonga • • Zomba • • Mchinji • • Chikwawa • • Mangochi • • Ntcheu • • Nkhata Bay • • Malawi ONSE Impact Evaluation: Baseline Report 25 “Smart” Capacity Building and Problem Solving A strategic principle employed by ONSE is the “smart” approach, comprised of smart capacity building and smart problem solving. Smart capacity building entails the co-development of capacity building plans with DHMTs to identify district-specific capacity building needs in MNCH, FP/RH, and malaria. Smart capacity building relies on performance improvement methods and onsite mentoring of healthcare workers and DHMT members. To help district health officials overcome challenges to improving health in their districts, ONSE also supports smart problem solving. One example of smart problem solving is the application of the District Health Program Improvement strategy, which helps district officials use locally collected data to identify health priorities and develop solutions to problems. Community Engagement and Mobilization A second strategic principle employed by ONSE is community engagement and mobilization to increase demand for and uptake of health services. ONSE’s community engagement and mobilization strategy involves working with community organizations and stakeholders to promote SBCC messages to strengthen knowledge about healthy behaviors and reduce harmful practices, with a focus on underserved groups. ONSE will work in collaboration with the national behavior change communications project to support activities in ONSE project districts. The ONSE Impact Evaluation The hypothesis of the impact evaluation is that ONSE will improve health outcomes by improving the quality of and access to health services. At baseline, we expect health and facility indicators to be similar in project and comparison districts. After completion of the project (e.g., at end line), we expect to see more improvement in health and facility outcomes in project districts than in comparison districts. The evaluation is designed to understand how the projects’ strategic approaches (i.e., smart capacity building, smart problem solving, and community engagement and mobilization) improve the effectiveness of project interventions. The ONSE impact evaluation aims to answer three primary research questions: 1) What is the impact of the ONSE project on changes in health and facility outcomes compared with changes in these outcomes in districts that did not receive the ONSE project? 2) How was smart capacity building and problem solving operationalized in each district? 3) What is the impact of ONSE’s community engagement and mobilization activities in communities where this intervention was implemented compared with communities that did not receive this intervention? Results of the evaluation will be used to inform future health programming in Malawi and will generate evidence related to approaches that tailor project activities at the district level based on varying needs. 26 Malawi ONSE Impact Evaluation: Baseline Report METHODS The ONSE impact evaluation employs a quasi-experimental design in which pre- and post-differences in outcomes will be compared between project and comparison districts to measure the impact of ONSE. Data collection methods are a baseline (2017) and end line (2021) quantitative household survey, a baseline and end line health facility assessment, implementation process monitoring (2018–2020), and an end line qualitative study. Each data collection method is summarized below and is described in more detail in Appendix D. The purpose of the baseline survey is to generate estimates of key outcome measures before the start of the ONSE project. The survey also identifies key differences between the outcomes in the comparison and project domains selected for the evaluation. This report presents the results of the baseline survey. Selection of Study Districts Three ONSE intervention districts (Machinga, Nkhotakota, and Salima) were chosen as the evaluation study districts. Mzimba, Nsanje, and Ntchisi, which are not supported by USAID, were selected as comparison districts. Details about the selection process are provided in Appendix C. Quantitative Household Survey A household survey was developed to measure population-level outcomes of interest among both married and unmarried WRA, ages 15 to 49. The household survey has a household questionnaire and a woman’s questionnaire, which were administered to all consenting WRA in selected households. The survey incorporated questions from the 2015‒16 Malawi DHS drawn from the household questionnaire and woman’s questionnaire modules on respondent background, reproduction, contraception, pregnancy and PNC, child health and nutrition, marriage and sexual activity, fertility preferences, husband’s background, and woman’s work. Additional modules were included on knowledge of maternal and newborn health, knowledge of child health, attitudes towards FP, and patient satisfaction.2 The quantitative household survey was conducted at baseline in all study districts, with the goal of measuring health outcomes at the population level in each study group (i.e., project and comparison districts). A summary of modules used are provided in Table C1 in Appendix C. The household survey instrument can be found in Appendix D. The household survey adopted a multi-stage cluster sampling design to obtain a random sample of households from the project and comparison districts. The two domains for sampling were project and comparison. Enumeration areas (EAs) were chosen randomly in each domain and then 30 households were chosen randomly in each EA (i.e., cluster). Forest reserves were excluded from these domains because they are not generally inhabited. 2 Custom knowledge and attitude indicators were developed by the evaluation team based on the Malawi National Health Communication Strategy and MEASURE Evaluation FP/RH indicators. Custom patient satisfaction indicators were developed by the evaluation team using ONSE’s Performance Management Plan. Malawi ONSE Impact Evaluation: Baseline Report 27 The study design requires an adequate sample of women with a birth in the past three years to measure changes in the percentage who received four or more ANC visits during pregnancy (ANC 4+). The sample size estimate is powered to detect a 10.0 percentage point difference in ANC 4+ between project and comparison districts over the five-year project period. The estimated minimum number of households to detect that change is a sample size of 8,030 households, resulting in approximately 6,633 WRA among whom an estimated 3,245 women would have had with a birth in the past three years.3,4 Information about the sampling frame and sampling weights is provided in Appendix C. The response rate for the household questionnaire was 98.6 percent in the project domain and 98.8 percent in the comparison domain. The response rate for the woman’s questionnaire was 97.5 percent in the project domain and 96.4 percent in the comparison domain (Table C2 in Appendix C). The primary outcomes of interest to the evaluation are population-level health outcomes. ANC 4+ was chosen as the main outcome due to its effect on ONSE’s long-term outcomes of interest for both women and infants―maternal and neonatal mortality and morbidity (MEASURE Evaluation, n.d.). In addition to health outcomes, secondary facility-level outcomes were also included and are described in the health facility assessment section below. Table C4 in Appendix C presents the primary population-level outcomes of interest estimated from the household survey. At baseline, population-level indicators for ANC, maternal health, PNC, and RH were calculated according to the Guide to DHS Statistics. 5 Some adjustments to recall times were made to accommodate the five-year duration of the ONSE project. Quantitative data analysis was conducted in Stata 14.2 (Stata Corp LP). The analysis of household survey data in the baseline report was limited to basic descriptive frequencies and cross tabulations. Emphasis is placed on the estimates of indicators and household and women’s characteristics in the project and comparison domains at baseline. Indicators are reported mainly as either percentages or means and are weighted using the sampling weights. 3 Powered at 80 percent; alpha=0.05; design effect = 2.2. A 10 percentage point increase is projected from an estimated 49 percent among women with a birth in the past three years (National Statistics Office [Malawi] and ICF, 2017). 4 It should be noted that the sample size calculations are approximations under the provided assumptions. The end line impact evaluation analysis will have components that increase the statistical power (e.g., individual covariates that reduce the unexplained variance) and some that may reduce it (e.g., unknown design effects), but they cannot be estimated with certainty in advance. 5 See https://dhsprogram.com/publications/publication-dhsg1-dhs-questionnaires-and-manuals.cfm. 28 Malawi ONSE Impact Evaluation: Baseline Report Health Facility Assessment A health facility assessment tool was developed to measure facility-level outcomes of interest. Although the facility outcomes are of secondary interest, they are critical outcomes along the casual pathway to improved health. The health facility survey was administered at all public dispensaries, health centers, and hospitals, 6 and at all CHAM health facilities7 in both project and comparison domains. The health facility assessment tool incorporated select questions from the Service Availability and Readiness Assessment (SARA)8 related to general service availability and specific service availability and readiness for the following topics of interest: FP; maternal health services, including ANC, normal delivery, cesarean delivery, and basic and comprehensive emergency obstetric and newborn care; preventative and curative child health services; and malaria services. The facility assessment tool can be found in Appendix D. The heath facility sample consisted of a census of district hospitals, health centers, and dispensaries in the six study districts. All CHAM facilities were also assessed. A list of facilities was obtained from the Ministry of Health and was updated during meetings with district health officers in each district. The response rate was 100 percent in all districts, resulting in a total sample of 139 health facilities (Table C3 in Appendix C). Improvements in access to and the quality of facility services contribute to improvements in health. Therefore, the impact evaluation includes measures of facility service availability and readiness as critical intermediate outcomes in the casual pathway to improving health. Table C5 in Appendix C provides a summary of facility-level outcomes derived from the SARA in project and comparison domains for the impact evaluation at baseline. Baseline health facility indicators were calculated according to the SARA Reference Manual9 and modified based on the needs of the evaluation stakeholders. SARA items that were not included in the Malawi national guidelines were omitted from the indicators and some additional questions were added. Stata 14.2 was used to analyze the facility data to generate descriptive frequencies and statistics according to the reference manual and specific requests of USAID/Malawi. Details about training for and implementation of the household and facility surveys can be found in Appendix E. 6 One central hospital and all military facilities were excluded from the facility sample. 7 CHAM health facilities are privately run facilities and comprise the largest network of nongovernmental healthcare facilities in Malawi. 8 See http://www.who.int/healthinfo/systems/sara_introduction/en//. 9 See http://www.who.int/healthinfo/systems/SARA_Reference_Manual_Chapter2.pdf?ua=1. Malawi ONSE Impact Evaluation: Baseline Report 29 Strengths and Limitations This evaluation provides for triangulation of data from several sources, including longitudinal data collection in households and facilities. This will allow for comparison of indicators in and across facilities and households over the course of the project and will strengthen the analyses for all research questions. There is some potential for unobserved factors between groups to bias the impact evaluation results. The DID approach is a rigorous quasi-experimental design that controls for observed and unobserved differences between project and comparison domains that are constant over time. Unobserved differences are assumed to be time-invariant, and thus, the DID approach does not account for any time-varying unobserved differences between the project and comparison domains. Implementation process monitoring data collection will collect additional information about potential shocks affecting the study groups differently and/or other outside factors that may influence the study’s outcomes. A variation in Country Development Cooperation Strategy (CDCS) integration between USAID implementing partners in the project domain has the potential to generate additional improvements in the project domain above and beyond the activities of the ONSE project. The extent of and variation in this type of integration will be monitored at regular intervals and robustness checks may be used during the impact analysis, as needed, to account for CDCS integration-related activities. 30 Malawi ONSE Impact Evaluation: Baseline Report RESULTS HOUSEHOLD POPULATION AND HOUSING CHARACTERISTICS Socioeconomic characteristics are important for understanding population-level health indicators and comparing the baseline status of project and comparison households. This section presents the demographics of the household population, by study domain, including the distribution of household members by age group and sex; household type and size; sex, age, and education level of household heads; and household wealth and possessions. Housing characteristics are also presented, by study domain. Household Population Age and Sex of Household Members The percentage distribution of the household population by age group and sex is presented in Table 2. The distribution by sex and age group was similar in the project and comparison domains. Females comprised 51.6 percent of household members in both domains. Approximately one-half of household members in both domains were under age 15. Malawi ONSE Impact Evaluation: Baseline Report 31 Table 2. Household population by age and sex (ONSE impact evaluation [IE] baseline, 2017) Characteristics Project Comparison Total study sample N Sex Male Female Total Male Female Total Male Female Sex 48.4 51.6 100.0 48.4 51.6 100.0 48.4 51.6 100.0 Number 9,025 9,736 18,761 9,072 9,659 18,731 18,097 19,395 37,492 Age Percent <5 17.3 15.7 16.5 15.4 14.4 14.8 15.7 5,859 5-14 34.3 34.5 34.3 33.7 34.0 33.6 34.1 12,715 15-19 11.3 8.6 9.9 11.0 8.6 9.8 9.8 3,690 20-24 7.4 8.7 8.0 7.8 8.2 7.9 8.0 2,981 25-29 6.0 6.2 6.1 6.0 5.9 6.0 6.0 2,268 30-34 4.9 5.7 5.4 5.2 5.7 5.6 5.4 2,059 35-39 4.7 4.7 4.9 4.6 5.1 4.9 4.8 1,829 40-44 3.9 3.3 3.4 3.8 3.5 3.7 3.6 1,330 45-49 2.4 2.5 2.5 3.0 2.8 2.9 2.7 1,008 Age Percent 50-64 4.5 5.9 5.2 5.9 7.3 6.7 5.9 2,237 64+ 3.4 4.3 3.9 3.7 4.5 4.2 4.0 1,516 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 37,492 Age groups Under five 17.4 15.7 16.5 15.2 14.4 14.8 15.7 5,859 Adolescents1 27.0 25.8 26.5 27.0 26.5 26.6 26.6 9,955 10-14 15.8 17.3 16.6 16.0 17.9 17.0 16.8 6,265 15-19 11.3 8.5 9.9 11.0 8.6 9.8 9.9 3,690 WRA - 39.7 - - 39.8 - - 15,165 N 25.0 9,736 18,761 9,072 9,659 18,731 18,731 37,492 1Adolescents include members ages 10 to 19 years of age. 32 Malawi ONSE Impact Evaluation: Baseline Report Household Type and Size The average household size was 4.7 members in the project and comparison domains, and the average number of WRA (1.0) was the same in both domains (Table 3). The proportion of households in the lowest wealth quintile was higher in the project domain (26.3 percent) than in the comparison domain (20.2 percent). At the same time, the proportion of households in the highest wealth quintile was higher in the comparison domain (20.6 percent) than in the project domain (16.4 percent). Nearly 90 percent of households in the project domain resided in rural areas as compared to 96.3 percent of households in the comparison domain. Table 3. Household by type, size, and select characteristics (ONSE IE baseline, 2017) Characteristic Project Comparison N Household size Average household size 4.7 4.7 7,929 Average number of adults ages 18 to 64 1.9 2.0 7,929 Average number of elderly ages 65 and older 0.2 0.2 7,929 Average number of WRA 1.0 1.0 7,929 Average number of children under five 0.8 0.7 7,929 Wealth quintile Lowest 26.3 20.2 1,834 Second 21.8 19.4 1,599 Middle 18.5 19.8 1,528 Fourth 17.0 20.1 1,475 Highest 16.4 20.6 1,493 Total 100.0 100.0 7,929 District Machinga 38.9 - 1,841 Nkhotakota 25.8 - 1,120 Salima 35.3 - 1,001 Mzimba - 66.4 2,671 Nsanje - 14.3 586 Ntchisi - 19.2 710 Total 100.0 100.0 7,929 Malawi ONSE Impact Evaluation: Baseline Report 33 Characteristic Project Comparison N Rurality Peri-urban 10.2 3.7 527 Rural 89.8 96.3 7,402 Total 100.0 100.0 7,929 N 3,962 3,967 7,929 Sex, Age, and Education of Household Heads Tables 4 and 5 present findings related to the sex, age, and education of household heads. Although just under two-thirds of households in the project domain were male-headed, nearly 70 percent of comparison households were male-headed (Table 4). In both domains, the proportion of female-headed households decreased as the wealth of the household increased (Table 5). The age distribution of household heads was similar in project and comparison domains (Table 4). In both study domains, household heads most frequently reported that they had some primary education or had completed their primary education (59.5 percent in the project domain and 63.0 percent in the comparison domain) (Table 4). In the project domain, no education among household heads (21.6 percent) was more frequently reported than in the comparison domain, where only 12.5 percent of household heads had no education. In both domains, the proportion of household heads with some secondary or higher education increased as the wealth quintile increased (Table 5). 34 Malawi ONSE Impact Evaluation: Baseline Report Table 4. Sex, age, and education of household heads (ONSE IE baseline, 2017) Characteristic Project Comparison N Sex of household head Male 65.7 69.8 5,379 Female 34.3 30.2 2,547 Total* 100 100 7,926 Age of household head 0-14** 0.0 0.0 1 15-19 1.1 1.2 91 20-24 9.6 8.9 721 25-29 14.2 12.1 1,035 30-34 13.8 13.9 1,110 35-39 13.3 12.5 1,039 40-44 10.8 10.5 830 45-49 7.8 8.7 648 50-64 15.6 18.5 1,359 64+ 13.8 13.7 1,092 Total 100.0 100.0 7,926 Education of household head No education 21.6 12.5 1,356 Some /completed primary 59.5 63.0 4,809 Some /completed secondary 15.7 21.0 1,479 More than secondary 3.1 3.5 282 Total 100.0 100.0 7,926 N 3,961 3,965 7,926 * Three households did not report head of household. **One 13-year-old household head is represented Malawi ONSE Impact Evaluation: Baseline Report 35 Table 5. Sex, age, and education of household heads by wealth quintile (ONSE IE baseline, 2017) Project Comparison N Lowest quintile Second quintile Middle quintile Fourth quintile Highest quintile Lowest quintile Second quintile Middle quintile Fourth quintile Highest quintile Sex of household head Male 23.9 20.5 19.3 17.6 18.8 18.6 19.2 19.8 20.7 21.8 5,377 Female 30.7 24.3 17.1 15.9 12.0 23.7 19.9 19.9 18.7 17.9 2,546 Total* 26.2 21.8 18.5 17.0 16.4 20.2 19.4 19.8 20.1 20.6 7,923 Age of household head 13-19 64.8 11.9 14.2 2.6 6.5 20.3 22.1 12.6 15.0 30.1 92 20-24 45.9 25.0 13.9 9.2 6.0 34.1 21.8 22.2 14.0 7.8 721 25-29 34.6 20.8 20.3 10.6 13.6 25.5 22.9 18.3 18.7 14.6 1,034 30-34 29.7 19.5 16.1 16.0 18.7 19.9 21.3 19.7 19.0 20.1 1,109 35-39 19.7 22.8 16.3 19.2 21.9 18.7 17.7 21.5 17.6 24.5 1,038 40-44 19.5 19.7 18.9 22.1 19.8 15.6 18.8 16.6 22.9 26.1 830 45-49 17.5 20.8 21.7 17.1 22.9 14.6 15.1 19.2 22.5 28.6 648 50-64 18.8 22.0 20.9 19.4 18.8 14.8 18.8 20.6 21.6 24.3 1,427 64+ 22.3 24.9 20.0 22.6 10.1 23.0 18.0 20.3 23.4 15.4 1,024 Total 26.2 21.8 18.5 17.0 16.4 20.2 19.4 19.8 20.1 20.6 7,923 Education of household head No education 36.0 23.9 18.3 16.6 5.2 36.4 22.0 18.2 16.6 6.8 1,355 Some/completed primary 27.3 24.8 19.8 16.9 11.1 20.8 21.8 21.6 20.6 15.2 4,807 36 Malawi ONSE Impact Evaluation: Baseline Report Project Comparison N Lowest quintile Second quintile Middle quintile Fourth quintile Highest quintile Lowest quintile Second quintile Middle quintile Fourth quintile Highest quintile Some/completed secondary 13.6 12.0 17.5 19.3 37.6 12.1 13.4 17.6 22.2 34.6 1,479 More than secondary 1.0 0.0 0.6 9.9 88.5 0.0 2.3 5.6 9.8 82.3 282 Total 26.2 21.8 18.5 17.0 16.4 20.2 19.4 19.8 20.1 20.6 7,923 N 1,064 837 759 661 640 768 762 769 813 853 7,926 * Three households did not report head of household and three households did not report wealth information. Malawi ONSE Impact Evaluation: Baseline Report 37 Household Possessions More than 30 percent of households in both study domains had radios (Table 6). Approximately one-half of households in the project domain and 57.9 percent in the comparison domain reported ownership of a mobile phone. Bicycles were the most common form of transport owned by households in both study domains. However, a larger proportion of households in the project domain (43.3 percent) reported ownership of a bicycle as compared with households in the comparison domain (29.9 percent). Land ownership was more frequently reported by comparison households (90.6 percent) than project households (85.2 percent). Similarly, ownership of livestock was more commonly reported by comparison households (60.0 percent) than project domain households (51.7 percent). Table 6. Household possessions (ONSE IE baseline, 2017) Possession Project Comparison Household effects Torch 84.8 80.3 Mobile phone 50.3 57.9 Radio 30.8 33.4 Bed with mattress 20.6 22.9 Sofa set 7.8 13.3 Television 7.5 9.3 Wrist watch 7.2 6.8 Paraffin lamp (not Koloboyi) 5.1 4.9 Koloboyi 4.1 2.8 Refrigerator 3.7 2.5 Computer 1.6 1.6 Means of transport Bicycle 43.3 29.9 Motorcycle or motor scooter 2.8 2.9 Animal drawn cart 0.6 4.7 Car or truck 0.9 1.4 Boat or motor 0.2 0.1 Ownership of agricultural land Has land 85.2 90.6 38 Malawi ONSE Impact Evaluation: Baseline Report Possession Project Comparison No land 14.8 9.4 Total 100.0 100.0 Ownership of farm animals Has livestock1 51.7 60.0 No livestock 48.3 40.0 Total 100.0 100.0 N 3,962 3,967 1 Livestock includes milk cows, bulls, other cattle, horses, donkeys, mules, goats, sheep, pigs, chicken, and other poultry. Housing Characteristics Electricity was reported by 11.0 percent of project households and 15.4 percent of comparison households (Table 7). In the project domain, 80.4 percent of households had an earth/sand floor, and 19.0 percent had a cement floor. In the comparison domain, 73.0 percent of households had an earth/sand floor, and 26.6 percent had a cement floor. Households in the comparison domain more frequently reported three or more sleeping rooms compared with those in the project domain (35.2 percent and 30.6 percent, respectively), and were also more likely to report a separate building for cooking than those in the project domain (73.7 percent and 54.7 percent, respectively). Project households (41.8 percent) were more likely to cook outdoors than comparison households (23.5 percent). Wood was the main source of cooking fuel in both project (82.1 percent) and comparison (92.1 percent) households. Malawi ONSE Impact Evaluation: Baseline Report 39 Table 7. Housing characteristics (ONSE IE baseline, 2017) Characteristic Project Comparison N Electricity Yes 11.0 15.4 1,108 No 89.0 84.6 6,821 Total 100.0 100.0 7,929 Type of Floor Earth/sand 80.4 73.0 6,046 Cement 19.0 26.6 1,837 Dung 0.4 0.3 31 Other 0.3 0.1 15 Total 100.0 100.0 7,929 Number of rooms for sleeping No sleeping room 0.0 0.1 8 1 sleeping room 29.0 24.9 2,114 2 sleeping rooms 40.3 39.7 3,174 3 or more sleeping rooms 30.6 35.2 2,633 Total 100.0 100.0 7,929 Main cooking location In a separate building 54.7 73.6 5,112 Outdoors 41.7 23.5 2,556 In the house 3.4 2.6 245 Other/no food cooked in household 0.1 0.3 16 Total 100.0 100.0 7,929 Main source of cooking fuel Wood 82.1 92.1 6,836 Charcoal 16.4 7.2 1,012 Straw/shrubs/grass/agricultural crop 0.9 0.1 30 Electricity 0.6 0.5 44 No food cooked in household 0.1 0.1 7 Total 100.0 100.0 7,929 N 3,962 3,967 7,929 40 Malawi ONSE Impact Evaluation: Baseline Report Water and Sanitation Characteristics A slightly higher proportion of comparison households (90.8 percent) reported drinking water from an improved source compared with project households (87.8 percent) (Table 8). The most common improved water source in both domains was a borehole, reported by 65.8 percent of project households and 76.9 percent of comparison households that reported use of an improved source. Just under one-third of project households reported treating their drinking water; of these, 85.6 percent used an appropriate method. Only 22.0 percent of comparison households reported treating their drinking water; of these, 79.9 percent used an appropriate method. Improved toilet facilities were reported by 47.7 percent of project households and 47.1 percent of comparison households. Pit latrines with slabs were the most commonly reported type of improved facility, reported by 75.1 percent of project households and 84.2 percent of comparison households that had improved sanitation. Table 8. Water and sanitation (ONSE IE baseline 2017) Project Comparison N Main source of drinking water Percentage of households with improved drinking water source Improved water source 87.8 90.8 7,023 Non-improved water source 12.2 9.2 906 Total 100.0 100.0 7,929 Improved source Piped into dwelling/yard/neighbor 7.3 5.7 532 Public standpipe 7.3 5.4 451 Protected well / springs 7.2 2.45 352 Tube well/borehole 65.8 76.9 5,667 Protected spring/rainwater/bottled water 0.2 0.3 21 Total improved sources 3,432 3,591 7,023 Unimproved source Unprotected well/springs 8.8 6.1 637 Surface water/rain 3.2 3.1 262 Other 0.2 0.0 7 Total unimproved sources 530 376 906 Malawi ONSE Impact Evaluation: Baseline Report 41 Project Comparison N Household treatment of drinking water Do not treat water 68.6 78.0 5,873 Treat water 31.4 22.0 2,056 Total 100.0 100.0 7,929 Of households that treat drinking water, treatment method used Boil 28.3 33.4 642 Add bleach/chlorine 54.3 44.7 1,007 Strain through a cloth 2.9 1.6 55 Use water filter (ceramic, sand, other filter) 0.2 0.2 4 Let it stand and settle 7.5 5.9 131 Other 7.0 14.2 217 Total 100.0 100.0 2,056 Of households that treat drinking water, percentage using an appropriate method1 Inappropriate water treatment method 14.4 20.1 348 Appropriate water treatment method 85.6 79.9 1,708 Total 100.0 100.0 2,056 Households using appropriate drinking water treatment method among all households1 No treatment or inappropriate water treatment method 73.2 82.4 6,221 Appropriate water treatment method 26.8 17.6 1,708 Total 100.0 100.0 7,929 Toilet facility Improved toilet facilities 47.7 47.1 3,827 Non-improved toilet facilities 52.3 52.9 4,102 Total 100.0 100.0 7,929 Of those using improved toilet facility, type: Flush to piped sewer system 2.0 0.0 36 Flush to septic tank 1.3 0.9 84 Flush to pit latrine 0.1 0.1 9 Flush, don't know where 0.1 0.0 3 Ventilated improved pit latrine 0.3 0.6 34 Pit latrine with slab 75.1 84.2 6,451 42 Malawi ONSE Impact Evaluation: Baseline Report Project Comparison N Composting toilet 0.1 0.0 6 Of those using non-improved toilet facility, type: Pit latrine without slab/open pit 13.5 6.7 739 Hanging toilet/hanging latrine 0.0 0.0 1 No facility/bush/field 7.6 7.4 566 Total 100.0 100.0 7,929 Toilet use For household members only 56.3 50.9 4,284 Other households use toilet of this household 36.7 41.8 3,107 No toilet facility 7.0 7.3 538 Total 100.0 100.0 7929 N 3,962 3,967 7,929 1 Appropriate water treatment methods are boiling, bleaching, filtering, and solar disinfection. Assistance Provided to Households Households were asked whether they were receiving or had received support or assistance from a health project or program (e.g., government, nongovernmental organizations, faith-based organizations) in the past twelve months. Just over one-third of households in both study domains reported receiving health-related support/assistance (Table 9). Of those households that had received assistance, assistance/support related to malaria was most commonly reported (68.0 percent of project households and 74.6 percent of comparison households that received any type of support). WASH services were the next most commonly reported type of assistance received (14.7 percent of project households and 12.1 percent of comparison households that received any type of assistance). Malawi ONSE Impact Evaluation: Baseline Report 43 Table 9. Health-related assistance/support provided to households (ONSE IE baseline, 2017) Project Comparison N Household received assistance in past 12 months Yes 36.9 35.6 2,888 No 63.1 64.4 5,041 Total 100.0 100.0 7,929 Of households that received assistance, type of assistance received Malaria services 68.0 74.6 2,031 WASH services 14.7 12.1 363 Nutrition services 9.4 7.7 266 Maternal and child health services 4.1 3.6 125 FP services 1.6 1.4 61 Other 2.1 0.6 42 Total 100.0 100.0 2,888 N 1,434 1,454 2,888 44 Malawi ONSE Impact Evaluation: Baseline Report CHARACTERISTICS OF WOMEN OF REPRODUCTIVE AGE This section presents the demographic and socioeconomic characteristics of WRA: age group, number of children, education, literacy, and exposure to mass media. It also presents results on women’s satisfaction with health services received for themselves or their child(ren) in the past three months. Characteristics of WRA Tables 10 and 11 provide information on the age, number of children, education, and wealth of households where WRA resided. The age distribution of WRA was similar in the project and comparison domains, with more than 40 percent of women respondents ages 15 to 24 in both study domains (Table 10). Approximately 80 percent of WRA in both domains had living children, with about one-quarter of WRA reporting that they had two to three living children. In the project domain, 42.4 percent of WRA gave birth in the past three years, as did 47.2 percent of WRA in the comparison domain. WRA in the project domain (13.0 percent) were more likely to have no education compared with WRA in the comparison domain (6.4 percent). Attainment of some primary education or completion of primary education was similar across domains, with 71.1 percent of project WRA and 72.0 percent of comparison WRA reporting that they had some/completed primary education. A larger proportion of comparison WRA (20.2 percent) had some/completed secondary education as compared with the project WRA (14.6 percent). The proportion of literate10 women was higher in the comparison domain (77.1 percent) than in the project domain (64.0 percent). In both domains, younger women were more likely to have attended school and have higher educational attainment (Table 11). For example, in the project domain, only 3.7 percent of 15 to 19-year-olds had no education, while 14.4 percent had some/completed secondary education. By contrast, 38.3 percent of 45 to 49-year-olds had no education, whereas only 2.8 percent had some/completed secondary education. In both domains, the proportion of women with higher levels of educational attainment increased with household wealth (Table 11). For example, in the lowest wealth quintile, 19.0 percent of project WRA had no education, 77.0 percent had some/completed primary education, 4.0 percent had some/completed secondary education, and none had more than secondary education. By contrast, in the highest wealth quintile, 3.2 percent of project WRA had no education, 52.7 percent had some/completed primary education, 37.3 percent had some/completed secondary education, and 6.8 percent had more than secondary education. 10 WRA with secondary education were assumed to be literate. Other respondents were given a sentence to read and were considered literate if they could read all or part of the sentence. Malawi ONSE Impact Evaluation: Baseline Report 45 Table 10. Characteristics of WRAs (ONSE IE baseline, 2017) Characteristic Project Comparison N Age groups 15-19 20.9 21.1 1,586 20-24 21.8 20.7 1,605 25-29 15.8 15.1 1,170 30-34 14.4 14.4 1,106 35-39 12.2 12.9 954 40-44 8.3 9.1 631 45-49 6.4 6.8 490 Total 100.0 100.0 7,542 Gave birth in past three years No 57.6 52.8 4,143 Yes 42.4 47.2 3,399 Total 100.0 100.0 7,542 Number of living children No living children 20.6 21.1 1,558 1 child 14.3 15.1 1,136 2-3 children 24.9 26.6 1,946 4-5 children 19.2 20.6 1,532 6 or more children 21.0 16.7 1,370 Highest grade of education completed No formal schooling 13.0 6.4 733 Standard 1-4 31.1 16.0 1,748 Standard 5-8 40.0 56.0 3,620 Secondary 1-2 7.6 12.3 755 Secondary 3-4 7.0 7.9 576 University or above 1.4 1.3 110 Total 100.0 100.0 7,542 Literacy Literate 64.0 77.1 5,327 Illiterate 36.0 22.9 2,215 Total 100.0 100.0 7,542 46 Malawi ONSE Impact Evaluation: Baseline Report Characteristic Project Comparison N Wealth quintile Lowest 23.7 17.7 1,561 Second 20.3 18.0 1,430 Middle 18.3 20.1 1,436 Fourth 17.6 21.3 1,458 Highest 20.0 22.9 1,652 Total 100.0 100.0 7,5421 District Machinga 35.5 -- 1,649 Nkhotakota 27.6 -- 1,143 Salima 36.9 -- 984 Mzimba -- 67.8 2,589 Nsanje -- 12.5 488 Ntchisi -- 19.7 689 Total 100.0 100.0 7,542 N 3,776 3,766 7,542 1 Five women were from households that did not have wealth information Malawi ONSE Impact Evaluation: Baseline Report 47 Table 11. Education of WRA by age, wealth, and district (ONSE IE baseline, 2017) Project Comparison N No education Some/ completed primary education Some/ completed secondary education More than secondary education Total No education Some/ completed primary education Some/ completed secondary education More than secondary education Total Age 15-19 3.7 81.9 14.4 0.0 100.0 0.6 78.6 20.5 0.2 100.0 1,586 20-24 4.7 73.2 20.3 1.8 100.0 2.0 64.0 32.5 1.5 100.0 1,605 25-29 9.6 71.1 16.5 2.7 100.0 4.5 66.7 26.5 2.4 100.0 1,170 30-34 13.6 71.5 13.0 1.9 100.0 6.5 76.2 14.9 2.4 100.0 1,106 35-39 20.0 65.9 13.3 0.8 100.0 11.4 73.3 14.8 0.6 100.0 954 40-44 33.4 56.2 9.1 1.3 100.0 14.7 72.6 10.1 2.6 100.0 631 45-49 38.3 57.2 2.8 1.7 100.0 20.7 75.6 3.1 0.6 100.0 490 Total 13.0 71.1 14.5 1.4 100.0 6.3 72.0 20.2 1.4 100.0 7,542 Wealth quintile Lowest 19.0 77.0 4.0 0.0 100.0 10.6 80.6 8.8 0.0 100.0 1,561 Second 13.3 81.1 5.7 0.0 100.0 7.4 80.7 11.8 0.2 100.0 1,430 Middle 14.3 74.0 11.7 0.0 100.0 7.1 75.5 16.8 0.6 100.0 1,436 Fourth 14.3 69.5 15.8 0.4 100.0 5.7 70.6 23.4 0.3 100.0 1,458 Highest 3.2 52.7 37.3 6.8 100.0 2.2 56.9 35.6 5.2 100.0 1,652 Total* 13.0 71.1 14.5 1.4 100.0 6.3 72.0 20.2 1.4 100.0 7,537 District Machinga 14.7 74.6 9.7 1.0 100.0 -- -- -- -- 100.0 1,649 48 Malawi ONSE Impact Evaluation: Baseline Report Project Comparison N No education Some/ completed primary education Some/ completed secondary education More than secondary education Total No education Some/ completed primary education Some/ completed secondary education More than secondary education Total Nkhotakota 6.5 71.1 21.0 1.5 100.0 -- -- -- -- 100.0 1,143 Salima 16.2 67.7 14.3 1.8 100.0 -- -- -- -- 100.0 984 Mzimba -- -- -- 0.0 100.0 2.8 73.0 22.7 1.5 100.0 2,589 Nsanje -- -- -- 0.0 100.0 20.8 64.7 13.9 0.6 100.0 488 Ntchisi -- -- -- 0.0 100.0 9.4 73.3 15.7 1.7 100.0 689 Total 13.0 71.1 14.5 1.4 100.0 6.3 72.0 20.2 1.4 100.0 7,542 N 480 2,684 556 56 3,776 246 2,684 775 61 3,766 7,542 *Five WRA were from households that did not report wealth information. Malawi ONSE Impact Evaluation: Baseline Report 49 Exposure to Mass Media The most common form of media exposure among WRA was radio. In the project domain, 43.4 percent of WRA reported listening to the radio at least once per week, as did 41.5 percent of WRA in the comparison domain (Table 12). About one-half of WRA in both domains (45.1 percent of project WRA and 50.1 percent of comparison WRA) reported that they did not access newspaper, television, or radio on a weekly basis. The proportion of women listening to the radio at least once per week was constant across age groups in both domains. For all three forms of media (newspaper, radio, and television), the proportion of women accessing each media at least once per week increased with educational attainment in both domains. For example, in the project domain, only 28.9 percent of the women with no education listened to the radio at least once a week, whereas 41.2 percent of those with some/completed primary education, 63.4 percent with some/completed secondary education, and 78.0 percent of those with more than secondary education did so. Similarly, listening to the radio at least once per week increased with household wealth. For example, in the project domain, only 22.0 percent of the women in the lowest wealth quintile listened to the radio at least once a week, whereas 43.4 percent of those in the middle quintile, and 75.9 percent of those in the highest quintile did so. 50 Malawi ONSE Impact Evaluation: Baseline Report Table 12. Percentage of WRA who were exposed to specific media on a weekly basis (ONSE IE baseline, 2017) Project Comparison N Read newspaper at least once/ week1 Watched television at least once/ week Listened to the radio at least once/ week Accessed all three media at least once/ week Accessed none of the three media once/ week Read newspaper at least once/ week1 Watched television at least once/ week Listened to radio at least once/ week Accessed all three media at least once/ week Accessed none of the three media once/ week Age 15–19 12.3 10.8 41.7 2.6 45.5 12.5 10.8 40.1 1.7 48.9 1,586 20–24 13.4 10.4 46.3 3.9 44.9 9.1 9.4 43.9 2.4 49.9 1,605 25–29 12.6 10.9 45.7 3.5 41.8 9.0 8.5 41.2 2.1 51.7 1,170 30–34 9.4 8.1 42.9 2.7 46.3 7.3 9.2 40.3 3.0 53.3 1,106 35–39 10.2 9.0 43.8 1.8 45.8 8.7 10.9 42.2 2.9 48 954 40–44 9.4 7.7 41.9 1.3 47.8 8.2 10.8 41.8 3.4 49.4 631 45–49 9.0 6.2 35.1 2.3 46 6.3 12.1 39.6 1.1 49.6 490 Educational attainment No education 0.0 2.7 28.9 0.0 74.3 7.4 3.0 29.9 0.0 53.0 726 Some/ completed primary 8.6 5.8 41.2 1.1 51.1 6.2 7.3 37.9 1.0 55.5 5,368 Some/ completed secondary 18.0 29.0 63.4 7.5 28.0 16.4 18.8 55.6 5.2 38.0 1,331 More than secondary 51.6 64.0 78.0 36.9 11.1 37.2 58.0 75.2 25.8 13.4 117 Wealth quintile* Lowest 4.8 1.9 22.0 0.3 69.6 5.7 2.0 22.8 0.2 70.7 1,561 Second 8.2 2.6 32.2 0.1 61.2 6.5 2.5 27.6 0.5 67.3 1,430 Malawi ONSE Impact Evaluation: Baseline Report 51 Project Comparison N Read newspaper at least once/ week1 Watched television at least once/ week Listened to the radio at least once/ week Accessed all three media at least once/ week Accessed none of the three media once/ week Read newspaper at least once/ week1 Watched television at least once/ week Listened to radio at least once/ week Accessed all three media at least once/ week Accessed none of the three media once/ week Middle 7.1 3.7 43.4 0.5 52.6 7.2 3.5 33.7 0.5 60.1 1,436 Fourth 11.2 6.9 48.1 1.9 43.4 8.3 4.2 47 0.9 47.8 1,458 Highest 21.7 33.4 75.9 9.6 14.4 15.5 33.3 68.5 7.6 23.6 1,652 District Machinga 10.6 7.1 41.9 2.5 48.1 -- -- -- -- -- 1,649 Nkhotakota 10.7 14.1 50.6 3.4 38.2 -- -- -- -- -- 1,143 Salima 13.4 8.4 39.3 2.7 48.4 -- -- -- -- -- 984 Mzimba -- -- -- -- -- 8.3 11.7 42 2.2 49.7 2,589 Nsanje -- -- -- -- -- 12.5 7.3 40 2.6 48.8 488 Ntchisi -- -- -- -- -- 12.0 6.0 40.6 2.9 52.6 689 Total 11.6 9.5 43.4 2.8 45.1 9.4 10.1 41.5 2.4 50.1 7,542 N 2,418 3,772 2,906 3,765 7,542 1Only asked of literate women. *Five WRA were from households that did not report wealth information 52 Malawi ONSE Impact Evaluation: Baseline Report Exposure to Moyo ndi mpamba, Usamalireni! Moyo ndi mpamba: Usamalireni! (“Life is precious/capital: take care of it!”) is a SBCC campaign launched in 2013, implemented by the Support for Service Delivery Integration (SSDI)-Communication project, which connects wellness to prosperity. ONSE will be conducting community-level activities in support of this campaign and messaging. WRA were asked whether they had heard the slogan Moyo ndi Mpamba: Usamalireni! in the past 12 months. In the project domain, 88.0 percent of WRA recalled hearing the slogan as compared with only 75.7 percent of WRA in the comparison domain (Table 13). Table 13. Exposure to Moyo ndi mpamba, Usamalireni! (ONSE IE baseline, 2017) Heard the slogan Moyo ndi mpamba, Usamalireni! in the past 12 months Project Comparison N Yes 88.0 75.7 6,200 No 12.0 23.9 1,323 Don’t know 0.1 0.4 19 Total 100.0 100.0 7,542 N 3,776 3,766 7,542 Satisfaction with Health Services Women were asked a series of questions about their satisfaction with health services received for themselves or their children in the past three months. Just under two-thirds of women reported that they had visited a health facility in the past three months in both the project and comparison domains (Table 14). Among those who had visited a facility, approximately 60 percent in both domains were very satisfied with the time they waited to see a healthcare provider, whereas about 12 percent were very dissatisfied with their wait time. Nearly three-quarters of respondents in both domains were very satisfied with their ability to discuss their problem with the healthcare provider and with the explanation they received about their problem or treatment. Only about 4 percent in both domains were very dissatisfied in this regard. Nearly 80 percent of the women in both domains were also very satisfied with the audio and visual privacy at the health facility, whereas less than 4 percent were very dissatisfied. In both domains, nearly 70 percent of the women were very satisfied with the availability of medicines, approximately 60 percent were very satisfied with service hours, and more than 75 percent were very satisfied with facility cleanliness. Approximately 72 percent were very satisfied with their overall treatment by staff. Malawi ONSE Impact Evaluation: Baseline Report 53 Women in the project domain (36.1 percent) more frequently reported paying a fee for service than women in the comparison domain (27.4 percent). Among those who paid a fee, women in the project domain (65.7 percent) more frequently reported that they were very satisfied with the cost of services than women in the comparison domain (59.3 percent). Women in the comparison domain (12.4 percent) were more likely to report that they were very dissatisfied with the cost of services than women in the project domain (8.4 percent). Table 14. Satisfaction with health services among WRA (ONSE IE baseline, 2017) Project Comparison N In the past 3 months, visited a health facility for care for self or child Yes 64.7 62.2 4,740 Among those who visited a facility: Extent satisfied with the time waited to see a provider Very satisfied 58.4 63.1 2,904 Somewhat satisfied 16.1 14.3 701 Neither 1.3 0.9 51 Somewhat dissatisfied 10.6 10.1 477 Very dissatisfied 13.6 11.7 607 Total 100.0 100.0 4,740 Extent satisfied with the ability to discuss problem with provider Very satisfied 73.4 75.2 3,532 Somewhat satisfied 16.1 15.6 723 Neither 1.3 0.8 53 Somewhat dissatisfied 5.1 4.5 224 Very dissatisfied 4.2 3.9 208 Total 100.0 100.0 4,740 Extent satisfied with explanation received about problem or treatment Very satisfied 74.5 74.1 3,521 Somewhat satisfied 15.6 15.4 724 Neither 0.8 0.8 42 Somewhat dissatisfied 4.7 5.4 229 Very dissatisfied 4.4 4.3 224 Total 100.0 100.0 4,740 54 Malawi ONSE Impact Evaluation: Baseline Report Project Comparison N Extent satisfied with the audio and visual privacy Very satisfied 77.5 80.6 3,766 Somewhat satisfied 14.0 12.8 609 Neither 1.6 0.7 57 Somewhat dissatisfied 3.3 3.1 152 Very dissatisfied 3.6 2.8 156 Total 100.0 100.0 4,740 Extent satisfied with the availability of medicines Very satisfied 68.0 70.2 3,283 Somewhat satisfied 15.1 12.3 642 Neither 1.4 1.1 61 Somewhat dissatisfied 7.9 8.0 364 Very dissatisfied 7.7 8.4 390 Total 100.0 100.0 4,740 Extent satisfied with the facility service hours Very satisfied 60.3 62.9 2,966 Somewhat satisfied 19.2 18.9 873 Neither 1.6 1.3 67 Somewhat dissatisfied 9.6 8.4 413 Very dissatisfied 9.4 8.6 421 Total 100.0 100.0 4,740 Extent satisfied with the facility cleanliness Very satisfied 78.2 77.0 3,696 Somewhat satisfied 12.3 13.4 595 Neither 1.2 0.5 43 Somewhat dissatisfied 4.1 5.0 213 Very dissatisfied 4.2 4.2 193 Total 100.0 100.0 4,740 Extent satisfied with the overall staff treatment Very satisfied 71.4 73.1 3,431 Somewhat satisfied 18.2 17.1 817 Malawi ONSE Impact Evaluation: Baseline Report 55 Project Comparison N Neither 1.1 1.0 60 Somewhat dissatisfied 5.2 4,7 230 Very dissatisfied 4.1 4.1 202 Total 100.0 100.0 4,740 Paid a fee Service was free 63.9 72.6 3,225 Paid fee 36.1 27.4 1,515 Total 100.0 100.0 4,740 Among those who paid a fee, extent satisfied with any cost associated with treatment Very satisfied 65.7 59.3 977 Somewhat satisfied 14.5 12.8 197 Neither 3.4 4.8 62 Somewhat dissatisfied 8.0 10.8 130 Very dissatisfied 8.4 12.4 149 Total 100.0 100.0 1,515 N 2,427 2,313 4,740 56 Malawi ONSE Impact Evaluation: Baseline Report FAMILY PLANNING This section presents information on current use of FP methods and beliefs about FP. Current Use of Family Planning Among all WRA, 48.2 percent reported current use of any method of FP in both the project and comparison domains (Table 15). Among married WRA, just under 58 percent in both domains reported using any method of FP. Approximately 46 percent of all WRA and 55 percent of married WRA reported using a modern method. The most commonly reported method was injectables. A larger percentage of women in the project domain reported using injectables than in the comparison domain. Injectable use was reported by 49.2 percent of all WRA and 51.4 percent of all married WRA in the project domain, and by 40.2 percent of all WRA and 41.1 percent of all married WRA in the comparison domain. Implants were the second most commonly reported method in both domains and were more popular among women in the comparison domain (25.5 percent of all WRA and 26.0 percent of married WRA in the comparison domain as compared with 18.3 percent of all WRA and 18.6 percent of married WRA in the project domain, respectively). Table 15. Current use of contraception among WRA (ONSE IE baseline, 2017) Project Comparison N Project Comparison N FP method All WRA1 Married WRA2 Any method 48.2 48.2 7,158 57.7 57.9 5,240 Any modern method 45.6 45.8 7,158 55.5 53.7 5,240 Injectables 49.2 40.2 1,537 51.4 41.1 1,391 Implant 18.3 25.5 786 18.6 26.0 699 Female sterilization 13.7 13.4 453 15.1 13.9 421 Male condom 9.8 8.1 297 5.2 5.6 166 Pill 2.5 5.2 131 2.7 5.6 128 Other traditional method 1.7 0.7 39 1.9 0.8 38 Intrauterine device (IUD) 1.5 1.9 61 1.6 2.1 56 Lactational amenorrhea method 1.2 0.8 31 1.3 0.8 28 Withdrawal 1.0 3.1 76 1.2 3.2 73 Rhythm method 0.6 0.8 25 0.5 0.9 21 Male sterilization 0.2 0.0 1 0.2 - 1 Standard days method 0.2 0.2 9 0.2 0.2 8 Malawi ONSE Impact Evaluation: Baseline Report 57 Project Comparison N Project Comparison N FP method All WRA1 Married WRA2 Female condom 0.1 0.0 1 0.0 - - Other modern method 0.1 0.0 2 0.1 - 1 Emergency contraception 0.0 0.1 1 0.0 0.1 1 Total 100.0 100.0 3,450 100.0 100.0 3,032 N 3,582 3,576 7,158 2,522 2,718 5,240 1 All women with potential need for contraception, i.e., married or sexually active unmarried women. 2 Includes women who are married or living with a man. Table 16 presents results related to contraceptive use among married or sexually active women who were not pregnant and who did not desire more children. Contraceptive use among these women was reported by 61.1 percent in the project domain and 58.3 percent in the comparison domain. Contraceptive use by married or sexually active women who were not pregnant and who did not desire more children was reported by a higher proportion of women with no education in the comparison domain than in the project domain (61.5 percent and 51.6 percent, respectively). There was little difference in the proportion of married or sexually active women who were not pregnant and who did not desire more children using contraception with regard to wealth quintile in the comparison domain. In the project domain, only 56.1 percent of the women in the lowest quintile reported contraceptive use, whereas 67.7 percent of the women in the highest quintile reported the same. Table 16. Contraceptive use among married or sexually active women who were not pregnant and who did not desire more children (ONSE IE baseline, 2017) Characteristics Project Comparison Contraceptive use N Contraceptive use N Age groups 15-19 23.0 21 19.7 24 20-24 72.4 73 62.7 72 25-29 72.5 146 67.7 130 30-34 68.5 221 67.0 234 35-39 66.1 237 61.3 281 40-44 54.2 149 50.4 187 45-49 33.2 121 44.6 150 Total 61.1 968 58.3 1,078 58 Malawi ONSE Impact Evaluation: Baseline Report Characteristics Project Comparison Contraceptive use N Contraceptive use N Educational attainment No education 51.6 213 61.5 114 Some/completed primary 63.2 638 57.3 805 Some/completed secondary 68.6 105 60.2 144 More than secondary 60.6 12 73.7 15 Total 61.1 968 58.3 1,078 Wealth quintile Lowest 56.1 219 60.7 155 Second 61.6 187 59.1 199 Middle 63.0 198 55.9 210 Fourth 56.7 167 57.8 246 Highest 67.7 196 58.8 268 Total 61.1 967 58.3 1,078 District Machinga 67.3 417 -- -- Nkhotakota 56.9 268 -- -- Salima 58.4 283 -- -- Mzimba -- -- 55.0 797 Nsanje -- -- 68.9 99 Ntchisi -- -- 66.5 182 Total 61.1 968 58.3 1,078 Table 17 presents results related to FP counseling among current users of modern methods ages 15 to 49 who started the last episode of contraceptive use in the three years before the survey (excluding sterilized women). The provision of information about side effects and other methods was similar in project and comparison domains, with about 60 percent of women receiving information about side effects and 70 percent of women receiving information about other methods. A higher percentage of middle-aged and more educated women received information than did women in other groups. Malawi ONSE Impact Evaluation: Baseline Report 59 Table 17. Informed choice among current users of modern methods ages 15 to 49 who started the last episode of contraceptive use in the three years before the survey (ONSE IE baseline, 2017) Project Comparison Possible side effects/ problem s What to do if side effects/ problem s occurre d Other contraceptiv e methods Possible side effects/ problem s What to do if side effects/ problem s occurre d Other contraceptiv e methods Age group 15–19 39.3 34.6 46.4 48.7 45.2 52.9 20–24 55.8 55.6 68.0 64.0 61.7 72.3 25–29 62.5 58.8 73.3 59.7 63.5 71.7 30–34 68.0 67.5 77.8 63.9 64.3 78.8 35–39 69.3 65.4 78.9 62.7 60.6 75.4 40–44 62.4 64.4 67.6 53.3 51.0 71.9 45–49 56.1 50.1 62.1 59.2 60.8 79.1 Total 67.2 63.0 66.6 57.7 58.4 81.0 Educational attainment No education 67.2 63.0 66.6 57.7 58.4 81.0 Some/completed primary 60.8 58.5 71.2 61.8 61.5 71.0 Some/completed secondary 43.5 44.0 62.1 57.6 56.6 72.1 More than secondary 65.1 67.7 60.6 55.9 52.4 69.6 Total 67.2 63.0 66.6 57.7 58.4 81.0 Wealth quintile Lowest 62.2 61.8 72.2 60.1 58.6 70.9 Second 60.7 56.2 70.8 64.9 62.2 74.2 Middle 60.0 61.7 71.5 62.3 63.9 76.8 Fourth 49.4 47.4 58.8 58.7 60.1 70.1 Highest 60.9 55.2 70.2 56.8 55.6 65.8 Total 67.2 63.0 66.6 57.7 58.4 81.0 60 Malawi ONSE Impact Evaluation: Baseline Report Project Comparison Possible side effects/ problem s What to do if side effects/ problem s occurre d Other contraceptiv e methods Possible side effects/ problem s What to do if side effects/ problem s occurre d Other contraceptiv e methods District Machinga 66.5 67.0 76.0 -- -- Nkhotakota 56.6 53.0 66.0 -- -- Salima 52.0 48.0 63.3 -- -- Mzimba -- -- 63.9 63.8 71.6 Nsanje -- -- 65.9 63.5 77.8 Ntchisi -- -- 47.2 47.3 68.2 Total 59.1 57.1 69.2 60.6 60.2 71.8 N 1,162 1,162 1,162 1,165 1,165 1,165 Malawi ONSE Impact Evaluation: Baseline Report 61 Beliefs about Family Planning WRA were asked to indicate their agreement with eight statements about FP drawn from the National Health Communication Strategy, 2015‒2020 and the Moyo ndi mpamba: Usamalireni! media campaign. Table 18 reports the percentage of WRA who believed that each statement was “completely true.” For all eights statements, a somewhat higher percentage of women in the comparison domain than in the project domain reported that the statements were “completely true.” In both domains, approximately 90 percent or more of women completely agreed with the statement, “There are family planning methods available at the clinic for everybody.” Approximately 80 percent or more of the women in both domains completely agreed with six additional statements: • “Planning the family is the responsibility of both men and women.” • “Getting pregnant before you are 18 puts your health and that of the baby in danger.” • “Long-acting family planning methods help to conveniently space pregnancies.” • “Family planning should be used by husbands and wives for the health of the entire family.” • “Becoming pregnant after 40 years of age can be dangerous to your health.” • “Talking openly and honestly to your children about the consequences of unprotected sex is important.” Approximately 70 percent or more of women believed that the statement “Long-term and permanent family planning methods provide safe and healthy ways to temporarily or permanently stop having children” was completely true. The remaining statement, “Family planning methods are safe,” was reported to be completely true by only 52.9 percent of project women and 54.7 percent of comparison women. In both domains, younger women, ages 15 to 19, were less likely to believe the statements were completely true than women in other age groups. In both domains, women with more than secondary education (as compared with women with less education) were more likely to believe each of the statements was completely true, except for the statement, “Family planning methods are safe.” Women with more than secondary education were the least likely to believe that this statement was completely true. In the comparison domain, women with more than secondary education were also less likely to report that they believed the statement, “Long-term and permanent family planning methods provide safe and healthy ways to temporarily or permanently stop having children” was completely true. g by in danger 62 Malawi ONSE Impact Evaluation: Baseline Report Table 18. Beliefs about FP, percentage of WRA who believed statements were completely true (ONSE IE baseline, 2017) FP methods are safe Planning the family is the responsibility of both men and women Getting pregnant before you are 18 puts your health and that of the baby in danger Long-acting FP methods help to conveniently space pregnancy FP should be used by husbands and wives for the health of the entire family Long-term and permanent FP methods provide safe and healthy ways to temporarily or permanently stop having children There are FP methods available at the clinic for everybody Becoming pregnant after age 40 can be dangerous to your health Talking openly and honestly to your children about the consequences of unprotected sex is important Project Age 15–19 43.1 81.1 84.7 72.1 82.6 58.1 83.5 69.7 83.4 20–24 53.9 92.4 91.3 89.0 93.6 72.1 95.9 81.6 88.9 25–29 54.7 92.3 91.1 89.8 94.6 75.4 96.3 90.5 91.1 30–34 63.8 94.0 90.8 91.8 95.9 80.0 97.1 91.1 93.6 35–39 59.5 92.6 91.5 90.1 95.8 79.4 97.3 91.0 93.7 40–44 58.7 88.5 94.1 89.8 95.3 80.7 94.4 93.0 93.9 45–49 60.1 94.9 94.7 88.6 96.2 77.1 94.9 93.3 95.1 Education No education 59.9 90.1 90.5 89.0 92.5 79.6 94.4 88.7 90.8 Some/completed primary 55.6 89.0 89.3 85.0 91.6 71.3 93.0 82.7 88.9 Some/completed secondary 47.1 94.4 94.3 89.0 95.5 72.7 94.7 89.5 95.2 Malawi ONSE Impact Evaluation: Baseline Report 63 FP methods are safe Planning the family is the responsibility of both men and women Getting pregnant before you are 18 puts your health and that of the baby in danger Long-acting FP methods help to conveniently space pregnancy FP should be used by husbands and wives for the health of the entire family Long-term and permanent FP methods provide safe and healthy ways to temporarily or permanently stop having children There are FP methods available at the clinic for everybody Becoming pregnant after age 40 can be dangerous to your health Talking openly and honestly to your children about the consequences of unprotected sex is important More than secondary 39.1 100.0 98.0 92.0 97.8 83.1 97.2 100.0 96.1 District Machinga 59.5 93.1 92.9 89.7 95.8 79.1 96.7 89.2 92.2 Nkhotakota 52.5 89.0 90.1 83.8 91.2 68.7 91.7 84.3 90.6 Salima 51.7 88.0 87.9 84.5 89.9 69.6 91.7 80.8 87.9 Total 52.9 84.5 85.1 79.9 87.3 68.3 89.1 80.3 86.3 Comparison Age 15–19 38.5 67.8 76.1 59.3 71.7 46.7 70.3 60.9 77.0 20–24 56.1 88.8 84.4 82.4 89.3 70.4 93.0 83.4 87.5 25–29 56.9 91.9 90.0 90.0 94.8 76.9 96.5 84.0 89.7 30-34 53.2 90.2 88.0 87.1 93.9 75.1 94.8 86.2 90.9 35-39 57.9 86.9 88.4 83.2 90.1 73.9 91.9 85.7 89.5 40-44 61.4 87.8 89.1 86.8 90.4 75.4 94.8 88.5 88.4 45-49 57.6 86.0 86.9 83.3 89.8 75.2 94.3 89.0 85.8 64 Malawi ONSE Impact Evaluation: Baseline Report FP methods are safe Planning the family is the responsibility of both men and women Getting pregnant before you are 18 puts your health and that of the baby in danger Long-acting FP methods help to conveniently space pregnancy FP should be used by husbands and wives for the health of the entire family Long-term and permanent FP methods provide safe and healthy ways to temporarily or permanently stop having children There are FP methods available at the clinic for everybody Becoming pregnant after age 40 can be dangerous to your health Talking openly and honestly to your children about the consequences of unprotected sex is important Education No education 63.4 84.5 85.3 85.6 88.3 76.5 93.5 84.1 87.6 Some/completed primary 52.8 82.5 83.5 78.6 85.7 68.4 87.6 78.5 84.1 Some/completed secondary 50.7 90.7 89.7 82.3 91.8 65.4 92.3 84.5 93.0 More than secondary 43.3 100.0 98.6 88.4 99.1 66.3 96.6 91.4 98.6 District Mzimba 53.1 85.3 85.1 80.3 87.7 68.5 89.0 81.1 85.7 Nsanje 50.7 83.8 83.3 78.5 86.0 64.7 86.6 77.1 84.9 Ntchisi 53.3 82.3 86.2 79.6 86.9 69.8 90.9 79.3 89.2 Total 54.7 90.1 90.3 86.2 92.4 72.8 93.5 84.7 90.2 Malawi ONSE Impact Evaluation: Baseline Report 65 MATERNAL HEALTH Reducing maternal mortality and morbidity are central goals of the ONSE project. This section presents information on ANC and delivery care, and women’s PNC for the 3,399 women who had live births in the three years preceding the survey. Antenatal Care Nearly all women (n=3,361) who had a birth in the past three years received ANC for their last birth. ANC was provided by a skilled provider (i.e., doctor, clinical officer, medical assistant, nurse, or midwife) for 98.6 percent of the women in the project domain and 96.9 percent of the women in the comparison domain (Table 19). In the project domain, 52.6 percent and 26.4 percent of the women who had received ANC for their most recent birth in the past three years reported that they received care at a government health center or hospital, respectively. In the comparison domain, 57.0 percent and 20.4 percent of the women reported the same, respectively. In both domains, approximately 48 percent of women who had received ANC for their most recent birth in the past three years reported that their first ANC visit occurred in their fourth or fifth month of pregnancy. Women in the comparison domain (35.2 percent) were somewhat more likely to have had their first ANC visit in their first three months of pregnancy than women in the project domain (30.8 percent). A larger proportion of the women in the comparison domain (55.8 percent) reported four or more ANC visits than women in the project domain (51.8 percent). The most frequently reported component of ANC received was having a blood sample taken, reported by 95.1 percent of women in the project domain and 94.4 percent of women in the comparison domain. Having blood pressure taken and being given iron tablets or syrup was reported by more than 80 percent of women in both domains. The least commonly received component of ANC in both domains was having a urine sample taken, reported by approximately one-third of women who had ANC for their most recent birth in the past three years. The World Health Organization (WHO) recommends intermittent preventive treatment of malaria in pregnancy (IPTp) at least three times during pregnancy for women in malaria endemic areas (WHO, 2012). Most women received treatment with sulfadoxine- pyrimethamine (SP)/Fansidar during their last pregnancy (87.5 percent of women in the project domain and 88.8 percent of women in the comparison domain) (Table 20). This treatment was provided almost universally during ANC visits. However, only 32.4 percent of women in the project domain and 33.5 percent in the comparison domain received the recommended minimum three doses of IPTp during their last pregnancy. 66 Malawi ONSE Impact Evaluation: Baseline Report Table 19. Women’s ANC in the past three years (ONSE IE baseline, 2017) Received ANC at last birth Project Comparison N Yes 98.7 99.2 3,361 First ANC visit in the first trimester 30.8 35.2 1,119 Received 4+ ANC visits 51.8 55.8 1,796 ANC from skilled provider 98.6 96.9 3,287 No 1.3 0.8 38 Total 100.0 100.0 3,399 Care provider among those who received ANC at last birth Skilled provider Doctor/clinical officer/medical assistant 22.7 22.9 765 Nurse/midwife 75.9 74.0 2,522 Other provider Patient attendant 0.3 0.6 15 Health surveillance assistant 1.0 2.4 56 Traditional birth attendant 0.1 0.0 2 Other 0.0 0.1 1 Total 100.0 100.0 3,361 Location of ANC Her home/other home 0.1 0.0 2 Government hospital 26.4 20.4 792 Government health center 52.6 57.0 1,838 Government dispensary 4.8 2.2 119 Government health post 1.1 2.2 55 Public mobile clinic 1.2 2.5 62 Other public sector 0.8 0.5 23 CHAM hospital 2.5 6.6 149 CHAM health center 6.1 4.9 186 CHAM other facility 0.3 0.1 7 Private facility 2.4 1.3 64 Banja la Mtsogolo health center/other 1.6 2.2 64 Total 100.0 100.0 3,361 Malawi ONSE Impact Evaluation: Baseline Report 67 Received ANC at last birth Project Comparison N Timing of first ANC visit among those who received ANC at last birth > 4 months 30.8 35.2 1,119 4-5 months 47.8 48.8 1,618 6-7 months 20.2 15.1 518 8+ months 1.3 1.0 43 Total 100.0 100.0 3,361 Components of ANC received among those who received ANC at last birth Blood sample taken 95.1 94.4 3,185 Iron tablets or iron syrup 82.3 89.4 2,878 Blood pressure taken 81.1 82.8 2,753 Took drug for intestinal worms 50.2 48.1 1,654 Urine sample taken 31.0 33.0 1,073 Total 100.0 100.0 3,361 N 1,801 1,598 3,399 Table 20. Use of SP/Fansidar during last pregnancy in the past three years (ONSE IE baseline, 2017) Use of SP/Fansidar during last pregnancy in past three years Project Comparison N Yes 87.5 88.8 2,948 No 12.5 11.2 395 Total 100.0 100.0 3,343 Number of times SP/Fansidar used among users Once 31.2 30.9 905 Twice 36.5 35.6 1,073 Three or more times 32.4 33.5 970 Total 100.0 100.0 2,948 Source of SP/Fansidar during last pregnancy ANC visit 99.1 99.7 2,930 Another facility visit 0.8 0.3 16 Other source 0.0 0.0 2 Total 100.0 100.0 2,948 68 Malawi ONSE Impact Evaluation: Baseline Report Assistance during Delivery Nearly all women (n=3,361) who had a birth in the past three years gave birth at a health facility (approximately 95 percent) for their most recent birth and were assisted by a skilled provider (approximately 94 percent) (Table 21). Approximately 70 percent of women in both the project and comparison domains were assisted by a nurse or midwife, and approximately one-quarter were assisted by a doctor, clinical officer, or medical assistant. Approximately 80 percent of women in both domains who gave birth in the past three years delivered their last child at a government health center or hospital. Table 21. Women’s delivery care in the past three years (ONSE IE baseline, 2017) Labor and delivery at last birth Project Comparison N Skilled birth attendant1 93.5 94.7 3,199 Doctor/clinical officer/medical assistant 23.5 26.6 808 Nurse/midwife 70.0 68.0 2,391 Other birth attendant Patient attendant 0.4 0.9 22 Traditional birth attendant 1.7 0.6 42 Relative/friend 2.9 2.2 86 Other 0.3 0.5 16 No one assisted 1.1 1.2 34 Total 100.0 100.0 3,399 Labor and delivery at last birth Project Comparison N Facility birth at last birth 94.6 95.8 3,239 Government hospital 39.0 33.9 1,169 Government health center 38.9 46.1 1,519 Government dispensary 2.6 1.7 68 Government health post/outreach 0.3 0.0 6 Other public sector 1.1 0.3 20 CHAM hospital 2.2 6.6 164 CHAM health center 5.3 3.8 157 CHAM other 0.2 0.4 9 Private hospital/clinic 1.3 0.6 33 Other private medical sector 0.5 0.2 9 Malawi ONSE Impact Evaluation: Baseline Report 69 Labor and delivery at last birth Project Comparison N Banja la Mtsogolo health center 3.2 2.1 85 Other location at last birth Her home 4.3 3.4 126 Other home 1.1 0.8 34 Total 100.0 100.0 3,399 N 1,801 1,598 3,399 1Includes women who received a check from a doctor, clinical officer, medical assistant, nurse, or midwife. Postnatal Care Approximately 67 percent of WRA who had a birth in the past two years at a health facility received PNC for their most recent birth (Table 22). Nearly all those who gave birth at a facility and received PNC got the PNC from a skilled provider—over 26 percent in both domains received care from a doctor/clinical officer/medical assistant, and approximately 70 percent received the care from a nurse/midwife. More than 85 percent of newborns who were delivered at a facility received PNC (88.2 percent in project and 85.0 percent in comparison domains) (Table 23). Newborn PNC was delivered by skilled providers almost universally in both the project and comparison domains. Nurses/midwives provided PNC to approximately three-fourths of newborns delivered at a facility, whereas doctors/clinical officers/medical assistants provided PNC to approximately one-fourth of newborns delivered at a facility. Table 24 presents information on the 177 women who gave birth in the past two years who did not deliver at a heath facility for their last birth. In the project domain, 48.7 percent of these women received PNC as compared with only 40.4 percent of women in the comparison domain. Among women who received PNC, 45.4 percent of women in the project domain and 40.4 percent in the comparison domain received PNC within two days. Among those who did not deliver at a facility, PNC was received from a skilled provider by 80.2 percent of women in the project domain and 93.2 percent of women in the comparison domain. Among the few newborns not delivered at a facility, only 41.4 percent of newborns in the project domain, and 30.0 percent in the comparison domain received PNC (Table 25). PNC was provided by skilled providers for 79.3 percent of newborns in the project domain and 92.6 percent in the comparison domains. 70 Malawi ONSE Impact Evaluation: Baseline Report Table 22. Women’s PNC in the past two years among those who delivered at a health facility (ONSE IE baseline, 2017) Project Comparison N Received PNC Received PNC 67.2 67.7 1,542 Did not receive PNC 32.8 32.3 761 Total 100.0 100.0 2,303 Time to PNC Within two days 62.5 60.3 1,396 3-41 days 4.7 7.3 146 Never/did not receive PNC 32.8 32.4 761 Total 100.0 100.0 2,303 PNC care provider Skilled provider1 97.6 96.8 1,501 Doctor/clinical officer/medical assistant 29.0 26.6 401 Nurse/midwife 68.6 70.2 1,100 Other provider 2.4 3.2 22 Patient attendant/health surveillance assistant 0.0 0.1 2 Traditional birth attendant/other 0.8 2.2 20 Don't know 1.6 0.9 19 Total 100.0 100.0 1,542 N 1,228 1,075 2,303 1Includes women who received a check from a doctor, clinical officer, medical assistant, nurse, or midwife. Malawi ONSE Impact Evaluation: Baseline Report 71 Table 23. Table 23. Newborn PNC in the past two years among those who were delivered at a health facility (ONSE IE baseline, 2017) Project Comparison N Received PNC Received PNC 88.2 85.0 1,957 Did not receive PNC 11.8 15.0 294 Total 100.0 100.0 2,251 Time to PNC Within two days 84.0 79.1 1,853 3-41 days 4.1 5.9 104 Never/did not receive PNC 11.8 15.0 294 Total 100.0 100.0 2,251 PNC care provider Skilled provider1 99.3 99.4 1,945 Doctor/clinical officer/medical assistant 25.0 25.5 466 Nurse/midwife 74.3 73.9 1,479 Other provider 0.7 0.7 12 Patient attendant/health surveillance assistant 0.3 0.2 5 Traditional birth attendant/other 0.4 0.5 7 Total 100.0 100.0 1,957 N 1,196 1,055 2,251 1Includes women who received a check from a doctor, clinical officer, medical assistant, nurse, or midwife. 72 Malawi ONSE Impact Evaluation: Baseline Report Table 24. Women’s PNC in the past two years among those who did not deliver at a health facility (ONSE IE baseline, 2017) Project Comparison N Received PNC Received PNC 48.7 40.4 73 Did not receive PNC 51.3 59.6 104 Total 100.0 100.0 177 Time to PNC Within two days 45.5 40.4 69 3-41 days 3.2 0.0 4 Never/did not receive PNC 51.3 59.6 104 Total 100.0 100.0 177 PNC care provider Skilled provider1 80.2 93.2 62 Doctor/clinical officer/medical assistant 14.2 12.7 14 Nurse/midwife 66.0 80.5 48 Other provider 19.8 6.9 11 Patient attendant/health surveillance assistant 1.4 0.0 1 Traditional birth attendant/other 18.4 6.9 10 Total 100.0 100.0 73 N 105 72 177 1Includes women who received a check from a doctor, clinical officer, medical assistant, nurse, or midwife. Malawi ONSE Impact Evaluation: Baseline Report 73 Table 25. Newborn PNC in the past two years among those who were not delivered at a health facility (ONSE IE baseline, 2017) Project Comparison N Received PNC Received PNC 41.4 30.0 76 Did not receive PNC 58.6 70.0 155 Total 100.0 100.0 231 Time to PNC Within two days 37.3 30.0 69 3-41 days 4.1 0.0 7 Never/did not receive PNC 58.6 70.0 155 Total 100.0 100.0 231 PNC care provider Skilled provider1 79.3 92.6 63 Doctor/clinical officer/medical assistant 13.0 11.1 13 Nurse/midwife 66.3 81.5 50 Other provider Patient attendant/health surveillance assistant 17.9 5.0 11 Traditional birth attendant/other 2.8 2.4 2 Total 100.0 100.0 76 N 139 92 231 1Includes women who received a check from a doctor, clinical officer, medical assistant, nurse, or midwife. Women’s Knowledge about Pregnancy and Childbirth The baseline survey tested women’s knowledge about pregnancy, childbirth, and newborn danger signs that are targeted by the National Health Communication Strategy through a SBCC campaign. One of the objectives of the SBCC campaign is to educate women about the types of information they should plan for before the birth of their child. Of the four items recommended for women’s birth preparedness planning, the most frequently reported by women was where she would get money for transportation, although less than one-fifth of women reported this item (Figure 2). The item that women least commonly reported was that they should consider who will care for their other children when they are in labor. 74 Malawi ONSE Impact Evaluation: Baseline Report Figure 2. Percentage of WRA who knew the four recommended components of birth preparedness planning (ONSE IE baseline, 2017)* *Results for women with a birth in the past three years differed occasionally but not in any systematic way. Figure 3 (and Tables A1 and A2 in Appendix A) present results on women’s knowledge of the danger signs of pregnancy. About one-half of WRA identified vaginal bleeding as a serious warning/danger sign during pregnancy. The second most frequently reported warning sign for women in the project and comparison domains was swollen hands, feet, or face (29.1 percent and 25.4 percent, respectively). Between 10 to 20 percent of women in both domains reported vaginal discharge, high fever, and severe headache as warning signs during pregnancy. Less than 10 percent of women knew that difficulty breathing, fatigue, and pale hands or eyes were warning signs during pregnancy. 13.2 15.5 19.7 3.6 13.1 14.1 15.0 3.8 0 20 40 60 80 100 What to do if you notice danger signs How you will get to the clinic Where you will get money for transportation Who will care for the other children in your home Percent Project Comparison Malawi ONSE Impact Evaluation: Baseline Report 75 Figure 3. Percentage of WRA who identified serious warning/danger signs during pregnancy (ONSE IE baseline, 2017) Figure 4 (and Tables A3 to A6 in Appendix A) present results on women’s knowledge of danger signs during childbirth. Knowledge of severe bleeding as a serious warning sign during childbirth was 54.9 percent in the project domain and 50.8 percent in the comparison domain. The second most frequently reported warning sign during childbirth was prolonged labor (lasting more than 12 hours), with 16.3 percent and 14.5 percent of women knowing this fact in the project and comparison domains, respectively. Figure 5 (and Tables A7 to A10 in Appendix A) present results on women’s knowledge of danger signs of newborn complications. Women knew little about the warning signs of newborn complications. Difficulty breathing, high fever, jaundice, pus or bleeding around the umbilical cord, and poor feeding were the most frequently reported warning signs, although none of these danger signs were reported by even one-third of women. Bleeding, pallor, convulsions, lethargy, loss of consciousness, green vomit, no stool in the first 24 hours, and swollen abdomen were reported by five percent or less of women. 53.9 11.8 29.1 17.9 6.0 7.8 5.5 19.8 49.2 10.5 25.4 12.4 4.3 7.7 6.9 18.9 0 20 40 60 80 100 Percent Project Comparison 76 Malawi ONSE Impact Evaluation: Baseline Report Figure 4. Percentage of WRA who identified serious warning/danger signs during childbirth (ONSE IE baseline, 2017)* *Results for women with a birth in the past three years differed occasionally but not in any systematic way. 11.1 54.9 5.5 1.7 8.4 10.7 16.3 5.9 11.6 9.5 4.7 4.2 1.0 11.2 50.8 3.0 1.1 7.7 7.6 14.5 7.4 10.0 7.5 6.5 3.4 1.1 0 20 40 60 80 100 Percent Project Comparison Malawi ONSE Impact Evaluation: Baseline Report 77 Figure 5. Percentage of WRA who identified serious warning/danger signs of newborn complications (ONSE IE baseline, 2017)* *Results for women with a birth in the past three years differed occasionally but not in any systematic way. 31.7 16.6 19.3 18.7 7.2 1.3 3.6 3.6 29.1 3.9 5.0 1.9 2.2 2.8 25.9 16.1 23.0 14.9 9.8 1.9 2.9 5.0 20.9 4.8 4.1 2.0 3.0 3.2 0 20 40 60 80 100 Percent Project Comparison 78 Malawi ONSE Impact Evaluation: Baseline Report CHILD HEALTH This section presents information on breastfeeding and the use of health services for fever, acute respiratory infection (ARI), and diarrhea for children under three. Women’s knowledge and beliefs about the symptoms and causes of malaria, pneumonia, and diarrhea are also presented. Background Characteristics of Children Under Three A total of 3,514 children under three years of age lived in the study households. The sex and age distribution (by months) of these children were similar across the project and comparison domains. A higher proportion of mothers were ages 20 to 24 (34.1 percent) in the comparison domain than in the project domain (29.7 percent) (Table 26). Table 26. Characteristics of children under three years of age (ONSE IE baseline, 2017) Characteristic Project Comparison N Sex Female 47.3 48.6 1,688 Male 52.7 51.4 1,826 Total 100.0 100.0 3,514 Age in months < 6 16.9 16.4 582 6-11 14.7 17.4 561 12-17 18.7 17.0 608 18-23 17.9 15.4 595 24-35 31.9 33.9 1,168 Total 100.0 100.0 3,514 Mother's age 15-19 12.1 10.3 381 20-24 29.7 34.1 1,085 25-29 22.4 22.8 756 30-34 17.9 16.5 595 35-39 11.3 11.7 397 40-44 4.6 3.5 132 Malawi ONSE Impact Evaluation: Baseline Report 79 Characteristic Project Comparison N 45-49 2.0 1.1 53 Total 100.0 100.0 3,399 District Machinga -- 38.6 873 Nkhotakota -- 24.6 510 Salima -- 36.8 483 Mzimba 64.2 -- 1,077 Nsanje 15.2 -- 256 Ntchisi 20.6 -- 315 Total 100.0 100.0 3,514 N 1,866 1,648 3,514 80 Malawi ONSE Impact Evaluation: Baseline Report Breastfeeding In the project domain, 73.7 percent of mothers with children born in health facilities in the past two years reported that they were observed breastfeeding by a healthcare worker, as did 70.5 percent of mothers in the comparison domain (Table 27). More than 88.4 percent of mothers in the project domain also reported that they were counseled on breastfeeding by a healthcare worker, as did 86.6 percent of mothers in the comparison domain. Observation of breastfeeding increased as education of the mother increased. It also tended to increase as the wealth of the mother increased, except for mothers in the lowest quintile in the project domain. Table 27. Breastfeeding observation and counseling among last births in facilities in the past two years (ONSE IE baseline, 2017) Background characteristic Observed Counseled Project Comparison Project Comparison Age group of the mother 15-19 70.0 74.1 88.3 84.5 20-24 75.8 68.0 87.5 83.8 25-29 76.1 74.4 89.5 91.1 30-34 71.9 68.9 87.4 89.4 35-39 72.0 72.6 91.4 84.4 40-44 71.3 63.0 86.0 86.9 45-49 65.2 59.7 85.9 83.3 Educational attainment of the mother No education 64.8 66.6 81.4 86.7 Some/completed primary 74.5 68.9 88.8 85.8 Some/completed secondary 76.1 76.3 92.1 88.9 More than secondary 93.4 90.6 100.0 97.7 Wealth quintile of the mother Lowest 74.0 65.8 89.0 85.7 Second 68.4 68.1 86.1 83.3 Middle 72.9 71.7 87.7 86.9 Fourth 78.7 73.4 92.1 89.4 Highest 77.2 74.6 87.9 87.8 Malawi ONSE Impact Evaluation: Baseline Report 81 Background characteristic Observed Counseled Project Comparison Project Comparison District Machinga 74.1 -- 87.3 -- Nkhotakota 78.3 -- 87.0 -- Salima 70.1 -- 90.4 -- Mzimba -- 73.7 -- 87.6 Nsanje -- 64.6 -- 83.3 Ntchisi -- 64.9 -- 85.9 Total 73.7 70.5 88.4 86.6 N 1,659 1,496 1,659 1,495 Approximately 45 percent of mothers who gave birth outside a health facility in the past two years in both study domains reported being observed breastfeeding by a healthcare worker (Table 28). Being counseled on breastfeeding by a healthcare worker was reported by 53.8 percent of mothers in the project domain and 58.0 percent in the comparison domain among women who gave birth outside a facility. Observation of breastfeeding increased as education of the mother increased. Observation of breastfeeding across the mothers’ wealth quintiles varied in both the project and comparison domains. Table 28. Breastfeeding observation and counseling among last births outside of health facilities in the past two years (ONSE IE Baseline, 2017) Background characteristic Observed Counseled Project Comparison Project Comparison Age group of the mother 15-19 61.9 29.8 65.7 29.8 20-24 32.1 51.0 30.4 59.6 25-29 45.0 46.8 64.7 82.2 30-34 54.9 45.1 58.3 46.8 35-39 28.7 50.4 46.3 57.9 40-44 25.5 29.2 25.5 41.6 45-49 75.5 0.0 75.5 40.6 82 Malawi ONSE Impact Evaluation: Baseline Report Background characteristic Observed Counseled Project Comparison Project Comparison Educational attainment of the mother No education 42.8 21.6 44.5 25.5 Some/completed primary 42.0 43.4 52.1 58.4 Some/completed secondary 89.7 70.1 89.7 76.7 More than secondary -- -- -- Wealth quintile of the mother Lowest 39.8 44.2 43.0 48.1 Second 36.2 44.2 43.4 53.3 Middle 61.3 42.7 65.8 58.9 Fourth 34.2 49.4 44.3 66.2 Highest 63.3 47.7 82.7 69.1 District Machinga 34.8 -- 35.1 -- Nkhotakota 57.0 -- 70.7 -- Salima 42.3 -- 51.2 -- Mzimba -- 51.7 -- 63.8 Nsanje -- 20.4 -- 31.0 Ntchisi -- 42.1 -- 60.8 Total 45.4 45.3 53.8 58.0 N 142 102 142 102 The percentage of children under three still being breastfed was approximately two-thirds in both domains (Table 29). More than 90 percent of children under one were still being breastfed. Among one-year-olds, 81.2 percent of children in the project domain and 85.4 percent in the comparison domain were still being breastfed, as compared with only 14.9 percent and 22.1 percent of two-year-olds, respectively. In general, breastfeeding was more commonly reported by younger women and women with less education. Malawi ONSE Impact Evaluation: Baseline Report 83 Table 29. Percentage of children who were still being breastfed among last births in the past three years (ONSE IE Baseline, 2017) Characteristics Project Comparison Yes N Yes N Age of the child Under 1 year old 92.7 583 91.5 536 1 year old 81.2 639 85.4 506 2 years old 14.9 491 22.1 485 Total 66.4 1,713 67.6 1,527 Age of the mother 15-19 74.2 211 79.7 170 20-24 64.0 556 65.4 529 25-29 60.2 396 65.3 360 30-34 69.9 312 57.3 282 35-39 58.4 211 69.6 186 40-44 63.1 79 64.2 52 45-49 33.7 36 33.5 18 Total 64.1 1,801 65.6 1,597 Educational attainment of the mother No education 64.6 233 71.8 95 Some/completed primary 64.3 1,341 67.4 1,178 Some/completed secondary 63.8 210 59.1 302 More than secondary 49.4 17 33.5 22 Total 64.1 1,801 65.6 1,597 Wealth quintile of the mother Lowest 65.2 572 70.8 349 Second 64.3 393 69.7 308 Middle 67.0 330 67.1 326 Fourth 58.3 249 63.5 326 Highest 63.9 256 54.5 288 Total 64.1 1,800 65.6 1,597 District Machinga 63.3 839 - 84 Malawi ONSE Impact Evaluation: Baseline Report Characteristics Project Comparison Yes N Yes N Nkhotakota 63.7 502 - Salima 65.4 460 - Mzimba - 66.1 1,049 Nsanje - 64.4 258 Ntchisi - 65.0 290 Total 64.1 1,801 65.6 1,597 Prevalence and Treatment of Fever, Acute Respiratory Infection, and Diarrhea Malaria, ARI, and diarrhea are three common childhood illnesses that contribute to child mortality (WHO, n.d.). This section presents results on children under three who had fever (the primary symptom of malaria), symptoms of ARI (short, rapid breathing), or diarrhea in the two weeks before the survey. Prevalence and Treatment of Fever Fever in the past two weeks was reported for a larger proportion of children under three in the project domain (38.1 percent) than in the comparison domain (31.4 percent) (Table 30). In both domains, treatment was sought for approximately 80 percent of these children. Treatment was most commonly sought on the second day (52.8 percent and 60.4 percent of children under three in the project and comparison domains, respectively). Among children for whom treatment was sought, approximately 94 percent of children in both domains were prescribed drugs. Acetaminophen/panadol/paracetamol were the most commonly prescribed drugs (70.7 percent of children in the project domain and 65.5 percent of children in the comparison domain). Antibiotics were also prescribed for 11.1 percent of children in the project domain and 17.8 percent of children in the comparison domain. Approximately six percent of children under three in both domains were prescribed an antimalarial. Malawi ONSE Impact Evaluation: Baseline Report 85 Table 30. Prevalence and treatment of fever in children under three in the past two weeks (ONSE IE baseline, 2017) Project Comparison N Child had fever Yes 38.1 31.4 1,196 No 61.9 68.6 2,318 Total 100.0 100.0 3,514 Sought treatment for fever Yes 78.8 81.2 948 No 21.2 18.8 248 Total 100.0 100.0 1,196 Days before treatment was sought Same day 10.8 10.7 111 2 days 52.8 60.4 513 3 days 23.5 20.3 218 4 or more days 12.9 8.7 106 Total 100.0 100.0 948 Among those who sought treatment for fever, prescribed drugs Yes 93.8 94.6 892 No 6.2 5.4 56 Total 100.0 100.0 948 Summary of drugs prescribed Acetaminophen/panadol/paracetamol 70.7 65.5 610 Antibiotics (pill/syrup, injection/intravenous) 11.1 17.8 133 Antimalarial 6.2 6.3 56 Aspirin/Cafenol 3.5 2.0 21 Ibuprofen 1.3 0.5 6 Other 6.0 5.4 49 Don’t know 1.3 2.6 17 Total 100.0 100.0 892 Total 1,866 1,648 3,514 86 Malawi ONSE Impact Evaluation: Baseline Report Prevalence and Treatment of Acute Respiratory Infection Symptoms of ARI in the past two weeks were reported for 15.7 percent of children under three in the project domain and 17.6 percent of children under three in the comparison domain (Table 31). Treatment was sought for more than 86 percent of children in both domains. Among children for whom treatment was sought, 53.6 percent of children in the project domain and 66.5 percent in the comparison domain received treatment on the same day or the next day. Drugs were prescribed for approximately 94 percent of children in both domains for whom treatment was sought. More than one-half of the children were prescribed aspirin/Cafenol. In the project domain, 21.7 percent of children were prescribed an antibiotic, as were 25.6 percent of children in the comparison domain. Malawi ONSE Impact Evaluation: Baseline Report 87 Table 31. Prevalence and treatment of ARI in children under three in the past two weeks (ONSE IE baseline 2017) Project Comparison N Child had symptoms of ARI Yes 15.7 17.6 579 No 84.3 82.4 2,935 Total 100.0 100.0 3,514 Sought treatment for ARI Yes 86.5 88.2 498 No 13.5 11.8 81 Total 100.0 100.0 579 Days before treatment was sought Same day 6.3 9.0 44 2 days 47.3 57.5 256 3 days 30.1 22.9 133 4 or more days 16.3 10.7 65 Total 100.0 100.0 498 Among those who sought treatment for ARI, prescribed drugs Yes 94.1 93.8 466 No 5.9 6.2 32 Total 100.0 100.0 498 Summary of drugs prescribed Acetaminophen/panadol/paracetamol 4.7 3.9 19 Antibiotics (pill/syrup, injection/intravenous) 21.7 25.6 107 Antimalarial 2.9 2.1 10 Aspirin/Cafenol 55.8 57.5 268 Ibuprofen 0.4 0.5 4 Other 13.4 7.1 45 Don’t know 1.2 3.5 13 Total 100.0 100.0 466 Total 1,866 1,648 3,514 Prevalence and Treatment of Diarrhea In the two weeks before the survey, 21.3 percent of children under three in the project domain and 23.7 percent in the comparison domain had diarrhea (Table 32). In both domains, treatment was sought for just over three-quarters of these children. Among children for whom treatment was sought in the project domain, 59.2 percent were treated with local fluids, 29.1 percent were treated with pre-packaged oral rehydration salts 88 Malawi ONSE Impact Evaluation: Baseline Report (ORS) or government recommended ORS, and 35.3 percent were given zinc tabs. In the comparison domain, 71.5 percent were treated with local fluids, 27.4 percent were treated pre-packaged ORS or government recommended ORS, and 41.7 percent were given zinc tabs. Table 32. Prevalence and treatment of diarrhea in children under three in the past two weeks (ONSE IE baseline, 2017) Project Comparison N Child had diarrhea Yes 21.3 23.7 787 No 78.8 76.3 2,727 Total 100.0 100.0 3,514 Sought treatment for diarrhea from any source Yes 77.7 76.9 599 No 22.3 23.1 188 Total 100.0 100.0 787 Days before diarrhea treatment was sought Same day 5.5 5.9 35 2 days 29.7 31.7 180 3 days 16.6 14.2 93 4 or more days 48.2 48.3 291 Total 100.0 100.0 599 Of those who sought treatment, treatment prescribed1 Local fluids 59.2 71.5 384 Pre-packaged ORS or government recommended ORS 29.1 27.4 169 Zinc tabs 35.3 41.7 234 Among those who did not seek treatment, treatment used1 Local fluids 15.0 7.1 24 Pre-packaged ORS or government recommended ORS 15.8 9.7 25 Zinc tabs 10.6 8.1 16 N 1,866 1,648 3,514 1Multiple responses allowed. Malawi ONSE Impact Evaluation: Baseline Report 89 Women’s Knowledge about Common Infectious Diseases in Children Knowledge of Symptoms and Causes of Malaria Nearly 83 percent of women in the project domain and 76.7 percent of women in the comparison domain knew that fever was a sign of malaria and just under one-half of women knew that chills were also a sign of malaria (Figure 6). Headache and joint pain were also reported as symptoms of malaria by approximately one￾third of women in both domains. The least reported symptom of malaria was poor appetite (Figure 6 and Tables A11 and A12 in Appendix A). Figure 6. Percentage of WRA who knew the signs and symptoms of malaria (ONSE IE baseline, 2017) Table 33 shows the percentage of women who identified mosquitos as the cause of malaria in the project and comparison domains. Overall, approximately 80 percent of women in both domains knew that mosquitos cause malaria. The percentage of women who correctly identified mosquitos follows an upward gradient for both education and wealth; the smallest percentage of women who had this knowledge were in the least educated (67.5 percent) and lowest wealth quintile (70.4 percent) categories in the project domain. 82.8 44.7 36.3 32.5 11.4 76.7 46.8 34.1 28.2 8.1 0 20 40 60 80 100 Fever Chills Headache Joint pain Poor appetite Percent Project Comparison 90 Malawi ONSE Impact Evaluation: Baseline Report Table 33. Percentage of WRA who correctly identified mosquitos as the cause of malaria (ONSE IE baseline, 2017) Project Comparison N Age 15-19 76.5 75.1 1,586 20-24 79.0 82.5 1,605 25-29 81.3 84.7 1,170 30-34 81.1 81.4 1,106 35-39 81.9 80.8 954 40-44 80.4 79.9 631 45-49 77.4 77.6 490 Education No education 67.5 71.6 726 Some/completed primary 78.0 77.8 5,368 Some/completed secondary 95.9 90.6 1,331 More than secondary 97.7 100.0 117 Wealth index* Lowest 70.4 78.6 1,561 Second 74.1 75.4 1,430 Middle 80.9 77.2 1,436 Fourth 86.8 81.3 1,458 Highest 88.2 87.3 1,652 District Machinga 78.4 -- 1,649 Nkhotakota 83.7 -- 1,143 Salima 77.4 -- 984 Mzimba -- 80.5 2,589 Nsanje -- 77.3 488 Ntchisi -- 81.6 689 Total 79.5 80.3 7,542 N 3,776 3,766 7,542 *Five households are missing wealth information. Malawi ONSE Impact Evaluation: Baseline Report 91 Knowledge of Symptoms and Causes of Pneumonia The National Health Communications Strategy aims to increase knowledge about the four signs and symptoms and three causes of ARI and pneumonia. Almost 60 percent of the women in the project domain and 55 percent of the women in the comparison domain knew that fast, difficult, or noisy breathing was a sign of pneumonia (Figure 7). Knowledge that cough, lethargy, and refusal to eat or breastfeed were signs of pneumonia was low in both domains. A higher percentage of WRA in the project domain reported that not dressing warmly enough was a cause of pneumonia than in the comparison domain (69.9 percent and 61.3 percent, respectively) (Figure 8). In addition to Figures 7 and 8, see Tables A13 to A15 in Appendix A. Knowledge of Symptoms and Causes of Diarrhea About 40 percent of women in the project and comparison domains knew that loose and watery stool for more than three days was a sign of diarrhea (Figure 9). Knowledge of the causes of diarrhea was slightly better, with between 35 percent and 40 percent of women stating that lack of safe drinking water and lack of food protection against contamination could cause diarrhea (Figure 10). About one-fifth of women stated that eating rotten food, touching food without washing with soap and water, and not washing hands after defecation could cause diarrhea. In addition to Figures 9 and 10, see Tables A16 to A19 in Appendix A. Figure 7. Percentage of WRA who knew the signs and symptoms of pneumonia (ONSE IE baseline, 2017) 59.1 15.9 2.4 2.1 54.2 16.5 4.3 2.6 0 20 40 60 80 100 Fast, difficult or noisy breathing Cough Lethargy Refusal to eat or breastfeed Percent Project Comparison 92 Malawi ONSE Impact Evaluation: Baseline Report Figure 8. Percentage of WRA who knew the causes of pneumonia (ONSE IE baseline, 2017) Figure 9. Percentage of WRA who knew the signs and symptoms of diarrhea (ONSE IE baseline, 2017) 69.9 0.7 2.1 61.3 0.4 0.9 0 20 40 60 80 100 Not dressed warmly enough Household air pollution Inadeqaute household ventilation Percent Project Comparison 41.1 0.8 18.3 4.6 39.6 0.8 12.7 4.6 0 20 40 60 80 100 Loose and watery stools for more than 3 days Fast or noisy breathing Lethargy Refusal to eat or breastfeed Percent Project Comparison Malawi ONSE Impact Evaluation: Baseline Report 93 Figure 10. Percentage of WRA who knew the causes of diarrhea (ONSE IE baseline, 2017) 39.8 4.5 18.9 35.3 22.3 21.7 34.7 3.5 19.7 38.0 20.4 23.2 0 20 40 60 80 100 Lack of safe drinking water Defacating or urinating in open spaces, especially in and around water sources Eating rotten food Lack of food protection against contamination Touching food without washing hands with soap Not washing hands after defecation Percent Project Comparison 94 Malawi ONSE Impact Evaluation: Baseline Report SERVICE AVAILABILITY AND READINESS ASSESSMENT This section presents the service availability and readiness indicators for a census of public and CHAM hospitals, health centers, and dispensaries in the project and comparison domains. Indicators presented are those for basic services, FP, maternal health, child health, and malaria. Characteristics of Health Facilities The health facility sample was 139 facilities. Just under 80 percent were health centers, 13.0 percent were hospitals, and 7.9 percent were dispensaries (Table 34). Three-quarters were public facilities, and one-quarter were CHAM facilities. Project facilities comprised 39.6 percent of the sample, and comparison facilities comprised 60.4 percent. In addition to Table 34, see Table A20 in Appendix A. Table 34. Characteristics of sampled health facilities (ONSE IE SARA survey, 2017) Background characteristics Percentage distribution of surveyed facilities Number of facilities surveyed Facility type Hospital 13.0 18 Health center 79.1 110 Dispensary 7.9 11 Total 100.0 139 Managing authority Government 74.8 104 CHAM 25.2 35 Total 100.0 139 District Machinga 15.1 21 Nkhotakota 14.4 20 Salima 10.1 14 Mzimba 41.0 57 Nsanje 10.1 14 Ntchisi 9.4 13 Total 100.0 139 Study domain Project 39.6 55 Comparison 60.4 84 Total 100.0 139 Malawi ONSE Impact Evaluation: Baseline Report 95 Basic Client Services and Amenities The basic client services assessed were the availability of child curative care, child growth monitoring services, any modern methods of FP, and ANC services. All basic client services were offered by 89.1 percent of project facilities and 85.7 percent of comparison facilities (Table 35). Basic amenities for client services consisted of regular electricity, an improved water source, visual and auditory privacy, a client latrine, communication equipment, a computer with Internet, and emergency transport. The results for the project and comparison domains were similar for several of the basic amenities examined. Approximately 70 percent had regular electricity, more than 92 percent had visual and audio privacy and a client latrine, and just under 20 percent had emergency transport (Table 36). A higher percentage of comparison facilities (32.1 percent) had an improved water source than project facilities (18.2 percent). On the other hand, a higher percentage of project facilities (69.1 percent) had communication equipment (phone or shortwave radio) than comparison facilities (53.6 percent). In addition, a somewhat higher percentage of project facilities (18.2 percent) than comparison facilities (14.3 percent) had a computer with Internet. In addition to Table 36, see Table A21 in Appendix A. Table 35. Availability of basic client services (ONSE IE SARA survey, 2017) Background characteristics Child curative care Child growth monitoring services Any modern methods of FP ANC services All basic client services1 Number of facilities Facility type Hospital 94.4 94.4 66.7 94.4 66.7 18 Health center 100.0 100.0 96.4 99.1 96.4 110 Dispensary 100.0 100.0 90.9 27.3 27.3 11 Total 99.3 99.3 92.1 92.8 87.1 139 Managing authority Government 99.0 99.0 99.0 92.3 92.3 104 CHAM 100.0 100.0 71.4 94.3 71.4 35 Total 99.3 99.3 92.1 92.8 87.1 139 District Machinga 100.0 100.0 90.5 90.5 81.0 21 Nkhotakota 100.0 100.0 100.0 95.0 95.0 20 Salima 100.0 100.0 92.9 100.0 92.9 14 Mzimba 98.2 98.2 91.2 91.2 86.0 57 Nsanje 100.0 100.0 85.7 100 85.7 14 Ntchisi 100.0 100.0 92.3 84.6 84.6 13 Total 99.3 99.3 92.1 92.8 87.1 139 96 Malawi ONSE Impact Evaluation: Baseline Report Background characteristics Child curative care Child growth monitoring services Any modern methods of FP ANC services All basic client services1 Number of facilities Study domain Project 100.0 100.0 94.5 94.5 89.1 55 Comparison 98.8 98.8 90.5 91.7 85.7 84 Total 99.3 99.3 92.1 92.8 87.1 139 1 Basic client services were child curative care, child growth monitoring services, any modern methods of FP, and ANC services. Table 36. Availability of basic amenities for client services (ONSE IE SARA survey, 2017) Background characteristics Regular electricity Improved water source Visual and auditory privacy Client latrine Commu￾nication equip￾ment Com￾puter with Internet Emer￾gency trans￾port Number of facilities Facility Type Hospital 88.9 5.6 88.9 94.4 88.9 72.2 83.3 18 Health center 65.5 27.3 97.3 93.6 57.3 7.3 8.2 110 Dispensary 81.8 54.5 90.9 90.9 36.4 9.1 18.2 11 Total 69.8 26.6 95.7 93.5 59.7 15.8 18.7 139 Managing authority Government 67.3 29.8 95.2 92.3 60.6 8.7 9.6 104 CHAM 77.1 17.1 97.1 97.1 57.1 37.1 45.7 35 Total 69.8 26.6 95.7 93.5 59.7 15.8 18.7 139 District Machinga 52.4 14.3 95.2 90.5 81.0 14.3 14.3 21 Nkhotakota 85.0 20.0 95.0 90.0 80.0 25.0 20.0 20 Salima 78.6 21.4 92.9 100.0 35.7 14.3 21.4 14 Mzimba 66.7 40.4 98.3 91.2 47.4 12.3 17.5 57 Nsanje 78.6 14.3 85.7 100.0 57.1 21.4 21.4 14 Ntchisi 69.2 15.4 100.0 100.0 76.9 15.4 23.1 13 Total 69.8 26.6 95.7 93.5 59.7 15.8 18.7 139 Study domain Project 70.9 18.2 94.5 92.7 69.1 18.2 18.2 55 Comparison 69.0 32.1 96.4 94.0 53.6 14.3 19.0 84 Total 69.8 26.6 95.7 93.5 59.7 15.8 18.7 139 Malawi ONSE Impact Evaluation: Baseline Report 97 Availability of Priority Medicines for Mothers and Children Tables 37 and 38 (and Tables A22 and A23 in Appendix A) provide a snapshot of the percentage of facilities that had priority medicines, by medicine, for mothers and children on the day of the survey. There were variations, by facility type, in which medicines were available for women’s and children’s priority medicines. The least available medicines for mothers were ampicillin powder and cefixime capsules/tablets whereas the most available medicines were benzathine benzylpenicillin powder for injection, gentamicin, and oxytocin. The least available medicines for children were ampicillin powder for injection and paracetamol, and the most available medicines for children were gentamicin injection, artesunate rectal or injection dosage forms, and artemisinin combination therapy (ACT). 98 Malawi ONSE Impact Evaluation: Baseline Report Table 37. Percentage of facilities with priority medicines for mothers (ONSE IE baseline, 2017) Background characteristics Oxytocin Sodium chloride inj. solution Calcium gluconate inj. Magnesium sulphate inj. Ampicillin powder (inj.) Gentamicin injection Metronidazole injection Misoprostol Azithromycin Cefixime Benzathine benzylpenicillin powder Betamethasone/ Dexamethasone injection Nifedipine Hydralazine injection Methyldopa tablet N Facility Type Hospital 89.5 73.7 27.8 89.5 57.9 94.7 77.8 50.0 66.7 11.1 94.7 78.9 72.2 83.3 83.3 18 Health center 93.6 70.9 -- 74.5 6.4 98.2 -- -- -- -- 100.0 18.2 -- -- -- 110 Dispensary 30.0 50.0 -- 10.0 0.0 70.0 -- -- -- -- 70.0 10.0 -- -- -- 11 Total 88.5 69.8 27.8 71.9 13.0 95.7 77.8 50.0 66.7 11.1 97.1 25.9 72.2 83.3 83.3 139 Managing authority Government 88.5 71.2 22.2 69.2 7.7 94.2 66.7 55.6 77.8 0.0 96.2 18.3 44.4 66.7 66.7 104 CHAM 88.6 65.7 33.3 80.0 28.6 100.0 88.9 44.4 55.6 22.2 100.0 48.6 100.0 100.0 100.0 35 Total 88.5 69.8 27.8 71.9 13.0 95.7 77.8 50.0 66.7 11.1 97.1 25.9 72.2 83.3 83.3 139 District Machinga 85.7 71.4 100.0 76.2 4.8 90.5 100.0 0.0 100.0 0.0 90.5 23.8 100.0 100.0 100.0 21 Nkhotakota 85.0 80.0 25.0 85.0 15.0 100.0 75.0 50.0 50.0 0.0 100.0 50.0 75.0 75.0 75.0 20 Salima 92.9 64.3 00. 85.7 35.7 100.0 100.0 100.0 100.0 0.0 100.0 21.4 0.0 100.0 100.0 14 Mzimba 87.7 71.9 37.5 57.9 8.8 94.7 62.5 37.5 50.0 12.5 96.5 22.8 62.5 75.0 75.0 57 Nsanje 100.0 50.0 0.0 85.7 21.4 92.9 100.0 66.7 100.0 33.3 100.0 21.4 100.0 100.0 100.0 14 Ntchisi 84.6 69.2 0.0 76.9 7.7 100.0 100.0 100.0 100.0 0.0 100.0 15.4 100.0 100.0 100.0 13 Total 88.5 69.8 27.8 71.9 13.0 95.7 77.8 50.0 66.7 11.1 97.1 25.9 72.2 83.3 83.3 139 Study domain Project 89.3 67.9 25.0 65.5 10.7 95.2 75.0 50.0 66.7 16.7 97.6 21.4 75.0 83.3 83.3 84 Comparison 87.3 72.7 33.3 81.8 16.4 96.4 83.3 50.0 66.7 0.0 96.4 32.7 66.7 83.3 83.3 55 Total 88.5 69.8 27.8 71.9 13.0 95.7 77.8 50.0 66.7 11.1 97.1 25.9 72.2 83.3 83.3 139 Malawi ONSE Impact Evaluation: Baseline Report 99 Table 38. Percentage of facilities with priority medicines for children (ONSE IE baseline, 2017) Background characteristics Amoxicillin Ampicillin powder for injection Ceftriaxone powder for injection Gentamicin injection Procaine benzylpenicillin powder for inj. ORS Zinc sulphate ACT Artesunate rectal or inj. forms Vitamin A Morphine granule, injection Paracetamol N Facility Type Hospital 68.4 57.9 83.3 94.7 21.1 89.5 72.2 89.5 94.7 47.4 44.4 57.9 19 Health center 52.7 6.4 -- 98.2 49.1 77.3 -- 91.8 94.5 60.0 -- 29.1 110 Dispensary 30.0 0.0 -- 70.0 40.0 80.0 -- 100.0 90.0 40.0 -- 20.0 10 Total 53.2 13.0 83.3 95.7 44.6 79.1 72.2 92.1 94.2 56.8 44.4 32.4 139 Managing authority Government 44.2 7.7 66.7 94.2 44.2 80.8 66.7 95.2 95.2 56.7 55.6 17.3 104 CHAM 80.0 28.6 100.0 100.0 45.7 74.3 77.8 82.9 91.4 57.1 33.3 77.1 35 Total 53.2 13.0 83.3 95.7 44.6 79.1 72.2 92.1 94.2 56.8 44.4 32.4 139 District Machinga 33.3 4.8 100.0 90.5 52.4 38.1 100.0 95.2 90.5 42.9 100.0 33.3 21 Nkhotakota 45.0 15.0 75.0 100.0 30.0 90.0 50.0 95.0 100.0 50.0 2.05 20.0 20 Salima 42.9 35.7 100.0 100.0 42.9 92.9 100.0 92.9 92.9 28.6 100.0 35.7 14 Mzimba 61.4 8.8 75.0 94.7 40.4 82.5 75.0 87.7 91.2 73.7 37.5 26.3 57 Nsanje 71.4 21.4 100.0 92.9 57.1 85.7 66.7 100.0 100.0 42.9 33.3 64.3 14 Ntchisi 53.8 7.7 100.0 100.0 61.5 92.3 100.0 92.3 100.0 61.5 100.0 38.5 13 Total 53.2 13.0 83.3 95.7 44.6 79.1 72.2 92.1 94.2 56.8 44.4 32.4 139 Study domain Project 61.9 10.7 83.3 95.2 46.4 84.5 75.0 90.5 94.0 66.7 41.7 34.5 84 Comparison 40.0 16.4 83.3 96.4 41.8 70.9 66.7 94.5 94.5 41.8 50.0 29.1 55 Total 53.2 13.0 83.3 95.7 44.6 79.1 72.2 92.1 94.2 56.8 44.4 32.4 139 100 Malawi ONSE Impact Evaluation: Baseline Report Family Planning Services Of the facilities surveyed, approximately 95 percent of project facilities and 91 percent of comparison facilities offered FP services (data not shown). All these facilities offered modern methods of FP (Table 39). More than 98 percent of both project and comparison facilities offered injectables, and 100 percent of project facilities and 96.1 percent of comparison facilities offered male condoms. Pills were offered by 96.2 percent of project facilities and 88.2 percent of comparison facilities. Female condoms were offered by 92.3 percent of project facilities and 86.8 percent of comparison facilities. Implants were also offered by most facilities: 88.5 percent of project and 93.4 percent of comparison facilities. Cycle beads were available at approximately two-thirds of project facilities and one-half of comparison facilities. About one-third of facilities offered IUDs. Female sterilization was offered by 31.0 percent and 38.3 percent of project and comparison facilities, respectively. Male sterilization was the least common modern method available, offered by only 18.3 percent and 21.3 percent of project and comparison facilities, respectively. In addition to Table 39, see Table A24 in Appendix A. Guidelines on FP were available at 92.3 percent and 94.7 percent of project and comparison facilities that offered FP services, respectively (Table 40). Just over 60 percent of project facilities and two-thirds of comparison facilities had at least one staff person trained in FP. A blood pressure apparatus was available at 84.6 percent and 78.9 percent of project and comparison facilities, respectively. Whereas 69.2 percent of project facilities had an examination light, only 42.1 percent of comparison facilities had this equipment. Malawi ONSE Impact Evaluation: Baseline Report 101 Table 39. Availability of FP methods among facilities that provided FP services (ONSE IE SARA survey, 2017) Background characteristics Provision of the following modern methods1 Any modern method4 Number of Pills facilities 2 Injectables 3 Female condoms Male condoms IUD Implant Cycle beads Male sterilization Female sterilization Facility type Hospital 91.7 100.0 91.7 100.0 66.7 91.7 83.3 83.3 83.3 100.0 12 Health center 92.5 98.1 89.6 97.2 34.0 93.4 57.5 12.3 28.3 100.0 106 Dispensary 80.0 100.0 80.0 100.0 -- 70.0 20.0 -- -- 100.0 10 Total 91.4 98.4 89.1 97.7 37.3 91.4 57.0 19.5 33.9 100.0 128 Managing authority Government 91.3 99.0 90.3 97.1 41.9 93.2 56.3 19.4 35.5 100.0 103 CHAM 92.0 96.0 84.0 100.0 20.0 84.0 60.0 20.0 28.0 100.0 25 Total 91.4 98.4 89.1 97.7 37.3 91.4 57.0 19.5 33.9 100.0 128 District Machinga 100.0 94.7 100.0 100.0 31.3 94.7 79.0 12.5 31.3 100.0 19 Nkhotakota 95.0 100.0 90.0 100.0 50.0 95.0 60.0 38.9 50.0 100.0 20 Salima 92.3 100.0 84.6 100.0 15.4 69.2 53.9 7.7 30.8 100.0 13 Mzimba 86.5 100.0 88.5 96.2 47.9 96.2 51.9 18.8 31.3 100.0 52 Nsanje 91.7 91.7 75.0 100.0 16.7 100.0 58.3 16.7 16.7 100.0 12 Ntchisi 91.7 100.0 91.7 91.7 27.3 75.0 41.7 18.2 45.5 100.0 12 Total 91.4 98.4 89.1 97.7 37.3 91.4 57.0 19.5 33.9 100.0 128 Study domain Project 96.2 98.1 92.3 100.0 39.4 88.5 65.4 18.3 31.0 100.0 52 Comparison 88.2 98.7 86.8 96.1 34.0 93.4 51.3 21.3 38.3 100.0 76 Total 91.4 98.4 89.1 97.7 37.3 91.4 57.0 19.5 33.9 100.0 128 1 Facility reported providing or prescribing the methods. 2 Facility provided or prescribed estrogen progesterone oral contraceptive pills or progestin-only contraceptive pills. 3 Facility provided or prescribed combined estrogen progesterone injectable contraceptives or progestin-only injectable contraceptives. 4 Facility provided or prescribed clients with any of the following: contraceptive pills (combined or progestin-only), injectables (combined or progestin-only), implants, IUDs, male condoms, female condoms, cycle beads for standard days method, female sterilization (tubal ligation), or male sterilization (vasectomy). 102 Malawi ONSE Impact Evaluation: Baseline Report Table 40. Availability of guidelines, trained staff, and equipment for FP services (ONSE IE SARA Maternal Health Services Of the facilities surveyed, approximately 87 percent of project facilities and 88 percent of comparison facilities offered maternal health services (data not shown). ANC services were offered by 94.5 percent of project facilities and 91.7 percent of comparison facilities (Table 41). Normal delivery services were offered by approximately 95 percent of both project and comparison hospitals and health centers. Cesarean delivery was available at two-thirds of both project and comparison hospitals. A provider of delivery care was available onsite or on-call 24 hours per day at 70.6 percent of hospitals overall. In addition to Table 41, see Table A25 in Appendix A. Background characteristics Percentage of facilities offering any modern FP that had: Number of facilities offering any modern FP methods Guidelines on FP Staff trained in FP Equipment Blood pressure apparatus Exam light Facility type Hospital 91.7 66.7 91.7 58.3 12 Health center 93.4 64.2 80.2 51.9 106 Dispensary 100.0 70.0 80.0 60.0 10 Total 93.8 64.8 81.3 53.1 128 Managing authority Government 93.2 67.0 77.7 47.6 103 CHAM 96.0 56.0 96.0 76.0 25 Total 93.8 64.8 81.3 53.1 128 District Machinga 84.2 73.7 78.9 73.7 19 Nkhotakota 100.0 50.0 100.0 80.0 20 Salima 92.3 61.5 69.2 46.2 13 Mzimba 96.2 69.2 82.7 46.2 52 Nsanje 91.7 50.0 75.0 41.7 12 Ntchisi 91.7 75.0 66.7 25.0 12 Total 93.8 64.8 81.3 53.1 128 Study domain Project 92.3 61.5 84.6 69.2 52 Comparison 94.7 67.1 78.9 42.1 76 Total 93.8 64.8 81.3 53.1 128 Malawi ONSE Impact Evaluation: Baseline Report 103 Table 42 presents information on signal functions critical to BEmONC and comprehensive emergency obstetric and newborn care (CEmONC) performed in the past 12 months. Signal functions for emergency obstetric and newborn care are the major interventions for averting maternal and neonatal mortalities. Parenteral administration of antibiotics and oxytocin and neonatal resuscitation with a bag and mask were performed by approximately 95 percent to 100 percent of hospitals and health centers in both domains in the past 12 months. More than 83 percent of hospitals and health facilities also performed parenteral administration of magnesium sulfate and blood transfusions. The least frequently performed signal function in the past 12 months was removal of retained products of conception (50.8 percent of hospitals and health centers). Approximately one-half of project facilities that offered maternal health services had at least one staff member trained in integrated management of pregnancy and childbirth (IMPAC) and had guidelines for IMPAC on hand. Among comparison facilities, only 45.8 percent had staff trained on IMPAC and only 29.2 percent had guidelines on hand. With regard to basic equipment for normal delivery at facilities that offered maternal health services, approximately 80 percent to 100 percent of both project and comparison facilities had five of nine recommended items: gloves, a partograph, a neonatal mask and bag, a suction apparatus, and a delivery pack (Table 43). More than 40 percent had an examination light, and approximately one-quarter of project facilities and one-third of comparison facilities had a manual vacuum extractor. Less commonly available delivery equipment were a vacuum aspirator or D&C kit (17.6 percent of project and 22.9 percent of comparison facilities) and emergency transport (17.6 percent of project and 18.8 percent of comparison facilities). In addition to Table 43, see Table A26 in Appendix A. 104 Malawi ONSE Impact Evaluation: Baseline Report Table 41. Availability of maternal health services (ONSE IE SARA survey, 2017) Background characteristics Percentage of facilities offering: Number of facilities Provider of delivery care available onsite or on￾call 24 hours per day Number of facilities offering normal delivery ANC services Normal delivery service Caesarean delivery ANC and normal delivery service ANC, normal delivery, and Caesarean delivery Facility type Hospital 94.7 89.5 66.7 94.4 66.7 18 70.6 17 Health center 99.1 95.5 -- 95.5 -- 110 0.0 105 Dispensary 20.0 -- -- -- -- 11 -- NA Total 92.8 95.3 66.7 95.3 66.7 139 9.8 122 Managing authority Government 92.3 96.8 66.7 96.8 66.7 104 6.6 91 CHAM 94.3 91.2 66.7 91.2 66.7 35 19.4 31 Total 92.8 95.3 66.7 95.3 66.7 139 9.8 122 District Machinga 90.5 94.4 100.0 94.4 100.0 21 5.9 17 Nkhotakota 95.0 94.4 50.0 94.4 50.0 20 11.8 17 Salima 100.0 100.0 100.0 100.0 100.0 14 7.1 14 Mzimba 91.2 94.2 50.0 94.2 50.0 57 8.2 49 Nsanje 100.0 100.0 100.0 100.0 100.0 14 21.4 14 Ntchisi 84.6 91.7 100.0 91.7 100.0 13 9.1 11 Total 92.8 95.3 66.7 95.3 66.7 139 9.8 122 Study domain Project 94.5 96.0 66.7 96.0 66.7 55 8.3 48 Comparison 91.7 94.9 66.7 94.9 66.7 84 10.8 74 Total 92.8 95.3 66.7 95.3 66.7 139 9.8 122 Malawi ONSE Impact Evaluation: Baseline Report 105 Table 42. Signal functions of BEmONC and CEmONC performed in the past 12 months BEmONC CEmONC N Parenteral administration Antibiotics Oxytocin Magnesium sulphate Assisted vaginal delivery Manual removal of placenta Removal of retained products of conception Neonatal resuscitation with bag and mask C-section Blood transfusion Facility type Hospital 100.0 100.0 94.1 94.1 88.2 94.1 100.0 70.6 88.2 17 Health center 94.3 99.0 81.9 53.3 70.5 43.8 98.1 -- -- 105 Total 95.1 99.2 83.6 59.0 73.0 50.8 98.4 70.6 88.2 122 Managing authority Government 95.6 100.0 81.3 58.2 72.5 49.5 98.9 75.0 75.0 91 CHAM 93.5 96.8 90.3 61.3 74.2 54.8 96.8 66.7 100.0 31 Total 95.1 99.2 83.6 59.0 73.0 50.8 98.4 70.6 88.2 122 District Machinga 100.0 100.0 88.2 70.6 100.0 58.8 100.0 100.0 100.0 17 Nkhotakota 88.2 100.0 70.6 52.9 70.6 52.9 100.0 50.0 75.0 17 Salima 100.0 100.0 92.9 85.7 78.6 57.1 100.0 100.0 100.0 14 Mzimba 95.9 100.0 77.6 61.2 61.2 49.0 98.0 57.1 85.7 49 Nsanje 100.0 100.0 100.0 50.0 78.6 50.0 100.0 100.0 100.0 14 Ntchisi 81.8 90.9 90.9 18.2 72.7 36.4 90.9 100.0 100.0 11 Total 95.1 99.2 83.6 59.0 73.0 50.8 98.4 70.6 88.2 122 Study domain Project 94.6 98.6 83.8 52.7 66.2 47.3 97.3 72.7 90.9 74 Comparison 95.8 100.0 83.3 68.8 83.3 56.3 100.0 66.7 83.3 48 Total 95.1 99.2 83.6 59.0 73.0 50.8 98.4 70.6 88.2 122 106 Malawi ONSE Impact Evaluation: Baseline Report Table 43. Availability of guidelines, trained staff, and equipment for delivery services (ONSE IE SARA survey, 2017) Background characteristics Percentage of facilities offering normal delivery service that had: Number of facilities offering delivery services Guide￾lines on IMPAC Staff trained in IMPAC Equipment Emer￾gency transport Exam light Delivery pack Suction apparatus (mucus extractor) Manual vacuum extractor Vacuum aspirator or D&C kit Neonatal bag and mask Parto￾graph Gloves Facility type Hospital 64.7 88.2 88.2 52.9 100.0 88.2 94.1 58.8 94.1 100.0 100.0 17 Health center 36.2 41.9 6.7 41.9 80.0 91.4 19.0 13.3 90.5 90.5 100.0 105 Total 64.7 48.4 18.0 52.9 82.8 88.2 94.1 58.8 94.1 100.0 100.0 122 Managing authority Government 36.3 47.3 9.9 40.7 76.9 90.1 28.6 17.6 92.3 92.3 100.0 91 CHAM 51.6 51.6 41.9 51.6 100.0 93.5 32.3 25.8 87.1 90.3 100.0 31 Total 40.2 48.4 18.0 43.4 82.8 91.0 29.5 19.7 910 91.8 100.0 122 Districts Machinga 11.8 52.9 17.6 47.1 88.2 100.0 41.2 29.4 82.4 94.1 100.0 17 Nkhotakota 41.2 35.3 17.6 41.2 94.1 70.6 41.2 23.5 100.0 100.0 100.0 17 Salima 35.7 50.0 21.4 35.7 85.7 85.7 21.4 14.3 71.4 64.3 100.0 14 Mzimba 42.9 44.9 16.3 44.9 75.5 93.9 28.6 22.4 91.8 91.8 100.0 49 Nsanje 71.4 57.1 21.4 64.3 100.0 100.0 21.4 7.1 100.0 100.0 100.0 14 Ntchisi 36.4 63.6 18.2 18.2 63.6 90.9 18.2 9.1 100.0 100.0 100.0 11 Total 40.2 48.4 18.0 43.4 82.8 91.0 29.5 19.7 91.0 91.8 100.0 122 Malawi ONSE Impact Evaluation: Baseline Report 107 Background characteristics Percentage of facilities offering normal delivery service that had: Number of facilities offering delivery services Guide￾lines on IMPAC Staff trained in IMPAC Equipment Emer￾gency transport Exam light Delivery pack Suction apparatus (mucus extractor) Manual vacuum extractor Vacuum aspirator or D&C kit Neonatal bag and mask Parto￾graph Gloves Study domain Project 47.3 50.0 17.6 44.6 78.4 94.6 25.7 17.6 94.6 94.6 100.0 74 Comparison 29.2 45.8 18.8 41.7 89.6 85.4 35.4 22.9 85.4 87.5 100.0 48 Total 40.2 48.4 18.0 43.4 82.8 91.0 29.5 19.7 91.0 91.8 100.0 122 108 Malawi ONSE Impact Evaluation: Baseline Report ANC at Facilities Offering Maternal Health Services Of the facilities surveyed, 95 percent of project facilities and 92 percent of comparison facilities offered ANC services (data not shown). A higher percentage of comparison facilities had three essential medicines available for ANC than project facilities. Iron tablets, folic acid tablets, and tetanus toxoid vaccine were available at more than 93 percent of comparison facilities that offered ANC services (Table 44). Among project facilities that offered ANC services, 78.8 percent had iron tablets, 71.2 percent had folic acid tablets, and 88.5 percent had tetanus toxoid vaccine. Table 45 provides information on basic infection control items at facilities that offered ANC services. Nearly all facilities had running water, latex gloves, and sharps containers. Between approximately 75 percent and 80 percent of facilities in both study domains also had soap. Only approximately 30 percent had a waste receptacle with a plastic bin liner. Alcohol-based hand disinfectant was available at 19.2 percent of project facilities and 40.3 percent of comparison facilities. Table 46 (and Table A27 in Appendix A) provide information on malaria services at facilities that offered ANC. Whereas nearly one-half (46.2 percent) of project facilities had a staff person trained in intermittent prevention treatment (IPT), only 27.3 percent of comparison facilities had an IPT-trained staff person. Medicines and commodities for malaria prevention were more commonly available at comparison facilities than at project facilities. Whereas IPTp and insecticide treated nets (ITNs) were available at 59.7 percent and 85.7 percent of comparison facilities, respectively, they were only available at 48.1 percent and 69.2 percent of project facilities, respectively. Malawi ONSE Impact Evaluation: Baseline Report 109 Table 44. Percentage of facilities offering ANC that had indicated medicines Background characteristics Percentage of facilities offering ANC that had indicated medicines Number of facilities offering ANC Iron tablets Folic acid tablets Iron or folic acid tablets Tetanus toxoid vaccine Facility type Hospital 100.0 100.0 100.0 88.2 17 Health center 88.1 81.7 89.0 95.4 109 Dispensary 100.0 100.0 100.0 100.0 3 Total 89.9 84.5 90.7 94.6 129 Managing authority Government 91.7 86.5 91.7 95.8 96 CHAM 84.8 78.8 87.9 90.9 33 Total 89.9 84.5 90.7 94.6 129 District Machinga 73.7 73.7 73.7 94.7 19 Nkhotakota 89.5 84.2 89.5 84.2 19 Salima 71.4 50.0 78.6 85.7 14 Mzimba 98.1 96.2 98.1 100.0 52 Nsanje 92.9 85.7 92.9 92.9 14 Ntchisi 100.0 90.9 100.0 100.0 11 Total 89.9 84.5 90.7 94.6 129 Study domain Project 78.8 71.2 80.8 88.5 52 Comparison 97.4 93.5 97.4 98.7 77 Total 89.9 84.5 90.7 94.6 129 110 Malawi ONSE Impact Evaluation: Baseline Report Table 45. Availability of items for infection control during provision of ANC (ONSE IE SARA survey, 2017) Background characteristics Percentage of facilities offering ANC that had items for infection control Number of facilities Soap offering ANC Running water Soap and running water Alcohol￾based hand disinfectant Soap and running water or alcohol￾based hand disinfectant Latex gloves Sharps container Waste receptacle Facility type Hospital 94.1 100.0 94.1 70.6 94.1 100.0 94.1 41.2 17 Health center 74.3 98.2 74.3 25.7 78.0 100.0 99.1 28.4 109 Dispensary 66.7 66.7 66.7 33.3 66.7 100.0 100.0 33.3 3 Total 76.7 97.7 76.7 31.8 79.8 100.0 98.4 30.2 129 Managing authority Government 70.8 96.9 70.8 25.0 75.0 100.0 99.0 25.0 96 CHAM 93.9 100.0 93.9 51.5 93.9 100.0 97.0 45.5 33 Total 76.7 97.7 76.7 31.8 79.8 100.0 98.4 30.2 129 District Machinga 94.7 94.7 94.7 5.3 94.7 100.0 100.0 31.6 19 Nkhotakota 63.2 100.0 63.2 21.1 63.2 100.0 100.0 26.3 19 Salima 85.7 100.0 85.7 35.7 85.7 100.0 100.0 35.7 14 Mzimba 75.0 98.1 75.0 46.2 80.8 100.0 98.1 34.6 52 Nsanje 78.6 100.0 78.6 28.6 78.6 100.0 92.9 28.6 14 Ntchisi 63.6 90.9 63.6 27.3 72.7 100.0 100.0 9.1 11 Malawi ONSE Impact Evaluation: Baseline Report 111 Background characteristics Percentage of facilities offering ANC that had items for infection control Number of facilities Soap offering ANC Running water Soap and running water Alcohol￾based hand disinfectant Soap and running water or alcohol￾based hand disinfectant Latex gloves Sharps container Waste receptacle Total 76.7 97.7 76.7 31.8 79.8 100.0 98.4 30.2 129 Study domain Project 80.8 98.1 80.8 19.2 80.8 100.0 100.0 30.8 52 Comparison 74.0 97.4 74.0 40.3 79.2 100.0 97.4 29.9 77 Total 76.7 97.7 76.7 31.8 79.8 100.0 98.4 30.2 129 112 Malawi ONSE Impact Evaluation: Baseline Report Table 46. Availability of malaria services in facilities offering ANC (ONSE IE SARA survey, 2017) Background characteristics Percentage of ANC facilities offering malaria services Staff trained in IPT Medicines and commodities Number of facilities offering IPTp ITNs ANC Facility type Hospital 100.0 50.0 72.2 88.9 17 Health center 100.0 33.0 53.2 78.0 109 Dispensary 100.0 -- -- 50.0 3 Total 100.0 34.9 55.0 79.0 129 Managing authority Government 100.0 29.2 54.2 77.1 96 CHAM 100.0 51.5 57.6 84.8 33 Total 100.0 34.9 55.0 79.1 129 District Machinga 100.0 42.1 15.8 78.9 19 Nkhotakota 100.0 47.4 57.9 57.9 19 Salima 100.0 50.0 78.6 71.4 14 Mzimba 100.0 23.1 71.2 82.7 52 Nsanje 100.0 35.7 35.7 92.9 14 Ntchisi 100.0 36.4 36.4 90.9 11 Total 100.0 34.9 55.0 79.1 129 Study domain Project 100.0 46.2 48.1 69.2 52 Comparison 100.0 27.3 59.7 85.7 77 Total 100.0 34.9 55.0 79.1 129 Child Health Child health services were available at all facilities surveyed. Nearly all facilities in both domains offered outpatient curative care for sick children, diagnosis and/or treatment of child malnutrition, routine vitamin A supplementation, growth monitoring, and treatment of pneumonia and malaria in children (Table 47). Malawi ONSE Impact Evaluation: Baseline Report 113 Guidelines for the integrated management of childhood illnesses (IMCI) and growth monitoring were present at a larger proportion of comparison facilities than project facilities (43.4 percent and 39.8 percent of comparison facilities, respectively, and only 36.4 percent and 29.1 percent of project facilities, respectively) (Table 48). Approximately 70 percent of both project and comparison facilities had at least one staff person trained in IMCI, and 55.4 percent of comparison facilities and 50.9 percent of project facilities had a staff person trained in growth monitoring, respectively. More than 95 percent of facilities had a length of height board, and approximately 90 percent had an infant scale and stethoscope. More than 85 percent had a child scale, and approximately 80 percent had a thermometer. Just under two-thirds had a growth chart (Table 48). Table 47. Availability of child health services (ONSE IE SARA survey, 2017) Background characteristics Percentage of facilities that offered: Number of facilities Outpatient curative care for sick children Diagnose and/or treat child malnutrition Routine vitamin A supple￾mentation Growth monitoring Treatment of pneumonia Treatment of malaria in children Facility Type Hospital 94.4 94.4 94.4 94.4 94.4 94.4 18 Health center 100.0 100.0 96.4 100.0 99.1 99.1 110 Dispensary 100.0 100.0 100.0 100.0 90.9 90.9 11 Total 99.3 99.3 96.4 99.3 97.8 97.8 139 Managing authority Government 99.0 99.0 97.1 99.0 98.1 98.1 104 CHAM 100.0 100.0 94.3 100.0 97.1 97.1 35 Total 99.3 99.3 96.4 99.3 97.8 97.8 139 District Machinga 100.0 100.0 90.5 100.0 100.0 100.0 21 Nkhotakota 100.0 100.0 95.0 100.0 100.0 100.0 20 Salima 100.0 100.0 100.0 100.0 92.9 92.9 14 Mzimba 98.2 98.2 98.2 98.2 96.5 96.5 57 Nsanje 100.0 100.0 100.0 100.0 100.0 100.0 14 Ntchisi 100.0 100.0 92.3 100.0 100.0 100.0 13 Total 99.3 99.3 96.4 99.3 97.8 97.8 139 114 Malawi ONSE Impact Evaluation: Baseline Report Background characteristics Percentage of facilities that offered: Number of facilities Outpatient curative care for sick children Diagnose and/or treat child malnutrition Routine vitamin A supple￾mentation Growth monitoring Treatment of pneumonia Treatment of malaria in children Study domain Project 100.0 100.0 94.5 100.0 98.2 98.2 55 Comparison 98.8 98.8 97.6 98.8 97.6 97.6 84 Total 99.3 99.3 96.4 99.3 97.8 97.8 139 Malawi ONSE Impact Evaluation: Baseline Report 115 Table 48. Availability of guidelines, trained staff, and equipment for child curative services (ONSE IE SARA survey, 2017) Background characteristics Among facilities offering curative care for sick children, percentage that had: Number of facilities offering outpatient curative care for sick children Guidelines Trained staff Equipment IMCI Growth monitoring IMCI Growth monitoring Child scale Infant scale Length of height board Thermometer Stethoscope Growth chart Facility Type Hospital 41.2 41.2 76.5 52.9 94.1 100.0 94.1 88.2 88.2 70.6 17 Health center 40.0 33.6 70.0 52.7 85.5 95.5 96.4 79.1 89.1 62.7 110 Dispensary 45.5 45.5 63.6 63.6 90.9 36.4 100.0 63.6 81.8 72.7 11 Total 40.6 35.5 70.3 53.6 87.0 91.3 96.4 79.0 88.4 64.5 138 Managing authority Government 40.8 31.1 74.8 54.4 84.5 90.3 95.1 71.8 85.4 61.2 103 CHAM 40.0 48.6 57.1 51.4 94.3 94.3 100.0 100.0 97.1 74.3 35 Total 40.6 35.5 70.3 53.6 87.0 91.3 96.4 79.0 88.4 64.5 138 Project districts Machinga 23.8 14.3 66.7 38.1 76.2 90.5 100.0 81.0 85.7 61.9 21 Nkhotakota 50.0 35.0 60.0 50.0 95.0 90.0 100.0 80.0 100.0 60.0 20 Salima 35.7 42.9 85.7 71.4 85.7 100.0 92.9 78.6 78.6 78.6 14 Mzimba 46.4 37.5 76.8 50.0 87.5 89.3 96.4 82.1 89.3 62.5 56 Nsanje 35.7 42.9 64.3 42.9 100.0 100.0 92.9 57.1 85.7 71.4 14 Ntchisi 38.5 46.2 53.8 92.3 76.9 84.6 92.3 84.6 84.6 61.5 13 Total 40.6 35.5 70.3 53.6 87.0 91.3 96.4 79.0 88.4 64.5 138 116 Malawi ONSE Impact Evaluation: Baseline Report Background characteristics Among facilities offering curative care for sick children, percentage that had: Number of facilities offering outpatient curative care for sick children Guidelines Trained staff Equipment IMCI Growth monitoring IMCI Growth monitoring Child scale Infant scale Length of height board Thermometer Stethoscope Growth chart Study domain Project 36.4 29.1 69.1 50.9 85.5 92.7 98.2 80.0 89.1 65.5 55 Comparison 43.4 39.8 71.1 55.4 88.0 90.4 95.2 78.3 88.0 63.9 83 Total 40.6 35.5 70.3 53.6 87.0 91.3 96.4 79.0 88.4 64.5 138 Malawi ONSE Impact Evaluation: Baseline Report 117 Malaria Of the facilities surveyed, all project facilities and 98 percent of comparison facilities offered malaria services (Table 49). Whereas 40 percent of project facilities had guidelines for the treatment of malaria on hand, only 20.7 percent of comparison facilities had malaria guidelines. A higher proportion of project facilities (83.6 percent) than comparison facilities (65.9 percent) had a staff person trained on malaria diagnosis and/or treatment. Nearly all facilities had malaria rapid diagnostic tests (RDTs) among their diagnostics. Only 18.2 percent of project and 30.0 percent of comparison hospitals and health centers had malaria microscopy available. In addition to Table 49, see Table A28 in Appendix A. Table 50 (and Table A29 in Appendix A) provide information on antimalarial medicines available at facilities. ACTs were available at 94.5 percent of project facilities and 91.5 percent of comparison facilities that offered malaria services. SP was the next most common antimalarial available, available at 45.5 percent of project facilities and 57.3 percent of comparison facilities. Oral quinine was available at 36.4 percent of project facilities and 29.3 percent of comparison facilities. 118 Malawi ONSE Impact Evaluation: Baseline Report Table 49. Availability of malaria services, and guidelines, trained staff, and diagnostic capacity (ONSE IE SARA survey, 2017) Background characteristics Percentage of facilities that offered malaria diagnosis/ treatment Number of facilities Among facilities that offered malaria diagnosis/treatment services: Number of facilities that offered malaria diagnosis/ treatment Guidelines Trained staff Diagnostics Guidelines for diagnosis/ treatment of malaria Staff trained in malaria diagnosis/ treatment Malaria RDT Malaria microscopy Any malaria diagnostics Facility type Hospital 94.4 18 41.2 88.2 100.0 88.2 100.0 17 Health center 100.0 110 26.4 70.9 99.1 12.7 99.1 110 Dispensary 90.9 11 30.0 70.0 100.0 -- 100.0 10 Total 98.6 139 28.5 73.0 99.3 22.8 99.3 137 Managing authority Government 98.1 104 24.5 69.6 100.0 17.2 100.0 102 CHAM 100.0 35 40.0 82.9 97.1 38.2 97.1 35 Total 98.6 139 28.5 73.0 99.3 22.8 99.3 137 District Machinga 100.0 21 42.9 85.7 95.2 44.4 95.2 21 Nkhotakota 100.0 20 35.0 85.0 100.0 16.7 100.0 20 Salima 100.0 14 42.9 78.6 100.0 28.6 100.0 14 Mzimba 96.5 57 18.2 70.9 100.0 17.7 100.0 55 Nsanje 100.0 14 28.6 50.0 100.0 21.4 100.0 14 Ntchisi 100.0 13 23.1 61.5 100.0 16.7 100.0 13 Total 98.6 139 28.5 73.0 99.3 22.8 99.3 137 Malawi ONSE Impact Evaluation: Baseline Report 119 Background characteristics Percentage of facilities that offered malaria diagnosis/ treatment Number of facilities Among facilities that offered malaria diagnosis/treatment services: Number of facilities that offered malaria diagnosis/ treatment Guidelines Trained staff Diagnostics Guidelines for diagnosis/ treatment of malaria Staff trained in malaria diagnosis/ treatment Malaria RDT Malaria microscopy Any malaria diagnostics Study domain Project 100.0 55 40.0 83.6 98.2 18.2 98.2 55 Comparison 97.6 84 20.7 65.9 100.0 30.0 100.0 82 Total 98.6 139 28.5 73.0 99.3 22.8 99.3 137 120 Malawi ONSE Impact Evaluation: Baseline Report Table 50. Availability of malaria medicines (ONSE IE SARA survey, 2017) Background characteristics ACTs SP Oral quinine Number of facilities that offered malaria diagnosis and/or treatment services Facility type Hospital 94.1 70.6 58.8 17 Health center 91.8 52.7 28.2 110 Dispensary 100.0 20.0 30.0 10 Total 92.7 52.6 32.1 137 Managing authority Government 96.1 52.0 19.6 102 CHAM 82.9 54.3 68.6 35 Total 92.7 52.6 32.1 137 District Machinga 95.2 14.3 14.3 21 Nkhotakota 95.0 55.0 40.0 20 Salima 92.9 78.6 64.3 14 Mzimba 89.1 67.3 29.1 55 Nsanje 100.0 35.7 35.7 14 Ntchisi 92.3 38.5 23.1 13 Total 92.7 52.6 32.1 137 Study domain Project 94.5 45.5 36.4 55 Comparison 91.5 57.3 29.3 82 Total 92.7 52.6 32.1 137 Malawi ONSE Impact Evaluation: Baseline Report 121 CONCLUSION The Malawi ONSE impact evaluation seeks to test the hypothesis that the interventions implemented by ONSE will improve health outcomes for women and children in the project domain compared with the comparison domain. The household survey conducted in 2017 as part of the Malawi ONSE impact evaluation establishes baseline indicators for household and women’s background characteristics, primary outcomes, and exposure to project or similar interventions in both the project and comparison domains. The health facility survey, conducted at the same time as the household survey, establishes baseline estimates for secondary outcomes related to the availability of health services and facility readiness to provide specific services in the project and comparison domains. Similarities and differences in these indicators and outcomes across domains are summarized below. Primary Outcomes The ONSE project targets FP/RH, MNCH, and WASH outcomes in the project domain with the goal of decreasing maternal, newborn, and child morbidity and mortality. One or several indicators for each project area was chosen to measure achievement of this goal. Family Planning The use of FP was very similar in the project and comparison domains, with approximately 55 percent of married women and 46 percent of all WRA using a modern contraceptive method. Injectables and implants were by far the most popular methods. More women in the project domain used injectables than women in the comparison domain. The opposite was true of implants; more women in the comparison domain used implants than women in the project domain. Maternal Health The use of ANC services at baseline was high, with almost all women attending at least one ANC visit during their last pregnancy. Attendance by a skilled provider was almost universal in both domains. However, meeting the minimum four ANC visit recommendations during pregnancy was low overall, and lower in the project domain than in the comparison domain (51.8 percent and 55.8 percent, respectively). Attendance at ANC during the first trimester of pregnancy was even lower, with just under one-third of the women in both domains attending an ANC visit during that time. The ONSE project focuses on early attendance for and retention in ANC, which should drive the percentage of women attending ANC during the first trimester and the overall number of ANC visits upward. The women surveyed almost universally delivered at health facilities with the attendance of skilled health professionals. Even so, only about two-thirds of the women received PNC. Approximately 60 percent of the women who gave birth received PNC within two days of birth. In sum, the majority of the women who received PNC received it in the recommended first two days after birth, but there is room for improvement with regard to the percentage of women receiving PNC. 122 Malawi ONSE Impact Evaluation: Baseline Report Newborns delivered at health facilities received PNC at a higher rate than their mothers, about 88 percent and 85 percent in project and comparison domains, respectively. Newborns who were delivered outside of a facility received a postnatal check less often than their mothers. Child Health A moderate percentage of children had been ill in the two weeks preceding the survey. Fever was the most frequent ailment reported, followed by diarrhea. The rate of care seeking for these children was around 80 percent for fever and diarrhea and close to 90 percent for children with symptoms of ARI/pneumonia. Most parents sought care for their children within two to three days, although the majority of parents waited four or more days to seek care for diarrhea. Patient Satisfaction Women were very satisfied with most aspects of services provided during their last visit to a health facility. The greatest percentage of WRA reported being less than “very satisfied” with the time they waited to see a provider and the facility service hours. The greatest percentage of women were “very satisfied” with facility cleanliness and audio and visual privacy during their visit. Women’s Knowledge and Health Beliefs Women’s knowledge about danger signs in pregnancy and childbirth was generally very low, although a higher percentage of the women in the project domain had knowledge of select items. Women who had a birth in the past three years did not exhibit a consistent pattern of higher knowledge. In terms of birth planning, between 13 percent and 20 percent of women in both the project and comparison domains knew that they should plan for what to do if they noticed danger signs, how to get to the clinic, and where to get money for transportation. A higher percentage of women in the project domain knew to plan for transportation costs. Regarding danger signs during pregnancy, vaginal bleeding; swollen hands, feet, or face; and vaginal discharge were most frequently reported. A higher percentage of women in the project domain knew that swelling was a danger sign. Just over one-half of women in the comparison and project domains knew that severe bleeding was a danger sign during childbirth, with a higher percentage of women in the project domain knowing this fact (54.9 percent and 50.8 percent, respectively). Knowledge about danger signs for newborns was also low in both domains. The most frequently reported danger signs in the project and comparison domains were breathing difficulty (31.7 percent and 25.9 percent, respectively) and high fever (29.1 percent and 20.9 percent, respectively). Knowledge about symptoms and causes of childhood illnesses was somewhat higher than knowledge of danger signs and symptoms in pregnancy, childbirth, and for newborns with many differences across domains. Malawi ONSE Impact Evaluation: Baseline Report 123 Secondary Outcomes There was less variation across domains for the secondary facility-level outcomes of interest to the evaluation, derived from the SARA. Almost all facilities provided child curative care and child growth monitoring. A somewhat smaller percentage offered ANC and FP. In terms of readiness to provide services, facilities fell short most frequently on the staffing and guidelines aspects of the index. The opposite was true with regard to basic obstetric and newborn care, although facilities did not meet the other requirements determined by WHO to consider a facility “ready” to provide those services. Two-thirds of hospitals provided cesarean sections. A majority of these facilities had provided the seven signal functions of BEmONC in the past twelve months, with assisted vaginal delivery and removal of retained products of conception being the least frequently provided. Much support is needed in this area in terms of having guidelines, training, and especially emergency transportation. The baseline estimates reveal few differences across project and comparison domain facilities. However, three differences were noted. Medicines and supplies for child health services were less available in the project domain, specifically ORS, amoxicillin, vitamin A, and zinc. The availability of ANC medicines and commodities was also lower in the project domain, specifically iron, folic acid, and ITNs (or vouchers for ITNs). Last, job aids and training for malaria services were more available in the project domain. Comparability of Project and Comparison Domains There is evidence that some key health outcomes differ between the project and comparison domains, including skilled ANC (p=0.03) and the number of women with four or more ANC visits (p=0.01). Although the prevalence of modern contraceptive use did not differ between domains, the method mix was different, specifically the use of injectables (p=0.00), implants (p=0.00), and oral contraceptives (p=0.00). The use of a skilled birth attendant, PNC, and care seeking for fever did not differ by domain. Health facilities were similar for almost all measured indicators, aside from supplies and commodities for child health and ANC, and for readiness to provide malaria services, specifically regarding staff and guidelines. Despite the differences noted, the planned methodology (the DID approach) was used under the assumption that the project and comparison domains would not be similar for all characteristics at baseline. The characteristics that remain constant over time, whether different or the same at baseline, will be differenced out by the model. Key differences in project and comparison domains that may change over time will be explored in the end line analysis. One option that can be explored during the end line analysis is to test the robustness of the findings of the DID model by re-running it on a subset of project and comparison areas that are more similar in key characteristics. 124 Malawi ONSE Impact Evaluation: Baseline Report Exposure to Other Interventions To examine the potential for contamination of the project and comparison domains, respondents were asked about assistance/support received by their household and/or community over the past year. More than 30 percent of households in both domains reported receiving some type of support, with more than two-thirds of them receiving support for malaria. WASH support was also received by between 12 percent and 15 percent of households in both domains. Interventions related to FP and MNCH were less frequently reported. It is also important to note that several of the study districts received support from the prior project, SSDI. Salima and Nkhotakota in the study domain, and Nsanje in the comparison domain all benefited from prior programming. This fact needs to be considered when interpreting the findings of the impact evaluation. Implications for the Impact Evaluation Complex interventions operating at some degree of scale in the real world raise several well-documented evaluation challenges, including the presence of other similar interventions implemented by other organizations. Our results suggest that a potential for contamination exists, primarily related to malaria outcomes. Interventions related to FP, MNCH, and WASH also present minimal risks and should be followed up and monitored. Implementation process monitoring will collect information periodically about other projects operating in the study domains to ensure current knowledge of contamination risks. Of concern is any widespread exposure in both the project and comparison domains to programming relevant to the project interventions and outcomes of interest, for example, malaria programming. Analysis of this exposure and the type and timing of malaria programming will need to be accounted for in the end line analysis to explore its potential implications for the evaluation’s findings. Ideally, exposure to other program activities at baseline would be negligent, but in practice, this is generally not realistic. Analysis of the small exposures to FP, MNCH, and WASH programming may also be included in the end line analysis to explore any potential implications for the evaluation’s findings. ONSE’s community engagement and mobilization work will focus on the specific, self-identified needs in each community, and therefore, dissemination of information about these topics will not be universal. As this aspect of the project rolls out, indicators may be refined to accurately capture outcomes of the ONSE project’s community activities. For interventions that are designed to tailor to the needs of specific subgroups (i.e., districts or communities) with differential activities and audiences, this flexibility is needed to measure the effects of the programming. Last, two primary outcomes—the percentage of women receiving ANC and giving birth with a skilled provider—were already above 90 percent in the study population. It is likely that these indicators will not increase significantly during the evaluation period. However, related outcomes, such as the percentage of women who attend four or more ANC visits and the percentage of women who receive PNC within two days (forty-eight hours) of birth, show room for improvement. Malawi ONSE Impact Evaluation: Baseline Report 125 Next Steps End line data collection is planned for 2021. The same households will be interviewed at that time to evaluate the impact of ONSE on the health outcomes of interest in the project domain. The DID approach will be used to compare pre- and post-intervention differences in outcomes between the project and comparison domains. Qualitative analysis will aim to describe and understand differences in how respondents in ONSE’s project communities were exposed to the SBCC campaign. Ongoing implementation process monitoring will occur annually through the time of the end line survey. This monitoring will focus on how the “smart” approach was operationalized in the project domain and will seek to identify pathways through which this approach affects project beneficiaries. Implementation process monitoring will also provide information about exposure to other activities that may affect the outcomes of the impact evaluation. 126 Malawi ONSE Impact Evaluation: Baseline Report REFERENCES DHIS 2. Retrieved from https://www.dhis2.org/ Government of Malawi. (2012). Findings from the 2012 baseline survey of 15 districts in Malawi. Lilongwe, Malawi: Government of Malawi. Retrieved from https://www.thehealthcompass.org/sites/default/files/project_examples/final_ssdi_baseline_survey_report _may_2012.pdf Government of Malawi. (2016). Findings from the 2016 endline survey of 15 districts in Malawi. Lilongwe, Malawi: Government of Malawi. Retrieved from https://www.k4health.org/sites/default/files/ssdi_endline_draft_report_dec_20_2016_final_with_appendic es.pdf Ministry of Health, Malawi & ICF International. (2014). Malawi Service Provision Assessment 2013-14. Lilongwe, Malawi, and Rockville, MD, USA: Ministry of Health & ICF International. Retrieved from https://dhsprogram.com/pubs/pdf/SPA20/SPA20[Oct-7-2015].pdf Ministry of Health, Central Monitoring and Evaluation Division, & Jhpiego. (2016). End line assessment report. Unpublished. National Malaria Control Programme [Malawi] & ICF International. (2015). Malawi malaria indicator survey (MIS) 2014. Lilongwe, Malawi, and Rockville, MD, USA: NMCP & ICF International. Retrieved from https://dhsprogram.com/publications/publication-MIS18-MIS-Final-Reports.cfm National Malaria Control Programme (NMCP) and ICF. (2018). Malawi Malaria Indicator Survey 2017. Lilongwe, Malawi, and Rockville, Maryland, USA: NMCP and ICF. Retrieved from https://dhsprogram.com/pubs/pdf/MIS28/MIS28.pdf National Statistical Office (NSO) & ICF Macro. (2011). Malawi demographic and health survey 2010. Zomba, Malawi, and Calverton, MD, USA: NSO & ICF Macro. Retrieved from https://dhsprogram.com/pubs/pdf/fr247/fr247.pdf National Statistical Office [Malawi] & ICF. (2017). Malawi demographic and health survey 2015-16. Zomba, Malawi, and Rockville, Maryland, USA. NSO & ICF. Retrieved from https://dhsprogram.com/pubs/pdf/FR319/FR319.pdf National Statistical Office (NSO). (2015). Malawi MDG endline survey 2014. Zomba, Malawi: NSO. Retrieved from http://www.nsomalawi.mw/index.php?option=com_content&view=article&id=210&Itemid=98 O'Hagan, R., Marx, M.A., Finnegan, K. E., Naphini, P., Ng'ambi, K., Laija, K., …Yosefe, S. (2017). National assessment of data quality and associated systems-level factors in Malawi. Global Health: Science and Practice, 5(3):367–381. Retrieved from http://www.ghspjournal.org/content/5/3/367 Malawi ONSE Impact Evaluation: Baseline Report 127 Tough, A. G. & Lihoma, L. (2017). Health information systems and medical record keeping in Malawi: A report on preliminary field research with recommendations. Retrieved from https://www.gla.ac.uk/media/media_541720_en.pdf Victora, C. G., Black, R. E., Boerma, J. T., & Bryce, J. (2011). Measuring impact in the Millennium Development Goal era and beyond: A new approach to large-scale effectiveness evaluations. The Lancet, 377: 85–95. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/20619886 World Health Organization (WHO). (2006). Reproductive health indicators: Guidelines for their generation, interpretation and analysis for global monitoring. Geneva, Switzerland: WHO. Retrieved from http://whqlibdoc.who.int/publications/2006/924156315X_eng.pdf. World Health Organization (WHO). (2002). WHO antenatal care randomized trial: Manual for the implementation of the new model. Geneva, Switzerland: WHO. Retrieved from http://whqlibdoc.who.int/hq/2001/WHO_RHR_01.30.pdf. World Health Organization (WHO). (2012). WHO Evidence Review Group: Intermittent preventive treatment of malaria in pregnancy (IPTp) with sulfadoxine-pyrimethamine (SP). WHO Headquarters, Geneva, 9-11 July 2012. Meeting report. Geneva, Switzerland: WHO. Retrieved from http://www.who.int/malaria/mpac/sep2012/iptp_sp_erg_meeting_report_july2012.pdf 128 Malawi ONSE Impact Evaluation: Baseline Report APPENDIX A. ADDITIONAL TABLES Table A1. Percentage of WRA who correctly identified the warning/danger signs in pregnancy in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Vaginal bleeding Severe headache Swollen hands, feet, or face High fever Difficulty breathing Severe weakness/ fatigue Pale hands or eyes Vaginal discharge N Age 15-19 24.3 7.0 21.5 8.3 3.2 4.7 2.8 10.2 790 20-24 52.8 12.1 27.8 18.6 5.7 8.4 5.6 21.9 836 25-29 60.5 13.0 34.7 19.7 6.6 9.7 6.3 24.6 598 30-34 69.3 11.5 34.0 23.2 7.3 8.2 4.7 23.8 554 35-39 69.8 15.8 31.5 21.9 7.1 8.1 5.6 19.4 469 40-44 60.4 15.1 33.2 22.0 7.3 9.1 10.2 16.8 295 45-49 65.5 11.9 23.4 18.4 8.1 7.7 7.1 26.8 234 Education No education 54.2 11.6 26.3 17.6 6.0 7.8 6.0 21.9 480 Some/completed primary 51.9 11.6 26.6 18.6 5.8 7.3 5.0 19.0 2,684 Some/completed secondary 62.5 13.1 42.1 14.4 6.8 9.4 6.3 21.5 556 More than secondary 66.5 10.7 43.3 24.0 11.6 15.2 15.6 18.3 56 Malawi ONSE Impact Evaluation: Baseline Report 129 Project Vaginal bleeding Severe headache Swollen hands, feet, or face High fever Difficulty breathing Severe weakness/ fatigue Pale hands or eyes Vaginal discharge N Wealth index* Lowest 48.7 10.4 24.2 16.1 6.8 8.5 5.0 21.1 917 Second 49.6 11.5 25.6 18.3 4.7 8.0 5.4 17.6 759 Middle 55.8 13.1 26.3 23.2 6.8 8.0 6.4 21.0 713 Fourth 57.3 11.2 34.8 17.3 7.0 6.0 5.3 21.5 631 Highest 60.2 12.9 36.1 15.7 4.9 8.1 5.4 17.8 752 District Machinga 46.2 12.8 26.3 20.3 3.5 8.6 5.8 18.7 1,649 Nkhotakota 65.7 11.7 36.5 18.2 5.2 6.5 5.9 20.2 1,143 Salima 52.6 10.9 26.2 15.5 9.1 7.9 4.9 20.5 984 Total 53.9 11.8 29.1 17.9 6.0 7.8 5.5 19.8 3,776 *Five households are missing wealth information. 130 Malawi ONSE Impact Evaluation: Baseline Report Table A2. Percentage of WRA who correctly identified the warning/danger signs in pregnancy in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Vaginal bleeding Severe headache Swollen hands, feet, or face High fever Difficulty breathing Severe weakness/ fatigue Pale hands or eyes Vaginal discharge N Age 15-19 18.5 5.2 12.8 6.0 1.8 4.1 2.2 6.2 796 20-24 49.3 10.4 27.0 12.3 5.2 7.1 4.2 18.2 769 25-29 58.9 11.4 31.9 15.0 3.3 8.4 7.4 26.1 572 30-34 61.4 12.8 27.9 16.8 5.1 8.9 9.9 22.6 552 35-39 63.0 12.1 31.1 15.8 4.9 9.0 10.0 24.8 485 40-44 60.4 14.3 28.9 13.9 6.2 11.0 12.5 21.3 336 45-49 55.7 12.9 24.8 9.4 5.7 9.6 8.5 22.9 256 Education No education 56.0 12.5 16.8 11.6 5.9 8.7 8.4 22.2 246 Some/completed primary 47.2 10.1 24.5 12.0 4.2 7.8 6.4 18.4 2,684 Some/completed secondary 52.3 11.2 30.0 13.2 4.1 7.3 6.9 19.5 775 More than secondary 73.3 16.9 46.5 25.8 2.7 5.3 23.8 22.4 61 Wealth index* Lowest 47.0 7.7 22.5 14.5 3.8 8.7 5.4 20.8 644 Second 45.6 11.4 24.2 13.3 5.2 9.0 6.1 16.5 671 Malawi ONSE Impact Evaluation: Baseline Report 131 Comparison Vaginal bleeding Severe headache Swollen hands, feet, or face High fever Difficulty breathing Severe weakness/ fatigue Pale hands or eyes Vaginal discharge N Middle 48.3 11.2 24.6 12.0 4.1 7.3 8.0 17.3 723 Fourth 52.2 10.8 24.3 9.5 4.1 6.4 5.1 19.0 827 Highest 51.8 11.2 30.4 13.2 4.2 7.4 9.3 20.7 900 District Mzimba 48.2 12.1 26.6 13.1 4.2 7.9 7.7 18.8 2,589 Nsanje 49.9 10.7 16.6 8.2 6.7 6.9 2.0 22.0 488 Ntchisi 52.1 5.0 26.7 12.9 3.2 7.4 7.0 17.4 689 Total 49.2 10.5 25.4 12.4 4.3 7.7 6.9 18.9 3,766 *One household is missing wealth information 132 Malawi ONSE Impact Evaluation: Baseline Report Table A3. Percentage of WRA who correctly identified the warning/danger signs of maternal complications during childbirth in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Age 15-19 5.5 25.1 3.2 0.3 3.8 5.4 9.0 2.0 6.1 4.8 1.9 2.6 0.5 790 20-24 13.2 53.4 5.3 1.0 7.8 8.7 14.8 3.5 9.9 9.2 3.9 2.6 0.9 836 25-29 12.5 61.8 6.1 2.2 9.7 14.6 19.8 7.7 14.1 9.6 6.5 5.6 1.1 598 30-34 12.5 68.6 5.7 1.9 10.2 12.8 18.7 8.7 14.6 11.8 6.0 4.8 0.7 554 35-39 11.9 67.7 6.6 2.6 12.2 12.6 18.0 7.7 15.8 11.8 5.7 4.9 1.4 469 40-44 10.2 68.1 6.7 2.4 10.3 13.2 18.3 9.7 11.4 10.6 7.6 6.6 1.1 295 45-49 15.4 67.2 8.5 4.6 8.4 13.8 25.0 7.3 15.1 15.1 3.3 5.3 2.0 234 Education No education 12.8 53.5 6.2 3.1 8.0 12.5 18.4 6.5 9.4 8.2 4.7 5.7 1.2 480 Some/completed primary 11.4 53.4 5.1 1.6 8.3 10.6 16.3 6.0 11.9 9.4 4.9 4.3 1.0 2,684 Some/completed secondary 8.3 61.2 6.7 1.3 9.0 9.7 12.8 5.0 11.5 11.0 3.6 2.5 1.0 556 More than secondary 8.0 75.9 7.6 0.0 11.0 12.7 30.3 2.3 18.3 15.2 4.4 1.7 0.0 56 Wealth index* Lowest 13.4 50.0 5.0 1.6 9.3 8.8 15.3 7.7 11.3 9.7 4.7 3.7 0.9 917 Second 11.6 52.3 5.1 2.8 10.8 10.6 15.2 5.3 10.0 7.5 4.2 4.0 1.1 759 Middle 12.1 56.0 5.4 2.0 6.6 10.8 19.1 5.9 10.4 10.5 5.6 4.1 1.4 713 Fourth 10.5 55.1 5.5 0.6 6.6 13.0 17.3 5.4 14.5 13.0 4.6 5.1 0.9 631 Malawi ONSE Impact Evaluation: Baseline Report 133 Project Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Highest 7.6 62.1 6.6 1.4 8.1 11.1 14.9 4.7 12.4 7.6 4.3 4.2 0.7 752 District Machinga 11.7 49.6 3.3 1.2 6.5 9.7 17.0 7.2 9.4 7.5 7.2 4.9 1.3 1,649 Nkhotakota 8.6 63.0 7.8 1.6 10.1 12.7 13.8 5.0 15.4 9.8 4.6 4.4 0.8 1,143 Salima 12.5 53.9 5.9 2.3 8.4 10.2 17.5 5.2 10.9 11.3 2.3 3.3 0.8 984 Total 11.1 54.9 5.5 1.7 8.4 10.7 16.3 5.9 11.6 9.5 4.7 4.2 1.0 3,776 *Four households are missing wealth information. Table A4. Percentage of WRA who correctly identified the warning/danger signs of maternal complications during childbirth in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Age 15-19 5.5 19.5 1.5 0.4 2.4 3.6 5.6 1.7 3.9 4.0 1.1 1.9 0.5 796 20-24 12.5 47.1 2.5 1.7 7.6 7.7 13.4 7.5 9.2 8.1 6.1 3.1 1.2 769 25-29 12.4 61.4 2.5 2.0 10.2 8.7 18.8 9.3 11.0 7.0 8.2 4.3 1.7 572 30-34 12.7 65.7 4.4 1.2 10.8 8.3 16.8 9.5 15.0 8.4 7.6 4.8 0.6 552 35-39 14.8 64.7 4.2 0.9 9.4 8.7 17.5 9.1 12.5 10.4 9.3 4.0 2.0 485 134 Malawi ONSE Impact Evaluation: Baseline Report Comparison Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N 40-44 9.8 66.6 3.2 0.6 7.6 9.1 17.8 11.4 11.3 8.6 11.5 3.4 1.2 336 45-49 14.4 56.8 5.1 0.2 9.1 11.6 21.0 7.8 12.1 9.4 6.8 3.4 1.1 256 Education No education 15.1 55.3 2.8 0.7 5.5 8.4 11.0 5.8 11.2 10.5 9.5 5.0 2.0 246 Some/completed primary 11.7 49.9 3.2 1.2 8.0 7.6 14.7 7.9 9.8 7.1 6.2 3.2 1.2 2,684 Some/completed secondary 8.8 50.5 2.6 0.9 6.8 6.8 13.8 6.0 9.7 7.5 6.1 3.6 0.8 775 More than secondary 6.3 79.1 2.0 0.0 15.9 12.7 27.9 11.8 19.2 16.1 15.7 5.8 0.0 61 Wealth index* Lowest 11.2 49.1 1.9 1.1 7.4 7.4 12.8 6.4 10.2 8.9 6.2 4.4 1.5 644 Second 12.8 47.7 3.2 1.1 7.0 8.6 13.8 6.7 9.6 8.2 7.4 2.5 1.2 671 Middle 13.8 50.6 3.9 1.4 7.2 7.7 15.1 7.7 9.4 5.9 5.7 2.5 1.2 723 Fourth 9.0 50.8 3.0 0.9 7.9 6.1 15.7 9.3 8.9 7.1 5.7 4.1 1.3 827 Highest 9.9 54.7 2.9 1.0 8.6 8.3 14.6 6.7 11.7 7.9 7.5 3.6 0.6 900 District Mzimba 12.0 51.3 3.6 1.1 7.4 8.3 15.4 7.8 9.3 7.5 5.9 3.8 0.9 2,589 Nsanje 12.0 51.3 1.0 1.5 6.3 5.8 8.4 4.0 7.1 8.7 4.7 2.5 3.0 488 Ntchisi 8.3 48.5 2.1 0.8 9.5 6.2 15.3 8.2 14.0 7.0 9.8 2.9 0.7 689 Total 11.2 50.8 3.0 1.1 7.7 7.6 14.5 7.4 10.0 7.5 6.5 3.4 1.1 3,766 *One household is missing wealth information Malawi ONSE Impact Evaluation: Baseline Report 135 Table A5. Percentage of WRA with a birth in the past three years who correctly identified the warning/danger signs of maternal complications during childbirth in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Age 15-19 15.6 44.2 7.4 1.1 9.7 11.9 17.5 5.6 11.6 9.4 4.3 4.3 1.9 211 20-24 14.0 56.5 5.1 1.1 8.8 10.7 17.7 4.5 11.6 9.3 4.5 2.8 1.4 556 25-29 12.6 62.4 5.5 2.1 9.9 14.8 19.3 9.1 14.2 8.8 6.5 6.4 0.8 396 30-34 13.7 67.7 4.7 0.5 11.6 12.1 17.2 8.7 15.4 13.6 6.4 4.8 0.6 312 35-39 14.0 71.4 6.0 2.1 7.2 10.0 19.7 4.0 17.5 11.9 5.2 5.3 2.2 211 40-44 10.6 70.0 7.6 2.9 13.2 12.0 22.9 12.3 10.4 7.4 6.2 8.5 0.0 79 45-49 15.4 71.0 4.9 0.0 6.6 21.5 19.4 3.5 9.1 6.0 3.2 6.4 8.6 36 Education No education 15.7 56.7 4.2 1.9 7.8 10.3 20.3 7.8 7.2 4.9 4.3 7.9 0.3 233 Some/completed primary 13.6 60.1 5.4 1.2 9.7 12.1 18.9 6.5 14.1 10.8 5.9 4.6 1.6 1,341 Some/completed secondary 11.8 69.6 8.1 2.2 10.9 14.7 12.6 6.9 13.5 10.8 3.6 3.0 0.9 210 More than secondary 16.7 80.4 11.2 0.0 16.9 19.1 28.9 2.3 38.9 16.7 4.0 0.0 0.0 17 Wealth index* Lowest 16.8 54.4 6.3 1.4 11.1 10.4 17.0 9.3 12.7 10.5 5.4 4.2 1.2 572 Second 13.6 60.9 4.8 2.2 11.1 13.0 16.9 6.7 11.2 8.0 5.3 3.1 1.2 393 Middle 15.9 60.6 4.7 0.9 6.4 9.6 24.1 6.3 9.7 10.0 7.0 6.7 2.3 330 Fourth 11.3 59.3 6.3 1.0 5.5 13.7 21.4 4.3 16.4 12.0 5.6 6.4 1.1 249 136 Malawi ONSE Impact Evaluation: Baseline Report Project Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Highest 7.1 76.6 5.8 1.3 12.4 16.7 14.1 4.1 19.8 10.6 3.3 4.6 1.1 256 District Machinga 14.3 52.6 3.8 1.1 7.6 11.8 19.0 7.7 10.8 7.8 8.9 5.9 1.7 839 Nkhotakota 9.0 72.9 7.0 1.0 13.3 14.9 13.9 5.1 20.5 11.3 4.3 4.2 0.7 502 Salima 16.3 61.2 6.6 2.0 9.2 10.8 21.1 6.8 11.2 11.7 2.5 4.1 1.4 460 Total 13.7 60.9 5.6 1.4 9.6 12.2 18.5 6.7 13.4 10.1 5.4 4.8 1.4 1,801 *One household is missing wealth information Table A6. Percentage of WRA with a birth in the past three years who correctly identified the warning/danger signs of maternal complications during childbirth in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N Age 15-19 13.1 47.7 4.8 0.9 3.5 7.3 15.0 3.2 7.0 9.5 2.3 1.9 1.5 170 20-24 14.0 51.0 2.7 2.1 9.0 7.7 14.3 8.8 10.9 8.3 7.0 3.2 1.2 529 25-29 13.0 63.1 3.0 2.6 11.0 10.2 20.3 10.7 11.8 7.5 9.5 5.0 2.5 360 30-34 11.6 64.3 2.7 1.5 11.5 7.5 14.8 7.8 16.0 7.2 6.3 3.4 1.3 282 35-39 14.9 64.2 4.3 1.4 12.1 7.5 17.1 8.2 9.4 10.0 9.7 3.4 2.1 186 Malawi ONSE Impact Evaluation: Baseline Report 137 Comparison Abdominal pain Severe bleeding Severe headache Blurred vision Convulsions High fever Labor pain 12+ hours Placenta retained Loss of conscious- ness Difficulty breathing Changes in fetal movement Swollen hands, feet or face Problems urinating N 40-44 10.1 58.5 5.2 0.0 5.5 8.1 13.1 3.7 12.3 12.2 7.5 2.1 0.7 52 45-49 0.0 55.9 16.1 3.2 9.6 11.8 25.0 6.2 0.0 11.5 4.9 15.3 0.0 18 Education No education 14.5 53.8 2.9 1.9 4.6 8.4 7.8 8.5 7.1 9.8 11.1 6.8 3.4 95 Some/completed primary 13.5 57.0 3.6 2.0 9.5 8.6 16.2 8.6 11.4 7.8 7.1 3.5 1.4 1,178 Some/completed secondary 11.4 58.6 2.8 1.5 10.0 6.5 18.3 6.2 11.7 10.0 6.8 2.5 1.9 302 More than secondary 10.0 83.0 5.2 0.0 27.5 10.1 26.2 15.7 22.1 13.7 12.2 8.7 0.0 22 Wealth index* Lowest 12.2 53.9 2.6 2.0 7.6 6.7 14.3 8.6 11.9 9.1 7.4 4.0 2.1 349 Second 13.4 53.1 2.4 1.7 9.7 9.3 13.2 7.8 9.9 8.5 10.3 1.6 1.6 308 Middle 17.1 58.0 4.6 1.6 9.6 9.3 17.9 8.7 9.2 6.7 6.5 2.4 1.9 326 Fourth 9.4 60.8 4.9 2.2 9.7 6.8 17.8 8.6 12.9 8.1 5.5 5.6 1.8 326 Highest 13.5 62.5 2.5 1.9 11.7 9.6 18.5 7.1 12.8 10.0 7.1 4.5 0.5 288 District Mzimba 14.2 58.5 4.1 2.0 9.1 9.8 17.9 8.6 10.4 8.6 6.5 3.9 1.2 1,049 Nsanje 14.2 57.7 0.3 2.1 6.5 5.4 7.8 3.3 5.9 9.0 4.5 3.5 4.9 258 Ntchisi 8.6 53.7 3.9 1.2 13.4 5.4 17.5 11.0 18.8 7.5 12.2 2.9 0.3 290 Total 13.1 57.5 3.4 1.9 9.5 8.2 16.2 8.2 11.3 8.4 7.3 3.6 1.6 1,597 138 Malawi ONSE Impact Evaluation: Baseline Report Table A7. Percentage of WRA who correctly identified the warning/danger signs of complications for newborns in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions /spasms High fever Lethargy Loss of consciousn ess Green vomit No stool passed Swollen abdomen N Age 15-19 17.7 8.1 6.8 5.2 5.5 0.6 1.9 1.8 15.3 1.4 2.1 1.5 1.0 0.7 796 20-24 29.3 17.8 17.3 14.3 6.7 1.5 4.0 2.9 31.5 5.2 4.6 1.8 2.1 2.2 769 25-29 36.9 21.0 21.6 26.9 7.8 1.2 2.7 2.7 32.0 2.7 4.5 3.3 3.2 3.6 572 30-34 39.7 20.1 24.7 21.8 9.3 1.0 4.1 4.8 33.0 4.3 7.9 2.2 1.0 2.9 552 35-39 34.8 16.4 27.0 25.2 5.5 2.3 4.4 6.5 35.7 4.4 5.8 1.1 3.8 4.4 485 40-44 38.8 18.8 26.6 30.4 7.4 1.5 5.0 4.8 32.3 5.0 6.5 2.0 2.7 3.1 336 45-49 40.0 19.2 25.3 23.6 10.8 1.2 5.1 4.6 33.9 6.8 7.6 1.0 2.3 5.9 256 Education No education 36.3 17.4 21.9 22.7 7.7 2.7 3.4 3.3 27.9 5.4 5.2 3.0 1.1 4.6 246 Some/completed primary 30.1 15.9 18.3 18.9 7.0 1.1 3.5 3.6 29.4 3.5 5.0 1.7 2.0 2.7 2,684 Some/completed secondary 35.1 19.1 21.1 14.2 6.8 1.0 4.1 3.8 29.0 3.9 5.2 1.7 4.1 1.6 775 More than secondary 38.2 20.1 32.9 21.1 15.4 0.0 3.8 9.1 30.3 4.6 2.1 3.2 0.0 2.1 61 Wealth index* Lowest 31.1 16.2 18.0 16.8 7.3 2.0 4.0 3.1 28.8 2.5 4.8 1.1 1.3 2.0 644 Second 31.8 16.2 19.1 17.4 7.9 0.6 2.8 5.1 28.5 3.5 4.8 1.5 2.9 4.0 671 Middle 31.2 15.3 18.3 22.4 5.7 1.2 4.2 3.2 29.5 4.8 6.2 1.9 1.8 3.3 723 Malawi ONSE Impact Evaluation: Baseline Report 139 Project Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions /spasms High fever Lethargy Loss of consciousn ess Green vomit No stool passed Swollen abdomen N Fourth 31.6 16.4 20.6 21.5 8.0 1.5 2.7 4.2 28.0 4.1 5.9 3.5 2.8 3.5 827 Highest 33.2 18.9 21.1 16.7 6.9 1.0 4.1 2.7 31.0 4.8 3.7 1.8 2.2 1.5 900 District Mzimba 29.0 17.2 13.6 23.1 8.8 1.6 2.0 3.3 29.6 2.6 3.7 2.0 2.2 1.7 2,589 Nsanje 31.0 19.5 24.9 22.0 7.8 0.9 5.9 3.2 25.6 4.8 3.8 1.9 2.4 1.5 488 Ntchisi 34.9 13.8 20.7 12.1 5.0 1.3 3.4 4.3 31.4 4.3 7.2 1.8 2.0 4.8 689 Total 31.7 16.6 19.3 18.7 7.2 1.3 3.6 3.6 29.1 3.9 5.0 1.9 2.2 2.8 3,766 140 Malawi ONSE Impact Evaluation: Baseline Report Table A8. Percentage of WRA who correctly identified the warning/danger signs of complications for newborns in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions /spasms High fever Lethargy Loss of conscious- ness Green vomit No stool passed Swollen abdomen N Age 15-19 11.3 6.1 6.5 6.0 6.9 0.4 1.2 1.8 11.1 1.9 2.2 1.1 1.1 1.7 796 20-24 27.1 15.1 21.3 12.3 8.3 1.9 2.9 2.3 20.4 4.4 2.5 2.0 2.4 3.1 769 25-29 29.0 22.0 25.2 18.5 9.4 3.2 2.5 6.7 31.5 6.2 4.1 3.0 5.3 2.7 572 30-34 30.8 21.1 30.4 18.3 11.5 1.9 3.2 6.4 27.4 6.2 5.3 2.0 3.3 3.4 552 35-39 34.8 18.3 31.8 19.8 10.3 2.0 4.5 6.1 21.6 5.9 3.8 2.3 3.9 4.8 485 40-44 28.0 20.8 31.4 21.3 14.7 1.9 5.0 8.9 19.0 6.7 6.4 1.5 4.1 2.4 336 45-49 30.5 15.3 31.6 17.7 13.1 3.5 2.6 9.2 17.4 4.1 9.3 3.1 2.3 6.6 256 Education No education 32.0 12.4 27.8 14.7 8.1 1.8 4.5 5.4 22.2 1.8 6.8 1.8 2.8 1.7 246 Some/completed primary 24.6 15.4 22.6 15.2 9.6 2.0 2.8 5.3 21.1 5.1 3.6 2.1 2.9 3.5 2,684 Some/completed secondary 28.4 18.9 22.2 13.5 10.8 1.7 2.6 3.8 19.0 4.1 4.3 1.6 2.8 2.4 775 More than secondary 29.2 22.8 35.8 21.4 12.5 0.9 2.3 7.3 33.7 10.2 13.8 6.3 14.4 3.0 61 Wealth index* Lowest 24.9 13.6 23.8 16.3 9.1 2.2 2.6 5.1 21.9 4.7 2.5 1.7 3.5 3.7 644 Second 22.9 12.5 23.7 13.8 9.4 2.1 2.3 4.7 23.0 4.1 3.8 2.4 2.3 3.4 671 Middle 25.4 17.4 21.4 14.6 8.2 1.9 2.6 3.6 20.4 4.1 5.0 2.3 3.0 3.9 723 Malawi ONSE Impact Evaluation: Baseline Report 141 Comparison Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions /spasms High fever Lethargy Loss of conscious- ness Green vomit No stool passed Swollen abdomen N Fourth 25.9 17.7 23.0 16.1 10.6 2.2 3.6 4.6 21.0 4.6 3.9 2.3 2.9 1.8 827 Highest 29.4 18.1 23.2 13.7 11.3 1.3 3.1 6.8 19.1 6.1 4.9 1.6 3.3 3.3 900 District Mzimba 24.6 17.9 22.5 15.1 10.8 1.7 2.9 4.4 22.3 5.1 4.0 2.2 2.8 3.4 2,589 Nsanje 26.7 11.0 18.8 13.7 7.3 1.7 5.3 1.9 20.3 4.2 4.6 2.2 2.5 1.6 488 Ntchisi 29.7 13.0 27.6 14.8 7.8 2.6 1.4 9.0 16.6 4.0 4.0 1.3 4.3 3.5 689 Total 25.9 16.1 23.0 14.9 9.8 1.9 2.9 5.0 20.9 4.8 4.1 2.0 3.0 3.2 3,766 142 Malawi ONSE Impact Evaluation: Baseline Report Table A9. Percentage of WRA with a birth in the past three years who correctly identified the warning/danger signs of complications for newborns in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions/ spasms High fever Lethargy Loss of consciousnes s Green vomit No stool passed Swollen abdomen N Age 15-19 36.3 17.5 15.6 11.5 9.9 1.6 1.1 5.6 24.9 4.2 3.0 3.3 1.1 0.6 211 20-24 31.3 19.4 19.6 18.7 6.7 1.6 5.1 3.5 34.5 4.6 5.2 1.8 3.0 2.4 556 25-29 35.9 20.6 22.1 28.3 7.6 1.0 2.6 2.8 31 1.8 3.9 3.5 3.4 4.3 396 30-34 40.1 16.9 26.5 23.1 10.0 0.9 5.6 5.8 36.4 4.1 7.6 2.5 1.0 3.1 312 35-39 34.9 17.9 26.7 26.6 4.9 1.5 5.4 6.3 38.4 2.4 7.5 1.3 4.0 4.6 211 40-44 38.6 19.2 25.5 40.2 3.3 1.8 2.0 7.8 31.5 4.5 6.6 3.8 2.2 3.3 79 45-49 27.9 16.0 29.8 23.6 23.1 3.5 8.5 0.0 39.5 2.1 3.4 1.5 0.0 4.6 36 Education No education 36.6 19.1 24.4 26.5 8.6 2.1 4.6 3.3 28.7 3.9 5.1 5.1 0.0 4.5 233 Some/completed primary 34.6 18.6 20.9 22.4 7.9 1.4 4.2 4.4 33.3 3.6 5.3 2.2 2.6 2.9 1,341 Some/completed secondary 37.1 19.2 26.1 19.2 6.4 0.7 3.6 4.4 37.2 2.0 6.5 1.8 5.2 2.6 210 More than secondary 36.6 17.5 42.6 38.8 13.2 0.0 0.0 22.3 47.2 11.2 3.8 0.0 0.0 0.0 17 Wealth index* Lowest 34.3 20.6 20.2 19.3 9.3 2.4 5.3 3.8 30.3 2.6 5.3 1.6 1.7 2.4 572 Second 33.3 16.5 22.3 20.5 8.5 0.7 3.2 6.0 33.6 4.3 4.9 2.8 4 4.1 393 Malawi ONSE Impact Evaluation: Baseline Report 143 Project Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions/ spasms High fever Lethargy Loss of consciousnes s Green vomit No stool passed Swollen abdomen N Middle 34.2 19.2 20.7 26.5 5.6 0.4 3.5 3.7 34.6 3.5 5.2 2.1 3.1 3.9 330 Fourth 38.6 17.0 24.0 25.8 8.4 1.7 3.8 5.2 29.4 3.6 6.8 4.7 2.7 3.0 249 Highest 37.7 19.5 26.1 25.5 6.3 1.2 4.3 3.6 41.5 4.4 5.0 2.3 1.1 1.7 256 District Mzimba 34.3 19.4 14 27.4 10 1.7 2.9 4.3 33 2.6 3.6 2.7 2.5 2.1 839 Nsanje 34.1 22.1 29.5 24.4 9.3 1.0 8.3 4.4 30.5 3.7 6.4 1.8 2.7 1.8 502 Ntchisi 36.8 15.7 25.6 16.6 4.6 1.3 2.4 4.7 35.5 4.3 6.5 2.9 2.4 4.9 460 Total 35.2 18.7 22.2 22.7 7.9 1.4 4.1 4.4 33.3 3.5 5.4 2.5 2.5 3.1 1,801 144 Malawi ONSE Impact Evaluation: Baseline Report Table A10. Percentage of WRA with a birth in the past three years who correctly identified the warning/danger signs of complications for newborns in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions/ spasms High fever Lethargy Loss of conscious- ness Green vomit No stool passed Swollen abdomen N Age 15-19 21.1 15.2 17.8 16.5 8.9 1.2 3.0 4.3 26.2 3.1 4.6 2.0 1.9 3.7 170 20-24 28.3 16.8 22.4 14.7 8.5 1.5 3.4 2.7 22.6 4.4 2.7 2.1 2.6 3.5 529 25-29 30.3 24.3 23.6 19.4 8.6 3.9 2.7 7.4 34.7 6.8 3.3 2.8 6.8 3.3 360 30-34 26.2 21.3 25.5 18.6 8.8 2.2 0.6 5.1 32 6.3 6.5 1.6 4.6 2 282 35-39 31.8 17.5 29.4 25.6 9.8 3.4 5.0 4.9 25.7 5.6 2.2 2.5 3.3 1.9 186 40-44 30.3 20.4 24.3 20.5 20.9 2.1 4.5 3.4 14.3 1.2 4.1 0.0 8.6 2.9 52 45-49 38.1 10.7 36.6 27.3 11.8 00 0.0 16.6 25.4 5.1 15.6 0.0 0.0 17.8 18 Education No education 37.4 8.6 28.5 14 9.2 3.2 3.4 4.1 22.9 0.0 7.6 2.2 5.4 2.8 95 Some/completed primary 27.1 17.5 22.4 18.4 9.1 2.1 2.9 4.8 28.5 5.8 3.5 2 3.5 3.2 1,178 Some/completed secondary 30.8 28.3 26.2 17.3 9.9 3.4 2.9 4.5 23.7 3.7 2.9 1.9 5.2 3.2 302 More than secondary 16.5 32.3 45.6 37.4 6.8 0.0 0.0 12.4 39.3 16.9 15.2 13.9 13.9 0.0 22 Wealth index* Lowest 26.0 15.8 23.5 20.0 8.3 2.5 2.8 3.7 28.7 4.7 2.7 1.6 5.0 4.1 349 Second 26.3 13.5 24.0 14.2 10.6 3.0 1.4 3.3 26.2 5.5 3.5 3.9 2.6 4.4 308 Malawi ONSE Impact Evaluation: Baseline Report 145 Comparison Breathing difficulty Feeding poorly Yellow skin eyes Pus or bleeding Very small baby Pallor Bleeding Convulsions/ spasms High fever Lethargy Loss of conscious- ness Green vomit No stool passed Swollen abdomen N Middle 29.9 21.5 22.6 17.0 4.9 2.3 3.4 3.9 29.1 4.7 5.6 1.6 4.2 3.4 326 Fourth 27.8 21.5 27.3 20.5 11.2 1.9 3.7 4.9 26.2 5.5 4.2 2.0 3.7 1.6 326 Highest 32.1 24.9 21.1 18.9 11.6 2 3.1 8.8 26.3 5.6 2.8 1.7 4.9 2.1 288 District Mzimba 25.6 22.4 23.8 18.2 10.7 1.9 2.8 4.8 29.1 5.2 3.3 2.2 4.0 3.2 1,049 Nsanje 29.4 11.8 20.2 16.6 6.3 2.0 6.0 2.2 25.4 5.3 6.2 2.0 2.0 2.7 258 Ntchisi 36.1 14.8 26.6 19.3 6.8 4.0 0.7 6.9 23.3 4.9 3.4 2.3 5.9 3.3 290 Total 28.3 19.2 23.8 18.2 9.2 2.4 2.9 4.8 27.4 5.2 3.8 2.2 4.1 3.2 1,597 146 Malawi ONSE Impact Evaluation: Baseline Report Table A11. Percentage of WRA who correctly identified the symptoms of malaria in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Fever Chills Headache Joint pain Poor appetite N Age 15-19 69.4 46.1 37.2 34.6 9.2 790 20-24 84.1 43.0 34.9 32.8 10.9 836 25-29 88.8 43.2 37.3 31.2 11.9 598 30-34 88.9 46.5 33.7 29.5 11.4 554 35-39 85.3 45.3 36.1 29.3 13.2 469 40-44 86.4 46.1 39.9 36.2 12.0 295 45-49 84.6 43.0 37.0 36.3 14.9 234 Education No education 83.2 43.0 34.7 31.8 11.1 480 Some/completed primary 82.7 44.8 34.3 31.1 11.2 2,684 Some/completed secondary 82.3 46.5 45.9 39.7 12.5 556 More than secondary 92.5 39.6 49.5 37.9 13.4 56 Wealth index* Lowest 82.5 43.8 35.0 32.2 11.0 917 Second 83.1 41.6 31.6 27.7 10.4 759 Middle 82.0 48.7 36.8 32.1 14.3 713 Fourth 82.6 46.0 37.6 34.6 13.5 631 Highest 83.9 44.4 40.8 36.2 8.5 752 District Machinga 77.5 59.1 32.8 27.5 7.8 1,649 Malawi ONSE Impact Evaluation: Baseline Report 147 Project Fever Chills Headache Joint pain Poor appetite N Nkhotakota 87.3 42.0 41.3 36.1 10.0 1,143 Salima 84.6 32.9 35.9 34.7 16.0 984 Total 82.8 44.7 36.3 32.5 11.4 3,776 Table A12. Percentage of WRA who correctly identified the symptoms of malaria in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Fever Chills Headache Joint pain Poor appetite N Age 15-19 61.4 46.7 35.6 29.6 5.2 796 20-24 77.0 43.5 32.2 28.5 8.0 769 25-29 84.0 43.5 32.2 26.8 7.5 572 30-34 85.0 46.7 37.5 26.0 10.0 552 35-39 82.8 45.1 31.9 26.8 9.2 485 40-44 79.0 54.8 37.0 28.1 10.9 336 45-49 75.4 56.8 31.9 33.2 9.1 256 Education No education 80.5 33.3 35.4 30.5 6.1 246 Some/completed primary 75.9 46.5 32.6 26.6 7.6 2,684 Some/completed secondary 78.2 51.6 36.8 32.0 9.1 775 More than secondary 82.3 53.9 64.3 45.7 27.9 61 Wealth index* Lowest 82.5 43.8 35.0 32.2 11.0 917 148 Malawi ONSE Impact Evaluation: Baseline Report Comparison Fever Chills Headache Joint pain Poor appetite N Second 83.1 41.6 31.6 27.7 10.4 759 Middle 82.0 48.7 36.8 32.1 14.3 713 Fourth 82.6 46.0 37.6 34.6 13.5 631 Highest 83.9 44.4 40.8 36.2 8.5 752 District Mzimba 75.3 51.9 35.6 28.6 9.4 2,589 Nsanje 73.3 37.5 46.9 32.2 7.5 488 Ntchisi 83.7 34.9 20.7 24.2 4.2 689 Total 76.7 46.8 34.1 28.2 8.1 3,766 Malawi ONSE Impact Evaluation: Baseline Report 149 Table A13. Percentage of WRA who correctly identified the signs and symptoms of pneumonia in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Fast, difficult, or noisy breathing Cough Lethargy Refusal to eat or breastfeed N Age 15-19 27.4 6.8 2.0 0.5 796 20-24 54.8 14.9 2.3 2.8 769 25-29 66.4 18.2 1.8 1.7 572 30-34 72.9 21.3 3.6 2.5 552 35-39 75.2 18.2 2.7 3.4 485 40-44 74.2 19.9 2.8 2.2 336 45-49 77.9 22.4 2.0 1.9 256 Education No education 64.4 19.2 1.8 1.6 246 Some/completed primary 57.4 14.7 2.4 2.0 2,684 Some/completed secondary 62.7 19.3 3.0 2.6 775 More than secondary 61.1 15.9 1.9 5.5 61 Wealth index* Lowest 57.7 17.2 1.5 1.6 644 Second 57.3 13.4 3.5 2.9 671 Middle 59.2 16.8 2.9 1.2 723 Fourth 63.3 17.2 2.2 2.0 827 Highest 59.2 15.2 2.1 2.7 900 District Machinga 49.8 11.4 2.1 2.4 2,589 Nkhotakota 63.2 15.8 3.3 2.5 488 Salima 65.1 20.4 2.4 1.5 689 Total 59.1 15.9 2.4 2.1 3,766 150 Malawi ONSE Impact Evaluation: Baseline Report Table A14. Percentage of WRA who correctly identified the signs and symptoms of pneumonia in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Fast, difficult, or noisy breathing Cough Lethargy Refusal to eat or breastfeed N Age 15-19 23.8 8.5 3.6 1.3 796 20-24 51.4 14.8 4.7 2.4 769 25-29 58.9 16.4 4.7 4.1 572 30-34 66.0 19.3 4.9 4.5 552 35-39 71.1 19.9 4.9 2.3 485 40-44 70.8 21.6 2.9 2.1 336 45-49 68.0 27.9 3.6 1.7 256 Education No education 66.5 24.7 6.0 4.0 246 Some/completed primary 54.1 16.6 4.3 2.7 2,684 Some/completed secondary 49.5 12.7 3.8 2.2 775 More than secondary 72.0 29.7 2.4 1.8 61 Wealth index* Lowest 59.0 16.1 3.9 2.2 917 Second 53.2 18.0 5.0 3.9 759 Middle 51.5 17.3 4.0 2.4 713 Fourth 55.1 14.6 4.8 2.2 631 Highest 52.9 16.8 3.8 2.6 752 District Mzimba 46.9 16.1 4.1 2.7 2,589 Nsanje 65.3 25.9 6.3 1.2 488 Ntchisi 72.3 12.1 3.8 3.5 689 Total 54.2 16.5 4.3 2.6 3,766 Malawi ONSE Impact Evaluation: Baseline Report 151 Table A15. Percentage of WRA who correctly identified the causes of pneumonia, by background characteristics (ONSE IE baseline, 2017) Background characteristics Project Comparison Not dressed warmly enough Household air pollution Inadequate household ventilation N Not dressed warmly enough Household air pollution Inadequate household ventilation N Age 15-19 48.2 0.1 0.7 790 38.3 0.4 0.5 796 20-24 70.6 1.1 1.6 836 61.4 0.4 1.0 769 25-29 71.9 1.3 3.1 598 68.1 0.3 1.7 572 30-34 79.6 0.7 3.2 554 68.9 0.5 0.3 552 35-39 79.8 0.5 1.6 469 72.1 0.1 0.9 485 40-44 79.8 0.3 2.6 295 69.8 0.2 1.1 336 45-49 79.8 1.0 3.6 234 69.0 0.8 0.4 256 Education No education 68.3 0.2 2.8 480 73.3 0.2 1.3 246 Some/completed primary 68.3 0.6 1.9 2,684 58.5 0.4 0.9 2,684 Some/completed secondary 77.6 1.3 2.2 556 66.0 0.4 0.8 775 More than secondary 85.9 7.7 3.0 56 83.0 0.0 0.0 61 Wealth index* Lowest 66.3 0.6 1.9 917 65.7 0.0 1.3 644 Second 67.4 0.2 1.3 759 58.3 0.1 0.6 671 152 Malawi ONSE Impact Evaluation: Baseline Report Background characteristics Project Comparison Not dressed warmly enough Household air pollution Inadequate household ventilation N Not dressed warmly enough Household air pollution Inadequate household ventilation N Middle 72.0 0.5 2.7 713 57.6 0.8 0.9 723 Fourth 70.3 0.7 3.3 631 63.0 0.3 1.0 827 Highest 74.3 1.6 1.5 752 61.9 0.5 0.5 900 District Machinga 69.6 0.4 1.8 1,649 -- -- -- -- Nkhotakota 71.1 1.5 2.2 1,143 -- -- -- -- Salima 69.3 0.4 2.2 984 -- -- -- -- Mzimba -- -- -- -- 53.3 0.4 0.6 2,589 Nsanje -- -- -- -- 83.2 0.2 1.6 488 Ntchisi -- -- -- -- 74.8 0.3 1.2 689 Total 69.9 0.7 2.1 3,776 61.3 0.4 0.9 3,766 Malawi ONSE Impact Evaluation: Baseline Report 153 Table A16. Percentage of WRA who correctly identified the signs and symptoms of diarrhea in the project domain, by background characteristics (ONSE IE baseline, 2017) Project 3 or more loose/watery stools in 24 hours Loose/watery stools for 3+ days Fast or noisy breathing Lethargy Refusal to eat or breastfeed N Age 15-19 58.6 36.7 0.3 17.1 2.5 790 20-24 69.5 38.8 0.8 17.3 4.0 836 25-29 69.6 42.7 0.3 18.1 5.4 598 30-34 72.5 43.1 1.6 17.8 6.5 554 35-39 73.4 42.7 0.9 19.8 4.8 469 40-44 67.1 46.5 0.9 21.2 5.1 295 45-49 74.3 44.5 1.2 20.6 6.5 234 Education No education 66.8 46.8 0.9 15.7 2.8 480 Some/completed primary 67.9 40.3 0.8 18.6 4.6 2,684 Some/completed secondary 70.4 39.2 0.4 19.4 6.4 556 More than secondary 78.6 47.9 5.3 14.1 7.5 56 Wealth index* Lowest 68.3 43.3 0.5 14.2 4.3 917 Second 66.2 40.1 0.9 16.7 4.1 759 Middle 69.8 41.4 1.1 20.7 4.8 713 Fourth 66.5 43.7 0.6 22.2 5.1 631 Highest 70.7 36.8 0.9 18.8 5.0 752 154 Malawi ONSE Impact Evaluation: Baseline Report Project 3 or more loose/watery stools in 24 hours Loose/watery stools for 3+ days Fast or noisy breathing Lethargy Refusal to eat or breastfeed N District Machinga 70.1 35.2 0.5 18.3 4.3 1,649 Nkhotakota 68.0 37.7 0.7 19.2 5.4 1,143 Salima 66.7 49.3 1.1 17.7 4.4 984 Total 68.3 41.1 0.8 18.3 4.6 3,776 Table A17. Percentage of WRA who correctly identified the signs and symptoms of diarrhea in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison 3 or more loose/watery stools in 24 hours Loose/watery stools for 3+ days Fast or noisy breathing Lethargy Refusal to eat or breastfeed N Age 15-19 57.3 30.2 0.2 11.9 2.3 796 20-24 62.5 38.6 1.1 13.2 3.6 769 25-29 67.2 45.9 0.9 13.8 5.1 572 30-34 69.8 40.2 1.7 10.2 4.6 552 35-39 72.2 45.8 0.1 14.2 8.3 485 40-44 69.7 43.2 1.1 13.2 7.1 336 45-49 71.4 40.4 1.3 12.3 3.2 256 Malawi ONSE Impact Evaluation: Baseline Report 155 Comparison 3 or more loose/watery stools in 24 hours Loose/watery stools for 3+ days Fast or noisy breathing Lethargy Refusal to eat or breastfeed N Education No education 71.3 34.2 1.1 9.6 5.7 246 Some/completed primary 65.3 39.7 0.7 11.9 4.4 2,684 Some/completed secondary 64.9 40.4 0.8 15.4 4.7 775 More than secondary 71.6 45.6 6.0 25.3 6.6 61 Wealth index* Lowest 63.5 41.9 0.7 12.1 4.8 644 Second 65.1 37.6 0.2 13.2 3.9 671 Middle 66.5 38.3 0.8 11.1 4.2 723 Fourth 63.1 40.5 1.2 10.9 5.1 827 Highest 69.4 39.8 1.2 15.6 4.8 900 District Mzimba 2,589 40.5 0.9 12.2 4.5 2,589 Nsanje 488 36.5 1.0 12.9 3.3 488 Ntchisi 689 38.3 0.7 14.1 5.7 689 Total 3,766 39.6 0.8 12.7 4.6 3,766 156 Malawi ONSE Impact Evaluation: Baseline Report Table A18. Percentage of WRA who correctly identified the causes of diarrhea in the project domain, by background characteristics (ONSE IE baseline, 2017) Project Lack of safe drinking water Defecating/ urinating in open spaces Eating rotten food Lack of food protection against contamination Touching food without washing hands with soap Not washing hands after defecation N Age 15-19 34.4 3.0 19.4 31.0 21.8 22.8 790 20-24 38.7 3.5 16.8 33.1 20.2 19.3 836 25-29 38.4 5.1 18.3 39.1 24.7 24.8 598 30-34 42.9 4.7 18.3 36.8 23.2 21.0 554 35-39 43.2 6.0 19.4 35.6 22.1 20.4 469 40-44 39.1 6.7 22.9 40.6 24.6 18.1 295 45-49 52.3 4.6 21.0 36.6 21.0 27.7 234 Education No education 38.3 4.8 14.7 30.4 21.3 18.0 480 Some/completed primary 38.3 4.4 18.4 33.2 21.2 21.8 2,684 Some/completed secondary 47.5 4.2 22.6 47.9 27.5 24.4 556 More than secondary 51.8 6.7 43.1 54.8 38.1 24.7 56 Wealth index* Lowest 39.4 4.4 16.6 28.3 17.8 20.6 917 Second 39.4 5.3 17.2 30.0 19.4 20.6 759 Middle 38.4 5.4 21.1 37.6 23.4 23.0 713 Fourth 39.1 4.0 19.1 39.9 26.8 22.0 631 Highest 42.6 3.5 21.2 42.7 25.5 22.9 752 District Machinga 35.7 5.2 15.1 35.5 23.6 21.2 1,649 Nkhotakota 49.2 3.7 18.9 32.6 19.2 19.8 1,143 Salima 36.7 4.3 22.5 37.1 23.4 23.7 984 Total 39.8 4.5 18.9 35.3 22.3 21.7 3,776 Malawi ONSE Impact Evaluation: Baseline Report 157 Table A19. Percentage of WRA who correctly identified the causes of diarrhea in the comparison domain, by background characteristics (ONSE IE baseline, 2017) Comparison Lack of safe drinking water Defecating/ urinating in open spaces Eating rotten food Lack of food protection against contamination Touching food without washing hands with soap Not washing hands after defecation N Age 15-19 26.9 2.4 17.2 29.1 17.0 19.5 796 20-24 32.0 3.0 16.2 33.0 18.8 22.3 769 25-29 36.6 3.0 21.0 40.1 24.2 23.3 572 30-34 38.7 3.2 19.7 43.4 24.2 28.8 552 35-39 42.4 4.0 23.3 42.6 19.1 24.0 485 40-44 35.0 4.8 23.1 44.8 22.1 25.3 336 45-49 40.0 7.1 24.1 47.5 19.2 20.7 256 Education No education 34.3 4.6 17.0 26.9 11.1 17.8 246 Some/completed primary 33.2 3.2 19.4 36.3 20.1 22.4 2,684 Some/completed secondary 38.2 4.5 21.5 46.5 22.5 26.2 775 More than secondary 68.1 0.0 22.3 54.6 42.6 44.5 61 Wealth index* Lowest 34.0 3.3 18.5 34.0 16.0 17.5 644 Second 29.9 3.2 18.7 33.4 15.8 19.5 671 Middle 33.0 4.1 18.5 33.9 22.4 23.1 723 Fourth 34.5 2.9 20.2 40.5 21.6 25.4 827 Highest 40.9 3.8 21.9 46.2 24.4 28.6 900 District Mzimba 35.5 3.4 20.5 39.0 22.0 24.0 2,589 Nsanje 34.5 5.8 19.5 32.7 17.5 19.0 488 Ntchisi 32.3 2.5 16.9 38.0 16.4 22.9 689 Total 34.7 3.5 19.7 38.0 20.4 23.2 3,766 158 Malawi ONSE Impact Evaluation: Baseline Report Table A20. Characteristics of sampled health facilities (ONSE IE SARA survey, 2017) Facility type Percentage distribution of surveyed facilities Number of facilities surveyed Machinga Hospital 4.8 1 Health center 81.0 17 Dispensary 14.3 3 Total 100.0 21 Nkhotakota Hospital 20.0 4 Health center 70.0 14 Dispensary 10.0 2 Total 100.0 20 Salima Hospital 7.1 1 Health center 92.9 13 Dispensary 0.0 0 Total 100.0 14 Mzimba Hospital 14.0 8 Health center 77.2 44 Dispensary 8.8 5 Total 100.0 57 Nsanje Hospital 21.4 3 Health center 78.6 11 Dispensary 0.0 0 Total 100.0 14 Ntchisi Hospital 7.7 1 Health center 84.6 11 Dispensary 7.7 1 Total 100.0 13 Total 100.0 139 Malawi ONSE Impact Evaluation: Baseline Report 159 Table A21. Availability of basic amenities for client services (ONSE IE SARA survey, 2017) Facility type Regular electricity Improved water source Visual and auditory privacy Client latrine Communication equipment Computer with Internet Emergency transport N Machinga Hospital 100.0 0.0 100.0 100.0 100.0 100.0 100.0 1 Health center 47.1 5.9 100.0 88.2 76.5 11.8 11.8 17 Dispensary 66.7 66.7 66.7 100.0 100.0 0.0 0.0 3 Total 52.4 14.3 95.2 90.5 81.0 14.3 14.3 21 Nkhotakota Hospital 75.0 0.0 100.0 100.0 75.0 75.0 75.0 4 Health center 92.9 21.4 92.9 92.9 85.7 14.3 0.0 14 Dispensary 50.0 50.0 100.0 50.0 50.0 0.0 50.0 2 Total 85.0 20.0 95.0 90.0 80.0 25.0 20.0 20 Salima Hospital 100.0 0.0 100.0 100.0 100.0 100.0 100.0 1 Health center 76.9 23.1 92.3 100.0 30.8 7.7 15.4 13 Total 78.6 21.4 92.9 100.0 35.7 14.3 21.4 14 Mzimba Hospital 87.5 12.5 87.5 87.5 87.5 62.5 75.0 8 Health center 59.1 43.2 100.0 90.9 45.5 2.3 6.8 44 Dispensary 100.0 60.0 100.0 100.0 0.0 20.0 20.0 5 Total 66.7 40.4 98.2 91.2 47.4 12.3 17.5 57 Nsanje Hospital 100.0 0.0 66.7 100.0 100.0 66.7 100.0 3 Health center 72.7 18.2 90.9 100.0 45.5 9.1 0.0 11 Total 78.6 14.3 85.7 100.0 57.1 21.4 21.4 14 Ntchisi Hospital 100.0 0.0 100.0 100.0 100.0 100.0 100.0 1 Health center 63.6 18.2 100.0 100.0 81.8 9.1 18.2 11 160 Malawi ONSE Impact Evaluation: Baseline Report Facility type Regular electricity Improved water source Visual and auditory privacy Client latrine Communication equipment Computer with Internet Emergency transport N Dispensary 100.0 0.0 100.0 100.0 0.0 0.0 0.0 1 Total 69.2 15.4 100.0 100.0 76.9 15.4 23.1 13 Total 69.8 26.6 95.7 93.5 59.7 15.8 18.7 139 Table A22. Percentage of facilities with priority medicines for mothers (ONSE IE SARA survey, 2017) Facility type Oxytocin Sodium chloride inj. solution Calcium gluconate inj. Magnesium sulphate inj. Ampicillin powder (inj.) Gentamicin injection Metronidazole injection Misoprostol Azithromycin Cefixime Benzathine benzylpenicillin powder Betamethasone/ dexamethasone injection Nifedipine Hydralazine injection Methyldopa tablet N Machinga Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 0.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 1 Health center 94.1 76.5 - 88.2 0.0 100.0 - - - - 100.0 23.5 - - - 17 Dispensary 33.3 33.3 - 0.0 0.0 33.3 - - - - 33.3 0.0 - - - 3 Total 85.7 71.4 100.0 76.2 4.8 90.5 100.0 0.0 100.0 0.0 90.5 23.8 100.0 100.0 100.0 21 Salima Hospital 100.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 100.0 0.0 100.0 100.0 0.0 100.0 100.0 1 Health center 92.3 61.5 0.0 84.6 30.8 100.0 - - - - 100.0 15.4 - - - 13 Total 92.9 64.3 0.0 85.7 35.7 100.0 100.0 100.0 100.0 0.0 100.0 21.4 0.0 100.0 100.0 14 Nkhotakota Hospital 100.0 100.0 25.0 100.0 50.0 100.0 75.0 50.0 50.0 0.0 100.0 75.0 75.0 75.0 75.0 4 Health center 85.7 71.4 - 85.7 7.1 100.0 - - - - 100.0 42.9 - - - 14 Dispensary 50.0 100.0 - 50.0 0.0 100.0 - - - - 100.0 50.0 - - - 2 Total 85.0 80.0 25.0 85.0 15.0 100.0 75.0 50.0 50.0 0.0 100.0 50.0 75.0 75.0 75.0 20 Malawi ONSE Impact Evaluation: Baseline Report 161 Facility type Oxytocin Sodium chloride inj. solution Calcium gluconate inj. Magnesium sulphate inj. Ampicillin powder (inj.) Gentamicin injection Metronidazole injection Misoprostol Azithromycin Cefixime Benzathine benzylpenicillin powder Betamethasone/ dexamethasone injection Nifedipine Hydralazine injection Methyldopa tablet N Mzimba Hospital 87.5 75.0 37.5 87.5 37.5 87.5 62.5 37.5 50.0 12.5 87.5 62.5 62.5 75.0 75.0 8 Health center 95.5 75.0 - 59.1 4.6 97.7 - - - - 100.0 15.9 - - - 44 Dispensary 20.0 40.0 - 0.0 0.0 80.0 - - - - 80.0 20.0 - - - 5 Total 87.7 71.9 37.5 57.9 8.8 94.7 62.5 37.5 50.0 12.5 96.5 22.8 62.5 75.0 75.0 57 Nsanje Hospital 100.0 66.7 0.0 100.0 100.0 100.0 100.0 66.7 100.0 33.3 100.0 100.0 100.0 100.0 100.0 3 Health center 100.0 45.5 0.0 81.8 0.0 90.9 - - - - 100.0 0.0 - - - 11 Total 100.0 50.0 0.0 85.7 21.4 92.9 100.0 66.7 100.0 33.3 100.0 21.4 100.0 100.0 100.0 14 Ntchisi Hospital 100.0 0.0 0.0 100.0 100.0 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 1 Health center 90.9 81.8 - 81.8 0.0 100.0 - - - - 100.0 9.1 - - - 11 Dispensary 0.0 0.0 - 0.0 0.0 100.0 - - - - 100.0 15.4 - - - 1 Total 84.6 69.2 0.0 76.9 7.7 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 13 Total 88.5 69.8 27.8 71.9 13.0 95.7 77.8 50.0 66.7 11.1 97.1 25.9 72.2 83.3 83.3 139 162 Malawi ONSE Impact Evaluation: Baseline Report Table A23. Percentage of facilities with priority medicines for children (ONSE IE SARA survey, 2017) Facility type Amoxicillin Ampicillin powder for injection Ceftriaxone powder for injection Gentamicin injection Procaine benzylpenicillin powder for inj. ORS Zinc sulphate ACT Artesunate rectal or inj. forms Vitamin A Morphine granule, injection Paracetamol N Machinga Hospital 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 100.0 0.0 1 Health center 35.3 0.0 - 100.0 58.8 35.3 - 94.1 88.2 41.2 - 41.2 17 Dispensary 0.0 0.0 - 33.3 33.3 33.3 - 100.0 100.0 33.3 - 0.0 3 Total 33.3 4.8 100.0 90.5 52.4 38.1 100.0 95.2 90.5 42.9 100.0 33.3 21 Nkhotakota Hospital 75.0 50.0 75.0 100.0 0.0 100.0 50.0 100.0 100.0 50.0 25.0 50.0 4 Health center 42.9 7.1 - 100.0 35.7 85.7 - 92.9 100.0 50.0 - 14.3 14 Dispensary 0.0 0.0 - 100.0 50.0 100.0 - 100.0 100.0 50.0 - 0.0 2 Total 45.0 15.0 75.0 100.0 30.0 90.0 50.0 95.0 100.0 50.0 25.0 20.0 20 Salima Hospital 0.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 100.0 0.0 1 Health center 46.2 30.8 - 100.0 46.2 92.3 - 92.3 92.3 30.8 - 38.5 13 Total 42.9 35.7 100.0 100.0 42.9 92.9 100.0 92.9 92.9 28.6 100.0 35.7 14 Mzimba Hospital 50.0 37.5 75.0 87.5 12.5 87.5 75.0 75.0 87.5 50.0 37.5 50.0 8 Health center 61.4 4.6 - 97.7 45.5 79.6 - 88.6 93.2 79.6 - 18.2 44 Malawi ONSE Impact Evaluation: Baseline Report 163 Facility type Amoxicillin Ampicillin powder for injection Ceftriaxone powder for injection Gentamicin injection Procaine benzylpenicillin powder for inj. ORS Zinc sulphate ACT Artesunate rectal or inj. forms Vitamin A Morphine granule, injection Paracetamol N Dispensary 80.0 0.0 - 80.0 40.0 100.0 - 100.0 80.0 60.0 - 60.0 5 Total 61.4 8.8 75.0 94.7 40.4 82.5 75.0 87.7 91.2 73.7 37.5 26.3 57 Nsanje Hospital 100.0 100.0 100.0 100.0 66.7 66.7 66.7 100.0 100.0 0.0 33.3 100.0 3 Health center 63.6 0.0 - 90.9 54.6 90.9 - 100.0 100.0 54.6 - 54.6 11 Total 71.4 21.4 100.0 92.9 57.1 85.7 66.7 100.0 100.0 42.9 33.3 64.3 14 Ntchisi Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 54.6 0.0 - 100.0 63.6 90.9 - 90.9 100.0 63.6 - 36.4 11 Dispensary 0.0 0.0 - 100.0 0.0 100.0 - 100.0 100.0 0.0 - 0.0 1 Total 53.9 7.7 100.0 100.0 61.5 92.3 100.0 92.3 100.0 61.5 100.0 38.5 13 Total 53.2 13.0 83.3 95.7 44.6 79.1 72.2 92.1 94.2 56.8 44.4 32.4 139 164 Malawi ONSE Impact Evaluation: Baseline Report Table A24. Availability of FP methods among facilities that provided FP services (ONSE IE SARA survey, 2017) Facility type Provision of the following modern methods: Any modern method Number of Pills facilities Injec￾tables Female condoms Male condoms IUD Implant Cycle beads Male steriliza￾tion Female steriliza￾tion Machinga Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 100.0 93.3 100.0 100.0 26.7 93.3 86.7 6.7 26.7 100.0 15 Dispensary 100.0 100.0 100.0 100.0 - 100.0 33.3 - - 100.0 3 Total 100.0 94.7 100.0 100.0 31.3 94.7 79.0 12.5 31.3 100.0 19 Nkhotakota Hospital 75.0 100.0 100.0 100.0 25.0 75.0 75.0 50.0 50.0 100.0 4 Health center 100.0 100.0 85.7 100.0 57.1 100.0 57.1 35.7 50.0 100.0 14 Dispensary 100.0 100.0 100.0 100.0 - 100.0 50.0 - - 100.0 2 Total 95.0 100.0 90.0 100.0 50.0 95.0 60.0 38.9 50.0 100.0 20 Salima Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 91.7 100.0 83.3 100.0 8.3 66.7 50.0 0.0 25.0 100.0 12 Total 92.3 100.0 84.6 100.0 15.4 69.2 53.9 7.7 30.8 100.0 13 Mzimba Hospital 100.0 100.0 75.0 100.0 75.0 100.0 75.0 100.0 100.0 100.0 4 Health center 86.4 100.0 90.9 95.5 45.5 100.0 54.6 11.4 25.0 100.0 44 Dispensary 75.0 100.0 75.0 100.0 - 50.0 0.0 - 100.0 4 Total 86.5 100.0 88.5 96.2 47.9 96.2 51.9 18.8 31.3 100.0 52 Nsanje Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 90.9 90.9 72.7 100.0 9.1 100.0 54.6 9.1 9.1 100.0 11 Total 91.7 91.7 75.0 100.0 16.7 100.0 58.3 16.7 16.7 100.0 12 Ntchisi Hospital 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Malawi ONSE Impact Evaluation: Baseline Report 165 Facility type Provision of the following modern methods: Any modern method Number of Pills facilities Injec￾tables Female condoms Male condoms IUD Implant Cycle beads Male steriliza￾tion Female steriliza￾tion Health center 100.0 100.0 100.0 90.0 20.0 80.0 40.0 10.0 40.0 100.0 10 Dispensary 0.0 100.0 0.0 100.0 - 0.0 0.0 - - 100.0 1 Total 91.7 100.0 91.7 91.7 27.3 75.0 41.7 18.2 45.5 100.0 12 Total 91.4 98.4 89.1 97.7 37.3 91.4 57.0 19.5 33.9 100.0 128 Table A25. Availability of maternal health services (ONSE IE SARA survey, 2017) Facility type Percentage of facilities that offered: Number of facilities Provider of delivery care available onsite or on-call 24 hours per day Number of facilities offering normal ANC delivery services Normal delivery service Caesarean delivery ANC and normal delivery service ANC, normal delivery, and Caesarean delivery Machinga Hospital 100.0 100.0 100.0 100.0 100.0 1 100.0 1 Health center 100.0 94.1 - 94.1 - 17 0.0 16 Dispensary 33.3 - - - - 3 - 0 Total 90.5 94.4 100.0 94.4 100.0 21 5.9 17 Nkhotakota Hospital 100.0 100.0 50.0 100.0 50.0 4 50.0 4 Health center 100.0 92.9 - 92.9 - 14 0.0 13 Dispensary 50.0 - - - - 2 - 0 Total 95.0 94.4 50.0 94.4 50.0 20 11.8 17 166 Malawi ONSE Impact Evaluation: Baseline Report Facility type Percentage of facilities that offered: Number of facilities Provider of delivery care available onsite or on-call 24 hours per day Number of facilities offering normal ANC delivery services Normal delivery service Caesarean delivery ANC and normal delivery service ANC, normal delivery, and Caesarean delivery Salima Hospital 100.0 100.0 100.0 100.0 100.0 1 100.0 1 Health center 100.0 100.0 - 100.0 - 13 0.0 13 Total 100.0 100.0 100.0 100.0 100.0 14 7.1 14 Mzimba Hospital 87.5 87.5 50.0 87.5 50.0 8 57.1 7 Health center 100.0 95.5 - 95.5 - 44 0.0 42 Dispensary 20.0 - - - - 5 - 0 Total 91.2 94.2 50.0 94.2 50.0 57 8.2 49 Nsanje Hospital 100.0 100.0 100.0 100.0 100.0 3 100.0 3 Health center 100.0 100.0 - 100.0 -- 11 0.0 11 Total 100.0 100.0 100.0 100.0 100.0 14 21.4 14 Ntchisi Hospital 100.0 100.0 100.0 100.0 100.0 1 100.0 1 Health center 90.9 90.9 - 90.9 - 11 0.0 10 Dispensary 0.0 - - - - 1 - 0 Total 84.6 91.7 100.0 91.7 100.0 13 9.1 11 Total 92.8 95.3 66.7 95.3 66.7 139 9.8 122 Malawi ONSE Impact Evaluation: Baseline Report 167 Table A26. Availability of guidelines, trained staff, and equipment for delivery services (ONSE IE SARA survey, 2017) Facility type Percentage of facilities offering normal delivery service that had: Number of facilities offering delivery services Guide￾lines on IMPAC Staff trained in IMPAC Equipment Emer￾gency transport Exam light Delivery pack Suction apparatus (mucus extractor) Manual vacuum extractor Vacuum aspirator or D&C kit Neonatal bag and mask Parto￾graph Gloves Machinga Hospital 0.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 12.5 50.0 12.5 50.0 87.5 100.0 37.5 25.0 81.3 93.8 100.0 16 Total 11.8 52.9 17.6 47.1 88.2 100.0 41.2 29.4 82.4 94.1 100.0 17 Nkhotakota Hospital 75.0 50.0 75.0 50.0 100.0 50.0 100.0 50.0 100.0 100.0 100.0 4 Health center 30.8 30.8 0.0 38.5 92.3 76.9 23.1 15.4 100.0 100.0 100.0 13 Total 41.2 35.3 17.6 41.2 94.1 70.6 41.2 23.5 100.0 100.0 100.0 17 Salima Hospital 0.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1 Health center 38.5 46.2 15.4 30.8 84.6 84.6 15.4 7.7 69.2 61.5 100.0 13 Total 35.7 50.0 21.4 35.7 85.7 85.7 21.4 14.3 71.4 64.3 100.0 14 Mzimba Hospital 57.1 100.0 85.7 42.9 100.0 100.0 100.0 85.7 85.7 100.0 100.0 7 Health center 40.5 35.7 4.8 45.2 71.4 92.9 16.7 11.9 92.9 90.5 100.0 42 Total 42.9 44.9 16.3 44.9 75.5 93.9 28.6 22.5 91.8 91.8 100.0 49 Nsanje Hospital 100.0 100.0 100.0 100.0 100.0 100.0 66.7 0.0 100.0 100.0 100.0 3 Health center 63.6 45.5 0.0 54.6 100.0 100.0 9.1 9.1 100.0 100.0 100.0 11 168 Malawi ONSE Impact Evaluation: Baseline Report Facility type Percentage of facilities offering normal delivery service that had: Number of facilities offering delivery services Guide￾lines on IMPAC Staff trained in IMPAC Equipment Emer￾gency transport Exam light Delivery pack Suction apparatus (mucus extractor) Manual vacuum extractor Vacuum aspirator or D&C kit Neonatal bag and mask Parto￾graph Gloves Total 71.4 57.1 21.4 64.3 100.0 100.0 21.4 7.1 100.0 100.0 100.0 14 Ntchisi Hospital 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 1 Health center 30.0 60.0 10.0 20.0 60.0 90.0 10.0 10.0 100.0 100.0 100.0 10 Total 36.4 63.6 18.2 18.2 63.6 90.9 18.2 9.1 100.0 100.0 100.0 11 Total 40.2 48.4 18.0 43.4 82.8 91.0 29.5 19.7 91.0 91.8 100.0 122 Malawi ONSE Impact Evaluation: Baseline Report 169 Table A27. Availability of malaria services at facilities offering ANC (ONSE IE SARA survey, 2017) Facility type Percentage of ANC facilities offering malaria services Staff trained in IPT Medicines and commodities Number of facilities offering IPTp ITNs ANC Machinga Hospital 100.0 0.0 0.0 100.0 1 Health center 100.0 47.1 17.7 82.4 17 Dispensary 100.0 0.0 0.0 0.0 1 Total 100.0 42.1 15.8 78.9 19 Nkhotakota Hospital 100.0 50.0 50.0 75.0 4 Health center 100.0 50.0 64.3 50.0 14 Dispensary 100.0 0.0 0.0 100.0 1 Total 100.0 47.4 57.9 57.9 19 Salima Hospital 100.0 0.0 100.0 100.0 1 Health center 100.0 53.9 76.9 69.2 13 Total 100.0 50.0 78.6 71.4 14 Mzimba Hospital 100.0 71.4 85.7 100.0 7 Health center 100.0 15.9 68.2 81.8 44 Dispensary 100.0 0.0 100.0 0.0 1 Total 100.0 23.1 71.2 82.7 52 Nsanje Hospital 100.0 66.7 66.7 100.0 3 Health center 100.0 27.3 27.3 90.9 11 Total 100.0 35.7 35.7 92.9 14 Ntchisi Hospital 100.0 0.0 100.0 100.0 1 Health center 100.0 40.0 30.0 90.0 10 Total 100.0 36.4 36.4 90.9 11 Total 100.0 34.9 55.0 79.1 129 170 Malawi ONSE Impact Evaluation: Baseline Report Table A28. Availability of malaria services and guidelines, trained staff, and diagnostic capacity (ONSE IE SARA survey, 2017) Facility type Percentage of facilities offering malaria diagnosis/ treatment Number of facilities Among facilities that offered malaria diagnosis/treatment services: Number of facilities that offered malaria diagnosis/ treatment Guidelines Trained staff Diagnostics Guidelines for diagnosis/ treatment of malaria Staff trained in malaria diagnosis/ treatment Malaria RDT Malaria microscopy Any malaria diagnostics Machinga Hospital 100.0 1 100.0 100.0 100.0 100.0 100.0 1 Health center 100.0 17 35.3 82.4 94.1 41.2 94.1 17 Dispensary 100.0 3 66.7 100.0 100.0 -- 100.0 3 Total 100.0 21 42.9 85.7 95.2 44.4 95.2 21 Nkhotakota Hospital 100.0 4 25.0 75.0 100.0 75.0 100.0 4 Health center 100.0 14 42.9 92.9 100.0 0.0 100.0 14 Dispensary 100.0 2 0.0 50.0 100.0 -- 100.0 2 Total 100.0 20 35.0 85.0 100.0 16.7 100.0 20 Salima Hospital 100.0 1 100.0 100.0 100.0 100.0 100.0 1 Health center 100.0 13 38.5 76.9 100.0 23.1 100.0 13 Total 100.0 14 42.9 78.6 100.0 28.6 100.0 14 Mzimba Hospital 87.5 7 28.6 85.7 100.0 85.7 100.0 7 Health center 100.0 44 15.9 68.2 100.0 6.8 100.0 44 Dispensary 80.0 4 25.0 75.0 100.0 -- 100.0 4 Malawi ONSE Impact Evaluation: Baseline Report 171 Facility type Percentage of facilities offering malaria diagnosis/ treatment Number of facilities Among facilities that offered malaria diagnosis/treatment services: Number of facilities that offered malaria diagnosis/ treatment Guidelines Trained staff Diagnostics Guidelines for diagnosis/ treatment of malaria Staff trained in malaria diagnosis/ treatment Malaria RDT Malaria microscopy Any malaria diagnostics Total 96.5 55 18.2 70.9 100.0 6 100.0 55 Nsanje Hospital 100.0 3 66.7 100.0 100.0 100.0 100.0 3 Health center 100.0 11 18.2 36.4 100.0 0.0 100.0 11 Total 100.0 14 28.6 50.0 100.0 21.4 100.0 14 Ntchisi Hospital 100.0 1 0.0 100.0 100.0 100.0 100.0 1 Health center 100.0 11 27.3 63.6 100.0 9.1 100.0 11 Dispensary 100.0 1 0.0 0.0 100.0 -- 100.0 1 Total 100.0 13 23.1 61.5 100.0 16.7 100.0 13 Total 98.6 139 28.5 73.0 99.3 21.2 99.3 137 172 Malawi ONSE Impact Evaluation: Baseline Report Table A29. Availability of malaria medicines (ONSE IE SARA survey, 2017) Facility type ACTs SP Oral quinine Number of facilities that offered malaria diagnosis and/or treatment services Machinga Hospital 100.0 0.0 0.0 1 Health center 94.1 17.7 17.7 17 Dispensary 100.0 0.0 0.0 3 Total 95.2 14.3 14.3 21 Nkhotakota Hospital 100.0 50.0 50.0 4 Health center 92.9 64.3 35.7 14 Dispensary 100.0 0.0 50.0 2 Total 95.0 55.0 40.0 20 Salima Hospital 100.0 100.0 100.0 1 Health center 92.3 76.9 61.5 13 Total 92.9 78.6 64.3 14 Mzimba Hospital 85.7 85.7 57.1 7 Health center 88.6 68.2 22.7 44 Dispensary 100.0 25.0 50.0 4 Total 89.1 67.3 29.1 55 Nsanje Hospital 100.0 66.7 66.7 3 Health center 100.0 27.3 27.3 11 Total 100.0 35.7 35.7 14 Ntchisi Hospital 100.0 100.0 100.0 1 Health center 90.9 27.3 18.2 11 Dispensary 100.0 100.0 0.0 1 Total 92.3 38.5 23.1 13 Total 92.7 52.6 32.1 137 Malawi ONSE Impact Evaluation: Baseline Report 173 APPENDIX B. KEY INDICATOR DATA SYNTHESIS Malawi is rich with survey data sources that gather information on key FP/RH and MNCH outcomes. USAID/Malawi requested a compilation and synthesis of the estimates from different sources. This synthesis draws on data from several different sources. A brief summary of secondary sources of information is provided below. Data Sources District Health Information System, version 2 (DHIS 2) A paper-based system is used in Malawi for recording data at health facilities and for reporting data from the facility to the district level. However, at the district level, the data are electronically captured into the DHIS 2. DHIS 2 contains data from all public and faith-based facilities (i.e., CHAM). DHIS 2 indicators are commonly used for performance monitoring, and several relevant indicators are presented from DHIS 2 that were derived from SSDI reports. Malawi DHS, 2010 and 2015−2016 The DHS is a repeated cross-sectional survey conducted in many countries around the world with an average of five-year intervals between surveys. DHS surveys are weighted to produce national-level estimates and, for some key indicators, are also designed to produce district-level estimates. The 2010 Malawi DHS was implemented by the National Statistics Office which conducted interviews in 24,825 households and with 23,020 WRA. The 2015-16 Malawi DHS included interviews with 26,361 households and 24,562 WRA. Malawi Malaria Indicator Survey (MIS), 2014 and 2017 The 2014 and 2017 MIS were conducted by the National Malaria Control Programme, with funding from The Global Fund and the President’s Malaria Initiative, and support from ICF International. The two Malawi MIS are nationally representative surveys conducted to estimate malaria program coverage and to estimate malaria-related burden and anemia prevalence testing for children under five. Malawi MDG Endline Survey (MES), 2014 The MES was conducted in 2014 by the National Statistics Office with support from UNICEF and the MDG as a part of the Multiple Indicator Cluster Surveys (MICS) program. MICS collect information on the status of women and children for policy use, research, and for measurement against the MDGs. The goal of this MES was to produce end line estimates for indicators that show progress of the attainment of the MDGs. Many indicators collected for the MDG end line in the MES are relevant for the ONSE project and impact evaluation. 174 Malawi ONSE Impact Evaluation: Baseline Report Malawi Service Provision Assessment (SPA), 2013−2014 The Malawi SPA provides national and subnational information on the availability and quality of and readiness to provide services in all hospitals, health centers, dispensaries, maternities, clinics, and health posts in the country. Facilities from both public and private managing authorities were included, specifically government, CHAM, nongovernmental organizations, and private and faith-based organizations. The 2013‒ 14 SPA reports on child health, FP, maternal and newborn health care (antenatal and delivery care), sexually transmitted infections, tuberculosis, and HIV/AIDS services, and also included interviews with health providers and clients. Observations of provider-client interactions were also conducted. Support for Service Delivery Integration-Communications (SSDI-C), 2012 and 2016 The SSDI-C project conducted a cross-sectional baseline survey in 2012. This survey used a stratified random sample in 15 project and four control districts and interviewed 1,134 women and 1,099 men. The purpose of the survey was to generate baseline estimates for health outcomes and baseline indicator values for selected health practices, such as knowledge, self-efficacy, risk perceptions, and social normative perceptions. Results of interviews with females are used in this analysis to maintain comparability with the ONSE baseline survey and DHS woman’s questionnaires from which many indicators are drawn. SSDI-C also conducted a cross-sectional end line survey in 2016. The purpose of the end line survey was to estimate end line indicators and measures of change in selected health outcomes compared with outcomes at the beginning of the project. It had a similar design as the baseline survey with a stratified random sampling approach. 1,223 women were interviewed in the 15 intervention districts, which are included in the estimates below. Support for Service Delivery Integration-Services (SSDI-S), 2016 The SSDI-S Endline Assessment Report provides results from facilities in the 15 SSDI-supported districts and five comparison districts from the Rapid Situational Analysis Questionnaire and a Client Exit Interview Questionnaire. The end line assessment collected data from 9 percent of public and CHAM hospitals and health centers in the survey’s 20 districts to gather information on targeted integrated facility outcomes, such as the availability of supplies and commodities, the capacity of service providers, and clients’ perception of services. District-level estimates are not available in the report. Key Outcome Areas Key outcome areas for the ONSE project and IE are FP, maternal health, child health, water and sanitation, and malaria. Indicator summaries are listed below by outcome area. Malaria is a cross-cutting area that falls in maternal and child health. Malaria-related indicators are placed in those key outcome areas. Malawi ONSE Impact Evaluation: Baseline Report 175 Family Planning Modern contraceptive prevalence rates (MCPRs) for Malawi are available from several sources over time, including the Malawi DHS, SSDI end line survey, and ONSE IE baseline survey. Figure B1 shows the trend in the MCPR between 2010 and 2017. The trend is increasing over time through the 2015‒16 Malawi DHS. The ONSE IE baseline MCPR estimate is lower than the latest Malawi DHS estimate (56 percent versus 58 percent, respectively). Reviewing the district-level estimates from the 2015‒16 Malawi DHS reveals that the MCPRs in ONSE districts (46 percent in Machinga, 51 percent in Nkhotakota, and 53 percent in Salima) are much lower than the average Malawi DHS MCPR, providing some explanation for why the ONSE IE baseline MCPR aggregate estimate for those districts is lower than the Malawi DHS national-level estimate. The trend in couple years of protection during the SSDI was also increasing, which is in line with an increasing MCPR (Figure B2). Table B1 provides a comparison of FP method type from the Malawi DHS 2015‒16 and the ONSE IE baseline districts. The percentage of use by method type is very similar in these two sources. Figure B1. MCPR trend, 2010-2017 42.0 45.0 58.1 55.5 0 20 40 60 80 100 2010 DHS 2015 SSDI endline 2015-16 DHS 2017 ONSE IE 176 Malawi ONSE Impact Evaluation: Baseline Report Figure B2. Couple years of protection in SSDI districts, April 2012−March 2016 Source: SSDI Endline Report Source: SSDI Endline Report Malawi ONSE Impact Evaluation: Baseline Report 177 Table B1. Contraceptive method use, by type, 2015−2016 Malawi DHS and ONSE IE baseline, 2017 FP method Malawi DHS 2015–2016 ONSE IE baseline 2017 Any modern method 58.1 55.5 Injectables 51.6 51.4 Implants 19.8 18.6 Female sterilization 18.8 15.1 Male condom 3.3 5.2 Pill 4.1 2.7 IUD 1.9 1.6 Male sterilization 0.2 0.2 Other modern method 0.3 0.1 Total 100.0 100.0 N 16,130 2,522 Maternal Health Services Table B2 provides a summary of maternal health services received during pregnancy, childbirth, and postpartum between 2010 and 2017. Skilled ANC is high across all sources and hovers at around 95 percent for all years that data are available. The 2015‒16 Malawi DHS also provides district-level estimates of the percentage of women receiving ANC from a skilled provider during their last pregnancy. Pregnant women in Nkhotakota, Salima, and Machinga received skilled ANC at slightly higher rates than the national average according to the Malawi DHS: 96.4 percent in Nkhotakota, 97.5 percent in Salima, and 96.7 percent in Machinga. The percentage of women who received four or more ANC visits during their last pregnancy does not show a pattern over time or across data sources. The rate of skilled birth attendance shows an upward linear trend over time, from 71 percent in the 2010 Malawi DHS to 94 percent in the 2017 ONSE IE baseline survey. The rates of PNC do not exhibit a consistent pattern between the 2014 MES and the 2017 ONSE IE baseline survey. The 2015‒16 Malawi DHS reported rates of PNC for both women and newborns that are significantly lower than from the other sources. Figure B3 shows the percentage of women receiving four or more ANC visits during SSDI, as calculated from service statistics (i.e., DHIS 2). These data show a slow upward trend from around 18 percent to 24 178 Malawi ONSE Impact Evaluation: Baseline Report percent over the project period. The rate of women receiving four or more ANC visits, as shown by these facility data, is much lower than that reported by the women in household surveys. ANC data in the Malawi DHIS 2 are considered to be very accurate when compared with facility registers (O’Hagan, et al., 2017). This means that facility data are being transferred accurately through the reporting system and into the electronic database. It is possible that not all ANC patients are recorded in the register, or that registers are misplaced or damaged (resulting in missing data), which could account for some of the differences between facility and household reports of ANC attendance. Women also have documentation of their ANC visits in their health passports, so one way to check completeness of the ANC facility registers would be to compare them with women’s records. However, health passports also have limitations. They can become damaged, destroyed, or lost, and patients sometimes alter their health passport to hide health status or use of services (e.g., HIV status, use of FP), or have multiple health passports (Tough & Lihoma, 2017). Household survey data, such as the MES and DHS, are known to systematically under- and/or overreport certain statistics based on the household report. In some cases, estimates may be underestimated due to recall bias. In other cases, the estimates may be overreported due to social desirability bias. There is no clear indication of which estimates of ANC are most correct in this case, but the likelihood is that the real rate of ANC is between the DHIS 2 estimates and the household survey estimates. This would occur if we expect that the DHIS 2 estimates are underreported and the household estimates are overreported. Malawi ONSE Impact Evaluation: Baseline Report 179 Table B2. Percentage of women who received maternal health care services, 2012–2017 Malawi Indicator 2010 Malawi DHS 2012 SSDI baseline 2014 MES 2014 MIS 2015–2016 Malawi DHS 2016 SSDI end line 2017 MIS 2017 ONSE IE baseline Any ANC - 98 - - 98 83.0 - 90.0 - 99 Skilled ANC 95 - 96 - 95 - - 99 ANC 4+ 46 63 45 - 51 56 - 52 Skilled birth attendance 71 83 87 - 90 90 - 94 Any PNC woman (facility births) - - 75 - 45 - - 67 Any PNC child (facility births) - - 81 - 63 - - 88 Two+ doses of SP/Fansidar - - - 64 - - 76.7 69* Three+ doses of SP/Fansidar - - - 13 - - 42.6 32 *The ONSE IE baseline indicator reports on the number of women receiving exactly two doses. 180 Malawi ONSE Impact Evaluation: Baseline Report Figure B3. Percentage of women who attended four or more ANC visits in SSDI-supported facilities, April 2012−March 2016 Source: SSDI Endline Report Malawi ONSE Impact Evaluation: Baseline Report 181 Availability of Maternal Health Medicines The availability of select maternal health medicines is presented in Table B3. These indicators generally asked about the availability of medicines on the day of the survey and may not reflect general trends in the availability of medicines. Timing of the survey related to distribution of medicines by the Central Medical Stores may also influence the availability of drugs. There is no pattern of availability of medicines over time or across surveys. The medicines were available at the majority of facilities at all times, except for nifedipine in 2014, as report by the SPA. Table B3. Availability of select medicines for maternal health, percentage of facilities with medicines available on the day of the survey Medicines 2012 SSDI baseline 2014 SPA 2016 SSDI End line 2017 ONSE IE baseline Oxytocin 89 95 84 89 Magnesium sulphate 62 85 79 66 Antihypertensives 93 - 91 - Methyldopa - - - 76 Nifedipine - 17 - 73 Knowledge of Maternal Health Complications The SSDI baseline and end line, and the ONSE IE baseline survey, asked women about danger signs during pregnancy and childbirth. In 2012, at the time of the SSDI baseline, under 20 percent of the women reported that vaginal bleeding was a danger sign in childbirth, whereas 45 percent of the women reported that swollen hands, feet, or face was a danger sign during pregnancy. The percentage of women reporting these symptoms as danger signs by the end of SSDI had increased. However, at the time of the ONSE IE baseline, these estimates were quite different. Knowledge of vaginal bleeding had increased whereas knowledge of swollen hands, feet, or face had decreased (Table B4). Table B4. Percentage of women who reported vaginal bleeding or swollen hands, feet, or face as danger signs during pregnancy in SSDI-supported districts (2012 and 2016) and ONSE IE project districts (2017) Vaginal bleeding Swollen hands, feet, or face 2012 SSDI baseline 18 45 2016 SSDI end line* 43-44 50-57 2017 ONSE IE baseline 534 29 *The SSDI end line report provided ranges for these indicators. 182 Malawi ONSE Impact Evaluation: Baseline Report Child Health Care Seeking and Knowledge of Symptoms and Causes of Malaria and Diarrhea Table B5 shows the percentage of children ill with fever and diarrhea who were sick in the past two weeks who had been taken for treatment. For all surveys except the ONSE IE baseline, the population of interest is children under five. For the ONSE IE baseline, the population of interest is children under three. The DHS surveys found that care was sought for approximately two-thirds of children with fever, whereas the SSDI baseline, MES, and ONSE IE baseline found that care was sought for approximately three-quarters of children. The MIS found that care was sought for just over one-half of children. Table B5. Percentage of ill children in the past two weeks taken for treatment and women’s knowledge of causes of illnesses Indicator 2010 DHS 2012 SSDI baseline 2014 MES 2014 MIS 2016 SSDI end line 2015-16 DHS 2017 ONSE IE baseline Care seeking for fever Care seeking for children with fever in the past 2 weeks 65 75 75 54 - 67 79 Knowledge of symptoms and causes of malaria Reported fever as a symptom of malaria - 96 - 72 - - 83 Reported mosquitoes as cause of malaria 82 80 Care seeking for diarrhea Treatment/advice sought for children with diarrhea in the past two weeks 62 - 67 - - 66 78 Knowledge of causes of diarrhea Defecating/urinating in open spaces - 40 - - 76 - 5 Touching food without washing hands with soap - 46 - - 60 - 22 Not washing hands after defecation - 68 - - 68 - 23 Knowledge of fever as a symptom of malaria and mosquitoes as the cause of malaria were high across surveys. Care seeking for diarrhea hovered at around two-thirds for all sources, except for the ONSE IE baseline survey, which found that 78 percent of children with diarrhea were taken for treatment. Although the DHS Malawi ONSE Impact Evaluation: Baseline Report 183 and MES provide national-level estimates, the ONSE IE baseline survey covers only three districts where care seeking may differ from the national average. Knowledge of the causes of diarrhea increased or remained constant from the SSDI baseline to end line surveys but was found to be much lower at the time of the ONSE IE baseline. Availability of Child Health Medicines Estimates of the availability of priority medicines for children are provided in the SSDI-S surveys, the 2014 SPA, and the ONSE IE baseline. The SPA and ONSE IE baseline estimates of the availability of medicines were generally lower than the SSDI-S baseline and end line estimates. However, these types of indicators generally asked about the availability of medicines on the day of the survey and may not reflect general trends in the availability of medicines. Timing of the survey related to distribution of medicines by the Central Medical Stores may also influence the availability of drugs (Table B6). The availability of medicines was high during the 2012-2016 surveys, except for vitamin A in 2014. The ONSE IE baseline results are more varied. The availability of ORS and ACT was high but the availability of vitamin A, amoxicillin, and procaine benzylpenicillin was low, at between 45 percent and 57 percent. Table B6. Availability of select priority medicines for children, percentage of facilities with medicines available on the day of the survey, Malawi 2012−2017 Medicine 2012 SSDI baseline 2014 SPA 2016 SSDI end line 2017 ONSE IE baseline ORS 96 91 100 79 Vitamin A 79 43 83 57 First-line antimalarial 81 - 96 - ACT - 92 - 92 Oral antibiotics 99 - 100 - Amoxicillin - 81 - 53 Procaine benzylpenicillin - - - 45 Water and Sanitation Water and sanitation indicators are available in the Malawi DHS surveys, the MES, and the ONSE IE baseline survey (Table B7). Access to an improved source of water was fairly constant over time, with an increase of about 8 percentage points between 2010 and 2017. Use of an appropriate method for treatment of water was also fairly consistent across sources. 184 Malawi ONSE Impact Evaluation: Baseline Report Access to improved sanitation generally increased over time, although the 2015‒16 Malawi DHS estimate is higher than both the 2014 MES and 2017 ONSE IE baseline survey. This could be because of the geographic range of the surveys; the Malawi DHS provides a national average, whereas the MES and ONSE IE baseline surveys were conducted in a subset of districts throughout Malawi. Improved sanitation lags behind improved source of water and could be improving at different rates throughout the country. Table B7. Water and sanitation indicators from various sources, percentage of households that had improved source of water and sanitation, Malawi 2010−2017 Indicator 2010 DHS 2014 MES 2015–2016 DHS 2017 ONSE IE baseline Improved source of water 79.7* 86.2 87.2 87.8 Appropriate treatment** 31.9 27.8 26.2 26.8 Improved sanitation 6.4 40.6 51.8 47.7 Number 24,825 26,713 26,361 3,962 *Urban, 92.6%, rural 77.1% **Appropriate water treatment methods are boiling, bleaching, filtering, and solar disinfection. Malawi ONSE Impact Evaluation: Baseline Report 185 APPENDIX C. METHODS SUPPLEMENT Selection of Study Districts The three ONSE intervention districts—Machinga, Nkhotakota, and Salima (Figure C1)—were chosen for inclusion in the study in collaboration with USAID/Malawi. These districts were selected because they are receiving the full package of ONSE interventions (i.e., FHP, HSS, and malaria). The goal of the evaluation is to determine the impact of ONSE by comparing districts receiving the full intervention package with districts not receiving ONSE. Mzimba, Nsanje, and Ntchisi were chosen as comparison districts. To make a reliable comparison between project and comparison domains, a comparison domain should be carefully selected to optimize comparability with project domains before the project begins (i.e., at baseline). A data-driven approach was used to identify like comparison districts for the impact evaluation. Specifically, district-level statistics related to ONSE outcomes were used to create a database with data from the 2014 Malawi MDG end line survey. These statistics were imported into Stata and a linear probability model was fitted to predict selection into the ONSE project. These predicted probabilities for selection into the ONSE project were then compared in regions for ONSE and non-ONSE districts. Comparison districts were then chosen based on these fitted values, past participation in the preceding USAID health project (SSDI), and CDCS priority status.11 To validate the selection of comparison districts, the evaluation team also used data from the 2010 Malawi DHS. USAID/Malawi provided a geographic analysis of health statistics that mapped priority indicators using groupings. These groupings were converted into ranked categories as inputs into a count index for each district. The index value for each district was then compared for ONSE and non-ONSE districts to further validate that the selected project and comparison districts were broadly comparable. Quantitative Household Survey Instrument The quantitative survey instrument used modified the 2015‒16 Malawi DHS survey instruments from the household and woman’s questionnaires to gather characteristics and outcomes at two levels of interest: 1) Household questionnaires for all selected households 2) Woman’s questionnaires for all WRA in the selected households 11 The USAID Health, Population and Nutrition Project Appraisal Document Geographic Analysis using the 2010 DHS data was then used as a secondary source to assess the selections. The geographic analysis mapped priority indicators using groups (e.g., infant mortality rate >66, IMR<66) and these groupings were converted into ranked categories. From these ranked categories, a count index from the indicators was created. The index value for each district was then compared for ONSE and non-ONSE districts for selected evaluation districts as a second source of validation that the districts were broadly comparable. 186 Malawi ONSE Impact Evaluation: Baseline Report The survey instruments were adapted and modified for the evaluation context. DHS translations of the Malawi DHS were used for the original DHS questions, and additional or modified items were translated by the CSR. The questionnaire modules are shown in Table C1. Figure C1. Map of project and comparison districts Malawi ONSE Impact Evaluation: Baseline Report 187 Table C1. DHS household survey modules selected for the ONSE baseline survey Household questionnaire Woman’s questionnaire • Household identification • Informed consent • Household roster and demographics • Dwelling characteristics • Household and community programs • Identification of WRA • Woman’s background • Reproduction • Contraception • Pregnancy and PNC • Knowledge of child health • Child health • Marriage and sexual activity • Fertility preferences • Husband’s background and women’s work • Patient satisfaction • Tracking Sampling Frame The baseline sampling frame used for the household survey was drawn from the frame of the Malawi Population and Housing Census, which was conducted in 2008. A total of 268 EAs were randomly selected for inclusion in the study with probability proportional to size: 134 in each study domain. In each EA, a household listing was conducted and then 30 households were randomly selected for inclusion in the study.12 Once selected, a household was then visited to conduct the survey. A household questionnaire was administered regardless of whether the household contained any eligible women—i.e., WRA. If the household contained WRA, all WRA were asked to complete the woman’s questionnaires. Baseline Sampling Weights Sampling for the household survey was based on a stratified multi-stage sampling design. Design weights were calculated based on the separate sampling probabilities for each sampling stage. 12 For the survey, a household was defined as a group of people who live together and eat from the same kitchen. If there were multiple wives and one husband living together and eating from the same kitchen, they were considered one household. 188 Malawi ONSE Impact Evaluation: Baseline Report The first stage involved the selection of EAs in the project and comparison domains. The EAs were selected based on the probability proportional to the population size, as determined by the 2008 Malawi Population and Housing Census. The selection probability of i-th EA in domain h is: 𝑝1ℎ𝑖 = 𝑎ℎ × 𝑁ℎ𝑖 𝑁ℎ Where 𝑎ℎ: number of sample clusters selected in domain h, 𝑁ℎ𝑖 : total number of households in the frame for the i-th sample cluster in domain h, and 𝑁ℎ: total number of households in the frame in domain h. The second stage involved a random selection of households from each selected EA. The selection probability of j-th households in EA i in domain h is: 𝑃2ℎ𝑖𝑗 = 𝑏ℎ𝑖 𝑁ℎ𝑖 ∗ Where 𝑏ℎ𝑖 = number of sampled households selected for the i-th sample cluster in domain h. 𝑁ℎ𝑖 ∗ = number of eligible households listed in the household listing for the i-th sample cluster in domain h. The overall selection probability of each household in EA i of domain h is the product of the selection probabilities of the two stages: 𝑃ℎ𝑖𝑗 = 𝑃1ℎ𝑖 × 𝑃2ℎ𝑖𝑗 = 𝑎ℎ × 𝑁ℎ𝑖 𝑁ℎ × 𝑏ℎ𝑖 𝑁ℎ𝑖 ∗ The design weight for each household in EA i of domain h is the inverse of its overall selection probability: 𝑊ℎ𝑖𝑗 = 1 𝑝ℎ𝑖𝑗 = 𝑁ℎ × 𝑁ℎ𝑖 ∗ 𝑎ℎ × 𝑁ℎ𝑖 × 𝑏ℎ𝑖 Malawi ONSE Impact Evaluation: Baseline Report 189 Design Weight of Women All women were interviewed in each selected household. Therefore, the selection probability of a WRA equals the selection probability of her household multiplied by a cluster-level non-response adjustment for women, expressed as follows: 𝑃ℎ𝑖𝑗𝑓 = 𝑎ℎ × 𝑁ℎ𝑖 𝑁ℎ × 𝑏ℎ𝑖 𝑁ℎ𝑖 ∗ × 𝑓ℎ𝑖 𝑒ℎ𝑖 = 𝑃ℎ𝑖𝑗 Where 𝑓hi = number of women in the cluster bhi who were interviewed ehi = number of women in the cluster who were eligible to be interviewed The design weight for the woman j in EA i of domain h is the inverse of its overall selection probability: 𝑊ℎ𝑖𝑗𝑓 = 1 𝑝ℎ𝑖𝑗 Sampling Weight The sampling weight was calculated with the design weight corrected for unit non-response calculated at the level of the cluster as ratios of the number of interviewed units over the number of selected units, where units could be households or individual respondents. The household sampling weight was calculated by dividing the household design weight by the household response rate. The individual sampling weight was calculated by dividing the individual design weight by the individual response rate. Survey Response Rates Tables C2 and C3 provide results of the household and facility survey response rates. Table C2. Results of the household and women’s interviews Characteristics Project Comparison N Household interviews Number selected 4,020 4,017 8,037 Number not found/absent 54 43 97 Number refused/incomplete 4 6 10 Number interviewed 3,962 3,967 7,929 Response rate (%) 98.6 98.8 98.7 Interviews with WRA 190 Malawi ONSE Impact Evaluation: Baseline Report Characteristics Project Comparison N Number selected 3,872 3,908 7,780 Number refused/not found/incomplete 96 142 238 Number interviewed 3,776 3,766 7,542 Response rate (%) 97.5 96.4 96.9 Note: Response rate = number interviewed/number eligible. Table C3. Health facility sample District Public facilities Hospitals Health centers Dispensaries CHAM Total Project Machinga 1 11 3 6 21 Nkhotakota 2 10 2 6 20 Salima 1 8 0 5 14 Comparison Mzimba 3 38 4 12 57 Nsanje 1 9 0 4 14 Ntchisi 1 9 1 2 13 Total 9 85 10 35 139 Summary of Evaluation Outcomes Tables C4 and C5 provide a summary of primary and secondary evaluation outcomes. Table C4. Primary population-level outcomes of interest Category Outcomes Antenatal care • Percentage of women of WRA who had a birth in the past three years who received ANC from skilled providers (such as doctors, medical officers, clinical officers, medical assistants, nurses, and midwives) at least once for their most recent birth* • Percentage of WRA who had a birth in the past three years who had an ANC visit in their first trimester for their most recent live birth • Percentage of WRA who had a birth in the past three years who had at least four ANC visits for their most recent live birth Maternal health • Percentage of live births in the last three years that were attended by a skilled provider (doctor, nurse, midwife, and auxiliary nurse/midwife) for their most recent birth* Postnatal care • Percentage of women with a postnatal checkup within two days of birth for their most recent live birth in the past two years Reproductive health • Modern contraceptive prevalence rate among WRA who are married or living with a man • Modern contraceptive rate among all WRA Malawi ONSE Impact Evaluation: Baseline Report 191 Category Outcomes Care seeking for fever • Percentage of children under five who had a fever in the two weeks preceding the survey for whom advice or treatment was sought from a health facility or provider Patient satisfaction • Percentage of WRA who visited a health facility in the last three months for themselves or their children who reported that they were “very satisfied” with: ✓ Time they waited to see a provider ✓ Ability to discuss problems or concerns with a provider ✓ Explanation they received about their problem or treatment ✓ Audio and visual privacy ✓ Availability of medicines at the facility ✓ Facility service hours Facility cleanliness Maternal and newborn health knowledge • Knowledge of information that should be included in birth planning • Knowledge of warning/danger signs of pregnancy • Knowledge of primary warning/danger signs of maternal complications during childbirth ✓ Knowledge of primary warning/danger signs of newborn complications Child health knowledge • Percentage of four signs and symptoms of diarrhea correctly identified by WRA • Percentage of six causes of diarrhea correctly identified by WRA • Percentage of four signs and symptoms of pneumonia correctly identified by WRA • Percentage of three causes of childhood pneumonia correctly identified by WRA • Percentage of WRA who correctly identified fever as a primary sign of malaria • Percentage of WRA who correctly identified mosquitos as the cause of malaria Beliefs about FP • Percentage of WRA who believe each statement is “completely true”: ✓ FP methods are safe ✓ Planning the family is the responsibility of both men and women ✓ Getting pregnant before you are 18 puts your health and that of the baby in danger ✓ Long-acting FP methods help to conveniently space pregnancies ✓ FP should be used by husbands and wives for the health of the entire family ✓ Long-term and permanent FP methods provide safe and healthy ways to temporarily or permanently stop having children ✓ There are FP methods available at the clinic for everybody 192 Malawi ONSE Impact Evaluation: Baseline Report Category Outcomes ✓ Becoming pregnant after 40 years of age can be dangerous to your health • Talking openly and honestly to your children about the consequences of unprotected sex is important Exposure to messaging ▪ Percentage of WRA who recall hearing the slogan Moyo ndi mpamba: Usamalireni! in the last 12 months *These indicators reached almost 100 percent at baseline and are not expected to increase significantly over the project period. We will track them as a component of downstream indicators. Table C5. Secondary facility-level outcomes of interest Category Outcome General service availability • Basic amenities: mean score of seven items as a percentage • Basic equipment: mean score of six items as a percentage • Percentage of facilities with 15 priority medicines for mothers • Percentage of facilities with 12 priority medicines for children Specific service availability • Percentage of facilities offering specific services: ✓ FP ✓ Maternal health services, including ANC, normal delivery, caesarean delivery, and basic and comprehensive emergency obstetric and newborn care ✓ Preventative and curative child health services ✓ Malaria services Family planning • Percentage of facilities with all items: ✓ Staffing and guidelines ✓ Medicines and commodities ✓ Equipment Maternal health • Percentage of facilities with all items: ✓ Staffing and guidelines ✓ Medicines and commodities ✓ Diagnostics ✓ Equipment Preventative and curative child health services • Percentage of facilities with all items: ✓ Staffing and guidelines ✓ Medicines and commodities ✓ Equipment Malaria services • Percentage of facilities with all items: ✓ Staffing and guidelines ✓ Medicines and commodities ✓ Diagnostics Malawi ONSE Impact Evaluation: Baseline Report 193 Evaluation Next Steps The next step in the impact evaluation is implementation process monitoring (2018‒2020). This will be followed by the end line household survey, health facility assessment, and impact analysis in 2021. A qualitative study is also planned at end line. Each of these activities is described below. Implementation Process Monitoring The evaluation team will conduct implementation process monitoring to understand how the “smart” capacity building and problem-solving approach were operationalized in the intervention districts of Machinga, Nkhotakota, and Salima, and how the approaches affected change in health and service delivery outcomes. This component will involve ongoing review of ONSE documents, such as workplans and progress reports, and annual key informant interviews with ONSE staff, DHMT members, and other stakeholders. Impact Analysis The primary objective of the evaluation is to quantify the impact of ONSE. After the end line household survey and health facility assessment are conducted, the evaluation team will quantify the impact of the intervention using the DID approach with household fixed effects. This section describes the methods for this analysis. Because the ONSE project domain was purposively selected, a like comparison domain may be a challenge to identify. Use of the DID method in combination with the data-based selection process to identify the comparison domain helps to maximize the internal validity of the study. Although observed information was used to select a comparison domain, by design, the DID approach assumes that the domains are not the same at baseline. This method differences out observed and unobserved differences between the domains that remain constant over the project period, thus accounting for any time-invariant differences that exist after selection. For the DID estimator to be valid, the parallel trend assumption must be met. The parallel trend assumption states that the outcomes trends in the comparison domain are a good approximation of what would have been observed in the project domain in the absence of the project. Alternatively stated, the average change in outcome for the project domain in the absence of the project equals the average change in outcome for the comparison domain. Empirically, the DID estimator is specified as follows: for each outcome of interest 𝑌, let 𝑌 1 (𝑡) represent the potential outcome, or outcome that would have been observed, under the ONSE project at time 𝑡, and let 𝑡 ∈ {0, 0.5, 1}. 𝑌 1 (0) represents the potential outcome under the ONSE project at baseline, and 𝑌 1 (1) represents the potential outcome under the ONSE project at end line. Similarly, 𝑌 0 (𝑡) represents the potential outcome at time 𝑡 without the ONSE project. Let 𝐴 be an indicator of inclusion of the district in the ONSE project. Parameter: The parameter of interest is the difference in the difference of outcomes before and after the intervention period with the ONSE project and without the ONSE project. Formally, the difference in 194 Malawi ONSE Impact Evaluation: Baseline Report outcomes over the intervention period under the ONSE project can be written as 𝑌 1 (1) − 𝑌 1 (0). The difference in outcomes over the intervention period without the ONSE program can be written 𝑌 0 (1) − 𝑌 0 (0). The parameter of interest, the difference in differences, can then be written as 𝛿𝐷𝐷 = [𝑌 1 (1) − 𝑌 1 (0)] − [𝑌 0 (1) − 𝑌 0 (0)]. The expected value of 𝛿𝐷𝐷 is the true effect of the intervention. Estimation: Assuming that the trends observed at the comparison sites are proportional to trends that would have been observed at the intervention sites had they been comparison sites, 𝐸[𝑌 0 (1) − 𝑌 0 (0)] may be estimated as 𝐸[𝑌(1) − 𝑌(0)|𝐴 = 0], and one can estimate 𝐸[𝑌 1 (1) − 𝑌 1 (0)] as 𝐸[𝑌(1) − 𝑌(0)|𝐴 = 1]. 𝛿𝐷𝐷 can be equivalently estimated using a regression model for 𝑌: 𝐸(𝑌|𝐴,𝑡) = 𝛽0 + 𝛽1𝐴 + 𝛽2𝑡 + 𝛽3𝐴 × 𝑡 where 𝛽3 is the impact of the intervention on outcome 𝑌. Estimates for 𝛽3 will be generated for each outcome and reported to the Mission in the end line ONSE impact evaluation report. End line Qualitative Study In addition to the end line household survey and health facility assessment, an end line qualitative study will be developed to understand how ONSE’s community engagement and mobilization activities increased the demand for and uptake of services. Qualitative findings will be integrated in the end line impact evaluation report. 274 Malawi ONSE Impact Evaluation: Baseline Report APPENDIX D. DATA COLLECTION TOOLS Malawi ONSE Impact Evaluation: Baseline Report 275 276 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 277 278 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 279 280 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 281 282 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 283 284 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 285 286 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 287 288 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 289 290 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 291 292 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 293 294 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 295 296 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 297 298 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 299 300 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 301 302 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 303 304 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 305 306 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 307 308 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 309 310 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 311 312 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 313 314 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 315 316 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 317 318 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 319 320 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 321 322 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 323 324 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 325 326 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 327 328 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 329 330 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 331 332 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 333 334 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 335 336 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 337 338 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 339 340 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 341 342 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 343 344 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 345 346 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 347 348 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 349 350 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 351 352 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 353 354 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 355 356 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 357 358 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 359 360 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 361 362 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 363 364 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 365 366 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 367 368 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 369 370 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 371 372 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 373 374 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 375 293 Malawi ONSE Impact Evaluation: Baseline Report APPENDIX E. SURVEY IMPLEMENTATION The baseline household and facility surveys were implemented by CSR under the guidance of MEASURE Evaluation. Training of Supervisors and Healthcare Workers and Pilot Test MEASURE Evaluation and the CSR core team (comprised of the principal investigator and three co￾investigators) trained 25 supervisor candidates and 20 healthcare workers on the survey protocol. Sessions included an introduction to the ONSE project and the ONSE impact evaluation; interviewing techniques; human subjects’ protection; household listing techniques; use of tablets; and roles and responsibilities. The household and woman’s questionnaire were reviewed question-by-question in English and Chichewa. Several days of role-plays using the tablets were conducted for the household and women’s questionnaires. The health facility assessment tool was reviewed question-by-question by the healthcare workers in English. This tool, due to the abundance of medical terms, was implemented in English. Written quizzes were periodically administered to assess the supervisor and healthcare worker candidates. At the end of supervisor and healthcare worker training, a two-day pilot of the household and woman’s questionnaire was conducted in six villages outside Zomba town, with each supervisor candidate interviewing a minimum of two households per day. The facility assessment tool was piloted in seven health facilities in Zomba district on the same days. The team identified minor edits to the survey tools and the tablet and paper forms were subsequently revised and finalized. Training of Research Assistants and Data Collection Immediately following the pilot, 85 research assistant (data collector) candidates were trained by MEASURE Evaluation and the CSR core team. This training mirrored the supervisor training but focused exclusively on the household and women’s questionnaire (as healthcare workers were already trained during supervisor training). Eighteen supervisors were selected from among the 25 candidates and trained on sampling, with the remaining candidates becoming research assistants. The training concluded with three days of field practice – one day practicing the household listing and sampling, and two days conducting practice interviews and using control sheets to track progress and record final interview outcomes. Supervisors served as assistant trainers during the training of research assistants. Eighteen data collection teams were formed, each comprised of a supervisor, at least one healthcare worker, and five research assistants. Data collection occurred from April 28 – June 30, 2017. A MEASURE Evaluation team member accompanied the team on first week of data collection for quality oversight, visiting each of the 18 data collection teams twice to review implementation of the sampling protocol, ensure proper use of supervisor and enumerator control sheets, and respond to queries from the team. Daily debriefs were held. 294 Malawi ONSE Impact Evaluation: Baseline Report Data Quality Control Data quality was ensured at several levels. At the tablet level, the survey was programmed so that most questions could not be skipped. Data validation checks were also programmed into the survey which required review of unlikely values or prevented research assistants from moving forward with the survey until errors were corrected. Supervisors monitored research assistant performance by observing interviews and reviewing survey responses in the tablet for completeness and consistency with fieldwork data management forms before finalizing and transmitting questionnaires to a secure MEASURE Evaluation server. Throughout data collection, a MEASURE team member reviewed incoming data daily and communicated with site supervisors to correct any errors. Selected quality checks included ensuring the correct number of completed forms (30) were received for each cluster; ensuring all forms had a final outcome code; confirming all operational health facilities were included in the dataset; and communicating with supervisors to correct duplicate household ID codes, as needed. Confidentiality The tablets used for data collection were password protected and their hard drives were encrypted. Supervisors transmitted completed surveys (encrypted) to the secure MESAURE Evaluation server whenever they had Internet access. Once transferred, data were stored on the secure server at MEASURE Evaluation. To ensure data protection and confidentiality across the study, all field team members signed a confidentiality agreement and committed to using reasonable data protection measures to protect the data. When data collection was complete, tablets were returned to MEASURE Evaluation, checked for completeness of data delivery, and cleared of all survey data. All paper forms used during implementation of the survey were stored in locked file cabinets. Informed Consent The study protocol (#16-2690) was reviewed by the UNC Office of Human Research Ethics, which determined that the impact evaluation did not constitute human subjects research as defined under federal regulations [45CFR 46.102 (d or f) and 21 CFR 56.102(c)(e)(l)] and did not require IRB approval. The study protocol (#17/02/1738) was reviewed and approved in Malawi by the National Health Sciences Research Committee. All supervisors, healthcare workers, and research assistants were trained in human subjects’ protection and informed consent was obtained from all participants before their participation in the study. Malawi ONSE Impact Evaluation: Baseline Report 295 APPENDIX F. EVALUATION SCOPE OF WORK MEASURE Evaluation Phase IV Scope of Work for ONSE Impact Evaluation: • 2016 Design Phase • 2017 Baseline Survey • 2018 Implementation Process Monitoring July 1, 2016 – December 31, 2018 Revised October 2017 Malawi 296 Malawi ONSE Impact Evaluation: Baseline Report Carolina Population Center University of North Carolina at Chapel Hill 400 Meadowmont Village Circle, 3rd Floor Chapel Hill, NC 27517 USA TEL: 919-445-9350 FAX: 919-445-9353 http://www.cpc.unc.edu/measure Malawi ONSE Impact Evaluation: Baseline Report 297 Introduction USAID/Malawi requested that MEASURE Evaluation conduct an impact evaluation of the Organized Network of Services for Everyone’s (ONSE) Health project. ONSE will work in 16 districts to reduce maternal, under-five, infant, and neonatal mortality rates, and to increase the modern contraceptive prevalence rate. ONSE is the five-year follow-on to USAID’s flagship project in Malawi, Support for Service Delivery Integration (SSDI), which ends in early March 2017. ONSE is designed to effect change through improved access to and quality of priority health services, including maternal, neonatal, and child health including nutrition (MNCH); FP/RH; malaria; and water, sanitation and hygiene (WASH). ONSE will also work to strengthen district health systems in support of MNCH, FP/RH, malaria, and WASH, and to increase community demand for these priority services. Target populations include: women of reproductive age (15-49); children under 5 years, including a specific focus on newborns; very young adolescents aged 10 to 14; adolescents aged 15-24 years; and pregnant and breastfeeding women. ONSE has three activity areas that will be implemented in various combinations throughout the 16 targeted districts: • The family health package (FHP) is ONSE’s service delivery component and is focused on improving access to and quality of MNCH, FP/RH, and WASH services. ONSE will be the primary implementer of services in a subset of the 16 targeted districts; • A health system strengthening (HSS) component will be implemented in all 16 targeted districts to improve management, supervision of human resources for health, governance, policy implementation, and the use of data-based decision making at the district level; • Malaria services will be provided in a subset of the targeted districts based on need and will be integrated with the FHP. A strategic principle employed by ONSE that is of interest to the impact evaluation is “smart capacity building and problem-solving” which will occur at the district, facility, and community level. An example of smart capacity building at the district level is co-development of capacity building plans with District Health Management Teams (DHMTs) to identify district-specific capacity building needs in MNCH, FP/RH, and malaria. Smart capacity building in this context will rely on performance improvement methods and on-site mentoring of health facility staff and DHMT members and includes workshops and clinical trainings. To overcome challenges to improving health in their districts, ONSE also supports smart problem-solving. An example of smart problem solving is application of the District Health Program Improvement strategy which helps district officials use locally collected data to identify health priorities and develop solutions to problems. The ONSE impact evaluation has three primary research questions: 1) What is the impact of the ONSE project on changes in facility and health outcomes compared with changes in these outcomes in districts that did not receive the ONSE project? Did variation in the application of smart capacity building and problem solving in each district and facility impact targeted outcomes? 298 Malawi ONSE Impact Evaluation: Baseline Report 2) How is “smart capacity building and problem solving” operationalized in each district? 3) What is the impact of ONSE’s strategic community engagement and mobilization activities on targeted outcomes compared with outcomes in communities that did not receive community engagement and mobilization support? A mixed methods approach will most thoroughly answer these research questions. Proposed methods include a household survey, health facility assessment, an implementation process monitoring component (project document review followed by key informant interviews), and a qualitative study of community engagement activities. This mixed methods approach provides an opportunity to use qualitative and quantitative results to triangulate data sources and enrich interpretation of the evaluation’s findings. This revised October 2017 scope of work (SOW) for the ONSE impact evaluation includes the following phases to date: The design phase (July 1 – December 31, 2016) included a scoping trip to Malawi in preparation for development of the impact evaluation concept note, as well as submission and approval of the concept note, baseline survey protocol, and SOW for baseline data collection. Baseline data collection phase (January 1, 2017 – January 31, 2018) includes household surveys in approximately 8,200 households in six districts (three intervention and three comparison districts). These surveys include interviews with women aged 15-49 living in these households. The baseline also includes health facility assessments in public and Christian Health Association of Malawi (CHAM) facilities in the same six districts. The baseline phase includes submission of the draft and final baseline reports, as well as a dissemination meeting. Implementation process monitoring phase (October 15, 2017 – December 31, 2018) includes review of ONSE work plans, quarterly and annual progress reports, and other technical documents, combined with key informant interviews with ONSE staff, DHMT members, and other stakeholders. The goal of implementation process monitoring is to understand how smart capacity building and problem solving is operationalized in project districts. The evaluation team will document the roll out of activities related to smart capacity building and problem solving/planning. This revised SOW (October 2017) includes implementation process monitoring activities through December 31, 2018. It is anticipated that potential future, annual, modifications to this SOW will expand the SOW to include process monitoring phases in 2019 and 2020, as well as an end line data collection phase in 2021. Malawi ONSE Impact Evaluation: Baseline Report 299 Background: MEASURE Evaluation Phase IV The primary objective of MEASURE Evaluation is to enable countries to strengthen their systems to generate high quality health information that is used for decision making at local, national, and global levels. MEASURE Evaluation applies a systems approach to achieve this objective in a sustainable way. One application of this approach is to increase capacity for rigorous evaluation. MEASURE Evaluation’s results framework reflects the overarching implementation strategy whereby the project works through distinct activities to achieve results. Achievements in the four result areas shown below contribute to the overall project objective. In Malawi, MEASURE Evaluation’s work in this SOW will address Result 4. Result 1: Strengthened collection, analysis and use of routine health data; Result 2: Improved country-level capacity to manage health information systems, resources and staff; Result 3: Methods, tools and approaches improved and applied to address health information challenges and gaps; Result 4: Increased capacity for rigorous evaluation. This SOW will contribute to the achievement of Result 4. Capacity-building in rigorous evaluation and related technical skills is primarily through collaborative implementation of these activities (learning by doing). As such capacity-building is embedded in the process of implementing the activities. 4MW-001: ONSE Impact Evaluation Activity Leader: Emily Weaver Other Staff: Milissa Markiewicz, Bernard Agala Objectives 1. To design the Malawi ONSE impact evaluation. 2. To implement the 2017 baseline survey in three intervention districts (Machinga, Nkhotakota, and Salima) and three comparison districts (Mzimba, Nsanje, and Ntchisi) in Malawi. 3. To conduct implementation process monitoring through December 31, 2018. 4. To build the capacity of the local research partner in rigorous evaluation through collaborative implementation of the baseline survey and other targeted activities, with a focus on the technical and managerial competencies needed to implement a large household survey and health facility assessment. 300 Malawi ONSE Impact Evaluation: Baseline Report Summary During the design phase (2016), MEASURE Evaluation finalized the study protocol, finalized the data collection tools for the ONSE impact evaluation, and submitted the UNC IRB application. The recruitment process for a local research partner also commenced. During the baseline phase (2017), an agreement with a local research partner [Centre for Social Research, (CSR)] was executed, and they coordinated ethics approval in Malawi. MEASURE Evaluation trained CSR co-investigators on the study protocol and administration of the data collection instruments. Together, MEASURE Evaluation and CSR trained the data collection team. CSR managed field work and was responsible for responding to data queries from MEASURE Evaluation. MEASURE Evaluation is currently leading data analysis and preparing the baseline report, with input from CSR. An annual dissemination workshop is anticipated with details to be finalized in collaboration with USAID/Malawi. Under this revised SOW, an implementation process monitoring phase (2018) will begin and continue through December 2018. It will include review of ONSE project documents and interviews with ONSE staff, DHMT members, and other stakeholders. Interviews will be conducted in collaboration with CSR. Roll out of relevant ONSE activities will be documented using process evaluation methods and a report will be drafted in April 2018. Review of ONSE project documents will be ongoing through the end of 2018. Benchmarks Design Phase (2016) Benchmark 1: Scoping visit conducted MEASURE Evaluation team members will conduct a two-week scoping visit to gather detailed information about ONSE’s goals and activities to be used as input into the design of the impact evaluation, and will also gather information to inform data collection, the sampling plan, and analysis. Benchmark 2: RFA posted for local research partner MEASURE Evaluation will post an RFA to recruit a local research partner. Benchmark 3: SOW for baseline data collection submitted MEASURE Evaluation will submit a SOW for baseline data collection to USAID/Malawi. Baseline Phase (2017) Benchmark 4: Malawi ethics approval obtained MEASURE Evaluation will execute an agreement with a local research partner. The local research partner will translate and back translate data collection tools and consent forms and submit an ethics approval application in Malawi and respond to any comments from the local ethics board. Malawi ONSE Impact Evaluation: Baseline Report 301 Benchmark 5: Data collection complete MEASURE Evaluation will train master trainers from the local research partner and assist with piloting the data collection tools. MEASURE Evaluation will work with the local research partner to train the data collectors, finalize data collection tools, and conduct field work. Benchmark 6: Draft baseline report of findings ready for review MEASURE Evaluation and the local research partner will conduct data analysis and write a draft baseline report. Benchmark 7: Finalize the baseline report A dissemination workshop with stakeholders in country will be hosted and thereafter the baseline report will be finalized with stakeholder input incorporated. Implementation Process Monitoring (2018) Phase Benchmark 8: Submit implementation process monitoring report MEASURE Evaluation will review ONSE documents and will conduct interviews with ONSE staff, DHMT members, and other stakeholders in collaboration with the local research partner. A report will be drafted in April 2018. A dissemination meeting (in person or virtual) is anticipated with details to be finalized in collaboration with USAID/Malawi. 302 Malawi ONSE Impact Evaluation: Baseline Report 2016 Design Phase Timeline and Deliverables Tasks Responsible Party June 1 – December 31, 2016 Deliverables July 2016 Aug 2016 Sept 2016 Oct 2016 Nov 2016 Dec 2016 Benchmark 1: Scoping visit conducted 1.1. Review project documents 1.2. Conduct scoping visit Benchmark 2: RFA posted for local research partner 2.1. Prepare and post RFA Benchmark 3. SOW for baseline data collection submitted 3.1. Concept Note for overall impact evaluation revised and finalized 3.2. Baseline survey protocol developed, reviewed, and finalized Baseline protocol 3.3. Baseline SOW submitted to USAID/Malawi SOW for baseline Malawi ONSE Impact Evaluation: Baseline Report 303 2017 Baseline Timeline and Deliverables Tasks Responsi ble Party January 2017 – January 2018 Deliver￾ables Ja n Fe b Ma r Ap r Ma y Ju n Ju l Au g Se p Oc t No v De c Ja n 4MW-002: ONSE Impact Evaluation Baseline Survey Benchmark 4: Malawi ethics approval obtained. 4.1. Subcontra ct with a local research partner MEASURE 4.2. Translate and back translate data collection tools and consent forms Local partner 4.3. Submit ethics approval application in Malawi and respond to comments MEASURE / Local Partner Benchmark 5: Data collection complete. 5.1. Train the master trainers and pilot the data collection tools MEASURE / Local Partner 5.2. Train data collectors MEASURE / 304 Malawi ONSE Impact Evaluation: Baseline Report Tasks Responsi ble Party January 2017 – January 2018 Deliver￾ables Ja n Fe b Ma r Ap r Ma y Ju n Ju l Au g Se p Oc t No v De c Ja n 4MW-002: ONSE Impact Evaluation Baseline Survey Local Partner 5.3. Finalize data collection tools MEASURE Final tools 5.4. Conduct field work MEASURE / Local Partner Benchmark 6: Draft baseline report of findings ready for review. 6.1. Conduct analysis MEASURE 6.2. Write draft report MEASURE Benchmark 7: Finalize the baseline report. 7.1. Conduct a disseminati on workshop with stakeholde rs in country MEASURE / Local Partner Dissemin a-tion meeting 7.2. Finalize the baseline report (incorporat e MEASURE / Local Partner Final report due Feb 28, 2017 Malawi ONSE Impact Evaluation: Baseline Report 305 Tasks Responsi ble Party January 2017 – January 2018 Deliver￾ables Ja n Fe b Ma r Ap r Ma y Ju n Ju l Au g Se p Oc t No v De c Ja n 4MW-002: ONSE Impact Evaluation Baseline Survey stakeholde r input) 306 Malawi ONSE Impact Evaluation: Baseline Report 2018 Implementation Process Monitoring Timeline and Deliverables Tasks Responsi ble Party October 2018 – December 31, 2018 Delivera bles Oc t￾De c 201 7 Ja n 201 8 Fe b 201 8 Ma r 201 8 Ap r 201 8 Ma y 201 8 Jun 201 8 Jul 201 8 Au g 201 8 Se p 201 8 Oc t 201 8 No v 201 8 De c 201 8 Benchmark 8: Submit implementation process monitoring report. 8.1. Review ONSE project documents and create key informant interview guides MEASUR E 8.2. Conduct key informant interviews in Malawi MEASUR E/ Local partner 8.3. Write and submit 2018 implement ation process monitoring report Report and dissemin a-tion meeting 8.4. On￾going review of ONSE project documents in preparation for Jan/Feb 2019 interviews; create interview guides for 2019 MEASUR E Malawi ONSE Impact Evaluation: Baseline Report 307 308 Malawi ONSE Impact Evaluation: Baseline Report APPENDIX G. TEAM MEMBERS AND DISCLOSURES OF CONFLICTS OF INTEREST Emily Weaver, PhD is the activity lead and Principal Investigator for the evaluation. She is responsible for the overall development of the evaluation design as well implementation of the evaluation. She will have primary responsibility for collaboration with local partners and consultants as well as coordination with USAID/Malawi. Dr. Weaver is a Research Associate for the MEASURE Evaluation project based at the Carolina Population Center. She has worked for more than 10 years in research and evaluation of public health programs with expertise in maternal, newborn and child health. At MEASURE Evaluation, Dr. Weaver is leading evaluations of RMNCH programs in Tanzania and Malawi.. She has worked on impact evaluations for the Feed the Future FEEDBACK project, specifically evaluating the impact of integrated agricultural value chain and nutrition interventions on health outcomes in Malawi and Guatemala. Other work includes building technical capacity in impact evaluation and in measurement of maternal mortality using various data platforms in Africa and Asia. Dr. Weaver earned a PhD in Health Policy and Management from the UNC Gillings School of Global Public Health and a Master of International Affairs from the University of California, San Diego. Milissa Markiewicz, MPH, PMP provides project management, logistic, and qualitative research support to the evaluation. She oversaw the IRB application process and manages subcontracting. Ms. Markiewicz is currently a Research Associate at UNC’s Carolina Population Center (CPC), and serves as CPC’s project manager for several evaluations, including evaluations in Malawi, Uganda, Rwanda, Kenya, Tanzania, and Zambia. She previously served as project manager for the SE Region of the Network for Public Health Law and as a program director at the Terry Sanford Institute of Public Policy at Duke University. Ms. Markiewicz worked in Uganda for more than three years as an academic director and special projects coordinator for the School for International Training. Chris Bernard Agala, PhD is an analyst for the evaluation. He is responsible for data management and analysis for the evaluation, specifically producing indicator statistics and tables for the study groups. Dr. Agala is a research analyst for MEASURE Evaluation, based at the Carolina Population Center. He has more than 13 years of experience working with non-profit and academic institutions, local communities, community￾and local government leaders in research, intervention design, implementation, and monitoring in public health areas including: HIV/AIDS and reproductive health, orphans and vulnerable children, health systems and development in Tanzania, Kenya, Malawi, Ethiopia, India and Cambodia. He earned his PhD in Health Policy and Management from UNC Chapel Hill’s Gillings School of Global Public Health, his bachelor’s degree in finance at Kenyatta University in Nairobi, Kenya and an associate degree in computer science from Starehe Technical Institute in Nairobi, Kenya. John Kadzandira, M.Sc., is the local Co-Principal Investigator for the evaluation. He is responsible for the overall coordination of the study in Malawi, including ensuring timely deliverables, quality control, and reporting to MEASURE Evaluation on technical issues. Mr. Kadzandira also oversaw the recruitment and training of the field team. Mr. Kadzandira has more than 18 years of experience conducting field research in Malawi and other Southern African countries. His expertise is in both quantitative and qualitative research methods, baseline surveys, data analysis, report and proposal writing, as well as monitoring and evaluation Malawi ONSE Impact Evaluation: Baseline Report 309 (M&E). He has considerable experience with the Malawian health sector having undertaken recent studies including the 2015/16 MPHIA, the 2016 house to house meter verification survey for Lilongwe Water Board, and the 2007 Census of Human Resources for the Health Sector, as well as numerous baseline surveys, evaluations, and studies. He has completed investigations funded by clients including USAID, Johns Hopkins University, Save the Children International, World Bank, CDC, NAC, and CARE Malawi. 310 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 311 312 Malawi ONSE Impact Evaluation: Baseline Report Malawi ONSE Impact Evaluation: Baseline Report 313