CHOLERA OUTBREAKS IN GREATER ACCRA REGION, GHANA: THE ECONOMIC COSTS TO THE HEALTH FACILITY AND AFFECTED HOUSEHOLDS BY AWALIME KWESI DZIEDZOM (10444083) THIS THESIS IS SUMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FILFILMENT OF THE REQUIREMENT FOR THE AWARD OF MPHIL ECONOMICS DEGREE JULY, 2015 University of Ghana http://ugspace.ug.edu.gh i DECLARATION This is to certify that this thesis is the result of research undertaken by Dziedzom Kwesi Awalime towards the award of the Master of Philosophy (M.PHIL) degree in Economics at the Department of Economics, University of Ghana. ………………………………………………………………………………….. DZIEDZOM KWESI AWALIME (10444083) ……………………………….. ………..…………………………… DR. NKECHI S. OWOO DR. EDWARD NKETIAH AMPONSAH University of Ghana http://ugspace.ug.edu.gh ii DEDICATION I dedicate this work to my brother Mr. Walter Awalime and sister Mrs. Cynthia Boateng whose love and support is without measure. University of Ghana http://ugspace.ug.edu.gh iii ACKNOWLEDGEMENT Special recognition and thanks to my Father and Lord Jesus Christ for health, strength and insight to reach this height in my life. The completion of my thesis would have been with much difficulty without His constant provision. My sincere gratitude goes to my supervisors, Dr. Mrs. Nkechi S. Owoo and Dr. Edward Nketiah-Amponsah for their immense support and constant encouragement till the completion of this work. I could not have done it without their valuable contributions and great supervision, for which I am grateful. Special thanks also to my family; Mr. Cephas Kofi Awalime, Mrs. Grace Emefa Kanda-Awalime, Mr. Walter Awalime and Mrs. Cynthia Addo-Boateng for their constant prayers, encouragement and financial investments in my education thus far. God bless you all. I wish to express my heartfelt gratitude to Mrs. Patience Bedu of the La General Hospital and Mr. Alfred Manyeh of the Dodowa Health Research Centre for their immense support from start to finish of this work. Also, I am grateful to the Noguchi Memorial Institute, Post-Doctoral Office for awarding me a thesis grant to undertake my field work. It would have been an uphill without their support. Finally, I wish to appreciate all my colleagues and friends in the department who have made this journey very fulfilling and worthwhile. May God Almighty bless you all. University of Ghana http://ugspace.ug.edu.gh iv ABSTRACT Ghana experienced its worst cholera outbreak in the last three decades in 2014. However, evidence of the economic costs of the disease on the health system and affected households has not been fully documented. This study therefore sought to determine economic cost associated with recent outbreak the disease. Two districts which were high and low incidence areas (HIA and LIA) were selected for comparative cost studies. A total of 418 (282 HIA and 136 LIA) households which experienced at least one case of cholera infection were interviewed. The Cost of Illness (COI) method was adopted to identify and cost the disease at household and health facility levels. The Tobit Regression model was employed to determine the correlates of cholera incidence. STATA 13 software was used for analysis. The average direct cost to households in the HIA amounted to GH¢342.00 (USD 106.88), almost 2 folds higher than in the LIA GH¢198.47 (USD 62.02). Total cost saving of the episode of cholera was GH¢317,444.10 (USD 99,201.28) in the LIA but rose to almost 8 folds more in the HIA (GH¢2,504,357.12; USD 782,611.60). Total cost of treatment in health facilities were above 10 folds more in the HIA (GH¢66,745.00; USD 20,862.58) compared with LIA (GH¢6,300 USD 1,968.75). Households with the lowest income category had the greatest incidence of cholera (0.073) compared to other categories and drinking water sources such as from neighbouring homes as well as marital status of individuals were significant in the incidence of cholera. University of Ghana http://ugspace.ug.edu.gh v Table of Contents Page DECLARATION.................................................................................................... i DEDICATIOIN ..................................................................................................... ii ACKNOWLEDGEMENTS ................................................................................ iii ABSTRACT .......................................................................................................... iv List of Tables ...................................................................................................... viii List of Figures ....................................................................................................... ix List of Abbreviations .............................................................................................x CHAPTER ONE ....................................................................................................1 INTRODUCTION..................................................................................................1 1.0 Background of Study .......................................................................................... 1 1.1 Problem Statement .............................................................................................. 6 1.2 Objectives of the Study ..................................................................................... 10 1.3 Relevance of the Study ..................................................................................... 10 1.4 Organization of the Study ................................................................................. 12 CHAPTER TWO .................................................................................................14 OVERVIEW OF CHOLERA IN GHANA ........................................................14 2.0 Introduction ....................................................................................................... 14 2.1 History of Cholera............................................................................................. 15 2.2 Clinical Features of Cholera ............................................................................ 16 2.3 Transmission of Cholera ................................................................................... 17 2.4 Cholera Treatment ............................................................................................ 18 2.5 Prevention of Cholera ....................................................................................... 19 2.6 Cholera Epidemic and Outbreak ....................................................................... 19 2.7 Cholera in Ghana .............................................................................................. 20 2.7.1 Historical Trends ...................................................................................... 21 University of Ghana http://ugspace.ug.edu.gh vi CHAPTER THERE .............................................................................................27 LITERATURE REVIEW ...................................................................................27 3.0 Introduction ....................................................................................................... 27 3.1 Theoretical Foundation ..................................................................................... 27 3.1.1 Economic Consequences of Illness on Households .................................. 29 3.1.2 Health as component of Human Capital ................................................... 32 3.1.3 Application of Grossman Model to Study ................................................. 33 3.2 Cost of Illness Approach (COI) ........................................................................ 34 3.2.1 Costs Definitions ....................................................................................... 36 3.2.2 Healthcare evaluation costs concepts ....................................................... 37 3.3 Theoretical Framework ..................................................................................... 41 3.3.1 Direct Cost ................................................................................................ 43 3.3.2 Indirect Cost.............................................................................................. 44 3.3.3 Intangible Cost .......................................................................................... 48 3.4 Empirical Review.............................................................................................. 48 3.4.1 Costs Methods in the measurement of Cholera ........................................ 50 3.5 Cholera, Portable Water, Sanitation and Poverty linkages ............................... 56 3.6 Summary of Chapter ......................................................................................... 59 CHAPTER FOUR ................................................................................................60 METHODOLOGY ..............................................................................................60 4.0 Introduction ....................................................................................................... 60 4.1 Characteristics of Study Area ........................................................................... 60 4.1.1 Study Sites ................................................................................................. 62 4.2 Data and Sampling ............................................................................................ 63 4.3 Data Collection Procedure ................................................................................ 64 4.4 Ethical Considerations ...................................................................................... 64 4.5 Theoretical Framework ..................................................................................... 65 4.6 Methods of Analysis ......................................................................................... 68 4.6.1 Direct Costs .............................................................................................. 69 4.6.2 Indirect Costs ............................................................................................ 70 4.6.3 Model Specification (Objective 1)............................................................. 73 4.6.4 Model Specification (Objective 2)............................................................. 76 4.6.5 Marginal Effects and Cross Tabulations .................................................. 78 University of Ghana http://ugspace.ug.edu.gh vii 4.6.6 Selection of Variables ............................................................................... 79 4.7 Methodological Limitations .............................................................................. 84 CHAPTER FIVE .................................................................................................85 PRESENTATION AND DISCUSSION OF RESULTS ...................................85 5.0 Introduction ...................................................................................................... 85 5.1 Demographic Characteristics ............................................................................ 85 5.1.1 Cholera Infection ...................................................................................... 86 5.1.2 Infection; Household heads, Sex, Age, Education and Income level ........ 87 5.1.3 Infection; Location, Access to portable water sources and Access to Toilet facility 88 5.2 Direct and Indirect Costs of Cholera to Household .......................................... 89 5.2.1 Direct Cost of Cholera within Higher and Lower Incidence Areas .......... 90 5.2.2 Indirect Cost of Cholera within Higher and Lower Incidence Areas ....... 94 5.3 Facility Costs in High and High and Low Incidence areas ............................... 99 5.4 Correlates of Cholera incidence using Tobit Regression Model .................... 100 5.5 Summary of Chapter ....................................................................................... 105 CHAPTER SIX ..................................................................................................107 SUMMARY, CONCLUSION AND POLICY RECOMMENDATIONS .....107 6.0 Introduction ..................................................................................................... 107 6.1 Main findings of study .................................................................................... 107 6.2 Contributions of the Study and Policy Recommendations ............................. 108 6.3 Limitations of the study and areas for further research................................... 110 References ...........................................................................................................111 APPENDICES ....................................................................................................126 APPENDIX A: Sample Size Calculation ................................................................... 126 APPENDIX B: Test for Multicolinearity ................................................................... 127 APPENDIX C: Descriptive Statistics ......................................................................... 127 APPENDIX D: Cholera Infection Cross Tabulations ................................................. 132 APPENDIX E: Health Facility Questionnaire ............................................................ 133 APPENDIX F: Household Questionnaire ................................................................... 136 University of Ghana http://ugspace.ug.edu.gh viii List of Tables Page Table 1.1 Cholera Cases in Ghana by Region by Week, 2014 ................................6 Table 2.1 Cholera cases and deaths by region, Ghana 2014 ..................................23 Table 3.1 Illustrative health policy questions addressed by economic impact studies. ...................................................................................................................32 Table 4.1 Variables and their expected signs .........................................................83 Table 5.1 Descriptive Statistics of dummy and continuous variables ...................86 Table 5.2 Direct Costs in High and Low Cholera Incidence area .........................91 Table 5.3 Individual and Household Direct Average Costs for HIA .....................92 Table 5.4 Days missed by Patients and Caregivers with Indirect Costs in High and Low incidence areas ...............................................................................................96 Table 5.5 Tobit Regression Output showing coefficients, marginal effects for censored sample, standard errors and p-values. ...................................................102 University of Ghana http://ugspace.ug.edu.gh ix List of Figures Page Figure 1.1 & 1.2 Overview of Cholera cases and deaths in WHO African sub- region .......................................................................................................................4 Figure 2.1Cholera cases within 16 districts of the Greater Accra between 2012 and 2014. ................................................................................................................22 Figure 2. 2 Cholera in Ghana: 1980 to 2014 ..........................................................25 Figure 2.2 Distributions of Cholera Cases by Region, Ghana 2014. .....................26 Figure 3.1 Conceptual Framework for Costs of Cholera. ......................................42 Figure 5.1 Household Direct Average Costs in High and Low Incidence Areas ..94 Figure 5.5 Total Costs in High and Low cholera incidence areas. ........................97 Figure 5.6 Per capita cost in High and Low incidence area. ..................................98 Figure 5.1 Total Cost of 2014’s Cholera Outbreak in High and Low Incidence Area. .......................................................................................................................99 University of Ghana http://ugspace.ug.edu.gh x List of Abbreviations CFR Case Fatality Rate COI Cost of Illness GHS Ghana Health Service GSS Ghana Statistical Service HIA High Incidence Area LIA Low Incidence Area UNEP United Nations Environmental Programme WHO World Health Organization University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE INTRODUCTION 1.0 Background of Study Economic measurements of disease complement clinical and epidemiological approaches in assessment of disease burden. Economic analysis seeks to address a number of policy questions on consequences of disease or injury (WHO, 2009). Such questions normally fall within either microeconomic or macroeconomic remit of a study. Under microeconomic analysis, impact of illness is measured on household income or a firm’s profit while the macroeconomic analysis aggregates impact of the disease on the country’s Gross Domestic Product (GDP) or its future growth prospects. Such estimates for diminished health status due to specific illnesses can usefully inform decision makers about the overall magnitude of economic losses and their distribution across a number of key drivers or categories of costs. These costs can ultimately be viewed as benefits from health improvements in the form of avoiding adverse health effects. Morbidity and mortality measures provide key features in estimating disease burden but such considerations are limited in accounting for health ‘shocks’ such as unexpected increases in health expenditure, reduced functional capacity and lost income and productivity which are primary risk factors for impoverishment (WHO, 1999; Xu et al., 2003). Ultimately, economic burden studies may help identify possible strategies for reducing the cost of disease via appropriate action, prevention or treatment strategies. University of Ghana http://ugspace.ug.edu.gh 2 Cholera still remains a major public health concern in many parts of the world but especially in developing countries. However, it seems to receive attention only with the on-set of an outbreak where large populations are affected. In 2014, there were 58 public health events within the World Health Organization (WHO) African Region and out of these, infectious diseases formed 95% of all these events with Cholera being the most frequently reported (31%), ahead of Ebola (13%) which has seen its biggest epidemic in history recently (WHO, 2014). According to the WHO, there were 101,987 cholera cases of which 1,881 resulted in deaths giving rise to a Case Fatality Rate (CFR) of 1.8% within the African Sub region (WHO, 2014). These reported cases were more than double that of the previous year 2013. In total, 16 countries reported cholera cases of which Ghana was the second most affected country only topped by Nigeria. Ghana, Nigeria plus DR Congo accounted for 85% of all cases reported in 2014. Figures 1 and 2 give details of distribution of cholera cases across the African sub-region during the 2014 outbreak. Worldwide, there have been many advances in methods and development of new drugs and vaccines before the turn of the 20th century, despite this, there has been an apparent increase in the emergence of many new as well as more ancient infectious diseases throughout the world of which cholera is one (Smith et al., 2005). Cholera was eradicated from the developed world well over five decades now but in Africa, poor sanitation as well as unsafe drinking water sources driven by poverty creates huge causal linkages between the frequent incidences and easy spread of the disease. According to United Nations Environmental Programme University of Ghana http://ugspace.ug.edu.gh 3 (UNEP, 2005), Cholera though an ancient disease, is at present, one of the most important resurgent diseases plaguing the world. Cholera is an acute intestinal infection caused by the ingestion of food or water contaminated with the bacterium Vibrio Cholera serogroups O1 or O139. It has a short incubation period and produces watery diarrhea in patients that can quickly lead to severe dehydration. The disease condition is nearly unrivaled in terms of the speed with which it kills. Healthy adults can die in as little as three to twelve hours after the first showings of symptoms if treatment is not promptly administered (Briggs, 2002). University of Ghana http://ugspace.ug.edu.gh 4 Figure 1.1 & 1.2 Overview of Cholera cases and deaths in WHO African sub- region Source: WHO, 2014 Outbreak Bulletin Vol. 6 Cholera remains a serious public health disease because outbreaks can spread easily over large regions and quickly overwhelm the health services system. Despite these facts, approximately 80 percent of all persons infected with the cholera never develop symptoms. However, the bacterium stays in their fecal matter for 7 to 14 days, leaving others at risk of infection when proper fecal Fig1. Geographic distribution of cholera cases in the WHO African Region, 2014 Fig2. Distribution of cholera cases and deaths by country, January – September 2014 35996 28944 22203 6421 3355 2059 832 582 485 309 262 235 213 44 35 11 755 243 372 167 185 80 12 9 13 9 11 12 3 0 9 2 0 5000 10000 15000 20000 25000 30000 35000 40000 Nigeria Ghana DR Congo South Sudan Cameroon Niger Benin Burundi Namibia Uganda Togo Cot d'Ivoire Angola Liberia Kenya Guinea-Bissau Deaths Cases University of Ghana http://ugspace.ug.edu.gh 5 matter disposal and personal hygienic practiced are not observed. Also, of all those who develop symptoms, up to 80 percent only develop mild to moderate watery diarrhea while between 10 and 20 percent develop severe watery diarrhea. Case fatality quickly rises to as high as 50 percent if symptoms are left untreated (WHO, 2014). “Cholera represents an estimated burden of 1.4 to 4.3 million cases, and 28,000 to 142,000 deaths per year worldwide” (WHO, 2015). However, WHO claims that there could be as much as 100,000 to 120,000 deaths due to cholera every year but countries normally fail to report actual numbers because of fear of external implications on their economics in sectors like trade and tourism. Coupled with this are poor surveillance systems and weak public health systems in most developing countries which are normally the worst hit regions. Such estimates are supported by Ali et al. (2012) who has reported the same case numbers annually worldwide. In addition to human suffering caused by Cholera, outbreaks of the disease cause panic, disrupt social and economic structures of families and can impede development in affected communities. In some instances, it may result in panic- induced reactions by other countries that curtails or restricts trade and travel from countries where the outbreak is occurring. For example, in 1991 the outbreak in Peru cost the country US$ 770 million due to food embargoes and adverse effects on tourism (WHO, 2014). University of Ghana http://ugspace.ug.edu.gh 6 Outbreaks may also result in catastrophic health expenditures by affected households in addition to over-stretching health system services with the rapid spreading of the disease. In Ghana, this was particularly evident in the 2014 outbreak where some health facilities reported overwhelming number of cases daily. In some of these facilities like the La General Hospital, the Out-Patient- Department (OPD) had to be converted into a makeshift ward for cholera patients. Table 1.1 Cholera Cases in Ghana by Region by Week, 2014 Source; WHO Country Office Ghana: Situation Report on Cholera Outbreak in Ghana as of 05 October (Week 40) and 28 December, 2014 (Week 52). 1.1 Problem Statement In parts of Ghana, cholera has now become endemic and the country has been experiencing outbreaks of the disease about every five years since 1970 (Amankwah, 2011). In June 2014, the country reported its first 6 cases of the disease in the Greater Accra region. Within two weeks the number of reported cases had risen above 250 and began spreading to other regions. Impoverished 1-Jan to 13-Jul 20-Jul 27-Jul 3-Aug 10-Aug 17-Aug 24-Aug 31-Aug 7-Sep 14-Sep 21-Sep 28-Sep 5-Oct 1-Jan to 5 Oct 1-Jan to 28-Dec W1-28 W29 W30 W31 W32 W33 W34 W35 W36 W37 W38 W39 W40 W1-40 W1-W52 Ashanti 0 0 0 0 0 0 30 0 0 130 13 173 287 Brong Ahafo 0 0 0 1 4 1 4 19 20 17 27 16 67 176 1056 Central 0 6 2 22 148 155 160 262 306 487 163 1711 3846 Eastern 0 16 50 92 107 154 117 125 180 145 165 180 27 1358 1875 Greater Accra 99 251 436 947 1873 1640 2188 2386 2044 1745 1425 1099 837 16970 20199 Northern 0 0 0 0 0 0 0 0 0 0 0 0 0 0 282 Upper East 0 0 0 0 0 0 0 0 3 0 2 5 22 32 294 Upper West 0 0 0 0 0 0 0 0 0 1 1 5 7 36 Volta 0 0 0 33 6 57 39 69 33 32 16 72 357 651 Western 0 0 4 2 8 7 13 17 12 47 28 31 2 171 429 Ghana 99 273 492 1097 2146 2014 2551 2878 2598 2604 1840 1408 955 20955 28955 Cases Region University of Ghana http://ugspace.ug.edu.gh 7 communities within the capital city of Accra which lack adequate plumbing and where open defecation is normally rampant appeared to be the epicenter of the country’s outbreak of Cholera. By the close of 2014, a cumulative total of 28,955 cases with 243 deaths and a CFR of 0.8% were recorded. All ten regions in Ghana reported cases with 70% of all the cases from the Greater Accra region alone. The economics surrounding Cholera have always shown poverty, social amenities like portable water and poor sanitation plus personal hygiene nexus. A desk study carried out by the Water and Sanitation Program (WSP) estimated that poor sanitation costs Ghana GH¢420 million (USD 290 million) each year. This sum is equivalent of USD 12 per person per year and forms 1.6% of the national GDP (WSP, 2012). Usually, the highest economic burden falls disproportionately on the poorest and costs of poor sanitation are inequitably distributed. For example, the average cost associated with poor sanitation, constitutes a much greater proportion of a poor person’s income than that of a wealthier person. The Water and Sanitation report on the economic impact of sanitation also reports that the poorest 20% of the population are 22 times more likely to practice open defecation than the wealthiest 20% of the population. For the poorest therefore, poverty cuts in two ways; not only are they more likely to have poor sanitation but they have to pay proportionately more for the negative effects that are resultant from these practices. “An efficient and hygienic method of human waste disposal available in a dwelling unit is a critical indicator of the sanitary condition of the unit and is an indirect measure of the socio-economic status of a household” (GSS, 2013 p.388). University of Ghana http://ugspace.ug.edu.gh 8 In 2010, public toilet was the highest reported facility in urban localities while in rural areas the highest was the use of bushes, fields and beaches which are all forms of open defecation. Public toilets were used by 38.4 percent of urban dwelling units and 29.8 percent in rural localities (GSS, 2013). These statistics are not entirely reflective of the cholera incidence situation in Ghana; because the largest numbers of cases are normally reported in urban areas which equally lack sanitation facilities as rural areas and unhygienic practices such as open defecation are also common. One intractable challenge of both urban and rural areas in Ghana is adopting modern and hygienic solid waste disposal systems. People are stuck to traditional ways of disposing off refuse such as burying and burning. A modern method such as collection by waste management companies is normally shorn especially by poor communities because of cost implications. Appropriate waste management helps to prevent the spread of infections including cholera and improves the quality of the environment. Poor sanitation however undercuts households and communities economically causing sanitation-related diseases burden, force adults out of work, children out of school, lowering productivity and household incomes. The burden of Cholera during an outbreak is easily noticeable with increases in diarrheal related cases at health facilities, increased deaths of patients presenting symptoms and a rapid spread of the disease to other localities. Despite these observable features, without estimation of real costs, the disease burden of Cholera will remain mostly observatory without much ‘teeth’ to show actual University of Ghana http://ugspace.ug.edu.gh 9 impact of the disease. Costs of the disease normally comprise epidemiological costs where government health systems and other agencies must respond quickly in curbing the outbreak. WASH response to Ghana’s cholera outbreak is estimated to be USD 1.2 million each year (WSP, 2012). Productivity losses are also significant due to either debilitating impact or premature death of patients as well as losses attributed to their caregivers. Outbreaks also lead to the diversion of expenditures from other essential sectors which have more direct impact on the economic development of communities rather than outbreak management. Trade and tourism is normally hampered by Cholera because it is a disease associated with flirt plus being communicable. Hence, the government experiences revenue losses due to reductions in both trade and tourism. The economic costs of Cholera in Ghana; direct, indirect or intangible have received very little attention despite the fact that the government, households and health facilities are all impacted at various levels and in various degree by the disease. Though most facilities have data on costs incurred, these usually cover only direct costs and in most instances these costs are shared costs for all services rendered at the facility without conscious effort in segmenting them for various disease impact. Additionally, costs borne by households in seeking care are completely missed and in this situation, the true cost of the disease is lost or underestimated because these key components are excluded. University of Ghana http://ugspace.ug.edu.gh 10 Additionally, the suddenness of Cholera’s emergence and its ease of spread can quickly drive both households and the whole health system into catastrophic expenditure. This situation is a cause for concern to all stakeholders involved. 1.2 Objectives of the Study The study is to generally assess the economic costs of 2014’s outbreak of Cholera on households and health facilities in the Greater Accra region of Ghana. Specifically the study seeks; • To estimate comparative costs in one low and one high cholera incidence area. • To determine correlates of cholera affected households with key interest on income quintiles plus other characteristics. 1.3 Relevance of the Study The relationship between socioeconomic status and health indicators has been widely studied, and it is recognized as a cause-effect relationship and access to social services such as education, housing, food, as well as equitable income distribution are important determinants of individual health conditions. Research on cholera in Ghana has focused more on determinants and risk factors of outbreaks, without much attention to economic costs of such outbreaks. Some of these studies include; (de Magny et al. (2006) who observe the relationship between disease’s temporal patterns and climate; Osei et al. (2010) studied spatial dependency of cholera prevalence on potential cholera reservoirs in Kumasi. These two studies reviewed the environmental factors influencing cholera University of Ghana http://ugspace.ug.edu.gh 11 incidence in particular locations. Thompson et al. (2011) sought to understand the evolution of the cholera epidemic in Ghana by analyzing a collection of Cholera strains from the beginning of the seventh cholera pandemic in Ghana. To the best of my knowledge and literature searches done, there is no known research on Ghana measuring economic costs of cholera outbreaks even though outbreaks have now become nearly perennial during every rainy season. The closest study of this kind was one done by Aikins et al. (2010) which measure health care costs of diarrheal disease in Northern Ghana. Studies of this nature however, lack the specificity needed in assessing disease-specific interventions which normally is the central focus of most costing studies. In addition, studies such as this that mainly present a provider based perspective greatly limit the study by ignoring the burden that disease pose to families. These all buttress the point of an existence of an important gap in the literature on Ghana especially with the regard of economic perspective on cholera. On the African front, Kirigia et al. (2009) proposed the urgent need for further studies to determine localized- level economic burden of cholera. The present study aims to fill the gap and increase global knowledge on costing of cholera in Ghana. This shall highlight a number of relevant policy implications and applications; • Cholera is an acute disease and can quickly lead to death within hours when left untreated. But up to 80% of cases can be successfully treated with Oral Rehydration Salts (ORS) (WHO, 2014). Most cases of deaths result from late arrival of patients at the health facilities to seek care. University of Ghana http://ugspace.ug.edu.gh 12 Knowledge in household’s cost drivers and importance of economics in early healthcare seeking is immensely critical in reducing CFR. • Households’ economic status can have an important influence on lifestyle, prevention and treatment of cholera. The a-priori expectation is for households that are relatively wealthier to live healthier life-styles and hence having lesser infection rates or faster recoveries from the disease. Having knowledge on the distributional effects of cholera on various income quintiles can guide policy makers in using targeted approaches to alleviate the burden of the disease where it is heaviest felt. • During emergencies like a cholera outbreak, there is increased collaboration between health agencies like the Ghana Health Service (GHS), Ministry of Health (MOH), WHO, UNICEF, the Red Cross, etc. In these collaborative efforts it is important that such organizations know the severity of the disease not just in human suffering but economic consequence within affected localities. This guides the channeling of collaborative efforts in combating such outbreaks. 1.4 Organization of the Study The study is structured into six chapters. Chapter one gives a general introduction, statement of problem, objectives, relevance, methodology and the organization of the study. Chapter two looks at cost of illness study, the history of the disease and its implications. Chapter three reviews existing literature on the both theoretical and empirical studies done on cholera. Chapter four profiles the two districts being studied, the theoretical framework and the empirical model that underpin University of Ghana http://ugspace.ug.edu.gh 13 analysis of data. Chapter five presents analysis and discussions of the data collected and entered from the field. Finally chapter six deals with summary of findings, recommendation and policy implications of the study. University of Ghana http://ugspace.ug.edu.gh 14 CHAPTER TWO OVERVIEW OF CHOLERA IN GHANA 2.0 Introduction Industrialization and development is lessens poverty and makes access to basic social amenities such as safe drinking water and sanitation improvements a reality. Sanitation and water are two key features in in the fight against infectious diseases such as Cholera. Due to these same features resulting from development, cholera has been eradicated from the developed world for over half a century now. Sadly, the situation seems to be worsening in the developing world and has been aided by proliferation of urban slumps, poor housing conditions, poor sanitary practices like open defecation among other factors. Cholera has been evidently predominant in communities or countries where poverty levels are still quite high. A study done by Talavera & Pérez (2009) showed that countries with low Gross National Income per capita reported the highest numbers of cholera cases in contrast with middle and high income per capita countries that reported 34% folds lower incidence. This underscored the statement that “cholera is a disease of poverty”. Cholera reached Africa in the 1970’s, and since then has become endemic in many African countries especially causing many deadly, multinational epidemics in many sub-Saharan countries. The incidence has also shown a growing seasonal pattern with onset of rainy seasons where poor sanitary conditions within these countries are exposed including large prevalence in refugee camp situations within the region which are plagued by many wars displaced persons. University of Ghana http://ugspace.ug.edu.gh 15 In 2005, 31 (78%) of the 40 countries that reported indigenous cases of cholera to WHO was in sub-Saharan Africa. The reported incidence of indigenous cholera in sub-Saharan Africa in 2005 (166 cases/million population) was 95 times higher than the reported incidence in Asia (1.74 cases/million population) and 16,600 times higher than the reported incidence in Latin America (0.01 cases/million population). In that same year, the cholera case fatality rate in sub-Saharan Africa (1.8%) was 3 times higher than that in Asia (0.6%); no cholera deaths were reported in Latin America. 2.1 History of Cholera There have been records of cholera-like diseases as far back as in the time of Hippocrates (460-377 BC) and Galen (129-216 AD) and numerous cholera-like maladies were also known in the plains of the Ganges River since antiquity. However, modern knowledge about cholera dates only from the beginning of the 19th century when researchers such as John Snow began to make progress towards a better understanding of the causes of the disease and its appropriate treatment. But since that period great advances have been made in identifying root causes, spread, treatment and prevention of the disease. So far, throughout history there have been 7 cholera pandemics starting from 1817 and which have always originated from Asia and spreading to the rest of the world. The first six pandemics all seem to have originated from Ganges in Bangladesh, and are thought to be caused by infection of the Classical O1 biotype strain of the disease (Laws, 2006). In 1961 the current cholera pandemic started and unlike the previous six which began in Bangladesh this began in Indonesia University of Ghana http://ugspace.ug.edu.gh 16 spreading through Asia to Africa, Europe and finally to Latin America and resulting in large number of cholera cases and deaths per year since that time (Colwell, 2002). Currently, the disease affects 3 to 5 million people every year predominantly in Africa and Asia (WHO, 2010). The pandemic reached sub-Saharan Africa in 1970 and has remained entrenched ever since. Latin America recorded its first cases in 1991 and within 3 years the number of recorded cases had risen to nearly 1 million. Though there was a great explosion of the disease within the first few years of it reaching that region, in contrast to Africa, cholera was largely eliminated from Latin America within a decade but in Africa it remains a persisting problem. 2.2 Clinical Features of Cholera Cholera is an acute diarrheal illness caused by infection of the intestine with the bacterium Vibrio cholerae serogroup O1 or O139. Infection can be asymptomatic, mild, or severe: approximately 1 in 20 infected persons have severe disease characterized by profuse watery diarrhea, vomiting, and leg cramps. In these persons, rapid loss of body fluids leads to dehydration, electrolyte disturbances, and hypovolemic shock. Without treatment, death can occur within hours but treatment is simple and inexpensive for most cases. In severe cases, vigorous oral or intravenous fluid and electrolyte replacement is key to recovery and overall mortality can be reduced from as much as 50% to less than 1%, even in makeshift rural treatment centers. University of Ghana http://ugspace.ug.edu.gh 17 2.3 Transmission of Cholera Two routes of transmission of cholera have been recognized; the first one occurs from aquatic reservoirs in the environment (primary transmission) and the second occurs from previously infected individuals (secondary transmission); once the primary transmission has initiated an outbreak, secondary transmission depends the problem by causing epidemics in the endemic areas (Ruiz-Moreno et al. 2010). Patients infected who have no symptoms will generally shed-off the organism for only a few days; however, patients who are symptomatic shed-off the organism between 2 days and 2 weeks (Nelson et al. 2009; Weil et al. 2009). Transmission of cholera within households has been documented (Weil et al. 2009) this increases the risk of infection among household members. Vibrio cholerae are present in human stool and when proper fecal disposal is not practiced there is the likelihood of pollution of water sources which are used by poeple (Nelson et al. 2007; Faruque et al. 2006). In environmental water, organisms convert to conditionally viable environmental cells within 24 hours (Nelson et. al. 2008). These organisms are infectious upon reintroduction into humans, although the infectious dose in this form is not known. Thus, the major source of Vibrio cholerae is feces of persons acutely infected with the organism that reaches water most often through sewage. Individuals with reduced gastric acidity and blood group O are more susceptible to the infection, and in situations where poor environmental sanitation is coupled with poor domestic and personal hygiene, transmission results from ingestion of faecally contaminated water (as well as University of Ghana http://ugspace.ug.edu.gh 18 food), and hence it is usually a disease of developing countries or areas where improved water and adequate sanitation are lacking. In endemic areas, water is usually the main vehicle of transmission, although this may occur through food, and thus infection due to Vibrio cholerae begins with the ingestion of contaminated water or food. Transmission of cholera in non-endemic areas is more commonly associated with consumption of foods, such as raw or undercooked seafood, imported from cholera-endemic regions and also due to migration of persons from affected regions. 2.4 Cholera Treatment Without treatment the case-fatality rate for severe cholera is about 50%. However, treatment is very effective and simple and is based on the concept of replacing fluids as fast as they are being lost. It is required that replacement fluids be similar to electrolyte composition of the fluids being lost. Initially, the fluids must be given sufficiently and rapidly to make up for the volume that has already been lost so that blood circulating volume can be restored. Additionally, maintenance fluids are given to replace continuing losses as they occur this will forestall any deaths when fluids are given promptly and efficiently. Normally, signs of dehydration are not clinically apparent until the patient has already lost about 5% of his or her bodyweight. A patient with severe dehydration requires emergency intravenous polyelectrolyte solution for rehydration followed by oral rehydration solution (ORS) for maintenance hydration. In milder cases, ORS can be used for both rehydration and maintenance. Patients who are severely University of Ghana http://ugspace.ug.edu.gh 19 dehydrated are assumed to have lost 10% of their bodyweight, and this is the volume that needs to be replaced. There are cases where patients have no pulse but in such situations patients are to be given fluids as rapidly as possible and more than one intravenous line may be needed to infuse the fluid rapidly enough to restore the pulse. In patients with lesser degrees of dehydration, the simple use of ORS is effective for rehydration. 2.5 Prevention of Cholera Contaminated food and water are normally the main routes for infection and much can be done to keep transmission rates to a minimum. The measures include ensuring a safe water supply, improving sanitation, making food safe for consumption by thorough cooking, and health education through mass media. Some important messages for the media during outbreaks include the importance of purifying water and seafood, washing hands after defecation and before food preparation, recognition of the signs of cholera, and locations where treatment can be obtained to avoid delays in case of illness. Filtration of water through sari cloth is quite effective and reduces cholera transmission by nearly 50% (Colwell et al. 2003). The long-term prevention of cholera will require improved water and sanitation facilities, but these improvements are not happening rapidly in most regions where cholera is prevalent because many needs usually ‘juggle’ for the same limited resources. 2.6 Cholera Epidemic and Outbreak Epidemics or explosive outbreaks generally occur in underdeveloped areas with inadequate sanitation, poor hygiene, and limited access to safe water supplies, University of Ghana http://ugspace.ug.edu.gh 20 whereas in some countries, a seasonal relation for cholera epidemics has been observed (Koelle et al. 2005; Faruque et al. 2005). WHO described a dramatic increase in the number of cholera cases and outbreaks, in new communities and in communities where the disease had been absent for emerging disease. 2.7 Cholera in Ghana In West Africa, Guinea was the first country to report cases of cholera and though the government initially denied it, the WHO had to break protocol and establish for the first time that “the health of the world's people is more important than the sovereignty of member countries” (Ofori-Adjei & Koram, 2014). The first case of cholera in Ghana occurred on September 1, 1970 when a Togolese national in transit from Guinea collapsed at the Kotoka International Airport and was later diagnosed with cholera (Pobee & Grant 1970). Two of the worst hit sites in Ghana, subsequently, were the fishing villages of Akplabanya (in the then Ada District) and Nyanyano (Winneba District). Cholera in these areas appeared to have been “smuggled in” by relatives of dead Ghanaian fishermen from Togo and Guinea, respectively, for burial despite a sanitary cordon on Ghana's borders. By July 1971, the Ashanti region began to report cases, indicating that cholera had spread across the country (Ashitey, 1994). Since then there has been focal outbreaks every 4 to 6 years. In recent years outbreaks have become more frequent and protracted and now believed to be endemic in Greater Accra, Central and Western Regions. A pilot study by GHS (2014) within Accra Metropolis found that 3 out of every 10 University of Ghana http://ugspace.ug.edu.gh 21 diarrhea cases proved positive for Vibrio cholera stamping rife of the disease. Attempts at controlling cholera have not been successful because needed long term approaches, such as potable water supply, proper disposal of solid waste etc. have been grossly inadequate and inequitably distributed to capture disadvantaged populations. Cholera is now endemic parts of Ghana with cyclical epidemics. These epidemics are now predictable but sanitary reforms have been inactive, ineffective and local authorities have failed in applying necessary bye-laws on food hygiene, sanitation, environmental health and waste disposal. Cholera in Ghana is mainly an urban problem with high impact on the urban poor. The unprecedented unregulated growth of urban slums has resulted in poor environmental conditions, lack of access to clean potable water and extreme challenges in waste disposal. Hence, the routes of cholera problem Ghana is well- known. 2.7.1 Historical Trends With the current outbreak inclusive, there have been 5 major outbreaks in Ghana since 1980 and the current outbreak being the worst ever to affect the country. Figure 2.1 below shows incidences of outbreaks from 1980 to 2014 giving a picture of its resurgence every 4 to 6 years. Most of the outbreaks have normally shown a steady rise of the outbreak from the previous year but the 2014 outbreak had a sudden trend rising from an almost zero cases in the previous year to become the worst outbreak in 30 years in a short span of six months. Out of the 5 major outbreaks, 3 recorded over 10,000 cases of the disease. The years between 1985 and 1989 was the longest period recorded with less than 200 cases. University of Ghana http://ugspace.ug.edu.gh 22 However, the years 1988, 1998 and 2013 had the lowest number of cases in the history of the disease in Ghana, reporting less than 20 cases. Finally, apart from the year 1998, which had zero cases of the disease, all other years have reported cases of the disease. The Greater Accra region has consistently reported the largest number of cases in the country and within the region Accra Metro, Ga West, Ga South and Ledzokuku Krowor districts have recorded the highest number of cases over the past 3 years. In the recent outbreak, the region recorded 70% of all reported cases and 50% of deaths. Figure 2.1Cholera cases within 16 districts of the Greater Accra between 2012 and 2014. Source: Author’s computation Data Source: GHS, 2015. Regional Disease Control Unit. 10,504 103 1 19 68 152 190 2,082 1,286 280 1,907 770 1,398 32 315 1,092 20,199 0 5,000 10,000 15,000 20,000 25,000 ACCRA METRO ADA EAST ADA WEST ADENTAN ASHAIMAN GA CENTRAL GA EAST GA SOUTH GA WEST KPONE KATAMANSO LA DADEKOTOPON LA NKWANTANANG LEDZOKUKU KROWOR NINGO PRAMPRAM SHAI OSUDOKU TEMA TOTAL 2014 2013 2012 University of Ghana http://ugspace.ug.edu.gh 23 Table 2.1 Cholera cases and deaths by region, Ghana 2014 Cholera Cases and Deaths by Region, Ghana 2014 Region Cases Deaths CFR (%) Ashanti 287 3 1.05 Brong-Ahafo 1056 26 2.46 Central 3868 64 1.65 Eastern 1875 6 0.32 Greater Accra 20199 121 0.6 Northern 282 2 0.71 Upper East 294 9 3.06 Upper West 36 1 2.78 Volta 651 8 1.23 Western 429 7 1.63 Source: GHS, 2015. Regional Disease Control Unit. Table 2.2 shows the distribution of cases in the recent 2014 cholera outbreak by region. Apart from Greater Accra, Central region had the second highest number of cases recording over 3800 cases over the period. Though Greater Accra reported the largest number of deaths over the entire period, it reported the least CFR (0.6) and the Upper East Region which recorded 294 cases had the highest CFR (3.06) in the country. This might be attributable to the relative ease in accessing health facilities, proximity of health facilities and education on the disease which might all be better in Greater Accra compared to the other regions. The cholera outbreak in 2014 recorded 28,975 cases 243 deaths (CFR=0.8%) with all the 10 regions of the country reporting cases. Some triggering factored identified by the GHS as the cause of the cholera outbreak include; Shortage of potable water supply, Breakdown of waste management systems, Poor personal hygiene and poor food hygiene. Accra was the epicenter of the outbreak and all University of Ghana http://ugspace.ug.edu.gh 24 cases reported within the country were due to internal migration. Identifiable challenges in dealing with the outbreak were due to persistence of risk factors (personal and environmental), slow response by stakeholder agencies, overwhelming large number of cases during outbreak and inadequate cholera logistics and treatment centers. Between December 2010 and 2012 nine out of the ten regions of Ghana experienced a protracted outbreak and though very few cases were reported in 2013, 2014 saw an unprecedented surge in cases giving a clear indication that the situation is far from being abated. University of Ghana http://ugspace.ug.edu.gh 25 Figure 2. 2 Cholera in Ghana: 1980 to 2014 Source: Author’s computation Data Source: GHS, 2015. Regional Disease Control Unit. 261 940 11086 15032 1015 62 25 58 17 119 3464 13743 273 1470 2387 4443 1118 64 0 9463 3431 5483 3763 239 517 4387 3353 179 1126 1272 438 9370 9562 18 28975 0 5000 10000 15000 20000 25000 30000 35000 19 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 Fig 2.2 Cholera in Ghana: 1980 to 2014 NO. OF CASES University of Ghana http://ugspace.ug.edu.gh 26 Figure 2.2 Distributions of Cholera Cases by Region, Ghana 2014. Source: GHS, report on cholera outbreak in Ghana, 2014. University of Ghana http://ugspace.ug.edu.gh 27 CHAPTER THERE LITERATURE REVIEW 3.0 Introduction In this chapter, the theoretical foundations of economics of ill-health and its impact on households is presented. The concept of economic cost is defined and the different methods recommended by WHO guidelines to identifying the economic consequences of disease are expounded. The Cost of Illness (COI) method as developed by Rice (1966) and modifications by Shepard et al. (1991) is reviewed and its application to identifying cost components and its estimation. Also, Grossman’s (1972, 2000) model on the demand for health including its application as human capital is discussed. Finally, an empirical review is done in two parts. The first part reviews studies using different cost approaches in measuring costs of cholera and the second part delves into the linkages of poverty, sanitation, access to portable water and the incidence of cholera. 3.1 Theoretical Foundation Health care utilization and health seeking behaviour modeled by Andersen (1995) has been widely used in public health circles. It is assumed by the model that health seeking behavior results from interaction of individual’s characteristics, population and his environment, and that the use of medical service is dependent on three main factors namely; predisposing, enabling and need factors. Predisposing factors are features that influence an individual’s health as a result of certain peculiar behaviour patterns which include demography (sex, marital University of Ghana http://ugspace.ug.edu.gh 28 status, household headship and past ailments); social structure (education, race, occupation and religion); and lastly, belief systems (such as culture). Enabling factors are related to issues of availability, equity, accessibility, quality and costs. These tend to promote, inhibit or limit the use of health services. The third factor; need, encompasses the individual’s health use that is necessitated by disease or impairment such as intensity of illness measured by the number of sick days. Thus, Andersen’s model estimates the use of health care services as a function of these three factors. However, economic theories on the use of health services focus on the how individuals make decisions in consuming health care to produce health as in the case of Grossman (1972) and with further simplifications by Wagstaff (1986). “The theory develops a conceptual apparatus for analyzing the interaction of socioeconomic determinants of health and indicates how this can be used to shed light on a variety of policy issues such as socioeconomic inequalities in health and the design of prevention policies” (Wagstaff, 1986, p.1). Wagstaff hinges his theory on three basic economic concepts; the indifference map, the health production function and the budget constraint. An individual’s indifference map is made up of several indifference curves of which these curves are formed from the individual’s combination of commodities that gives him or her a certain amount of utility. Under the theory, the individual views ‘good health’ as a desirable commodity in addition to other commodities from which utility is derived but the individual does not place an overriding value on ‘health’ relative to the other commodities. The individual will therefore end up with an indifference curve University of Ghana http://ugspace.ug.edu.gh 29 showing a combination of health consumption as well as other goods that brings him or her maximum satisfaction. In the ‘health production function’, it is postulated that an individual produces an output ‘health’ by combing health inputs. These inputs include; education, medical care, food, exercise etc. and the interaction between these inputs in producing the output forms the ‘health production function’. Thirdly, the theory assumes the existence of a budget constraint resulting from costs of health inputs and its consumption process. Meaning the consumption of health inputs is not limitless and the quantity that can be accessed by an individual will be dependent on his or her income and the prices of the inputs. Deducing from these three principles, it can be assumed that an individual’s production of health will depend on factors such as his utility, technical know- how (education) and income. 3.1.1 Economic Consequences of Illness on Households Welfare economic theory suggests that individuals seek to maximize their utility or economic welfare subject to some constraints which include income and time. This is achieved by combining a range of goods and services; some of which can be bought or sold (including health care) and some of which cannot but still have value. Health, unlike other economic goods and services that yield direct utility, has an indirect utility, because consumers of health would have preferred not to incur University of Ghana http://ugspace.ug.edu.gh 30 these expenses in terms of money and or time. Therefore, health contributes to the individual’s utility in three ways. One, people prefer more healthier days to less healthier days. Two, the level of health partly influences the consumption of other goods and services. Three, poor health inhibits economic objectives like making income that allows people to consume market goods. Poor communities normally have high proportion of out-of-pocket health expenditure and ill-health drives up household’s consumption of health goods whilst lowering that of their basic needs such as food, clothing, housing and children’s education. For example, in Germany where per capita GDP is US$ 32,860, households bear 11.3% of all medical expenses and the rest by a social health insurance or by the government. In contrast, in DR. Congo where GDP per capita is only US$120, households bear about 90% of health care costs directly. This situation creates the tendency of deterring poorer people from seeking orthodox health services, truncating treatments or resorting to other unorthodox methods of treatment which sometimes leaves fatal results. Time spent seeking care further inhibits work and lowers financial earnings. Financing healthcare might transcend current income into households’ savings if available or if not, through a loan or sale of family assets. Reduced income, savings and assets will deplete investment and eventually affect future consumption possibilities. Xu et al. (2003, 2007) estimate that 44 million households worldwide are plunged into catastrophic health expenditures (40% of a household’s none-subsistence income on healthcare payments as defined by Xu et al. (2003) and about 25 million pushed into poverty each year. Clearly defined, University of Ghana http://ugspace.ug.edu.gh 31 Xu et al. (2003) states that health expenditure as being catastrophic if a household’s financial contributions to the health system exceed 40% of its income remaining after subsistence needs have been met. Wyszewianski (1986) alludes to the fact that, high medical costs were not necessary synonymous to catastrophic expenditures to households especially when there is the existence of medical insurance policy to cover either all or large portions of the costs. But conversely, small medical bills for common diseases could become financially dangerous to poor households with no insurance cover. A recent WHO report also noted that the likelihood of financial catastrophe is negligible if 15-20% of household’s total health expenditure is out-of-pocket (OOP). However, the average OOP is 40% for African countries, indicating the higher possibility and plausibility of financial catastrophe and resultant poverty. From the aforementioned, Cost of Illness studies try to answer a number of health policy questions under macro or micro levels of the economy. Table 3.1 captures questions at their respective levels and the economic unit these questions apply to. University of Ghana http://ugspace.ug.edu.gh 32 Table 3.1 Illustrative health policy questions addressed by economic impact studies. Source: WHO guide to identifying the economic consequences of disease and injury (2009). 3.1.2 Health as component of Human Capital In the broadest definition of the term asset, economists define an asset as a resource with economic value that an individual, corporation or country owns or controls with the expectation that it will provide future benefit. This asset can either become depleted with time or produce gains. In this sense, health can be viewed as economic asset capable of either depletion with time and other factors like high risk behaviour or enhanced through investments like healthy dietary practices and exercise. Hence health has the capacity to be produced through investments and also viewed as capital consumption due to aging and lifestyle. Health investment can also be seen through either curative or preventive methods. LEVEL QUESTION MACRO 1. What impact does ill-health have on gross domestic product or its rate of growth? Society 2. How much does society pay for medical and other expenses because of illness? 3. What is the impact on social product (i.e., both market and non-market consumption lost opportunities), or on social welfare more generally? MICRO Households 1. What impact does ill-health have on a household's income or consumption patterns (over a single year, or for a longer period of time)? 2. How much do households pay for medical or other expenses because of illness (for an episode, over a year, or over a lifetime)? Firms 1. What impact does ill-health have on a firm's operating costs, output or profit? 2. What is the impact of ill-health on productivity in the work place (including impaired performance while still at work, as well as absenteeism) ? Government 1. What proportion of government expenditure could have been saved and directed to an alternative use in the absence of illness? (e.g. what social security payments could be avoided by the prevention of or cure for disease?) 2. What impact does ill-health have on the government workforce and on the government's ability to provide services? University of Ghana http://ugspace.ug.edu.gh 33 Curative is where upon the onset of ailment, medical services are purchased to ensure the return to good health and preventive in the form of one introducing exercise to lifestyle and healthier eating habits that reduces the risk of certain disease conditions. The return on these health investments is lesser time spent in bad health. The healthy time gained can in-turn increase labour income and thus increase consumption or leisure. In economic theory rational beings prefer more utility to disutility and will seek to maximize this utility. This model is theorized by Grossman (1972) which suggest that given an individual’s initial stock of health inherited from their parents and their environment, the individual will make informed choices regarding their health improvements over their life cycle. The individual does so by investing in healthcare and practicing health enhancing behaviours to maximize utility but will be constrained by a budget. 3.1.3 Application of Grossman Model to Study According to Grossman (2000), the demand for health is split into two based on either for ‘Pure Consumption’ or ‘Pure Investment’ purposes. With pure consumption purpose, the return is mainly ‘psychic’ granted that it reduces disutility in terms of fewer sick days. On the other hand, pure investment in health is mainly for the purposes of increasing healthy days to participate in market and non-market activities. This study assumes based on this model that wealthier persons have greater positive returns on health and will hence face a lesser burden when an outbreak of cholera occurs. Grossman further postulates that price of medical goods have a negative effect on both consumption and investment into health. This is directly linked to factors of affordability and access. Wealthier University of Ghana http://ugspace.ug.edu.gh 34 homes are able to utilize and make investments into their health better than their less wealthy counterparts due to these factors and hence poorer homes continually face greatest burden of diseases. On education, similar arguments are put forward by the model. Stating that with improved education, individuals are more efficient in producing health, that is; more educated people are better off following instructions for medications and or are more knowledgeable about what is bad for their health. It has been empirically observed that population level (local-level) risk factors of cholera include poverty, lack of development, high population density, low education, and lack of previous exposure (Ackers et al., 1998; Ali et al., 2002). Penrose et al. (2010) also suggest that global urbanization plus urban poverty are associated with high risk of infectious diseases such as cholera among other studies. This study hinges on these assumptions and tries to determine the impact of the cholera outbreak based on the costs that result from treatment of the disease and its predisposing factors resulting from the income strata of cholera affected households. The COI method is used to identify costs and a Tobit model in determining the correlates of cholera for these households. 3.2 Cost of Illness Approach (COI) Cost of illness (COI) studies is a type of economic study common in medical literature, particularly in specialist clinical journals (Byford et. al., 2000). It is most usual to these studies when the economics of health is being observed. COI studies seek to identify and measure all costs of a particular disease; direct and University of Ghana http://ugspace.ug.edu.gh 35 indirect as well as intangible dimension in some cases. The determination of total cost of an illness provides two critical useful pieces of information. One, it tells us how much society is spending on a particular disease, and by implication the amount that would be saved if the disease were curtailed. Two, it identifies the different components of cost and the size of the contribution of each sector in society. Such information, it is argued, can help to determine research and funding priorities by highlighting areas where inefficiencies may exist and savings may be made (Rice, 1994; Ament & Evers, 1993). COI was formulated by Dorothy Rice and colleagues in the late 1960’s and subsequently revised (Rice, 1966, 1967; Cooper and Rice, 1976; Rice et al., 1985). A bulk of economic impact studies in health use some version of COI approach and it remains the most common measurement of illness approach. In its original form, measurements of costs associated with pain or suffering were not considered though they were described as ‘intangible costs’. The traditional COI approach estimates the societal impact of disease and injury combining ‘direct costs’ (medical diagnosis, treatment, transportation, follow-ups care, etc.) and ‘indirect costs’ (the value of lost production because of reduced working time). Two methods of COI have evolved; Prevalence and Incidence approaches. The commonest is the prevalence method and it estimates the total cost of a disease incurred in a given year. Incidence approach calculates the lifetime costs of cases first diagnosed in a particular year (Rice, 1994). In this study a prevalence University of Ghana http://ugspace.ug.edu.gh 36 approach is applied for the costs associated with the incidence of cholera within the 2014 outbreak. An important factor for consideration in estimating households’ direct and indirect costs in seeking healthcare is households’ ‘coping strategies’ in mitigating the impact of ‘shocks’. These strategies include adjustments of labour supply within the household to preserve production and income flows, drawing on savings, borrowing, social or health insurance and disposal of assets (Sauerborn et al., 1996). On the other hand, Russell (2004) shows that most of these coping strategies come with future repercussions and might in the long-run be more costly to the household’s livelihood. For this reason, coping strategies though might mitigate impact of disease burden in the short term, have long term effects which are equally problematic for households. Hence coping strategies are concluded as having little to no effect in the long-run. 3.2.1 Costs Definitions Economists define costs as opportunity cost. That is, a forgone benefit in using resources in their next best alternative. This concept is fundamental in economics because resources are scarce and choices need to be made among alternative uses. For example, during the 2014 cholera outbreak, the Out-Patient-Department (OPD) of the La General Hospital became a mid-shift admission ward and all other medical services suspended except emergencies. Hence, the opportunity cost of the outbreak was all other unattended health cases which were less acute. Therefore, opportunity cost presents a picture of competing needs and scaling of priorities. University of Ghana http://ugspace.ug.edu.gh 37 On the other hand, accountants’ costs comprise of only physical expenditure on goods and services such as payments for medical supplies, staff salaries, equipment etc. But accountants’ costs tend to underestimate costs because it does not capture costs of resources being used in other capacities as well as costs such as volunteers’ time and auxiliary services like free treatment of medical staff. Therefore economic costs capture both financial monetary costs as well as non- monetary costs. Costs studies are crucially dependent on the viewpoint of the measurements of interest. As in Table 3.1, economic impact on the whole economy might differ from that on a firm or a household. For example, the effect of disease or injury on a household's income or consumption opportunities (household question 1 in Table 3.1) could be interpreted to include only market production and cash purchases, in which case the quantity of interest is the value of market consumption opportunities foregone because of illness. On the other hand, the quantity of interest could be redefined to include the economic losses associated with both market and non-market production or consumption opportunities. Finally, it could be defined even more broadly to include the value of the lost welfare associated with illness, which would include not just the lost production and consumption, but the value of the changes in leisure and the decrement in welfare. Each of the questions measure different quantities of interest and estimates derived from each are normally viewed separately. 3.2.2 Healthcare evaluation costs concepts • Cost analysis University of Ghana http://ugspace.ug.edu.gh 38 Cost analysis break down a cost summary into its various constituents and investigate the parts of the whole and their relations in making up the whole. It focuses on assessing the costs of providing or consuming a service or intervention. It is useful for assessing the affordability of a programme and for guiding budgetary planning. It also provides useful understanding of incremental costs of interventions and the total resource needed to support current and future strategies. Cost analysis does not consider the effectiveness of an intervention or service. In its application in health economics, it identifies the key drivers of costs for particular diseases and measures its impact at the individual, household, firm, government or the entire society level. It is also relevant to conduct when a new intervention is sort to be introduced like a vaccine to deal with a health related problem. It provides useful information on where costs are heaviest felt and therefore needs policy focus. • Cost-minimization analysis (CMA) Cost-minimization analysis is used when identifying the least cost among alternative choices that produce similar outcomes in terms of benefits or effectiveness. It is a comparison of two or more strategies which have the same effectiveness but which are assumed to have different costs. It aims at achieving ‘value for money’ either through a desire to achieve a predetermined objective at least cost or a desire to maximize the benefit to the population of patients served from a limited amount of resources. It is a useful tool when benefits from alternative measures are considered to be equal. It becomes however challenging especially among human rights activists when priority of ‘value for money’ is placed above saving lives. This aside, this method seeks to achieve the best use of University of Ghana http://ugspace.ug.edu.gh 39 resources in producing maximum outcomes without compromising quality of healthcare. • Cost-effectiveness analysis (CEA) A cost-effectiveness analysis compares relative costs and outcomes/effects of two or more courses of action. The aim is to identify the strategy with the lowest cost per unit of output, or alternatively the strategy that delivers the highest output for a given fixed budget. In the health services field, it is normally useful in situations where it may be inappropriate to monetize health effect. It is expressed in terms of a ratio where the denominator is a gain in health such as years of life, sight years, premature births averted etc. and the numerator, the cost associated with the health gain. The effectiveness indicator is the same for each strategy being compared, and consists of a health outcome measure. • Cost-utility analysis (CUA) Cost-utility analysis is relevant when the aim of the study is to compare alternative health services or interventions that are associated with different costs and different outcomes especially with procurement decisions. In health economics it estimates the ratio between the costs of a health related intervention and the benefits it produces in terms of the number of years lived in full health by beneficiaries. The main distinguishing feature of cost-utility analysis from cost- effectiveness analysis is that it involves measurement of the “utilities” associated with different interventions. This estimation technique is based on “expected utility theory”, also referred to as von Neumann-Morgenstern utility theory, which is a theory of rational decision-making under uncertainty. The existence of University of Ghana http://ugspace.ug.edu.gh 40 uncertainty in the analysis captures the extent to which individuals are risk averse, risk-seeking or risk-neutral. This procedure allows for comparison across different health programmes and policies by using a common unit of measure and it also takes into account the quality of life of the individual has. • Cost-benefit analysis (CBA) Cost-benefit analysis is credited to the French Economist Jules Dupuit. He theorized this analysis by comparing the total expected costs of alternative intervention programmes against its benefits to see whether the benefits outweighed the costs. This technique determines the option that provides the best approach for adoption in terms of benefits or effectiveness. CBA assigns monetary value to health improvements which differ from other cost analysis like Cost Effective Analysis and Cost-Utility Analysis which mostly assigns Quality Adjusted Life Years (QALYs) as the unity of analysis. Cost analysis procedure is most suitable for this study because the study attempts to identify all cost components associated with of an episode of cholera outbreak and how each component contributes towards the overall costs of treatment to households and health facilities in dealing with the disease. Though direct costs are important and shows an obvious impact of illnesses on the unit under study, some studies such as Akazili et al. (2007), Poulos et al. (2012), Sachaetti et al. (2012) & Sarker et al. (2013) all found indirect costs to be higher than direct costs. Specially, Akazili et al. (2007) found indirect cost to be as high as 71% of the total cost of treatment of malaria within households in Northern Ghana. University of Ghana http://ugspace.ug.edu.gh 41 Cost analysis also measures the amount society spends in total upon an episode of the disease and in effect the cost-savings of the eradication or minimizing of infections of the disease. Other cost analyses don’t make conscious effort breaking down the costs into its base components to check its impact on specific units for example, CBA will weigh the costs and benefits of a programme to choose the programme but not consciously seeking to find impact on various cost components. CMA seeks the least cost in undertaking a venture while CEA seeks to maximize a given budget. But because this study seeks to identify the impact of costs on households especially identifying other none-direct costs which might play a significant role in impacting these households Cost analysis is the best method for analysis. Finally, this study does not examine the impact of an intervention like the introduction of a Cholera vaccine or interaction of various interventions such as combination of water source improvements with either a vaccination programme which all the other cost analysis might be suitable in providing relevant measures, they are not used here. 3.3 Theoretical Framework Mortality, morbidity and debility are costs associated with cholera but this study focuses on the impact of morbidity and to a lesser extent mortality (burial cost). Debility which deals with the impact of an inability to engage in work due to physical weakness is not investigated separately in this study but captured under morbidity. This is peculiar with cholera studies as the idiosyncrasies of the disease make all cases admission cases. Shephard et al. (1991) categorize various University of Ghana http://ugspace.ug.edu.gh 42 costs components borne by the individual, the household, the health facility, the government and or the economy into direct, indirect and intangible. Malaney, (2003) further comprehensively expressed by as: COI = Private Medical Cost + Non Private Medical Cost + Labour Loss + Risk Related Behavior Modification + Investment Lost + Non Economic Personal Burden Figure 3.1 shows the COI due to cholera is conceptualized in a framework adopted from Asante & Asenso-Okyere’s (2003) Economic Burden of Malaria in Ghana but with modifications to suite this study. These modifications exclude intangible costs and other direct costs to institutions such as Non-Governmental Organizations because they don’t fall within the scope of this particular study. Figure 3.1 Conceptual Framework for Costs of Cholera. Mortality Morbidity Costs of Illness Direct Costs • Household Treatment, Transportation & burial. • Facility Treatment Indirect Costs • Treatment time, waiting & travel time • Income & Production losses Household / Individual Costs Health Institute Costs Reduced Productivity • Loss of Labour hours • Absenteeism Cholera Effect University of Ghana http://ugspace.ug.edu.gh 43 Source: Asante & Asenso-Okyere (2003) with modifications by Author. 3.3.1 Direct Cost Direct costs are costs that can easily be identified and completely attributed to specific items in the production of specific goods or services but in this study it is seen as in the production of health by households and health facilities. The process of seeking treatment involves cost to the individual and his household. In addition, facilities giving out treatment also bear costs in providing such services to its consumers and these are also captured under the direct cost. Cash expenditures made by households in seeking care form the root of most direct costs in health economics and these costs normally include; medical expenses, cash expenditures on special foods, transportation and provisions for burials in event of deaths. On the other hand, facilities bear direct cost through both medical and non-medical supplies and services. Both households and facilities costs constitute resource consumption costs for the seeking and provision of healthcare. Direct costs to households can easily be obtained through recalls but that of health facilities become more complex due to how health systems are setup. Costs within such establishments are mostly shared among several activities which are not easily distinguishable. For the purposes of this study, only costs that went directly into logistical supplies in dealing with the outbreak at the facility level were considered. These supplies were items used directly in addressing the problem of the cholera outbreak in the facility. University of Ghana http://ugspace.ug.edu.gh 44 3.3.2 Indirect Cost There is a usual debate about the inclusion of indirect costs in costing procedures since it is not mostly tangible or directly associated with an event. But when included in studies they show the bigger view and truest costs because they mostly comprise opportunity costs which are fundamental costs to economists. Indirect costs are very significant with the incidence of cholera. Cholera unlike other diseases leaves the patient acutely dehydrated and unable to partake in any meaningful work, school or social activity. Additionally, caregiving for patients is crucial towards the full recovery of patients because patients are critically weak and need constant attention and sometimes special feeding. These characteristics of the disease condition make both patient and caregiver loose valuable hours which hitherto could have be engaged in participating in productive venture. Oxfam’s Cholera outbreak guidelines (2012) suggests that children under the age of five are among the three most vulnerable groups during an outbreak and these children mostly will require constant caregiving for the entire duration of their ailment until recovery. Indirect costs hence include costs incurred not for its own sake but as a result of another item or activity. These costs sometimes go beyond monetary expense to include productivity and time losses. Therefore, they are difficult to assign to or identify with a specific cost objects. In definition, indirect costs are ‘…costs associated with lost or impaired ability to work or to engage in leisure activities due to morbidity and lost economic productivity due to death’ (Gold et. al, 1996. University of Ghana http://ugspace.ug.edu.gh 45 p.181). It is simply defined as the value of time lost due to illness or death. These productivity loses will differ according to the age and occupation of an individual. During the period of sickness patients and their caregivers incur an opportunity cost in time spent on treatment, waiting and travel. This is not an out-of-pocket expenditure but productive time loss that could otherwise be used for attaining both economic and non-economic goods. Such productivity losses result from incapacitation, occupational disability, caregiving and premature death. Indirect costs are made up of basically three components; absence from paid work, reduced productivity at paid work and unpaid production such as reduced possibilities of performing usual activities at home. The first two components apply to the adult productive group normally between the ages of 15 years (age for employment under the Ghana Children’s Act 1998) and 60 years (retirement age in Ghana). Indirect costs become importantly crucial when the sick household member is the bread-winner or contributes substantially to the household income and his or her incapacitation results in huge financial strains on the family. The commonest measure of indirect cost is through the human capital approach based on the Neo-classical viewpoint that wage rate is equivalent to the value of marginal productivity of labour. Concluding that incomes earned can be equated to monetary value of production lost and therefore calculates the indirect cost using the individual’s gross wage that would have been earned except for the onset of disease multiplied by the duration of absence. University of Ghana http://ugspace.ug.edu.gh 46 A more recent approach is the friction method. This came in response to the assumption of full employment adopted by the human capital approach which did not take into account certain realities. A more societal view is to look at the gap between unemployment and frictional unemployment. In this case, an absentee worker is replaced by a previously unemployed person and hence his lost income transferred to the newly employed. On a societal level, there is no loss of income or productivity, the only loss of productivity is the period it takes to replace the sick individual and this is called the ‘friction cost’. This makes the friction method provide a less exaggerated production cost compared to the human capital approach. A second recent method recommended by the Washington Panel’s guidelines incorporates indirect non-medical cost as health effects rather the traditional methods of productivity losses due to ailment. It jumps a step further ahead of the friction approach by supplementing a friction effect with health effect. It recommends that productivity losses be measured through its impact on quality of life in terms of health effects plus friction costs. Unlike the human capital approach and the Friction methods which both capture in monetary terms indirect costs through productivity losses and income reduction, the Panel method further breaks into two the impact of reduced productivity; one, impairment in role functioning and two, reduced income conditioned by disease effect. The Panel’s method theorizes that quality of life of an individual is in part dependent on their role functionality in society or specifically on the job they are involved in. Therefore, the impact of role functioning is measured as health related quality of University of Ghana http://ugspace.ug.edu.gh 47 life-effects and the relationship between productivity and quality of life runs partially through the patient’s income. The Panel concludes that patients will base their estimation of quality of life reduction on their loss of income rather than directly on the loss of production. However, in response to this model, Brouwer et al. (1997) criticized the Panel’s method on three key issues; one, direct health effects might not be meaningful monetized on the quality of life and should only be considered as health effects and not productivity losses at the societal level. Two, the model does not take into account compensational facilities such as social security and private health insurance which can mitigate the impact of income losses. And finally, the model becomes most useful from the patient’s (individual or household) perspective when measuring the impact on quality of life and not for productivity costs from a societal viewpoint. From the above-mentioned, the traditional human capital approach has the greatest tendency of overestimating indirect costs because it assumes away many real life possibilities. The friction method deals with this potential overestimation by incorporating an augmenting factor of replacement of sick worker but is most useful in macro-level studies where a firm’s reduced productivity is being measured. Brouwer et al.’s third criticism of the Washington Panel’s method show this method has rather has advantages when doing micro level studies. Finally, other studies measure costs through a willingness-to-pay approach. This approach seeks to capture either the entire household by including adult University of Ghana http://ugspace.ug.edu.gh 48 household members who are not working and yet can incur some form of indirect cost. It normally focuses on individual’s own weighted perceived costs. Studies like Asante & Asenso-Okyere (2013) go further in using this approach in measuring intangible costs for burden of malaria in Ghana but this measure is clearly subjective and mostly criticized for not having any economic implication. WTP attaches monetary weights to costs which are not easily measurable such as psychological effects of ill-health and because it is subjective it can be influenced by many multiple factors. 3.3.3 Intangible Cost Intangible costs are costs which are not easily measured and associated with costs of pain, suffering, stigma and the fear of death as a result of a health condition. These costs are mostly subjective; limited to feelings and so might vary from person to person. In addition, there are difficulties in attaching real costs to them. Therefore, due to these measurement difficulties, this type of cost measurement is excluded from the study. 3.4 Empirical Review Literature on Cholera seems to lean towards its epidemiology, environmental, risk factors and socioeconomic correlates of the disease. This situation is no different for studies on Ghana, which have examined topics on cholera related to; climatic changes (global and regional) and its effects on cholera patterns observed in Ghana (de Magny et al. 2006). This study used cholera data from the WHO Weekly Epidemiological Record for Ghana from the years 1975 to 1995 and employed a wavelet analysis in running for its results. They found strong University of Ghana http://ugspace.ug.edu.gh 49 statistical coherence between cholera outbreak resurgence, temporal patterns of incidence and climatic parameters from the end of the 1980’s. The study revealed that there were two cyclical patterns of incidence every 4-5years and 7-8 years. However, this study just like others, presented only an environmental linkage with cholera incidence but recognizes in its literature review that other factors such as poverty and human density, may influence the spatial and temporal distribution of cholera cases. A similar work by Osei & Duker (2008), carried out a Geographic Information System based spatial analysis and statistical analysis to determine clustering of cholera in the Ashanti region of Ghana. Their study showed high cholera rates clustered around the Kumasi Metropolis (the central part of the region) showed direct relationship between cholera and urbanization as well as overcrowding but an inverse relationship with other neighborhood towns away from Kumasi. A later study by Osei et al. (2010) also did similar work on the relationship between cholera prevalence and environmental factors such as exposure of surface water to potential cholera reservoirs (like dump sites) in Kumasi the second largest city in the Ghana. This study set out to look for the impact of surface water contamination on cholera infection and spatial mapping of surface water to determine the relationship between cholera prevalence and proximity of upstream potential cholera reservoirs. Making use of statistical Ordinary Least Squares (OLS) and GIS spatial analysis they established that surface water pollution through run-offs from dump sites play significant role in cholera infection. It University of Ghana http://ugspace.ug.edu.gh 50 concluded that communities that were closer to dump site showed greater incidence of cholera compared with those further away. The closest attempt at observing an economic implication of cholera was a study by Davis-Teye et al. (2014) who looked at the socioeconomic factors associated with cholera outbreak in Southern Ghana. In their study, they conducted a descriptive and unmatched case-control study for persons who suffered from cholera during the 2012 outbreak within the Osu-Klottey district. Information on individual’s socio-economic, hygiene, food and water exposures were collected and used to determine the socioeconomic implication. They found that persons aged below 18 years, persons with educational level below tertiary, households that have exclusive toilet facility, the eating of cold or warm food either from home or outside home, eating food from home and access to pipe-borne water were associated all with cholera. But the study earns here and does not look at the impact in terms of the costs that households bore but only the socio and economic causal relationship of the disease within that district. This current study adds on to the thin number of studies of this nature by establishing an economic dimension of the disease in terms of its costs and going a step further to assessing its impact rather than only observing its causal routes. 3.4.1 Costs Methods in the measurement of Cholera From my literature search, I observed that studies focusing on the economic impact of the cholera is scant compared to that on other infectious diseases such as malaria. Most literature measuring economic impact of cholera has measured University of Ghana http://ugspace.ug.edu.gh 51 its impact at the macro level. These have mostly looked at the loss to Gross Domestic Product (GPD) or the impact on certain sectors of the economy like tourism, exports and domestic consumption. (Suarez & Bradford, 1993; Kimball et al., 2005; IVI, 2010). At the micro level, the impact are usually measured by either households’ or individuals’ expenditure during an outbreak (Sarker, 2013). Measurement of cholera’s economic impact have been done through different costing procedures including; Cost analysis, cost effectiveness analysis, cost benefit analysis and cost minimization analysis. A firth method of measurement; cost utility analysis, seems to be the rarely used as my literature search did not come across the use of this measure on assessing the impact of cholera. The earliest real attempt to measure cholera was a work done by Suarez and Brandford (1993) who conducted an extensive study of the costs on the Peruvian 1991 and 1992 cholera epidemic. They adopted a cost of illness approach in identifying the costs components of the outbreak at the macro level of the economy. They further calculated the impact on the supply channels and three demand effects: revenue from tourism, revenue from exports of goods and impact on domestic consumption. The total economic impact of the study was found to be about US$ 96.6 million; and when mortality costs are included the costs of illness increased to US$ 176.9 million. A drawback of the study however, is the use of secondary data from only public health institutions under the Peruvian Ministry of Health for its analysis. Private health facilities are not covered under the MOH and hence all patients treated in such facilities and all other facilities are University of Ghana http://ugspace.ug.edu.gh 52 completely missed. This has a potential of underreporting and hence underestimating the actual costs. Poulos et al. (2012) determined private (individual/household) and public (government) costs of Cholera by using a combination of hospital and community based studies in Mozambique, Bangladesh, India and Indonesia. The community based aspect of the study capture both slum and middle class neighbourhoods to present a more representative view of the study. Private costs were defined under treatment and hospitalization costs in the facility and public costs as payments for medical care and loss of earning of the patient and caregivers. They found that costs borne by the public sector was higher than private costs but private costs became significant when patients had been hospitalized. Costs of the illness to the patient formed 21% of the average monthly income and rises to as much as 65% when there is hospitalization. Direct costs among the children group exceeded indirect costs for both children and adults. This was explained by the fact that children engaged in less economic activity and therefore had lesser indirect impact in terms of costs incurred due to productivity losses. The study also found patients and caregiver lost productivity forming the largest component of indirect costs. A study by Kirigia et al. (2009) took another dimension by calculating the short- term and long-term costs of Cholera in WHO African region. Their estimates proved a total economic loss of US$53.2 million as a result of the 125,016 cases reported in the region in 2005. Their costs included; hospitalization and treatment, laboratory diagnosis of the disease, short-term loss of earnings of patients and University of Ghana http://ugspace.ug.edu.gh 53 their families and loss of productivity and output due to limited working ability and premature death. Similar to the work done by Suarez and Brandford (1993), reported cases were obtained from secondary sources but this time from the WHO but this presents a huge potential for underestimation due to large numbers of unreported cases. Underreporting of cholera cases to the WHO has come about as a result of a number of factors which include; challenges in conducting cholera surveillance in the developing world due to weak or poor health systems and concern in some countries about economic and social ramifications of reporting cases such as reduction in tourism and export of some goods (WHO, 2000). Naficy et al. (1998) used cost-effective analysis to assess alternative strategies in the control of cholera in a hypothesized Sub-Saharan refugee of 5000 persons. Refugee camp situation was hypothesized because of the frequent reporting of cholera within such camps. The study tested three intervention; Preemptive therapy, Reactive therapy and Adding vaccination to either therapies. The preemptive therapy sets in motion an intervention before even a cholera outbreak begins. The reactive therapy was interventions that were adopted once an outbreak occurs. Their results showed that a Reactive therapy cost almost $270 more per death averted compared to the preemptive therapy. The combination therapies of vaccination plus preemptive and vaccination plus reactive showed a cost of $1745 and $3833 respectively per death averted. These were hugely expensive. However, the combination became more cost-effective than the single therapies when the price of vaccine fell below $0.22 per dose. University of Ghana http://ugspace.ug.edu.gh 54 In 2009, there was an Oral Cholera mass vaccination campaign in Tansania to aid traditional preventive and treatment methods of Cholera in endemic parts of Zanzibar. To assess the cost-effectiveness of this campaign, Schaetti et al., (2012) estimated the health care provider and household costs of illness from cholera outbreaks and the costs of the mass vaccination campaign that used the two-dose Oral Cholera Vaccine (OCV) Dukoral. The results showed that the campaign was not cost-effective due to the pricing of the vaccine and the relatively low incidence of cholera. In 2010, the International Vaccine Institute conducted similar studies in Mozambique and Bangladesh. In this study, GDP is lowered by around 2-2.5% in the year of the epidemic and 0.5-1% in the following year. This represented a loss of more than US$ 2 billion for Bangladesh and US$257 million for Mozambique. The study concluded that the shocks to labour supply resulting from the cholera epidemics in these countries was very small and was unlikely to represent significant cost. On the other hand, Sarker et al., (2013) estimate the economic burden of cholera at the micro level focusing on costs incurred during treatment by households without looking at facility costs of treatment. The study was carried out in response to a large vaccine trial in an urban area in Bangladesh. Productivity losses due to forgone non-market activities including household chores, child care and leisure were captured by adopting either age-specific wage or an occupation- specific wage. But they ignored costing intangibles such as suffering and grief because these costs do not represent any tangible economic impact. The cost University of Ghana http://ugspace.ug.edu.gh 55 analyses used both prospective (likelihood of cases occurring) and retrospective (occurred cases) incidence-based analyses of cholera illness per episode per household. The study found that by preventing an episode of cholera, a total cost of US$ 30.40 is averted per household. Medicine costs formed the largest share of direct medical costs and transportation costs that of direct non-medical costs. For indirect costs caregiver lost productivity were found to be greater than that of patients. This particular study is largely similar to this current study but only differs in the addition of facility level costs and addition of an assessment of the impact of cholera by income groupings. On Ghana, Aikins et al. (2010) used cost-analysis on hospital health care cost of diarrheal disease in Northern Ghana. This study focused on the health sector perspective and captured only direct medical costs. Estimation of costs only focused on direct medical costs and only children below the age 5 were included in the study. The analysis was based on the WHO guidelines for estimating the economic burden of diarrheal disease and took a prospective approach in identifying costs. But studies like this have two main weaknesses; they are limited in capturing costs since household impact is lost and they are too generalized because diarrheal diseases are composed of different forms of intestinal diseases. These studies have estimated the impact of cholera using different costing approaches. My study goes a step further from estimating the economic costs of cholera in Ghana, it will seek to provide comparative costs of the 2014’s Cholera outbreak in Ghana using costs from one low and one high incidence district. Finally, in addition, the study will aim at assessing the burden of the outbreak on University of Ghana http://ugspace.ug.edu.gh 56 various income groupings by observing the individual and household characteristics that causes the individual’s likelihood of belonging to particular income strata. 3.5 Cholera, Portable Water, Sanitation and Poverty linkages Between 1848 and 1854 cholera caused a quarter of a million deaths in Great Britain, primarily among the poor who regularly had limited choices in using water that were contaminated with human and animal wastes in their everyday lives (Anbarci et al. 2006). In 1993, 70% of Brazil’s states reported confirmed cases of cholera with the hardest hit areas being the mostly impoverished North- eastern parts. These areas reported 87% of all cases and a 3 times higher probability of death over the national average once disease was contracted (Lima, 1994). These four features; cholera, portable water, sanitation and poverty are intertwined and have strong linkages. Poor sanitation, portable source of water supply and the incidence of cholera all seem to stem from poverty and is evidenced by the occurrence of cholera amongst poor slum communities. The eradication of cholera is one of the key indicators of social development and yet the disease remains a global threat especially in the developing world. While the disease no longer poses a menace to countries with minimum standards of hygiene, it remains a serious challenge to countries where access to safe drinking water and adequate sanitation cannot be guaranteed (Talavera & Perez, 2009). Access to portable water is a major determinant of good health and prevention of many diseases including cholera. Inadequate drinking water does not only result University of Ghana http://ugspace.ug.edu.gh 57 in much sicknesses and deaths but also increases health care expenditures, lowers worker productivity and school enrolment (World Bank, 1994). Talavera and Perez (2009) buttress this point stating that economic development is an important factor in the morbidity and mortality of cholera. Furthermore, Anbarci et al. (2006) concludes that cholera and other diarrheal diseases account for 11.3% of all deaths among the poor globally marking such illnesses the second leading cause of death among the poor. This study was at a macro level and it tested the theoretical model of the need for a collective action in fighting against cholera having an increasing function of both a country’s level of income and the income inequality distribution within the country. The analysis was done on 1032 annual observations from 17 relatively poor countries between the years 1980 and 2002. They found for example that collective action in providing clean portable water is an increasing function of income and inequality. Diarrheal diseases are caused by poor environmental hygiene of water and food. Water based disease transmission by drinking contaminated water is responsible for significant outbreaks of faeco-oral diseases such as cholera, typhoid, dysentery and diarrhea (Nketiah-Amponsah et al., 2009). But poverty seems to perpetuate these conditions were due to lack of portable water supply people are faced with increased risk of contracting diseases such as cholera. Naidoo & Patric (2002) also introduce another dimension stating that health education is difficult in countries with low literacy rates which is a common feature in developing countries. This evidence is supported in Grossman’s (2000) University of Ghana http://ugspace.ug.edu.gh 58 model which suggests that people with higher forms of education are better placed in understanding health implications and therefore make better choices regarding their health. It must be stated here that, though cholera is easily preventable if certain simple hygienic principles are followed but poor assimilation of education on the prevention and identification of the symptoms of the disease in low educated areas makes its impact great within such communities. Further, they state that poorer countries do not have sufficient resources to provide relief during an outbreak of cholera, let alone provide education to plan for one. Contrary to this, Nations and Monte (1996) did a study on reactions towards mass media campaigns and government efforts in controlling cholera outbreaks in Northeastern Brazil. Using intensive ethnographic interviews and participant observations the found out that the non-compliance with recommended regiments had more to do with impoverished residents revolting against accusatory attitudes and actions of the elite than an outright rejection of care by the poor. Therefore education is key in the battling of the disease but in some cases certain socio-cultural factors hinder the level of influence of education. In Ghana, Fobil et al. (2010) investigated the relationship between socioeconomic conditions and urban environmental quality in Accra drawing socioeconomic and environmental variables from the 2010 census data. The study established strong socioeconomic predictions of urban neighbourhood environmental quality and suggested that widening socioeconomic inequalities (e.g., urban unemployment, University of Ghana http://ugspace.ug.edu.gh 59 income gaps, etc.) at household level could worsen existing urban environmental health inequalities at the community level. GHS report of cholera outbreak in the Greater Accra (2014) identified these six risk factors; poor sanitation, unsafe drinking water, increased person to person transmission, poor personal hygiene, poor food hygiene, floods leading to contamination of domestic water sources, broken down water and waste disposal systems. All of these have their roots in poverty. 3.6 Summary of Chapter The chapter opened with the theoretical review underpinning the study. The Grossman (1972) model and modifications by Wagstaff (1986) were discussed and its application to this study. The COI approach was thoroughly discussed and the theoretical framework for this established. From empirical review, a close association was established between cholera incidence, poverty and the lack social amenities. Despite this, there exist a gap in literature on the costs of cholera in Ghana and the impact it has on various income strata for which this study is initiated to fill. University of Ghana http://ugspace.ug.edu.gh 60 CHAPTER FOUR METHODOLOGY 4.0 Introduction This chapter outlines the methodology used in the study on the costs of cholera. It describes the characteristics of the study sites, explains the selection procedure of the population and the sample size, the data collection procedure, sampling technique, ethical considerations for the work and finally, the method of analysis adopted. 4.1 Characteristics of Study Area The Greater Accra region is the capital city of the ten administrative regions of Ghana with a population of 4,530,905 (GHS, 2014). The region has remained the most densely populated region since 1960 and has increased from 167 in 1960 to 1,236 in 2010, nearly an eightfold increase. It lies in the South East of the country along the gulf of Guinea and has miles of coastline especially in the rural parts of the region. The shores in the capital city, Accra are however mostly polluted and sanitation practices within shoreline communities are poor. The Greater Accra region is challenged by problems of equitable access to an acceptable quality of health service. The number of households sharing a single room form 20.3% of the population and 8% of all the households has no toilet facility in their homes and the practice of open defecation is mostly common (GSS, 2012). La Dadekotopon Municipality was carved out of the Accra Municipal Area and forms part of the newly created municipalities in June 2012. The Municipality University of Ghana http://ugspace.ug.edu.gh 61 covers an area of 36.033 square kilometres. It shares boundaries with Accra Metropolitan to the west, east: Ledzokuku-Krowor Municipal, north: La Nkwantanang-Madina Municipal and south: Gulf of Guinea. The Municipality is entirely urban and is populated by 183,528 of which 52.7% are females. It has a total of 51,154 of households and an average household size of 3.6 persons per household. Within these households children constitute the largest proportion (35.2%). One room constitutes the highest percentage of room occupancy by households with about 18.1% of households with 10 or more members occupying a single room. The main source of drinking water for these households is pipe- borne water inside dwelling (31.9%) which are complemented by other sources of usually tap water (pipe-borne outside dwelling, 30.5%; Sachet water 24.0%, Public Standpipe; 9.2% and others 4.4%). Of the population 11 years and above, 94% are literate and 6% not literate. The proportion of non-literate female (6,712) far outnumbers their male counterparts (1,887) by more than three times. The main economic activity for men involves the operation of public transport system business while majority of the women are into petty trading as well as hawking. About 60 percent of the population of La Dade-Kotopon are self-employed and are mostly into food vending, mechanical works, hairdressing, tailoring and carpentry (GSS, 2014). Shai Osudoku district formerly Dangme West was carved out of the former Dangme district in 1988 as a result of a national re-demarcation exercise carried out in relation to decentralization reforms in the country. It is boarded by the Yilo Krobo, Manya Krobo and Asuogyaman districts in the north, to the east: Ada University of Ghana http://ugspace.ug.edu.gh 62 West, south: Ningo-Prampram and west: Akwapim North Municipal and Tema Metropolis. It has a total land area of 1,442 square meters representing 41.5% of the regional land, making it the largest of all 16 administrative districts. According to the 2010 population and housing census the population of the district is 51,913 with males making up 48.7% of the population. About 78.7% of the district population resides in rural communities. The housing stock in the district is 8,351 and the average number of persons per house is 6.2. Literacy in the district is fairly high with 70.7% of the population 11 years and older literate. The predominance of the rural population reflects in agricultural activities forming the largest type of occupation. Although agriculture dominates the district, the leading sectors in terms of provision of revenue to the District are the quarries. The main source of drinking water is pipe-borne water outside dwelling (37.3%) and the main type of toilet facility used is bush/beach/flied (31.2%), Public toilet (30%) followed by Pit latrine (21.1%). The Greater Accra region has been the epic-center of all outbreaks of cholera in the country since it was first reported in the country in 1970. 4.1.1 Study Sites The study is conducted within La-Dadekotopon and Shai-Osudoku districts of the Greater Accra Region. By the close of the year 2014, health facilities within La- Dadekotopon and Shai-Osudoku had recorded 1,907 and 315 cholera cases respectively. Among the top ten affected districts in Ghana, La-Dadekotopon ranked second while Shai-Osudoku rank second from bottom. Cholera incidence in the two districts, its costs plus other considerations such as accessibility, socio- University of Ghana http://ugspace.ug.edu.gh 63 economic conditions, rural-urban considerations, financial and time constraints of the study aided the selection criteria. Facility studies are conducted in the La General Hospital and Shai-Osudoku District Hospital. Both hospitals have same status as district hospitals for their various catchment areas and received the largest cases of patients reporting of suspected cholera symptoms throughout the outbreak period. 4.2 Data and Sampling Household data of patients were obtained from the GHS line list for cholera outbreak. This contained names, place of residence, sex, age, laboratory test result, outcome of treatment and the telephone contact of patient. The GHS used this database in their contact tracing of cases and this same tool was used in tracing patients to their households. For the purposes of this study, population is defined as all positive cases of cholera reported from a particular district. These form the bases for the inclusion criteria and all other households which did not suffer cases are excluded. A purposive sampling procedure is adopted in selecting a total of 418 households; 282 and 136 from a high and low incidence area respectively (See Appendix A for the calculation of the sample size). More specifically, the sampling technique takes the form of a homogenous purposive sampling because all households within the selection must have suffered at least a case of cholera to help answer the research objectives of this study. This sampling technique is adopted because costing of a specific disease requires only households that suffered cases to be University of Ghana http://ugspace.ug.edu.gh 64 sampled. The sampling technique is not wholly purposive as the total number of cases reported from these two districts forms the populations out of which the appropriate household samples are randomly selected. 4.3 Data Collection Procedure Field workers were selected and trained on the structured questionnaire. A pilot survey was then conducted for all loose ends and ambiguities within the questionnaire clarified. This ensured increased consistency among field workers and increased accuracy of responses. Contacts of patients or their caregivers were selected from the GHS cholera outbreak line list and were contacted. Facility level costs were obtained from hospital administration and the district public health officer interviewed. These costs were obtained from the hospital pharmacy cost list or from their records. 4.4 Ethical Considerations The study sought Ethical Approval from the Institutional Review Boards of the Institute of Statistical Social and Economic Research (ISSER; clearance number - ECH 033/14-15) and the Dodowa Health Research Centre (clearance number - DHRC/IRB/15//03). A letter from the Department of Economics and a copy of the proposal of the study were sent to the Greater Accra Regional Health Directorate who gave approval for participation of selected facilities. Ethical consideration is therefore given to all households where data would be collected. Consent was sought from members of households providing University of Ghana http://ugspace.ug.edu.gh 65 information in the form of oral and written, signed consent forms. During interviews, privacy of individual is assured and all responses were confidential. 4.5 Theoretical Framework Health is determined by several factors of which medical care is one. These factors range from social settings within families, environmental factors such as settlement patterns, incomes, education, diet and lifestyle, to mention a few. The main crux of this study is based on Grossman’s (2000) modeling of demand for health. In the model, an individual is assumed to be born with an amount of health stock that depreciates over time but can be augmented through investments in health. Health is not passively purchased from markets but is produced in combining time with purchased medical inputs. The model assumes health care to be state-dependent when an individual is in sub-optimal health. Demand for health is considered as a consumption good that is; the demand for “good health” because health is seen as a productive good by individuals that produces healthy days. On the other hand, demand for healthcare is considered as an investment good that is; demand for medical services to produce good health. On the bases of this an inter-temporal utility function for cholera patients adopted from Grossman (2000) is represented as; 𝑈𝑖 = 𝑈(∅𝑡𝐻𝑡,𝐻0,𝑍𝑡), 𝑡 = 0,1,2, … . ,𝑛 …………………………………….. (1) Where; 𝐻𝑡 = Health stock of the individual at time t ∅𝑡 = Service flow per unit of stock at time t University of Ghana http://ugspace.ug.edu.gh 66 ∅𝑡𝐻𝑡 = Total consumption of health services at time t 𝑍𝑡 = Consumption of other commodities at time t and 𝐻𝑜 = Initial stock of health Grossman (2000), states that the length of life is determined by the quantities of health capital that is maximized subject to the production and resource constraints. Death therefore occurs when an individual health stock at a particular time is less than or equal to some minimum amount of health stock required to survive. Owing to this assumption, the individual is assumed to produce health by the amount of time he/her spends making appropriate investments in his/her health in order to live. The health of the individual however depreciates by a certain rate ‘δ’ which is exogenous but is a function of individual’s age. The net investment in the stock of health is defined as the gross investment less depreciation of the health stock. That is; Net I = It – δt Ht = Ht+1 – Ht ……………………………………………….. (2) It = Gross investment at time t δt = the rate of depreciation at time t (0 < δt <1) Ht = the amount of health investment the individual makes at time t Ht = the amount of health investment the individual makes at time t+1 Similarly, consumers are assumed to produce gross investments in health and other commodities that enter into their production function as; Ii = Ii (Mt, THt; E)…………………………………………………………………. (3) Zt = Zt (Xt, Tt;E)………………………………………………………………….. (4) University of Ghana http://ugspace.ug.edu.gh 67 Where; Mt = a vector of medical inputs purchased from healthcare system that contribute to gross investment in health. Xt = a vector of inputs that contribute to the production of other goods (Z) THt = time inputs in the production of health Tt = time inputs for Z at time t E = human capital or an individual’s stock of knowledge that is assumed to be exogenous. The budget constraint for goods is equal to the present value of expenditure on goods and to the present value of income over the life cycle plus the consumer’s initial assets can be stated as; � 𝑃𝑡𝑀𝑡 + 𝑄𝑡𝑋𝑡 (1 + 𝑟)𝑡 = 𝑛 𝑡=0 � 𝑊𝑡𝑇𝑊𝑡 (1 + 𝑟)𝑡 𝑛 𝑡=0 + 𝐴0… … … … … … … … … … … … … … … … (5) Where; Pt = price of medical inputs (Mt) Wt = hourly wage rate THt = Time input for health at time t Tt = Time input for other communities (Zt) TLt = Time lost due to illness at time t A0 = initial assets To attain the optimal health capital from the health investments, we maximize the utility function in equation (1) subject to equations (3) to (6). Grossman (2000) finally arrives at a reduced form equation for the healthcare as; University of Ghana http://ugspace.ug.edu.gh 68 𝐼𝑛𝐻 = 𝛼𝐼𝑛𝑀𝑡 + 𝜌𝐻𝐸 − 𝛿𝑡 − 𝐼𝑛𝜕0… … … … … … … … … … … … … … … … … . . (7) Where 𝜌𝐻 measures the percentage increase in the health production function due to a unit increase in stock of knowledge (E) and 𝜕0 as the disturbance term. From equation (7), the health production function is a function of health inputs represented by Mt, human capital given by E and the depreciation rate of health (𝛿𝑡) based on age. Grossman (2000) states that wealth, wage rates, the price of medical care and age can influence health. Wealth and wage rates he said influence health positively by increasing the demand for good health but costs of medical care decrease demand for good health due to price. The effect of age and education is not so clear and ambiguous. Provided the rate of depreciation of health stock rises with age then ageing reduces both amounts of health stock demanded and health capital supplied by individuals. Based on these assumptions under the Grossman (2000) model, price of medical goods is equated to costs of cholera treatment which are captured with the COI method. These cost components are identified and their magnitude on households and facilities determined. In determining the impact of the disease based on proportion of infection, the Tobit model presents a good advantage with the premise by the Grossman model that factors such as wealth, education and age all influence health. 4.6 Methods of Analysis The study makes use of standard Cost-of-Illness (COI) by Rice (1966) and the WHO guidelines for estimating the economic consequences of disease and injury University of Ghana http://ugspace.ug.edu.gh 69 (WHO, 2009): (a) the direct medical costs including institutional costs, i.e. those directly borne by affected families in addressing cholera and those borne by the health providers; and (b) the indirect costs, i.e. loss of productivity caused by cholera, which is borne by the individual or the household. The study further examines the relationship between cholera affected households using a Tobit Regression model by observing the relevance of income categories and other household characteristics in relation to the proportion of household contraction. Household’s characteristics are examined within this framework testing which income groupings bore the greatest burden of the outbreak. The data is collected by the use of a structured questionnaire to determine households’ direct and indirect costs as a result of the episode of cholera. These include costs of medical treatment, drugs, laboratory tests, transportation etc. Indirect cost is estimated based on the time lost in seeking treatment or giving care to affected household members multiplied by the average earnings of the household to determine amount of productivity loss. Household average income is determined by the summing the incomes of employed household members and then divided by the household size. 4.6.1 Direct Costs Direct costs to the household include; first aid given to patients before the visit to the health facility; transportation costs for both patients and caregivers; consultation fees, drugs purchased, laboratory tests and treatment costs at the facility; under-the-table payments which are unofficial payments made either to University of Ghana http://ugspace.ug.edu.gh 70 receive preferential or early treatment; feeding costs for patients; admission costs and finally, burial costs in the event of death. Summation of these costs components will give the total cost attributable to the individual and by extension the household. Facility direct costs are measured on items used directly in the treatment of the condition including intravenous fluids, gloves (surgical and examination), given set, cannula and cotton wool. Other facility direct costs captured includes costs on Health education, supervision of burial, intervention programmes (such fumigation of toilets), disinfectants, fuel and any other payments to staff for extra hours working on patience. 4.6.2 Indirect Costs This study draws from the Washington Panel approach in measuring indirect costs of cholera on the household. This model recommends measurement of productivity losses through quality of life measurement in terms of health effects combined with some form of friction cost. The traditional Human Capital approach which use wage rate to value productivity losses drawing from the neo- classical view that wage rates are equivalent to marginal productivity and hence all incomes earned can be equated to productivity loss. This approach has a huge drawback and removed from reality because it assumes full employment of resources. The model also assumed that replacement of workers was impossible and hence production losses continue until recovery of patient or in the case of impairment of workers ability to work in full capacity then production losses will continue until retirement. It therefore uses various proxies in the measurement of University of Ghana http://ugspace.ug.edu.gh 71 indirect costs including average agricultural wage, income per capita, salaries, legislated minimum wage, marginal productivity calculated from a Cobb-Douglas production function among others. These methods are however poor in representing actual marginal product of labour and must be used with caution (Prescott, 1999). In addressing this problem within the human capital approach, the Friction model evolved. This model took account of the reality of less than full employment within an economy and therefore the possibility of replacing workers. Hence, production losses were measured up until recovery or when the worker is replaced and the possible costs of replacement such as training costs. Though the Friction model presents a useful way of capturing indirect costs with less likelihood for exaggeration of costs, it is most useful when measuring the societal impact of indirect costs. This is because the method takes a holistic view of the economy and when there is loss in production due to a worker’s ailment, that worker’s lost productivity and hence lost income is compensated for by the employment of a replacement. The method concludes that one’s loss of income is gained by another in the economy therefore there is no loss of income within the economy but the loss is reduced to the period it takes for a replacement to be found and the costs incurred in training the replacement. This method is not most complementary to this study because the study focuses on the individual and the household’s micro level cost plus facility costs. Brouwer et. al (1997) suggests that the Panel method represents the most useful approach in measuring the impact on quality of life at such a level. University of Ghana http://ugspace.ug.edu.gh 72 The Panel method is calculated on the time lost in seeking treatment or giving care to affected household members multiplied by the average earnings of the household to determine amount of productivity loss. This method is similar to that adopted by Sarker et al. (2013) in measuring indirect cost. The time components include; travel time to facility, time spent at facility till recovery and the number of work or school days missed after recovery. Waiting time with cholera treatment is reduced to zero because all cases brought into facilities are treated as emergencies and hence are not significant to this study. To ensure that time loss estimates are not overweighed, during data collection only actively employed patients and caregivers were assumed to be losing productive hours and unemployed patients and caregivers time loss assumed to be zero. Some studies like Sacker et al. (2013) and Asenso-Okyere and Dzator (1997) monetize loss associated with children using age-specific wages calculating the loss of productivity to teenagers by halving the average daily wage of their parents and for children three forth of the parent’s age. This calculation though could be useful these studies did not provide the bases for the choice of one half parent’s wage for teenagers and three forth for children. Therefore these estimates cannot be assumed as a standardized generalized loss of productivity to children. In addition, loss of productivity of children might vary from country to country because the value of children’s loss might be greater weighed in some countries than others. Hence, this kind of calculation is not considered in this study. University of Ghana http://ugspace.ug.edu.gh 73 4.6.3 Model Specification (Objective 1) The specification is based on work done by Asante & Asenso-Okyere (2003) but with modifications to suite this study. The cost of illness due to cholera constitutes resources spent on treatment, control and prevention of cholera by households and health institutions. It also includes the monetary value of output and services that are not performed as a result of the illness. These costs can be categorized into direct and indirect. The cost of illness can be expressed as; C = X + Y Where: C = Cost of Cholera, X = Direct costs associated with cholera, Y = Indirect costs. The direct cost of illness (X) is the combination of personal, household and institutional expenditures on both treatment and control of cholera. The direct cost is expressed as; X = H + I + G, Where; H = the household costs of cholera treatment, I = the institutional cost of cholera not borne by patients and G = costs borne by government in cholera control and supplies to facility. The household direct cost is expressed as; H = h1 + h2+ h3 + … + h9. Where; h1 = cost of first aid h2 = fees paid for registration and consultation h3 = laboratory test fees h4 = drugs University of Ghana http://ugspace.ug.edu.gh 74 h5 = transportation for patient and caretaker h6 = admission costs h7 = cost of feeding (special feeding and water) h8 = under-the-table payments; when there were such occurrences h9 = burial costs; in the event of death of patient The institutional cost of cholera is also expressed as; I= b1+ b2+ b3+ b4+…+ bn. Where; b1 = cholera treatment items b2 = extra payments to staff for overtime work where applicable b3 = intervention costs; Health education, supervision of burial, Fuel cost for case tracing and drugs given during outreach. The indirect cost of illness (Y) due to cholera is the cost incurred by persons in households because they could not participate in normal working activity or the hours dedicated for these activities were limited because they were sick or caring for the sick. The indirect cost due to cholera morbidity is expressed as: Y = µ (y1 + y2 + … + y4) Where; y1 = Days of incapacitation without admission days y2 = time spent by patient on admission at the facility, y3 = time spent by caregiver caring for the sick, y4 = travel time to and from facility µ = average household income’s daily wage The sum of y1 to y4 gives the value of productive time lost by the patient and the caregiver attributed to cholera morbidity. To do this estimation, the number of University of Ghana http://ugspace.ug.edu.gh 75 days lost from work is multiplied by the average household income. To ensure that costs are not overestimated, estimated earnings is used to calculate this value rather than an across board daily minimum wage. All household incomes is pooled and divided by the household size to obtain the average household income. All average household incomes are summed for the two districts separately and divided by the number of households per district to obtain district average incomes. Since household incomes are collected on monthly bases and the indirect cost were being estimated on daily bases this is then divided by 8 hours to obtain the household daily wage. This ensures a standardized measure for all individuals’ indirect cost calculation. This will be used in calculating the amount of wage loss or opportunity cost resulting from an episode of sickness or caregiving. In this study, we assume a working class aged between 15 years and 60 years. All ages below 15 years and above 60 years are classified as dependent group and hence in calculating indirect costs for the dependent groups, we assume that they are not formally employed and therefore have an opportunity cost of labour of zero. This procedure has been used in similar studies such as Akazili et al. (2007). Also, household members that fall within the working class but are currently unemployed or in school will also have a zero opportunity cost of labour. Indirect costs also include productivity lost due to premature mortality. This is defined as any death occurring before the age of 61 years, which is the average Human Development Index life expectancy at birth for Ghana. To obtain the mortality cost of cholera, the Years of Potential Life Lost (YPLL) method could University of Ghana http://ugspace.ug.edu.gh 76 be used to estimate the value of life lost but contrary, there have been both methodological and ethical criticisms to this method because it places monetary value to human life. This has been found to be very problematic. The focus of this study is not to consider the value of individuals as a production factor and thereby equating monetary value of their output to their life. 4.6.4 Model Specification (Objective 2) Tobit Regression model was adopted in the determination of the burden of cholera on affected households. This model is useful because of its ceiling effect and generating a model that predicts the outcome variable within the specified range. It supposes that there is an unobserved latent variable (𝑦∗) that is linearly dependent on 𝑥𝑖 via a parameter vector β plus an error term ui that is normally distributed. Where 𝑦𝑖 ∗ is the latent variable. 𝑦𝑖 ∗ = 𝛽𝑥𝑖 + 𝑢𝑖 ~ 𝑁 (0,𝜎 2) The dependent variable in this study is the proportion of household infected which is set between 0.09 (least proportion of household infection) and 1(complete household infection). This model is advantageous over models such as normal Binary models which only capture zeros or ones responses of the dependent variable and cannot cater for multiple responses. It is also a better model University of Ghana http://ugspace.ug.edu.gh 77 compared to truncated models because it takes into account the unobserved observations within the dependent sample rather than simply cutting off data as the truncated models do. Also unlike the Ordinary Least Squares method, this model yields unbiased coefficient estimates for each of the independent variables. The regression is modeled over a number of demographic and income category of households that were affected by the outbreak in order to access the impact of the outbreak. The general model is specified as; 𝑖𝐶 = 𝑋𝑖𝛽 + 𝜀 (I) 𝑖𝐶 = 𝑛 𝑁 → 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐻𝐻 𝑚𝑒𝑚𝑏𝑒𝑟𝑠 𝐼𝑛𝑓𝑒𝑐𝑡𝑒𝑑 𝐻𝐻 𝑆𝑖𝑧𝑒 (II) Where; 𝑖𝐶, the dependent variable is proportion of the household members that contracted cholera during the outbreak. The key variable of interest is household income category and additional demographic characteristics examined which include, sex, marital status, relationship with household head, age, education, health insurance status and access to portable water. 𝑋𝑖 is specified as 𝑋𝑖= 𝑋1… … … … … … … .𝑋𝑛 Following the Tobit Regression model, the empirical model is specified as: University of Ghana http://ugspace.ug.edu.gh 78 𝑖𝐶 = 𝛽 0 + 𝛽 1 𝐼𝑁𝐶 + 𝛽 2 𝑆𝐸𝑋 + 𝛽 3 𝑀𝑆 + 𝛽 4 𝑅𝐻𝐻 + 𝛽 5 𝐴𝐺𝐸 + 𝛽 6 𝐸𝐷𝑈 + 𝛽 7 𝐻𝐼 + 𝛽 8 𝐷𝑊𝑆 (III) 𝑖𝐶 = Proportion of Household Infected 𝐼𝑁𝐶 = Income category 𝑆𝐸𝑋 = Sex 𝑀𝑆 = Marital status RHH = relationship with household head AG = age EDU = education HI = health insurance status DWS = drinking water source 4.6.5 Marginal Effects and Cross Tabulations Marginal effects of the censored sample are calculated for the Tobit model to capture how the observed variable (proportion of household infected) changes with respect to the regressors. This will cater for the representativeness of the model by taking into account all observations whether observed or unobserved. Marginal effects are given by partial derivatives of the probability of each outcome of the dependent variable with respect to the particular explanatory variable being observed. Finally, crosstabs are run on the survey data serve as robustness and consistency checks for results observed from the regression output. University of Ghana http://ugspace.ug.edu.gh 79 4.6.6 Selection of Variables The dependent and independent variables for this model are chosen in accordance with findings from existing literature earlier reviewed and the objectives of this study. INCOME Cholera is basically seen as a disease of poverty and empirical evidence has pointed to the persistence of the disease in areas where poverty is entrenched and provision of basic amenities is deeply lacking. There is an inverse relationship between income and cholera contraction hence, increased income is expected to lessen the incidence of cholera. Households with higher levels of income can afford better sources of drinking water, more improved and less shared toilet facilities and are more likely to avoid the consumption from food vendors operating under insanitary conditions. Income is a categorical variable with persons earning the highest income (above GH¢750 per month) within the sample as the reference base. Other categories are those not earning any income at all, those earning less than GH¢ 100, between GH¢ 100 and GH¢ 350, above GH¢ 350 but less than GH¢ 750. SEX and AGE From literature cholera appears to have no discrimination in its infection with regard to sex or age but there exist some evidence supporting claims of gender and age disparities in the rate of infection especially due to particular behavioral and social characteristics. In this model, sex is a dummy variable with males as the reference group. Males are represented as 1 and females 0. A report by Oxfam University of Ghana http://ugspace.ug.edu.gh 80 on Sierra Leone’s 2012 outbreak suggests that “women and young girls appear to be exposed to contamination in the domestic realm, in particular during the role of caring for babies and the sick, whereas in some urban contexts, males appear to be more vulnerable to infection, presumably, due to their hygiene and eating practices on the street” (Oxfam, 2012, p.35). Particularly in Greater Accra Region, Ghana, during the recent outbreak, contraction was higher for males than females and infection seem to be clustered within the age range of 10 to 49 years but the highest infection among those between 20 to 29 years GHS (2014). Based on this, the proportion of infection among males will be expected to higher than their female counterparts. Age is a continuous variable measured in years describing the last birthday of household members. Age of an individual is considered important in the literature towards the demand for health. Grossman’s demand for health care postulates that the elderly make less investment into their health because of decreasing health stock but there exist a counterfactual where the elderly have an increased health expenditure because of depleting stock of health. Age is also an important demographic feature of individual and will demonstrate the risk and exposure of contraction of cholera. As shown by the outbreak in Greater Accra, Ghana, infections have been highest within the youthful age range and least among more dependent populations (those below 9 years and those above 60 years). Just as sex, cholera is age neutral and different studies have suggested different age ranges susceptibility. But for this study, impact is expected for increased ages and specifically for youthful populations. University of Ghana http://ugspace.ug.edu.gh 81 MARITAL STATUS Marital status is defined as a dummy variable. It classifies individuals into either married or not married. For the peculiarities of this study, people living together in consenting relationships are also considered under the married category. Unmarried persons have been shown to be at greater risks of contracting cholera due to their likelihood of eating from vending sources which has been identified by the GHS as one of the main factors leading to the incidence of the recent outbreak in Ghana. Also, married persons are presumed to be more careful about their lifestyles because they normally carry an extra responsibility in caring and catering for other household members and hence will be cautious in their out of home eating habits. The Ministry of Health profiling of affected men in Ghana’s last outbreak showed higher incidence among unmarried men than the married counterparts. RELATIONSHIP WITH HOUSEHOLD HEAD Household composition is an important determinant on the impact of disease. For example the impact on heads and spouses might have greater consequences than other household dependents because there are usually “bread winners”. Other dependents within the household are set as the reference base and impacted of cholera Head of the household, spouse and son or daughter regressed in relation to the base. EDUCATION Education also plays a key role in the contraction, prevention and cure of diseases. Cholera is no different and literature suggest higher incidence of cholera University of Ghana http://ugspace.ug.edu.gh 82 among people with lower levels or no education than people with higher levels of education. This is simple due to their ability to assimilate and make better sanitary choices that greatly influence the contraction of cholera. Education is a categorical variable and is categorized into No education, Primary Education, Middle School or Junior High School and Higher. Higher education is set as the reference category for which the impact is accessed. Since education is key in the fight against cholera, infections among lower levels of education is expected to higher than higher levels. HEALTH INSURANCE Health insurance is a dummy variable and those with insurance are set as the base. Jehu-Appiah et al. (2010) have observed that poor homes have limited access to health insurance due to the issue of costs. Hence, it is expected that individuals without insurance cover will have a greater impact of cholera to their insured counterparts. But contrary to this evidence, the theory of moral hazard suggests that people with some form of health insurance tend to live riskier lifestyles due to the knowledge that their health care costs will either be totally absorbed or at least partially by a third party. Therefore the impact can be either ways, impacting greater on those insured or those not. DRINKING WATER SOURCE Water sources are often classified as ‘improved’ or ‘unimproved’: Sources considered as improved are piped public water into homes, public standpipe, borehole, protected (lined) dug well, protected spring, and rainwater collection; unimproved are unprotected wells and springs, vendors, and tanker-trucks (WHO University of Ghana http://ugspace.ug.edu.gh 83 and UNICEF, 2000). Source of drinking water is a categorical dummy and inside plumbing is set as the reference base. The other categories are drinking water from water vendors, neighboring house pipe, public stand pipe and other sources. According to the WHO report on the progress of sanitation and drinking water update 2014, the safest source of drinking water is a piped water source on the household premises and most unimproved are surface water sources like ponds, rivers and dams. Hence, other water sources are compared to this reference base to assess the impact of cholera. It is expected that households using unimproved sources will have greater likelihood of cholera infections. Table 4.1 Variables and their expected signs Variable Variable Defined Expected sign 1. Income Categorical dummy; Base Highest Income group (Above GH¢750per month) Category 1: less than GH¢ 100 +/- Category 2: between GH¢ 100 & GH¢ 350 Category 3: above GH¢ 350 but less than GH¢ 750 2. Sex Dummy variable; base male - 3. Age Continuous variable +/- 4. Marital status Dummy variable: base married - 5. Relationship with HH head Categorical dummy; base Other dependents +/- Category 1: Head of Household Category 2: Spouse Category 3: Son/Daughter 6. Education Categorical dummy; base highest education (above Sec.) +/- Category 1: None Category 2: Basic Category 3: Sec/Tec/Voc. University of Ghana http://ugspace.ug.edu.gh 84 7. Health insurance Dummy variable with not enrolled on health insurance as base +/- 8. Source of Drinking Water Categorical dummy; base Inside Plumbing/Inside Standpipe +/- Category 1: Water vendor Category 2: Neighboring house pipe Category 3: Public Standpipe Category 4: Other sources 4.7 Methodological Limitations Some critiques have challenged the usefulness of the COI method claiming that the identification of costs does not provide enough information for policy. This argument is flawed to some degree because COI studies is the first line method in assessing economic burden in terms of costs of an ailment and its possible cost saving benefits when the disease is either eradicated or its prevalence reduced. Also, COI has an important feature of placing costs within their various segments and this helps to know which aspects of the costs are most important and attention focused. University of Ghana http://ugspace.ug.edu.gh 85 CHAPTER FIVE PRESENTATION AND DISCUSSION OF RESULTS 5.0 Introduction This chapter presents demographic statistics of households selected for the study. These characteristics include; age, sex, education, income classification and etc. It also presents descriptive statistics of key variables relevant to this study and ends with an econometric estimation of the data. All descriptive statistics were from empirical results obtained from the field and were analyzed using STATA version 13. 5.1 Demographic Characteristics The survey captured in total 1543 individuals from 418 sampled households in two districts. Out of these, 1052 and 491 individuals were within La-Dadekotopon district; the High Incidence Area (HIA) for this study and Shai-Osudoku district; Low Incidence Area (LIA) for this study respectively. In the HIA females formed 54.0% of the sample selected and in the LIA 48.9%. Within the household headship structure in these districts, 63.4% were males for the HIA and 75.5% for the LIA. In both districts, children and other dependents formed over 40% of the household composition. Females were dominant at lower levels of education (Primary or None) in both districts but in relation to health insurance, females formed a slightly higher proportion (over 52% in both districts) of those enrolled. The mean 0.46 and 0.51 for sex in the HIA and LIA show an almost near equal proportion of males and University of Ghana http://ugspace.ug.edu.gh 86 females sampled in the study. The mean age in both districts was 27 years. The largest household size was an 11 member household found in the HIA. The average room occupancy was 0.5 and 0.6 for the HIA and LIA respectively showing relatively more people per room in the HIA than the LIA. There is a maximum of 1 for the proportion of household infected for both districts. This indicate a situation in which either all members of a household were infected or a single member household infection but in this study the latter was the case. Table 5.1 and appendix C provide details of these variables. See appendix C, Section I for tables of demographic statistics. Table 5.1 Descriptive Statistics of dummy and continuous variables La-Dadekotopon Shai-Osudoku Variable Obs Mean Std. Dev. Min Max Obs Mean Std. Dev. Min Max sex 1052 0.460 0.499 0 1 491 0.511 0.500 0 1 age 1052 27.136 17.398 0 85 491 26.697 17.991 0 95 NHIS enrolled 1052 0.476 0.500 0 1 491 0.713 0.453 0 1 Cholera infection 1052 0.285 0.452 0 1 491 0.283 0.451 0 1 Marital status 1052 0.326 0.469 0 1 491 0.387 0.488 0 1 Proportion of HH infected 281 0.342 0.164 0.09 1 137 0.354 0.179 0.1 1 Number of rooms 281 1.491 0.975 1 13 137 2.015 1.430 1 10 HH size 281 3.733 1.627 1 11 137 3.577 1.670 1 10 5.1.1 Cholera Infection The total of 1542 persons were affected by the outbreak of which 439 (28.5%) were infected by cholera of which 300 (19.4%) were from the HIA and 139 (9.0%) from the LIA. Infection among females was slightly higher (52.3%) in the HIA but the inverse persisted in the LIA (51.1%). Infection rates were highest in University of Ghana http://ugspace.ug.edu.gh 87 the 20 to 29 years age category where as many as 109/438 (24.9%) persons were infected. The youthful age category (between 10years to 49 years) recorded the highest incidence of the disease and together they formed 69.4% of all cases. Adults above 60 years were the least people affected by the disease with not more than 6% of total infections. The mean proportion of household infection was 0.2 and 0.4 for the HIA and LIA respectively showing a higher relative proportion of household infection in the LIA than the HIA. This can be well explained by the slightly higher mean household size of 3.7 in the LIA compared to 3.6 in the LIA. See appendix C, Section II for tables on Cholera Infection. 5.1.2 Infection; Household heads, Sex, Age, Education and Income level Household heads play an important role in the social and economic setup of the family. In most cases they are “breadwinners” of their home and most often the impact of their incapacitation due to disease is heavy. In all households, 67% of the household heads were males and females 33%. Of the total 439 persons infected, 147 (33.5%) where heads of household and they formed the second highest percentage of those infected with the highest persons being dependents (36.5%) who were children of the head of household. For all those infected, there was an almost 50% split between males and females who contracted cholera but the females slightly edged the males by 2.6%. A large proportion of those that contracted cholera fell within age categories less than 50 years (over 85% of all infections). Literature shows an inverse relationship between cholera infection and education. With higher levels of education people are more likely to make better sanitary choices and avoid University of Ghana http://ugspace.ug.edu.gh 88 diseases than those with lower forms of education or no education. Household members with educational levels lower than secondary recorded 75.6% of all infections cases. In terms of infections and income quintiles, persons within households of the poorest income quintiles had 45.6% of total infection which was the highest percent of infection among the income categories. These comprised of children who did not earn any income and adults who were currently unemployed. The highest income category (that is those who earned above GH¢750) had the least percentage of infection (3.9%). Infection rates in the income categories that fall between the highest categories and lowest category showed mixed infection rates. Persons in the middle income category (those earning above GH¢100 but less then GH¢350) had 21.2% of all infection followed by those of the fourth category (17.3%) then followed by those of the second category (12.1%). This implied a mixed infection within income categories. See appendix C, Section III for tables. 5.1.3 Infection; Location, Access to portable water sources and Access to Toilet facility Inside household plumbing or inside house standpipe is viewed as the safest source of drinking water and surface water the least safe. Majority of households (176) resort to the use of public standpipes (42.1%) followed closely by 111 households having an inside plumbing or inside house standpipe (26.6%). These two sources together account for close to 70% of drinking water sources in households. Hence access to portable water is seemly not to be a major problem in these districts. This raises questions about the handling of drinking water or University of Ghana http://ugspace.ug.edu.gh 89 hygienic practices which might not be properly observed by households such as washing hands with soap before meals and after visiting the toilet. La- Dadetokopon had 32.4% of its households using inside plumbing whilst in Shai- Osudoku (14.6%). This evidence is counterfactual that cholera is a water related disease and is normally incidence by lack of portable drinking water. There was greater access to toilet facility within the homes in the LIA than in the HIA. In the LIA, there was 23.6% more access to a toilet facility at home than in the HIA. In total only close to 31% of households in both districts combined have access to toilet facility in their homes. With the type of toilet facility commonly used, both districts showed a high usage of KVIP (59.1%) than any other facility. Of the 29.9% that were infected in the HIA, 23.8% did not have access to toilet facility at home; as much as 17.8% more for homes without toilet facility but this gap was considerable lower in the LIA (5.8%). See appendix C, Section IV for tables. 5.2 Direct and Indirect Costs of Cholera to Household Direct costs were measured by simple accounting process of identifying the components of the costs and their respective expenditures summed. For example, first aid was calculated by summing all medical expenditures (specifically on drugs) made by households before patient was finally sent to the health facility for treatment. In calculating average costs at the individual level, costs incurred for each component (e.g. First aid) is divided by the number of cholera cases recorded per that district but at the household level the cost is divided by the number of households in a particular district to obtain the average household cost. University of Ghana http://ugspace.ug.edu.gh 90 For example, the total first aid cost incurred in La-Dadetopon was GH¢ 2,236.60. This amount was obtained by summing all first aid costs incurred by households in that district. This is then divided by 300 which is the total number of cholera cases in the sample to obtain an individual average of GH¢ 7.46 for first aid. At the household level the cost is divided by 282 which was the total household sampled for La-Dadekotopon to obtain GH¢ 7.93. Indirect costs on the other hand are calculated by transforming all missed time components into days. All time components collected in hours are summed to obtain the total hours mixed and then divided by 8 hours (official working hours) to obtain the number of days mixed. The average income for each district is calculated by summing all household incomes and dividing by the number of households within that particular district. This is then multiplied by the lost days for each component. For example, the total number of days missed by patients in Shai-Osudoku was 655, this is multiplied by GH¢26.90 which is the average income for Shai-Osudoku to obtain GH¢17,619.50, as the Household Daily Total Indirect Cost. This amount is further divided by 136 (the number of households in the district) to obtain the Household Daily Average Indirect Cost (GH¢129.56). 5.2.1 Direct Cost of Cholera within Higher and Lower Incidence Areas The total direct cost incurred by households in La-Dadekotopon and Shai- Osudoku amounted to GH¢96,444.30 (USD 30,138.84) and GH¢26,991.30 (USD 8,434.78) respectively. In both districts, treatment cost formed the largest component of all direct cost, 49.49% in La-Dadetopon and 47.74% in Shia- Osudoku. This observation is relevant because cholera treatment costs at facilities University of Ghana http://ugspace.ug.edu.gh 91 during the outbreak were supposed to be free, but hospitals reported shortages of treatment materials and hence patients needed to pay from out-of-pocket. When admission costs are added to treatment costs, household medical expenditure shoots up by 24% in both districts. Besides the medical (treatment and admission) costs, feeding costs represented the highest form of non-medical direct costs (GH¢11,658.00 and GH¢3,684.00 in La-Dedakotopon and Shai-Osudoku respectively). There were six death cases reported in La-Dadetokopon but none in Shai-Osudoku from sampled households. These deaths cost a total of GH¢18,820 (USD 5,881.25) for burial and when included in costs for the HIA it shots up to GH¢115,264.30, a 16.3% increase in the total cost. Table 5.2 Direct Costs in High and Low Cholera Incidence area Direct Cost La-Dadekotopon Shai-Osudoku Total GH¢ USD$ GH¢ USD$ GH¢ USD$ First Aid 2,236.60 698.94 283.80 88.69 3,219.34 1,006.04 (2.32) (1.05) (2.10) Treatment at Facility 47,729.00 14,915.31 12,885.00 4,026.56 75,529.31 23,602.91 (49.49) (47.74) (49.18) Feeding 11,658.00 3,643.13 3,684.00 1,151.25 18,985.13 5,932.85 (12.09) (13.65) (12.36) Admission 22,637.00 7,074.06 6,470.00 2,021.88 36,181.06 11,306.58 (23.47) (23.97) (23.56) Transportation (Patient) 5,912.00 1,847.50 1,410.00 440.63 9,169.50 2,865.47 (6.13) (5.22) (5.97) Transportation (Caregiver) 6,271.70 1,959.91 2,258.50 705.78 10,490.11 3,278.16 (6.50) (8.37) (6.83) Total 96,444.30 30,138.84 26,991.30 8,434.78 123,435.60 38,573.63 Source: Survey Data; Author’s computation from Excel *Italic & bracket figures are percentages *Exchange rate: 1USD = GH¢3.20 (Exchange rate as at December 31, 2014) University of Ghana http://ugspace.ug.edu.gh 92 At the individual level, a person incurred a total average cost of GH¢321.48 (USD 100.46) in direct costs per episode of cholera in the HIA. This cost rose to GH¢342.00 (USD 106.88) when the average cost is considered at the household level. In the LIA, the direct average cost to the individual was GH¢194.18 (USD 60.68) and GH¢198.47 (USD 62.02) at the household level but was much higher in the HIA, GH¢321.48 (USD 100.46) at the individual level and GH342.00 (USD 106.88) at the household level. These household costs were significantly higher (more than two folds higher in the LIA and more than three folds higher in the HIA) than the average costs reported by Sarker et al. (2013) in Bangladesh where average cost of cholera to the household amounted to USD 30.40. The average expenditure by the household on treatment in the HIA was GH¢169.25 (USD 52.89) per episode of cholera. This was markedly lower (GH¢94.74 or USD 29.61) in the LIA. Table 5.3 Individual and Household Direct Average Costs for HIA Individual Average Direct Cost HH Average Direct Cost La-Dadekotopon Shai-Osudoku La-Dadekotopon Shai-Osudoku GH¢ USD GH¢ USD GH¢ USD GH¢ USD First Aid 7.46 2.33 2.04 0.64 7.93 2.48 2.09 0.65 Treatment 159.10 49.72 92.70 28.97 169.25 52.89 94.74 29.61 Feeding 38.86 12.14 26.50 8.28 41.34 12.92 27.09 8.47 Admission 75.46 23.58 46.55 14.55 80.27 25.09 47.57 14.87 Transportation (Patient) 19.71 6.16 10.14 3.17 20.96 6.55 10.37 3.24 Transportation (Caregiver) 20.91 6.53 16.25 5.08 22.24 6.95 16.61 5.19 Total 321.48 100.46 194.18 60.68 342.00 106.88 198.47 62.02 Source: Survey Data; Author’s computation from Excel *Exchange rate: 1USD = GH¢3.20 (Exchange rate as at December 31, 2014) University of Ghana http://ugspace.ug.edu.gh 93 In all, average direct costs components were higher in the HIA than the LIA. This was due to the private purchases of treatment materials that occurred in the HIA, where because of shortages patients had to purchase some treatment materials from private sources rather than free treatment. First aid cost was relatively small (less than GH¢10.00) for both districts and represented the least amount spent during an episode of cholera. This is in consonance with the suddenness of the onset of cholera with healthy individuals becoming critically ill within few hours and hence need to be rushed quickly to the hospital. In general, average direct cost in the HIA is almost 2 times more than that of the LIA, signaling a proportional increase in costs with higher incidence. Cost compositions from findings of Sarker et al. (2013) study, suggest transportation costs to be the largest component of direct non-medical costs (USD 1.90) followed by caregiver costs (USD 1.50) before feeding costs (0.80). This situation is only somewhat different to that which persists in Ghana. When transportation costs are aggregated for patients and caregivers they cost USD 13.50 which is higher than feeding costs (USD 12.92) but there are no costs attributed to caregiver payments. This is not a normal practice in Ghana, because most often caregivers are relatives of sick patients and are not paid caregivers. Caregivers in turn only incur indirect costs associated with days missed away from normal economic activity. University of Ghana http://ugspace.ug.edu.gh 94 Figure 5.1 Household Direct Average Costs in High and Low Incidence Areas Source: Survey Data; Author’s computation from Excel 5.2.2 Indirect Cost of Cholera within Higher and Lower Incidence Areas The survey found average daily wage for households in Shai-Osudoku to be higher (GH¢26.90) but the district’s average household size (3.6) smaller than that in La-Dadekotopon (GH¢22.80; 3.7). These have bearings on the estimation of costs within these two districts since with higher average incomes the opportunity cost away from work would be much greater compared to that in the LIA. Also, because incomes are averaged by the household size; larger household sizes mean lower per capita income within that household therefore indicating a greater burden of the disease on that household. In both districts, days missed by patients during recovery formed the largest composition of indirect costs but was 1% higher in La-Dadetokopon. Indirect 0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00 First Aid Treatment Feeding Admission Transportation (Patient) Transportation (Caregiver) Total Household Direct Average Costs (GH¢) Shai-Osudoku La-dadekotopon University of Ghana http://ugspace.ug.edu.gh 95 costs associated with travel time was not significant in both districts and is explained by existence of fairly easy access to transportation in both districts and even though Shai-Osudoku is rural, its proximity to the capital Accra makes transportation means much better than that which pertains in most typical rural settings. Average admission days was the same for both districts (3 days) but admission days formed a larger composition of total indirect cost in the LIA (25%) than the HIA (19%) but interestingly, patients from LIA spent an average of 5 days away from normal activities, 2 days lesser than in the HIA (7 days). These suggest that complete recovery was much faster and better in the LIA than the HIA. Total productivity loss by patients was GH¢ 141, 656.40 and GH¢ 56,293.20 in HIA and LIA respectively. That of caregivers was GH¢56,293.20 (HIA) and GH¢28,809.90 (HIA). Together these costs formed more than 70% of indirect cost composition. Figure 5. 3 & 5. 4 Graph of the composition of Indirect Cost for high and low incidence areas. Source: Survey Data; Author’s computation from Excel 47% 28% 25% 0% Shai-Osudoku Days missed (Patient) Days missed (Caregiver) Admission Days Travel Time - Days 48% 33% 19% 0% La-Dadekotopon Days missed (Patient) Days missed (Caregiver) Admission Days Travel Time - Days University of Ghana http://ugspace.ug.edu.gh 96 From Table 5.4, an average 25 days were missed in total by patients and caregivers away from their normal economic activity in the LIA but this was almost doubled in the number of days missed in the HIA (48 days). This translated into GH¢11,055.07 and GH¢831.52 average indirect cost for the selected sample in La-Dadekotong and Shai-Osudoku respectively. Indirect costs in La-Dadekotong were greater for all components than in Shai-Osudoku with the exception of costs associated with productive days missed by caregivers. Table 5.4 Days missed by Patients and Caregivers with Indirect Costs in High and Low incidence areas Days missed (Patient) Days missed (Caregiver) Admission Days Travel Time (Days) Total La- Dadekopon 2071 1451 823 5 14409 Average 6.90 4.84 2.74 0.02 48.03 Shai- Osudoku 655 388 357 1 6816 Average 4.71 1.41 2.57 0.00 24.79 Total 2726 1839 1180 6 5751 GH¢ GH¢ GH¢ GH¢ GH¢ La- Dadekopon HH Total Indirect Cost 141,656.40 56,293.20 99,248.40 331.03 297,529.03 Shai- Osudoku HH Total Indirect Cost 52,858.50 28,809.90 31,311.60 107.21 113,087.21 La- Dadekopon HH Average Indirect Cost 502.33 199.62 351.94 1.17 1,055.07 Shai- Osudoku HH Average Indirect Cost 388.67 211.84 230.23 0.79 831.52 Difference 113.66 -12.22 121.71 0.39 223.54 Source: Survey Data; Author’s computation from Excel University of Ghana http://ugspace.ug.edu.gh 97 In Figure 5.5 all costs are summed and represented by the incidence area. It shows household costs of GH¢393,973.33 and GH¢140,078.51 for sample selected in La-Dadekotopon and Shai-Osudoku districts respectively during the 2014 cholera outbreak in Ghana. In La-Dadekotopon indirect costs was markedly greater (above GH¢200,000 more) but in Shai-Osudoku just slightly above GH¢65,000. Figure 5.5 Total Costs in High and Low cholera incidence areas. Source: Survey Data; Author’s computation from Excel La-Dadekotopon Shai-Osudoku Total Total Direct Cost (GH¢) 96,444.30 26,991.30 123,435.60 Total Indirect Cost (GH¢) 297,529.03 113,087.21 410,616.24 Sample Total Cost GH¢ 393,973.33 140,078.51 534,051.84 0.00 100,000.00 200,000.00 300,000.00 400,000.00 500,000.00 600,000.00 Co st s (G H ¢) Total Costs in High and Low Incidence area University of Ghana http://ugspace.ug.edu.gh 98 Figure 5.6 Per capita cost in High and Low incidence area. Source: Survey Data; Author’s computation from Excel In total, 2,222 cholera cases were reported in health facilities; HIA (1907) and LIA (315). These included cases of individuals who were from other districts but sought treatment within the selected districts. When costs are projected for the total number of cases reported for both districts, the total cost of the 2014 cholera outbreak in these areas GH¢2,504,357.12 (USD 782,611.60) and GH¢317,444.10 (USD 99,201.28) for HIA and LIA respectively. Hence, if cases are reduced to levels of the LIA, cost saving will be GH¢2,186,913.02 (USD 683,410.32). 321.48 194.18 515.66 991.76 813.58 1,805.34 1,313.24 1,007.76 2,321.00 La-Dadekotopon Shai-Osudoku Total Per capita Cost within two districts Percapita Total Direct Cost GH¢ Percapita Total Indirect Cost GH¢ Percapita Total Cost GH¢ University of Ghana http://ugspace.ug.edu.gh 99 Figure 5.1 Total Cost of 2014’s Cholera Outbreak in High and Low Incidence Area. Source: Survey Data; Author’s computation from Excel 5.3 Facility Costs in High and Low Incidence areas Facility costs in treating one cholera patient included items such as gloves, IV Fluids, given set, cotton wool, surgical gloves, spirit and cannular. Treatment of cholera during the outbreak was covered by the government and facilities were to bear no costs but facilities still bore extra costs in providing these items when there were shortages. The average cost of treatment in La General Hospital amounted to GH¢35 (USD 10.94) and GH¢20 (USD 6.25) in Shai-Osudoku District Hospital, these amounts when multiplied by the total number of cholera cases treated at these facility (regardless of which district patient came from; 1907 in HIA and 315 in LIA) amounted to a total of GH¢66,745 (USD 20,862.58) and La-Dadekotopon, GH¢2,504,357.12 Shai-Osudoku, GH¢317,444.10 Total, GH¢2,821,801.22 La-Dadekotopon, USD$782,611.60 Shai-Osudoku, USD$99,201.28 Total, USD$881,812.88 University of Ghana http://ugspace.ug.edu.gh 100 GH¢6,300 (USD 1,968.75) in La General Hospital and Shai-Osudoku District Hospital respectively. Other costs such as physician and nurses time costs and facility physical structure use were not easy to capture due to the shared nature of these costs. 5.4 Correlates of Cholera incidence using Tobit Regression Model The Tobit model estimates linear relationships between variables when there is either left or right-censoring in the dependent variable and also called censoring from below and above, respectively. Censoring from above takes place when cases with a value at or above some threshold, all take on the value of that threshold, so that the true value might be equal to the threshold, but it might also be higher. In the case of censoring from below, values those that fall at or below some threshold are censored. In this study, the dependent variable (proportion of household that contracted cholera) is censored from below at 0.09 proportion of household infection. The highest value in the censored range is 1 and represents an entire case of complete household infection and 0.09, the least proportion of household infection. These proportions represent the level of incidence within these households and this is calculated on the number of household members who contracted per the household size. The tobit regression model gives a useful advantage by truncating off the proportion of household that did not contract cholera from the regression in order to access more appropriately the impact incidence. Hence, because all households within the sample experienced at least a case of cholera, the regression captures the observation at the household level. After the estimation, University of Ghana http://ugspace.ug.edu.gh 101 marginal effects are calculated on the censored sample to adjust for the representativeness in the sample. A multicollinearity test is run using a correlation matrix to determine if two or more of the explanatory variables are highly correlated. Multicollinearity becomes a problem in a model when the pair-wise correlation coefficient is greater than or equal to 0.8 (Gujarati, 2006 pp.372). When the test was run for the model in this study, none of the variables had a correlation coefficient up to 0.8 with each other and hence the model passes this test. (See Appendix B for correlation matrix). In addition, a tobit regression estimation from STATA also runs Likelihood Ratio Chi-Square test in determining whether all predictors' regression coefficients in the model are simultaneously zero. Per the model, the probability chi2 is 0.000 and shows that the overall significance of model at 1 percent. Though the model reports a pseudo R2 but this is equivalent to the R2 found in an OLS regression which is the proportion of variance of the response variable explained by the predictors, the pseudo R2 is usually not attached much importance. Table 5.5 shows the tobit regression output and calculated marginal effects for the censored sample indicating the impact of the recent outbreak of cholera. University of Ghana http://ugspace.ug.edu.gh 102 Table 5.5 Tobit Regression Output showing coefficients, marginal effects for censored sample, standard errors and p-values. Tobit Regression Censored Sample Proportion of HH Coef. Std. Err. P>t dy/dx Std. Err. P>z Income Grouping (Above GH¢750) C1: No Income 0.073** 0.032 0.022 0.019** 0.009 0.036 C2: Less than GH¢100 0.002 0.033 0.960 0.000 0.009 0.960 C3: >GH¢100 GH¢350 chi2 = 0.0000 University of Ghana http://ugspace.ug.edu.gh 103 Log likelihood = -3801.5065 Pseudo R2 = 0.0398 Significance levels = ***p<0.01, **p<0.05, *p<0.1 Source: Author’s Survey, 2015. Socioeconomic and demographic factors have been reported to significantly influence the vulnerability of infection among populations and contribute to epidemic spread (Ali et al., 2002a, 2002b; Borroto and Martinez-Piedra, 2000; Ackers et al., 1998; Sasaki et al., 2008; Sur et al., 2005). Also, a research report by Oxfam on gender and vulnerability to cholera in Sierra Leone states factors such as sex, age and social status are all factors that may contribute to individuals’ vulnerability to cholera (Oxfam, 2013). The regression result supports the fact that the impact of cholera is felt largest by the lower income categories. The Tobit regression output shows that with a unit increase of individuals in households of category 1 (those earning no income) there is 0.07 increase incidence of cholera infection compared to those in the highest income category (above GH¢750). The impact reduces for the categories 3 and 4 where the proportion of infection is 0.06 in both categories. All are statistically significant at 10%. These results corroborate the percentages in the crosstab calculations on the infection rates among these income categories. See results in appendix D. Both Borroto & Martinez-Piedra (2000) Talavera & Pérez (2009) studies support impacts of cholera being heaviest on the least income people and people burdened by poverty. On the other hand, when the marginal effects are calculated for the censored sample, the impact of a unit change in University of Ghana http://ugspace.ug.edu.gh 104 income in all categories leads to the same proportion of infection (0.019) in relation to the highest income category. Unmarried folks are more associated with eating outside of home and street food vending was seen as a key risk factor in the incidence of cholera in Accra during the outbreak. Households with unmarried persons had 0.14 greater incidence than their married counterparts. The marginal effects on the censored sample show relatively similar impact. An Oxfam research report on gender and vulnerability to cholera in Sierra Leone (2013) showed that infection rates were higher in males that were unmarried compared those that were married. Within the household composition, the impact of cholera was greatest among dependents of household heads in the sample. There was 0.15 higher proportion of cholera incidence among this group in relation to the head of household. Spouses however had 0.04 less incidence of the disease during the outbreak. With the level of incidence in relation to age, despite the results show statistical significance at 1% the level of impact of age is small. With an additional year in age of a household member, there is barely 0.002 greater incidence in proportion of cholera infection. Indicating the impact of age on cholera incidence is very low. But Deen et al. (2008) in a study of three cholera endemic areas (Jakarta, Kolkata and Biera) found that in all three area the impact of the disease was highest among children under five. The incidence of cholera and sources of water cannot be exaggerated and studies such as Crooks and Hailegiorgis (2014) corroborate this fact. The safest source of University of Ghana http://ugspace.ug.edu.gh 105 drinking water as stipulated by the WHO is piped water on premises. Nketiah- Amponsah et. al’s (2009) study on the socioeconomic determinants of drinking water source in Ghana found that income increases access to piped water in residence by 29 percentage points. Asante (2003) also found a significant statistical relationship between income and access to safe or portable water. Based on these, inside plumbing and inside standpipe was set as the reference base for which other sources were estimated. In the sample for this study close to 70% of the households get their drinking water from a piped source, that is, an inside plumbing or in-house stand pipe (27.5%), tab water in neighbouring house (18.1) or public stand pipe (42.0%). These together formed 87.6% of drinking water sources that qualify as portable sources. The only water source that showed significance in relation to infection of cholera was water from neighbouring house. There was 0.058 incidence in cholera in households that had their water sources from a neighbouring house as compared to household having inside plumbing as source of drinking water. 5.5 Summary of Chapter This chapter has discussed the findings from the study. Cholera’s economic impact in Ghana is not well documented and this study set out to investigate the costs of the country’s most recent outbreak at the household and facility level plus impact of that the disease had on households. These costs were assess based on direct and indirect costs resulting from cholera infection and impact on households assessed on the income categories of households plus some demographic characteristics of the households. Results from the study revealed University of Ghana http://ugspace.ug.edu.gh 106 that costs of the disease were significantly higher in the HIA as expected but contrary to expectation, the impact of treatment costs was large despite government’s efforts in making treatment of the disease free during the outbreak. Average treatment cost per household to amounted GH¢47.74 in LIA and GH¢49.49 in HIA which were almost twice the average household income in the LIA (GH¢26.90) but more than twice in the HIA (GH¢22.80). The tobit regression output clearly showed increased proportion of infection within income categories 1 (0.073), 4 (0.061) and 3 (0.059) respectively. Factors such as age, marital status, role in household and drinking water source were significant and characterized various levels of impact of the cholera episode. University of Ghana http://ugspace.ug.edu.gh 107 CHAPTER SIX SUMMARY, CONCLUSION AND POLICY RECOMMENDATIONS 6.0 Introduction This chapter summarizes the study and presents conclusions and recommendations for policy consideration. These findings and recommendations can be the bases for action against the menace that cholera causes. Finally, the limitations of this study are outlined in the last section including areas for which future research can target. 6.1 Main findings of study This study suggests that preventing one episode of cholera outbreak in a HIA will avert a total cost of GH¢2,504,357.12 (USD 782,611.60) or when reduced to LIA rates will reduce costs to GH¢317,444.10 (USD 99,201.28). Key findings are detailed as follows; Firstly, total direct cost in HIA amounted to GH¢96,444.30 (USD 30,138.84) about four times lesser in the LIA (26,991.30; USD 8,434.78). Treatment costs were the largest component of direct costs in both districts (49.49% in HIA and 47.74% in LIA). With admission costs included, direct medical costs rose by 24% in both districts. Secondly, transportation costs in both districts were not important indirect cost factors. Average admission days were the same in both districts (3 days) but University of Ghana http://ugspace.ug.edu.gh 108 admission days formed higher composition of indirect cost in LIA (33%) compared to that of HIA (28%). Fourly, facility costs were 10 folds higher in the HIA than in LIA. Total costs for all cases reported in the HIA health facility was GH¢66,745 (USD 20,862.58) and GH¢ 6,300 (USD 1,968.75). Finally, factors such as income quintile, marital status, age and some drinking water sources were significant correlates with the incidence of cholera. 6.2 Contributions and Policy Recommendations of the Study The main contribution of the study is the detailing and capturing of costs associated with cholera incidence on Ghana where a gap in literature exited. Based on detailed findings, the following recommendations are made; One, costs of cholera for Ghana have been clearly defined by this study hence, policy implementers have a firm footing backed with figures from the study to make claims for appropriate interventions like investments into cholera vaccines or investments in appropriate drainage systems seeing the grave financial impact these outbreaks cause. Two, findings revealed that treatment costs formed the greatest composition of direct cost regardless of the fact that treatment costs for cholera during the outbreak was declared free by the government. The study has shown a gap in policy and its implementation and the need for government and other related non- governmental organizations making investments into cholera related issues to ensure correct implementation of their policies. Additionally, the government needs to do an assessment on whether its policy intervention such as free treatment is faced with challenges and if so, the best strategies in overcoming this. University of Ghana http://ugspace.ug.edu.gh 109 This information is critical in guiding government and other agencies financial budgeting in curbing the impact of future outbreaks. Three, in both districts indirect cost were more than twice that of direct costs on households. This prelude that impact of the disease leads to huge indirect income losses to both patients and caregivers. Therefore to lessen such impacts, social intervention packages such as free feeding can be put in place to mitigate the impact of the disease on the household. Four, the study proved that travel time was not an important factor and that most patients could easily get to facilities. Contrastingly, most deaths were as a result of patients arriving late to health facility. This indicates that most deaths can be prevented if education on treatment and first aid procedures is intensified. Most affected persons will able to make better informed choices and needless deaths will be averted. This education needs to be under by hospitals and government agencies like National Commission for Civic Education and the Information Services Department. Five, disparities in infections at variously income strata plus other characteristics such as age, education, sex, access to portable drinking water etc. shows that issues of discrimination and equity of resources distribution needs a re- examination by government agencies if the efforts into tackling cholera incidence are to be holistically addressed. This will effectively address the problem of cholera being a disease of poverty. University of Ghana http://ugspace.ug.edu.gh 110 6.3 Limitations of the study and areas for further research A nation-wide study would have given a complete picture of the costs associated with the cholera outbreak in Ghana because the 2014 outbreak affected all ten regions but time and financial constraints inhibited widening the scope of the study. These same constraints; time and finances also limited the study from doing further comparative with other disease conditions or a control group of households where there were no incidence and hence assess the relative impact of the disease and thereby establishing priorities. Costs borne by the government is important in assessing the total cost of the disease but this information was difficult to ascertain due to the complex distributional channels hospitals receive their supplies. Moving forward, cholera in mainly a water and waste management problem and there is a need for cost-benefit analysis of interventions such increased access to pipe water on premises, management of drainage system in preventing flooding that leads to contamination of water and costs of investments in public education to increase sanitation knowledge. University of Ghana http://ugspace.ug.edu.gh 111 References Ackers, M.L., Quick, R.E., Drasbek, C.J, Hutwagner, L. & Tauxe, R.V. (1998). Are there national risk factors for epidemic cholera? The correlation between socioeconomic and demographic indices and cholera incidence in Latin America. International Journal of Epidemiology; 27(2):330-334. Aikins, M., Armah, G., Akazili, J. & Hodgson, A. (2010). Hospital health care cost of diarrheal disease in Northern Ghana. S130. Journal of Infectious Diseases; 202 (1): 126 -129. Akazili, J., Aikins, M. & Binka, F.N. (2007). Malaria treatment in Northern Ghana: What is the treatment cost per case to households? African Journal of Health Sciences; 14:70-79. Ali, M., Lopez, A.L., You, Y.A., Kim, Y.A., Sah, B., Maskery, B., & Clemens, J. The global burden of cholera. (2012). Bulletin of the World Health Organization; 90: 209–218. Ali, M., Emch, M., Donnay, J.P. Yunus, M. & Sack, R.B. (2002b). The spatial epidemiology of cholera in an endemic area of Bangladesh. Social Science & Medicine; 55(6):1015-1024. Ali, M. Emch, M., Donnay, J.P, Yunus, M. & Sack R.B. (2002a). Identifying environmental risk factors of endemic cholera: a raster GIS approach. Health & Place; 8(3):201-210. University of Ghana http://ugspace.ug.edu.gh 112 Amankwa, J. (2011). Cholera epidemic kills 60 in Ghana. Ghanaweb; 18th March, 2011. Source:http://www.ghanaweb.com/GhanaHomePage/NewsArchive/artikel.php?I D=205245 Accessed: Friday, November 21, 2014 09:13:12 AM. Ament, A. & Evers, S. (1993). Cost of illness studies in health care: a comparison of two cases. Health Policy; 26:29-42. Anbarci, N., Escaleras, M., & Register, C. A. (2012). From cholera outbreaks to pandemics: the role of poverty and inequality. American Economist; 57(1): 21 - 27 Andersen, R.M. (1995). “Revisiting the behavioural model of and access to medical care: Does it matter?” Journal of Health and Social Behavior. 36(1): 1-10 Asante, F.A. & Asenso-Okyere, K. (2003). Economic Burden of Malaria in Ghana. A Technical Report Submitted to the World Health Organization (WHO), African Regional Office (AFRO). Asante, F. A. (2003). Economic Analysis of Decentralization in Rural Ghana. Development Economics and Policy, 32, Peter Lang, Frankfurt am Main. Ashitey, G.A. (1994). An epidemiology of disease control in Ghana, 1901-1990. Ghana Universities Press. University of Ghana http://ugspace.ug.edu.gh 113 Bardasi, E. & Wodon, Q. (2008). Who Pays the Most for Water? Alternative Providers and Service Cost in Niger. Economics Bulletin; 9(20):1-10. Borroto, R.J. & Martinez-Piedra, R. (2000). Geographical patterns of cholera in Mexico; 1991-1996. International Journal of Epidemiology; 29(4):764- 772. Briggs, L.C. (2002) Stories in the time of cholera: Racial profiling during a medical nightmare. University of California Press. Brouwer, W.B.F., Koopmanschap, M.A. & Rutten F.F.H. (1997). Productivity costs measurement through quality of life? A response to the recommendation of the Washington panel. Health Economics; 6:253–259. Byford, S., Torgerson, J.D. & Raftery J. (2000). Cost of Illness studies. British Medical Journal; 320(7245):1335. Chima, R.I., Goodman, C. & Mills, A. (2003). The economic impact of malaria in Africa: a critical review of the evidence. Health Policy; 63:17–36. Christine, P., Riewpaiboon, A., Stewart, J. F., Clemens, J., Guh, S., Agtini, M., Sur, D., Islam, Z., Lucas, M., Whittington, D., & DOMI Cholera COI Study Group. (2008). Costs of Illness Due to Endemic Cholera. Epidemiology and Infection; 140(3):500-509. Colwell, R.R. (2002); A voyage of discovery: cholera, climate and complexity. Environmental Microbiology; 4:67-69. University of Ghana http://ugspace.ug.edu.gh 114 Colwell, R.R., Huq, A., Islam, M.S., Aziz, K.M.A., Yunus, M., Huda-Khan, N., Mahmud, A., Sack ,R.B., Nair, G.B., Chakraborty, J., Sack, D. A. & Russek-Cohen, E. (2003). Reduction of cholera in Bangladeshi villages by simple filtration. Proceedings of the National Academy of Sciences of the USA; 100:1051–55. Cooper, B.S. & Rice, D.P. (1976). The economic cost of illness revisited. Social Security Bulletin, 39:21-36. Crooks, A.T., & Hailegiorgis, A.B. (2014). An agent-based modeling approach applied to the spread of cholera. Environmental Modelling & Software; 62; 164-177. Davis-Teye, B.B.K., Vanotoo, L., Yabani, J.B. & Kwaakye-Maclean C. (2014). Socio-economic Factors associated with Cholera Outbreaks in Southern Ghana. 20th IEA World Congress of Epidemiology. Global Epidemiology in a changing environment: The circumpolar perspective. August 17-21, 2014. Anchorage, Alaska USA. De Magny, G.C., Cazelles, B. & Guegan J. (2006). Cholera Threat to Humans in Ghana Is Influenced by Both Global and Regional Climatic Variability. EcoHealth; 3:223–231 Deen, J.L., von Seidlein, L., Sur, D., Agtini, M., Lucas, Lopez, A.L., Kim, D.R., Ali, M., & Clemens. J.D. (2008). The High Burden of Cholera in Children: Comparison of Incidence from Endemic Areas in Asia and Africa. PLoS Neglected Tropical Diseases; 2(2): e173. University of Ghana http://ugspace.ug.edu.gh 115 Drummond, M.F., Sculpher, M.J., Torrance, G.W., O'Brien, B.J. & Stoddart, G.L. (1987). Methods for the Economic Evaluation of Health Care Programmes, Oxford University Press. Faruque, S.M., Biswas, K., Udden, S.M., Ahmad, Q.S., Sack, D.A., Nair, G.B. & Mekalanos J.J. (2006). Transmissibility of cholera: in vivo-formed biofilms and their relationship to infectivity and persistence in the environment. Proceedings of the National Academy of Sciences of the USA; 103: 6350–55. Faruque, S.M., Naser, I.B. & Islam, M.J. (2005). Seasonal epidemics of cholera inversely correlate with the prevalence of environmental cholera phages. Proceedings of the National Academy of Sciences of the USA; 102: 1702- 1707. Fobil, J., May J. & Kraemer, A. (2010). Assessing the Relationship between Socioeconomic Conditions and Urban Environmental Quality in Accra, Ghana. International Journal of Environmental Research and Public Health; 7: 125-145. GHS (2014). Report of cholera outbreak in the Greater Accra Region. Ghana Health Service, Disease Control Unit, Accra. GHS (2014). Report of Cholera Outbreak Investigation, Greater Accra Region, June-August 2014. Ghana Health Service, Accra. University of Ghana http://ugspace.ug.edu.gh 116 GHS (2015). Report on cholera outbreak in the Greater Accra Region: June to December, 2014. Ghana Health Service, Accra. Gold, M.R., Siegel, J.E., Russell, L.B. & Weinstein, M.C. (eds) (1996). Cost- Effectiveness in Health and Medicine. Oxford: Oxford University Press. Government of Ghana. (1998). Children’s Act, 1998. Act 560, Part V – Employment of Children Sub-Part I – Child Labour; 89: 27. Grossman, M. (1972). On the Concept of Health Capital and the Demand for Health. The Journal of Political Economy; 80(2): 223-225. Grossman, M. (2000). “The Human Capital Model” in Handbook of Health Economics, Edited by Culer, A.J. & Newhouse J.P. Elsevier Science; 1:347-408. GSS (2012). 2010 Population and Housing Census; Summary Report of Final Results. Ghana Statistical Service, Accra. GSS (2013). 2010 Population and Housing Census. National Analytical Report. Ghana Statistical Service, Accra. GSS (2014). 2010 Population and Housing Census; District Analytical Report. La-Dadekotopon. Ghana Statistical Service, Accra. GSS (2014). 2010 Population and Housing Census; District Analytical Report. Shai-Osudoku, Ghana Statistical Service. University of Ghana http://ugspace.ug.edu.gh 117 Harwood, H. J. (1994). Analytical principles and issues in making cost of illness estimates for substance abuse. In International Symposium on the Economic Costs of Substance Abuse, Banff, Canada. pp. 11-13. International Vaccine Institute. (2010). Economic impact of a cholera epidemic on Mozambique and Bangladesh. Oxford Economics; 4-12. Jeuland, M., Whittington, D. (2009). Cost-benefit comparisons of investments in improved water supply and cholera vaccination programs. Vaccine; 27(23): 3109-3120. Kimball, A. M., Wong, K. Y., & Taneda, K. (2005). An evidence base for International Health Regulations: quantitative measurement of the impacts of epidemic disease on international trade. Revue scientifique et technique (International Office of Epizootics); 24(3): 825-832. Kirigia, J. M., Sambo, L. G., Yokouide, A., Soumbey-Alley, E., Muthuri, L. K., & Kirigia, D. G. (2009). Economic burden of cholera in the WHO African region. International Health and Human Rights; 9(1): 8. Koelle, K., Pascual, M., & Yunus, M. (2005). Pathogen adaptation to seasonal forcing and climate change. Proceedings of the Royal Society of London B: Biological Sciences, 272(1566), 971-977. Lamond, E. & Kinyanjui, J. (2012). Cholera outbreak Guidelines (Preparedness, Prevention and Control). Oxfam. p 7. Laws, E. (2006). Case study: cholera. Oceanography - Washington DC- Oceanography Society; 19:81-83 University of Ghana http://ugspace.ug.edu.gh 118 Malaney, P. (2003). Micro-economic Approaches to evaluating the Burden of Malaria. Centre for Infectious Diseases working paper; No. 99. Naficy, A., Rao, M.R., Paquet, C., Antona, D., Sorkin, A. & Clemens, J.D. (1998). Treatment and Vaccination strategies to control cholera in sub- Saharan refugee settings: A cost-effectiveness analysis. Journal of the American Medical Association; 279(7): 521-525 Nelson, E.J., Chowdhury, A., Flynn, J., Schild, S., Bourassa, L., Shao, Y., LaRocque, R.C., Calderwood, S.B., Qadri, F. and Camilli, A. (2008). Transmission of Vibrio cholerae is antagonized by lytic phage and entry into the aquatic environment. PLoS Pathogens; 4(10): 187. Nelson, E.J., Chowdhury, A., Harris, J.B., Begum, Y.A., Chowdhury, F., Khan, A.I., LaRocque, R.C., Bishop, A.L. , Ryan, E.T., Camilli, A., Qadri, F. & Calderwood, S.B. (2007). Complexity of rice-water stool from patients with Vibrio cholerae plays a role in the transmission of infectious diarrhea. Proceedings of the National Academy of Sciences of the USA; 104: 19091–19096. Nelson, E.J., Harris, J.B., Morris, J.G. Jr, Calderwood, SB & Camilli, A. (2009). Cholera transmission: the host, pathogen and bacteriophage dynamic. National Review Microbiology; 7: 693–702. Nketiah-Amponsah, E., Aidam, P.W. & Senadza, B. (2009). Socio-economic Determinants of Sources of Drinking Water: Some Insight from Ghana. University of Ghana http://ugspace.ug.edu.gh 119 Conference on International Research on Food Security, Natural Resource Management and Rural Development. October 6-8, 2009. Hamburg, Germany. Nketiah-Amponsah E., Arthur, E. and Abuosi, A. (2012). Correlates of Contraceptive use among Ghanaian women of Reproductive age (15-49 years). African Journal of Reproductive Health; 16(3): 154-167. Ofori-Adjei, D. and Koram, K. (2014). Editorial commentary of Cholera and Ebola virus disease in Ghana. Ghana Medical Journal; 48(3): 120. Osei, F. B., Duker, A. A., & Stein, A. (2012). Evaluating spatial and space-time clustering of cholera in Ashanti-Region-Ghana. INTECH Open Access Publisher. Source: ttp://www.intechopen.com/books/cholera/evaluating-spatial-and-space- time-clustering-of-cholera-in-ashanti-region-ghana. Date Accessed: Tuesday, February 10, 2015 10:45:57 AM Osei, F.B., Duker, A.A., Augustijn, E. & Stein, A. (2010). Spatial dependency of cholera prevalence on potential cholera reservoirs in an urban area, Kumasi, Ghana. International Journal of Applied Earth Observation and Geoinformation; 12:331-339. Osei, A.A. & Duker, A.A. (2008). Spatial and demographic patterns of Cholera in Ashanti region – Ghana. International Journal of Health Geographics; 7: 44. University of Ghana http://ugspace.ug.edu.gh 120 Oxfam (2012). Lamond, E. & Kinyanjui J. Cholera outbreak guidelines. Preparedness, Prevention and Control. Pp.7 Oxfam (2013). Gender and vulernability to cholera in Sierra Leone. Gender analysis of the 2012 cholera outbreak and an assessment of Oxfam’s response. Penrose, K., Castro, M.C., Werema, J. & Ryan, E.T.. (2010). Informal Urban Settlements and Cholera Risk in Dar es Salaam, Tanzania. PLoS Neglected Tropical Diseases; 4(3): e631. Pobee, J.O.M & Grant, F. (1970). Case report of Cholera. Ghana Medical Journal; 9(4):306-309 Poulos, C., Riewpaiboon, A., Stewart, J. F., Clemens, J., Guh, S., Agtini, M., Sur, D., Islam, Z., Lucas, M. & Whittington, D. (2012). Costs of Illness Due to Endemic Cholera. Epidemiology & Infection; 140(3): 500-509. Prescott, N. M. (1999). “The economics of Schistosomiasis and Development,” World Development: 7(1):1-14. Rice, D.P. (1966). Estimating the Cost of Illness. Health Economics Series, No. 6. DHEW Pub. No. (PHS) 947-6. Rockville, MD: U.S. Department of Health, Education and Welfare. Rice, D.P. (1967). Estimating the Cost of Illness. American Journal of Public Health, 57: 424-440. Rice, D.P. (1994). Cost-of-illness studies: fact or fiction? Lancet; 344:1519-20. University of Ghana http://ugspace.ug.edu.gh 121 Rice, D.P., Hodgson, T.A. & Kopstein, A.N. (1985). The economic costs of illness: a replication and update. Health Care Financing Reviews; 7: 61- 80. Ruiz-Moreno, D., Pascual, M., Emch, M. & Yunus, M. (2010). Spatial clustering in the spatio-temporal dynamics of endemic cholera. Infectious Diseases; 10: 51. Russell, S. (2004). The economic burden of illness for households in developing countries: a review of studies focusing on malaria, TB and HIV/AIDS. American Journal of Tropical Medicine and Hygiene; 71(2):147-155. Sarker, A.R., Islam, Z., Khan, I.A., Saha, A., Chowdhury, F., Khan, A.I., Qadri, F. & Khan, J.A.M. (2013): Cost of illness for cholera in a high risk urban area in Bangladesh: an analysis from household perspective. Infectious Diseases; 13:518. Sasaki, S., Suzuki, H., Igarashi, K., Tambatamba, B. & Mulenga, P. (2008). Spatial Analysis of Risk Factor of Cholera Outbreak for 2003-2004 in a peri-urban Area of Lusaka, Zambia. American Journal of Tropical Medicine and Hygiene; 79(3):414-421. Sauerborn, R., Adams, A. & Hien, M. (1996). Household strategies to cope with the economic costs of illness. Social Science and Medicine; 43: 291-301. Schaetti, C., Weiss, M.G., Ali, S.M., Chaignat ,C., Khati, A.M., Reyburn, R., Tebbens, R.J.D & Hutubessy, R. (2012). Costs of Illness Due to Cholera, University of Ghana http://ugspace.ug.edu.gh 122 Costs of Immunization and Cost-Effectiveness of an Oral Cholera Mass Vaccination Campaign in Zanzibar. PLoS Neglected Tropical Diseases; 6(10): e1844. Shepard, D.S., Ettling, M.B., Brinkmann, U. & Sauerborn, R. (1991). The economic cost of malaria in Africa. Tropical Medicine and Parasitology; 42:199-203. Source: http://www.gdnonline.org/resources/oxfam-gender-cholera-vulnerability- sierra-leone-170613-en.pdf Accessed: Thursday, November 20, 2014 10:17:50 PM Smith, K.F., Dobson, A.P., McKenzie, F.E., Real, L.A., Smith, D.L. & Wilson, M.L. (2005). Ecological Theory to enhance infectious disease control and public health policy. Frontiers in Ecology and the Environment; 3:29–37. Suarez, R. & Bradford, B. (1993). The Economic Impact of the Cholera Epidemic in Peru: An Application of the Cost of Illness Methodology. Water and Sanitation for Health (WASH) Project Field Report 415 Sur, D. Deen, J. Manna, B. Niyogi, S. Deb, A. Kanungo, S. Sarkar, B. Kim, D. Danovaro- Holliday, M. Holliday, K. Gupta, V. Ali, M. von Seidlein, L. Clemens, J. & Bhattacharya, S. (2005). The burden of cholera in the slums of Kolkata, India: data from a prospective; community based study. Achieves of Diseases of Childhood; 90(11): 1175-1181. Talavera, A. & Pérez, E.M. (2009). Is cholera disease associated with poverty? Journal of Infections of Developing Countries; 3(6):408-411. University of Ghana http://ugspace.ug.edu.gh 123 Thompson, C.C., Freitas, F.S., Marin, M.A., Fonseca, E.L., Okeke, I.N. & Vicente, A.C.P. (2011). Vibrio cholerae O1 lineages driving cholera outbreaks during seventh cholera pandemic in Ghana. Infection, Genetics and Evolution; 11:1951-1956. UNEP (2005). Emerging Challenges – New Findings. In: Geo Year Book 2004 – 2005: An Overview of Our Chnaging Environment, New York: United Nations Publications; pp. 71-79. UNICEF (2015). WASH in communities. Empowerment through water and sanitation. Source: http://www.unicef.org/ghana/wes.html. Date Accessed: Thursday, November 20, 2014 11:23:52 PM Wagstaff, A. (1986). “The Demand for Health: Theory and Applications”. Journal of Epidemiology and Community Health; 40:1-11. Weil, A.A., Khan, A.I., Chowdhury, F., LaRocque, R.C., Faruque, A.S.G., Ryan, E.T., Calderwood, S.B., Qadri, F. & Harris, J.B. (2009). Clinical outcomes in household contacts of patients with cholera in Bangladesh. Clinical Infectious Diseases; 49:1473-79. WHO (1999). The World Health Report, 1999: Making a difference. Geneva, Switzerland: World Health Organization. WHO (2000). WHO Report on Global Surveillance of Epidemic-prone Infectious Diseases. Geneva, Switzerland: World Health Organization, 1-43. University of Ghana http://ugspace.ug.edu.gh 124 WHO (2007); Cholera 2006. Weekly Epidemiological Record 2007; 82: 273-284. WHO (2009). Guide to identifying the economic consequences of disease and injury. Department of Health Systems Financing Health Systems and Services World Health Organization Geneva, Switzerland. WHO (2010); Cholera vaccines: WHO position paper. Weekly Epidemiological Record 2010; 85: 117-28. WHO (2014). WHO fact sheet on Cholera. Source: http://www.who.int/mediacentre/factsheets/fs107/en. Accessed: Thursday, November 20, 2014 12:29:59 PM WHO (2014). WHO Outbreak Bulletin, 31 October 2014; 4 (4). WHO, (2014); WHO 10 facts on Cholera. Source: http://www.who.int/features/factfiles/cholera/facts/en/index2.html. Date Accessed: Wednesday, November 12, 2014 12:13:06 PM WHO, (2015) Global Health Observatory (GHO) data. Source: http://www.who.int/gho/epidemic_diseases/cholera/cases_text/en/. Date Accessed: Wednesday, November 12, 2014 9:17:15 PM WHO and UNICEF (2000). The Global Water Supply and Sanitation Assessment 2000 Report. Geneva: World Health Organization. World Bank (1994) Infrastructure for Development at Target. University of Ghana http://ugspace.ug.edu.gh 125 Source: http//www.target.com/world-Development-Report-1994-Infrastructure/ap Date Accessed: Thursday, November 20, 2014 09:36:12 PM WSP (2012). Economic Impacts of Poor Sanitation in Africa. Source: http://www.wsp.org. Date Accessed: Wednesday, November 12, 2014 11:25:10 AM Wyszewianski, L. (1986). “Financially Catastrophic and High Cost Cases: Definitions, Distinctions and their Implications for Policy Formulation.” Inquiry; 23(4): 382-94. Xu, K, Evans, D. B., Carrin, G., Aguilar-Rivera, A.M., Musgrove, P. & Evans, T. (2007). Protecting households from catastrophic health expenditures. Health Affairs; 26:972-983. Xu, K., Evans, D.B., Kawabata, K., Zeramdini, R., Klavus, J. & Murray, C.J. (2003). Household catastrophic health expenditure: a multi-country analysis. Lancet, 362:111-7. University of Ghana http://ugspace.ug.edu.gh 126 APPENDICES APPENDIX A: Sample Size Calculation S = X2 NP(1− P) ÷ d2(N −1) + X2P(1− P). S = required sample size. X2 = the table value of chi-square for 1 degree of freedom at the desired confidence level (3.841). N = the population size. (In this case the population is the number positive cases from a particular district) P = the population proportion (assumed to be .50 since this would provide the maximum sample size). d = the degree of accuracy expressed as a proportion (0.05). Sample Size (Shai-Osudoku District); Population Size (N = 211) S = X2 NP(1− P) ÷ d2(N −1) + X2P(1− P) S = (3.841) (211) (0.5) (1-0.5) ÷ (0.05)2 (211-1) + (3.841) (0.5) (1-0.5) S = 202.61275 ÷ 1.48525 S = 136 Sample Size (La-Dadekotopon District); Population Size (N = 1063) S = X2 NP(1− P) ÷ d2(N −1) + X2P(1− P) S = (3.841) (1063) (0.5) (1-0.5) ÷ (0.05)2 (1063-1) + (3.841) (0.5) (1-0.5) S = 1020.74575 ÷ 3.61525 S = 282 University of Ghana http://ugspace.ug.edu.gh 127 APPENDIX B: Test for Multicolinearity incomeq sex marita~a hhhrel age edulevel nhisen~d DWS incomeq 1.0000 sex 0.0633 1.0000 maritalsta 0.5313 -0.0106 1.0000 hhhrel -0.4866 -0.1193 -0.4267 1.0000 age 0.5559 -0.0643 0.5216 -0.5309 1.0000 edulevel 0.4143 0.0966 0.2296 -0.2052 0.3467 1.0000 nhisenrolled 0.0123 -0.0923 -0.0163 0.0500 -0.0255 -0.0214 1.0000 DWS -0.0949 0.0357 -0.0466 -0.0226 -0.0662 -0.1266 -0.0601 1.0000 APPENDIX C: Descriptive Statistics I. DEMOGRAPHIC STATISTICS Gender distribution within the two districts La-Dadekotopon Shai-Osudoku Sex Freq. Percent Cum. Freq. Percent Cum. 0. Female 568 53.99 53.99 240 48.88 48.88 1. male 484 46.01 100 251 51.12 100 Total 1,052 100 491 100 Household Headship and Gender distribution within the two districts La-Dadekotopon Shai-Osudoku Relationship with HH Head Sex Sex Female Male Total Female Male Total Head of HH 107 185 292 35 108 143 Percent 36.64 63.36 100 24.48 75.52 100 Total 107 185 292 35 108 143 36.64 63.36 100 24.48 75.52 100 Household composition distribution within the two districts Relationship with HH Head La-Dadekotopon Shai-Osudoku Freq. Percent Cum. Freq. Percent Cum. Head of HH 292 27.76 27.76 143 29.12 29.12 Spouse 122 11.6 39.35 76 15.48 44.6 Son/Daughter 441 41.92 81.27 203 41.34 85.95 Other Dependents 197 18.73 100 69 14.05 100 Total 1,052 100 491 100 University of Ghana http://ugspace.ug.edu.gh 128 Education and Gender distribution within the two districts Educational La-Dadekotopon Shai- Osudoku level Sex Sex complete Female 1. male Total Female 1. male Total 1. None 159 127 286 95 91 186 Percent 15.11 12.07 27.19 19.35 18.53 37.88 2. Primary 103 78 181 38 17 55 Percent 9.79 7.41 17.21 7.74 3.46 11.2 3. Middle/JSS 197 125 322 79 84 163 Percent 18.73 11.88 30.61 16.09 17.11 33.2 4. Higher 109 154 263 28 59 87 Percent 10.36 14.64 25 5.7 12.02 17.72 Total 568 484 1,052 240 251 491 Percent 53.99 46.01 100 48.88 51.12 100 II. CHOLERA INFECTION STATISTICS Cholera Infection per District Cholera District Infection La-Dadekotopon Shai-Osudoku Total Not Infected 752 352 1,104 Percent 48.74 22.81 71.55 Infected 300 139 439 Percent 19.44 9.01 28.45 Total 1,052 491 1,543 Percent 68.18 31.82 100 Cholera infection by Gender distribution within the two districts La-Dadekotopon Shai-Osudoku Sex Sex Female Male Total Female Male Total Shai-Osudoku HH members currently enrolled on any health insurance B2 Sex 1. Yes Total 1. Yes Total 0. Female 296 296 185 185 Percent 59.08 59.08 52.86 52.86 1. male 205 205 165 165 Percent 40.92 40.92 47.14 47.14 Total 501 501 350 350 100 100 100 100 La-Dadekotopon Gender and Health Insurance enrolment within the two districts University of Ghana http://ugspace.ug.edu.gh 129 Infected 157 143 300 68 71 139 Percent 52.33 47.67 100 48.92 51.08 100 Infection among Age categories Age Category Total Infected Percent Cum Below 5yrs 43 9.82 9.8 5yrs to 9yrs 32 7.31 17.1 10yrs to 19yrs 71 16.21 33.3 20yrs to 29yrs 109 24.89 58.2 30yrs to 39yrs 75 17.12 75.4 40yrs to 49yrs 49 11.19 86.5 50yrs to 59yrs 35 7.99 94.5 60yrs & above 24 5.48 100.0 Total 438 100 III. CHOLERA INFECTION STATISTICS, HEAD OF HOUSEHOLD, SEX, AGE, EDUCATIONAL AND INCOME LEVEL Gender distribution in household headship Sex Head of HH Female Male Total Total 142 293 435 Percent 32.64 67.36 100 Cholera Infection within HH Relationship Number Infected with HH Head Infected Percent Head of HH 147 33.49 Spouse 51 11.62 Son/Daughter 160 36.45 Other Dependents 81 18.45 Total 439 100 Cholera Infection by gender Female Male Total Infected 225 214 439 Percent 51.25 48.75 100 Cholera Infection by Education Educational Level Infected Cum. Percent None 126 Percent 28.7 28.7 University of Ghana http://ugspace.ug.edu.gh 130 Basic 206 Percent 46.92 75.62 Sec/Tec/Voc 96 Percent 21.87 97.49 Higher 11 Percent 2.51 100 Total 439 Percent 100 Cholera Infection and Income Quintiles Infected Cum. Percent C1: No Income 200 Percent 45.56 45.56 C2: Less than GH¢100 53 Percent 12.07 57.63 C3: >GH¢100 GH¢350GH¢750GH¢100 GH¢350GH¢750 SECTION B: HOUSEHOLD INFORMATION HH ID No. B1: NAME Please give the names of persons who usually l ive and eat in your household starting with the head of household. Add *Name(s) of a caregiver(s) who is not a member of this HH at the bottom B2:SEX M…..1 F…….2 B3: RELATIONSHIP WITH HH HEAD Head of HH…...........……........1 Spouse…………........2 Son/ Daughter.......3 Son/Dau-in-law.....4 Grandchild……......5 Parent………...........6 Parent inlaw.........7 Bro/Sis…................8 Other relative...….9 Not related…….....10 Caregiver not part of HH........................11 B4:AGE Age at last birthday B5:MARITAL STATUS Married….…......1 Divorced…..…….2 Separated…….…3 Widowed…..…….4 Living togthter……….…5 Married but not l iving together……..….6 Never Married………...7 N/A………………99 B6:RELIGION Christian.…...1 Muslim.......…2 Traditional….3 None………….4 Other…………..5 B7:EDUCATION AL LEVEL COMPLETED None………..….1 Primary………..2 Middle/JSS……3 Secondary….…4 Tech/Voc………5 Higher…………..6 B8: Is HH member currently enrolled on any health Insurance Yes……..1 No……..2 Check insurance card for current enrolment Skip B9 if "2" is chosen for HH member B9: Type of Health Insurance NHIS……..1 PHIS……..2 SECTION A: IDENTIFICATION PANEL House Number: NAME OF INTERVIEWER: START TIME: : HH No.: Household NumberHH ID: (LD or SO/IntID/HH No.) : / / DATE: / / DISTRICT NAME & CODE: La-Dadekotopon(LD) ..…..……..…..…......1 Shai-Osudoku(SO)……..……………..….....2 University of Ghana http://ugspace.ug.edu.gh 137 HH ID No. C1: MAIN OCCUPATION Public Sector……..1 Private Formal…...2 Private Informal...3 Agric……………….….4 Self Employed..…5 Retired………….....6 Student……………..7 Unemployed.…...8 Other Specify…….9 Skip to C5 If 6, 7or 8 is chosen C2: APPROXIMATELY HOW MUCH INCOME DOES YOUR MAIN OCCUPATION GENERATE IN A MONTH? Less than GH¢100……...……1 >GH¢100 GH¢350GH¢750GH¢1000GH¢1500GH¢2000.............................7 C3: SECONDARY OCCUPATION Public Sector……..1 Private Formal…...2 Private Informal...3 Agric……………….….4 Self Employed..…5 Student……………..6 Retired………….....7 Unemployed.…...8 Other Specify…….9 NA…………………….99 If NA skip to C5 C4: APPROXIMATELY HOW MUCH INCOME DOES YOUR SECONDARY OCCUPATION GENERATE IN A MONTH? Less than GH¢100……...……1 >GH¢100 GH¢350GH¢750GH¢1000GH¢1500GH¢2000.............................7 C6: Approximately how much did you receive from other income sources like Pension, Relatives or Friends working & living elsewhere over the past 12 months? Amount in GH cedis ID ITEM AMOUNT AMOUNT C5: HOW MUCH DID YOU EARN IN KIND OVER THE LAST MONTH? Indicate item & amount in GH cedis SECTION C: SOCIOECONOMIC & EMPLOYMENT STATUS ( >15YEARS HH MEMBERS) University of Ghana http://ugspace.ug.edu.gh 138 D1: TYPE OF DWELLING Single Family House………….…………..1 Apartment/Flat………………….……….…..2 Compound House……………………….....3 Family House…………………………………4 Other Specify…………………………….……5 D9: What is the main fuel used by household for cooking? Wood…………………...1 Charcoal……………….2 Gas………………………..3 Electricity……………..4 Kerosene……………….5 Other (specify)……..6 None……………………99 D2: WHAT IS THE MAIN ROOF MATERIAL OF DWELLING Thatch/straw…………………………………1 Wood………………………………...…………..2 Corrugated Iron sheets………………….3 Iron sheets……………………………………..4 Cement…………………………………………..5 Asbestos………………………………………….6 Brick Tiles………….……………………………7 Other (specify)…………………………………8 D10: How does your HH dispose off refuse? Waste Mgt Com…...….1 Burning……………………2 Rubbish dump…………3 Burying …………………..4 Others (specify)……….6 D3: HOW MANY ROOMS DOES HH HAVE? Number of rooms: D11: Do you have a toilet facil ity in your house? Yes……………………..1 No…...…………………2 D4: MAIN FLOOR MATERIAL OF ROOMS Earth/sand………………………………………1 Rudimentary/Wood Planks…….…….2 Linoleum…………………………………………3 Cement……………………………………………4 Cement Tiles……………………………….…5 Wall-to-wall carpet…………………..….6 Other (specify)…………………………….….7 D12: What kind of toilet facil ity do most members of your household use? Flush Toilet……...………1 Pit Latrine………….…….2 Pan/Bucket….…………..3 K.V.I.P……………..……….4 Open defecation………….……5 Other (specify)...……..6 D5: MATERIAL OF WALLS OF DWELLING Earth/Mud………………………………………1 Wood………………………………………………2 Cement……………………………………………3 Straw………………………………………………4 Burnt Bricks……………………………...…..5 Other (specify)………………………………..6 D13: Do you normally wash hands with soap under runnning water before meals? Always………..………….1 Most of the time..…..2 Sometimes……………..3 Never……………………..4 D6: WHAT IS THE OWNERSHIP STATUS OF DWELLING Built on squatter land……………………1 Given by relative/Rent-free…………..2 Provided by government/Rent-free…3 Rented………………………….…………….……4 Owned………………………………………….…5 Other (specify)………………………….…..6 D14: Do you normally wash hands with soap under runinng water after visiting the toilet? Always………..………….1 Most of the time..…..2 Sometimes……………..3 Never……………………..4 D7: What is the main source of drinking water in your household Indoor Plumbing/Inside Standpipe.....1 Water vendor……………………....……...…..2 Neighbouring house……….........…......…3 Public Standpipe……………….......……….4 Covered Well…………………….........…......5 Uncovered well......................................6 River/Lake/Spring/Pond……......………..7 Rain water……………………………......…….8 Other (Specify)……………………...............9 D8:How long does it take to fetch water to & fro? HOURS…………………………………………. MINUTES……………………………………… SECTION D: HOUSEHOLD FEATURES (RECORD OBSERVATION WHERE APPLICABLE) University of Ghana http://ugspace.ug.edu.gh 139 NO. E1: Which member(s ) of your Household suffered from Cholera las t year? Suffered………...1 Did not suffer....2 Did not suffer but a caregiver…...3 If "Did not suffer (2)" END for that HH member Ski t E20 if E2: Where did the person "first" seek Treatment from? Sel f- treatment…....1 Hospi ta l……….. 2 Tradi toina l Methods…….... 3 Other (Speci fy)……....4 E3: Did the person receive any fi rs t a id or medication before going to the faci l i ty? Yes………….. 1 No…………...2 If No skip to E5 E4: How much did you spend on the fi rs t a id? Amount in GH cedis E5: Did you pay (out of pocket cash) for consul tation, drugs , treatment and laboratory test at faci l i ty Yes…..…………1 No……………….2 If No skip to E7 E6: How much did you pay for these services in tota l? Amount in GH cedis If answered skip to E8 E7: Why did you not pay for service? Used Heal th Insurance……1 FreeTreatmen t ……..……...….2 Could not afford………...3 Others (Speci fy)……..4 E8: Did you make any under- the-table payment to receive treatment ? Yes…..….1 No………..2 If No skip to E10 E9: How much did you pay for under- the-table to receive treatment ? Amount in GH cedis E10: How much did you spend on feeding for the s ick patient? Amount in GH cedis E11: How many nights was HH member on admiss ion ? Number of nights the patient spent admitted E12: How many work/school days did s ick person(s ) miss as a resul t of cholera? SECTION E: HEALTHCARE UTILIZATION & COSTS (For HH members that contracted cholera) University of Ghana http://ugspace.ug.edu.gh 140 NO. E13: How much did the admissio n cost you? Amount in GH cedis E14: By what means did you get to the facil ity? Taxi….…..….. 1 Trotro..…..... 2 Walked….….. 3 Private Car…4 Ambulance …5 Other Specify.……6 E15: How long did it take you to get to the health facil ity? Time in HR:MINS E16: How much did you pay for transpor tation? (Indicate separate amounts for multiple patients) Amount in GH cedis E17: Did any member of your HH die as a result of Cholera? Yes……...1 No….……2 If No to E19 E18: How much did you pay for the burial in Total? Amount in GH cedis E19: How did your HH pay for the all expenses? Paid cash….…1 Sold Assets…..2 Borrowed money…….….3 Took a loan……….....4 Others (Specify)……..5 E20: Did caregiver(s) do multiple visits to facil ity to cater for the patient? (Either moving between the home and facility more than once within a day or over a number of days) Yes…………...….1 No………………..2 If No skip to E22 E21: How many times did you visit patient before he/she was discharge d? E22: How many days did you stay away from your normal work/scho ol activities to take care of the patient? E23: How much did caregiver( s) spend in total on transporta tion for these visits? E24: How much time in total did caregiver(s) spend at the facil ity in multiple visits to the facil ity? Time in HR:MINS E25: What is the main consequence of costs incurred as a result of cholera to the HH and the caregiver? (Accept Multiple Responses) Cash Shortage……….....1 Less savings……………...2 Loss of investment…….…....3 Less consumption…….....4 Couldn't pay school fees.......5 Kids dropped out of sch.......6 HH members had to work more for extra cash...........7 Lost Job............................8 No Effect..........................9 Other (Specify).................10 : : : : : : SECTION E: HEALTHCARE UTILIZATION & COSTS (For HH members that contracted cholera) Caregivers Only University of Ghana http://ugspace.ug.edu.gh