http://ugspace.ug.edu.gh/ UNIVERSITY OF GHANA HEALTH SEEKING BEHAVIOUR IN THE AKUAPEM-NORTH MUNICIPALITY OF GHANA BY AMPONSAH, KOJO COMMEY NTSEFUL ANNANG (10442951) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF MPHIL GEOGRAPHY AND RESOURCE DEVELOPMENT DEGREE. JULY, 2016. 1 http://ugspace.ug.edu.gh/ DECLARATION This is to certify that this thesis is the result of research undertaken by Kojo Amponsah Commey Ntseful Annang towards the award of a Master of Philosophy degree in Geography and Resource Development in the Department of Geography and Resource Development, University of Ghana. ......................................... ………………………. (Signature) (Date) Kojo Amponsah Commey Ntseful Annang (Candidate’s Name) …………………………….. ……………………... Dr. Joseph Kofi Teye (Date) (Principal Supervisor) ……………………………… ……………………... Dr. Charlotte Wrigley-Asante (Date) (Co-Supervisor) i http://ugspace.ug.edu.gh/ DEDICATION This work is dedicated, first of all, to the Almighty God for His mercies, grace, guidance, favour and gift of life. Secondly, to my late father who should have lived to see me come this far. Thirdly, to my family especially my mother and wife, Mrs. Diana Boateng Amponsah Commey, whose immense contributions, consistent love and support have given me the strength to continue surging forward even in times of hardship and difficulty. Finally, to all those who have been part of my success life. ii http://ugspace.ug.edu.gh/ ACKNOWLEDGEMENT I first acknowledge the Almighty God for His grace that has brought me this far. In fact, He has been faithful to me. To Him be the glory forever and ever. I wish to express my heart-felt gratitude to my supervisors, Dr. Joseph Kofi Teye and Dr. Charlotte Wrigley-Asante. In fact, their invaluable suggestions, directions and guidance have made this research work a success. To them, I say a very big thank you. I am also grateful to the Head of Department, teaching and non-teaching staff, Geography and Resource Development for the opportunity to be part of their family and the necessary and sufficient conditions provided for the success of this piece. I extend my gratitude to Madam Alice Johnfiah, Aretha, Monica, Sarah, Michael, Stephanie, Stephen, Patrick, Thelma, Delorise, Seth, Mark, Lydia and Diana. Your prayers, love, support and inspiration have made me what I am today and contributed to the reality of this work. God richly bless you all. A friend in need, they say, is a friend indeed. The support given me by Thomas Padi Appai and Noah Ogyiri Adu cannot be forgotten. You were there for me when I least expected. Colleagues, I really appreciate what you did for me. God bless you all. I cannot forget the respondents who, in spite of their tight schedules, made themselves available to be interviewed. Without them this research work would not have been possible. Their efforts are greatly acknowledged. Finally, my special gratitude goes to Emmanuel, Frank, Rebecca, and all others through whom God blessed me. iii http://ugspace.ug.edu.gh/ TABLE OF CONTENTS DECLARATION i DEDICATION ii ACKNOWLEDGEMENT iii TABLE OF CONTENTS iv LIST OF TABLES ix LIST OF FIGURES xi LIST OF ABBREVIATIONS xii ABSTRACT xiii CHAPTER ONE INTRODUCTION 1.1 Background of the study 1 1.2 Statement of the Problem 4 1.3 Research Questions 5 1.4 Objective of the study 6 1.5 Hypotheses 6 1.6 Significance of the study 7 1.7 Limitations of the study 8 1.8 Definition of concepts 10 1.8 Organisation of the study 10 CHAPTER TWO REVIEW OF LITERATURE AND THEORETICAL FRAMEWORK 2.0 Introduction 12 iv http://ugspace.ug.edu.gh/ 2.1 Theoretical models to explain health seeking behaviour 12 2.1.1 The health care utilisation model 14 2.1.2 Kroeger’s model of health service utilisation 15 2.2 Factors influencing Individual’s health seeking behaviour 17 2.2.1 Age 17 2.2.2 Sex 19 2.2.3 Education 21 2.2.4 Employment status and occupation 22 2.2.5 Income 24 2.2.6 Health insurance 26 2.2.7 Severity of illness 27 2.2.8 Availability of health providers 30 2.2.9 Distance to health providers 32 2.2.10 Quality of service 34 2.2.11 Health service cost and financing 36 2.2.12 Place of residence and health service provider options 39 2.3 Health facility attended when ill 41 2.4 The nature of the health system in Ghana 43 2.5 The nature of the health system in Akuapem-North Municipality 45 2.6 Conceptual framework for the study 46 2.7 Summary 49 CHAPTER THREE PROFILE OF THE STUDY AREA AND METHODOLOGY v http://ugspace.ug.edu.gh/ 3.1 PROFILE OF THE STUDY AREA 50 3.1.0 Introduction 50 3.1.1 Physical characteristics 50 3.1.1.1 Location and size 50 3.1.1.2 Climate 51 3.1.1.3 Vegetation 51 3.1.1.4 Geology and soil 52 3.1.1.5 Topography and drainage 52 3.1.2.0 Socio-economic characteristics 54 3.1.2.1 Population characteristics 54 3.1.2.1.1 Age and sex distribution 54 3.1.2.1.2 Dependency ratio 55 3.1.2.2 Settlement 55 3.1.2.3 Education 56 3.1.2.4 Health 56 3.1.2.5 Water and sanitation 57 3.1.2.5.1 Water 57 3.1.2.5.2 Sanitation situation 58 3.1.2.6 Economic activities 58 3.2 METHODOLOGY 60 3.2.0 Introduction 60 3.2.1 Selection of study sites 60 3.2.2 Research design 61 vi http://ugspace.ug.edu.gh/ 3.2.3 Primary data sources 66 3.2.3.1 Individual survey 66 3.2.3.2 In-depth interviews 69 3.2.4 Secondary data sources 71 3.2.5 Data analysis and presentation 71 3.2.6 Challenges encountered in the fieldwork 71 3.3 Summary 72 CHAPTER FOUR TYPE OF ILLNESS AND INDIVIDUALS’ HEALTH SEEKING BEHAVIOUR 4.0 Introduction 74 4.1 Socio-demographic characteristics of respondents 75 4.2 Major type of illness and source of treatment 81 4.3 Severity of illness and when reported for treatment 85 4.4 Availability of health service providers 95 4.5 Individuals’ health seeking behaviour in the municipality 97 4.6 Summary 112 CHAPTER FIVE UTILISATION OF HEALTH SERVICES IN THE AKUAPEM NORTH MUNICIPALITY 5.0 Introduction 116 5.1 Most preferred health service provider of respondents 116 5.2 Distance and means of transport to health service providers 120 vii http://ugspace.ug.edu.gh/ 5.3 Determinants of individuals’ health seeking behaviour 125 5.4 Barriers to the choice and use of health services 139 5.5 Health service cost and financing 147 5.6 Summary 159 CHAPTER SIX SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 6.0 Introduction 162 6.1 Summary of key findings 162 6.2 Conclusions 168 6.3 Recommendations 169 REFERENCES 173 APPENDIX A: INDIVIDUAL QUESTIONNAIRE 185 APPENDIX B: INTERVIEW GUIDE FOR HEADS OF HEALTH 205 PROVIDERS APPENDIX C: INDIVIDUAL INTERVIEW GUIDE 207 viii http://ugspace.ug.edu.gh/ LIST OF TABLES 3.1.1 Age and sex distribution 55 3.1.2 Contributions of the sectors to employment 59 3.2.1 Research design 65 4.1.1 Distribution of respondents by age, sex, level of education and religion 78 4.1.2 Distribution of respondents by marital status, occupation and income 80 4.2.1 Respondents’ major type of illness and source of treatment 82 4.2.2 Respondents’ major type of illness by source of treatment 84 4.3.1 Severity of illness and place of residence 86 4.3.2 Respondents’ severity of illness and when reported for treatment 88 4.3.3 Respondents’ major type of illness by severity of illness 91 4.3.4 Major type of illness by when reported for treatment 92 4.3.5 Health facility attended when ill by major type of illness 94 4.4.1 Available health providers respondents sought treatment from 97 4.5.1 Health facility attended when ill 99 4.5.2 Respondents’ specific first action taken when ill 101 4.5.3 Health facility attended when ill by socio-demographic characteristics 105 4.5.4 Multinomial logistic regression coefficients for independent variables of 105 predictors of the health facility respondents attended when ill 5.1.1 Health facility usually accessed by respondents by place of residence 117 5.1.2 Respondents’ most preferred health provider by place of residence 119 5.2.1 Respondents’ means of transport, payment for transport, travel time and 124 waiting time at health facilities ix http://ugspace.ug.edu.gh/ 5.3.1 Determinants of individuals’ health seeking behaviour 127 5.3.2 Multinomial logistic regression coefficients for independent variables of 138 predictors of the health facility respondents attended when ill 5.4.1 Difficulty in accessing health services 141 5.4.2 Respondents’ major barriers to health service use 142 5.5.1 Cost of and payment for health service use 150 5.5.2 Cost of health service use by health facility attended when ill 151 5.5.3 Respondents’ place of residence and the health facility attended when ill by 153 mode of payment for health service use 5.5.4 Socio-demographic variables by payment for health service use 157 x http://ugspace.ug.edu.gh/ LIST OF FIGURES 2.6.1 Conceptual framework for the study 48 3.1.1 Map of the study area 53 5.2.1 Distance to nearest health providers 121 xi http://ugspace.ug.edu.gh/ LIST OF ABBREVIATIONS ANC Antenatal Care ANDA Akuapem-North District Assembly ANMA Akuapem-North Municipal Assembly ANMHD Akuapem-North Municipal Health Directorate CHN Community Health Nurse CHPS Community Health Planning Services CSHS Community Senior High School CWSA Community Water and Sanitation Agency FMPs Formal Medical Providers GHS Ghana Health Services GSS Ghana Statistical Services IMPs Informal Medical Providers JHS Junior High School MOH Ministry of Health NHIA National Health Insurance Authority NHIS National Health Insurance Scheme PHN Primary Health Nurse SAP Structural Adjustment Programme SHS Senior High School WHA World Health Authority xii http://ugspace.ug.edu.gh/ ABSTRACT People’s response to illness symptoms whenever they are perceived is dependent on them and influenced by a plethora of factors which makes them behave differently towards treatment seeking. With this, the study examined the health seeking behaviour in the Akuapem-North Municipality. Drawing on Kroeger’s (1983) health service utilisation model, it specifically examined individuals’ socio-demographic characteristics and other factors that influence the choice and use of particular health facilities. It also examined the barriers to health service use and the modes of health service financing in the study area. The study was a cross-sectional one and adopted the mixed method approach. A four-step multi-stage cluster sampling was used to select 205 respondents for the individual survey while respondents for the qualitative facet of the study were randomly selected. The study reports of general body pains/headache as the predominant illness in the municipality. It showed higher utilisation of informal than formal medical facilities. Sex, educational level, income, health insurance, severity of illness and distance to health providers are some of the factors that significantly influence individuals’ health seeking behaviour in the Akuapem-North Municipality. However, availability, proximity, longer waiting times and attitude of staff of health providers are some of the restrictive barriers to utilising formal health facilities whereas NHIS and out-of-pocket payment are the main modes of health service financing in the municipality. Conclusively, the study confirms the positive association between perceived morbidity, illness response and access to and utilisation of health services emphasised by Kroeger (1983). The study recommends an integration of both the formal and informal medical systems in order to help improve and assist health service delivery in the country and especially in the municipality. xiii http://ugspace.ug.edu.gh/ CHAPTER ONE 1.0 INTRODUCTION 1.1 Background of the study Health seeking behaviour, which has to do with peoples’ response to perceived illness symptoms, differ from person to person (Anwar, Green & Norris, 2012; Roy, Torrez & Dale, 2004) and is influenced by a multiplicity of variables categorized into geographical, economic, cultural, social and institutional factors (Shaikh & Hatcher, 2005; MacKian, 2003). These factors facilitate the use of health services by individuals and groups and influence them to behave differently in relation to their health (MacKian, 2003). As some go for treatment at formal medical providers, others seek care at informal medical sources irrespective of their socio-demographic and economic background (Aboagye & Agyemang, 2013; Ghosh et al., 2013; MacKian, 2003). Treatment seeking from the informal medical system is reported even in the highly advanced countries among the working class and people of high socio-economic status (Sandman, Simantov & An, 2000) and also among minority groups and children (Roy et al., 2004) albeit the greater proportion of the population engage with the formal medical system for care whenever they perceive illness symptoms (NHRA, 2006; Roy et al., 2004; Sandman et al., 2000). Studies done in developing countries report of high treatment seeking at informal medical facilities where it is more profound and devastating in the rural areas (Abubakar et al., 2013; Ghosh et al., 2013; Robyn et al., 2012). Higher proportions of developing countries’ populations have been found to ‘self-treat’ as first and major action taken whenever sick 1 http://ugspace.ug.edu.gh/ or perceived symptoms of illness (Robyn et al., 2012; Mashreky et al., 2010; Yanagisawa, Mey & Wakai, 2004). For instance, majority of the study population in Bangladesh were found to take their burned children to unqualified service providers such as medicine shopkeepers as first action taken (Mashreky et al., 2010). Yanagisawa et al. (2004), in a similar study in Cambodia, found more of the better-off than their poor counterparts to have resorted to home remedy applications and self-medication, likewise Georgians (Gotsadze et al., 2005) and Bangladeshis (Ahmed et al., 2005) where it is more detrimental in rural areas irrespective of the individuals’ age. In West Bengal, Ghosh et al. (2013) report of gender inequality in terms of child care seeking as more of male than female children were taken to a health provider for treatment. This, they attributed to a myriad of factors. Even in places where people have access to formal medical providers, treatment seeking at informal medical sources still dominates (Mazumda et al., 2009). In addition, similar trends of health seeking have been found in other countries and parts of same countries where despite the large numbers that sought some form of treatment, it was mainly done at informal medical facilities (Haque et al., 2012; Rahman et al., 2012; Khabiri et al., 2011; Mushtaq et al., 2011; Siddiqui et al., 2011). Health seeking at informal medical providers predominate in Africa especially among rural dwellers where factors such as unavailability of formal medical facilities coupled with poor road networks, proximity, traditional beliefs and practices and low socio-economic status compel patients to go for treatment at such sources (Fomundam et al., 2012; Oluwatuyi, 2010; Atuyambe et al., 2009). However, reports also indicate higher health seeking at formal medical providers especially in the urban centres on the continent (van der Hoeven et al., 2012; Hjelm & Atwine, 2011) though large proportions of urbanites are found to 2 http://ugspace.ug.edu.gh/ self-treat (Rai et al., 2012; Awusabo-Asare & Anarfi, 1997). This results in delays to care seeking from the formal health system (Afolabi et al., 2013) and further worsens the health conditions of the people, which sometimes costs them their lives of diseases that could have been prevented should contacts with the biomedical services have been made earlier. Ghana, ever since independence in 1957, has made several strides to increase access to health services at formal health facilities on behalf of her populace. This has seen the promulgation and institution of the National Health Insurance Scheme (NHIS) in the year 2004 with the main aim of ensuring “equitable and universal access for all residents of Ghana to an acceptable package of quality health service” (Agyapong & Adjei, 2008). This should have sought to the health needs of the people especially the rural poor, since it would have curtailed the problem of ‘point-of-service-use’ (Agyapong & Adjei, 2008) but reports indicate delays in reporting times at formal medical facilities and increases in treatment seeking at informal health facilities especially in the rural areas even after health insurance (Annan et al., 2013; Laar et al., 2013; Blanchet, Fink & Osei-Akoto, 2012). This is attributed to the unavailability of formal medical facilities coupled with poor transportation systems and low socio-economic status especially in the rural areas (Aboagye & Agyemang, 2013; Laar et al., 2013; Danso-Appiah et al, 2010). Urbanites, because of the availability of formal medical facilities, mainly seek health at such formal medical providers though substantial proportions go for care at informal medical providers especially from drug stores, traditional healers and apply home-made remedies (Aboagye & Agyemang, 2013; Annan et al., 2013; Blanchet et al., 2012). It is against this backdrop that the research into the health seeking behaviour of people especially in the Akuapem-North Municipality of Ghana has been undertaken. 3 http://ugspace.ug.edu.gh/ 1.2 Statement of the problem There is quite an extensive literature regarding the concept of health seeking behaviour. While some have examined the concept in rural communities (Iyalomhe & Iyalomhe, 2012; Oluwatuyi, 2010; Abdulraheem & Parakoyi, 2009), others have done so in urban areas (Afolabi et al, 2013; Ajiboye & Adebayor, 2012; Russell, 2008), but literature is limited in analysing the concept of health seeking behaviour in both rural and urban communities in a single research especially in Ghana. Also, previous studies have looked at how people with specific diseases and/or health problems such as malaria (Laar et al., 2013), tuberculosis (Annan et al., 2013), HIV/AIDS and STIs (Adanu et al., 2008; Awusabo-Asare & Anarfi, 1997), schistosomiasis-related symptoms (Danso-Appiah et al., 2010), and pregnancy threats (Dako-Gyeke et al., 2013) engaged with health services. Nevertheless, taking a non-disease-specific outlook is needful, since in some cases even the cultural dynamics of the people influence their health seeking behaviour in diverse ways (Abubakar et al., 2013). The Akuapem-North Municipality is one of the numerous communities in the country with people of different socio-demographic, economic and cultural background with the Twi- speaking Akans and Guans being the dominant ethnic groups (GSS, 2012). Geographical access to formal health providers in the municipality shows some form of inequality with most of the formal health facilities located in the urban areas leaving the rural communities with only few, which are mostly health centres and Community Health Planning Services (CHPS) compounds. These normally provide ‘first aid’ care to patients and cannot accommodate substantive patients for even hours. Due to this, most people especially those in the rural areas delay in engaging with formal medical providers as they adopt a “wait- 4 http://ugspace.ug.edu.gh/ and-see” strategy as a way of buying time (ANDA, 2006) to find out whether their sicknesses would vanish or not. It is evident and vivid from the above that individuals differ in their choice of treatment sources. This brings the question of what might have triggered this trend of health seeking. Though much has been done regarding the concept of health seeking behaviour in the country, little has been done on the subject, especially in the Akuapem-North Municipality hence the need for this study to fill such a gap. 1.3 Research questions The research questions were fixated on the individuals’ socio-demographic and socio- economic background information such as age, sex, education, occupation, income, health insurance and illness severity and how they influenced their decisions to seek health from a particular medical system. The questions further examined the respondents’ view on other factors as kinds of health service providers available, distance to health providers, attitude of health providers’ staff and the perceived quality of service rendered by various health providers within the communities. It also ascertained the various barriers to and the mode of financing health service use in the municipality. Thus, given the dimension of the problem, the following were the research questions that emerged to serve as the basis for the objectives:  What is the preferred health seeking system of people in the study area?  What relationship exist between individuals’ socio-demographic characteristics and their health seeking behaviour in the municipality? 5 http://ugspace.ug.edu.gh/  What factors determine peoples’ choice and use of particular health facilities in the study area?  What obstacles bar individuals from utilising health services and by what means did patients financed their health service use in the municipality? 1.4 Study objectives The aim of the research was to explore the health seeking behaviour of both urban and rural dwellers in the Akuapem-North Municipality. Specifically the study sought to:  Identify the preferred health seeking system of rural and urban dwellers in the study area.  Examine the relationship between individuals’ socio-demographic characteristics and their health seeking behaviour.  Investigate the factors that influenced individuals’ choice and use of particular health service providers in the municipality.  Assess the barriers to and mode of payment for health service use in the municipality. 1.5 Hypotheses The study was guided by the following hypotheses: H0: There is no significant difference in the health facility rural and urban dwellers attended whenever illness symptoms were perceived. H1: There is a significant difference in the health facility rural and urban dwellers attended whenever illness symptoms were perceived. 6 http://ugspace.ug.edu.gh/ H0: The availability of a health provider will not significantly influence their utilisation in both the rural and urban communities. H1: The availability of a health provider will significantly influence their utilisation in both the rural and urban communities. H0: There is no significant relationship between illness severity and the type of health facility rural and urban dwellers sought treatment from whenever illness symptoms were perceived. H1: There is a significant relationship between illness severity and the type of health facility rural and urban dwellers sought treatment from whenever illness symptoms were perceived. 1.6 Significance of the study Health service providers have, in most cases, been baffled by people’s deceptively absurd choices regarding their search for treatment. In fact, health seeking behaviour, which is the action people take when dealing with illness not only affect their health situations directly, but also is influenced by multiple factors. With this, information on peoples’ health seeking behaviour is an expression of social, economic, geographical, institutional and cultural mixed standpoints. Therefore, a better understanding of health seeking behaviour forms the basis to build health policies and programmes that will meet the health requirements of the people. Thus, probing into the health seeking behaviour among the people of Akuapem- North Municipality will help authorities, both at the national and local levels, to design policies; monitor and evaluate healthcare developments, and support the achievement of national healthcare objectives. 7 http://ugspace.ug.edu.gh/ Socio-economic characteristics of individuals and populations have been found to highly influence their health seeking behaviour, but with the advent of the National Health Insurance Scheme (NHIS) programme even marginalised individuals and communities have been granted economic access to health services. It, therefore, becomes expedient to investigate into people’s health seeking behaviour in the country and especially in the study area to ascertain, if it is so, other possible factors that may affect people’s decision to seek health from a particular medical system in order to inform policies. The study of people’s health seeking behaviour will enable policy makers at both the national and local levels to arrange different health programmes for different segments of the population, and improve efficacy and efficiency of the organisation and delivery of health service in the country. Again, education is seen as one of the best tools for all and sundry. Thus, giving people information about the healthcare system and directing them to use the formal health system more efficiently will lead to healthy individuals, communities and populations at large, thereby, promote economic growth and development, both at the local and national levels respectively. Lastly, the study will serve as a platform for further research as regards the availability of the numerous forms of treatment channels and the factors that inhibit people’s decisions to engage with such medical systems in the study area and the nation as a whole. 1.7 Limitations of the study The problems that hindered the progress of this survey were inadequate funds, the quantification of some variables and the time allotted for the study. Funds were needed to print questionnaires, transport and remunerate the field assistants, acquire the needed 8 http://ugspace.ug.edu.gh/ research tools such as audio recorders and other essential materials needed to facilitate the collection of data, and document the research findings. This made the cost of data collection more expensive. Also, some of the variables of the study such as respondents’ perceptions of health personnel attitude and quality services rendered by health providers were difficult to quantify and thus affected the quantitative representation of how they influenced health seeking behaviour in the study communities. In such cases, a qualitative approach was employed for the analysis to ensure precision and avoid bias. Again, the time frame for the survey is also challenging. There must be the need for ample time for reconnaissance survey of the study area, proposal writing, literature review, data collection, data analysis and for writing of chapters. Other unforeseen circumstances such as the dispersed nature of the rural community made access to those areas extremely difficult. However, appropriate measures were put in place to overcome the anticipated limitations of the survey. Adequate funds were raised from family members to supplement the financial support offered by the University for the purpose of this research. Also, a proper schedule of work was designed to contain all the activities involved in the research, so as to work in accordance with time in order to complete the work within the stipulated time scheduled by the University. More so, respondents, especially in the rural area, were educated about the essence of the research and its intended benefits for the country in order to win their maximum co-operation for data collection. Where respondents were illiterates, the interview method of data collection was used to avoid call-backs and the interpretations were done by both the researcher and the field assistants. 9 http://ugspace.ug.edu.gh/ 1.8 Definition of concepts Health Seeking Behaviour refers to peoples’ response to illness (MacKian, 2003). For the purpose of this study, health seeking behaviour refers to the health service provider/facility patients attended when symptoms of illness were perceived (Anwar et al., 2012). Formal medical/health providers include hospitals, clinics, health centres, CHPS compounds and/or health posts, be it public or private. Hence engaging with any of these was seen as treatment seeking at a formal medical/health provider or facility. Informal medical/health providers encompass all forms of traditional healers, home remedies, drug/chemical shops, drug peddlers, and pharmacy shops (Ahmed, 2005). Hence treatment at any of these (which include self-treatment and/or self-medication) denoted treatment seeking at informal medical/health providers. 1.9 Organisation of the study This research work comprises six chapters. The first chapter deals with the general introduction of the study. It comprises the background of the study, statement of the problem, research questions, objectives of the study, hypotheses, significance of the study, limitation of the study, definition of concepts, and the organization of the study. This helped to determine and define the boundary of the study. Chapter two presents a review of related literature and theoretical frameworks. Here, the views of other researchers that are related to health seeking behaviour, theories and models used in understanding health seeking behaviour as a concept and determining factors that influence people to behave differently towards treatment seeking whenever illness 10 http://ugspace.ug.edu.gh/ symptoms are perceived have been discussed. Also, literature regarding the health system in Ghana and the mode of health service use financing were reviewed to help put the current study into its right perspective. Chapter three presents the profile of the study area – Akuapem-North Municipality – its location, physical characteristics including its drainage and topography, geology and soil, vegetation, climate, population characteristics, and the socio-economic situations of the area where the educational situation, health condition, water and sanitation situations, and finally, the economic sectors and activities of the populace were discussed. The chapter also discussed the methodology used in conducting the study. Chapter four and five dealt with the presentation and analysis of the data that was garnered from the field. This has been discussed in line with already existing literature showing confirmatory and contradictory findings and significant relationships established between the dependent and some independent variables. Finally, the study ends with chapter six, which looked at the summary of findings, conclusions and appropriate recommendations to the study. 11 http://ugspace.ug.edu.gh/ CHAPTER TWO REVIEW OF LITERATURE AND THEORETICAL FRAMEWORK 2.0 Introduction This chapter aims at placing the study in a scholarly context by reviewing the main contributions made by researchers on the concept of health seeking behaviour and the main factors that influence individuals’ decision to engage with a particular medical system when sick or perceived symptoms of illness. The literature review comprises three main sections. The first section looks at conceptualizing health seeking behaviour for the purpose of the study and further reviews literature on possible theoretical and/or conceptual frameworks that have been used in the study and where necessary, identify the strengths and problems associated with them leading to their use or non-use in this current study. The second section focuses on the factors that influence individuals’ decision to seek treatment from a particular health provider. The third and final section looks at the medical system in Ghana and in the Akuapem-North Municipality in brief. 2.1 Theoretical models to explain health seeking behaviour The choice of a particular health provider and the subsequent utilisation of health services is a comprehensive course of action (Ghosh et al., 2013; Mushtaq et al., 2011; Danso- Appiah et al., 2010), especially in the context of a developing nation like Ghana. Whether individuals and populations seek treatment from either the formal or informal medical system have been attributed to a plethora of factors including economic, cultural, social, geographical and institutional (Shaikh & Hatcher, 2005; MacKian, 2003). The focus here 12 http://ugspace.ug.edu.gh/ is to review related frameworks on health seeking behaviour in order to understand individuals’ choice of health service providers when sick or perceived symptoms of illness. A number of scholars from different fields of study including Medical Geography have explored the reasons behind the choice of a particular health service provider by individuals (MacKian et al., 2004; Hausmann-Muela et al., 2003). This has led to the formulation of different frameworks and/or models to explain the concept of health seeking behaviour as a tool for describing how individuals engage with particular health providers (Shaikh, Hatcher & Haran, 2008; MacKian, 2003) and serves as a platform for identifying loopholes in health service delivery for appropriate health service measures to be meted-out (Ahmed, 2005; MacKian et al., 2003). Researchers, more often, adapt these models to meet the requirements of their fields of study, or to a higher extent, merge a number of them with the aim of augmenting the range of possible key factors that may influence individuals’ health seeking decisions (Hausmann-Muela et al., 2003). With respect to health seeking behaviour of individuals, more emphasis is placed on the socio-economic conditions of the individual and the prevailing health system characteristics (Shaikh et al., 2008; Shaikh & Hatcher, 2005). As a result, the hub of health seeking behaviour research from a behavioural point of view is on identifying the most appropriate factors that may influence the individual’s choice of the health service provider (Ahmed, 2005; Hausmann-Muela et al., 2003). With this, numerous frameworks have been proposed to explain the health seeking behaviour of an individual, but from the behavioural viewpoint, the Andersen & Newman’s model (1973) and the Kroeger’s model (1983) are the major ones used on which literature has been reviewed regarding this study. 13 http://ugspace.ug.edu.gh/ 2.1.1 The health care utilisation model Considered as one of the conceptual models aimed at understanding the factors that lead to the use of health services, Andersen and Newman’s (1973) framework categorises the factors that influence an individual’s choice of a particular medical system when symptoms of illness are perceived into three: “predisposing”, “enabling” and “need” factors respectively (Andersen & Newman, 1973). Predisposing factors, which include the socio- economic and demographic characteristics of an individual tries to explain the individual’s preference of a health provider prior to the occurrence of an illness. Enabling factors, which include the availability of services, health insurance and income explains how an individual’s need of health services is fulfilled. Need factors focus on the basis for utilising health services by an individual and consist mainly of the perceived severity of the illness (Andersen & Newman, 1973). These factors, according to Andersen & Newman (1973), will influence an individual’s health or treatment seeking behaviour by coming out with actions which may include home remedy applications, over-the-counter purchases of drugs, visits to traditional healers, drug peddlers, faith healers, hospitals, clinics and health centres (Andersen & Newman, 1973; Andersen, 1995). The model, despite its applicability in various fields and study areas, has been criticized for not paying enough attention to culture and social interaction. This has been empirically established in both the developed (Roy et al., 2004) and developing nations (Annan et al., 2013; Ghosh et al., 2013) where due to the unfriendly attitude of staff, treatment seeking at formal medical facilities is reduced. Andersen (1995) however argues they are included in the predisposing characteristics of their model. 14 http://ugspace.ug.edu.gh/ Another criticism is the over stressing of ‘need’ factors as a major determining factor in health service utilisation. This is because the model initially was developed for developed nations where enabling and inhibitive factors do not influence patients’ health service access and utilisation. However, in the context of developing nations like Ghana where health service factors like poor attitude of personnel, low quality of service rendered and inhibitive factors such as unavailability of transport facilities coupled with poor road networks impact health service utilisation than need (Annan et al., 2013; Russel, 2008). However, Andersen (1995) justified this by acknowledging need as a social structure and represented it as “perceived need” and “evaluated need.” Further criticism has to do with the model’s inability to consider the health system as a possible factor capable of determining individual’s access to and utilisation of services. The unavailability of health facilities in especially the rural areas of developing countries pose a challenge to effective and appropriate health care seeking (Aboagye & Agyemang, 2013; Anwar et al., 2012; Mazumdar et al., 2009) but this is not the case in the developed nations where the model was initially designed to be used (Hausmann-Muela et al., 2003). 2.1.2 Kroeger’s model of health service utilisation Kroeger (1983), after a detailed review of literature modified Andersen & Newman’s (1973) model of health care utilisation to include health service factors in addition to ‘predisposing’ and ‘enabling’ factors to propose a theoretical framework suitable for studying health seeking and healthcare seeking behaviours in both developed and developing countries (Shaikh et al., 2008). The framework consist of individuals’ characteristics, illness characteristics, individuals’ perception and health service 15 http://ugspace.ug.edu.gh/ characteristics as the key factors influencing health service use (Hausmann-Muela et al., 2003; Kroeger, 1983). Individuals’ characteristics is similar to predisposing factors proposed by Andersen and Newman which include “demographic and society-related” variables. Illness characteristics consist of the severity and nature of illness. Individuals’ perception consists of impressions about expected benefits of treatment, views about illness type and opinions about aetiology of diseases. Health service characteristics is similar to enabling factors and has to do with service accessibility, acceptability, appeal; quality, communication and cost of use (Hausmann-Muela et al., 2003; Kroeger, 1983). The interaction of these factors, according to Kroeger, guides the selection of health service resources by an individual which he posited to include modern treatment, traditional healers, drug sellers and self/no treatment (Hausmann-Muela et al., 2003; Kroeger, 1983). Though Kroeger’s model omits the “need factors” postulated by Andersen and Newman which are instrumental in understanding the health seeking behaviour of an individual, it has the advantage of having a variety of factors organised in categories that make treatment seeking at a health service provider possible or impossible. The model allows for relationships among the varying factors with good certainty to be established, though it does not illustrate how these different factors affect curative selection (Ahmed, 2005; Hausmann-Muela et al., 2003). However, a major short-coming of the model is its failure to acknowledge the impact of health insurance and spatial variation regarding the choice and use of specific health service providers (Iyalomhe & Iyalomhe, 2012; van der Hoeven et al., 2012) due to the disparity in terms of formal health facilities in especially a developing country like Ghana. Health insurance and the place of residence of individuals have been identified to significantly 16 http://ugspace.ug.edu.gh/ determine the choice and use of health services at a particular health provider (El Kahi et al., 2012; Rai et al., 2012; El-Kak et al., 2009). 2.2 Factors influencing individuals’ health seeking behaviour The models have been applied on individuals suffering from specific ailments to treatment seeking in general without focusing on the nature and/or aetiology of ailments (Hausmann- Muela et al., 2003; MacKian, 2003). The present study is concerned with health seeking behaviour in general without necessarily considering a particular ailment. This review of literature looks at the application of the models for general health provider choices and looks at the factors that may influence the individual’s choice of a health provider for treatment seeking in addition to other factors that have been added to the already existing ones through modifications made by researchers due to their use of the models in different environments. With this, confirming and contrasting results have been documented at different times and in varying contexts hence, a review of these factors assisted in building the required framework for this study. 2.2.1 Age Age is one of the demographic variables mentioned in Andersen & Newman’s (1973) and Kroeger’s (1983) models to influence individual’s decision to engage with a particular medical system whenever illness symptoms are perceived. Based on that, majority of past studies (Haque et al., 2012; Rahman et al., 2012; Ahmed et al., 2005) conducted in both the developed and developing nations adopting the frameworks have empirically tested and reported the significance of age as a determinant factor for individual’s choice of health providers though others (Ghosh et al., 2013; Mushtaq et al., 2011) have not found any 17 http://ugspace.ug.edu.gh/ significant associations. Studies across the globe have reported of treatment seeking from both the formal and informal medical sources for the young and elderly alike. In the more advanced countries such as the United States of America, Sandman et al. (2000) found adults of the working age group (18-64 years) to seek care at hospitals than the elderly (65 years and above) due to the former’s ability to pay for services than the latter. Ahmed et al. (2005) did identify self-care/self-treatment as the most common treatment choice for both the young and the elderly in rural Bangladesh as Biswas et al. (2006) researching on the elderly in rural Bangladesh found them to mainly self-treat or visit traditional healers due to their proximity and flexibility in treatment payments and only visit the doctor when illness gets more severe. Also in rural Bangladesh, more of older women (greater than 35 years) than their younger counterparts (less than 20 years) sought maternal care from either a doctor or a nurse (Chakraborty et al., 2003). In Cambodia, the 20-39 years age group was found mainly to resort to home remedies and self-medication than individuals of all the other age groups (Yanagisawa et al., 2004). In the same way, younger-aged women (15-29 years) in Beirut were found to seek reproductive healthcare services the more than those in the 30-59 years age category (El-Kak et al., 2009). In Africa, Taffa & Chepngeno (2005) found the ages of mothers and children in Nairobi to influence children’s health seeking behaviour. Older mothers were less likely to take their sick children for treatment than younger mothers. In the same study, treatment was mainly sought for younger children (below 4 years) than those above 4 years. Also, Abdulraheem & Parakoyi (2009) found mothers to seek care the more for their 0-12 month’s old babies who were sick. This, according to the researchers, gradually declined for older age groups and became lowest for the 49-60 months old babies. Similarly, Aboagye & Agyemang 18 http://ugspace.ug.edu.gh/ (2013) confirmed the influence of age on individual’s health seeking behaviour in Ghana. They found younger women with first time experience to receive antenatal care (ANC) and assisted delivery from a medically trained provider whereas the older and more experienced women mainly receive less or no ANC visits and deliver in the house with or without help. Also in Ghana, Danso-Appiah et al. (2010) identified an inverse association between age and health seeking behaviour. According to them, treatment was sought the more for young children (aged 0-9) than for teenagers and to a lesser extent for adults. As reported by Sreeramareddy et al. (2006) and Pronyk et al. (2001), the current study expected no significant association between age and individuals health seeking behaviour in the municipality. 2.2.2 Sex The sex of individuals has been identified to influence their treatment seeking options at particular health service providers (El Kahi et al., 2012; Khabiri et al., 2011; Mashreky et al., 2010). Sex, normally goes beyond the biological differences between men and women as it is attached with social and cultural strings (Prosser, 2007). This induces men to be seen as superior over women as men are ranked higher in terms of abilities hence command the higher positions in society due to their access to almost all the economic resources (Shaikh et al., 2008; Hausmann-Muela et al 2003). With this, the power to make decisions, which includes health decisions regarding themselves and other family members, lies in the domain of men (Atuyambe et al., 2009; Shaikh et al., 2008). This has led to a huge disparity in terms of health seeking behaviour and sex especially in the developing nations where it is more profound and devastating, causing women than men to be poorer in health (Shaikh et al., 2008; Hausmann-Muela et al., 2003). This, as a result, cause women than 19 http://ugspace.ug.edu.gh/ men to engage the more in self-treatment and home remedy applications (Shaikh et al., 2008; Hausmann-Muela et al., 2003). But this is not the case in the highly industrialised nations where equal attention is given to the health needs of both men and women hence no disparity exist regarding sex and health service utilisation (Hausmann-Muela et al., 2003). Sandman et al. (2000) buttressed this where they rather found men to be “out of touch with the health care system” as men more than women did not see a physician in the year prior to the study and again more men than women did not have a regular doctor to go to when sick and needed medical care in the United States of America. Studies (Ghosh et al., 2013; Teerawichitchainan & Phillips, 2007: Taffa & Chepngeno, 2005) indicate treatment seeking at especially formal medical facilities for sick male children by parents more than for their sick female children. This reflects gender disparity even at the very early stages of life as some parents prioritise the well-being of their male children because they are the future bread-winners of their families (Ghosh et al., 2013; Shaikh et al., 2008). There are instances where females, both young and aged, have engaged the more with qualified service providers whenever ill than their male counterparts (El Kahi et al., 2012; Khabiri et al., 2011; Mashreky et al., 2010). Also, Fomundam et al. (2012) found females than males to seek treatment the more at informal health providers and attributed such treatment seeking trend to a female-dominating study. El Kahi et al. (2012), Khabiri et al. (2011) and Teerawichitchainan & Phillips (2007) established a significant association between sex and health seeking behaviour. The study expected a significant difference in terms of sex and health seeking behaviour in the municipality due to a male-dominating population and males having access the more to economic resources in the municipality (GSS, 2012). 20 http://ugspace.ug.edu.gh/ 2.2.3 Education The educational status of people, all over the world, has been found to be associated with their health seeking behaviours. Whether to seek treatment or not and the choice of health providers at which care is sought is greatly influenced by the educational level of an individual (Anwar et al., 2012; Shaikh et al., 2008). It is therefore unsurprising that education is included in the models of both Andersen & Newman (1973) and Kroeger (1983) as one of the variables that may influence an individual’s health provider choice and utilisation of health services. In the highly developed countries, what seems to be a positive association between education and healthcare utilisation exists where individuals with higher educational background mainly go for treatment at formal medical providers as their low educated counterparts usually self-treat and/or self-medicate as first actions taken whenever they perceive symptoms of illness (Cropley, 2004; Roy et al., 2004; Sandman et al., 2000). For instance, Roy et al. (2004) and Sandman et al. (2000) found less educated men more than their high educated counterparts in the United States of America to seek treatment the more at informal medical sources. The opposite, according to the researchers, is true. Treatment seeking at formal than informal medical providers by the high educated compared with their less educated and no education counterparts has been reported to predominate in developing countries (Haque et al., 2012; Rahman et al., 2012; van der Hoeven et al., 2012) though substantial proportions of individuals with higher educational background have been found seeking treatment at informal medical providers (Fomundam et al., 2012; Yanagisawa et al., 2004). For instance, Oluwatuyi (2010) and Kakai et al. (2009) noticed “an inverse relationship between education and health seeking behaviour 21 http://ugspace.ug.edu.gh/ among rural dwellers in Ekiti State, Nigeria and Kenya. They found the more educated compared with the no education and less educated to seek treatment the more from qualified medical personnel, while self-treatment and visits to traditional and faith healers increased with decreasing educational status as first action taken when ill. Similarly, the educational background of individuals in Ghana, has been found to influence the type of health providers patients frequent most (Aboagye & Agyemang, 2013; Annan et al., 2013; Dako-Gyeke et al., 2013). The more educated, according to Blanchet et al. (2012), were more likely to be enrolled on the National Health Insurance Scheme (NHIS) and also sought formal medical care than the less educated. Aboagye & Agyemang (2013) found the more educated women in the Bosomtwe district of Ghana to receive ANC and assisted delivery from a medically trained professional whereas the illiterate women mainly visit TBAs for antenatal and delivery care. Similar pattern of health seeking is expected to be found in the Akuapem-North Municipality where the high educated than their low educated and ‘no education’ counterparts seek treatment the more at formal than informal medical providers. While some studies (Oluwatuyi, 2010; Russell, 2008; Teerawichitchainan & Phillips, 2007) found the educational background of individuals to significantly influence their health seeking behaviour others (van der Hoeven et al., 2012; Taffa & Chepngeno, 2005) did not find any significant association. 2.2.4 Employment status and occupation The employment status and occupation of individuals has been found to influence greatly their health seeking behaviour since it has been established to determine their income levels (Kakai et al., 2009; Ahmed et al., 2005). People in active employment are expected to seek 22 http://ugspace.ug.edu.gh/ treatment the more at formal medical providers since they can afford treatment cost and will seek treatment the first day of symptoms identification. The unemployed are expected to seek treatment the more at informal medical healers whenever sick as they cannot afford the cost of service at especially formal medical facilities, hence delay in treatment at such health facilities which worsens their health conditions. Studies done by Mashreky et al. (2010) in Bangladesh, Kakai et al. (2009) in Kenya as well as Sudharsanam & Rotti (2007) in India respectively have reported of the significant association between employment status and health seeking behaviour. As the employed parents took their sick children to qualified medical providers for treatment, their unemployed counterparts, because of their inability to afford treatment costs at formal health facilities, sought treatment at informal medical sources for their sick children. Similarly, Atuyambe et al. (2009) likewise Chakraborty et al. (2003) found in Uganda and Bangladesh that women with some form of work doing and those with husbands working in businesses and in government organisations go for treatment mainly at formal medical facilities as first action taken whenever illness symptoms were perceived while their unemployed counterparts and those whose husbands were farmers or not working mainly self-treated and visited traditional healers for treatment as first actions when ill. Notwithstanding, studies have reported of treatment seeking at informal medical sources by individuals who are gainfully employed as government officials, merchants, farmers, and fishermen (Nahar, 2010; Yanagisawa et al., 2004; Asenso-Okyere et al., 1998). This pattern of health seeking, according the researchers, is mainly as a result of the non-severity of illness and the aetiology of the illness which make the gainfully employed seek treatment the more from the informal medical system. However, the current study expected the 23 http://ugspace.ug.edu.gh/ employed more than their unemployed counterparts, in both communities, to seek treatment the more at formal medical facilities. 2.2.5 Income Income is considered as the major socio-economic determinant of peoples’ health seeking behaviour (Adanu et al., 2008; Ahmed, 2005; Ahmed et al., 2005) as portrayed by the frameworks regarding the study of individual’s health provider choices. Income influences the ability to spend on healthcare which in turn influences the type of health provider an individual chooses to seek care from whenever ill (Ahmed, 2005). All over the world, people in the higher income groups have been found to be healthier than those in the lower income group because of their ability to pay for health services especially from the more qualified medical practitioners. Studies (NRHA, 2006; Roy et al., 2004; Sandman et al., 2000) conducted in the developed nations have confirmed the positive correlation between income and treatment seeking at formal medical providers as high income earners compared with their low income counterparts seek treatment the more at formal than informal medical providers for either themselves or their families as first action taken when ill. Also, studies (Mushtaq et al., 2011; Nahar, 2010; Gotsadze et al., 2005; Owusu-Daaku & Smith, 2005) have identified frequent and increased visits to hospitals, clinics and health centres (formal medical providers) in the developing countries by individuals belonging to the high income groups whereas majority of the poor (low income earners) mainly seek treatment from the informal medical channel whenever illness symptoms are perceived. This is true even in the rural areas (Rai et al., 2012; Oluwatuyi, 2010; Ahmed et al., 2005) 24 http://ugspace.ug.edu.gh/ where a dichotomy exists in treatment seeking between the rich and the poor. As the rural rich, in most cases, seek care at formal health facilities their poor counterparts, who are the majority, mostly self-treat, apply a home remedy, visit traditional and faith healers for cure and only go for treatment at formal health providers whenever their illness symptoms are perceived to be severe (Rai et al., 2012; Oluwatuyi, 2010; Biswas et al., 2006; Taffa & Chepngeno, 2005). According to Ahmed et al. (2005), “poverty is the single most important factor” that causes almost all the people in rural Bangladesh to either choose self-care/self- treatment or seek care at a drug store salesperson. However, the rich are sometimes found to engage with the informal medical system even the more than the poor (Afolabi et al., 2013; El Kahi et al., 2012; Yanagisawa et al., 2004; Davidson & Schattner, 2003). Among university students in Nigeria (Afolabi et al., 2013) and Lebanon (El Kahi et al., 2012) who were found to be rich, self-treatment and/or self- medication was/were the commonest form of care sought with medicines purchased over the counter or obtained through friends. Davidson & Schattner (2003) identified among sick medical doctors with very high income levels to self-treat as their first and predominant choice of care seeking of acute minor illnesses and explained that they (medical doctors) “often feel embarrassed to attend another doctor” for treatment. Also in Cambodia, Yanagisawa et al. (2004) found that for both first and second actions taken toward treatment seeking, the better-off compared with the very poor resorted the more to home remedies and self-medication respectively while the very poor were found to seek treatment the more at a health centre than the better-off. As a significant association was found between income and individuals’ health seeking behaviour (Danso-Appiah et al., 2010; Oluwatuyi, 2010; Kakai et al., 2009; Taffa & 25 http://ugspace.ug.edu.gh/ Chepngeno, 2005), other studies (Rai et al., 2012; Hjelm & Atwine, 2011; Pronyk et al., 2001) did not find any significant association between individuals’ income levels and their health seeking behaviour. The study, however, expected the high income earners than their low income counterparts in both communities to seek treatment the more from formal than informal medical sources. 2.2.6 Health insurance Earlier studies on health seeking behaviour and healthcare service utilisation, all over the world, have reported of both direct and indirect correlation between health insurance and health provider choices by the individual ( Blanchet et al., 2012; Robyn et al., 2012; El- Kak et al., 2009; NRHA, 2006; Roy et al., 2004). The presence of insurance guarantees the individual of economic access to healthcare services whereas its absence means increased cost of service use, hence determines the type of health provider an individual seeks health at whenever illness symptoms are perceived. Studies done in the developed nations (NRHA, 2006; Roy et al., 2004; Sandman et al., 2000) found that medical/health insurance influenced the use of professional medical services for both adults and children where the insured mostly seek care from qualified medical personnel while the uninsured normally self-treat and/or self-medicate or do not seek healthcare services at all. Similarly, studies (Ghosh et al., 2013; Robyn et al., 2012; Khabiri et al., 2011; El-Kak et al., 2009) done in developing countries have also confirmed the association between health insurance and treatment seeking at a particular medical channel. For instance, Ghosh et al. (2013) found in West Bengal that majority of insured mothers (77.8%) than their uninsured counterparts sought treatment at formal medical facilities for their sick children. El-Kak et 26 http://ugspace.ug.edu.gh/ al. (2009) found in Beirut that more insured women than their uninsured counterparts did seek reproductive healthcare services. In Pakistan, Siddiqui et al. (2011) found lack of insurance as one of the major reasons why a high proportion of the studied population self- treated and applied home remedies. Khabiri et al. (2011) found some patients in Iran to self-treat because they were uninsured hence could not afford healthcare services. Interestingly, studies done in Ghana (Aboagye & Agyemang, 2013; Annan et al., 2013; Blanchet et al., 2012) and Africa (Robyn et al., 2012) on the whole, found contrasting results regarding individuals’ health insurance coverage and their health seeking behaviour. In some cases more of the Ghanaian insured sought treatment at informal medical providers due to reasons such as proximity, poor road network and inadequate means of transport (Aboagye & Agyemang, 2013) likewise other parts of Africa (Robyn et al., 2012). Also, treatment seeking at formal medical facilities was the dominant health provider choice for the Ghanaian insured while the uninsured mainly visited prayer camps, fetish shrines, herbalists and pharmacy shops for treatment (Annan et al., 2013; Blanchet et al., 2012). Health insurance coverage was found to significantly influence individuals’ health seeking behaviour (Annan et al., 2013; Khabiri et al., 2011; El-Kak et al., 2009). 2.2.7 Severity of illness How severe and/or serious an illness is has been found to greatly determine individuals’ choice of a particular medical system whenever illness symptoms are perceived (Anwar et al., 2012; Rahman et al., 2012; Danso-Appiah et al., 2010) and also when treatment is sought for a particular symptom of illness (Taffa & Chepngeno, 2005). Studies done in the developed nations such as the United States of America (NRHA, 2006; Roy et al., 2004; 27 http://ugspace.ug.edu.gh/ Sandman et al., 2000) have identified the non-severity of illness as one of the major reasons why patients mainly self-treat and/or self-medicate hence, delay in care seeking especially at the formal medical providers. In developing countries where the health system, compared to the developed countries, is far from advanced and are still at an early stage of addressing basic healthcare requirements (Anwar et al., 2012), treatment seeking from the informal health sector is a common place (Awusabo-Asare & Anarfi, 1997). Majority of patients delay in seeking treatment and in the end, resort to self-treatment and/or self- medication, traditional healers, and faith healers when they finally decide to seek some form of medication (Asenso-Okyere et al., 1998; Awusabo-Asare & Anarfi, 1997). Treatment seeking from the formal medical system is only sought when illness symptoms become highly severe either for themselves or for a family member (Rahman et al., 2012; Khabiri et al., 2011; Danso-Appiah et al., 2010; Oluwatuyi, 2010). In some instances, individuals are found to wait for over three (3) days (Rahman et al., 2012; Danso-Appiah et al., 2010) and as long as ten (10) days (Sudha et al., 2003) before seeking some form of treatment all because their illnesses were seen not to be severe while in some cases treatment is sought immediately illness symptoms are perceived due to their severity (Khabiri et al, 2011; Teerawichitchainan & Phillips, 2007; Sreeramareddy et al, 2006). Also, there are instances where patients wait for months after symptoms identification before care is sought. This is attributed to the type and nature of the illness (Subba, 2004). In contrast, Chakraborty et al. (2003) found a significant proportion of pregnant women in rural Bangladesh suffering from very severe and high-risk diseases such as excessive vomiting, high-risk conditions of fever and cough and severe conditions 28 http://ugspace.ug.edu.gh/ of abdominal pains not to seek any form of care. Only a few, according to the researchers, sought care from doctors, nurses or family welfare visitors (FWVs). In Nairobi, Kenya, mothers belonging to the lowest socio-economic group sought treatment mainly at drug/chemical shops for their sick children or applied a home remedy and only took their sick children to a health facility when they perceived their wards’ illnesses to be severe (Taffa & Chepngeno, 2005). Similar trends have been reported in urban Nigeria among university students (Afolabi et al., 2013), among rural dwellers in Ekiti State, Nigeria (Oluwatuyi, 2010) and among diabetic patients in Uganda (Hjelm & Atwine, 2011). In addition to this, studies in Ghana have found some patients to wait for, as long as, four to five days after symptoms identification before some form of treatment was sought while others sought care mainly the first two days symptoms of illnesses showed up due to their perceived severity (Annan et al., 2013; Laar et al., 2013; Danso-Appiah et al., 2010). Again, Blanchet et al. (2012) and Russell (2008) have identified in Ghana that even after being covered by health insurance, people still delay in health seeking when sick or perceived symptoms of illness because of lack of severity of the so called illness. Though most studies (Annan et al., 2013; Laar, et al., 2013; Khabiri et al., 2011; Danso-Appiah et al., 2010) have found individuals’ illness severity as a significant predictor of their health seeking behaviour, some (Mushtaq et al., 2011; Russell, 2008) have not established such a significant association. However, the current study envisaged a similar trend of health seeking regarding individuals’ illness severity. 29 http://ugspace.ug.edu.gh/ 2.2.8 Availability of health providers The notion has been that individuals will seek treatment at health service providers that are readily available in their communities, and to a larger extent, at the ones that are closer to their houses the more, as the current study expects. It is also perceived that people will seek care from particular health service providers because they are the only ones providing health services in their localities. Numerous studies (Aboagye & Agyemang, 2013; Annan et al., 2013; van der Hoeven et al., 2012) have confirmed these assertions though some (Nahar, 2010; Mazumda et al., 2009; Crommett, 2008; Russell, 2008) have disproved such ideologies on certain grounds as illness needing special care, health insurance, quality of service provided and among others. Globally, treatment seeking at more available and closest health providers especially in rural areas has been found to predominate for general illness except situations where special services or treatment is required (Aboagye & Agyemang, 2013; Nahar, 2010; Crommet, 2008; Russell, 2008). Rural dwellers, due to certain hindrances such as distance, low socio-economic status, lack of insurance, aetiology of illness, cultural beliefs and practices, age and gender of patients, and patients’ position in household mainly engage with health providers that are found in their localities and are more closer to their houses (Ghosh et al., 2013; Al-Mandhari et al., 2009; Atuyambe et al., 2009; Shaikh et al., 2008). This made pregnant adolescents in Uganda to resort to home-prepared concoctions and TBAs since these were found in one’s immediate surroundings and comes with no or less cost compared with care seeking at public and private clinics and health centres that are distantly located from their houses (Atuyambe et al., 2009). However, in extreme cases where sickness calls for special treatment, patients are insured and quality of service 30 http://ugspace.ug.edu.gh/ becomes a priority, rural dwellers travel long distances going for care far away from their homes (Aboagye & Agyemang, 2013; Ghosh et al., 2013; Mazumda et al., 2009; Cropley, 2004). On the other hand, visits to formal medical sources that are mainly the closest to individuals’ home predominate in urban areas (Dominic et al., 2013; van der Hoeven et al., 2012; Danso-Appiah et al., 2010; Sudha et al., 2003) though some urban dwellers also resort to home remedies, self-medication and traditional healers for being close to their homes (Annan et al., 2013; Mashreky et al., 2010; Yanagisawa et al., 2004). As almost all the study respondents in Iran sought treatment at hospitals and clinics for being available in the locality (Khabiri et al., 2011), majority of rich urbanites, according to Yanagisawa et al. (2004), resorted to home remedies and self-medication as first and second actions taken to treat illnesses because no transportation cost was borne in accessing them. In the same way, van der Hoeven et al. (2012) in South Africa have confirmed the predominant treatment seeking at formal health facilities by urbanites due to their availability. Similar results have been documented in Ghana where the presence of hospitals and clinics, both public and private in urban areas, made the inhabitants go for treatment the more at such facilities (Dako-Gyeke et al., 2013; Danso-Appiah et al., 2010; Adanu et al., 2008) whereas others (Aboagye & Agyeman, 2013; Annan et al., 2013; Laar et al., 2013) report of care seeking at informal medical providers mainly in rural Ghana due to their availability in such rural communities. Contrary, studies (Aboagye & Agyemang, 2013; Dako-Gyeke et al., 2013; Russell, 2008) indicate visits to health service providers that are not in patients’ locality due to reasons including quality of care, severity of illness, nature of illness, health insurance and provider’s attitude and behaviour towards patients. Some studies in the 31 http://ugspace.ug.edu.gh/ country (Annan et al., 2013; Asenso-Okyere et al., 1998; Awusabo-Asare & Anarfi, 1997) however, have reported of urbanites seeking treatment at all sort of informal medical providers due to their closeness to them. 2.2.9 Distance to health providers All over the world distance has been identified as the key geographical factor that greatly influences and determines the use of health services at a particular medical system (Anwar et al., 2012; Al-Mandhari et al., 2009; Shaikh & Hatcher, 2005). Physical distance to a health service provider, in addition to its influence on individuals’ health provider choices, determines also the means of transportation and the cost of service use. If the distance to the facility is very far, one may have to board a vehicle or motorbike before care can be accessed thereby increasing the total cost of service utilisation. Physical accessibility has been found not to be an obstacle to treatment seeking in the advanced countries like USA (NRHA, 2006; Roy et al., 2004). However, in developing countries such as Pakistan (Mushtaq et al., 2011; Siddiqui et al., 2011), Bangladesh (Haque et al., 2012; Rahman et al., 2012; Mashreky et al., 2010; Biswas et al., 2006), Oman (Al-Mandhari et al., 2009), Bengal (Ghosh et al., 2013), Nepal (Sreeramareddy et al., 2006; Subba, 2004), India (Mazumda et al., 2009; Sudharsanam & Rotti, 2009; Sudha et al., 2003), Cambodia (Yanagisawa et al., 2004), Vietnam (Teerawichitchainan & Phillips, 2007), Georgia (Gotsadze et al., 2005) and Iran (Khabiri et al., 2011), long physical distance people have to cover before seeking treatment from especially the formal medical sources make them cut down their healthcare service consumptions and resort in the end to self-treatment/self- medication, traditional healers, pharmacy and drug store purchases and home-made remedies. This, according to the researchers, is more profound in the rural than the urban 32 http://ugspace.ug.edu.gh/ areas of the developing world. For instance, majority of individuals in Pakistan (Mushtaq et al., 2011), India (Mazumda et al., 2009), and Georgia (Gotsadze et al., 2005) travel over ten (10) kilometres before reaching the nearest formal medical facility for care hence, usually self-treat/self-medicate and/or seek treatment at traditional healers as first actions taken whenever illness symptoms are perceived. Studies conducted in Africa (Robyn et al., 2012; Oluwatuyi, 2010; Abdulraheem & Parakoyi, 2009; Atuyambe et al., 2009) and Ghana (Annan et al., 2013; Laar et al., 2013; Adanu et al., 2008) to be specific, confirm reports from other parts of the globe. The words of Kakai et al. (2009) in rural Kenya buttresses this; “the nearer the treatment venue is to the home the more likely the treatment will start early and the earlier the treatment starts after diagnosis of signs/symptoms of the disease, the better for reduction of complications and eventual death.” This was the responsibility of mothers and care givers who mainly resort to home remedy applications due to the distance they have to ply before treatment at a formal medical facility can be accessed (Kakai et al., 2009). Physical distance caused pregnant women in especially rural Ghana to deliver in the house with or without TBAs’ supervision even after being enrolled on the country’s health insurance scheme (Aboagye & Agyemang, 2013; Blanchet et al., 2012). Also, patients in northern (Annan et al., 2013) and southern (Russell, 2008) Ghana sometimes seek care at prayer camps, fetish shrines, herbalists and other traditional healers and home remedies due to over 30 minutes’ of walk to nearby hospitals, clinics and health centres in order to access formal healthcare services. 33 http://ugspace.ug.edu.gh/ 2.2.10 Quality of service Quality of care, according to Uzochukwu & Onwujekwe (2004) is a “subjective variable” hence, becomes difficult to assign to it an object and measureable indicator since ‘quality’ is a function of who defines it. Thus, what may seem to be ‘quality’ to someone may not be so to another. However, researchers have tried measuring ‘quality of care’ provided by different health providers by looking at it from the respondents’ view-point as ‘perceived quality’ (Ghosh et al., 2013; Anwar et al., 2012; Russell, 2008; Uzochukwu & Onwujekwe, 2004). With this, different criteria have been used in measuring individuals’ perceived quality of a health provider but, in this study respondents’ satisfaction with attitude of staff of health providers, whether other members of the community take the same actions respondents took and whether respondents recommend to other community members the health providers they took actions at were used as the indicators for individuals’ perceived quality of care rendered by a health provider. Studies all over the world have indicated low quality of service of formal medical facilities as one of the main determinants of seeking care at informal medical providers especially among people of high socio-economic status (Khabiri et al., 2011; Mushtaq et al., 2011; Biswas et al., 2006; Roy et al., 2004). In West Bengal, majority of mothers did not seek healthcare for their sick children because they were dissatisfied with the services provided and the inhumane attitude of staff of healthcare providers (Ghosh et al., 2013). In Iran, more people were pushed to seek care at the expensive private-for-profit facilities, self- treat or did not seek health at all due to low quality of service of the public health sector (Khabiri et al, 2011) as lack of trust and/or confidence, preferential treatment and inhumane attitude of health care personnel compelled people to self-treat/self-medicate even in the 34 http://ugspace.ug.edu.gh/ developed countries (NRHA, 2006; Roy et al., 2004; Sandman et al., 2000). However, some studies have also affirmed the high quality of care provided by the formal than the informal medical system being the reason for their corresponding high patronage (Dominic et al., 2013; van der Hoeven et al., 2012; Subba, 2004; Uzochukwu & Onwujekwe, 2004). This is so even in rural areas where majority go for treatment at hospitals, clinics, health centres and CHPS compounds, be it public or private though they might have consulted other providers initially (Dominic et al., 2013; van der Hoeven et al., 2012; Hjelm & Atwine, 2011). Contrasting results have been documented regarding the quality of service offered by different health providers on the continent as a whole and in Ghana in particular. Whereas some individuals (67%) consulted drug stores for treatment because they could not get treated and that their conditions worsened after care seeking from clinics (Fomundam et al., 2012) others self-treated/self-medicated due to long waiting times, expensive care provided at hospitals, poor attitudes of hospital staff and unavailability of medicines at hospitals (Afolabi et al., 2013). Also, poor hygienic practices, poor health personnel attitude and behaviour, and total poor quality of care at public health facilities are some of the determinant factors that discourage adolescents from seeking antenatal care (ANC) and delivery services in Uganda (Atuyambe et al., 2009). In the same study (Atuyambe et al., 2009), some adolescents were of the view that the good quality of care provided by some public healthcare facilities, better hygienic conditions and the good personnel-patients rapport motivated them to go for treatment at such health providers. Russell (2008) found in Ghana that poor quality of care offered by especially public healthcare providers caused majority of the people to seek treatment at private health providers to the extent of some 35 http://ugspace.ug.edu.gh/ visiting faith healers and practising over-the-counter sales. Similar results have been found by Blanchet et al. (2012), Danso-Appiah et al. (2010); Adanu et al. (2008); in the country, whereas others (Annan et al., 2013; Dako-Gyeke et al., 2013; Laar et al., 2013) have reported of high utilisation of formal medical facilities in the country due to patients’ satisfaction with services offered, comfort with services, short wait times at facilities, good personnel-patient rapport and perceived quality of care provided. In the current study, quality of services of health providers is expected to be high in the urban than rural community and from especially formal medical facilities than informal medical providers. 2.2.11 Health service cost and financing The cost of health service use and the means through which patients pay for their health service use have been identified to influence their health seeking behaviour (Aboagye & Aggyemang, 2013; Atuyambe et al., 2009; El-Kak et al., 2009; Russell, 2008; Biswas et al., 2006). Relevant as it is, Kroeger (1983) includes the cost of health service use in his model as one of the numerous factors that influence individuals to make decisions regarding their health whenever illness symptoms are perceived. The cost of health service use, in most cases, has deterred people from seeking treatment at formal medical facilities than at informal medical providers due to the relative high costs of service use at the former (Atuyambe et al., 2009; Adanu et al., 2008; Biswas et al., 2006; Subba, 2004). This causes majority of patients to go for treatment at informal medical providers in the end (Subba, 2004; Yanagisawa et al., 2004; Asenso-Okyere et al., 1998). For instance, an overwhelming proportion (73.2%) of patients in Nepal engaged with informal medical providers due to their inability to afford the cost of service use at formal medical facilities (Subba, 2004). In the same way, majority of pregnant adolescents in Uganda (Atuyambe 36 http://ugspace.ug.edu.gh/ et al., 2009) and poor likewise better-off Cambodians (Yanagisawa et al., 2004) resorted to home remedies and self-medication as first actions taken when ill because they were free hence borne no cost for their use. Also, treatment seeking at informal medical providers is on the ascendancy due to their flexibility in terms of payment for health service use (Nahar, 2010; Biswas et al., 2006). Though costs of health service use at formal medical facilities has been found to be expensive relative to that of informal medical providers, studies (van der Hoeven et al., 2012; Russell, 2009; Cropley, 2004) report of cost of health service use being relatively cheaper at public than at private formal medical facilities. In studies conducted in South Africa by van der Hoeven et al. (2012), Ghana by Russell (2008), and Belize, Central America by Cropley (2004), patients sought treatment the more at public than private formal medical facilities because costs of service use at the former was found to be relatively less. This was the case even among rural dwellers (van der Hoeven et al., 2012). Also, mothers in Belize, Central America sought treatment the more at government hospitals and clinics for their sick children suffering from malaria than at private formal health facilities because care seeking at the former was free (Cropley, 2004). Likewise, poor patients in Cambodia were found to seek treatment the more at public health centres than at private clinics and health centres due to their relative low cost of service use (Yanagisawa et al., 2004). In certain circumstances, patients rather seek treatment the more at formal medical facilities than at informal medical providers due to the relative high cost of service use at the latter (Nahar, 2010; Cropley, 2004; Yanagisawa et al., 2004). The poor in Cambodia were found to seek treatment the more at public health centres than at traditional healers due to the 37 http://ugspace.ug.edu.gh/ relative high cost of service use at the latter. Also in Bangladesh, Nahar (2010) found the cost of service use of informal medical providers in the immediate communities to be very expensive hence, sought care at informal medical providers in other localities. Cost of health service use has been found to significantly determine individuals’ health seeking behaviour (Russell, 2008). On the other hand, the mode through which patients finance their health service use also predicts their health seeking behaviour (Aboagye & Agyemang, 2013; Khabiri et al., 2011; Danso-Appiah et al., 2010). Globally, health/medical insurance and out-of-pocket payment are the two main modes by which people finance their health service use. Health/medical insurance, for a long time, has been the major means of health service use payment in the highly industrialised nations (Roy et al., 2004; Sandman et al., 2000). This has increased the probability of patients’ treatment seeking at formal medical facilities hence, enhances prompt and appropriate health care seeking (Sandman et al., 2000). On the part of developing countries like sub-Saharan African countries where self-medication is the main treatment option (Awusabo-Asare & Anarfi, 1997), out-of-pocket payment was the major mode of financing health service use even at formal medical facilities especially before the advent of health insurance (Asenso-Okyere et al., 1998; Awusabo-Asare & Anarfi, 1997). The situation became worse after the Structural Adjustment Programme (SAP) in sub- Saharan Africa where the introduction of the “cash and carry” system pushed most patients away from seeking treatment at formal medical facilities (Agyemang & Adjei, 2008). The implication is that, treatment seeking at informal medical providers increased hence, rendering out-of-pocket payment as a disincentive for care seeking at especially formal medical facilities. 38 http://ugspace.ug.edu.gh/ Presently, health insurance has been the major mode of financing health service use especially at formal medical facilities (Annan et al., 2013; Ghosh et al., 2013; Khabiri et al., 2011). This has been effective after nations, including third world countries, subscribed to the World Health Authority’s (WHA) health care financing policy of universal health coverage, which aims at increasing access to health care services (Mazumda et al., 2009). Ghana is no exception of such policy which has seen the introduction of the National Health Insurance Scheme (NHIS). The NHIS, according to Aboagye & Agyemang (2013), Annan et al. (2013) and Saeed et al. (2013) has been the key health care service financing mode in the country ever since its implementation irrespective of patients place of residence. The scheme has given patients, whether poor or rich, financial access to health care service use hence, has aided in bridging the disparity gap as regards health service use (Blanchet et al., 2012). Nevertheless, payment for treatment seeking at informal medical providers has been mainly through out-of-pocket (Fomundam et al., 2012; Nahar, 2010; Biswas et al., 2006). Mode of health service financing has been found to significantly predict individuals’ health seeking behaviour (Saeed et al., 2013). However, the study expected treatment seeking at informal medical providers to be financed by out-of-pocket payments whiles treatment seeking at formal medical facilities is expected to be financed mainly through health insurance. 2.2.12 Place of residence and health service provider options The location or area where one lives has been found to greatly influence his/her health seeking behaviour (Laar, et al., 2013; Anwar et al., 2012; van der Hoeven et al., 2012; Mashreky et al., 2010). To highlight the importance of individuals’ place of residence in determining his/her choice of health service provider for consultations when sick or 39 http://ugspace.ug.edu.gh/ perceived symptoms of illness, Andersen & Newman (1973) have included it in their healthcare utilisation framework. Studies conducted across the globe indicate that a mark difference exist between the health seeking behaviour of rural and urban dwellers. Whereas urban dwellers are noted for seeking treatment from the formal medical system like hospitals, clinics and health centres, rural folks, on the other hand, are mostly found resorting to the informal medical channel such as traditional healers, drug store salespersons, drug peddlers and prayer camps as first responses to illness (van der Hoeven et al., 2012; Mashreky et al., 2010; Nahar, 2010; Biswas et al., 2006). Also, in cases where rural dwellers seek care at formal medical facilities, reports indicate that they usually do so at the public medical facilities while their urban counterparts normally seek treatment at the private health facilities (Dominic et al., 2013; van der Hoeven et al., 2012; Mushtaq et al., 2011). This finds prominence in developing countries like Ghana where a significant proportion of the population (49.1%) live in rural areas (GSS, 2012). This, in addition to the low socio-economic status of rural dwellers, the complete dearth of health facilities, poor transportation, high expenditure on treatment, high adherence of cultural beliefs and practices, non-severity of illness, poor attitude of health personnel, low quality of care provided at formal health facilities, and the ever presence of the informal health sector in the rural areas of the world give rise to the rural folk interacting the more with informal health providers (Aboagye & Agyemang, 2013; Anwar et al., 2012; Rai et al., 2012; Nahar, 2010) resulting in delays to seek treatment at formal medical facilities which may worsen their health conditions (Afolabi et al., 2013; Oluwatuyi, 2010; Shaihk et al., 2008) and facilitate their early mortality. 40 http://ugspace.ug.edu.gh/ Notwithstanding, reports indicate treatment seeking at informal medical providers such as self-care/self-medication, traditional healers, faith healers or spiritualists and home remedies by urbanites (Afolabi et al., 2013; Annan et al., 2013; Mashreky et al., 2010; Yanagisawa et al., 2004). For instance, Afolabi et al. (2013) found in a university community in urban Nigeria that self-treatment was the commonest form of care by respondents and that self-medication was the initial choice of care among university students when sick or perceived symptoms of illness. Similarly, Annan et al. (2013) found among urbanites in the central region of Ghana that over half (66%) of TB patients sought care at fetish shrines, herbalists, prayer camps and pharmacies. This, according to the researchers, was facilitated by close proximity, lack of health insurance, healthcare staff attitude and cultural beliefs regarding the cause and transmission of the disease. 2.3 Health facility attended when ill The decisions people do take whenever they perceive symptoms of illness have major consequences for both the occurrence and development of the illness (Iyalomhe and Iyalomhe, 2012). Such decisions are said to be influenced by a plethora of factors (Shaikh & Hatcher, 2005; Roy et al., 2004; MacKian, 2003) and cause people to either visit informal medical providers (traditional healers, drug/chemical shops, drug peddlers, prayer camps and home remedies) for cure or seek treatment at formal medical providers (hospitals, clinics, health centres, CHPS compounds and/or health posts) whenever illness symptoms are perceived (Anwar et al., 2012; Mashryky et al., 2010; Shaikh & Hatcher, 2005). Studies globally report of health seeking on the part of individuals from both the formal and informal medical systems but reports indicate the dominance of treatment seeking at formal medical providers in the highly industrialised countries than developing 41 http://ugspace.ug.edu.gh/ countries though substantial proportions of developed countries citizens visit informal medical providers for care (NRHA, 2006; Roy et al., 2004; Sandman et al., 2000). For instance, a substantial proportion (40%) of minority groups in the United States of America treated their sick children with home remedies (Roy et al., 2004) as first action taken whenever illness symptoms were perceived while Sandman et al. (2000) found men more than women, in the same country, to be “out of touch with the healthcare system” as more men than women were not having a regular doctor to go to when sick and needed medical care. Similarly, treatment seeking at informal medical providers tend to predominate in developing countries (Ghosh et al., 2013; Khabiri et al., 2011; Mushtaq et al., 2011) where it is more profound and devastating in the rural areas (Dominic et al., 2013; Haque et al., 2012; Rahman et al., 2012). A higher proportions of the population buy medicines from pharmacies and drug/chemical shops, visit traditional healers for care, apply home-made remedies and also buy medicines from drug peddlers as first actions taken whenever symptoms of illness are perceived though considerable proportions seek treatment at formal medical facilities (Ghosh et al., 2013; Siddiqui et al., 2011; Mashreky et al., 2010; Gotsadze et al., 2005). Health seeking behaviour in Africa is not different from what pertains in other parts of the world where majority of the people have been found to self-treat/self-treatment (Awusabo- Asare & Anarfi, 1997). An overwhelming proportion (75.6%) has been found to self- medicate as first action taken when ill even after health insurance (Robyn et al., 2012). Likewise, treatment seeking at other informal medical sources is paramount on the continent (Fomundam et al., 2012; Uzochukwu & Onwujekwe, 2004) where it is even 42 http://ugspace.ug.edu.gh/ worse in the rural areas (Oluwatuyi, 2010; Atuyambe et al., 2009) though studies (van der Hoeven et al., 2012; Taffa & Chepngeno, 2005) also show higher rates of treatment seeking at formal medical facilities especially in urban areas of the continent. In Ghana, similar pattern is found where the informal medical system is the main health seeking approach by most rural people due to inadequate formal health facilities coupled with poor road networks linking the few existing ones, low socio-economic status, low quality of service, inhumane health personnel attitude towards patients and a very high observance of cultural beliefs and practices whereas treatment is mainly sought at formal medical providers by urban dwellers (Agyemang & Aboagye, 2013; Crommett, 2008; Russell, 2008). This delays rural dwellers’ reporting times at formal health facilities and make them go for treatment at such facilities only when illness becomes severe. This can worsen their health conditions both presently and years to come, thereby, putting into jeopardy their entire existence. With this, the expectation of the current study is a confirmation of what the vast majority of literature has reported where health seeking at informal medical providers dominated. Therefore, individuals in the Akuapem-North Municipality were expected to seek treatment the more at informal medical providers than at formal medical providers as first action taken when ill. 2.4 The nature of the health system in Ghana The health system in Ghana comprises the formal and informal medical systems respectively as it is in other parts of the world especially, in developing countries (Fomundam et al., 2012). The formal medical system consists of both public and private healthcare facilities including hospitals, clinics, health centres and health posts or CHPS 43 http://ugspace.ug.edu.gh/ compounds, whereas the informal medical system include traditional medicine practitioners (whether registered or not), pharmacies, drug/chemical shops, drug peddlers, quack doctors and faith healers (Aboagye & Agyemang, 2013; Asenso-Okyere et al., 1998). Though the operations of some health providers of the informal medical system are, to a lesser extent, not recognized in the country, they play vital roles in healthcare service delivery to quite a significant proportion of the populace, especially those in the rural areas of the country (ANMHD, 2013). Although private entities also provide healthcare to portions of the populace, the main provider of healthcare services in Ghana is the government. This is administered through the Ministry of Health (MOH) and the Ghana Health Service (GHS) (Aboagye & Agyemang, 2013). The healthcare system has five levels of providers: Health posts – which are first level primary care for rural areas: Health Centres and Clinics, District Hospitals, Regional Hospitals and Tertiary Hospitals. There are about 3,316 healthcare facilities in the country of which 1,818 are public facilities, 1,294 are private facilities and 204 belong to religious institutions. Of this, the majority provide general services while very few provide specialised services (ANMHD, 2013). The implication is that, patients would have to travel to certain hospitals and sometimes outside the country if special services are required. Currently, only seven (7) hospitals specialise in Gynaecology, two (2) in Paediatrics and nine (9) specialise in Surgery (ANMHD, 2013). Unlike the private healthcare facilities that have adequate skilled personnel, more modern equipment and provide relatively better quality service, the public healthcare facilities are known for providing poor quality service and are inclined towards the core (Adanu et al., 44 http://ugspace.ug.edu.gh/ 2008; Russel, 2008). Averagely, the entire country is poorly served with formal health facilities as majority of both the public and private healthcare facilities are found in the urban areas with only those belonging to religious institutions found in the rural areas of the country (Asenso-Okyere et al., 1998). This explains the reason why a lot of people still rely on traditional healers, faith healers, home remedies, and over-the-counter sales especially in rural Ghana (Annan et al., 2013; Awusabo-Asare & Anarfi, 1997). Due to patients’ inability to access healthcare services under the “cash and carry” system operated by Ghana’s healthcare system some decades ago, the National Health Insurance Scheme (NHIS) programme was instituted. The NHIS is supposed to guarantee the individual of free access and use of healthcare services but interestingly care must be taken when healthcare services are accessed. This is because not all healthcare facilities accept the NHIS card and also not all treatments are covered by public insurance (Blanchet et al., 2012). 2.5 The nature of the health system in the Akuapem-North Municipality The health system in the Akuapem-North Municipality is not different from that of the entire country. Both the formal and informal medical systems are present in the municipality with the latter predominating especially in the rural areas, though some with their operations are not factored into the mainstream healthcare system. With over 70% of the formal medical facilities present in the few urban areas, health seeking from these sources is high on the part of the urbanites than their rural counterparts. Majority of the rural dwellers are believed to mainly engage the more with the informal medical system 45 http://ugspace.ug.edu.gh/ since they mostly delay towards treatment seeking at formal health facilities in the municipality (ANMHD, 2013). The distribution of formal medical facilities in the municipality is skewed towards the core as almost all of them are found in the top twenty (20) communities of the municipality (ANDA, 2006). This leaves the number of formal medical facilities to cater for peoples’ health needs in the rural areas to be very few compared with that of fetish shrines, drug/chemical shops, drug peddlers, prayer camps, herbalists and other forms of traditional healers. Being aware of its very limited number of qualified health personnel to cater for the health needs of the growing population, the municipality’s Health Directorate does recognise the operations of over 97 trained traditional birth attendants (TBAs) to assist in the reproductive healthcare service delivery for women, especially in the rural areas. They help in offering antenatal care and assisted delivery for women who unfortunately find themselves in rural areas of the Municipality, thereby reducing maternal and infant mortality. In furtherance to this, there are five (5) Primary Health Nurses (PHN) and about sixty (60) Community Health Nurses (CHN) who assist in providing especially the rural inhabitants with basic health needs. These, according to the Health Directorate, are not enough (ANMHD, 2013). 2.6 Conceptual framework for the study Based on the above review, Kroeger’s model of healthcare utilisation, which posits that health seeking behaviour is influenced by a plethora of factors (Shaikh et al., 2008; Hausmann-Muela et al., 2003), was adapted in this study with some modification. Here the model is used in the context of two communities (an urban and a rural community each) 46 http://ugspace.ug.edu.gh/ with populations of varying socio-demographic and economic background to identify the factors that influence an individual’s choice and use of health service at a particular health service provider taking into consideration the deficiencies of the model regarding its applicability in especially a developing country like Ghana. Also, the reason for adapting Kroeger’s model in this study is that it fits the objectives set for the study as it helped in wholly identifying the various factors that influence patients’ decisions to engage with a particular health service provider whenever illness symptoms are perceived. According to the model, an individual’s predisposing factors: age, sex, education, employment status and occupation, and place of residence; need factors: disease severity, aetiology and expected benefit of treatment; and enabling factors including health service availability, distance between individuals’ house and health providers, health insurance, service quality, and health service use and financing are the main determinants of individuals’ health seeking behaviour. These factors were perceived to determine the decisions individuals in the Akuapem-North Municipality took to engage with particular health service providers as first actions taken whenever they perceived symptoms of illness. These sets the theoretical base for the current study, hence the study into the health seeking behaviour of individuals in the Akuapem-North Municipality of Ghana. Figure 2.6.1 illustrates the adapted form of Kroeger’s (1983) model for the current study. 47 http://ugspace.ug.edu.gh/ Figure 2.6.1: Conceptual framework for the study INDEPENDENT VARIABLES DEPENDENT VARIABLES Predisposing factors (Age, sex, education, employment status and occupation, and HEALTH SEEKING place of residence) BEHAVIOUR Informal Medical System Characteristics of --Self-treatment the disorder and --Self-medication their perception --Drug/chemical shops (Disease severity, --Pharmacy shops Choice aetiology, diagnosis, --Traditional healers of health and expected benefit Perceived --Drug peddlers service of treatment) morbidity --Home-made remedies provider Characteristics of the Formal Medical System health service and --Hospitals enabling factors --Clinics --Health centres (HS availability, --CHPS compounds distance, income, --Health posts health insurance, service quality, health service use and financing) Source: Adapted from Kroeger (1983). 48 http://ugspace.ug.edu.gh/ 2.7 Summary This literature review identifies that Andersen & Newman’s (1973) and Kroeger’s (1983) models of healthcare utilisation are the most adaptable behavioural frameworks used by researchers in the study of individuals’ health service provider choices in rural and urban areas of both developed and developing nations. The review has provided an outline of the development of the literature on health seeking behaviour and has helped in identifying the relevant variables for understanding individuals’ health service provider choice patterns in both rural and urban areas whenever illness symptoms are perceived. Studies show the availability of formal medical providers in urban areas with well- developed infrastructure to take care of the health needs of the urbanites compared to rural areas. This facilitates treatment seeking at formal medical providers by urbanites the more than their rural counterparts who, due to the dearth of these formal medical facilities coupled with poor transport system, inadequate human and material resources, compel them to mainly go for treatment at informal medical healers that are ever present and easily accessible socially, economically and geographically. The chapter below presents the study area in which this current study was conducted and the methodology that was used to obtain the required data to determine which factors influence individuals’ choice of particular health service providers when sick or perceived symptoms of illness in both the urban (Adukrom) and rural (Korkormu) communities in the Akuapem-North Municipality in the eastern region of Ghana. 49 http://ugspace.ug.edu.gh/ CHAPTER THREE PROFILE OF THE STUDY AREA AND METHODOLOGY 3.1 PROFILE OF THE STUDY AREA 3.1.0 Introduction This chapter presents the background of the study area – Akuapem-North Municipality. It touches on the physical and demographic characteristics of the area. This includes the population size, age and sex distribution, dependency ratio and settlement patterns. It also discusses the socio-economic background of the municipality. The chapter also discusses the methodology adopted for the study. Here, both the cross- sectional research design and the mixed method approach of gathering data that were adopted for the study have been fully explained. 3.1.1 Physical characteristics 3.1.1.1 Location and size The Akuapem-North Municipality is located in the south-eastern part of the Eastern region of Ghana lying between latitude 6° 05’ South and 6° 45’ North and Longitude 0° 05’ East and 0° 30’ West. It is approximately 58km from Accra, the capital city of Ghana. The Akuapem-North Municipality shares boundaries to the northeast with Yilo Krobo District, north with New Juaben Municipal, southeast with Dangme West District, southwest with Akuapem South Municipal, and in the west with the Suhum-Kraboa-Coaltar District. The boundary stretches from Obosomase in the South, through to Abonse in the east to 50 http://ugspace.ug.edu.gh/ Okrakwadwo and Asamang in the North to Okorase and Mangoase in the West. The Municipality covers a land area of about 450 sq.km representing 2.3% of the total land area of the Eastern Region. The Akuapem-North Municipality has about 248 settlements with Akropong-Akuapem as its capital (ANDA, 2006). 3.1.1.2 Climate There are two rainy seasons with the major rainfall occurring between May and August and the minor rainfall in October. The mean annual rainfall is about 1,270mm. The mean annual temperature is 23.88ºC (75 ºF) however, mean daily temperatures range between 24ºC and 30ºC during the day and between 13ºC and 24ºC during the night. Also, the major dry season is from December to February with the minor occurring in August. 3.1.1.3 Vegetation The vegetation of the municipality is forest-with-shrub and semi-forest. The hills tops are characterised by broken forests. Secondary forests are mainly found on slopes and valleys. Scrub and bush dominate along the motor roads and main footpaths, while the slopes facing the Accra plains are predominantly thickets. Also, there are two major forest reserves though there are a lot of forest patches and sacred groves scattered all over the municipality. Notable ones can be found in Gyamfiase, Larteh Junction, Bankana near Tutu, Akropong, Obosomase, Mampong and Saforo. Most towns and villages are located on a mountain and visibility is very poor in the morning, because of the tall trees. It has also got deep valleys which make farming activities very difficult. Rainfall averages 1270mm, and the weather reflects the invigorating and salubrious, mild cold mountainous climate (ANDA, 2006). 51 http://ugspace.ug.edu.gh/ 3.1.1.4 Geology and soil The municipality consists of two main types of rocks of pre-Cambrian age namely; the Togo sandy shales series and the Birimian series. The Togo series are found in the north- east to the south-west from the Senya-Breku part of the Akuapem Range, which is west of Accra. The Birimian series, on the other hand, are found in Adawso area comprising of gneisses and schist with granites and pegmatites, which are metamorphic rocks. Sandy shales are found in Kwamoso area and Mamfe – Adawso road near mile 36. Sand series are found in the area stretching from the north-west slope of the Akuapem Range to the Nyensi and Nsaki Valleys. Phylites are found in stream valleys such as the Bump Valleys near Larteh. Also, quartzite covers greater parts of the Akuapem Range, Mampong, Tutu, Amanokrom and Larteh (ANDA, 2006). 3.1.1.5 Topography and drainage The Municipality is mountainous with hills ranging between 381meters and 488 meters in height, although the highest peak reaches 500 meters above sea level. There is one main hill range, Akuapem Range with height between 1,250ft (381m) and 16000ft (487.7m). The highest point is at Amanokrom near the water tank, which is about 1642ft (500m) and the lowest point is about 500ft. The municipality can be divided physically into two ridges, which are semi-urban, and the lowland area, which is rural (ANDA, 2006). 52 http://ugspace.ug.edu.gh/ Figure 3.1.1: Map of the study area Source: Akuapem-North Municipal Assembly (2006) 53 http://ugspace.ug.edu.gh/ 3.1.2.0 Socio-economic characteristics 3.1.2.1 Population characteristics The Akuapem-North Municipality had a total population of 68,247 in 1970 and increasing at an annual rate of 1.6%, thus increasing by a percent change of approximately 25% to 85,131 in 1984. It is estimated that the annual growth rate of the Municipality’s population between 1984 and 2000 was 1.8% while the total population as at 2000 was 104,753 made up of 48,942 males and 55,811 females which translates into a sex ratio of 0.88 males to 1 female (that is, 88 males to 100 females). Currently, the Municipality has a population of 136,483 made up of 64,028 males and 72,455 females with a growth rate of 1.8% (based on 2010 Population and Housing Census). This also reflects a sex ratio of 0.88 male to 1 female. This shows that the population is growing moderately as compared to the national population growth rate of about 2.6% per annum (GSS, 2012). There are two major ethnic groups in the Municipality, namely; Twi and Guan-speaking Akuapems. However, there are a number of settlers including Ewes, Ga-Adangmes, and people of northern extraction living in the municipality (ANDA, 2006). 3.1.2.1.1 Age and sex distribution The age and sex structure basically represents the distribution of population in terms of age and sex. In the municipality, the dominant sex is females constituting about 53.09% of the population whilst the males represent about 46.91% of the total population. The sex ratio of the municipality is about 0.88:1. The age structure of the municipality is basically youthful thus indicating that the working age in the municipality is more than those in the dependant age. The working age is about 63.4% of the population whilst the dependant age 54 http://ugspace.ug.edu.gh/ forms about 36.6% of the population. The age-dependency ratio in the municipality thus substantiates the fact that there are more people in the working age than the dependant age (GSS, 2012). Table 3.1.1: Age and sex distribution Age group Males (%) Females (%) Total (%) 0-14 18.66 18.00 36.66 15-64 25.32 29.84 55.16 65+ 2.93 5.25 8.18 Total 46.91 53.09 100.0 Source: GSS, 2012 3.1.2.1.2 Dependency ratio The dependency ratio – the ratio of the dependent-age population (the young and the old) to the working-age population – of the municipality is 1:0.8 which means that each 100 persons in the active population group are being depended upon by 80 persons in the inactive population group. This shows that the working population group is not over- burdened since one person in the active group takes care of himself and almost another person in the inactive group (GSS, 2012). 3.1.2.2 Settlement The municipality has a total of 248 settlements. Currently, there are four urban settlements namely, Akropong, Larteh, Mampong, and Adukrom with Akropong, Larteh and Mampong being the only first order settlements with populations of over 10,000 while four 55 http://ugspace.ug.edu.gh/ other settlements with populations of 5000 and over come out as the second order communities. Nine towns follow as third order with populations ranging between 2000- 2999. The rest have populations below 2000 and these towns are concentrated in the eastern part of the municipality. The western part is however, characterized by dispersed small settlements with populations far below one thousand (1000), except for a few communities (ANDA, 2006). 3.1.2.3 Education The municipality boasts of one of the only two schools for the visually impaired in the whole country. This is located in Akropong-Akuapem – the capital of the municipality. It also boasts of having one of the few schools for the deaf and dump at Mampong-Akuapem. Efforts are being made to help raise the standard of education in the municipality. The municipality has 147 Primary Schools, 87 Junior High Schools, 12 Senior High Schools, one College of Education and two private universities (ANDA, 2010). 3.1.2.4 Health The municipality can boast of a number of clinics and health centres scattered throughout towns and villages. Most of them have received support like the training of Medical Assistants and the provision of infrastructure. The municipality has a well-resourced Hospital which was named after Tetteh Quashie (Tetteh Quashie Memorial Hospital) and is privileged to have the largest Orthopaedic hospital – Bryan Lowe Orthopaedic Hospital – in the West African sub-region all at Mampong-Akuapem. The municipality also has the country’s most renowned research centre; the “Centre for Scientific Research into Plant Medicine” also at Mampong-Akuapem, which deals with herbal and traditional treatments 56 http://ugspace.ug.edu.gh/ of ailments. There are currently 26 health posts in the municipality, 8 sub-district clinics, 13 CHPS compounds, 4 RCH Clinics and one (1) Pharmacy – Finco Pharmacy at Mampong-Akuapem. Also, there are 11 doctors, 6 Medical Assistants and over 450 health workers in the municipality and one Sri-Plant Medicine. The doctor-patient ratio in the municipality is 1:12,408, which is higher than that of the entire country being 1:11,111 which is an indication of extra work on the part of the doctors. 3.1.2.5 Water and sanitation 3.1.2.5.1 Water There are five major sources of water in the municipality and these are pipe-borne, well, river, borehole and rain water respectively. About 26% of the people in the Akuapem- North Municipality have their main source of water from stream/river and well. The percentage of people having access to potable water is 74%. The implication is that a substantial number of the people are prone to diseases especially as the municipality is mountainous geographically with lots of chemicals and more especially as a large number do not treat their water before use. On the whole, the municipality faces serious water problems. About 10.9% have problems with the quality of the water they use with 26.7% of the people facing the problem of irregular supply. The remaining 23.3% face problems ranging from high cost to long distance (ANDA, 2010). 57 http://ugspace.ug.edu.gh/ 3.1.2.5.2 Sanitation situation Sanitation is a basic service that adds up to the quality of human life. Accessibility to adequate sanitation facilities will impact positively on human health. The KVIP constituting 56.7% is the most commonly found toilet facility in patronage in the municipality followed by the pit latrine which constitutes 22.7%. An assessment of the state of housing facilities and services in the entire district demonstrates that about 45.6% of the toilet facilities are well and appreciably maintained whilst 54.4% are poorly maintained contributing to the insanitary conditions of the municipality. Disposal of waste (solid and liquid) in the municipality is an issue of major concern. It is estimated that approximately 90% of the people dispose waste water through the free range and the disorganized drains systems while 76.4% dispose refuse through open dump. These methods of disposing off effluent are likely to cause diseases since a substantial number of the people (43.2%) live in compound houses. However, efforts are underway in curbing the problems posed by water and sanitation in the municipality. On the whole, the Water and Sanitation programme in the municipality has provided facilities for 46 communities - 29 for Akropong constituency and 17 for Okere constituency. Also, since even water supply to towns from Ghana Water Company is still a problem, the Community Water and Sanitation Agency (CWSA) is providing potable water for rural communities which have drastically reduced the incidence of water-borne diseases (ANDA, 2010). 3.1.2.6 Economic activities The main economic activities in the Akuapem-North Municipality include farming and commerce. Major crops grown are cassava, maize, yam, plantain, potatoes, fruits, citrus, 58 http://ugspace.ug.edu.gh/ oil palm cocoa and vegetables. Vegetables produced include tomatoes, lettuce, cabbage, pepper as well as squash. Vegetable production is irrigated. Major vegetable producing areas are Mampong, Aseseeso and Kwamoso. Non-traditional products, particularly snails and mushrooms, are also being produced. Commerce, the other economic activity of the municipality is rightly next to agriculture in terms of patronage. Over 60% of people engaged in commercial activities are into petty trading, 36.4% are into retailing whilst only 3.6% are into wholesale trading. This shows that majority of the people employed in the commercial sector are petty traders hence earn meagre incomes. The structure of the Akuapem-North Municipality’s economy remains an agrarian one. This shows that the agriculture sector dominates in terms of employment with about 67.4% of the people being employed in the sector as shown in table 3.1.2 below. There are three main sectors of the economy of Akuapem-North Municipality – agriculture, industry and service – and each contributes to employment (ANDA, 2010). Table 3.1.2: Contribution of the sectors to employment Sector District (%) National (%) Agriculture 67.4 55.0 Industry 9.0 11.7 Service 23.6 33.3 Total 100.0 100.0 Source: ANDA, 2010 In terms of the contribution of the sectors to total income, the service sector contributes the highest income in the municipality providing 38.7% of the total income while the 59 http://ugspace.ug.edu.gh/ agricultural sector is the least contributor to the economy contributing 29.8%. Even though most of the people within the district are engaged in agriculture, its contribution to income is relatively low, thus contributing to the low living standards of the people (ANDA, 2010). 3.2 METHODOLOGY 3.2.0 Introduction This section deals with the procedure followed to carry out the study. It covers the selection of the study sites, research design, primary data sources, which include individual surveys and in-depth interviews likewise the secondary data sourced for the study. Also, how the primary data garnered from the field was presented and analysed has been well discussed in this section. Lastly, the various challenges that were encountered during the data collection process are enshrined in this section. 3.2.1 Selection of study sites The communities in the entire municipality were categorised into two: thus, urban and rural communities respectively as defined in the municipality’s profile and mapping survey (ANDA, 2006). The categorisation was done to ensure that communities from different geographical locations in the municipality were given equal chance of selection for data collection. In order to ensure a higher level of precision and representation, the lottery approach of the simple random sampling method was used to select a community each from the urban and rural categories respectively. The urban community selected was Adukrom while Korkormu was the rural community chosen for the study. Thus, two communities, one urban (Adukrom) and one rural (Korkormu), were selected in all for the 60 http://ugspace.ug.edu.gh/ study. These were afterwards clustered into zones, based on location and settlement patterns, representing the existing enumeration areas delineated by the Electoral Commission of Ghana (ANDA, 2010). With this, the urban community (Adukrom) was clustered into five zones where three were randomly selected whereas the rural community (Korkormu) was divided into two zones with all being selected for data collection. Though a patient may seek health service from a facility outside his/her locality, health service providers, formal and informal, available in the selected communities were sampled for the study. This was done to examine whether the availability of a health service provider necessarily warrant utilisation. With this, a health service provider, either formal or informal, was randomly selected from each sampled enumeration area in both communities for data collection. Thus, a health centre and two drug/chemical shops were selected from the urban community (Adukrom) while a traditional birth attendant and a herbalist were also randomly chosen from Korkormu, the rural community. In all, five (5) health service providers were selected for the study. The selection of these health service providers were as a result of the individual survey where it was realised that they were mostly patronised by patients as first actions taken whenever illness symptoms were perceived. 3.2.2 Research design The study adopted a population-based cross-sectional study design to assess peoples’ health seeking behaviour in the Akuapem-North Municipality. Cross-sectional design, which deals with the collection of data on more than one case and at a particular point in time (Kothari, 2004, pp. 31), became useful for the study as data for the current study was collected on more than one case and at a single specified time. Also, Kothari (2004, pp. 31) noted that cross-sectional study, as a research design, is good for determining variations, 61 http://ugspace.ug.edu.gh/ patterns of association and may indicate causation through the examination of relationships between and among variables. This justifies why the study adopted a cross-sectional study design in its conduction. Also, in order to gain a comprehensive understanding of the issues examined, the study adopted the mixed method approach. This approach has been admitted by Teye (2012) as a sure way of scrutinising multifaceted phenomena. Thus, by implication, the study depended on both quantitative and qualitative data sources in order to realise the set objectives. The quantitative approach, which resonates with positivism, argue that science is neutral, objective and value-free hence aim at explaining general or broad behavioural patterns (Teye, 2012). On the other hand, the qualitative approach expresses the interpretivists’ ideology which gives credence to subjective interpretation of especially social reality rather than the robust and rigid generalisations of phenomena opined by positivism (Teye, 2012). Thus, for qualitative approach, much relevance is accorded to the perceptions and/or opinions, actions and behaviours of an individual or a group of people instead of the objective way pioneered by positivism, which is fact-based, bias-free and thus exist interdependently of the individual’s mind or perception. Based on these, the study adopted both approaches to unveil the complexities of the issues understudy. The relevance of mixed method approach in health service utilisation studies has been emphasised by Anwar et al. (2012), Grundy & Annear (2010) and Tipping & Segall (1995). According to the researchers, the qualitative approach is mainly used to identify patterns with respect to treatment seeking while the quantitative approach helps in determining the rate of occurrence connected with such patterns. Based on the intent of the researcher (Anwar et al., 2012) either approach can precede the other. The situation where the 62 http://ugspace.ug.edu.gh/ quantitative approach precedes predominate (Anwar et al., 2012; Grundy & Annear, 2010) but there are instances where the qualitative approach precedes (Tipping & Segall, 1995). In the study conducted by Ghosh et al. (2013) in West Bengal and Russell (2008) in the Dangme West District in Ghana, qualitative interviews were used to complement quantitative data already garnered via questionnaires. They reiterated that the qualitative interviews provided a thorough examination and evaluation of essential data for the study’s analysis. Conversely, Tipping & Segall (1995) preceded the quantitative method of data collection with qualitative interviews where the questionnaire survey was afterwards used to determine the association between the main patterns of care seeking initially identified through the qualitative interviews. However, to better comprehend the various issues and themes associated with the health seeking behaviour of people in the Akuapem-North Municipality of Ghana, the sequential explanatory strategy of research design was adopted by the study as recommended by Grundy & Annear (2010) and Prosser (2007). Thus, the quantitative method of data collection preceded the qualitative approach. The questionnaire survey was first used in garnering information especially regarding respondents’ behaviour towards their health and the health provider options in case of illness symptoms identification. This aided in determining frequencies, patterns, associations and predicting of illness response and the kind of health services providers they sought treatment at. The qualitative method was used to gather a thorough and comprehensive data for further understanding of the issues examined. Also, this gave a thorough and vivid insight about some of the themes that needed further examination from the quantitative facet of the study. 63 http://ugspace.ug.edu.gh/ Again, statistical techniques which are characteristic of the quantitative method for model specification, analysing quantifiable data, generalisations, predictions and establishing correlations between and among variables (Kothari, 2004, pp. 131) were used to test for causalities and establish associations among some selected variables. The study sought to examine the contextual influence of peoples’ health seeking behaviour in the municipality. This include respondents’ experiences, perceptions, emotions, beliefs and behaviours toward treatment seeking at a particular health service provider of which the quantitative approach is not very good at (Teye, 2012). Hence, the need to complement it with the qualitative approach in order to obtain thorough and comprehensive understanding of the issues and themes that emanated from the study. Against this backdrop, the study adopted the mixed method approach in order to gain complete understanding of the issues to be unfolded. 64 http://ugspace.ug.edu.gh/ Table 3.2.1 Research design Research Target Sample Intended Outcome Test Tool Population Size Questionnaire 1. Response to perceived Cross-tabulations, illness Chi-square tests Household 2. Choice and use of of significance respondents health service and Multinomial (Persons above 205 providers logistic regression 20 years) 3. Barriers to health service use 4. Modes of health service financing In-depth Personnel of 1. Health service Frequency and Interviews selected Health coverage perceptionist service 5 2. Health service approach providers delivery In-depth 1. Reasons associated Frequency and Interviews with first action taken perceptionist Individual when ill approach respondents 15 2. Barriers to health service use 3. Mode of health service financing 65 http://ugspace.ug.edu.gh/ 3.2.3 Primary data sources In order to provide a vivid and comprehensive understanding of the issues examined, the study relied heavily on primary data sources. The use of questionnaire has been identified by Kothari (2004, pp. 100) as the best method of primary data collection especially in cross- sectional surveys hence, its use as the main primary data collection tool in this study. With this, a semi-structured questionnaire was therefore used for the current study. In addition to this, the study further conducted fifteen (15) in-depth interviews with the various respondents in order to obtain detailed information about their engagement with particular health service providers whenever symptoms of illness were perceived. Also, five (5) in- depth interviews were conducted with various heads of health service providers where detailed information regarding their coverage and delivery of service was gathered. The decision to select the twenty respondents for the in-depth interviews was partly due to limited time and budgetary constraints and wholly because the information given was relevant and reflective of the issues examined. 3.2.3.1 Individual survey The study was based on an individual survey which has extensively been used in health seeking behaviour studies (Blanchet et al., 2012; Haque et al., 2012). Individuals aged twenty (20) years and above, representing the unit of inquiry (respondents), were randomly selected from the two communities chosen for the study. The selection of adults became necessary for the study to determine accurately their health seeking behaviour using their first actions taken whenever illness symptoms were perceived (Blanchet et al., 2012; Haque et al., 2012; Danso-Appiah et al., 2010) and how their demographic and socio-economic 66 http://ugspace.ug.edu.gh/ status influenced such actions. Children were excluded in the study due to issues of consent and the fact that decisions regarding children’s health mainly rest upon their parents especially their fathers (Shaikh et al., 2008). Among the factors the study sought to examine its relationship with individuals’ health seeking behaviour was sex due to the dearth of understanding regarding sex and health service utilisation especially in the developing countries (Abubakar et al., 2013; Shaikh et al., 2008). Hence the study included sex to examine the differences and similarities in the health seeking behaviour of both males and females in the selected communities in the Akuapem-North Municipality. A semi-structured questionnaire was used to gather data from individual respondents during the survey. In order to have a fair representation of the selected communities since the urban population (8,615) far out-number that of the rural community (724) (ANDA, 2010), one hundred and seventeen (117) and eighty-eight (88) individual respondents from both the urban (Adukrom) and rural (Korkormu) communities respectively were randomly selected for the study using the Yamane (1967 in Israel, 2013) formulawith a confidence level of 95% and a precision limit of 10%. This corresponds with the urban-rural population distribution variation in the study area. With this, a total of two hundred and five (205) individual respondents were sampled for the study. The semi-structured questionnaire was strategically designed in such a way that respondents answer varying questions regarding their socio-demographic background, their health seeking behaviour and health provider options, utilisation of health services, barriers to health service use, and the mode of health service use financing. The all- inclusive nature of the semi-structured questionnaire made it possible for the study to gather a lot of useful data. The research instrument was pretested in two different 67 http://ugspace.ug.edu.gh/ communities, an urban and a rural community in the municipality. The field assistants were taken through three days of training – due to their tight schedules – to acquaint themselves with the research instrument before the pretesting was done. This gave room for correcting certain errors, rephrasing of some questions and identifying local names for certain terms that needed translation in the local dialect before the actual survey. The actual data collection was done in both English and Twi with the help of four field assistants. The data collection was done within two months, from February to March 2015. Initially, the expectation was to undertake the data collection exercise within a month but certain challenges including what was encountered in Korkormu and data gathering from heads of health service providers made it last for two months. The actual data collection exercise began with a reconnaissance survey of the two selected communities for the study. With the help of the four research assistants and given that the study sites and the enumeration areas have already been selected (see section 3.2.1), the lottery approach of the simple random sampling method was used in selecting the various individual respondents for the study. In order to get a sample frame for the study, various houses within the selected enumerated areas in the study communities were listed systematically for the study. The first house was randomly selected after which each fifth house in Adukrom and third house in Korkormu were respectively selected for the survey. This gave houses from different geographical locations of the communities an equal chance of selection for data collection. With this, a uniform sample of thirty-nine (39) houses in Adukrom and forty-four (44) houses in Korkormu were selected from each enumeration area respectively. In addition, the lottery approach of the simple random sampling method was then used to select the respondents for the study. From this, one hundred and seventeen (117) and eighty-eight 68 http://ugspace.ug.edu.gh/ (88) individuals aged twenty years and above were randomly chosen from both the urban (Adukrom) and rural (Korkormu) communities respectively for the study. After this, respondents were visited from house-to-house in the study communities for adequate and relevant data where the interview method of data collection was used in most cases. In each case questions were structured and read out to respondents in their own local dialect for clarity and understanding. This was necessary to ensure that the data collection process was interactive and participatory, and that respondents understood well the questions to which they were to respond. This was done because majority of the respondents especially those in the rural community (Korkormu) had no formal education while more of the others including those in the urban community (Adukrom), were less educated. Hence a direct face-to-face interaction with them made data collection more effective and reliable. The interview method further helped to avoid call-backs and non- response to questions. Also, it gave the individual respondents who were the unit of inquiry, the opportunity to participate fully in the survey process. Besides, the respondents consented to participate in the study and were enthusiastic with most of the issues influencing their health seeking behaviour. This unveiled certain intriguing themes that called for further probing at the in-depth interview phase. 3.2.3.2 In-depth interviews Most studies of health seeking behaviour have been criticised for focusing only on formal health service utilisation hence interview respondents at out-patients departments (OPD’s) of formal health facilities and administrators of such facilities (Anwar et al., 2012). However, this study interviewed respondents at their respective homes and also engaged 69 http://ugspace.ug.edu.gh/ varying health providers and health institutions in order to ascertain the various ways the operations of various health providers impact and shape the individuals’ decision to engage with particular health providers in the Akuapem-North Municipality. This helped in providing an all-inclusive insight into the complex study of health seeking behaviour (Ghosh et al., 2013). Health service providers belonging to both the formal and informal medical systems were selected for the interview. In all these, an interview was conducted with the heads, and where the head was unavailable, a representative was appointed to be interviewed. The interview guide were in two forms; one for the various individual respondents selected for the qualitative aspect of the study and the other for the heads of health service providers selected for the study. The institutional interview guide mainly looked at issues regarding type and nature of illness for which treatment was sought, time of illness patients visited their facility for treatment, severity of illness, mode of health service financing, service coverage and quality of service rendered and was conducted through in-depth interview with the heads of the health service providers. This was done to identify certain health service and institutional factors that impact individuals’ decision to seek treatment at a particular health service provider in the municipality. On the part of the respondents, the interview guide mainly looked at common health problems faced, availability of health service providers, level of knowledge about their health problems, mode of health service financing, barriers to treatment seeking, and quality of health service rendered. With this, the study conducted twenty in-depth interviews with the various participants in order to obtain detailed information regarding health seeking behaviour in the Akuapem-North Municipality. 70 http://ugspace.ug.edu.gh/ 3.2.4 Secondary data sources The secondary data for the study were obtained from books, journals, articles, periodicals, Akuapem-North Municipal Assembly (ANMA), Akuapem-North Municipal Health Directorate (ANMHD), Ghana Health Service (GHS), and various health providers selected for the study. Data mainly regarding the nature of the health system in Ghana and the municipality were sourced on-line from the Ghana Statistical Services (GSS) and Ghana Health Service (GHS) database. Again, other relevant information was obtained from written publications, books via the internet. 3.2.5 Data analysis and presentation The Statistical Package for Service Solution (SPSS) version 16.0 was used to analyse the quantitative data gathered from the field for the study. The data collected was manually coded and entered into the SPSS software where frequency tables and percentage charts were generated to help in the analysis and discussion of especially key variables for the study. Again, Pearson’s Chi-square tests were used to test levels of association or variations among the variables while multinomial logistic regression was used in predicting pattern of illness severity and health seeking behaviour. Lastly, the qualitative data gathered were analysed manually by categorising them into themes. 3.2.6 Challenges encountered in the fieldwork Most of the problems were encountered in Korkomu – the rural community – where the entire community is divided on political front due to the political stand of the two main leaders being the Assemblyman and the traditional ruler (that is, the “Odikro.”). This made 71 http://ugspace.ug.edu.gh/ data collection in the initial stages very difficult. However, both opinion leaders were contacted again when the problem was identified which afterwards, paved the way for a successful data collection exercise. Aside that, most of the rural participants were reluctant in providing information regarding their health because they thought such information was going to be disclosed to the general public for ridiculing. With this, the purpose of the study was again made known to them to the extent of buttressing it with our student’s identity cards since they were unwilling to give any information without first seeing the ID cards though some still resisted with the opinion that permission ought to be sought from their husbands who were nowhere to be found at that time. This, in fact, prolonged the stipulated time for the exercise especially in the rural community since some of the trained field assistants were not students of the University of Ghana. Again, obtaining data from some of the institutions was a major challenge. The “go to the internet” response by some officials of these institutions was disturbing and was even compounded by the aggregated and fragmented nature of the data after a successful on-line surfing. Where the data were fragmented, some portions of it were not available to ascertain patterns and thus enhance trend analysis. 3.3 Summary The chapter discussed the background of the study area likewise the methodology used. The former provides a general overview of the physical, economic and social environment of the study area where the location, size, demographic characteristics, education, health, water and sanitation and major economic activities and their contributions to the economy 72 http://ugspace.ug.edu.gh/ of the study area have been discussed. The latter, on the other hand, discussed the steps generally adopted to systematically solve the research problem. 73 http://ugspace.ug.edu.gh/ CHAPTER FOUR TYPE OF ILLNESS AND INDIVIDUALS’ HEALTH SEEKING BEHAVIOUR 4.0 Introduction Socio-demographic factors have been identified to significantly influence individuals’ choice of health service providers whenever illness symptoms are perceived (El Kahi et al., 2012; Haque et al., 2012; Rahman et al., 2011; Danso-Appiah et al., 2010) hence this chapter of the study examines the distribution of respondents according to their socio- demographic characteristics. These socio-demographic variables include age, sex, level of education, marital status, religion, occupation and average monthly income. The chapter also examines the type of illness symptoms respondents perceived in the study communities, severity of illness and the day of illness treatment was sought. Also, the relationship between respondents’ cause of illness and the severity of their illness and the day of illness treatment was sought is established in this chapter. This was to find out the kind of illness symptoms respondents rated as severe and vice versa and the particular day of their illness they visited health facilities for treatment. Again, to find out whether respondents sought treatment at available health service providers in their communities, the kind of health service providers available in their communities was assessed in the chapter. Lastly, the chapter examined the particular action respondents took when they first perceived symptoms of illness and sought to establish the relationship between their socio- demographic variables and their health seeking behaviour. 74 http://ugspace.ug.edu.gh/ 4.1 Socio-demographic characteristics of respondents Table 4.1.1 shows the distribution of respondents according to selected socio-demographic variables, namely sex, age, level of education, religion, marital status, occupation and average monthly income. Contrary to the 2010 Population and Housing Census report for the municipality (GSS, 2012), males (57.6%) in general, dominated the study albeit female (52.3%) dominance in the rural area. The study projects that males more than females would engage with formal than informal medical providers due to their dominance in terms of economic resources (Atuyambe et al., 2009; Shaihk et al., 2008) as is found by the current study. From Table 4.1.1, such categorisation regarding respondents’ age became necessary for the study when it was realised after the data had been collected that the age range was too wide should it be categorised into an interval of five. This would have increased the age groupings to about ten (10) which would have made analysis a bit difficult since there would have been more divisions under age alone to deal with. With this, the 30-39 years age group (31.7%), in general, predominated among all the age groupings followed closely by the 20-29 years age group (27.8%) with the least (9.8%) being the 60 years and above age category. Among the urbanites, the 30-39 years age category (32.5%) dominated whereas the 20-29 years age group (36.4%) became paramount among the rural dwellers. The study foresees the young (below fifty years) than the aged (above fifty years) to engage the more with formal medical facilities since they have the energy to work and especially for longer periods of hours. Majority (87.8%) of the sampled population had received some form of formal education while 12.2% had received no formal education at the time data was collected. This is expected to affect health seeking behaviour in the municipality positively as a direct 75 http://ugspace.ug.edu.gh/ association between such a high educational background and treatment seeking from formal medical facilities is anticipated. In general, the urbanites had high level of education relative to their rural counterparts as more of the urbanites had Secondary and Tertiary education than their rural counterparts (see Table 4.1.1). In the same way urbanites have had more years of schooling than their rural counterparts. For urbanites, 57.2% and 42.8% have schooled for above and below ten (10) years compared with 17.1% and 82.9% of their rural colleagues respectively. Also, 90.2% of the sampled population had dependants with those having and living with three children (30.2%) predominating the study. Hence urbanites compared with their rural counterparts, are likely to seek treatment the more from formal than informal medical providers. Christians (60%) were the predominant religious group. This was followed by Muslims (32.7%) with the least being Traditionalists (1.5%). This is reflective of the general religious trend in the municipality (ANDA, 2010). With this, 62.9% said they attend worship services very often while as low as 12.2% do not attend worship services at all. Information regarding the ethnicity of respondents was gathered and Ewes (33.7%) were found to predominate followed very closely by Guans (32.2%) and distantly by the Ga Adangmes (2.0%). Data on the employment status of respondents was elicited and from Table 4.1.2, slightly more of the rural dwellers (94.3%) than the urbanites (93.2%) were gainfully employed at the time data was gathered. Majority (49.4%) of the rural employed were Farmers with Civil Servants (2.4%) forming the least while more of their urban counterparts were Traders (29.4%) with the least (5.5%) being Drivers as at the period data was collected. Respondents in the study area are expected to seek treatment the more from formal than 76 http://ugspace.ug.edu.gh/ informal medical facilities due to the very high employment status (Kakai et al., 2009). However, there is the possibility that treatment seeking at informal medical providers will dominate due to the type and nature of their occupations (Sudharsanam & Rotti, 2007; Chakraborty et al., 2003) due to its effects on income levels and the ability to access and utilise health services (Kakai et al., 2009; Ahmed et al., 2005). The study found the urbanites to be far better-off compared with their rural counterparts. Majority (43.6%) of the urbanites, as at the time data was collected, earned, on the average, between GHȼ201.00-GHȼ500.00 while the least (6.0%) earning averagely, below GHȼ100.00. The rural inhabitants, on the other hand, had the majority (63.7%) of them earning below GHȼ100.00 on the average with the least (3.4%) that earned above GHȼ500.00 averagely every month (see Table 4.1.2). Relatively, respondents’ average monthly income levels is low hence it is expected to translate into their health seeking behaviour where treatment seeking from informal medical providers dominate. On the other hand, urbanites compared to their rural counterparts are expected to engage the more with formal medical facilities than informal medical providers due to their relative high income levels. Interestingly, more of the rural respondents (39.8%) than the urbanites (16.2%) reported being very satisfied with their average monthly incomes. To help identify the income levels of respondents in the selected communities, data on the amount of money they spend in a day including their dependent was elicited. With this, majority of the urbanites (50.4%) were found spending GHȼ11.00-GHȼ20.00 daily while for their rural counterparts, the majority (59.1%) spent between GHȼ6.00-GHȼ10.00 daily including their dependents. 77 http://ugspace.ug.edu.gh/ Table 4.1.1: Distribution of respondents by age, sex, level of education and religion Socio-demographic variables Place of residence Total Urban Rural N % N % N % Age 20-29 years 25 21.4 32 36.4 57 27.8 30-39 years 38 32.4 27 30.7 65 31.7 40-49 years 25 21.4 15 17.0 40 19.5 50-59 years 17 14.5 6 6.8 23 11.2 60 years and above 12 10.3 8 9.1 20 9.8 Total 117 100.0 88 100.0 205 100.0 Sex Male 76 65.0 42 47.7 118 57.6 Female 41 35.0 46 52.3 87 42.4 Total 117 100.0 88 100.0 205 100.0 Level of Education Neve schooled 5 4.3 20 22.7 25 12.2 Primary/JHS 43 36.8 55 62.5 98 47.8 Secondary 46 39.3 10 11.4 56 27.3 Tertiary 23 19.7 3 3.4 26 12.7 Total 117 100.0 88 100.0 205 100.0 Religion None 9 7.7 3 3.4 12 5.9 78 http://ugspace.ug.edu.gh/ Christian 101 86.3 22 25.0 123 60.0 Muslim 6 5.1 61 69.3 67 32.7 Traditional 1 0.9 2 2.3 3 1.5 Total 205 100.0 88 100.0 205 100.0 Source: Field Work, 2015. 79 http://ugspace.ug.edu.gh/ Table 4.1.2: Distribution of respondents by marital status, occupation and average monthly income Socio-demographic variables Place of residence Total Urban Rural N % N % N % Marital status Single 22 18.8 8 9.1 30 14.6 Married 65 55.6 51 58.0 116 56.6 Divorced 5 4.3 1 1.1 6 2.9 Widowed 8 6.8 5 5.7 13 6.3 Cohabiting 17 14.5 23 26.1 40 19.5 Total 117 100.0 88 100.0 205 100.0 Occupation Farmer 6 5.1 41 46.6 47 22.9 Trader 32 27.4 31 35.2 63 30.7 Artisan 19 16.2 3 3.4 22 10.7 Civil Servant 14 12.0 2 2.3 16 7.8 Business 16 13.7 6 6.8 22 10.7 Casual worker 8 6.8 0 0.0 8 3.9 Driver 14 12.0 0 0.0 14 6.8 None 8 6.8 5 5.7 13 6.3 Total 117 100.0 88 100.0 205 100.0 Average monthly income 80 http://ugspace.ug.edu.gh/ Below GHȼ50 5 4.3 32 36.4 37 18.0 GHȼ50-GHȼ100 2 1.7 24 27.3 26 12.7 GHȼ100-GHȼ200 14 12.0 25 28.4 39 19.0 GHȼ201-GHȼ500 51 43.6 4 4.5 55 26.8 GHȼ501-GHȼ1000 32 27.4 3 3.4 35 17.1 GHȼ1000 and above 13 11.1 0 0.0 13 6.3 Total 117 100.0 88 100.0 205 100.0 Source: Field Work, 2015. 4.2 Major type of illness and source of treatment In order to find out the type of perceived illness symptoms respondents sought treatment for they were initially asked whether they had ever fallen ill in the past twelve months as was the case of Blanchet et al. (2012) and El Kahi et al. (2012). Two hundred and two (202) respondents representing 98.5% indicated they had perceived illness symptoms, in one way or the other, in the twelve-month period. These were further asked to indicate the major type of illness symptoms they perceived and unlike Robyn et al. (2012) in Burkina Faso; Gotsadze et al. (2005) in Tbilisi; Uzochukwu & Onwujekwe (2004) in Nigeria and Yanagisawa et al. (2004) in Cambodia where self-reported malaria or fever was the most commonly reported health problem, general body pains/headache (33.7%) was by far the predominant illness perceived by respondents in both the urban and rural communities. This is followed by malaria/fever (23.9%) with the least (1.4%) indicating acute respiratory infections (ARIs) (see Table 4.2.1). This reverberates with the findings of Fomundam et al. (2012) in Africa and Subba (2004) in Nepal where majority of the studied population cited headaches, and body pains and weaknesses as the main reported perceived illness. 81 http://ugspace.ug.edu.gh/ Thus, whereas more of the urbanites (40.2%) named general body pains/headache as the main perceived illness symptoms, majority (30.8%) of their rural counterparts claimed they did not know of the cause of their illness (see Table 4.2.1). Table 4.2.1: Respondents’ major type of illness and source of treatment Place of residence Variables Urban Rural Total Major type of illness N % N % N % Don’t know 21 17.9 27 30.8 48 23.4 Malaria/Fever 28 23.9 21 23.9 49 23.9 Stomach/abdominal pains 17 14.6 15 17.0 32 15.6 General body pains/headache 47 40.2 22 25.0 69 33.7 Spiritual illness 2 1.7 2 2.2 4 2.0 Acute Respiratory Infections (ARIs) 2 1.7 1 1.1 3 1.4 Total 117 100.0 88 100.0 205 100.0 Source of treatment Self-diagnosis 65 55.6 37 42.0 103 50.3 Friend/Neighbour/Relative 1 0.9 2 2.3 3 1.5 Person in drug/chemical shop 18 15.4 7 8.0 25 12.2 Professional medical personnel 31 26.5 27 30.7 58 28.3 Traditional healer 2 1.6 15 17.0 16 7.8 Total 117 100.0 88 100.0 205 100.0 Source: Field Work, 2015. 82 http://ugspace.ug.edu.gh/ On how they got to know of their illness, half of the respondents in general said they did so themselves followed by a substantial proportion (28.3%) that said a professional personnel at formal health facilities diagnosed them (see Table 4.2.1). This echoes the findings of Cropley (2004) and Uzochukwu & Onwujekwe (2004) where self-diagnosis and diagnosis by professional medical personnel dominated. Studies (Dominic et al., 2013; Khabiri et al., 2011; Mushtaq et al., 2011; Siddiqui et al., 2011) conducted in other parts of the world likewise in Ghana (Annan et al., 2013; Dako-Gyeke et al., 2013; Laar et al., 2013) report of more than half of the studied population taking some form of action whenever illness symptoms were perceived relative to those that do not take any action at all. This induced some patients to go for treatment at formal health providers like hospitals, clinics, health centres and CHPS compounds while others resorted to treatment seeking at informal health providers such as drug stores, traditional healers, and drug peddlers. With this, the study found whether respondents took any action or not upon perceiving symptoms of illness and it was found that all the respondents, in one way or the other, did something to help them feel better, either to reduce the pain or wholly cure the illness. The study further found out the type of illness symptoms perceived by respondents and how these were diagnosed. From Table 4.2.2, it was identified generally that half of the respondents that said they did not know the cause of their illness got diagnosed by professional medical personnel followed distantly by those that got diagnosed by traditional healer (27.1%). In the same way, respondents that reported suffering from malaria/fever got diagnosed by a professional medical personnel (42.9%) followed by those that diagnosed themselves (40.8%) confirming the findings of Gotsadze et al. (2005) and Uzochukwu & Onwujekwe (2004). Also, majority (62.5%) of the respondents that 83 http://ugspace.ug.edu.gh/ reported suffering from stomach/abdominal pains got diagnosed by themselves likewise those that suffered from general body pains/headache (82.6%). Again, majority (75.0%) of the participants that reported the type of their illness to be spiritual were diagnosed by a traditional healer whereas all those that claimed suffering from acute respiratory infections (ARIs) got diagnosed by professional medical personnel (see Table 4.2.2). It is worth noting that none of the respondents that suffered from acute respiratory infections (ARIs) got diagnosed by themselves, a relative, traditional healer or any other person. This might be due to the severity of such illness (see Table 4.2.2). Table 4.2.2: Respondents’ major type of illness by source of treatment Source of treatment Type of Self- Friend Drug Qualified Traditiona Total illness diagnosis or store personn l healer Relati person el ve nel Don’t know 3(6.3%) 1(2.1%) 7(14.6%) 24(50.0%) 13(27.1%) 48(100%) Malaria 20(40.8%) 2(4.1%) 6(12.2%) 21(42.9%) 0(0.0%) 49(100%) Stomach pains 20(62.5%) 0(0.0%) 4(12.5%) 7(21.9%) 1(3.1%) 32(100%) Headache 57(82.6%) 0(0.0%) 8(11.6%) 4(5.8%) 0(0.0%) 69(100%) Spiritual 1(25.0%) 0(0.0%) 0(0.0%) 0(0.0%) 3(75.0%) 4(100%) ARIs 1(33.3%) 0(0.0%) 0(0.0%) 2(66.7%) 0(0.0%) 3(100%) Total 102(49.8%) 3(1.5%) 25(12.2%) 58(28.3%) 17(8.3%) 205(100%) Chi-square (χ2) = 1.300E2, df (36), p = 0.000 Source: Field Work, 2015. (Figures in bracket represent percentages in a row). 84 http://ugspace.ug.edu.gh/ 4.3 Severity of illness and when reported for treatment The severity or seriousness of individuals’ illness has been found to determine their choice and use of health services at a particular health provider (Annan et al., 2013; Ghosh et al., 2013; Haque et al., 2012; Rahman et al., 2012). Patients with severe and more severe illness symptoms have been found to seek prompt care and especially at formal medical facilities whiles those with less severe and non-severe illness symptoms mainly visit informal medical healers for treatment (Annan et al., 2013; Rahman et al., 2013; Oluwatuyi, 2010; Taffa & Chepngeno, 2005). The severity of individuals’ illness is also found to determine when patients reported for treatment at health service providers (Rahman et al., 2012; Biswas et al., 2006; Taffa & Chepngeno, 2005). Information regarding the severity of respondents’ illness was elicited and the result is shown in Table 4.3.1. Respondents severity of illness was measured on a four-point Likert scale: ‘not severe,’ ‘less severe,’ ‘severe,’ and ‘more severe’ and did indicate the intensity and/or seriousness of their illness themselves by selecting one of the above. In general, majority (34.2%) of the respondents said it was severe. Also, 28.3% of the respondents in both communities were of the view that their illness were less severe whereas 19%, and 18.5% reported of their illness to be more severe and not severe correspondingly (see Table 4.3.1). With this, an equal proportion (28.4%) of the rural respondents concurrently reported of their illness to be severe and more severe alike while 27.3% and 15.9% (being the least) said they perceived the symptoms of their illness to be less severe and not severe in that order. For the urbanites, majority (38.5%) reported of their illness to be severe followed by those that reported of their illness to be less severe (29.1%) with the least (12.0%) reporting of their 85 http://ugspace.ug.edu.gh/ illness to be more severe (see Table 4.3.1). This was statistically significant in both the urban and rural communities of the study (χ2 = 14.102, df (3), p = 0.003). Table 4.3.1 Severity of illness and place of residence Place of Residence Urban Rural Total Severity of Illness N % N % N % Not Severe 24 20.5 14 15.9 38 13.7 Less Severe 34 29.1 24 27.3 58 28.2 Severe 45 38.4 25 28.4 70 34.1 More Severe 14 12.0 25 28.4 39 19.0 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 14.102, df (3), p = 0.003 When reported for treatment First day of illness 16 13.7 10 11.4 26 12.7 Second day of illness 8 6.8 17 19.3 25 12.2 Third day of illness 19 16.3 24 27.3 43 21.0 Fourth day of illness 50 42.7 21 23.8 71 34.6 Fifth day of illness and above 24 20.5 16 18.2 40 19.5 Total 117 100.0 88 100.0 205 100.0 Chi-square (χ2) = 33.805, df (4), p = 0.000 Source: Field Work, 2015. 86 http://ugspace.ug.edu.gh/ Also, the number of days it took respondents to go for treatment at a particular health provider after symptoms of illness were perceived was solicited. Unlike in India where patients waited for over ten days before some treatment was sought (Sudha et al., 2003), the study found majority (34.6%) of the respondents, in general, to wait for four days before going for treatment at health service providers. This is followed by those that went for care on the third day (21.0%) of illness identification while the least (12.2%) claimed going for treatment on the second day illness symptoms were perceived (see Table 4.3.1). This also is in consonance with the findings of Rahman et al. (2012) and Khabiri et al. (2011) where patients sought care on the first three days of illness identification. When asked concerning their reasons for seeking treatment at such days of their illness majority of both the urban (45.3%) and rural (47.7%) respondents said their illnesses were not severe. Other reasons given were ‘illness was very severe’ (15.4% and 22.7%) and ‘wait to see if I will be fine’ (38.5% and 18.2%) by both the urban and rural respondents correspondingly. Regarding respondents’ major response for seeking treatment at the time they did, some participants in the in-depth interview had this to say: “Sometimes you are so pressed by work that before you realise the health centre likewise the drug/chemical shops here have all closed. And the intensity of the illness the next morning might not be as serious as earlier so you take it to work and you know the nature of our work; you have to set off very early in the day else you will not get much” (A 31 year old driver at Adukrom, Individual Interview). “It has never been my wish to go for treatment at such times of my illness but the thing is; when the money to purchase health service is not available you take a hot 87 http://ugspace.ug.edu.gh/ bath in the night and go to sleep hoping to wake up strong the next day” (A 29 year old farmer at Korkormu, Individual Interview). The day of illness respondents reported for treatment was significant (χ2 = 33.805, df (4), p = 0.000) in both communities according to the Chi-square goodness-of-fit test. For a clearer explanation and understanding regarding why respondents went for treatment from health service providers the particular days they did, the study sought to establish the relationship between when they reported for treatment and the severity of their illness. The result is illustrated by Table 4.3.2. Table 4.3.2: Respondents’ severity of illness and when reported for treatment Severity of illness Not severe Less severe Severe More severe Total When reported for treatment 1st day 3(4.5%) 2(7.7%) 13(50.0%) 8(30.8%) 26(100.0%) 2nd day 2(8.0%) 8(32.0%) 7(28.0%) 8(32.0%) 25(100.0%) 3rd day 7(16.3%) 12(27.9%) 19(44.2%) 5(11.6%) 43(100.0%) 4th day 14(19.7%) 23(32.4%) 21(29.6%) 13(18.2%) 71(100.0%) 5th day & above 12(30.0%) 13(32.5%) 10(25.0%) 5(12.5%) 40(100.0%) Total 38(18.5%) 58(28.3%) 70(34.1%) 39(19.0%) 205(100.0%) Chi-square (χ2) = 21.010 df (12) Sig.=0.000 Source: Field Work, 2015. (Figures in bracket represent percentages in a row). Studies (Rahman et al., 2012; Sreeramareddy et al., 2006) conducted elsewhere show that patients seek treatment from health service providers the early days of symptoms 88 http://ugspace.ug.edu.gh/ identification when such illness symptoms are perceived to be more severe and/or severe while treatment is delayed especially where identified symptoms are perceived not to be severe and/or less severe. From Table 4.3.2, exactly half of the respondents (being the majority) that sought treatment from health service providers the first day of symptoms identification had their illness to be severe. This is followed by those who perceived their illness to be more severe (30.8%) while the least (7.7%) had their illness symptoms to be less severe. With regard to respondents that reported at health services providers on the second day of symptoms identification, majority (32.0%) perceived their illness to be more severe and less severe concurrently. Respondents that had their illness to be severe followed with the least (8.0%) perceiving their illness symptoms no to be severe (see Table 4.3.2). Also, majority (44.2%) of the respondents that reported for treatment on the third day of their symptoms identification had their illness to be severe followed by those that perceived their illness to be less severe (27.9%). The least (11.6%) perceived their illness symptoms to be more severe (see Table 4.3.2). Again from Table 4.3.2, majority (4.3.2, majority (32.4%) of the respondents that reported at health facilities for treatment on the fourth day of symptoms identification perceived their illness to be less severe. This is followed by 29.6% that perceived their illness to be severe while the least (18.3%) had their illness symptoms to be more severe. Lastly, for respondents that reported for treatment after the fifth day of symptoms identification, majority (32.5%) of them had their illness symptoms to be less severe followed by those that perceived their illness symptoms not to be severe (30.0%) with the least (12.5%) who perceived their illness symptoms to be more severe. The above picture is in consonance with the findings of Rahman et al. (2012) and Sreeramareddy et al. (2006) where patients 89 http://ugspace.ug.edu.gh/ reported at health facilities earlier for severe and more severe illness but for non-severe and less severe illness, treatment was delayed. From the Chi-square test of association, a significant (χ2 = 21.010, df (12), p = 0.000) relationship was established between the day respondents reported for at health service providers for treatment and the severity of their illness. The study also sought to establish the relationship between the kind of illness symptoms respondents perceived and the severity of such illness. This was done to examine the type of illness symptoms was perceived to be more severe, severe, less severe and not severe. The results are shown in Table 4.3.3. From Table 4.3.3, majority (43.8%) of the respondents who were of the view that they knew not of their illness had their illness symptoms to be more severe. This is followed distantly by those who perceived their illness symptoms to be severe (27.1%) while those that perceived their illness symptoms to be less severe (14.6%) and not severe (14.6%) simultaneously became the least. For respondents who suffered from malaria/fever, 46.9% was found to be severe, 22.5% was identified as not severe while the least (12.2%) was acknowledged to be more severe (see Table 4.3.2) confirming the findings of Rahman et al. (2012), Taffa & Chepngeno (2005) and Uzochukwu & Onwujekwe (2004). Also, the study found respondents who suffered from stomach/abdominal pains to have the majority (40.6%) of them being severe followed by those who had theirs to be more severe (28.1%). The least (6.3%) had their perceived stomach/abdominal pains not to be severe (see Table 4.3.3). 90 http://ugspace.ug.edu.gh/ Table 4.3.3: Respondents’ major type of illness by severity of illness Severity of illness Type of illness Not severe Less severe Severe More severe Total Don’t know 7(14.6%) 7(14.6%) 13(27.1%) 21(43.8%) 48(100.0%) Malaria/Fever 11(22.5%) 9(18.4%) 23(46.9%) 6(12.2%) 49(100.0%) Stomach pains 2(6.3%) 8(25.0%) 13(40.6%) 9(28.1%) 32(100.0%) General body 18(26.1%) 34(49.3%) 16(23.2%) 1(1.4%) 69(100.0%) Spiritual 0(0.0%) 0(0.0%) 2(50.0%) 2(50.0%) 4(100.0%) ARIs 0(0.0%) 0(0.0%) 3(100.0%) 0(0.0%) 3(100.0%) Total 38(18.5%) 58(28.3%) 70(34.1%) 39(19.0%) 205(100.0%) Urban: χ2=61.537 df (18) Sig.=0.000 Source: Field Work, 2015. (Figures in bracket represent percentages in a row). Respondents who suffered from general body pains/headache had the majority (49.3%) to be less severe followed by those that had theirs not to be severe (26.1%) with the least (1.4%) claiming theirs was more severe. Again, none of the respondents who perceived their illness to be ‘spiritual’ had them to be less severe and not severe. However, an equal proportion (50.0%) of the respondents who suffered from spiritual illnesses perceived them to be both severe and more severe respectively. Unsurprisingly, all the respondents that suffered from acute respiratory infections (ARIs) perceived them to be severe (see Table 4.3.3). The study established a significant relationship between respondents’ major type of illness perceived and the severity of their illness in both communities according to the Chi- square test of association (χ2=61.537, df (18), p=0.000). 91 http://ugspace.ug.edu.gh/ Given that the type of illness suffered by patients determines where and when they go for treatment (Rahman et al., 2012; Taffa & Chepngeno, 2005; Uzochukwu & Onwujekwe, 2004), the study sought to establish the relationship between the type of illness respondents’ reported and the day of illness treatment was sought. The results are shown in Table 4.3.4. Table 4.3.4: Major type of illness by when reported for treatment When reported for treatment Type of 1st day 2nd day 3rd day 4th day 5th day & Total illness above Don’t know 8(16.7%) 10(20.8%) 11(22.9%) 13(27.1%) 6(12.5%) 48(100%) Malaria 9(18.4%) 7(14.3%) 9(18.4%) 14(28.6%) 10(20.4%) 49(100%) Stomach pain 5(15.6%) 5(15.6%) 7(21.9%) 10(31.3%) 5(15.6%) 32(100%) Headache 4(5.8%) 4(5.8%) 13(18.8%) 31(44.9%) 17(24.6%) 69(100%) Spiritual 0(0.0%) 0(0.0%) 2(50.0%) 1(25.0%) 1(25.0%) 4(100%) ARIs 0(0.0%) 0(0.0%) 2(66.7%) 0(0.0%) 1(33.3%) 3(100%) Total 26(12.7%) 26(12.7%) 44(21.5%) 69(33.7%) 40(19.5%) 205(100%) Chi-square (χ2) = 31.058 df (24) Sig.=0.001 Source: Field Work, 2015. (Figures in bracket represent percentages in a row). From Table 4.3.4, majority of the respondents who had no idea regarding the type of illness they suffered from (60.4%), those who perceived their illness to be malaria/fever (51.1%), those who suffered from stomach/abdominal pains (53.1%) and those who suffered from acute respiratory infections (66.7%) sought treatment at health service providers within the first three days of symptoms identification whereas the remaining proportions waited for 92 http://ugspace.ug.edu.gh/ four to five days and over before seeking some form of care. This echoes the findings of Rahman et al. (2012) and Taffa & Chepngeno (2005) where patients waited for two to three days before treatment was sought for illness types such as fever and stomach problems. For respondents who suffered from general body pains/headache, the majority (75.9%) waited till four to five days and over before going for care at health service providers (see Table 4.3.4). This may be attributed to the non-severity of such illnesses which has been echoed by Haque et al. (2012), Rahman et al. (2012) and Mustaq et al. (2011). Again, an equal proportion (50.0%) each of the respondents that suffered from spiritual illnesses waited for three days and over four days before seeking treatment (see Table 4.3.4). The longer days it took patients before seeking care may be due to the fact that they wanted to be sure about the cause of their illnesses so that treatment be sought at the appropriate health provider. A significant (χ2 = 31.058, df (24), p = 0.001) relationship was established between respondents’ major type of illness perceived and the time they sought treatment. In Nepal, Subba (2004) established that treatment was sought at a particular health service provider based on the type of illness suffered by respondents. Patients that suffered from headaches, body aches and weakness mainly sought treatment at informal health providers (self-medication and alternative medication) whereas those that suffered from ARIs, Gastritis, Asthma, Tuberculosis, Arthritis, Fever and Diarrhoea mainly sought care at formal medical facilities (modern medication) (Subba, 2004). Based on this, the study attempted to establish the relationship between the type of illnesses respondents perceived and the particular action they took to ascertain the specific health service provider at which treatment was sought for which illness type. The results are presented in Table 4.3.5. 93 http://ugspace.ug.edu.gh/ Table 4.3.5: Health facility attended when ill by major type of illness Health facility attended when ill IMPs FMPs Type of illness Urban Rural Urban Rural Total Don’t know 7(9.2%) 17(28.8%) 14(34.1%) 10(34.4%) 48(23.4%) Malaria/Fever 13(17.1%) 12(20.3%) 15(36.6%) 9(31.0%) 49(23.9%) Stomach pains 12(15.8%) 7(11.9%) 5(12.2%) 8(27.6%) 32(15.6%) General body pains 42(55.3%) 21(35.6%) 5(12.2%) 1(3.5%) 69(33.7%) Spiritual 2(2.6%) 2(3.4%) 0(0.0%) 0(0.0%) 4(2.0%) ARIs 0(0.0%) 0(0.0%) 2(4.9%) 1(3.5%) 3(1.5%) Total 76(100%) 59(100%) 41(100%) 29(100%) 205(100%) Chi-Square (χ2) = 38.492 df (6) Sig. = 0.000 Source: Field Work, 2015. (Figures in bracket represent percentages in a column). From Table 4.3.5, majority (55.3%) of the urbanites that sought treatment at informal medical providers suffered from general body pains/headache confirming the findings of Subba (2004). This is followed by those that suffered from malaria/fever (17.1%) with the least (2.6%) who suffered from all forms of spiritual illnesses. On the part of rural dwellers that sought care at informal medical providers as first action taken when ill, 35.6% suffered from general body pains/headache also buttressing the results of Subba (2004), followed by those who did not know of the cause of their illness (28.8%). The least (3.4%) suffered from spiritual illness (see Table 4.3.5). 94 http://ugspace.ug.edu.gh/ On the part of treatment seeking at formal medical facilities, majority (36.6%) of the urbanites suffered from malaria/fever confirming the findings of Subba (2004). This is followed closely by those that had no idea regarding the type of illness they suffered from (34.1%), while the least (4.9%) claimed they suffered from acute respiratory infections (ARIs) (see Table 4.3.5). For their rural counterparts, those that said they did not know of their illnesses (34.4%) dominated. This is followed closely by those that reported suffering from malaria/fever (31.0%) as reported by Subba (2004), while an equal proportion (3.5%) concurrently that claimed suffered from general body pains/headache and acute respiratory infections (ARIs) formed the least. Whereas none of the respondents that suffered from acute respiratory infections (ARIs) visited an informal medical provider for treatment in both communities resonating the findings of Subba (2004), none of those who claimed they suffered from all forms of spiritual illnesses in both communities sought treatment at formal medical providers (see Table 4.3.5). This may be attributed to the severity of both spiritual illness and ARIs as was found by the study, the belief and knowledge respondents hold and have concerning the aetiology of such illness, hence need further probing. According to the Chi-square test of association (χ2 = 38.492, df (6), p = 0.000), a significant relationship was established between the health facility respondents attended when the perceived symptoms of illness and the major type of illness they suffered from. 4.4 Availability of health providers The study solicited for information regarding the forms of health providers available in respondents’ communities and it was identified that both formal and informal health service providers exist in the urban community while in the rural community, only informal health providers are available. Adukrom, the urban community for the study, boasts of 95 http://ugspace.ug.edu.gh/ having a health centre, drug/chemical shops, various forms of traditional healers (traditional birth attendants (TBAs), bone setters, herbalists, spiritualists and fetish priests), drug peddlers and a prayer camp. The rural community (Korkormu), on the other hand, is full of differing kinds of traditional healers. Globally treatment seeking at health service providers available and closest to patients house has been found to dominate irrespective of place of residence for general illness (Aboagye & Agyemang, 2013; Annan et al., 2013; Dominic et al., 2013) except where special services or treatment is required (Nahar, 2010; Mazumda et al., 2009; Russell, 2008). The study confirmed this finding as majority of both the urbanites (96.6%) and rural dwellers (81.8%) admitted seeking treatment at health service providers that are available in their respective communities (see Table 4.4.1). The Chi-square goodness-of-fit test conducted (χ2 = 185.488, df (1), p < 0.05) showed that treatment seeking at health service providers available in both the urban and rural communities selected for the study was significant (see Table 4.4.1). This is in consonance with the findings of Annan et al. (2013) and Dominic et al. (2013). This implies that the availability of a health service provider influences respondents’ health seeking behaviour in the Akuapem-North Municipality. With this, the alternate hypothesis which states that the availability of a health provider will significantly influence their utilisation in both the rural and urban communities is accepted while the null hypothesis which states otherwise is rejected. 96 http://ugspace.ug.edu.gh/ Table 4.4.1: Available health providers respondents sought treatment from Place of residence Urban Rural Total N % N % N % Yes 113 96.6 72 81.8 185 90.2 No 4 3.4 16 18.2 20 9.8 Total 117 100.0 88 100.0 205 100.0 Chi-square (χ2) = 185.488, df (1), Sig. = 0.000 Source: Field Work, 2015. Also, from Table 4.4.1, 9.8% of the respondents said they go for treatment at health service providers elsewhere. It was identified in the in-depth interview that reasons such as illness needing special treatment, health insurance, quality of service rendered, severity of illness, attitude of provider staff and nature of illness as echoed by Dako-Gyeke et al. (2013), Russell (2008) and Asenso-Okyere et al. (1998) made them seek treatment at such providers that are not available in their communities. 4.5 Individuals’ health seeking behaviour in the municipality The study by Roy et al. (2004) in the United States of America found a considerable proportion (40%) of minority groups treating their sick children with home remedies with the belief that some home remedies “are better for curing child illness than prescribed medicines from qualified medical providers.” Also, Mashreky et al. (2010) report of about 60% of Bangladeshi parents seeking treatment at unqualified service providers for their burned children as first action taken in burn-injury management. Of this figure, 41% took 97 http://ugspace.ug.edu.gh/ their wards to medicine shopkeepers for treatment. In the same way, Yanagisawa et al. (2004) found the rich (71.7%) than the poor (63.6%) in Cambodia to apply home remedies as first actions taken when symptoms of illness were perceived. Georgians, according to Gotsadze et al. (2005) mainly self-treat (65.5%) as first action taken when ill likewise rural Bangladeshis (Ahmed et al 2005). Also in India, though a higher proportion (80%) of pregnant women sought care at formal medical facilities, substantial proportions (65% and 61%) of those with and without access to formal medical facilities respectively went for treatment at informal medical providers (mainly TBAs). But for West Bengali mothers, the majority (73.8% and 72.4%) sought treatment at formal medical providers for their sick male and female children respectively. It is however not surprising that self-treatment (an informal means of care seeking) predominates in Africa (Fomundam et al., 2012; Taffa & Chepngeno, 2005; Awusabo-Asare & Anarfi, 1997). Given these empirical evidence, this study ascertained participants’ response to perceived illness symptoms by asking them to state the health service provider/facility they attended whenever they perceived illness symptoms since it was used as the measure for individuals’ health seeking behaviour for the study. From Table 4.5.1, majority of both the urban (65%) and rural (67%) dwellers reported visiting informal medical providers for care. This shows that individuals in the Akuapem-North Municipality, irrespective of their place of residence, mainly seek treatment at informal medical providers as first action taken whenever illness symptoms are perceived buttressing the findings of Fomundam et al. (2012), Mashreky et al. (2010), Gotsadze et al. (2005) and Yanagisawa et al. (2004) where more than half of the sampled populations understudy engaged with informal medical providers for treatment. The health facility respondents attended after perceived illness 98 http://ugspace.ug.edu.gh/ symptoms was found to be statistically significant (χ2 = 20.610, df (1), p < 0.05) in both the urban and rural communities selected for the study per the Chi-square goodness-of-fit test (see Table 4.5.1). This resonates with the findings of El Kahi et al. (2012), Rai et al. (2012) and Mushtaq et al. (2011) where a significant association was established between place of residence and health seeking at a particular health service provider whenever illness symptoms were perceived. This pattern of health seeking behaviour did not happen by chance. Thus, no difference exists regarding the health seeking behaviour of people in the Akuapem-North Municipality irrespective of their place of residence. Therefore, the alternate hypothesis is rejected whereas the null hypothesis which assumes that no significant difference exist in the health facility rural and urban dwellers attended whenever illness symptoms were perceived is accepted. Table 4.5.1: Health facility attended when ill Place of Residence Urban Rural Total Health facility attended when Ill N % N % N % Treatment seeking from IMPs 76 65.0 59 67.0 135 65.9 Treatment seeking from FMPs 41 35.0 29 33.0 70 34.1 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 20.610 df (1) Sig.= 0.000 Source: Field Work, 2015. With regard to the specific actions respondents in the selected communities took when illness symptoms were perceived, majority (43.4%) of them said they bought medicines 99 http://ugspace.ug.edu.gh/ from drug/chemical shops with more of the urbanites (60.7%) than their rural counterparts (20.5%) involved in such action. This buttresses the findings of Fomundam et al. (2012) in Africa, Siddiqui et al. (2011) in Bangladesh and Mashreky et al. (2010) also in Bangladesh where pharmacy and drug store purchases of medicines predominated, but is in antithesis to the findings of Annan et al. (2013) in northern Ghana, Yanagisawa et al. (2004) in Cambodia and Nahar (2010) in Bangladesh where home remedy applications and treatment seeking at traditional healers predominated as participants’ first actions taken when illness symptoms were perceived. This is followed by treatment seeking at a hospital/clinic/health centre (34.1%) also with more of the urban (35.0%) than the rural (33.0%) dwellers taking this action resonating with earlier reports by van der Hoeven et al. (2012) in South Africa, Mazumda et al. (2009) in India and Yanagisawa et al. (2004) in Cambodia. The least (2.9%) bought medicines from drug peddlers as first action taken when ill. The study found none of the urbanites to engage in this form of treatment seeking. 100 http://ugspace.ug.edu.gh/ Table 4.5.2 Respondents’ specific first action taken when ill Place of Residence Urban Rural Total First Action Taken N % N % N % Apply a home remedy 2 1.7 13 14.8 15 7.3 Buy medicine from a drug/chemical shop 71 60.7 18 20.5 89 43.4 Visit a traditional healer for cure 3 2.6 22 25.0 25 12.2 Seek treatment at Hospital/Clinic/HC 41 35.0 29 33.0 70 34.1 Buy medicine from a drug peddler 0 0.0 6 6.8 6 2.9 Total 117 100.0 88 100.0 205 100.0 Source: Field Work, 2015. When asked of their major reason for taking such actions, majority (23.4%) of the respondents were of the view that their illness was not severe. This is followed closely by those that said they always got cured of their illness when such actions were taken (20.5%) with the least (1.5%) saying that the health providers understand their problems and believed their illness were spiritually induced concurrently. This is what some participants had to say in the individual interview: “I don’t remember the last time I went to a hospital or the Health Centre here for treatment though I have enrolled on the NHIS. I always get the medicines I need from the drug/chemical shop at very affordable prices where I don’t have to be insulted by a ‘small girl’ I can give birth to and also wait for longer hours before being attended to. Besides, the most important thing is getting cured which I am always assured of” (A 48-year old man at Adukrom, Individual Interview). 101 http://ugspace.ug.edu.gh/ “As for general body pains and headaches, they are normal after a hard day’s work. So why waste money and time travelling to either Adawso or Kwamoso (nearby villages with health centres) to seek health care. For me I most at times buy pain killers, and balms and ointments especially at Adawso market days because of the nature of my work. So I just have to apply the balm/ointment and take my pain killers and the next day I am more than strong for a new day’s work” (A 40-year old farmer at Korkormu, Individual Interview). In a Cambodian study, more than half of the study population (60.3%) did not get cured after the first action taken hence took a second action (Yanagisawa et al., 2004). When asked whether the first action taken cured their illness, majority (64.9%) of the respondents in general said they got cured where slightly more of the urbanites (65.8%) than their rural (63.6%) counterparts responded so. Out of the seventy-two respondents who indicated they did not get cured of their perceived illness symptoms after taking the first action, 56.9% went to a formal medical facility for care as second action taken while the remaining 43.1% visited informal medical providers for treatment. More of the urbanites (65.0%) than the rural dwellers (46.9%) engaged in this care seeking option. This contradicts the results of Yanagisawa et al. (2004) where majority of both the poor and better-off still sought care at informal medical providers (bought medicines from drug sellers) as second action taken. Given that the health facility respondents attended when ill is known, the study examined the relationship between their socio-demographic characteristics and the kind of health service providers they sought treatment from (see Table 4.5.3). These socio-demographic characteristics including age, sex, educational level, and average monthly income have been reported by Andersen & Newman (1973) and further corroborated by Kroeger (1983) 102 http://ugspace.ug.edu.gh/ to influence individuals’ health seeking behaviour. The health service providers were categorised into formal medical/health providers and informal medical/health providers. Formal medical/health providers, as used in this study, include all public and private hospitals, clinics, health centres and CHPS compounds hence seeking treatment at any of the above means treatment seeking at a formal medical/health provider/facility. Informal medical providers, in this context, include all categories of traditional healers, prayer camps, drug peddlers, pharmacy shops and drug/chemical shops. Also, self-treatment and/or self-medication, in this study, is regarded as buying of medicines from a drug/chemical shop hence, care seeking in any form at any of the above is classified as treatment seeking at an informal medical/health provider/facility. Most studies report of the significant association between certain socio-demographic characteristics and response to illness symptoms whenever they are perceived. Danso- Appiah et al. (2010), Oluwatuyi (2010) and Sreeramareddy et al. (2006) indicated that age, sex, level of education, and income significantly influenced individuals’ first action taken whenever illness symptoms were perceived. In rural Ekiti State, Nigeria, Oluwatuyi (2010) identified age; educational level and income to predict the type of health service providers an individual sought treatment from. He identified a positive association between educational level and treatment seeking at formal medical facilities likewise severity of illness. Similarly, Sreeramareddy et al. (2006) found maternal education and perceived severity of illness to significantly predict ‘appropriate care’ seeking while income proved significant for seeking ‘prompt care.’ ‘Appropriate care’ was mainly sought by highly educated mothers and when a child’s illness was perceived to be serious (Sreeramareddy et al., 2006). Also, Danso-Appiah et al. (2010) found in Ghana that the likelihood of 103 http://ugspace.ug.edu.gh/ seeking formal health care increases as income and severity of schistosomiasis-related symptoms increased. Contrariwise, the findings of Pronyk et al. (2001) showed that with the exception of household size, other socio-demographic characteristics like age, sex, level of education, and income did not significantly influence individuals’ choice and use of particular health service providers as first action taken when illness symptoms were perceived. 104 http://ugspace.ug.edu.gh/ Table 4.5.3: Health facility attended when ill by socio-demographic characteristics Health facility attended when ill Variables Informal Medical Formal Medical Total Providers Providers Urban Rural Urban Rural Age 20-29 years 19(25.0%) 25(42.4%) 6(14.6%) 7(24.1%) 57(27.8%) 30-39 years 27(35.5%) 16(27.1%) 11(26.8%) 11(37.9%) 65(31.7%) 40-49 years 14(18.2%) 10(16.9%) 11(26.8%) 5(17.2%) 40(19.5%) 50-59 years 10(13.2%) 4(6.8%) 7(17.1%) 2(6.9%) 23(11.2%) 60 years and above 6(7.9%) 4(6.8%) 6(14.6%) 4(13.8%) 20(9.8%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-square (χ2) = 6.361, df (4), Sig.=0.174 Sex Male 49(64.5%) 22(37.3%) 27(65.9%) 20(31.0%) 118(57.6%) Female 27(35.5%) 37(62.7%) 14(34.1%) 9(69.0%) 87(42.4%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-square (χ2) = 4.055, df (1), Sig.=0.044 Level of Education Never schooled 4(5.3%) 15(8.5%) 1(2.3%) 5(17.2%) 25(12.2%) Primary/JHS 31(40.8%) 36(61.0%) 12(30.0%) 19(65.5%) 98(47.8%) Secondary 28(36.8%) 8(13.6%) 18(43.9%) 2(6.9%) 56(27.3%) Tertiary 13(17.1%) 0(0.0%) 10(24.4%) 3(10.3%) 26(12.7%) 105 http://ugspace.ug.edu.gh/ Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-square (χ2) = 4.387, df (3), Sig.=0.223 Average monthly income Below GHȼ5.00 5(6.6%) 24(40.7%) 0(0.0%) 8(27.6%) 37(18.0%) GHȼ51-GHȼ100 2(2.6%) 16(27.1%) 0(0.0%) 8(27.6%) 26(12.7%) GHȼ101-GHȼ200 12(15.8%) 16(27.1%) 2(4.9%) 9(31.0%) 39(19.0%) GHȼ201-GHȼ500 32(42.1%) 3(5.1%) 19(46.3%) 1(3.4%) 55(26.8%) GHȼ501-GHȼ1000 19(25.0%) 0(0.0%) 13(31.7%) 3(10.3%) 35(17.1%) Above GHȼ1000 6(7.9%) 0(0.0%) 7(17.1%) 0(0.0%) 13(6.3%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-square (χ2) = 7.772, df (5), Sig.=0.169 Source: Field Work, 2015. (Figures in bracket represent percentages in a column) Age has been one of the main socio-demographic variables that determine peoples’ choice and use of specific health service providers included in the models of Andersen & Newman (1973) and Kroeger (1983). Although significant association between age and individuals’ health seeking behaviour has been reported by some studies (Haque et al., 2012; Rahman et al., 2012; Abdulraheem & Parakoyi, 2009; Ahmed et al., 2005) findings from both the Chi-square tests of association (χ2 = 6.361, df (4), p > 0.05) (see Table 4.5.3) and the multinomial logistic regression model employed for the study (p > 0.05) (see Table 4.5.4) suggests no significant relationship between age and individuals’ health seeking behaviour in both communities. This confirms findings of Ghosh et al (2013), Mushtaq et al (2011), Al-Mandhari et al (2009) and Sreeramareddy et al (2006). However, no spatial variation exists between respondents’ age and their health seeking behaviour in the study area albeit 106 http://ugspace.ug.edu.gh/ both the young and the aged were more likely to seek treatment at informal than formal medical providers in both communities (see Table 4.5.4). The study found an inverse relationship between age and treatment seeking at both medical providers where treatment seeking at both medical providers decreases as age increases. Thus, majority of both the young (below 50 years) and aged (50 years and above) sought treatment at both formal and informal medical providers in both communities where the young predominated (see Table 4.5.3). This confirms the findings of El-Kak et al. (2009), Ahmed et al. (2005), Taffa & Chepngeno (2005) and Sandman et al. (2000). In the same way, more of the aged in both the urban (31.7% and 21.1%) and rural (20.7% and 13.6%) communities sought treatment at formal medical facilities than at informal medical healers (see Table 4.5.3) confirming the findings of Ahmed et al. (2005) and Chakraborty et al. (2003) but is in antithesis with that of Aboagye & Agyemang (2013), and Abdulraheem & Parakoyi (2009). More of the urban aged compared with their rural counterparts sought treatment at both formal and informal medical providers as has been reported by Biswas et al. (2006) in Bangladesh. Sex has been reported to have a significant relationship with the choice and use of health service providers with respect to illness type and perceived severity (El Kahi et al., 2012). Evidence from this study shows that sex is a significant predictor of individuals’ health seeking behaviour in the municipality (χ2 = 4.055, df (1), p < 0.05) confirming the findings of El Kahi et al. (2012), Khabiri et al. (2011) and Teerawichitchainan & Phillips (2007). Contrary to the findings of El Kahi et al. (2012) where Lebanese women more than their men sought care at formal medical facilities, the study found men, in both the urban (0.06) and rural (1.32) communities of the study, than women to seek treatment the more at formal medical facilities (see Table 4.5.4). This can be attributed to males dominating the study 107 http://ugspace.ug.edu.gh/ and having the economic power hence, make health decisions regarding themselves and other family members as echoed by Atuyambe et al. (2009), and Shaikh et al. (2008). For treatment seeking at informal medical providers, males (64.5%) than females (35.5%) dominated in the urban community contradicting the findings of Fomundam et al. (2012) and El Kahi et al. (2012), while females (62.7%) than males (37.3%) dominated in the rural area which confirms the findings of Fomundam et al. (2012) and Nahar (2010). Females dominating in treatment seeking from informal medical providers in the rural area can be attributed to their dominance in the rural community and in the municipality (GSS, 2012). From Table 4.5.4, sex significantly influenced individuals’ health seeking behaviour in both communities (p < 0.01) hence no spatial variation exists in terms of respondents’ health seeking behaviour and sex. The educational status of individuals world-wide has been found to be associated with their health seeking behaviour (Mashreky et al., 2010; Kakai et al., 2009; Roy et al., 2004; Chakraborty et al., 2003). The study established a positive association between education and treatment seeking at formal medical facilities in both communities (see Table 4.5.3) as reported by Haque et al. (2012), Siddiqui et al (2011), Mashreky et al. (2010) and Kakai et al (2009) where treatment seeking at formal medical facilities increased with increasing education. Similarly, a positive association was determined between education and treatment seeking at informal health providers (see Table 4.5.3) also confirming the findings of Mashreky et al. (2010), Nahar (2010), Sandman et al. (2000), and Asenso- Okyere et al. (1998). This they attributed to illness type and perceived severity of the illness as it is the case in other studies. The study found a significant difference regarding respondents educational level and their health seeking behaviour as education was found 108 http://ugspace.ug.edu.gh/ to significantly influence respondents health seeking behaviour in the rural community (χ2 = 7.499, df (3), Sig. = 0.034) but not in the urban community (χ2 = 2.516, df (3), Sig. = 0.472) according to the Chi-square test of association (see Table 4.5.3) and the multinomial logistic regression model (rural p < 0.01 and urban: p > 0.05) for the study (see Table 4.5.4). Notwithstanding, respondents’ educational level, in general, did not significantly influence their health seeking behaviour according to the Chi-square test of association (χ2 = 4.387, df (3), p > 0.05) (see Table 4.5.3) likewise the multinomial logistic regression model employed for the study (p < 0.01) (see Table 4.5.4). This affirms the findings of Oluwatuyi (2010) and Taffa & Chepngeno (2005). The model indicates that both urbanites and rural dwellers with either formal or no formal education were more likely to seek treatment from informal medical providers than from formal medical facilities (see Table 4.5.4). This confirms the findings of Mashreky et al. (2010) and Nahar (2010). Hence, a spatial variation exists in terms of individuals’ health seeking behaviour and educational level in the study area. Income has been identified as the major socio-economic determinant of individuals’ health seeking behaviour (Danso-Appiah et al., 2010; Ahmed et al., 2005; Sandman et al., 2000) as it greatly influences one’s ability to spend on health service use. Ahmed et al. (2005) found income to be the single most important determinant of individuals’ treatment seeking at health service providers. Studies (Oluwatuyi, 2010; Roy et al., 2004; Sandman et al., 2000) have established a positive association between income and treatment seeking at formal medical facilities where the rich have been found to seek treatment more than the poor. Contrary to this, the study identified an inverse relationship between income and treatment seeking at formal medical facilities as majority of the low income earners (< 109 http://ugspace.ug.edu.gh/ GHȼ500.00) in both the urban (51.2%) and rural (89.7%) communities rather sought care at formal medical facilities (see Table 4.5.3). This buttresses the findings of van der Hoeven et al. (2012), Oluwatuyi (2010), and Ahmed et al. (2005) where the rural rich than their poor counterparts sought treatment the more at formal medical facilities. Also, treatment seeking at informal health providers was dominated by the low income earners in both communities confirming the findings of El Kahi et al. (2012) and Subba (2004) but is in antithesis to the findings of Yanagisawa et al. (2004) where majority of the rich were found mainly to apply a home remedy and self-medicate. The study established a significant association between respondents’ income and their health seeking behaviour in the rural community according to both the Chi-square test of association (χ2 = 7.499, df (3), Sig. = 0.034) (see Table 4.5.3) and the multinomial logistic regression model ((p < 0.01) (see Table 4.5.4) but not in the urban community. Such finding reverberates that of Danso- Appiah et al. (2010), Kakai et al. (2009), and Ahmed et al. (2005). Rural dwellers belonging to all the categories of income classifications were more likely to seek treatment from informal medical providers than their urban counterparts (see Table 4.5.4). Per the model, rural dwellers who earned below GHȼ50.00 every month on the average, were twenty-one (21.49) more likely to seek treatment from informal medical providers (see Table 4.5.4). On the part of their urban counterparts, the likelihood of going for treatment from informal medical providers decreases with increasing income (see Table 4.5.4). 110 http://ugspace.ug.edu.gh/ Table 4.5.4: Multinomial Logistic Regression coefficients for independent variables of predictors of the health facility respondents attended when ill Health facility attended when ill Informal Medical Providers Formal Medical Providers Urban Rural Urban Rural Variables Odds ratio Odds ratio Odds ratio Odds ratio Age 20-29years 1.437 2.019* -1.437 -2.019* 30-39yeas 1.153 0.740 -1.153 -0.740 40-49years 0.403 1.479 -0.403 -1.479 50-59years 0.431 1.107 -0.431 -1.107 60years and above 0.258 0.539 -0.258 -0.539 Intercept 1.197 1.171 -1.197 -1.171 Sex Male -0.061 -1.318 0.061 1.318 Female -0.022 -1.192 0.022 1.192 Intercept 0.657* 1.414** -0.657* -1.414** Educational Level Never Schooled 1.124 20.572** -1.124 -20.572** Primary/JHS 0.687 20.112** -0.687 -20.112** Secondary 0.179 20.859** -0.179 -21.859** Tertiary 0.124 16.085** -0.124 -16.085** Intercept -0.262 -19.473** 0.262 19.473** 111 http://ugspace.ug.edu.gh/ Income (GHȼ) Below 50.00 19.486 21.572** -19.486 -21.572** 51.00-100.00 19.293 21.116** -19.293 -21.116** 101.00-200.00 1.946* 21.048** -1.946* -21.048** 201.00-500.00 0.675 21.572** -0.675 -21.572** 501.00-1000.00 0.534 19.248** -0.534 -19.248** Above 1000.00 0.347 - -0.347 - Intercept -0.154 -20.473** 0.154 20.473** Source: Field Work, 2015 *P < 0.05 **P < 0.01 4.6 Summary The chapter examined the distribution of respondents according to their socio-demographic characteristics which included age, sex, education, marital status and income. It also examined the type of illness respondents suffered from in the study communities and the means of diagnosis of such illnesses. Again, the severity and the day of illness treatment was sought was examined likewise the health service providers available in the study communities. More so, the chapter assessed the health seeking behaviour of individuals in the municipality and established the relationship between respondents’ socio-demographic characteristics and their health seeking behaviour. Males (57.6%), in general, dominated the study likewise the 30-39 years age category (31.7%). Majority (87.8%) of the respondents had some form of formal education. This was high among the urbanites (58.9%). The most predominant religion professed is Christianity (60.0%) with Traditional religion (1.5%) being the least professed. Though the 112 http://ugspace.ug.edu.gh/ study area is an Akan state, Ewes (33.7%) predominated. The employment situation in the study area is very encouraging with the rural dwellers (94.3%) slightly ahead of their urban counterparts (93.2%). Majority (30.7%) of the employed in general, engaged in trading. Urbanites were by far better-off than their rural counterparts as majority (38.5%) of them compared with the rural dwellers (3.4%) earned more than GHȼ500.00 per month, on the average. Conversely, an overwhelming proportion of the rural inhabitants (96.6%) earn below GHȼ500.00 averagely each month. In general, respondents who earned below GHȼ500.00 (76.6%) averagely dominated with only 23.4% earning, on the average, above GHȼ500.00. As was the case in Africa (Fomundam et al., 2012) and Nepal (Subba, 2004), general body pains/headache (33.7%) was the predominant illness type perceived by respondents in both communities, followed by malaria/fever (23.9%) with the least (1.4%) being acute respiratory infections (ARIs). The predominant means of diagnosis was by oneself. This is followed distantly by diagnosis by professional medical personnel. Unsurprisingly, majority of the respondents that did not know of the type of their illness got diagnosed by professional medical personnel likewise majority of those that suffered from malaria/fever (42.9%) and all those that suffered from ARIs. Contrary, self-diagnosis predominated among respondents that suffered from stomach/abdominal (62.5%) and general body pains/headache (82.6%). Respondents with severe (34.2%) illness symptoms dominated the study followed by those with less severe (28.3%) illness symptoms with the least (18.5%) that claimed it was not severe. Treatment seeking on the fourth day (34.6%) of symptoms identification predominated. As expected, majority of the respondents who did not know of the type of 113 http://ugspace.ug.edu.gh/ illness, suffered from malaria/fever, stomach/abdominal pains, and all those that suffered from ARIs, had their symptoms to be severe and also, sought treatment within the first three days of symptoms identification. On the other hand, majority of the respondents that suffered from general body pains/headache had such illnesses to be less severe hence, waited for four to five days and over before going for treatment. This was significant in both the urban and rural communities selected for the study. Subba (2004) reports of treatment seeking at a particular health facility based on the type of illness suffered. With this, treatment seeking at informal medical providers was dominated by respondents that suffered from general body pains/headache whereas care seeking at formal medical facilities was dominated by those that claimed suffered from malaria/fever and those who did not know of their illnesses. It is worth noting, however, that none of the respondents that suffered from acute respiratory infections (ARIs) visited an informal medical provider for care. The study found both medical systems to be available in Adukrom while for Korkormu, only the informal medical system is available in the community. With this, treatment seeking at available health providers in both communities, according to the Chi-Square goodness-of-fit test, was significant as care seeking at these available health providers dominated the study hence, the acceptance of the hypothesis that the availability of a health service provider will influence their health service provider use. Like other studies (Gotsadze et al., 2005; Yanagisawa et al., 2004), as health facility respondents attended when ill, treatment seeking at informal medical providers (65.9%), in general, dominated where more of the rural dwellers (67.0%) than the urbanites (65.0%) 114 http://ugspace.ug.edu.gh/ engaged in such practice. Besides, more urbanites than their rural counterparts sought care at formal medical facilities affirming the findings of van der Hoeven et al. (2012). The health facility respondents attended when ill was significant in both communities hence, the rejection of the hypothesis that a difference exists in the health seeking behaviour of urban and rural dwellers in the municipality. Also, a significant difference was found between health seeking behaviour and education and income in the municipality whereas for sex and individuals’ health seeking behaviour, no difference exists as sex significantly predicted their health seeking behaviour in both the urban and rural communities selected for the study. 115 http://ugspace.ug.edu.gh/ CHAPTER FIVE UTILISATION OF HEALTH SERVICES IN THE AKUAPEM NORTH MUNICIPALITY 5.0 Introduction The utilisation of health services, as reported by Andersen & Newman (1973) and further emphasised by Kroeger (1983), is determined by a host of factors, which induce individuals and populations to behave differently towards treatment seeking whenever symptoms of illness are perceived (MacKian, 2003). This chapter sought to assess respondents’ most preferred health facilities in the municipality. This was to find out where they would want to go for treatment whenever illness symptoms are perceived should they get the opportunity. Also, to help ascertain the predictive factors of individuals’ health seeking behaviour, the chapter sought to examine the determinants of individuals’ health seeking behaviour in the municipality. It again examined the barriers that hinder patients from seeking treatment at various health facilities. This helped to explore the specific health service provider respondents encountered difficulties the more as service was utilised. In the same way, the various means by which patients financed their health service use is examined to help find out how service use at particular health providers was financed. 5.1 Most preferred health service provider of respondents The particular health provider respondents usually sought treatment from was identified and the majority (46.8%) indicated they usually seek treatment at a health centre with more of the rural dwellers (52.2%) than the urbanites (42.8%) reporting so. This is followed distantly by 18.5% that said they usually bought medicines at drug/chemical shops with 116 http://ugspace.ug.edu.gh/ the least (1.5%) that purchased drugs from drug peddlers (see Table 5.1.1). Where respondents usually go for care generally showed significant (χ2 = 112.049, df (3), p = 0.000) in both the urban and rural communities of the study. Table 5.1.1: Health facility usually accessed by respondents by place of residence Place of residence Urban Rural Total Variables N % N % N % Where treatment is usually sought Hospital 30 25.6 7 8.0 37 18.1 Health centre 50 42.8 46 52.2 96 46.8 Traditional healer 2 1.7 29 33.0 31 15.1 Drug/chemical shop 35 29.9 3 3.4 38 18.5 Drug peddler 0 0.0 3 3.4 3 1.5 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 112.049, df (3), Sig. = .000 Source: Field Work, 2015. Regarding their major reasons for usually seeking treatment at such health facilities, majority of both the urban (29.9%) and rural (21.6%) respondents said they did so because they always get cured of their illness. Other reasons given by respondents in the urban area include ‘provide quality care’ (21.4%), with the least being ‘illness needed special care’ (1.7%) while in the rural area ‘provide quality care’ (19.3%), ‘understands my health problem’ (18.2%) and ‘illness needed special care’ (2.3%) being the least. 117 http://ugspace.ug.edu.gh/ This became evident in the qualitative facet of the study as participants in both the urban and rural communities had these to say: “I am always assured of getting cured of my illness whenever I go there for treatment due to the quality service provided. They don’t just get up and give you medicines. The last time I went there for instance, series of laboratory tests were conducted on me before medication was given out and I was cured after three days of taking the medicines” (A 36 year old trader at Adukrom, Individual Interview) “For me whenever I am pregnant, I go to the health centre for treatment though it is a bit far from here. There are a lot of complications with pregnancy which sometimes need special treatment from qualified medical personnel. I have seen people die during pregnancy and child birth so that is why I always seek treatment from the health centre when I am pregnant” (A 27 year old petty trader at Korkormu, Individual Interview). Information regarding where respondents would want to go for treatment whenever they perceive symptoms of illness was elicited and majority of both the urbanites (76.9%) and rural inhabitants (47.8%) said they would prefer seeking treatment at a hospital. In addition, 12.8% and 28.4% of both the urban and rural dwellers opted for treatment seeking at a health centre with the least in both the urban (3.4%) and rural (2.2%) communities respectively that preferred to seek treatment at a drug/chemical shop (see Table 5.1.1). This is an indication that treatment seeking from the formal medical system is the preferred health seeking system in both communities as an overwhelming majority of both the urbanites (89.7%) and rural dwellers (76.2%) reported they prefer to seek treatment at formal medical facilities whenever symptoms of illness are perceived (see Table 5.1.1). 118 http://ugspace.ug.edu.gh/ Generally, the Chi-Square goodness-of-fit test conducted showed that respondents’ most preferred health service provider showed significant across their place of residence (χ2 = 276.195, df (4), p = 0.000) and also in both the urban (χ2 = 241.590, df (4), p = 0.000) and rural (χ2 = 68.364, df (4), p = 0.000) communities of the study (see Table 5.1.1). This indicates that no difference exists regarding respondents’ preference for health service providers in the Akuapem-North Municipality. Table 5.1.2: Respondents’ most preferred health provider by place of residence Place of residence Urban Rural Total Variables N % N % N % Most preferred health provider Hospital 90 76.9 42 47.8 132 64.4 Health centre 15 12.8 25 28.4 40 19.5 Traditional healer 7 6.0 19 21.6 26 12.7 Drug/chemical shop 5 4.3 2 2.2 7 3.4 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 276.195, df (4), Sig. = .000 Source: Field Work, 2015. Both the urban and rural respondents gave the following as their reasons for choosing a particular health provider as their most preferred health provider: ‘provide quality care always’ (35% and 40.9%), ‘have the best medical personnel’ (29.1% and 29.5%) and 119 http://ugspace.ug.edu.gh/ ‘always get cured of their illness’ (15.4% and 8%). Participants in the in-depth interview had these to say: “I prefer going for treatment there because that is where you will get all the best professional medical personnel in the municipality to take care of you” (A 33-year old civil servant at Adukrom, Individual Interview). “They always give good medicines that are able to cure whatever illness you go there with. They also have medical personnel who are professionals and well experienced in handling all medical cases and on top of it all, they accept NHIS” (A 28-year old woman at Korkormu, Individual Interview). The study further found whether their preferred health providers were where they sought treatment at and 36.8% and 33.0% of both urban and rural respondents indicated seeking treatment at their preferred health providers while 34.2% and 29% urbanites and 31.8% and 35.2% of the rural dwellers responded ‘no’ and ‘sometimes’ respectively. The views of those who said ‘no’ and ‘sometimes’ were sought regarding their inability to go for care from their most preferred health providers and majority of both the urban (60.8%) and rural (76.3%) respondents said it is because such providers are not available in their communities. 5.2 Distance and means of transport to health providers Though numerous health providers were found to be present in both the rural and urban communities, information was gathered on particular ones that were closest to respondents’ houses. For the rural community, different kinds of traditional healers were reported being 120 http://ugspace.ug.edu.gh/ closest to their houses. On the part of the urbanites, majority (53.0%) reported having a drug/chemical shop nearest to their houses followed by those who said a health centre (31.6%) was closest to their houses. Respondents that said a traditional healer (14.5%) was the closest to their houses followed suit with the least (0.9%) having a prayer camp closest to their houses. Respondents were further asked about the distance, in terms of kilometres, to these health providers from their houses. From Figure 5.2.1, all the rural dwellers said it is less than a kilometre likewise majority of the urbanites (88.9%) with the remaining 11.1% guessing it to be between one to four kilometres. Figure 5.2.1 Distance to nearest health provider 100% 88.9% 11.1% Urban Rural Less than 1km Btn 1km-4km Source: Field Work, 2015. Regarding the means of transport respondents used in reaching the usual health providers they sought treatment from, majority (56.1%) said they go there by vehicles buttressing the report of Mazumda et al. (2009) but contradicts that of Russell (2008) and Subba (2004) where walking and bicycle were the major means of transport to health facilities 121 http://ugspace.ug.edu.gh/ respectively. This is followed by those that go there on foot (39.0%). The means of transport respondents used in reaching the health facilities they visited for treatment when ill showed significant (χ2 = 179.527, df (3), p = 0.000) in both communities (see Table 5.2.1). Regarding the amount respondents spent on transportation to these health service facilities they visited for treatment, majority (44.9%) of the respondents in both communities said they paid less than GHȼ5.00 which is relatively higher than what has been reported by Russell (2008) in the country. This is followed closely by those that said they paid nothing while the least (13.2%) claimed paying more than GHȼ5.00 (see Table 5.2.1). The amount of money respondents spent on transportation to and from health facilities was statistically significant in both communities (χ2 = 209.761, df (5), p = 0.000) (see Table 5.2.1). Information regarding the one who paid for respondents’ transportation expenses to and from health service providers they usually visited for treatment was assessed and the majority (44.8%) reported footing their transport expenses themselves. This was followed closely by those who were of the view that nobody paid for their transportation expenses (42.0%) because no transport cost was borne. This also was statistically significant in both communities (χ2 = 190.098, df (4), p = 0.000) (see Table 5.2.1). Also, the time taken by respondents to reach the particular health service provider where actions were usually taken was determined and generally, those that reported taking less than thirty (30) minutes (82.9%) predominated. This was followed distantly by those that spent between thirty (30) minutes to one (1) hour (15.6%) while the least (1.5%) reported taking more than one (1) hour (see Table 5.2.1). This was significant across respondents’ place of residence (χ2 = 99.605, df (3), p = 0.000). Moreover, respondents were asked to 122 http://ugspace.ug.edu.gh/ indicate the average time it took them to receive treatment the last time they utilised health services and majority (38.0%) indicated they spent more than one hour contradicting the findings of Russell (2008). This is followed closely by those who claimed spending less than thirty (30) minutes (37.6%) while the least (24.4%) said they spent between thirty minutes to one hour at the health facility (see Table 5.2.1). More of the rural (42.0%) than the urban (35.0%) inhabitants reported waiting at health facilities for longer times before finally being attended to. The test statistic shows that this was statistically significant in both communities (χ2 = 31.707, df (4), p = 0.000). 123 http://ugspace.ug.edu.gh/ Table 5.2.1: Respondents’ means of transport, payment for transport, travel time and waiting time at health facilities. Place of residence Urban Rural Total Variables N % N % N % Means of transport Foot 56 47.9 24 27.3 80 39.0 Bicycle 0 0.0 1 1.1 1 0.5 Motorbike 1 0.9 8 9.1 9 4.4 Vehicle 60 51.3 55 62.5 115 56.1 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 179.527, df (3), p = 0.000 Amount spent on transport (GHȼ) None 59 50.4 27 30.7 86 42.0 GHȼ5.00 19 16.2 8 9.1 27 13.2 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 37.766, df (2), p = 0.000 Payment for transportation None 58 49.8 28 31.8 86 42.0 Self 52 44.4 40 45.5 92 44.9 Spouse 1 0.9 12 13.7 13 6.3 Parent 0 0.0 4 4.5 4 2.0 124 http://ugspace.ug.edu.gh/ Other relative 6 5.1 4 4.5 10 4.9 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 190.098, df (4), p = 0.000 Time taken to reach health service provider <30minutes 92 78.6 78 88.6 170 82.9 30mins-1hr 24 20.5 8 9.1 32 15.6 >1hour 1 0.9 2 2.3 3 1.5 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 233.044, df (2), p = 0.000 Waiting time at health facility <30minutes 40 34.2 37 42.0 77 37.6 30mins-1hr 36 30.8 14 16.0 50 24.4 >1hour 41 35.0 37 42.0 78 38.0 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 31.707, df (4), p = 0.000 Source: Field Work, 2015. 5.3 Determinants of individuals’ health seeking behaviour As reported by Andersen & Newman (1973) and further emphasised by Kroeger (1983) a plethora of variables, apart from socio-demographic characteristics, influence the health seeking behaviour of people. These variables dictate the choice of a particular health service provider an individual engages with whenever symptoms of illness are perceived (Shaikh & Hatcher, 2005; Roy et al., 2004; MacKian, 2003; Andersen, 1995). Given that respondents’ first action taken whenever illness symptoms are perceived is known (see 125 http://ugspace.ug.edu.gh/ Table 4.5.1), the study examined the relationship between these variables, which include health insurance, severity of illness, and waiting time and the kind of health service providers they sought treatment at. The result is shown in Table 5.3.1. The health service providers were grouped into formal medical providers and informal medical providers. Formal medical providers include all forms of government and private hospitals, clinics, health centres and CHPS compounds and care seeking from any of the above denotes treatment seeking from a formal medical provider/facility. Conversely, informal medical providers include all forms of traditional healers, faith healers, drug peddlers, pharmacy shops, drug/chemical shops, and self-treatment and/or self-medication. Engaging with any of the above denotes care seeking from an informal medical provider/facility. Most studies report of the significant association between the utilisation of health services and health insurance (Annan et al., 2013; Blanchet et al., 2012; El-Kak et al., 2009), severity of illness (El Kahi et al., 2012; Rahman et al., 2012; Danso-Appiah et al., 2010), distance to health providers (Mushtaq et al., 2011; Al-Mandhari et al., 2009; Sreeramareddy et al., 2006), and quality of health service (Russel, 2008; Uzochukwu & Onwujekwe, 2004). For instance, Danso-Appiah et al (2010), Oluwatuyi (2010) and Sreeramareddy et al. (2006) indicated that severity of illness significantly influenced individuals’ health seeking behaviour. In rural Ekiti State, Nigeria, Oluwatuyi (2010) identified a positive association between severity of illness and treatment seeking at formal medical providers. A similar study by Sreeramareddy et al. (2006) found the severity of a child’s illness as a significant predictor for seeking ‘appropriate care.’ Also, Danso-Appiah et al. (2010) found in Ghana that the likelihood of seeking formal health care increases as the severity of schistosomiasis-related symptoms increases. 126 http://ugspace.ug.edu.gh/ Table 5.3.1: Determinants of individuals’ health seeking behaviour Health facility attended when ill Determinants Informal Medical Formal Medical Total Providers Providers Urban Rural Urban Rural Health Insurance Insured 71(93.4%) 48(81.4%) 41(100.0%) 29(100.0%) 189(92.2%) Uninsured 5(6.6%) 11(18.6%) 0(0.0%) 0(0.0%) 16(7.8%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-Square (χ2) = 8.999 df (1) Sig. = 0.003 Severity of illness Not severe 18(23.7%) 14(23.7%) 6(14.6%) 0(0.0%) 38(18.5%) Less severe 32(42.1%) 20(33.9%) 2(4.9%) 4(13.8%) 58(28.3%) Severe 24(31.6%) 18(30.5%) 21(51.2%) 7(24.3%) 70(34.2%) More severe 2(2.6%) 7(11.9%) 12(29.3%) 18(62.1%) 39(19.0%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-Square (χ2) = 53.110 df (3) Sig. = 0.0000 Distance to health service providers < 1kilometre 71(93.4%) 59(100.0%) 33(80.5%) 29(100.0%) 192(93.7%) Btn 1km to 4km 5(6.6%) 0(0.0%) 8(19.5%) 0(0.0%) 13(6.3%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-Square (χ2) = 4.632 df (1) Sig. = 0.031 Quality of service 127 http://ugspace.ug.edu.gh/ Excellent 24(31.6%) 24(40.7%) 11(26.8%) 15(51.7%) 74(36.1%) Very good 35(46.1%) 28(47.5%) 15(36.6%) 12(41.4%) 90(43.9%) Good 12(15.8%) 7(11.9%) 14(34.1%) 2(6.9%) 35(17.1%) Fair 5(6.6%) 0(0.0%) 1(2.4%) 0(0.0%) 6(2.9%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-Square (χ2) = 3.618 df (3) Sig. = 0.306 Waiting time < 30 minutes 37(48.7%) 28(47.5%) 3(7.3%) 9(31.0%) 77(37.6%) Btn 30mins to 1hour 24(31.6%) 10(16.9%) 12(29.3%) 4(13.8%) 50(24.4%) > 1hour 15(19.7%) 21(35.6%) 26(63.4%) 16(55.2%) 78(38.0%) Total 76(100.0%) 59(100.0%) 41(100.0%) 29(100.0%) 205(100.0%) Chi-Square (χ2) = 25.362 df (2) Sig. = 0.000 Source: Field Work, 2015. (Figures in bracket represent percentages in a column). Health insurance has been found to influence individuals’ choice and use of specific health service providers both directly and indirectly (Aboagye & Agyemang, 2013; Ghosh et al., 2013; Blanchet et al., 2012; Robyn et al., 2012). Studies (Aboagye & Agyemang, 2013; Blanchet et al., 2012; Roy et al., 2004) have reported of a positive correlation between health insurance and treatment seeking at formal medical facilities. Conversely, other studies (Annan et al., 2013; Laar et al., 2013; Robyn et al., 2012) have reported of increased treatment seeking at informal medical providers on the part of the insured for such reasons as proximity, poor road network and inadequate means of transport to formal health facilities. Annan et al. (2013), Khabiri et al. (2011), and El-Kak et al. (2009) identified a significant association between health insurance and treatment seeking at health service 128 http://ugspace.ug.edu.gh/ providers. The study found majority of both the urban (93.4%) and rural (81.4%) insured to seek treatment at formal medical providers as none of the uninsured in both communities sought treatment at formal medical facilities (see Table 5.3.1). This confirms the findings of Ghosh et al. (2013), Khabiri et al. (2011), Siddiqui et al. (2011) and El-Kak et al. (2009). Also, substantial proportions of both the urban (6.6%) and rural (19.6%) insured visited informal health providers for care as has already been reported in Ghana (Aboagye & Agyemang, 2013; Annan et al., 2013; Laar et al., 2013) and in Burkina Faso (Robyn et al., 2012). The urban insured were of the view that their illness symptoms were not severe while their rural counterparts cited the unavailability of a formal medical facility in their community as the main reasons for engaging with informal medical providers. For the uninsured, more of the rural dwellers than the urbanites visited informal medical healers for care (see Table 5.3.1) which buttresses the findings of Siddiqui et al. (2011) and Khabiri et al. (2011). This is described by a participant in one of the in-depth interviews: “I don’t see why I should again waste money at other health facilities knowing that I can go for treatment at the health centre on the ticket of NHIS. Oh yes! But I think it is as a result of tiredness due to over working. They are human beings so it is possible they can be rude sometimes but for me, that is not a reason to pay for health service I can access freely through NHIS. I have seven (7) children and without NHIS it would be difficult to cater for their health needs. They all access health service on the ticket of NHIS” (A 47 year old driver at Adukrom, Individual Interview). Health insurance was found to significantly influence respondents’ health seeking behaviour in both communities (χ2 = 8.999, df (1), p = 0.003) according to the Chi-square 129 http://ugspace.ug.edu.gh/ test of association confirming the findings of Annan et al. (2013), Khabiri et al. (2011), and El-Kak et al. (2009). In the same way, the multinomial logistic regression model established a significant association between health insurance and individuals’ health seeking behaviour in the municipality (p < 0.01) as the insured urbanites were sixteen times (16.78) whereas the rural insured were seventeen times (17.87) more likely to seek treatment at formal health facilities (see Table 5.3.2). The severity of individuals’ illness has been found to greatly determine their choice for a particular medical system whenever illness symptoms are perceived and also when treatment is sought for a particular symptoms of illness (Rahman et al., 2012; Danso- Appiah et al. 2010; Biswas et al., 2006; Sreeramareddy et al., 2006; Taffa & Chepngeno, 2005). Reports indicate a positive association between individuals’ illness severity and treatment seeking at formal medical facilities. Thus, treatment seeking at formal medical facilities increases as individuals’ illness symptoms get severe and more severe while individuals seek treatment at informal medical providers the more when they perceive their illness symptoms not to be severe and/or less severe (Afolabi et al., 2013; Rahman et al., 2012; Khabiri et al., 2011; Danso-Appiah et al., 2010; Oluwatuyi, 2010). Studies conducted by El Kahi et al. (2012), Rahman et al. (2012), Danso-Appiah et al. (2010), and Sudha et al. (2003) have identified individuals’ illness severity as a significant predictor of their health seeking while some (Haque et al., 2012; Rai et al., 2012; Roy et al., 2004) have not established such a significant association. The study confirmed the positive association between individuals’ illness severity and their health seeking behaviour as more of both the urban (80.5%) and rural (84.2%) dwellers sought treatment at formal medical facilities 130 http://ugspace.ug.edu.gh/ as illness symptoms were perceived as severe and more severe (see Table 5.3.1). This is what a participant in an interview had to say; “It is not intentional and has never been so. Reporting earlier at the health centre here after symptoms identification comes with two things: either you are insulted or you are given ‘pain killers’ like paracetamol which can be obtained easily and cheaply from the drug/chemical shops. So to avoid the inhumane attitudes by the personnel at the health centre, I normally buy drugs at the drug store and only go to the health centre for treatment when I do not get cured. So definitely this will take some days” (A 38 year old driver at Adukrom, Individual Interview). This is similar to findings of other similar studies conducted in Uganda (Hjelm & Atwine, 2011), Ghana (Danso-Appiah et al., 2010), Bangladesh (Biswas et al., 2006) and Nepal (Sreeramareddy et al., 2006). This shows that the situation is not peculiar to Ghana even with time lapse. Also, more of both the urban (65.8%) and rural (57.6%) respondents visited informal medical healers for cure whenever illness symptoms were perceived as ‘not severe’ and ‘less severe’ (see Table 5.3.1) confirming the findings of Rahman et al. (2012), Oluwatuyi (2010), Danso-Appiah et al. (2010), Biswas et al. (2006), and Taffa & Chepngeno (2005). The study also found more of the urbanites (75% and 73.6%) than their rural counterparts (57.2% and 45.8%) respectively with illness symptoms not to be severe and less severe to wait for five days and above before seeking some form of treatment from health providers confirming the findings of Rahman et al. (2012), Danso-Appiah et al. (2010) and Laar et al. (2013). Similarly, more of the urbanites (53.4% and 50%) compared with the rural 131 http://ugspace.ug.edu.gh/ dwellers (28% and 44%) with severe and more severe symptoms waited for five and more days before going for treatment from health providers also confirming the findings of Chakraborty et al. (2003). This is quite better than in India where chest symptomatics in both rural and urban areas waited as long as ten (10) days before seeking some form of treatment all because their illness were seen not to be severe (Sudha et al., 2003). This became evident in the in-depth interview: “Though I have enrolled on the National Health Insurance Scheme (NHIS), accessing health service in a formal medical facility is not that easy since such formal health facilities are not available here. Sometimes too the illness is not that serious and you have lots of work to do on your farm. So you would want to go to the farm and only go to a formal medical facility for treatment after some days of illness still persisting” (A 33 year old farmer at Korkormu, Individual Interview). The severity of respondents’ illness was identified as a significant predictor of their health seeking behaviour in both the urban and rural communities according to both the Chi- square test of association (χ2 = 53.110, df (3), p = 0.000) and the multinomial logistic regression model for the study (p < 0.05) (see Tables 5.3.1 and 5.3.2). This resonates with the reports of Rahman et al. (2012), El Kahi et al. (2009), Sudha et al. (2006) and Taffa & Chepngeno (2005). Hence, the alternative hypothesis which states that a significant relationship exist between illness severity and the type of health facility rural and urban dwellers sought treatment from whenever illness symptoms were perceived is accepted while the study rejects the null hypothesis which states otherwise. 132 http://ugspace.ug.edu.gh/ Distance, all over the world, has been identified as a key geographical factor that greatly influences and determines the choice and use of health services at a particular medical provider (Sudharsanam & Rotti, 2009; Al-Mandhari et al., 2008; Mazumda et al., 2009; Russell, 2008). It also determines the means and cost of transportation to and from health facilities (Russel, 2008; Sreeramareddy et al., 2006; Yanagisawa et al., 2004; Subba et al., 2004). The study found overwhelming proportions (93.4% and 80.5%) of the urbanites to seek treatment at both informal and formal medical providers because they were very close (less than a kilometre) to their houses. For the rural dwellers, 72.9% visited informal health providers for care for being the closest (less than a kilometre) to their houses (see Table 5.3.1). A participant in the in-depth interview had this to say; “Oh my friend! Why should I walk or travel over a very long distance just to access health service of which I can access one very close to my house and would not need to pay for transportation. I also get cured whenever treatment is sought from these health service providers very close to my house and their staff too are very friendly and treat me well” (A 46 year old businessman at Adukrom, Individual Interview). This confirms the findings of Aboagye & Agyemang (2013), Annan et al. (2013), Khabiri et al. (2011), and Mashreky et al. (2010). Also, a considerable proportion (27.1%) of the rural dwellers went to formal medical facilities for care even though they (formal medical providers) were not available in their locality hence over four (4) kilometres away from their houses (see Table 5.3.1). From the qualitative facet of the study, this was attributed to especially the severity of respondents’ illness and the illness needing special care. This also supports the reports of Ghosh et al. (2013), Laar et al. (2013) and Rahman et al. (2012) where patients travel over ten (10) kilometres in order to utilise services at formal health 133 http://ugspace.ug.edu.gh/ facilities that are not available in their communities. Distance was found to be a significant predictor of respondents’ health seeking behaviour in both the urban and rural communities according to the Chi-Square test of association (χ2 = 4.631, df (1), p = 0.013) (see Tables 5.3.1). This confirms the findings of Mushtaq et al. (2011), Al-Mandhari et al. (2009), and Sreeramareddy et al. (2006). Also, respondents whose houses were less than a kilometre away from closest health service providers were more likely to seek treatment there as those whose houses were more than a kilometre away from health service providers were less likely to seek treatment there according to the multinomial logistic regression model for the study (see Table 5.3.2). The model also established a significant association between distance and health seeking behaviour in the municipality (p < 0.05) (see Table 5.3.2). As reported by other studies in other parts of the world, the low quality of service of formal medical providers especially in the urban area makes treatment seeking at informal medical providers paramount (Annan et al., 2013; Ghosh et al., 2013; Khabiri et al., 2011; Mushtaq et al., 2011). This is even the case in the highly industrialised nations (Roy et al., 2004; Sandman et al., 2000). Similar finding is depicted by the current study. The study identified the service quality in the rural community to be higher compared with that provided in the urban community as more of the rural respondents (92.1% and 88.2%) than their urban counterparts (63.4% and 77.7%) rated the quality of service provided by both formal and informal medical providers as excellent and very good respectively (see Table 5.3.1). This was confirmed in the in-depth interview: “The drug/chemical shop salespersons are more friendly and respectful compared with staff of the health centre here. Even if you do not know the medicine to buy, 134 http://ugspace.ug.edu.gh/ they take their time in finding out from you exactly what your problem is and give you the right medicines. I always get cured of any illness too any time I buy drugs from the drug/chemical shop” (A 27 year old trader at Adukrom, Individual Interview). “Personnel of health service providers here are good, very caring, loving and respectful. They are always prepared and ready to help you get cured of your illness and for me I always get cured of my illness whenever I seek treatment from them” (A 31 year old farmer at Korkormu, Individual Interview). In spite of this, respondents’ perceived quality of service rendered in both communities was found not to significantly influence their health seeking behaviour according to the Chi-square test of association (χ2 = 3.618, df (3), p = 0.306) and the multinomial logistic regression model employed for the study (p > 0.05) (see Table 5.3.1 and 5.3.2). This contradicts the findings of Russell (2008) and Uzochukwu & Onwujekwe (2004) where service quality significantly influenced individuals’ health seeking behaviour. The average time spent at health service facilities as individuals utilise health services has been found to influence their health service provider options (Afolabi et al., 2013; Fomundam et al., 2012; van der Hoeven et al., 2012; Russell, 2008). Patients have been found to spend shorter times in receiving treatment at informal than formal health providers (Afolabi et al., 2013; Fomundam et al., 2012; Russell, 2008). This increases health service use at such facilities since short waiting times at these facilities means prompt treatment and recovery after diagnosis, and increased access to income due to longer hours of work (Russell, 2008). Also, reports indicate short waiting times at private formal health facilities 135 http://ugspace.ug.edu.gh/ than at public health care facilities hence, predominant treatment seeking at the former than the latter (van der Hoeven et al., 2012; Russell, 2008). Conversely, Annan et al. (2013), Dako-Gyeke et al. (2013) and Laar et al. (2013) reports of higher rates of treatment seeking at formal medical facilities in Ghana due to short waiting times. Similar finding was established in both the quantitative and qualitative facets of the study. Some respondents in both the urban and rural communities selected for the study had these to say in the in- depth interview: “Unlike the health centre, I do not spend more than five minutes anytime I go to the drug/chemical shop for medicines. Sometimes even less than a minute. But I spent more than two (2) hours the last time I went to the health centre here for treatment. So getting all forms of medicines you are looking for in less than five (5) minutes means you can do other things like working as well” (A 36-year old trader at Adukrom, Individual Interview). “I spend more time seeking antenatal care at any of the health centres around us than at the traditional birth attendant (TBA) here” (A 28-year old trader at Korkormu, Individual Interview). Although majority (73.3%) of the respondents sought care at informal health facilities due to lesser waiting times than at formal medical facilities (Afolabi et al., 2013; Fomundam et al., 2012; Russell, 2008), a significant difference was determined between individuals’ health seeking behaviour and waiting time in the municipality. As a significant association was established between respondents’ waiting time and the type of health facility they sought treatment from according to the Chi-square test of association in both communities 136 http://ugspace.ug.edu.gh/ (χ2 = 30.871, df (4), p = 0.000) (see Table 5.3.1), no significant association was established between respondents’ waiting time and the type of health facility they sought treatment from when ill (p < 0.05) according to the multinomial logistic regression model employed for the for the study though the urbanites were three times (3.06) less likely to spend 30 minutes at formal medical facilities (see Table 5.3.2). 137 http://ugspace.ug.edu.gh/ Table 5.3.2: Multinomial logistic regression coefficients for independent variables of predictors of the health facility respondents attended when ill Health facility attended when ill Informal Medical Providers Formal Medical Providers Urban Rural Urban Rural Variables Odds ratio Odds ratio Odds ratio Odds ratio Health Insurance Insured -16.782** -17.865** 16.782** 17.865** Uninsured -14.583** -15.743** 14.583** 15.743** Intercept 17.331** 18.361** -17.331** -18.361** Severity of illness Not severe 2.890** 22.314* -2.890** -22.314* Less severe 4.564** 2.554** -4.564** -2.554** Severe -1.925* -1.889** 1.925* 1.899** More severe -1.037** -1.316* 1.037** 1.316* Intercept -1.792* -0.944* 1.792* 0.944* Distance to Health Provider < 1kilometre 1.287* 0.873** -1.287* -0.873** Btn 1km to 4km 0.729* - -0.729* - Intercept -0.470* 0.873** 0.470* -0.873** Quality of service Excellent -0.829 -0.783 0.829 0.783 Very good -0.762 -0.405 0.762 0.405 138 http://ugspace.ug.edu.gh/ Good -1.764 -0.216 1.764 0.216 Fair -0.234 - 0.234 - Intercept 1.607 1.253 -1.607 -1.253 Waiting time < 30 minutes 3.062** 0.863 -3.062** -0.863 Btn 30mins to 1hr 1.243* 0.644 -1.243* -0.644 >1 hour -0.423 0.639 0.423 -0.639 Intercept -0.550 0.272 0.550 -0.272 Source: Field Work, 2015 *P < 0.05 **P < 0.01 5.4 Barriers to the choice and use of health services The choice and use of particular health service facilities by individuals whenever illness symptoms are perceived are sometimes hindered by certain barriers (Abubakar et al., 2013; Iyalomhe & Iyalomhe, 2012; El Kahi et al., 2012; Andersen, 1995). As some of these barriers emanate from the health service providers others arise as a result of patients’ demand for treatment at health service providers (Ghosh et al., 2013; Grundy & Annear, 2010; Russell, 2008). Better still, some are attributed to disparities in health service distribution, low socio-economic status, cultural beliefs and practices, religious differences, geographical accessibility and longer waiting times associated with treatment seeking at health providers (Blanchet et al., 2012; El Kahi et al., 2012; Mazumda et al., 2009; Crommett, 2008). With regard to this, the study sought to assess the various barriers that militate against effective and appropriate use of health services at various health service providers in the 139 http://ugspace.ug.edu.gh/ Akuapem-North Municipality. Information regarding respondents’ perceived barriers to health service use was mainly garnered through the qualitative facet of the study. This was to help identify the major difficulties respondents in both the urban and rural communities were confronted with in their quest to utilise health services and health service providers in their attempt to provide genuine and unquestionable services to their clients. Given that all respondents engaged with a health service provider for first actions taken whenever illness symptoms were perceived (see Table 4.5.1), the study sought the need to find out if respondents encountered any difficulty as they tried seeking treatment at the health service providers they usually visited for cure. Of the total, slightly more than half (55.6%) of the respondents said they did face certain barriers while the remaining 44.4% said they did not encounter any difficulty in relation to their health service use (see Table 5.4.1). Respondents’ difficulty in accessing health services was significant in general (χ2 = 73.005; df (1); p < 0.05) and also in both the urban (χ2 = 46.154; df (1); p = 0.000) and rural (χ2=28.932; df (1); p=.000) communities of the study area. This can be attributed to the fact that majority (59.0%) of the urbanites reported facing barriers in accessing formal health care services relative to their rural counterparts (41.0%). 140 http://ugspace.ug.edu.gh/ Table 5.4.1: Difficulty in accessing health services Place of Residence Urban Rural Total Difficulty in utilising health services N % N % N % Yes 69 59.0 45 51.1 114 55.6 No 48 41.0 43 48.9 91 44.4 Total 117 100.0 88 100.0 205 100.0 Chi-Squre (χ2) = 73.005, df (1), p = 0.000 Source: Field Work, 2015. After this, respondents were asked to indicate exactly the major barrier that hindered their health service access and use in the municipality. Participants mentioned a host of inhibitive factors including long distance to health service providers, longer wait times at especially formal health facilities, unavailability of formal health facilities, poor attitudes of health provider staff, lack of professional medical expertise, cost of transportation, and poor quality of service provided. Of the one hundred and fourteen (114) respondents that reported facing difficulties in utilising health services at various health service providers, majority (34.8%) of the urbanites indicated longer wait time spent at especially formal medical facilities as the major barrier to effective health service use. This is followed by those that were of the view that the services rendered at formal medical facilities are of poor quality (23.2%) with the least (2.9%) citing longer distance to health facilities and transportation cost concurrently (see Table 5.4.2). For the rural respondents, the majority (42.2%) indicated the unavailability of formal health facilities in their locality as what hindered them the most 141 http://ugspace.ug.edu.gh/ from seeking treatment at such facilities. Also, the long distance one needs to traverse before utilising formal health services (20.0%) followed distantly while the least (4.4%) cited the poor attitude of health service staff (see Table 5.4.2) as some of the factors that barred them from the use of especially formal health facilities. Table 5.4.2: Respondents’ major barrier to health service use Place of Residence Barriers Urban Rural Total Unavailability of health provider 3(4.3%) 19(42.2%) 22(19.3%) Distance to health provider 2(2.9%) 9(20.0%) 11(9.6%) Poor quality of service rendered 16(23.2%) 3(6.7%) 18(15.8%) Longer waiting times 24(34.8%) 5(11.1%) 29(25.4%) Transportation cost 2(2.9%) 7(15.6%) 9(7.9%) Poor attitude of health service staff 15(21.7%) 2(4.4%) 17(14.9%) Lack of professional medical expertise 7(10.2%) 0(0.0%) 7(6.1%) Total 69(100.0%) 45(100.0%) 114(100.0%) Source: Field Work, 2015. (Figures in bracket represent percentages in a column). Most of the respondents in the rural community cited unavailability of formal health facilities in their locality as the main reason why they engaged the more with informal medical healers whenever illness symptoms were perceived. During the in-depth interviews, most of the rural dwellers bitterly talked about the stress they go through before treatment seeking at Health Centres at nearby villages (Adawso and Kwamoso). This, according to them, comes with extra cost since they have to board vehicles before they can utilise formal health care. Hence, treatment is sought mainly at informal health providers 142 http://ugspace.ug.edu.gh/ such as herbalists and traditional birth attendants (TBAs) who are with them in the community and only go for care at formal health facilities when illness symptoms become severe and need special care. Formal health facilities in the municipality itself are skewed toward the core posing a major barrier to health care seeking for those in the periphery and rural areas. This influences the kind of health service provider an individual seek treatment at. A farmer had this to say: “There is no hospital, clinic, health centre or CHPS compound (formal medical facility) here.” But the nearest ones are at Adawso and Kwasmoso which are all very far from here. Also, there isn’t any drug/chemical shop here but there is one at Nyensiso (the village before ours). That is where I normally go for drugs on my bicycle. Over here one only prays that his/her illness does not become severe in the night. Though the village is located on the main Accra-Koforidua road getting a vehicle at night is difficult because of the activities of unscrupulous people. In the presence of all these one has no option than to go for some herbs from herbalists and the especially the pregnant women doing the same from untrained traditional birth attendants (TBAs) that are available and closest to us” (A 32-year old farmer at Korkormu, Individual interview). This confirms the findings of Aboagye & Agyemang (2013) in the Bosomtwe District in the Ashante region, Ghana and Mazumda et al. (2009) in India where pregnant women in the rural areas due to unavailability of formal medical facilities, long distance to these formal medical facilities and poor road networks linking the periphery to the urban centres either receive antenatal care and assisted deliveries from untrained traditional birth attendants or elderly women in the locality or not at all. 143 http://ugspace.ug.edu.gh/ Also, most of the rural respondents lamented on how long distance to formal health facilities and cost of transportation inhibit them from seeking treatment at such health service providers. Because these formal facilities are not located in their community but in the nearby communities, respondents had to travel over longer distances which comes at a cost before they can utilise health care services. This is extra cost and increases the overall cost of health service use. This is what a respondent had to say: “Getting a vehicle to any of the health centres nearest to us; either at Adawso or Kwamoso is not that difficult. The road linking our village to either of these villages is in good condition but you will not spend less than GHȼ7.00 on transportation. This, in addition to what you pay at the health facility, increases your overall cost of health service use” (A 26-year old woman at Korkormu, Individual interview). This is consistent with the findings of Mushtaq et al. (2011); Mazunda et al. (2009); Gotsadze et al. (2005) where more than half of the studied population sought care at informal medical providers (self-treated and visited traditional healers) as first actions taken due to over ten (10) kilometre distance they had to travel before utilising health care services. It also confirms the results of Yanagisawa et al. (2004) where treatment seeking at formal health facilities increased with decreasing distance in Cambodia. The study found none of the Cambodians residing over ten (10) kilometres away from formal health facilities to seek treatment at such facilities due to high transport cost. For respondents in the urban community, longer waiting times, poor attitude of staff and perceived low quality of service rendered especially by the only formal medical provider (Health Centre) compelled them to go for care at informal medical providers the more. Urbanites lamented of the longer hours they usually spent at especially the Health Centre 144 http://ugspace.ug.edu.gh/ in the community that induced them to go for treatment at other available sources. A driver had this to say: “I spent more than thirty (30) minutes the last time I went to the Health Centre here before a personnel attended to me just to check my temperature and pulse. I almost spent the whole day at the facility and could not go to work. So, if I can get the needed medicine at the drug/chemical shop which would not take me more than five (5) minutes and also would get cured of my illness, why not. This will enable me work for more hours” (A 46-year old driver at Adukrom, Individual interview). This is in accordance with the results of Annan et al. (2013) and Russell (2008) in Ghana and Fomundam et al. (2012) in Africa where patients mainly sought treatment at informal medical providers such as drug stores and fetish priests instead of care seeking at formal medical facilities due to shorter wait times at the former. Also, shorter wait times spent at private formal medical facilities made respondents opt for treatment seeking at such facilities at the expense of public formal facilities (van der Hoeven et al., 2012). Another major barrier urbanites spoke extensively about has to do with the poor attitude of health provider staff towards patients and the poor quality of service rendered. All these are in line with treatment seeking at the formal medical facility (Health Centre) which made them seek care at other available health providers. The kind and/or rate of maltreatment and insults received at especially the only health centre in the community from the staff demotivated patients from seeking treatment there. Also, respondents perceived the health service offered at the health centre (formal health facility) to be of poor quality hence indicated that they do not usually recommend treatment seeking at that facility to other community members whenever they perceive illness symptoms. A trader reiterated; 145 http://ugspace.ug.edu.gh/ “There is a drug/chemical shop here where whenever the man sees that your illness is complicating, takes a sample of your blood and conduct some tests before giving you medicines. This he normally does for especially children and I tell you he always gets it right. I don’t remember the last time this was done on me at the health centre. They only listen to you talking and by the time you are done the physician has also finished with the prescription” (A 36-year old trader at Adukrom, Individual interview). “Imagine a big man like me being shouted at and insulted by a young girl in the name of a nurse. The medicines I am always given on the ticket of health insurance are pain killers like paracetamol and the likes. For important medicines capable of curing the illness they always tell you they don’t have hence will prescribe them for you to buy outside. So why should I waste my precious time going to the health centre for treatment when I can get all the necessary medicines at the drug/chemical shops and get cured always when I take them” (A 56-year old teacher at Adukrom, Individual interview). This is in accordance with the findings of van der Hoeven et al. (2012) in South Africa and Khabiri et al. (2011) in Iran where more people especially in the urban areas were pushed to seek treatment at expensive private formal health facilities, self-treat or did not seek care at all due to low quality of care rendered by public health care facilities and the inhumane attitude of formal health provider staff. It also confirms the findings of Robyn et al. (2012) in Burkina Faso, Russell (2008) in Ghana and Sandman et al. (2000) in the United States of America where lack of trust and/or confidence, inhumane attitude of health care personnel and poor quality of service provided at health care facilities compelled patients 146 http://ugspace.ug.edu.gh/ to go for treatment at private formal health providers, engage in over-the-counter sales, and even visit faith healers for treatment. However, it contradicts the findings of Annan et al. (2013), Dako-Gyeke et al. (2013), Laar et al. (2013) in Ghana; Dominic et al. (2013) in India and Hjelm & Atwine, (2011) in Uganda where majority of the studied population rather visited formal health facilities for treatment due to their high quality of service rendered. Again, urbanites cited the lack of experienced medical expertise as one of the barriers that inhibit them from seeking treatment at the only formal medical provider (Health Centre) in the community. The absence of professional personnel with specialty such as gynaecologists, surgeons etc. at the health centre in the community induce urbanites to seek care at formal medical providers elsewhere. The study found most of the urbanites that go for treatment at formal medical providers to do that at the municipality’s hospital at Mampong-Akuapem. A respondent indicated: “I stopped going to the health centre here for treatment when I realised that no one is capable of treating me. This was made known to me when I last visited the facility. The woman who attended to me said the specialist to treat me is at Tetteh Quashie Memorial Hospital (the municipality’s hospital) hence referred me to him. Ever since, I have being going there for treatment” (A 29-year old woman at Adukrom, Individual interview). 5.5 Health service cost and financing The cost of health service use and the means by which individuals finance their health service use have been found to impede appropriate and effective treatment seeking 147 http://ugspace.ug.edu.gh/ especially at formal medical facilities (Asenso-Okyere et al., 1998; Andersen, 1995). In view of this, studies conducted elsewhere (El-Kak et al., 2009; Biswas et al., 2006; Cropley, 2004; Subba, 2004) and in Ghana (Dako-Gyek et al., 2013; Blanchet et al., 2012; Danso-Appiah et al., 2010) have acknowledged financial barriers as one of the major determinants of individuals’ choice and use of health service providers. For instance, Danso-Appiah et al. (2010) found that treatment seeking at formal medical facilities for schistosomiasis-related symptoms increased when cost of service use was borne by someone else other than the patient. This, according to the researchers, “showed some effect but no clear impact.” Similarly, Cropley (2004) reports in Belize, Central America that majority of mothers sought treatment at government hospitals and clinics for their malarial sick children because health service use at such formal facilities was free. Also, Subba (2004) reports in Nepal that a higher proportion (73.2%) of the studied population sought treatment at informal health providers such as drug store due to their inability to afford the cost of care at formal medical facilities. Given that health insurance and out-of-pocket payment are regarded globally as the main modes of health service financing, the study assessed respondents’ cost of service use and the various ways through which they financed their health service use in the municipality (see Table 5.5.1). Also, the study attempted to establish the relationship between respondents’ socio-demographic variables such as sex, age, educational level and income and the mode of health service financing. This was done to ascertain how specific actions taken when illness symptoms were perceived were financed and by what category of respondents in the municipality. The results are illustrated in Table 5.5.3. 148 http://ugspace.ug.edu.gh/ From Table 5.5.1, majority (47.8%) of the respondents reported spending below GHȼ10.00 followed closely by 45.9% that spent between GHȼ10.10 to GHȼ50.00 while the least (6.3%) spent above GHȼ50.00 the last time they visited a health facility for care (see Table 5.5.1). This implies that respondents in the municipality compared with those in the Dangme West Municipality (Russell, 2008) in the same country spent less in using health services but comparatively, expensive to patients who sought care at government hospitals and clinics in Belize, Central America (Cropley, 2004). The cost of health service use, in general, was significant according to the Chi-Square goodness-of-fit test (χ2 = 67.327, df (2), p < 0.05) and also in both the urban (χ2 = 41.692, df (2), p = 0.000) and rural (χ2 = 51.841, df (2), p = 0.000) communities selected for the study (see Table 5.5.1). 149 http://ugspace.ug.edu.gh/ Table 5.5.1: Cost of and payment for health service use Variables Place of residence Urban Rural Total N % N % N % Cost of health service use < GHȼ10.00 40 34.2 58 65.9 98 47.8 GHȼ10.10 GHȼ50.00 67 57.3 27 30.7 94 45.9 >GHȼ50.00 10 8.5 3 3.4 13 6.3 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 67.327, df (2), Sig.=0.000 Mode of Payment Self 31 26.5 30 34.1 61 29.8 NHIS 72 61.5 44 50.0 116 56.6 Spouse 1 0.9 9 10.2 10 4.9 Parent 9 7.7 5 5.7 14 6.8 Relative 4 3.4 0 0.0 4 2.0 Total 117 100.0 88 100.0 205 100.0 Chi-Square (χ2) = 221.561, df (4), Sig.=0.000 Source: Field Work, 2015. On the part of respondents’ mode of payment for health service use, NHIS (56.6%) as a mode of health service financing predominated confirming the findings of Annan et al. (2013); Blanchet et al. (2012); El-Kak et al. (2009); and Roy et al. (2004) where health insurance was reported as the major payment mode for health care service use. This was 150 http://ugspace.ug.edu.gh/ followed by respondents that financed their health service use themselves (29.8%) while the least (2.0%) said a relative paid for their health service use (see Table 5.5.1). The means through which respondents financed their health service use became significant according to the Chi-square goodness-of-fit test (χ2 = 221.561; df (4); p < 0.05) resonating with the findings of Saeed et al. (2013) that an individuals’ choice and use of health service at a particular health provider is dependent on the mode of payment for the service use. Table 5.5.2: Cost of health service use by health facility attended when ill Health facility attended when ill Amount spent (GHȼ) IMPs FMPs Total <10.00 78(57.8%) 20(28.6%) 98(47.8%) 10.10-50.00 50(37.0%) 44(62.9%) 94(45.9%) >50.00 7(5.2%) 6(8.6%) 13(6.3%) Total 135(100.0%) 70(100.0%) 205(100.0%) Chi-Square (χ2) = 15.761, df (2); p = 0.000 Source: Field Work, 2015. Given that cost of health service use influence individuals’ health seeking behaviour (Atuyambe et al., 2009; El-Kaka et al., 2009; Adanu et al., 2008; Russell, 2008) the study sought to establish the relationship between respondents’ cost of health service use and their health seeking behaviour. This was to ascertain the particular health provider at which cost of treatment was higher and vice versa. The results are shown in Table 5.5.2. As substantial proportions of the studied population visited informal medical providers due to their inability to afford treatment cost at formal health facilities (Adanu et al., 2008; Subba, 2004), majority of pregnant adolescents in Uganda (Atuyambe et al., 2009) likewise poor 151 http://ugspace.ug.edu.gh/ and better-off people in Cambodia (Yanagisawa et al., 2004) sought care at informal health providers because no cost was borne in utilizing them. The study found the cost of health service use at informal medical providers to be less expensive relative to treatment seeking at formal health facilities. This confirms the findings of Atuyambe et al. (2009), and El- Kak et al. (2009) but in contrast to that of Nahar (2010) and Yanagisawa et al. (2004). From Table 5.5.2, majority (57.8%) of the respondents that sought treatment at informal medical providers spent less than GHȼ10.00 as against 28.6% on the part of formal medical facilities. Also, for respondents that sought treatment at formal health facilities, majority (71.5%) reported spending more than GHȼ10.00 compared with those that visited informal providers (42.2%) for treatment (see Table 5.5.1). Cost of health service use, in general, significantly predicted individuals’ health seeking behaviour in the municipality (χ2 = 15.761, df (2); p = 0.000). This buttresses the finding of Russell (2008) in Ghana where a significant association was established between cost of service use and health seeking behaviour. 152 http://ugspace.ug.edu.gh/ Table 5.5.3: Respondents’ place of residence and the health facility attended when ill by mode of payment for health service use. Factors Payment for health service use Total Self NHIS Spouse Parent Relative Place of residence Urban 31(26.5%) 72(61.5%) 1(0.9%) 9(7.7%) 4(3.4%) 117(100.0%) Rural 30(34.1%) 44(50.0%) 9(10.2%) 5(5.7%) 0(0.0%) 88(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 4(2.0%) 205(100.0%) Chi-Square (χ2) = 14.506 df (4) Sig. = .006 Health facility attended when ill IMPs 51(68.9%) 0(0.0%) 8(10.8%) 12(16.2%) 3(4.1%) 74(100.0%) FMPs 10(7.6%) 116(88.6%) 2(1.5%) 2(1.5%) 1(0.8%) 131(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 4(2.0%) 205(100.0%) Chi-Square (χ2) = 21.125 df (4) Sig. = .000 Source: Field Work, 2015. (Figures in bracket represent percentages in a row) Of the 116 respondents that financed their health service use through NHIS, 72 (62.1%) were from the urban community with the remaining proportion (37.9%) being from the rural community. Surprisingly, none of the respondents from the rural community, where family ties are considered very high, had a relative paying for their health service use (see Table 5.5.3). Nevertheless, mode of payment for health service use was significant in both the urban (χ2 = 149.795; df (4); p = 0.000) and rural (χ2 = 45.727; df (3); p = 0.000) communities. Respondents in the qualitative facet of the study cited cost saving as the main 153 http://ugspace.ug.edu.gh/ reason for accessing health care services using NHIS at the expense of poor quality of service and longer waiting times. Also, an overwhelming proportion of 88.6% of the respondents in both communities that sought treatment at formal medical providers whenever illness symptoms were perceived paid for their health service use through NHIS confirming the findings of Annan et al. (2013) and Blanchet et al. (2012). This is followed by those that paid for their health service use themselves (7.6%) while the least (0.8%) said a relative paid for their health service use (see Table 5.5.3). This finding contradicts that of Danso-Appiah et al. (2010) where respondents were highly probable to seek treatment at formal medical providers when someone either than themselves paid for health care. For respondents that sought treatment at informal medical providers as first action taken when illness symptoms were perceived in both communities, the study found the majority (68.9%) to foot their health service expenses themselves followed by those whose parents (16.2%) paid for their health service use. However, it is worth noting that all patients that sought treatment at formal medical facilities paid for their health service use through NHIS whereas none of the patients that sought treatment at informal medical providers paid for their health service use through NHIS (see Table 5.5.3). Once again, as reported by Saeed et al. (2013) this showed statistically significant (χ2 = 21.125; df (4); p < 0.000). This became apparent during the individual and institutional interviews conducted for the study. The health personnel at the Adukrom Health Centre indicated: “It is very rare seeing an uninsured person coming for treatment here because per our records, almost all patients who come here for treatment, in one way or the other, come with NHIS. Ever since I came here some five years ago, I can count the 154 http://ugspace.ug.edu.gh/ number of patients who have been here without NHIS. And I don’t think you will get more than twenty (20). Yes; I mean it” (Health personnel at Adukrom Health Centre, Individual interview). Also, a drug/chemical shop sales personnel had this to say: “Whether the drug one is looking for is expensive or not NHIS cannot be used to purchase drugs here simply because this drug/chemical shop is not an NHIS accredited on hence if you need drugs of any kind from this shop, then you are prepared to pay for them instantly. I have heard that some pharmacy shops especially those in the big cities accept NHIS as a means of paying for drugs bought but for Adukrom I don’t think this can be the case” (A drug/chemical shop personnel at Adukrom, Individual interview). Given that health insurance and out-of-pocket were the two main modes of health service payment in the municipality, majority of both males (57.6%) and females (55.2%) paid for their health service use through NHIS. The remaining proportions (males (42.4%) and females (44.8%)) financed their health service use through out-of-pocket payment (see Table 5.5.4). The study established a significant relationship between respondents’ sex and their mode of health service use financing in general according to the Chi-square test of association (χ2 = 16.045; df (4); p < 0.05). However, no significant association was established between sex and mode of health service use financing in the urban community (χ2 = 2.489; df (4); p = .647) but such a significant association was established in the rural community (χ2 = 11.266; df (3); p = .010). This is an indication that a difference exists in terms of sex and mode of health service use financing in the municipality. 155 http://ugspace.ug.edu.gh/ Also from Table 5.5.4, NHIS is identified as the main health service payment mode for respondents in both study communities irrespective of respondents’ age. However, age was not found to be significant with the various modes of health service payment in the selected study communities (χ2 = 25.913; df (16); p > 0.05) though majority of the aged – those who are 50 years and above – (65.1%) than the young – those below 50 years – (54.3%) financed their health service use through the NHIS. Like sex and educational level, the study established a significant relationship between respondents’ average monthly income and the various modes through which they financed their health service according to the Chi- square test of association (χ2 = 6.920; df (4); p < 0.05) as depicted by Table 5.5.4. Also, NHIS as a mode of health service financing, predominated among the various income categories, where it was high for both the above GHȼ500.00 (68.8%) and below GHȼ500.00 (52.9%) income groups (see Table 5.5.4). This may be the case because the study found majority of the respondents to be registered with the scheme (92.2%) and have their membership status to be active (86.2%). 156 http://ugspace.ug.edu.gh/ Table 5.5.4: Socio-demographic variables by payment for health service use Variables Payment for health service use Total Self NHIS Spouse Parent Relative Sex/Gender Male 40(33.9%) 68(57.6%) 0(0.0%) 7(5.9%) 3(2.5%) 118(100.0%) Female 21(24.1%) 48(55.2%) 10(11.5%) 7(8.1%) 1(1.1%) 87(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 4(2.0%) 205(100.0%) Chi-Square(χ2) = 16.045 df = 4 Sig. = .003 Age(years) 20-29 20(35.1%) 24(42.1%) 2(3.5%) 10(17.5%) 1(1.8%) 57(100.0%) 30-39 21(32.3%) 39(60.0%) 2(3.1%) 2(3.1%) 1(1.5%) 65(100.0%) 40-49 12(30.0%) 25(62.5%) 2(5.0%) 1(2.5%) 0(0.0%) 40(100.0%) 50-59 3(13.0%) 15(65.4%) 3(13.0%) 1(4.3%) 1(4.3%) 23(100.0%) 60+ 5(25.0%) 13(65.0%) 1(5.0%) 0(0.0%) 1(5.0%) 20(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 42.0%) 205(100.0%) Chi-Square(χ2) = 25.913 df (16) Sig. = .055 Educational Level None 8(32.0%) 7(28.0%) 7(28.0%) 2(8.0%) 1(4.0%) 25(100.0%) Prim/JHS 33(33.8%) 55(56.1%) 3(3.1%) 5(5.1%) 2(2.0%) 98(100.0%) Secondary 12(21.4%) 37(66.1%) 0(0.0%) 6(10.7%) 1(1.8%) 56(100.0%) Tertiary 8(30.8%) 17(65.4%) 00.0%) 1(3.8%) 0(0.0%) 26(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 4(2.0%) 205(100.0%) Chi-Square(χ2) = 41.896 df (12) Sig. = .000 157 http://ugspace.ug.edu.gh/ Average monthly income (GHȼ) Below 50 14(37.8%) 16(43.2%) 5(13.5%) 2(5.4%) 0(0.0%) 37(100.0%) 51-100 12(46.2%) 9(34.6%) 2(7.7%) 3(11.5%) 0(0.0%) 26(100.0%) 101-200 1128.2%) 21(53.9%) 2(5.1%) 5(12.8%) 0(0.0%) 39(100.0%) 201-500 11(20.0%) 37(67.3%) 1(1.8%) 3(5.3%) 3(5.3%) 55(100.0%) 501-1000 11(31.4%) 22(62.9%) 0(0.0%) 1(2.9%) 1(2.9%) 35(100.0%) 1000 2(15.4%) 11(84.6%) 0(0.0%) 0(0.0%) 0(0.0%) 13(100.0%) Total 61(29.8%) 116(56.6%) 10(4.9%) 14(6.7%) 4(2.0%) 205(100.0%) Chi-Square(χ2) = 32.663 df = 20 Sig. = .037 Source: Field Work, 2015. (Figures in bracket represent percentages in a row) The study also examined how respondents’ educational background impacted their means of health service financing. This was done in the sense that educational level of people have been found to influence greatly their income levels (Owusu-Dako & Smith, 2005) and access to formal health service use (Asenso-Okyere et al., 1998; Awusabo-Asare & Anarfi, 1997). Education, in general, showed significant relationship with the various modes of health service use financing (χ2 = 41.896, df (12), p < 0.01). This is illustrated by Table 5.5.4. The study also established a significant relationship between education and the various modes through which respondents financed their health service us in both the urban (χ2 = 36.036, df (12), p = .000) and rural (χ2 = 29.177, df (12); p = .007) communities selected for the study. Once again, NHIS was the predominant health service payment mode for respondents with some form of formal education (60.6%) in the study area while for those with no form of formal education, out-of-pocket payment (72.0%) dominated (see Table 5.5.4). This implies that respondents with some form of formal education than those 158 http://ugspace.ug.edu.gh/ with no form of formal education are gainfully employed hence have registered with the NHIS and are also literate health-wise, thus understand the implications of being enrolled on the NHIS in Ghana. From the individual interviews, availability, proximity and flexibility in terms of payment were identified as the main reasons that accounted for this pattern of health service financing for all respondents but especially those in the rural community. Rural dwellers cited the availability of informal medical providers in their locality and the flexibility regarding payment for health service use at informal health providers as the main reasons why out-of-pocket payment was their main means through which health service use was financed. On the part of the urbanites, the availability of a health centre in their community, which is an accredited NHIS facility, accounted for the high use of NHIS as the main health service use financier. 5.6 Summary This chapter examined the most preferred health service providers of individuals and the distance to health service providers from respondents’ house in the Akuapem-North Municipality. It also assessed the determinants of individuals’ health seeking behaviour and the restrictive barriers that hindered individuals from utilising service at specific health service providers. Lastly, the chapter examined the cost of health service use and the modes of financing health service use in the municipality. Respondents in both the urban (89.7%) and rural (76.2%) communities cited formal medical facilities (hospitals, clinics, health centres, and CHPS) as their most preferred health service providers. The Chi-Square goodness-of-fit conducted showed that 159 http://ugspace.ug.edu.gh/ respondents’ most preferred health service provider was significant in both the urban and rural communities selected for the study. Such reasons as provision of quality service, have the best medical personnel, and always get cured when care is sought were given by respondents irrespective of their place of residence as to why these are their most preferred health providers. The unavailability of these most preferred health providers in both communities made the majority of the urbanites (63.2%) and rural dwellers (67.0%) unable to seek treatment at such facilities. Also, all the rural than the urban (88.9%) respondents had the closest health service providers to their houses to be within a kilometre range. Whereas all the rural dwellers had a traditional healer closest to their houses, majority (53.0%) of the urbanites had a drug/chemical shop closest to their houses followed by 31.6% that had a health centre closest to their house with the least (0.9%) having a prayer camp closest to their houses. The use of vehicles predominated as the means of transportation in both the urban (51.3%) and rural (62.5%) communities where the majority (44.9%) spent GHȼ5.00 and below as transport cost. This was borne mainly by the respondents themselves. Travel time to health service providers is relatively less as the majority (82.9%) claimed spending less than thirty minutes before reaching health service providers where care was mainly sought. With regard to waiting time, majority (38.0%) reported spending as long as over an hour before treatment is sought with more of the rural dwellers experiencing this. The qualitative facet of the study identified longer wait times to be characteristic of treatment seeking at formal than informal health facilities in the study communities. 160 http://ugspace.ug.edu.gh/ Regarding the determinants of individuals’ health seeking behaviour in the municipality, the study found health insurance, severity of illness, and distance to health providers to significantly predict their health seeking behaviour in the municipality. Also, the skewed nature of especially formal health facilities and militating factors such as long distance, high cost of service, poor nature of roads, means of transport, longer wait times at formal health facilities barred individuals’ health seeking behaviour in the municipality. The negative effect of these were on the ascendancy in the rural than the urban community hence explains the trend and pattern of health seeking behaviour by respondents in both communities of the study. The individual interviews conducted vividly affirmed the fact that the unavailability of formal medical providers, type and nature of illness, flexible terms of payment were some of the reasons rural dwellers mainly visited informal medical healers for treatment whereas perceived severity of illness, perceived quality of service offered and longer wait time led the bulk of reasons why urbanites also engaged the more in over-the-counter sales of medicines. Again, out-of-pocket payment and NHIS were identified as the main modes of health service financing in the municipality with the latter being the predominant. The study empirically established a significant relationship between sex and educational level and the various modes of health service financing in the municipality. 161 http://ugspace.ug.edu.gh/ CHAPTER SIX SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 6.0 Introduction This chapter presents a summary of the major findings of the study. It draws conclusions on the findings, and the theoretical and/or conceptual framework that underpropped the direction of the study and suggests areas for further research. Lastly, the chapter provides recommendations based on the findings of the study. 6.1 Summary of key findings The study mainly focused on exploring and comparing the health seeking behaviour of individuals in both urban and rural communities in the Akuapem-North Municipality of Ghana. Specifically, it looked at the most preferred health seeking system of urban and rural dwellers in the municipality, and the relationship between individuals’ socio- demographic characteristics and their health seeking behaviour. It also focused on the factors that influence individuals’ choice and use of particular health service providers in the municipality, and the barriers to health service use and mode of health service financing in the study area. The study adopted a mixed method approach in order to gain comprehensive understanding of the issues examined. A four-step multi-stage cluster sampling was used to select two hundred and five (205) respondents for the individual survey whereas a simple random sampling method was used to select twenty (20) respondents for the qualitative facet of the study. 162 http://ugspace.ug.edu.gh/ General body pains/headache (33.7%) was identified as the most predominant perceived illness symptom that respondents in both communities suffered from followed by malaria/fever (23.9%) with acute respiratory infections (1.4%) being the least. The major means of diagnosis was by oneself (50.3%) followed promptly by those who were diagnosed by a professional medical personnel (28.3%). Majority of respondents that did not know of the aetiology of their illness likewise majority of those that suffered from malaria/fever (42.9%) and all those that suffered from ARIs got diagnosed by a professional medical personnel. Contrary, majority of the respondents that suffered from stomach/abdominal pains (62.5%) and general body pains/headache (82.6%) got diagnosed by themselves. As majority of the respondents that suffered from general body pains/headache and malaria/fever perceived them to be less severe and not severe hence sought treatment at informal medical providers, more of the respondents that suffered from ARIs and stomach/abdominal pains perceived them to be severe and more severe hence went for care at formal medical facilities. The study found both the formal and informal medical systems to be available in the urban community whereas only the informal medical system was found available in the rural community. All the rural dwellers than the urbanites (88.9%) had the closest health service provider to be within a kilometre range to their houses. The availability of these health service providers warranted a very high utilisation in both the urban (96.6%) and rural (81.8%) communities. The Chi-square goodness-of-fit test conducted (χ2 = 185.488, df (1), p = 0.000) showed that the availability of health service providers significantly influenced individuals’ choice and use of health services in the Akuapem-North Municipality as has been emphasised by Kroeger (1983) in his health service utilisation model. 163 http://ugspace.ug.edu.gh/ Health seeking behaviour in the Akuapem-North Municipality was measured as the health service provider/facility respondents attended whenever illness symptoms were perceived in a twelve month period. Health seeking at informal health providers (65.9%), in general, dominated likewise in both the urban (65.0%) and rural (67.0%) communities selected for the study. The health facility respondents attended whenever illness symptoms were perceived was significant (χ2 = 20.610, df (1), p = 0.000) according to the Chi-square goodness-of-fit test. Therefore, the acceptance of the null hypothesis since there was no significant difference regarding the health facility rural and urban dwellers attended whenever illness symptoms were perceived but a rejection of the alternate hypothesis. Respondents cited the formal medical system as their most preferred health seeking system in both the urban (89.7%) and rural (76.2%) communities of the study. Such reasons as provision of quality services, having the best medical personnel and always getting cured of ones’ illness were given as the main reasons why respondents prefer going for treatment from formal medical facilities. As indicated by Andersen & Newman (1973) and further corroborated by Kroeger (1983), certain predisposing, enabling, need and health system characteristics determine individuals’ choice and use of health services. The study showed confirming and contrasting results regarding these factors and health seeking behaviour. Whereas age was found not to significantly influence individuals’ health seeking behaviour, the study established a significant association between sex and the health facility respondents attended when ill according to both the Chi-square test of association ((χ2 = 4.055, df (1), p = 0.044) and the multinomial logistic regression model (p < 0.01). For education and income, the study established a significant difference between them and the health facility 164 http://ugspace.ug.edu.gh/ they attended as a significant relationship was established between respondents’ educational level (χ2 = 7.499, df (3), Sig. = 0.034), (p < 0.01) and their income (χ2 = 7.499, df (3), Sig. = 0.034), (p < 0.01) and their health seeking behaviour in the rural community but not in the urban community according to both the Chi-square test of association and the multinomial logistic regression model employed for the study. Though respondents with and without formal education in both communities visited informal medical providers than formal medical facilities for treatment, their rural counterparts were twenty (20.57) times more likely to take such action. Also, with regard to income, rural low income earners (˂ GHȼ500.00) likewise their rich (˃ GHȼ500.00) counterparts were more likely to attend an informal medical facility than a formal medical facility for treatment and also than their urban counterparts. Also, other factors such as health insurance, severity of illness, and distance to health providers significantly predicted individuals’ health seeking behaviour in the Akuapem- North Municipality as has been indicated by the frameworks of Andersen & Newman (1973) and Kroeger (1983). The study identified a positive association between health insurance, severity of illness and treatment seeking at formal medical facilities. As none of the insured was found seeking treatment at formal health facilities majority of the respondents with severe and more severe illness symptoms sought care at formal medical facilities. The urban insured were sixteen times (16.78) whereas the rural insured were seventeen times (17.87) more likely to seek treatment at formal health facilities but less likely to seek treatment from informal health facilities. The Chi-square test of association (χ2 = 8.999, df (1), p = 0.003) likewise the multinomial logistic regression model (p < 0.01) conducted both show that a significant relationship was established between health 165 http://ugspace.ug.edu.gh/ insurance and the health facility respondents attended when ill. Also, urbanites whose illness symptoms were not severe (-2.89) and less severe (-4.56) were less likely to go for treatment at formal health facilities while their rural counterparts were twenty-two times (- 22.31) and two times (-2.53) less likely to seek treatment at formal medical facilities, whereas those with severe and more severe illness symptoms were more likely to seek treatment at formal medical facilities. Individuals’ illness severity was found to significantly predict their health seeking behaviour per the Chi-square test of association (χ2 = 53.110, df (3), p = 0.000) and the multinomial logistic regression (p < 0.01) model. Therefore, the study rejects the null hypothesis which states that no significant difference relationship exist between patients’ illness severity and the type of health facility rural and urban dwellers sought treatment from whenever illness symptoms were perceived but accepts the alternative hypothesis which state otherwise. In the same way, more of both the urbanites and rural dwellers whose illness symptoms were not severe and less severe waited for over five days before seeking some form of treatment likewise those with severe and more severe illness symptoms. Also, a positive association between distance and health seeking behaviour in the municipality was identified. Majority of both the urban (66.6%) and rural (67.0%) sought treatment at health service providers closest to their houses. Thus, urbanites with an informal health provider closest to their houses were more likely (1.29) to seek treatment there while rural dwellers without a formal medical facility in their locality were less likely (-0.73) to seek treatment there. Distant significantly influenced individuals’ health seeking behaviour according to the Chi-square test of association (χ2 = 4.631, df (1), p = 0.013) and the multinomial logistic regression (p < 0.05) model for the study. Quality of service, on 166 http://ugspace.ug.edu.gh/ the other hand, did not significantly influence individuals’ health seeking behaviour in the municipality likewise waiting time though both urbanites (3.06) and rural dwellers (0.86) were more likely to spend less than thirty (30) minutes at health facilities the last time they went there for treatment. Being widely documented, the study assessed the various barriers that impeded patients’ choice and use of health services in the municipality and a host of factors were identified. For the urbanites, such barriers as longer waiting times, low quality of service rendered, poor attitude of staff and lack of professional medical expertise restricted their utilisation of especially formal health facilities in the municipality. Urbanites lamented especially about the services of the health centre in the community which impelled them to mainly seek treatment at drug/chemical shops, which they expressed much satisfaction with their services and claimed to be cured of their illnesses whenever treatment was sought there. On the part of the rural inhabitants, indicators such as unavailability of health facilities, distance to health facilities, transportation cost and longer waiting times were the major barriers that inhibited them from utilising especially formal health facilities. These made them engage the more with informal medical providers like traditional healers as first action taken whenever illness symptoms were perceived. The study found the cost of health service use in the municipality to be relatively less expensive as majority (47.8%) of the respondents reported spending below GHȼ10.10. Those who spent more than GHȼ100.00 (6.3%) formed the least. The cost of health service use, per the Chi-square goodness-of-fit test, was significant (χ2 = 67.327, df (2), p < 0.05) in the study area. This means that the cost of health service use determined whether individuals utilise health services or not in the Akuapem-North Municipality. Again, out- 167 http://ugspace.ug.edu.gh/ of-pocket payment and NHIS were identified as the main modes of health service use financing in the municipality with the latter being the predominant. The study empirically established a significant relationship between sex (χ2 = 16.045, df (4), p = 0.003), educational level (χ2 = 41.896, df (12), p = 0.000) and income (χ2 = 32.663, df = 20, p = 0.037) and the various modes of health service financing in the municipality. 6.2 Conclusion As indicated by both Andersen & Newman (1973) and Kroeger (1983) in their frameworks, findings from this study show that certain predisposing factors as sex, and education and enabling factors as income and health insurance as well as some disorder characteristics as illness severity influence the choice and use of health services. Also, health system characteristics such as availability of health service providers, health service cost and financing determine individuals’ health seeking behaviour in the municipality as suggested by Kroeger’s (1983) model of health service utilisation. As again documented by Andersen & Newman (1973) and Kroeger (1983) and corroborated by this study, individuals’ choice and use of either formal or informal medical facilities or deciding not to go for treatment at all are often dependent on perceived morbidity. The study therefore concludes that a positive correlation exist between perceived morbidity, illness response, access to and use of health services as emphasised in the frameworks of Andersen & Newman (1973) and Kroeger (1983). As a major finding, treatment seeking from informal medical providers (drug/chemical shops, drug peddlers, traditional healers and home remedy applications) predominated the study in both the urban and rural communities in the Akuapem-North Municipality though substantial proportions 168 http://ugspace.ug.edu.gh/ of the people sought health from the formal medical system (hospitals, clinics, health centres and CHPS compounds). A great limitation of the current study is its inability to include social distance and cultural variables as factors that influence individuals’ health seeking behaviour as included in Kroeger’s (1983) model. It is suggested therefore that future studies should examine how the interplay of social distance and cultural factors influence peoples’ choice and use of health services. Also, as has been identified by the current study, further studies should be conducted regarding the nature and type of diseases and the utilisation of health service for a more vivid insight and understanding. Again, the study found respondents in the urban community (Adukrom) with a formal medical provider available (health centre) to rather seek treatment the more at informal medical providers than those in the rural area (Korkormu) where only informal medical providers are available. With this, there is the need for future studies to redirect attention to examining the health seeking behaviour of people in urban communities to unearth the underlying factors. More of the insured in the study were found seeking treatment from informal health facilities than from formal medical facilities hence it is suggested that future studies centre on health insurance and how it imparts health service utilisation. 6.3 Recommendations  It is recommended that more research works be conducted in the study area with larger sample sizes representative of the municipality’s and/or the nation’s population, as the current study’s sample size is considered small, to help vividly unearth the realities regarding the health seeking behaviour of people from different 169 http://ugspace.ug.edu.gh/ spatial locations to facilitate the design and implementation of nation-wide health policies and programmes that will bring to bear the health needs and concerns of the rural inhabitants who are mainly at a disadvantage as far as social intervention policies are concerned.  The study discovered that majority of the people in the selected communities especially those in the rural area have schooled for below six (6) years. This is an indication of an unhealthy community since majority of the population in the study area will be illiterates health-wise, lacking the ability to read and comprehend health information. It is therefore recommended, based on this, that educational policies and programmes such as free senior high school (SHS) education and the establishment of community senior high schools (CSHS) precisely in the study communities be pursued by all concerned to assist Junior High School graduates to upgrade themselves. This will make them ‘health-literates,’ thereby reducing the high rates of illness prevalence in the study area. Also, the people must be educated on the need to go for health services especially at formal medical facilities as early as illness symptoms are perceived to help reduce mortality from preventable and curable morbidities as the majority are seen to wait for five and more days before some form of treatment is sought which is even mainly done at informal medical providers especially over-the-counter purchases of drugs.  The study also discovered that majority of enrolees of the National Health Insurance Scheme (NHIS) with their insurance status to be active mainly resorted to treatment seeking at informal medical providers such as drug/chemical shops; traditional healers; drug peddlers and home remedy applications than health 170 http://ugspace.ug.edu.gh/ seeking at the formal medical system in both the rural and urban communities. Distance to and availability of especially formal health facilities are found to be the main determinant factors preventing majority of the people to seek health at formal medical facilities albeit insured. This rather facilitated treatment seeking at informal medical providers. Therefore, the study recommends that government and other stakeholders join forces to establish more of the formal medical facilities such as Clinics, Health Centres and CHPS compounds or Health Posts especially in the rural areas where none exist and equip them, in addition to already existing ones at nearby villages, with adequate resources to make them attractive to the people. This will reduce longer travel distances and times to access service at these formal medical facilities hence, reduce the problem of expensive health expenditure especially in the rural communities of the Akuapem-North Municipality.  To curb the problem of longer waiting time at especially formal medical facilities in the municipality, it is recommended that the necessary and adequate logistics as well as qualified medical personnel be posted to such health facilities to beef up existing hands. This will shorten the minutes patients spend at the facility hence help increase the quality of service rendered in the municipality.  Finally, the study recommends, as has been echoed by other researchers, the integration of both the formal and informal medical systems in our health system delivery, due to extensive visits to these informal medical healers by the populace, to help improve and assist health service delivery in the country especially in the Akuapem-North Municipality. To be able to do this, the informal medical providers 171 http://ugspace.ug.edu.gh/ must be given the needed education and training likewise retraining to enable them to diagnose and prescribe medication for the people. 172 http://ugspace.ug.edu.gh/ REFERENCES: Abdulraheem, I.S and Parakoyi, D.B., (2009): Factors affecting mothers’ healthcare‐ seeking behaviour for childhood illnesses in a rural Nigerian setting. Early Child Development and Care, 179:5, 671-683, DOI: 10.1080/03004430701500885 Aboagye, E., and Agyemang, O.S., (2013): Maternal Health-Seeking Behaviour: The Role of Financing and Organization of Health Services in Ghana. Global Journal of Health Science; Vol. 5, No. 5, pp. 67-79. Abubakar, A., van Baar, A., Fischer, R., Bomu, G., Gona, J.K and Newton, C.R., (2013): Socio-Cultural Determinants of Health-Seeking Behaviour on the Kenyan Coast: A Qualitative Study. PLoS ONE; 8(11): e71998. Adanu, R.M.K., Hill, A.G., Seffah, J.D., Darko, R., Anarfi, J.K., and Duda, R.B., (2008): Sexually Transmitted Infections and Health Seeking Behaviour among Ghanaian Women in Accra. African Journal of Reproductive Health; Vol. 12, No. 3, pp. 151- 158 Andersen, R.M. (1995): Revisiting the behavioral model and access to medical care: does it matter? Journal of health and social behavior: 36, 1-10. Andersen, R.M., & Newman, J.F. (1973): Societal and individual determinants of medical care utilization in the United States. Milbank Memorial Fund Quarterly: 51, 95- 124. Afolabi, M.O, Daropale, V.O., Irinoye, A.I and Adegoke, A.A., (2013): Health-seeking behaviour and student perception of health care services in a university community in Nigeria. Health; Vol. 5, No. 5, pp. 817-824 173 http://ugspace.ug.edu.gh/ Agyepong, I. A, and Adjei, S., (2008): Public Social Policy Development and Implementation: A case study of the Ghana National Health Insurance Scheme. Oxford University Press, USA. Ahmed, S.M., (2005): Exploring health-seeking behaviour of disadvantaged populations in rural Bangladesh.” Karolinska University Press; Stockholm, Sweden Ahmed, S.M., Tomson, G., Petzold, M and Kabir, Z.N., (2005): Socioeconomic Status Overrides Age and Gender in Determining Health-Seeking Behaviour in Rural Bangladesh. Bulleting of the World Health Organisation, 83(2) Ajiboye, O.E and Adebayor, K.A., (2012): Socio-Cultural factors affecting pregnancy outcome among the Ogu speaking people of Badary area of Lagos State, Nigeria. International Journal of Humanities and Social Science, Vol. 2, No. 4, pp. 133-144 Akuapem-North District Assembly, (2006): District Annual Report. Akuapem-North District Assembly, (2010): District Annual Report. Akuapem-North Municipal Health Directorate (2013): Akuapem-North Municipal Health Report. Al-Mandhari, A., Al-Adawi, S., Al-Zakwani, I., Al-Shafaee, M and Eloul, L., (2008): Impact of Geographical Proximity on Health Care Seeking Behaviour in Northern Oman. Sultan Quboos University Medical Journal; 8(3): pp 310-318. 174 http://ugspace.ug.edu.gh/ Annan, A.A., Singh, A., Dogbe, J.A., Asante, D., and Owusu-Dabo, E., (2013): Health- Seeking Behavior of Tuberculosis Patients and Related Factors in the Central Region of Ghana. Journal of Science and Technology, Vol. 33, No. 3, pp. 27-38 Anwar, M., Green, J., and Norris, P., (2012): Health-Seeking Behaviour in Pakistan: A narrative review of the existing literature. Public Health, 126 (2012) 507-517. Also available at: doi:10.1016/j.puhe.2012.02.006 Asenso-Okyere, W.K., Anum, A., Osei-Akoto, I., and Adukonu, A., (1998): Cost recovery in Ghana: are there challenges in health care seeking behaviour? Health Policy and Planning; 13(2): 181-188. Atuyambe, L., Mirembe, F., Annika, J., Kirumira, E.K., and Faxelid, E., (2009): Seeking safety and empathy: Adolescent health seeking behaviour during pregnancy and early motherhood in Central Uganda. Journal of Adolescence; 32, 781-796 Awusabo-Asare, K., and Anarfi, J.K., (1997): Health-seeking behaviour of persons with HIV/AIDS in Ghana. Health Transition Review, Supplement to Vol. 7, 243-256 Biswas, P., Kabir, Z.N., Nilson, J and Zaman, S., (2006): Dynamics of Health Care Seeking Behaviour of Elderly People in Rural Bangladesh. International Journal of Ageing and Later Life: 1(1): 69-89. Blanchet, N.J., Fink, G and Osei-Akoto, I., (2012): The Effects of Ghana’s National Health Insurance Scheme on Health Care Utilisation. Ghana Medical Journal; Vol. 46, No. 2, pp 76-84. 175 http://ugspace.ug.edu.gh/ Chakraborty, N., Islam, M.A., Chowdhury, R.I., Bari, W and Akhter, H.H., (2003): Determinants of the use of Maternal Health Services in Rural Bangladesh. Health Promotion International; Vol. 18, No. 4, pp 327-337. Crommett, M., (2008): Confronting Religion: Perceptions and Health-Seeking Behaviours of Devout Adolescents when faced with a Sexually Transmitted Infections in Ghana. Georgetown University Journal of Health Studies; Vol. 5, No. 1, pp. 105- 118 Cropley, L., (2004): The Effect of Health Education Interventions on Child Malaria Treatment-Seeking Practices among Mothers in Rural Refugee Villages in Belize, Central America. Health Promotion International; Vol. 19, No. 4, pp 445-452. Doi:10.1093/heapro/dah406 Dako-Gyeke, P., Aikins, M., Aryeetey, R., Mccough, L and Adongo, P.B., (2013): The Influence of Socio-cultural Interpretations of Pregnancy threats on Health-seeking Behaviour among Pregnant women in urban Accra, Ghana. BMC Pregnancy and Childbirth 2013, 13:211 Danso-Appiah, A., Stolk W.A., Bosompem, K.M., Otchere, J., Looman, C.W.N., Habbena, D.F., and de Vlas, S.J., (2010): Health Seeking Behaviour and Utilization of Health Facilities for Schistosomiasis-Related Symptoms in Ghana. PLoS Negl Trop Dis 4(11): e867. doi:10.1371/journal.pntd.0000867 Davidson, S.K., & Schattner, P.L., (2003): Doctor’s health-seeking behaviour: a questionnaire survey. MJA: Vol 179: 302-305. 176 http://ugspace.ug.edu.gh/ Dominic, R.A., Shashidhara, Y.N., and Nayak, M.G., (2013): Health seeking behaviour of rural adults. Nitte University Journal of Health Science; Vol. 3; No. 3, pp. 77-82. El-Kak, F., Khawaja, M., Salem, M and Zurayk, H., (2009): Care-Seeking Behaviour of Women with Reproductive Health Problems from Low-income Areas of Beirut. International Journal of Gynaecology and Obstetrics; 104, 60-63. El Kahi, H.A., Rizk, G.Y.A., Hlais, S.A and Adib, S.M., (2012): Health-care-seeking Behaviour among University Students in Lebanon. Eastern Mediterranean Health Journal; Vol. 18, No. 6, pp 598-606. Fomundam, H., Tesfay, A., Maranga, A., Chanetsa, L., Muzoola, V and Oyaro, F., (2012): Identifying Treatment and Healthcare Seeking Behaviour as a Means of Early HIV/AIDS Intervention in Africa. World Journal of AIDS; 2, 165-173. Ghana Statistical Service (2012): 2010 Population and Housing Census: Summary Report of Final Results, Accra, Ghana. Ghosh, N., Chakrabarty, I., Chakraborty, M and Biswas, R., (2013): Factors affecting the healthcare-seeking behaviour of mothers regarding their children in a rural community of Darjeeling District, West Bengal. International Journal of Medicine and Public Health; Vol. 3, Issue 1, pp 12-16. Gotsadze, G., Bennette, S., Ranson, K and Gzirishvili, D., (2005): Health Care-Seeking Behaviour and Out-of-pocket Payments in Tbilisi, Georgia. Oxford University Press and the London School of Hygiene and Tropical Medicine. 177 http://ugspace.ug.edu.gh/ Grundy, J and Annear, P., (2010): Health-seeking behaviour studies: a literature review design and methods with a focus on Cambodia. Health Policy and Health Financing Knowledge Hub: The Nossal Institute for Global Health, University of Melbourne, Working Paper Series; No. 7 Haque, S.E., Rahman, M., Mostofa, G and Zohan, S., (2012): Reproductive Health Care Utilisation among Young Mothers in Bangladesh: Does Autonomy Matter? Women’s Health Issues; 22-2: e171-e180. Hausmann-Muela, S., Ribera, J.M and Nyamongo, I., (2003): Health-Seeking Behaviour and the Health System Response. Disease Control Priorities Project; Working Paper No. 14 Hjelm, K and Atwine, F., (2011): Health-care Seeking Behaviour among Persons with Diabetes in Uganda: An interview study. International Health and Human Rights: 11:11. Accessed from http://www.biomedcentral.com/1472-698X/11/11 Israel, G.D., (2013): Determining Sample Size. Programme Evaluation and Organisational Development (PEOD6). Institute of Food and Agricultural Sciences, University of Florida, USA Iyalomhe, G.B.S and Iyalomhe, S.I., (2012): Health-Seeking Behaviour of Rural Dwellers in Southern Nigeria: Implications for Healthcare Professionals. International Journal for Tropical Disease and Health; Vol. 2, No. 2, pp. 62-71 178 http://ugspace.ug.edu.gh/ Kakai, R., Menya, D., and Odero, W., (2009): Supporting formal education to improve quality of health care provided by mothers of children with malaria in rural western Kenya. J. Infect Dev Ctries; 3(7): 548-553 Khabiri, R., Pourreza, A., Foroushani, A.R., Sari, A.A., Arab, M and Kavosi, Z., (2011): Health Care-Seeking Behaviour in Tehran, Islamic Republic of Iran. World Applied Sciences Journal, 14(8): 1190-1197. Kothari, C.R., (2004): Research Methodology: Methods and Techniques. (Second Revised Edition); New Age International (P) Limited, Publishers. New Delhi, India. Kroeger, A., (1983): Anthropological and socio-medical healthcare research in developing countries. Soc. Sci. Med., 17, 147-161. Laar, A.S., Laar, A.K., and Dalinjong, P.A., (2013): Community Perception of malaria and its influence on health-seeking behaviour in rural Ghana: a descriptive study. Malaria World Journal, Vol. 4, No. 1, pp. 1-6 MacKian, S., (2003): A Review of Health Seeking Behaviour: Problems and Prospects. Health System Development Programme: University of Manchester, HSD/WP/ 05/03 MacKian, S., Bedri, N and Lovel, H., (2004): Up the garden path and over the edge: where might health-seeking behaviour take us? Health Policy and Planning; 19(3): 137- 146. Mashreky, S.R., Rahman, A., Chowdhury, S.M., Svanstrm, L., Shafinaz, S., Khan, T.F., and Rakman, F., (2010): Health seeking behaviour of burned children in 179 http://ugspace.ug.edu.gh/ Bangladesh is related to family socioeconomics. Injury, Int. J. Care Injured, 41, 528-532 Mazumda, P.G., Kanjilal, B., Barman, D and Mandal, A., (2009): Revisiting the Role of Geographical Accessibility in Women’s Access to Healthcare. ‘Future Health Systems: Innovations for Equity.’ India Series, Working Paper IV, India Mushtaq, M.U., Gull, S., Shad, M.A and Akram, J., (2011): Socio-demographic Correlates of the Health-Seeking Behaviours in two Districts of Pakistan’s Punjab Province. Journal of Pakistan Medical Association; Vol. 61, No. 12, pp 1205-1209. Nahar, Papreen (2010): Health seeking behaviour of childless women in Bangladesh: An ethnographic exploration for the special issue on: Loss in child bearing. Social Science and Medicine, 71, 1780-1787 National Rural Health Association, (2006): Racial and Ethnic Health Disparities. NRHA Issue Paper, USA Oluwatuyi, O., (2010): Health Seeking Behaviour among the Rural Dwellers in Ekiti State, Nigeria. African Research Review; Vol. 4(2): pp. 125-138 Owusu-Daaku, F.T.K., and Smith, F.J., (2005): Health-seeking behaviour: perspectives of Ghanaian women in London and Kumasi. International Journal of Pharmacy Practice; 13: pp. 71-76. DOI 10.1211/0022357055632 Prosser, T. (2007): Utilisation of health and medical services: Factors influencing health care seeking behaviour and unmet health needs in rural areas of Kenya. (PhD thesis 180 http://ugspace.ug.edu.gh/ Edith Cowan University, Australia): Accessed on April 21, 2014 from www.ro.ecu.edu.au/cgi/viewcontent.cgi1046&context=theses Rahman, S.A., Kielmann, T., McPake, B and Normand, C., (2012): Healthcare-Seeking Behaviour among the Tribal People of Bangladesh: Can the Current Health System Really Meet their Needs? Journal of Health, Population and Nutrition; 30(3): pp 353-365. Rai, R.K., Singh, P.K and Singh, L., (2012): Utilization of Maternal Health Care Services among Married Adolescent Women: Insights from the Nigerian Demographic and Health Survey, 2008. Women’s Health Issues; 22-4: e407-e414 Robyn, P.J., Fink, G., Sie, A., and Sauerborn, R., (2012): Health Insurance and Health- Seeking Behaviour: Evidence from a randomized community-based insurance roll- out in rural Burkina-Faso. Social Science and Medicine, 75, 595-603 Roy, L.C., Torrez, D and Dale, J.C., (2004): Ethnicity, Traditional Health Beliefs, and Health-Seeking Behaviour: Guardians’ Attitude Regarding their Children’s Medical Treatment. Journal of Paediatric Health Care; 18, 22-29. Russell, S., (2008): Demand-side Factors Affecting Health Seeking Behaviour in Ghana. Georgetown University Journal of Health Sciences, Vol. 5, No. 1 Saeed, I. I.B., Abdul-Aziz, A.R., & XicangZhaoa, (2013): Assessing the Influential Factors on the Use of Healthcare: Evidence from Ghana. International Journal of Business and Social Science, 4 (1), 12-20. 181 http://ugspace.ug.edu.gh/ Sandman, D., Simantov, E and An, C., (2000): Out of Touch: American Men and the Health Care System. Louis Harris and Associates Inc., The Commonwealth Fund. Shaikh, B.T., Haran, D., Hatcher, J and Azam, S.I., (2008): Studying Health-Seeking Behaviour: Collecting Reliable Data, Conducting Comprehensive Analysis. Journal of Biosocial Science; Vol. 40, Issue 01, Pp. 53-68. Shaikh, B.T., Haran, D., and Hatcher, J., (2008): Women’s Social Position and Health- Seeking Behaviour: Is the Health Care System Accessible and Responsive in Pakistan? Health Care for Women International; 29:8-9, 945-959 Shaikh, B.T., and Hatcher, J., (2005): Health seeking behaviour and health service utilization in Pakistan: challenging the policy makers. Journal of Public Health, Vol. 27, No.1, pp. 49-54 Siddiqui, M.S., Siddiqui, M.K., and Sohag, A.A., (2011): Health Seeking Behaviour of the People; Knowledge, Attitude and Practice (KAP) study of the people of urban slum areas of Karachi. Professional Med. Journal, 18(4): 626-631 Sreeramareddy, C.T., Shankar, R.P., Sreekumaran, B.V., Subba, S.H., Joshi, H.S and Ramachandran, U., (2006): Care Seeking Behaviour for Childhood Illness – A questionnaire survey in Western Nepal. International Health and Human Rights; 6:7. Accessed from: www.biomedcentral.com/1472-698X/6/7 Subba, N.R., (2004): Health Seeking Behaviour of Rajbanshi Community in Katahari and Baijanathpur of Morang District, Nepal. Journal of Nepal Health Research Council; Vol. 2, No. 2, pp 14-18. 182 http://ugspace.ug.edu.gh/ Sudha, G., Nirupa, C., Rajasakthivel, M., Sivasusbramanian, K., Sundaram, V., Bhatt, S., Subramanian, K., Thiruvalluvan, E., Mathew, R., Renu, G and Santha, T., (2003): Factors Influencing the Care-Seeking Behaviour of Chest Symptomatics: A community-based study involving rural and urban population in Tamil Nadu, South India. Tropical Medicine and International Health; Vol. 8, No. 4, pp 336-341. Sudharsanam, M.B and Rotti, O.S., (2007): Factors Determining Health Seeking Behaviour for Sick Children in a Fisherman Community in Pondicherry. Indian Journal of Community Medicine; 32:71-2. Taffa, N and Chepngeno, G., (2005): Determinants of Health Care Seeking for Childhood Illness in Nairobi Slums. Tropical Medicine and International Health; Vol. 10, No. 3, pp 240-245. Teerawichitchainan, B and Phillips, J.F., (2007): Differentials in Parental Health Seeking for Childhood Illness in Vietnam. Poverty, Gender, and Youth Working Paper No. 3; Population Council Inc., New York, USA Teye, K. J. (2012). Benefits, challenges and dynamism of positionalities associated with mixed methods research in developing countries, evidence from Ghana. Journal of mixed methods research: Sage, 3, 1-13. Tipping, G. & Segall, M. (1995): Health Care Seeking Behaviour in Developing Countries: An Annotated Bibliography and Literature Review. Development Bibliography 2: Institute of Development Studies, Sussex University. 183 http://ugspace.ug.edu.gh/ Uzochukwu, B.S.C., and Onwujekwe, O.E., (2004): Socio-economic differences and health seeking behaviour for the diagnosis and treatment of malaria: a case study of four local government areas operating the Bamako initiative programme in south-east Nigeria. International Journal for Equity in Health. 3:6. van der Hoeven, M., Kruger, A., and Greeff, M., (2012): Differences in health care seeking behaviour between rural and urban communities in South Africa. International Journal for Equity in Health, 11:31 Yanagisawa, S., Mey, V., and Wakai, S., (2004): Comparison of Health-seeking behaviour between poor and better-off people after health sector reform in Cambodia. Journal of the Royal Institute of Public Health, 118, 21-30 184 http://ugspace.ug.edu.gh/ APPENDIX A UNIVERSITY OF GHANA COLLEGE OF SOCIAL SCIENCES DEPARTMENT OF GEOGRAPHY AND RESOURCE DEVELOPMENT QUESTIONNAIRE FOR DATA COLLECTION This structured interview guide has been designed to help find out the factors influencing the health seeking behaviour of individuals in the Akuapem-North Municipality. The researcher would be very grateful if you could provide answers to these questions. All information given shall be used solely for academic purposes and will be treated confidentially. SECTION A: SOCIO-DEMOGRAPHIC DATA 1. Place of residence Code Answer Tick 1 Urban [ ] 2 Rural [ ] 2. Sex Code Answer Tick 1 Male [ ] 2 Female [ ] 3. Age ……………………………………………………………………………………. 185 http://ugspace.ug.edu.gh/ 4. Religion Code Responses Tick 1 None [ ] 2 Christian [ ] 3 Muslim [ ] 4 Traditional [ ] 5 Other: specify…………................. [ ] 5. How often do you attend worship services? Code Responses Tick 1 Very often [ ] 2 Often [ ] 3 Less often [ ] 4 Not at all [ ] 5 Other: specify………........................... [ ] 6. Do you agree that your religion influence your decision regarding treatment seeking? Code Responses Tick 1 Strongly agree [ ] 2 Agree [ ] 3 Disagree [ ] 4 Strongly disagree [ ] 186 http://ugspace.ug.edu.gh/ 7. Ethnicity Code Responses Tick 1 Akan [ ] 2 Guan [ ] 3 Ewe [ ] 4 Ga Adangme [ ] 5 Dagomba [ ] 6 Other: specify………........................... [ ] 8. Employment status Code Answer Tick 1 Employed [ ] 2 Unemployed [ ] 9. Occupation Code Responses Tick 1 Farmer [ ] 2 Trader [ ] 3 Artisan [ ] 4 Civil servant [ ] 5 Other: specify………………………………... [ ] 187 http://ugspace.ug.edu.gh/ 10. Educational level? Code Responses Tick 1 Never schooled [ ] 2 Primary/JHS [ ] 3 Secondary [ ] 4 Tertiary [ ] 5 Other: specify…………………………… [ ] 11. Years of schooling ……………………………………………………………………………………………. 12. Marital status Code Responses Tick 1 Single [ ] 2 Married [ ] 3 Divorced [ ] 4 Widowed [ ] 5 Cohabiting [ ] 6 Other: specify………………………… [ ] 188 http://ugspace.ug.edu.gh/ 13. Number of children including other dependents. Code Responses Tick 1 None [ ] 2 1 [ ] 3 2 [ ] 4 3 [ ] 5 4 and above [ ] 14. Monthly income …………………………………………………………………………………………….. 15. How satisfied are you with your monthly income? Code Description Tick 1 Very satisfied [ ] 2 Satisfied [ ] 3 Somehow satisfied [ ] 4 Dissatisfied [ ] 5 Very dissatisfied [ ] 16. How much do you spend a day including your dependents? …………………………………………………………………………………………. 189 http://ugspace.ug.edu.gh/ 17. Are you the head of a household? Code Responses Tick 1 Yes [ ] 2 No [ ] 3 Shared [ ] 18. Are you enrolled on the National Health Insurance Scheme? Code Answer Tick 1 Yes [ ] 2 No [ ] 19. If yes, what is the state of your NHIS status? Code Insurance status Tick 1 Active [ ] 2 Expired [ ] 20. How long have you been a registered member of the NHIS? Code Years with NHIS Tick 1 Less than 1 year [ ] 2 1-3 years [ ] 3 4-7 years [ ] 4 8 years and above [ ] 190 http://ugspace.ug.edu.gh/ 21. If no, why have you not registered yet? Code Responses Tick 1 Unable to afford premium payment [ ] 2 NHIS does not cover all my health needs [ ] 3 Limited quality of care at public health facilities [ ] 4 Do not fall sick very often [ ] 5 Do not get time to go to NHIS office [ ] 6 Other: specify………………………………… [ ] 22. What is the main source of drinking water for you? Code Source of water Tick 1 Covered source {Well/Piped/Roof catchment} [ ] 2 Uncovered source {River/Spring/Stream/Pond} [ ] 23. What kind of toilet do you have and use? Code Amenity Tick 1 Own flush toilet [ ] 2 Shared flush toilet [ ] 3 Traditional pit toilet [ ] 4 Ventilated improved pit [ ] 5 Latrine [ ] 6 No facility/bush/field [ ] 7 Other: specify………………………...... [ ] 191 http://ugspace.ug.edu.gh/ 24. Please describe the main material used for your home. Code Home Material Used Tick 1 Roof 1. Corrugated iron [ ] 2. Thatch [ ] 2 Walls 1. Wood [ ] 2. Mud [ ] 3. Corrugated iron [ ] 4. Concrete/stone/bricks [ ] 3 Floor 1. Mud [ ] 2. Concrete/stone/bricks [ ] 25. Where you live, do you Code Housing Tick 1 Own your house [ ] 2 Rent your house [ ] 3 Don’t own house, but don’t pay rent [ ] 4 Other: specify…………………………………….. [ ] SECTION B: HEALTH PROVIDER CHOICES AND HEALTH SEEKING BEHAVIOUR 26. Which health service providers are readily available in your community? 192 http://ugspace.ug.edu.gh/ ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 27. Which of these health service providers is the closest to your house? ………………………………………………………………………………………….. 28. Are these available health service providers where you mostly seek treatment? Code Answer Tick 1 Yes [ ] 2 No [ ] 29. Have you been ill in the past twelve (12) months? Code Answer Tick 1 Yes [ ] 2 No [ ] 30. Did you know what was making you ill? (Please tick major illness only). Code Major illness Tick 1 Don’t know [ ] 2 Malaria/Fever [ ] 3 Sexually Transmitted Infections (STIs) [ ] 4 Stomach/abdominal pains [ ] 5 General body pains/Head-ache [ ] 6 Spiritual [ ] 193 http://ugspace.ug.edu.gh/ 7 Other: specify……………..................................... [ ] 31. How did you know what was making you ill? Code Source of treatment Tick 1 Self-diagnosis [ ] 2 Friend/Neighbour/Relative [ ] 3 Pharmacist [ ] 4 Person in drug/chemical shop [ ] 5 Hospital/Clinic/Medical personnel [ ] 6 Traditional healer [ ] 7 Other: specify…………………………………… [ ] 32. Severity of the illness? Code Severity of illness Tick 1 Not severe [ ] 2 Less severe [ ] 3 Severe [ ] 4 More severe [ ] 33. Did you do something to help you feel better? Code Answer Tick 1 Yes [ ] 2 No [ ] 194 http://ugspace.ug.edu.gh/ 34. What major action do you take, in an attempt to seek health, when you first perceived symptoms of illness? Code First action taken when ill Tick 1 Apply a home remedy [ ] 2 Buy medicine from drug store [ ] 3 Go to a traditional healer for cure [ ] 4 Seek treatment from a medical facility [ ] 6 Allow nature to take its course [ ] 7 Consult a pharmacist for medicine [ ] 8 Other: specify………………………………......... [ ] 35. What is your major reason for taking this action? ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 36. Do you always get cured when this action is taken? Code Answer Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 195 http://ugspace.ug.edu.gh/ 37. If no/sometimes; what again do you do with the idea of seeking health? Code Second action taken when ill Tick 1 Apply a home remedy [ ] 2 Buy medicine from drug store [ ] 3 Go to a traditional healer for cure [ ] 4 Seek treatment from a medical facility [ ] 6 Allow nature to take its course [ ] 7 Consult a pharmacist for medicine [ ] 8 Other: specify………………………………........................................... [ ] 38. Do you at this time get cured? Code Answer Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 39. What is your main reason for taken this action? ……………………………………………………………………………………………… ……………………………………………………………………………………………… 40. At what time of your illness do you mostly go for treatment from your health providers? Code Time of illness Tick 1 At first day of illness [ ] 2 At second day of illness [ ] 196 http://ugspace.ug.edu.gh/ 3 At third day of illness [ ] 4 At fourth day of illness [ ] 5 Other: specify………………………… [ ] 41. What is your main reason for seeking treatment at that time? ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… SECTION C: UTILISATION OF HEALTH SERVICES 42. Where do you usually seek health/treatment when ill? Code Health Provider Tick 1 Hospital [ ] 2 Clinic [ ] 3 Health centre/CHPS compound [ ] 4 Traditional healer [ ] 5 Drug/chemical shop [ ] 6 Prayer camp [ ] 7 Other: specify………………………………... [ ] 43. What is your main reason for seeking treatment from this provider? 197 http://ugspace.ug.edu.gh/ ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 44. If you have a choice, where would you seek treatment? Code Most preferred health provider Tick 1 Hospital [ ] 2 Clinic [ ] 3 Health centre/CHPS compound [ ] 4 Traditional healer [ ] 5 Drug/chemical shop [ ] 6 Prayer camp [ ] 7 Drug peddler [ ] 8 Other: specify…………………………………….... [ ] 45. Why is this health provider your most preferred in treatment seeking? ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 46. Is your most preferred health provider where you seek treatment? Code Answer Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 198 http://ugspace.ug.edu.gh/ 47. If no/sometimes; what prevents you from seeking treatment there? ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 48. How far, in kilometres, is the nearest health provider in the community from your house? Code Distance Tick 1 Less than 1km [ ] 2 Between 1km to 4km [ ] 3 Between 5km to 9km [ ] 4 More than 10km [ ] 5 Other: specify………………………… [ ] 49. Is that the health service provider you normally seek treatment at when ill? Code Answer Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 50. If no/sometimes; where else, apart from this health provider, do you go for treatment? ……………………………………………………………………………………………… ……………………………………………………………………………………………… 51. What is your major reason for seeking care from this health provider? 199 http://ugspace.ug.edu.gh/ ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 52. Which is the way you would normally get to the health service provider you mostly visit? Code Mode of transportation Tick 1 Foot [ ] 2 Bicycle [ ] 3 Motorbike [ ] 4 Vehicle [ ] 5 Other: specify…………………….. [ ] 53. How much do you pay as transportation to this health provider you mostly visit? ……………………………………………………………………………………………… 54. Who pays for your transportation to the health provider you mostly visit? Code Payment Tick 1 Self [ ] 2 Spouse [ ] 3 Parent [ ] 4 Other relative [ ] 5 Other: specify…………………….. [ ] 55. Using the way you would normally, how long does it take to get to this health provider? ................................................................................................................................................ 200 http://ugspace.ug.edu.gh/ 56. How long, on the average, does it take to receive treatment at the health provider you mostly visit? ……………………………………………………………………………………………… 57. Are you always cured of your illness anytime you seek care from these health providers? Code Answer Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 58. Are you always treated well by personnel of health providers you mostly visit for care? Code Well treated Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 4 Don’t know [ ] 59. How satisfied are you with the attitude of personnel at health providers you mostly visit? Code Personnel attitude Tick 1 Very satisfied [ ] 2 Satisfied [ ] 3 Dissatisfied [ ] 4 Very dissatisfied [ ] 5 Cannot tell [ ] 201 http://ugspace.ug.edu.gh/ 60. How satisfied are you with the services provided by the health providers you mostly visit? Code Satisfaction Tick 1 Very satisfied [ ] 2 Satisfied [ ] 3 Dissatisfied [ ] 4 Very dissatisfied [ ] 61. How would you rate the overall quality of service delivery by the health providers you mostly seek treatment at? Code Overall quality Tick 1 Excellent [ ] 2 Very good [ ] 3 Good [ ] 4 Fair [ ] 5 Poor [ ] 6 Very poor [ ] 62. Do you recommend the health providers you visit to other people in the community? Code Recommendation Tick 1 Yes [ ] 2 No [ ] 3 Sometimes [ ] 202 http://ugspace.ug.edu.gh/ SECTION D: BARRIERS TO AND PAYMENT FOR HEALTH SERVICE USE 63. Have you faced any difficulty in getting to the health provider you mostly seek care at? Code Answer Tick 1 Yes [ ] 2 No [ ] 64. If yes, what difficulty was it? ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………………………………… 65. Have you faced any difficulty in using the health service provider you mostly visit for treatment? Code Answer Tick 1 Yes [ ] 2 No [ ] 66. If yes, what difficulty was it? ……………………………………………………………………………………………… ……………………………………………………………………………………………… 67. How much, on the average, did you spend the last time you visited your usual health service provider for treatment? ……………………………………………………………………………………………… 203 http://ugspace.ug.edu.gh/ 68. How do you pay for your health service use? ……………………………………………………………………………………………… 69. Who pays for your health service use? Code Payment Tick 1 Self 2 NHIS [ ] 3 Spouse [ ] 4 Parent [ ] 5 Relative [ ] 6 Other: specify…………………….. [ ] 204 http://ugspace.ug.edu.gh/ APPENDIX B UNIVERSITY OF GHANA COLLEGE OF SOCIAL SCIENCES DEPARTMENT OF GEOGRAPHY AND RESOURCE DEVELOPMENT INTERVIEW GUIDE FOR HEADS OF HEALTH SERVICE PROVIDERS This interview guide has been designed for a study on the health seeking behaviour of people in the Akuapem-North Municipality. The researcher would be very grateful if you could provide answers to these questions. All information given shall be used solely for academic purposes and will be treated confidentially. BACKGROUND INFORMATION 1. Place of residence…………………………………………….... 2. Sex……………………………………………………………... 3. Age (optional)………………………………………………….. 4. Qualification…………………………………………………… 5. Marital status…………………………………………………… 6. Number of children…………………………………………….. HEALTH SEEKING BEHAVIOUR AND HEALTH SERVICE USE 1. Which illness symptoms do patients report the most at your facility?  Which category of people frequently report of these illness symptoms?  How do you regard the severity of their illness?  At what stage of their illness symptoms is treatment sought at your facility? 2. How do you regard the people that come to your facility for treatment?  Which category of people mainly visits your facility for treatment? 205 http://ugspace.ug.edu.gh/ 3. How does your presence in the community and proximity to peoples’ house influence their decision to seek treatment at your facility whenever they perceived symptoms of illness?  When do you start work in the day and when do you close?  Are your services limited to this community alone?  Do you have patients visiting this facility from other towns/villages? 4. How do your operations influence the peoples’ decision to seek health at your facility?  Do your patients always get the needed medicines?  How do you relate with your patients?  Do your patients mainly complain about your services?  How would you describe the efficiency of your staff? 5. What do you think are some of the barriers to the patronage of your services?  Who do you think bears the impact of these barriers?  How do these barriers affect the operations of your facility?  How can these barriers be revamped?  Do you face challenges in providing health services in this community?  What are some of them and by what means are you coping with these? 6. What is the main mode of health service use payment at your facility?  Is there any other way patients pay for their health service use? 206 http://ugspace.ug.edu.gh/ APPENDIX C UNIVERSITY OF GHANA COLLEGE OF SOCIAL SCIENCES DEPARTMENT OF GEOGRAPHY AND RESOURCE DEVELOPMENT INDIVIDUAL INTERVIEW GUIDE This interview guide has been designed for a study on the health seeking behaviour of people in the Akuapem-North Municipality. The researcher would be very grateful if you could provide answers to these questions. All information given shall be used solely for academic purposes and will be treated confidentially. BACKGROUND INFORMATION 1. Place of residence……………………………………………............................................... 2. Sex………………………………………………………………………………………….. 3. Age…………………………………………………………………………………………. 4. Educational level…………………………………………………………………………… 5. Years of schooling………………………………………………………………………….. 6. Employment status…………………………………………………………………………. 7. Occupation…………………………………………………………………………………. 8. Average monthly income…………………………………………………………………... 9. Marital status……………………………………………………………………………….. 10. Number of children including other dependents……………………………………………. 11. NHIS member……………………………………………………………………………… 12. How long have you been with the scheme………………………………………………….. 207 http://ugspace.ug.edu.gh/ HEALTH SEEKING BEHAVIOUR AND HEALTH SERVICE USE 1. What are some of the commonest illness symptoms that you mostly perceive?  How frequent have you perceived this illness symptom?  What do you think is the major cause of this perceived illness symptoms? 2. How do you deal with the illness symptoms whenever they are perceived?  What is your reason for taken such a decision? 3. Do you have preferences in terms of treatment seeking from health service providers?  Why do you prefer such health service providers to others? 4. What are some of the factors that influence your choice and use of health services at a particular health service provider?  How do these factors influence your choice and use of health services?  What do you wish could be done to solve such problems? 5. What are some of the difficulties you encounter in your health service use at a particular health service provider?  How do these hinder your health service use at a particular health service provider?  How can these difficulties be overcome? 6. How much money did you pay the last time you visited a health facility for treatment?  What can you say about it; is it expensive or not? 7. Which is the main means by which you finance your health service use?  Was it difficult financing your health service use through this means?  Is there any other means by which you pay for health service us?  If you would, how would finance your health service use? 208