SCHOOL OF PUBLIC HEALTH COLLEGE OF HEALTH SCIENCES UNIVERSITY OF GHANA RESILIENCE AND SOCIAL SUPPORT AS INFLUENCING FACTORS ON THE RELATIONSHIP BETWEEN PSYCHOLOGICAL DISTRESS AND QUALITY OF LIFE IN PERSONS LIVING WITH CANCER. A CROSS-SECTIONAL STUDY AT THE KORLE-BU TEACHING HOSPITAL BY SWITHIN MUSTAPHA SWARAY (10933674) THIS DISSERTATION IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON, IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF MASTER OF PUBLIC HEALTH DEGREE DECEMBER, 2022 University of Ghana http://ugspace.ug.edu.gh ii DECLARATION I, Swithin Mustapha Swaray, declare that this dissertation is my original work and has not been taken from the contents of others save to the extent that such work has been cited and acknowledged within the text of my own elaboration and development. This research was under the guidance of Dr. Benedict Weobong, Department of Social and Behaviour Change, School of Public Health, University of Ghana. The matter embodied herein has not been submitted in part or whole to any other university for the award of any other academic degree. All materials cited have been acknowledged by means of a complete reference. Name of Student: SWITHIN MUSTAPHA SWARAY Signature: Date: 19/10/2023 Name of supervisor: DR. BENEDICT WEOBONG Signature: Date: 21/10/2023 University of Ghana http://ugspace.ug.edu.gh iii DEDICATION In loving memory of my dear father Dr. Swithin Mustapha Swaray Snr, Mr. Alexis Wulluh Nakaar, Angelina Nyamekye, Raymonda Okeke-Macauley, Prof. Frederick Rodrigues, Halima Amoa, Nutifafa Kanawatey and all the beautiful minds and loving hearts who were swept away by the storm of cancer. Your footprints remain in our hearts. University of Ghana http://ugspace.ug.edu.gh iv ACKNOWLEDGEMENT To Prof. Alfred Edwin Yawson, thanks for opening my eyes to new stages of opportunity and strength. I will forever be grateful for your guidance and kindness. To Dr. Benedict Weobong, thanks for your encouragement and guidance. You always seek to bring out the best in your students. You are everything one could look for in a good supervisor. To my buddies John Tetteh, Ewurabena Oduma Duker, Alex Ackon, Isabella Asamoah, Prosper Kaba, and Adwoa Kumiwa A Afrane, when we have friends to support and wish us well, we will most certainly excel. Thanks for making this journey worth the while. To you Mr. John Tetteh, I could not have completed this in good time without your immense support. You are simply phenomenal, the one person who gives meaning to data, that I could run to no matter the hour. I owe you a debt of gratitude. To my SPH SOBS class of 2021/2022, you were simply incredible. Gloria Fiifi Yankson, you became my best buddy. To my mothers, Felicia Boye-Doe and Prof. Onike Rodrigues, your support has simply been amazing. To my loving wife, Ama Dede, my awesome boys, Malachi, Micaiah, and Zion, and my supportive sister Madia, you are the wind in my sails. It was never going to be easy being away from family for the whole run. Thanks for your love, understanding, and support. You remain my world. University of Ghana http://ugspace.ug.edu.gh v ABSTRACT Background: The psychological well-being of persons living with cancer (PLWC) could be enhanced if factors influencing the effect of psychological distress (distress) on quality of life (QoL) are well understood. Distress in PLWC is reshaped over the course of the disease, prompting the need for awareness and constant monitoring by health professionals and caregivers to identify and understand factors that mitigate the long-term implications for QoL. The influencing roles of resilience and social support on the distress-QoL relationship amongst PLWC in Ghana have not been widely explored. This study was conducted to assess the mediating role of resilience and social support in the relationship between distress and QoL. Methods: This research employed a facility-based cross-sectional analytical design. The exposure and primary outcomes were distress and QoL respectively. Resilience and social support were the mediators. A causal path analysis was adopted to assess the mediation effect of resilience and social support on the relationship between distress and QoL. All estimations were done using the 95% confidence interval. Results: The prevalence of high distress amongst respondents was 45.4% (95%CI=40.95- 49.96) with overall good QoL being 52.7% (95%CI=48.12-57.16). There was no indirect effect of resilience and social support on the relationship between distress and QoL [(β(95%CI)p- value = -0.0001(-0.002 to 0.002)0.877 and 0.004 (-0.01 to 0.02)0.630 respectively]. Generally, amongst persons with high distress, the adjusted odds of good QoL was 0.18 (95%CI=0.11- 0.30) statistically significant. However, it significantly increased to 0.28(95%CI=0.12-0.64) when resilience was high. Conclusion: Approximately 2 out of 5 respondents experienced high distress and a little more than half experienced good QoL. There was insufficient evidence to conclude that resilience and social support mediate the relationship between distress and QoL. However, resilience was found to be an effect modifier in the relationship. This study provides evidence supporting the University of Ghana http://ugspace.ug.edu.gh vi need for a more holistic approach to cancer care and management, focusing on facilitating positive traits like resilience and improving social support. University of Ghana http://ugspace.ug.edu.gh vii TABLE OF CONTENTS DECLARATION..................................................................................................................... ii DEDICATION........................................................................................................................ iii ACKNOWLEDGEMENT ...................................................................................................... iv ABSTRACT .............................................................................................................................. v TABLE OF CONTENTS ..................................................................................................... vii LIST OF ABBREVIATIONS ................................................................................................. x LIST OF FIGURES ................................................................................................................ xi LIST OF TABLES ................................................................................................................ xii CHAPTER ONE ...................................................................................................................... 1 1.0 INTRODUCTION.......................................................................................................... 1 1.1 Background of the study ........................................................................................... 1 1.1.1 Cancer and Psychological Distress ........................................................................... 2 1.1.2 Cancer and Quality of Life ....................................................................................... 3 1.1.3 Cancer and Resilience ............................................................................................... 4 1.1.4 Cancer and Social Support ........................................................................................ 4 1.2 Problem Statement ....................................................................................................... 5 1.3 Research Questions ...................................................................................................... 7 1.4 Aims and Objectives .................................................................................................... 7 1.5 Justification .................................................................................................................. 8 1.6 Relevance of the study ................................................................................................. 8 1.7 Hypothesis.................................................................................................................... 9 1.8 Theoretical and Conceptual Framework ...................................................................... 9 CHAPTER TWO ................................................................................................................... 14 2.0 LITERATURE REVIEW ........................................................................................... 14 2.1 Introduction ................................................................................................................ 14 2.2 Prevalence of Psychological distress in persons living with cancer .......................... 15 University of Ghana http://ugspace.ug.edu.gh viii 2.2.1 Factors associated with psychological distress in persons living with cancer ........ 16 2.3 Level of Quality of life in persons living with cancer ............................................... 17 2.3.1 Factors associated with quality of life in persons living with cancer ..................... 18 2.4 Mechanisms of Resilience and Social Support in Persons living with cancer .......... 19 2.5 Knowledge Gap ......................................................................................................... 20 CHAPTER THREE ............................................................................................................... 22 3.0 METHODS ................................................................................................................... 22 3.1 Introduction ................................................................................................................ 22 3.2 Description of site ...................................................................................................... 22 3.3 Research Approach .................................................................................................... 22 3.4 Research design ......................................................................................................... 23 3.5 Study population ........................................................................................................ 23 3.6 Inclusion and exclusion criteria ................................................................................. 23 3.7 Sample size ................................................................................................................ 23 3.8 Sampling method and data collection procedure ....................................................... 24 3.9 Questionnaire piloting/pretesting ............................................................................... 25 3.10 Measurement Scales................................................................................................. 25 3.11 Data analysis ............................................................................................................ 30 3.12 Ethical considerations .............................................................................................. 31 CHAPTER FOUR .................................................................................................................. 32 4.0 RESULTS ................................................................................................................. 32 4.1 Introduction ................................................................................................................ 32 4.2 Sociodemographic and cancer related characteristics of study participants .............. 32 4.3 Prevalence of distress and quality of life among persons living with cancer ............ 35 4.4 Factors associated with high psychological distress and good quality of life amongst study participants. ............................................................................................................ 39 4.5 Indirect effect of resilience and social support on the relationship between distress and quality of life amongst persons living with cancer. .................................................. 43 University of Ghana http://ugspace.ug.edu.gh ix 4.6 How resilience and social support modify the effect of distress on quality of life among persons living with cancer.................................................................................... 44 CHAPTER FIVE ................................................................................................................... 46 5.0 DISCUSSION ............................................................................................................... 46 5.1 Introduction ................................................................................................................ 46 5.2 Prevalence of high distress and associated factors in persons living with cancer ..... 46 5.3 Prevalence of good QoL and associated factors amongst persons living with cancer .......................................................................................................................................... 51 5.4 Mediation effect of resilience and social support on the relationship between psychological distress and quality of life in persons living with cancer.......................... 52 5.5 Effect modification of resilience and social support on the relationship between psychological distress and QoL ....................................................................................... 53 5.6 Study implications for clinical and public health practice ......................................... 54 5.7 Strength and limitations ............................................................................................. 55 CHAPTER SIX ...................................................................................................................... 56 6.0 CONCLUSION AND RECOMMENDATION ......................................................... 56 6.1 Introduction ................................................................................................................ 56 6.2 Conclusion ................................................................................................................. 56 6.3 Recommendation ....................................................................................................... 57 REFERENCE ......................................................................................................................... 60 APPENDICES ........................................................................................................................ 85 Appendix 1: Level of social support and resilience amongst study respondents ................ 85 Appendix 2: Informed Consent Form ................................................................................. 86 Appendix 3: Voluntary Agreement Form ............................................................................ 88 Appendix 4: Questionnaire .................................................................................................. 89 Appendix 5: Ethical Consideration ...................................................................................... 94 University of Ghana http://ugspace.ug.edu.gh x LIST OF ABBREVIATIONS Abbreviation Meaning ACS American Cancer Society ANOVA Analysis of Variance APS American Psychological Society CD-RISC Conor-Davidson Brief Resilience Scale CVDs Cardiovascular diseases GLOBOCAN Global Cancer Observatory/Data GSEM Generalized Structural Equation Modelling HBV Hepatitis B virus HCC Hepatocellular carcinoma HPV Huma papilloma virus HRQoL Health related quality of life IARC International Agency for Research on Cancer KBTH Korle-Bu Teaching Hospital MSPSS Multidimensional Scale of Perceived Social Support NCDs Non-communicable diseases PD Psychological distress PDI Psychological Distress Inventory PDI-R Psychological Distress Inventory - Revised PLWC Persons living with cancer PsyCap Psychological Capital QoL Quality of Life SF-8 Short Form-8 SSA Sub-Saharan Africa WHO World health organization University of Ghana http://ugspace.ug.edu.gh xi LIST OF FIGURES Figure 1.1: Conceptual framework defining the causal pathway of distress and QoL among PLWC ...................................................................................................................................... 13 University of Ghana http://ugspace.ug.edu.gh xii LIST OF TABLES Table 3 1: Study variable and definitions ................................................................................ 28 Table 4.1: Sociodemographic and cancer related characteristics of study participants………33 Table 4.2: Level of psychological distress and quality of life amongst study respondents ..... 35 Table 4.3: Prevalence of high psychological distress and good quality of life by sociodemographic characteristics of participants .................................................................... 36 Table 4.4: Prevalence of high psychological distress and good quality of life by cancer related characteristics of participants ................................................................................................... 38 Table 4.5: Crude and adjusted estimates showing sociodemographic characteristics associated with high psychological distress and good quality of life amongst study participants ............ 40 Table 4.6: Crude and adjusted estimates showing cancer related characteristics associated with high psychological distress and good quality of life amongst study participants .................... 42 Table 4.7: Pairwise correlations showing the relationship between psychological distress, quality of life, resilience, and social support amongst study participants ............................... 44 Table 4.8: Direct and Indirect effect of resilience and social support on quality of life ......... 44 Table 4.9: Adjusted logistic regression showing association between psychological distress and quality of life modified by level of resilience .......................................................................... 45 University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE 1.0 INTRODUCTION 1.1 Background of the study Globally, after cardiovascular diseases (CVDs), cancer is the second principal cause of premature deaths and significant disability-adjusted life years (DAILYs) (Roberts et al., 2022; WHO, 2022). In 2020, using data from the International Agency for Research on Cancer (IARC), GLOBOCAN projected 19.3 million new cases of cancer were reported globally, with resulting mortality of 9.9 million. In that same period, the African region alone accounted for over 1 million new cases, with a resultant 711,429 deaths (Roberts et al., 2022). This alarming trend presents the public health community with a huge crisis, especially in Western Sub- Saharan Africa where most people with cancer report for treatment in the advance stages of the disease and epidemiologic data points to a steady rise in mortalities over the next ten years (Abda et al., 2017; Kingham et al., 2013; Twahir et al., 2021). Although breast, cervical, and prostate cancers account for most malignancies in the region, the cancer profile in Sub-Saharan Africa is highly varied and public health data suggest deaths due to breast and cervical cancers are amongst the highest worldwide (ACS, 2022; Bahnassy et al., 2020; Roberts et al., 2022). The IARC stipulates the number of new cases in Ghana approximates 24, 009, with breast cancer (18.7%), hepatocellular carcinoma (14.4%), cervix uteri (11.6%), prostate cancers (8.9%) and non-Hodgkin lymphoma (5.0%) topping the chart for reported cancers, cancer mortality and prevalence for all ages (IARC, 2021). Poor diagnosis and limited access to appropriate management are important variables that contribute to poor prognosis. This is affirmed by the low availability of pathology services in the region, with just about 26% present in low-income countries in 2017 (Ayandipo et al., 2020). The rising expense of cancer treatment is a public health concern, particularly in developing nations (Colonio et al., 2021; Twahir et al., 2021). The impact of the disease can be so overwhelming that the World Health University of Ghana http://ugspace.ug.edu.gh 2 Organization’s (WHO) global health observation reports an alarming 703,000 persons living with cancer commit suicide yearly with many others having suicidal ideations (WHO, 2022). Cancer has been associated to mental disorders with systematic reviews linking specific cancers to distress and post-traumatic stress disorders (Arnaboldi et al., 2017; Caruso et al., 2017; Watts et al., 2014). Data on cancer in Ghana may not be indicative of the actual burden as they were sourced from a single cancer registry. Despite this likely under reporting, figures are in line with the global picture (Mensah & Mensah, 2020; Roberts et al., 2022). The rise is strongly linked to the growing behavioural risk factors observed in Ghanaian society today. These include overweight and obesity, physical inactivity, poor consumption of fruits and vegetables, excessive consumption of alcoholic drinks (Juma et al., 2020; Mensah & Mensah, 2020) as well as indiscriminate sex (Binka et al., 2017). The epidemiology of cancer in sub-Saharan Africa has also been linked to urbanization and the development and exposure to more hazardous waste as a result of inadequate disposal systems (Bickler et al., 2018; Fasinu & Orisakwe, 2013). 1.1.1 Cancer and Psychological Distress Persons living with cancer (PLWC) are often overwhelmed due to the psychological impact of the diagnosis, the high symptom burden associated with the disease and the related side effect of treatment (Grassi et al., 2017). These symptoms usually occur simultaneously, negatively affecting functional status and making PLWC susceptible to psychological distress (distress) (Kroenke et al., 2013; Matzka et al., 2016). These enfeebling symptoms aside, PLWC battle mental health challenges on a regular basis (Lim et al., 2014; Matzka et al., 2016). The taxing, unpleasant, and lengthy treatment can cause a loss of autonomy, physical fragility, and death anxiety leading to significant cancer-related discomfort with clinically relevant symptoms of various mental illnesses (Graf & Stengel, 2021). University of Ghana http://ugspace.ug.edu.gh 3 The financial toxicity caused by cancer is strongly linked to distress as well, with implications for families of patients and their wellbeing (Chabowski et al., 2018; Yu et al., 2022). In some instances, cancer diagnosis has led to loss of employment (Munir et al., 2009; J.-H. Park et al., 2009; Paul et al., 2016; Taskila & Lindbohm, 2007), a predictor of psychiatric comorbidity that correlates with at least threefold increased risk of mental disorder (Nakash et al., 2014; Roick et al., 2022). Distress indeed is pervasive in PLWC and occurs at varying degree throughout the course of the illness highlighting the magnitude of this public health crisis (Islam, 2019; Lavelle et al., 2017; Matzka et al., 2016; Pásztor et al., 2019; Zamanian et al., 2018). 1.1.2 Cancer and Quality of Life Quality of life (QoL) has been reported as one of the most critical variables related to survival in PLWC and in recent years, has become a key indicator of health service performance as it is influenced by the state of affairs of the healthcare system (Sheikhalipour et al., 2019; Stefani et al., 2016). This construct is defined by the World Health Organization (WHO) as “an individual's perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns” (Megari, 2013). In their integrative review of literature amongst a population of African female cancer survivals, Muliira et al., (2017) highlighted psychosocial, socio-demographic, illness-related, treatment related and cultural related factors influencing QoL. These point to the construct as a multi-dimensional concept reflecting the subjective nature of an evaluation which is also rooted in sociocultural as well as environmental context. With regards to cancer, illness related factors responsible for the limitations and strain on QoL have been found across types of cancer and at all phases of the medical process (Ruiz-Rodríguez et al., 2022). The distress associated with receiving a cancer diagnosis, the side effect of treatment and the high symptom burden are strongly linked to poor QoL (Liao et al., 2008; Peters et al., 2020). Such burden inundate persons with fear, shattering their dreams and confining them to a state of depressive University of Ghana http://ugspace.ug.edu.gh 4 rumination negatively affecting functionality, cognition and in effect, QoL (Hajigholami et al., 2021). 1.1.3 Cancer and Resilience The American Psychological Association defines resilience as “the ability to adapt successfully in the face of adversity, trauma, tragedy, danger, or severe causes of stress, such as family and relationship issues, serious health issues, or employment and financial difficulties” (APS, 2020). A growing body of research has firmly linked this dynamic process of adapting to adverse life changing circumstances to lower distress, better cancer adjustment, improved QoL, and better mental health and treatment results in PLWC (Duan-Porter et al., 2016; Martin et al., 2021; Matzka et al., 2016; Ye et al., 2017). In the context of cancer, resilience is a term used to describe a person's personal qualities that are thought to be adjustable and facilitate successful adaption to cancer. It involves, among other things, a sense of coherence, optimism, self-efficacy, self-esteem, cognitive flexibility, coping, social support, as well as spirituality (Helmreich et al., 2017). This positive psychological state of an individual in the face of adversity is influenced by social context and environmental factors interacting with several personality traits (Celik et al., 2021; Matzka et al., 2016) and confirmed to be a protective factor for good QoL emphasizing its usefulness in psycho-oncology (Seiler & Jenewein, 2019). 1.1.4 Cancer and Social Support Social support is crucial for the psychological adjustment of PLWC as it has been found to provide a safety net, significantly influencing prognosis and improving QoL (Alizadeh et al., 2018; Quiroga et al., 2018). The presence of this support has implications that transcends treatment outcomes; promoting physical and mental well-being, encouraging healthy living, and aiding in the management of negative thoughts, depressed moods, and hazardous behaviours (Ekem-Ferguson et al., 2020). It has ramifications for financial worries, emotional University of Ghana http://ugspace.ug.edu.gh 5 burden as well as assistance with daily routine (Adam & Koranteng, 2020). Researchers (Breuer et al., 2017; Ko et al., 2013; Melguizo-Garín et al., 2019; Morelli et al., 2015; Pejner et al., 2012) categorised social support into emotional (important during times of stress or feeling lonely), instrumental (important for addressing immediate needs) and informative (important when making major decisions). They evinced that due to its chronic debilitating nature, PLWC and their families may be exposed to a wide range of stressful events over time, presenting occasions where patients' social support requirements have not been met because their support system is inadequately prepared. Aside the type of support, the source and quality of support are key (Yıldırım & Tanrıverdi, 2021). Patients value support from partners, family, and friends. They consider this source of support essential to cope with the diagnosis and treatment (Pfaendler et al., 2015). Sometimes support source may include health care workers who constantly interreact with patients throughout their treatment and management. For the PLWC, social support can be therapeutic not just in the short term, but through the uncertain anxious journey. In summary, adjustment to living with cancer is a tumultuous everyday process with different challenges at each stage of the illness. As indicated by Agostinelli et al., (2022) and Muzzatti et al., (2020) distress and QoL may be reshaped over the course of the disease prompting the need for awareness and constant monitoring by health professionals and care givers. To reduce the impact of the distress-QoL relation on PLWC and promote mental health and wellbeing in the face of serious illness, interventions must be informed by the understanding of factors potentially influencing the causal pathway of this association. 1.2 Problem Statement Compared to the general population, PLWC have worse QoL (Alam et al., 2020; Quinten et al., 2015) as they are highly susceptible to distress (Momenimovahed et al., 2021). Such vulnerability has been strongly linked to poor QoL (Bitew et al., 2021; Kugbey et al., 2019). University of Ghana http://ugspace.ug.edu.gh 6 Ideally, clinical practice should screen for distress in PLWC however this does not appear to be a routine practice in many parts of Sub-Saharan Africa (SSA) (Kagee, 2022). With the alarming rise in incidence observed in Africa, SSA in particular, trends could soon mirror that of developed nations where cancer is fast becoming the leading cause of mortality, responsible for twice as many deaths as cardiovascular diseases (Mahase, 2019). In Ghana, indications are that cancer incidence is on the rise as behavioural risk factors become pronounced (Juma et al., 2020; Kagee, 2022; Mensah & Mensah, 2020). Ghana has seen a rise in cancer cases from 61,388 in 2010, to 95,000 in 2019 (Sasu, 2022). This implies greater burden of chronic non- communicable diseases (NCDs) related mental health challenges for Ghana as PLWC battle mental health issues affecting their functionality, wellbeing and QoL (Lim et al., 2014; Matzka et al., 2016; Stein et al., 2019). It is particularly concerning that the average age of breast cancer diagnosis in Ghana is almost two times lower compared to Europe and America (Thomas et al., 2017). This could be attributed to variations in the expression of breast cancer genes specific to each population. Ekwueme et al., (2014) estimates about one fourth of cancer survivals feel less productive at work and nearly one third face limitations in carrying out daily routine. This has led to employment induced depression in an already precarious situation (Munir et al., 2009; J.-H. Park et al., 2009; Paul et al., 2016; Qan’ir et al., 2022; Taskila & Lindbohm, 2007) and contributed to poor QoL. The effect of cancer is immense not just for persons living with the condition, but for their families as well as employers, reflecting the public health burden of the problem. Cancer treatment cost has been reported to have negative impact on the individual and family placing them at high risk of bankruptcy (Ekwueme et al., 2014; Yu et al., 2022). In the race to improve cancer care, studies have attempted to better understand the distress dynamics (Krasnoselskyi et al., 2022). For instance, Kong & Guan, (2019) established a strong reciprocal link between distress of PLWC and their support system. This social safety net has been shown to improve resilience. Despite extensive work on distress University of Ghana http://ugspace.ug.edu.gh 7 and QoL, there is scanty research on the causal pathway of the relationship by resilience and social support from Ghana. Kugbey, Oppong Asante, et al., (2020) examined the mediating roles of social support and religiosity with respect to depression and anxiety on QoL in a population of Ghanaian women with breast cancer and found significant associations though social support. Meanwhile, a systematic review of resilience in chronic illness, cited social support as a protective factor involved in resilience (Cal et al., 2015). Quality of life is one of the most important health outcomes used to assess healthcare quality and survivorship and this study seeks to evaluate how the effect of distress on QoL is influenced by resilience and social support. 1.3 Research Questions 1. What is the level of distress and quality of life amongst persons living with cancer? 2. What are the factors associated with distress and quality of life amongst persons living with cancer? 3. What is the mediation effect of resilience and social support on the relationship between distress and quality of life among persons living with cancer? 4. How does resilience and social support moderate the effect of distress on quality of life among persons living with cancer? 1.4 Aims and Objectives The main aim of the study is to assess the mechanistic effect of resilience and social support on the association between psychological distress and quality of life among persons living with cancer. 1.4.1 Specific Objectives The specific objectives are to: University of Ghana http://ugspace.ug.edu.gh 8 1. Estimate the prevalence of distress and quality of life among persons living with cancer. 2. Determine factors associated with distress and quality of life amongst persons living with cancer. 3. Determine the indirect effect of resilience and social support on the relationship between distress and quality of life amongst persons living with cancer. 4. Determine how resilience and social support modify the effect of distress on quality of life among persons living with cancer. 1.5 Justification The ever-increasing incidence, high prevalence, alarming death rate, and probable traumatic significance of cancer necessitates more research into identifying and understanding variables influencing QoL. Psychological distress has a substantial impact on the QoL for patients living with cancer. Evidence have shown that patients with moderate and severe distress experience worse outcomes. Although QoL of PLWC have been extensively studied and generally found to be adversely affected by distress, in the Ghanaian context, factors involved in the causal pathway have not been widely explored. Little is known about how much of the variance in QoL amongst PLWC in Ghana can be accounted for by resilience and social support. QoL is key to improving the mental health and well-being of PLWC and this study will not only contribute to the body of knowledge from Ghana on psycho-oncology but provide an evidence base platform for developing effective interventions. 1.6 Relevance of the study This research provides a source of valuable knowledge for stakeholders in cancer care and management. It contributes to the understanding of QoL challenges faced by PLWC and serves as an evidence-based platform for informed interventions that build resilience at the individual, University of Ghana http://ugspace.ug.edu.gh 9 family, and community levels. It strengthens the rich body of knowledge relating to cancer care and management, and aides in promoting comprehensive and meaningful social support systems through understanding the mechanistic effects of resilience and social support on the relationship between distress and QoL. 1.7 Hypothesis H1: Resilience does not mediate the relationship between distress and quality of life in persons living with cancer. H2: Social Support does not mediate the relationship between distress and quality of life in persons living with cancer. H3: Effect of distress on quality of life in persons living with cancer is not significantly different by resilience. H4: Effect of distress on quality of life in persons living with cancer is not significantly different by social support. 1.8 Theoretical and Conceptual Framework 1.8.1 Theoretical framework underpinning the study 1.8.1.1 The Psychological Capital Theory This study is rooted in theories of psychological capital (PsyCap) and biopsychosocial model. Influenced by Bandura’s theory in which an individual uses his abilities and strengths to succeed in life, the PsyCap theory encompasses an individual’s psychological development and evolution on a daily basis. According to the theory credited to Luthans and his colleagues (Luthans et al., 2007), a combination of personal qualities can be adjusted to improve well- being and performance. Studies reveal that PLWC with comparable illness severity and treatment status frequently exhibit considerably varying levels of psychological stress, most University of Ghana http://ugspace.ug.edu.gh 10 likely as a result of differences in resilience, optimism, self-efficacy, and hope (Leung et al., 2016; MacDonald et al., 2021; Moreno et al., 2018). Psycap idea is centered on a combination of these four-character traits that can be adjusted to enhance performance and well-being. The essence is the positive perspectives, perceptions, and consequence of an individual’s behaviour translating into good QoL. Developing even one of the PsyCap resources can help an individual persevere through adversity. For this study, the trait of interest is resilience, which is one of the two intermediate outcome variables. Resilience is a psychological construct that is frequently connected to social support (the second intermediate outcome variable) and life satisfaction. It was described by Liu et al., (2017) as "the ability to adapt to stress and adversity." Empirical evidence has shown resilience as a valuable psychological resource that can aid in the recovery of persons who have experienced turmoil (Wolf et al., 2018; Yıldırım & Arslan, 2022). Key resources required to cope with adversity are what the positive PsyCap theory offers. These resources are ‘state-like’, implying unlike fixed features, they can be developed and changed through intervention (Luthans et al., 2007). Given that social support serves as a buffer against the negative effects of distress on mental and physical health and well-being, as well as stimulates positive influence on resilience, it is the other intermediate outcome variable considered in this study. Social support is exceptionally important for maintaining good QoL and mental health as it reduces depression, anxiety, and stress (Ruiz-Rodríguez et al., 2022). 1.8.1.2 The Biopsychosocial model of health The other major theory underpinning this study is the biopsychosocial (BPS) model credited to George Engel (Engel, 1977). This theory advocates for a holistic approach to health, one that is mindful of the confluence of biological, psychological, and social connections defining all facets of human illness. This central proposition of the BPS model provides the impetus for this research. According to this model, the interplay between these three elements will University of Ghana http://ugspace.ug.edu.gh 11 determine the relative distress and in effect, the QoL experienced by PLWC. In PLWC, biological factors (e.g., infection due to weakened immunity or deep vein thrombosis as a side effect of chemotherapy), psychological factors (e.g., distress or negative mood) and social factor (e.g., social support or loss of job) interact to influence overall wellbeing or QoL. In this study biological factors such as demographic, disease, and treatment factors; psychological factors like distress, QoL, resilience and psychological services as well as social support and social resources comprising social factors, interact to influence the origin, progression, and resolution of cancer outcomes. Since these three factors are interdependent, any adverse effect in one domain frequently has negative implications for the others. It is thus crucial that health professionals involved in cancer care move beyond the traditional biomedical approach to recognizing the complex dynamics of factors influencing health for which the BPS model provides a more comprehensive approach. Combining these two models, highlights the significance of understanding the role personal traits as well as psychological and social factors play in the chronic illness trajectory thereby informing integrated interventions and improving psycho-oncology management. Based on these theoretical frameworks, a conceptual framework was developed as presented in figure 1.1 1.8.2 Conceptual framework underpinning the study Studies have confirmed that in comparison with the general population, PLWC experience higher level of distress resulting either from the unilateral impact of the diagnosis, the treatment being received, and the functional impairment or because of their combined effect (Leong Abdullah et al., 2015, 2015; Taghizadeh et al., 2018; Tareke et al., 2022). It is important to acknowledge that a patient's personality, psychological make-up, family, and social support system, as well as any physical impairments, will impact on how they react to receiving a cancer diagnosis and these factors can also affect their treatment outcome. Several other factors University of Ghana http://ugspace.ug.edu.gh 12 have been implicated as sources of high distress in PLWC. These include sociodemographic and economic factors, disease factors and treatment modalities (Choi et al., 2022; Roick et al., 2022; Velure et al., 2022; Yu et al., 2022). These cascade of factors, along with the high incidence and growing mortality rate, can be quite distressing. Long-term distress has been associated to high symptom burden, which translates to poor health and poor QoL (Nayak et al., 2017; Seiler & Jenewein, 2019; Zou et al., 2016). The study hypotheses and the corresponding objectives were conducted using the conceptual framework as presented in Figure 1.1. The framework attempts to establish the dynamics of the relationship between distress and QoL among PLWC (mixed cancers). It explores the mechanistic effect of resilience and social support on this relationship. The solid red lines indicate the hypothetical idea of the study. The blue lines indicate covariates' influences on the study outcomes, which were adjusted during statistical analysis. The framework depicts that, the relationship between distress and QoL is probably mediated and or moderated by one’s resilience and or social support. It is worth noting that both resilience and social support (perceived) have been found to correlate negatively with distress (Nigah et al., 2019; Yasien et al., 2016). This study held that factors including socio-demographic characteristics, disease factor and treatment modality were associated with distress and QoL as well as resilience and social support. All identified associated factors were adjusted to assess the possible mediating effect of resilience and social support as well as how these constructs modify the relationship between distress and QoL. University of Ghana http://ugspace.ug.edu.gh 13 Figure 1.1: Conceptual framework defining the causal pathway of distress and QoL among PLWC Psychological Distress (Primary Exposure) Quality of Life (Primary Outcome) Socio-demographic Factors (Confounders) Sex Age Level of education Employment status Asset Index NHIS Civil Status Number of children alive Religion Region of residence Number of people living at home Social Support (Intermediate variable) Resilience (Intermediate variable) Disease factor (Confounders) Cancer Type Cancer Stage Duration since diagnosis Family history of cancer Symptom(s) burden Comorbidities Treatment Modality Surgery Chemotherapy Radiation Palliative Care Number of interventions received. Psychological services received. D irect E ffect Indirect Effect Indirect Effect University of Ghana http://ugspace.ug.edu.gh 14 CHAPTER TWO 2.0 LITERATURE REVIEW 2.1 Introduction This chapter provides a summary of earlier studies on the subject matter of this research. The review of literature covered the prevalence of distress and QoL amongst PLWC as well as factors associated with these constructs in the population. Applying their appropriate search protocols, literature search was thoroughly done using electronic resources like PubMed/MEDLINE, Google scholar, Google, African journal online (AJOL), African Index Medicus and Scopus. Taking PubMed as an example, using Boolean operators, keywords and phrases adopted during the literature search included, but not limited to: 1. ((((Psychological distress)) OR (distress)) AND (cancer)) OR (persons with cancer) 2. (((Quality of life) OR (health related quality of life)) AND (cancer)) OR (persons living with cancer) 3. (((((psychological distress) OR (distress)) AND (quality of life)) AND (cancer)) AND (Africa)) OR (Sub-Saharan Africa) Search terms were identified in the title, abstract or the full text and resulting articles were examined to determine their appropriateness for inclusion. For searches conducted using the Google search engine, symbols such as (" "), around a phrase to look for the exact phrase, and (..) to look up a date within the search period were necessary. Search terms were explored in webpages and documents using the operators "allintext:", "filetype:", and "OR:" in this option as well. To a large extent, methodical search was carried out on publications from the year 2012 – 2022 and articles relevant to the focus of this work were thoroughly examined for information in line with this study. University of Ghana http://ugspace.ug.edu.gh 15 2.2 Prevalence of Psychological distress in persons living with cancer Cancer diagnosis and treatment continue to be extremely stressful events that can cause significant psychological distress because it is frequently viewed as a life-threatening condition (Ebob-Anya & Bassah, 2022; Grassi et al., 2017; Mason et al., 2019; Mehnert et al., 2018a). In PLWC, distress has been linked to poor QoL stemming from diminished functional status, reduced adherence to treatment and high symptom burden (Blenkiron et al., 2014; Peters et al., 2020a; Yee et al., 2017), all of which increase the risk of mental health challenges in this population (Anuk et al., 2019; Mehnert et al., 2014). The impact of distress on QoL of PLWC is well chronicled in literature for different cancer types and for different stages of the disease course (Nayak et al., 2017; Oh & Cho, 2020; Peters et al., 2020a; Prapa et al., 2021; Qan’ir et al., 2022). The evidence shows that cancer-related events are accompanied with extremely distressing emotional states that adversely affect health and life quality. Prevalence of distress in PLWC have yielded ambiguous findings, with Peters et al., (2020) reporting distress levels ranging from 30% to 60%. Yet Ebob-Anya & Bassah, (2022) in their descriptive cross-sectional hospital base study, reported considerable clinical distress in 69.2% of in patients in two prominent cancer management centres in Cameroon. To ascertain the overall prevalence of distress among geographically different cancer programs, Carlson et al., (2019) reported 46% significant distress amongst PLWC from 55 North American cancer treatment centers. In a systematic review of the prevalence of distress among PLWC in the Southeast Asia region, Ostovar et al., (2022) found distress measured as anxiety and depression to range from 7% to 88% and 3% to 65.5% respectively. Mason et al., (2019) in a hospital based observational cross study carried out in a tertiary care facility in northern India, observed distress of 38.5%. Mehnert et al., (2018) using participants drawn from five diverse study centers across Germany which included respondents from outpatient cancer care facilities as University of Ghana http://ugspace.ug.edu.gh 16 with this study, reported distress prevalence of 52%. While Arts et al., (2021) reported distress of 17% amongst participants diagnosed with lymphoma from 13 hospitals in Holland. Some studies focused on the magnitude of distress with Peters et al., (2020) reporting high distress of 65.9% among participants with mixed cancers from their retrospective study. Mehnert et al., (2018) identified 52% also in a population of mixed cancers which included patients receiving ambulatory care as with this study. All these findings confirm the ambiguity in distress prevalence rate. The differences in prevalence rates might be somewhat explained by the study populations, study settings (inpatient or outpatient), duration of study, and measures/scales employed. 2.2.1 Factors associated with psychological distress in persons living with cancer Studies have identified several risk factors associated with distress in PLWC. While reports show consistency with some factors, others have been equivocal. In their cross-sectional multicenter study in 5 diverse cancer centers across Germany, Mehnert et al., (2018) reported high distress in women, patients 60 years and above, the unemployed, advanced cancer stage as well as gynecological and pancreatic cancers. Interestingly, in comparing cancer-related distress in young adults (18 – 39 years) with older adults (middle age :40 - 65 years and senior adults 66 – 90 years), Burgoyne et al., (2015) observed higher distress in young and middle- aged adults compared with senior adults. Similar findings were observed in Peters et al., (2020) who noticed significant high distress in patients above 65 years, compared to patients between ages 50 and 64 years. Furthermore, patients with children were significantly less distress than their counterparts without children. Quality of life, symptom burden, financial status and education level have also been linked to distress. Ebob-Anya & Bassah., (2022) in assessing the level of distress and QoL in PLWC in their study in Cameroon, found distress to be University of Ghana http://ugspace.ug.edu.gh 17 significantly linked with QoL, and fatigue (p<0.001) as well as functionality, sleep disturbance and financial hardship (p<0.05). Kim et al., (2017) in investigating the prevalence and prognostic significance of psychological distress in patients with gastric cancer found significant higher distress in females (p=0.024), patients who were jobless (p=0.02), patients with lower educational background (p=0.021) as well as those in advanced stage of the disease (p=0.008). Patients with low educational background had 2.39 higher odds of experiencing high distress compared to the well-educated and those with advanced cancer reported 2.72 times higher odds of experiencing distress than those in the early stage.Abu-Odah et al., (2022) reported women with breast cancer to have higher odds of significant distress compared to stomach and colon cancers (OR = 4.81,95%CI1.22-18.94;2.62,95%CI 1.11-6.16). Patients who were newly diagnosed were also more likely to experience high distress while physical challenges independently influenced distress (OR=0.16,95%CI 0.04-0.60). 2.3 Level of Quality of life in persons living with cancer In relation to cancer, QoL has been cited as indicative of a patient’s social, psychological and physiological status, as well as their well-being (Tian & Hong, 2014). Studies have shown varying QoL levels in PLWC by geographic location and scale of measurement (Alam et al., 2020; Almigbal et al., 2019; Nayak et al., 2017). In a systematic review and meta-analysis of health-related QoL and its determinants amongst patients with breast cancer in Africa, the pooled estimates mean score of general QoL based on the EORTC QLQ-C30 standard instrument was 52.77 (95% CI: 42.199 to 63.345; I2 = 99.21%, P < 0.001) (Bitew et al., 2021). ElMokhallalati et al., (2022) on the other hand reported mean overall QoL to be 70.5 (SD 19.9) in their assessment of the association between sociodemographic and disease related characteristics, symptom burden and QoL amongst patients receiving outpatient services in University of Ghana http://ugspace.ug.edu.gh 18 Gaza. Ramasubbu et al., (2020) also found overall mean score of QoL to be 61.93±5.86 in their cross-sectional analytic study of QoL and factors affecting it in PLWC undergoing treatment in a tertiary care hospital. In a hospital-based cross-sectional analytical study of cervical cancer survivals in Ghana, overall QoL scores was 79.6 (SD 16.0) with 74.5% of survivals reporting good QoL score within the median follow-up time of 3 years 5 months following cancer diagnosis (Amo-Antwi et al., 2022). In a pilot study assessing QoL and association with nutritional and performance status of PLWC Alam et al., (2020) found the prevalence of low and very low QoL were 53.76% and 28.67% respectively. Nayak et al., (2017) in an exploratory survey of patients with mixed cancers in selected reputable cancer hospital in Karnataka, India, reported the prevalence of low QoL to be as high as 82.3%. 2.3.1 Factors associated with quality of life in persons living with cancer Studies have confirmed that QoL in PLWC is adversely affected by the psychological impact of the diagnosis, the treatment, and symptoms burden (Sharma & Purkayastha, 2017; Sheikhalipour et al., 2019). In a systematic review of quantitative studies on QoL among PLWC and their family caregivers in Sub-Saharan region, age and marital status were the two demographic characteristics that most frequently affected QoL. Cancer stage and type of treatment were cited as common cancer-related variables, while social determinants of health such as education, access to services and information, financial hardship, and place of residence (rural vs. urban) were the most common factors affecting QoL and its subdomains (Qan’ir et al., 2022). Qan’ir et al., (2022) further identified overall good QoL to be associated with younger age, the female sex, being married, early stage of the disease, higher education level and receiving a combination of chemo and radiotherapy. However, QoL correlated negatively with distress in the form of anxiety and depression as well as with challenges financing health needs. As with Qan’ir and colleagues, Ramasubbu et al., (2020) found education to be an University of Ghana http://ugspace.ug.edu.gh 19 important protective factor for QoL. In their study aimed at assessing QoL and factors affecting it in PLWC receiving chemotherapy, cancer treatment was also found to decrease functionality, increase fatigue, and consequently affect QoL. Illiteracy, financial constraints and unemployment were also found to negatively affect QoL. Nayak et al., (2017) also confirmed income to have significant association with QoL (P=0.006). Symptom burden (pain, sleep disturbance and fatigue) were also found to affect QoL. Megari, (2013) likewise documented limitations in daily activities as well as psychosocial elements such as troubled intermate relationship, sexual intimacy, distorted body image, maladaptive coping strategies, lack of social support and type of surgery (mastectomy) to be strongly associated with QoL. Interestingly, single participants, not receiving radiotherapy and end stage cancer were found to be linked with better QoL in a cross-sectional study carried out by ElMokhallalati et al., (2022). They also linked lower education level to poor QoL. Numerous literature have implicated distress as a key factor affecting QoL in PLWC (Ebob-Anya & Bassah, 2022; Kugbey et al., 2020; Ostovar et al., 2022; Prapa et al., 2021). 2.4 Mechanisms of Resilience and Social Support in Persons living with cancer A thorough review of the mediation and moderation literature from Ghana on cancer related QoL found one study (Kugbey et al., 2020), that focused on the indirect effect of anxiety and depression on QoL through social support and religiosity, but not resilience. Social support was found to mediate the effect of anxiety and depression on QoL. Search engines did not produce any more results using key terms and Boolean operators combined. This poses a challenge presenting review of literature from Ghana. Nevertheless, in other jurisdictions, resilience and social support have been researched in relation to their mediating and/or moderating roles. While in the United States Wu et al., (2015) observed resilience to mediate the impact of distress resulting from cancer symptoms on QoL, Lee & Kim., (2018) established University of Ghana http://ugspace.ug.edu.gh 20 no mediating effect of resilience on the relationship among uncertainty, distress, and health- related QoL in Korea. In a population of patients from Italy with prostate cancer, social support was found to mediate the relationship of body image distress and depressive symptoms which inevitably affects QoL (Scandurra et al., 2022). Poręba-Chabros et al., (2020) observed social support to moderate function in the formation of cancer stress response. Huang & Hsu., (2013) also confirmed social support as a moderator between depression and QoL in survivals of breast cancer. With regards to resilience, Groarke et al., (2020) found the impact of stress on distress was greatest when resilience was low and watered down when resilience was high. 2.5 Knowledge Gap Persons living with cancer are more at risk of developing high distress, a fact well chronicled in literature offering evidence of the debilitating effect of cancer related distress on QoL (Bitew et al., 2021; Ebob-Anya & Bassah, 2022; Kugbey et al., 2020; Ostovar et al., 2022). In addition to demographic, disease and treatment factors, resilience and social support have been cited by systematic reviews as factors influencing QoL in PLWC (El Haidari et al., 2020; Sousa et al., 2019), however their possible mediating or moderating roles on the relationship between distress and QoL has not been thoroughly explored in the Ghanaian context. A thorough review of literature found one study that focused on social support and religiosity as mediators (Kugbey et al., 2020). Elsewhere, studies have found resilience to mediate cancer related QoL with respect to cancer symptom distress (Wu et al., 2015) and psychological predictors (Zhou et al., 2022). Zhang et al., (2017) found social support to partially mediate the relationship between resilience and QoL. A systematic review of resilience in chronic illness, cited social support as a protective factor involved in resilience (Cal et al., 2015). Ruiz-Rodríguez et al., (2022) found social support and resilience to improve health quality in PLWC. Social support University of Ghana http://ugspace.ug.edu.gh 21 improved patients’ general health, improved coping, and reduced patients’ symptoms while resilience enhanced general health, functioning and improved symptoms. To improve QoL in the study population, it is crucial that we understand the mechanistic effect of these constructs on health in order to develop appropriate evidence-based interventions. Furthermore, seeing that QoL of life has a sociocultural context and considering the dearth of literature from Ghana examining the study’s objectives, there is a compelling need to bridge the research gap. University of Ghana http://ugspace.ug.edu.gh 22 CHAPTER THREE 3.0 METHODS 3.1 Introduction Using an analytical facility-based cross-sectional study design, this study sought to assess the mechanistic effect of resilience and social support on the association between psychological distress and quality of life among PLWC. This chapter provides information by which the validity of the study is judged. It describes the study site, the research approach, the study population, sample size calculation and sampling method, the research tool and scales used, as well as the analysis plan to address the research objectives. 3.2 Description of site The study was carried out at Ghana's top tertiary medical centre, the Korle Bu Teaching Hospital. The facility, which was set up on October 9, 1923, currently has over 2000 bed capacity. On a daily basis, ambulatory services account for 1,500 patients and inpatient admissions account for 250. The Hospital has 21 clinical and diagnostic departments and three Centres of Excellence, namely the National Radiotherapy Oncology and Nuclear Medicine Centre, the National Reconstructive Plastic Surgery and Burns Centre and the National Cardiothoracic Centre. Departments include Internal Medicine and Therapeutics, Child Health, Surgery, Obstetrics and Gynaecology, Anaesthesia, Family Medicine/Polyclinic, Accident & Emergency, Psychiatry and Accident & Orthopaedics. Others are Pharmacy, Pathology, Laboratory, and Radiology (KBTH, 2020). The NRONMC offers a range of cancer care services supported by cutting-edge technology catering for about 1500 patients from various parts of the country and the sub region. 3.3 Research Approach The study approach consisted of a quantitative method with the use of a questionnaire purposely designed for this research (Appendix 4). University of Ghana http://ugspace.ug.edu.gh 23 3.4 Research design An analytical facility-based cross-sectional study design was adopted to assess the causal pathway of psychological distress on quality of life by resilience and social support. To assess the study’s constructs, this approach was adopted given the ease of studying multiple variables simultaneously to draw a picture of the current happenings in the population regarding the objectives. This approach enables description and explanation of possible relationships and thus provides a basis for making inferences and gathering preliminary data to support further research. 3.5 Study population The study participants were PLWC undergoing clinical care at the National Radiotherapy Oncology and Nuclear Medicine Centre (NRONMC) of KBTH. 3.6 Inclusion and exclusion criteria Eligibility criteria included adults living with cancer aged 18 years and above who were undergoing clinical care at the NRONMC in Korle-Bu and gave consent to participate in the study. PLWC who were critically ill and not able to communicate or had previous history of any mental illness as per their medical records, were excluded. 3.7 Sample size Since the study sought to answer the question “what is the mechanistic effect of resilience and social support on the distress-QoL relationship?”, sample size calculation was done taking distress and QoL as outcome variables. Considering the wide range of reported prevalence for these constructs, the 50% assumption was applied to give the largest sample size or maximum variability, which gives the study sufficient power. This study adopted the Cochran formula (1963:75) for the sample size calculation (Cochran, 1963). University of Ghana http://ugspace.ug.edu.gh 24 n= (𝑍𝛼/2) 2𝑝𝑞 𝑒2 ; Where n = sample size required, Zα/2 = Zscore (normal deviates for Type I error), p = Proportion population, q = 1-p, e = margin of error. For this study the level of confidence is pegged at 95%, giving an ∝ = 0.5 with a standard normal deviate value of 1.96. The assumed proportion of population experiencing exposure outcome = 50.0%, therefore P = 0.50. The study aims to estimate the prevalence within 5 percentage points of the true population proportion, therefore the margin of error (e) = 0.05. The formula thus reads: n = 385 The sample size arrived at was thus 385. Taking into account a non-response rate of 5%, the minimum sample size required for this study was 405. 3.8 Sampling method and data collection procedure Systematic sampling technique with a calculated sampling interval was employed to sample participants for this study. The idea behind this approach is that the study was conducted in tertiary healthcare facility operating a first come, first serve system for outpatients. At the NRONMC, even though patients are given appointments, consultation is based on this systematic process. The centre provides service to patients from Monday to Friday with an average daily attendance of 80. Sampling interval was calculated as below: Number of days estimated for data collection = 10 Daily participants to be interviewed = 405/10 ≈ 41 Approximate daily attendance = 80 Therefore, daily sampling interval = 41/80 ≈1/2 This daily sampling interval implied that on the average, half of the daily attendance were to be recruited for the study; for every two patients attending the clinic, one was sampled. This sampling 𝑛 = (1.96)2 𝑥 0.5(1 − 0.5) (0.05)2 University of Ghana http://ugspace.ug.edu.gh 25 frame was generated daily throughout the period of data collection based on this sampling methodology. To be able to sample the patient, a simple random sampling method was adopted among the first 2 patients and subsequently adopted the interval to select participants sequentially. This is to say, systematic sampling was used to recruit participants. However, simple random sampling was adopted to determine the starting point for the systematic sampling frame. The opportunity was given to the next patient if identified respondent refused to partake. Data collection was done on every clinic day starting from the first week of August to the first week of September 2022. Data collectors were trained on how to effectively collect the data and special codes were used to distinguish participants and to detect duplicates. About 15-20 minutes were spent interviewing and taking participants through the data collection tool. 3.9 Questionnaire piloting/pretesting A total of eight research assistants were recruited and trained for data collection. The data collection instrument was pre-tested on a group of patients living with cancer who were attending the Haematology Day Care and Breast Clinic, both located within the Korle-bu Teaching Hospital. This was carried out in the last week of July 2022 (25th to 27 of July 2022). After pretesting, appropriate modifications were made to the questionnaires based on feedback from the sampled population. 3.10 Measurement Scales The NRONMC is a facility that attracts interest from researchers all year round. In order to reduce the response burden or fatigue of participants in the study population who are already experiencing significant symptom burden, this study opted for shorter versions of standardized tools where applicable. As such, the revised psychological distress inventory (PDI-R), the short form-8 health survey for HRQoL, and the Conor-Davidson brief Resilience scale (CD-RISC 10) with minimum items were adopted along with the Multidimensional Scale of Perceived Social Support (MSPSS). University of Ghana http://ugspace.ug.edu.gh 26 3.10.1 The Revised Psychological Distress Inventory (PDI-R) Psychological distress was measured using the Revised Psychological Distress Inventory (PDI-R). This is an 8-items scale in which each item was scored on a five-point scale ranging from 1 to 5, with 1 representing least distress experienced ("not at all”) and 5 indicating highest distress experienced ("very much”). The composite score was standardized and converted into percentages and scores below 50 was considered as low and those above 50 indicative of high (Abdi, 2007; Frey, 2018). This scale is the revised version of Morasso et al., 1996 psychological distress inventory (PDI). Rossi et al., 2022 investigated the psychometric properties of the original PDI and found it to lack a clear structure. This led to the provision of this revised version, showing more robust psychometric qualities, stronger factorial structure, and is age and gender invariant. It is a valid indicator of distress in various populations of patients with cancers is highly recommended for use in clinical and research settings to enhance accuracy of diagnosis and management of psychological distress in patients with oncological issues. The revised 8 items PDI scale was effective at discriminating between subjects with high or low level of internal and external distress. A Cronbach coefficient of 0.776 for its internal distress subscale and 0.754 for its external distress subscale. The general distress scale showed a Cronbach’s alpha of 0.853. The tool can be found in Appendix 5. 3.10.2 Short Form-8 (SF-8) Health Survey for HRQoL Quality of life was gauged using the Short Form-8 (SF-8) Health Survey. This is an 8-items scale with scores ranging from 0 to 100 for each domain. Six of the items (2, 3, 5, 6, 7, 8) were scored on a scale of 1 to 5, with 1 denoting the best QoL and 5 the worst. Two of the items (1, 4) had score ranging from 1 to 6. A composite score was generated, and scores were standardized and generated into percentage. The SF-8 produces two summary measures: Physical component summaries (PCS) and mental component summaries (MCS). As with all PDI-R, higher PCS and MCS scores suggest University of Ghana http://ugspace.ug.edu.gh 27 a higher QoL. SF-8 demonstrated very good reliability in a population-based study conducted in Spain, with a Cronbach's alpha coefficient of 0.92 and high item-total correlations ranging from 0.57 to 0.93. In a different study on the Chinese population, the item-total correlations for all eight items, with the exception of VT (r=0.39), were moderate to high (r > 0.5). The aggregate Cronbach's alpha for these items was 0.85. 3.10.3 Conor-Davidson brief Resilience scale (CD-RISC) Resilience was measured using the Conor-Davidson brief Resilience scale (CD-RISC-10) (Seiler & Jenewein, 2019). The CD-RISC-10 is a 10-item, unidimensional self-report scale that measures five components of resilience. Items relate to flexibility (1 and 5), sense of self-efficacy (2, 4 and 9), capacity to manage emotion (10), optimism (3, 6 and 8), and cognitive focus under stress, the scale primarily measures hardiness (7). A score of 0 indicates that the resilience statement is not true at all, while a score of 4 indicates that the statement is true almost always. Each item is graded on a five-point scale ranging from 0 to 4. The sum of the 10 items yields the final score. Therefore, the result can extend from 0 and 40. Higher scores indicate higher resilience, whereas lower scores indicate less resilience or a harder time overcoming distress. For this study, scores were standardized and converted into percentages and scores below 50 was considered as low and those above 50 indicative of high. Resilience may be measured well with CD-RISC thanks to its high psychometric quality, good internal consistency, and robust psychometric features (Ekem-Ferguson et al., 2022). Several research measuring resilience in patients with cancer have made used of the CD-RISC. Data obtained from the CD-RISC scale using PLWC authenticate findings from Connor and Davidson and demonstrates that PLWC who are resilient, are well adjusted to their condition and experience less psychological distress(Seiler & Jenewein, 2019). University of Ghana http://ugspace.ug.edu.gh 28 3.10.4 Multidimensional Scale of Perceived Social Support (MSPSS) This is a 12-item, seven-point scale ranging from 1 to 7 (Zimet et al., 1988). The response scale ranged from 12 to 84, however, the scores were standardized and converted into percentages ranging 1-100. The MSPSS has demonstrated good internal consistency and correlated negatively with depression. Additionally, support for convergent validity was determined by assessing the correlation between perceived family and friends support and satisfaction with family and friends’ measure. Similar to the subscales of family and friends, univariate testing revealed a significant gender difference for each of these domains, with males exhibiting stronger perceptions of support in both. The MSPSS's psychometric features have revealed that it is reliable for measuring support perceptions in the Ghanaian setting (Wilson et al., 2017). Table 3.1: Variable description, measurement, and scale of measurements Variable Type of variable Description Measurement Scale of measurement Quality of life Primary outcome/ dependent Patient Quality of Life Raw counts as a continuous measure and categorical. Continuous Binary Resilience Immediate outcome measures/Mediator Resilience level among participants Raw counts Continuous Social support Immediate outcome measures/ Mediator Level of social support among participants Raw counts Continuous Psychological Distress Exposure outcome/Independent variable Patient psychological distress Raw counts as a continuous measure and categorical. Continuous Binary Age Confounder Age group of the participant ≤39, 40-49,50-59, 60- 69, 70+ Categorical Sex Confounder Sex differential of participant Male or Female Binary Education level Confounder Educational level of participant None, basic, SHS, Tertiary Categorical Employment Status Confounder Employment status of participant Unemployed, student, employed (non-self), Categorical University of Ghana http://ugspace.ug.edu.gh 29 Variable Type of variable Description Measurement Scale of measurement Employed (self), pensioner Asset Index Confounder Income range of participant Low, Medium, High Categorical Valid NHIS card Confounder Validity of NHIS card Yes or No Categorical Civil status Confounder The marital status of the participant Divorced/separated, Single, Married, Widow/Widower Categorical Number of Children alive Confounder Participant’s number of children alive None, 1-2, 3-5, 6+ Categorical Religion Confounder Religious affiliation of the participant Christian, Islam Categorical Region Confounder Participant’s region of residence Greater Accra, Central, Eastern, other Categorical Cancer type Confounder Participant’s confirmed cancer diagnosis Breast, Prostate, Gynae, Head &neck, Unaware, Others Categorical Duration since diagnosis Confounder Duration since participant was diagnosed < 1 year, 1 year, 2+ years Categorical Cancer stage Confounder Participant’s cancer stage Stage I, stage II Advanced, unaware Categorical Family history of cancer Confounder Participant’s family history of cancer No, yes, unaware Categorical Symptom burden Confounder Symptom(s) burden of participants None, 1-2, 3-5, 6+ Categorical Comorbidity Confounder If participant has any other comorbid condition(s) No, yes Binary Therapeutic Intervention Confounder Number of interventions received by participants Yet to start, 1-2, 3+ Categorical University of Ghana http://ugspace.ug.edu.gh 30 Variable Type of variable Description Measurement Scale of measurement Ever received psychological service at facility Confounder Participant’s report of ever receiving psychological service at the facility No, yes Binary 3.11 Data analysis Data was collected into EpiData version 3.1 and then exported to Stata version 16.1 for data cleaning, processing, and analysis. Details of analysis for each objective are presented below. 3.11.1 Objective 1: The level of psychological distress and QoL amongst PLWC To estimate the level of distress and QoL amongst PLWC, descriptive tabulation of the various outcomes was performed. Test of nonlinear hypotheses after prevalence estimation was adopted to assess significant difference within independent variables. 3.11.2 Objective 2: factors associated with psychological distress and QoL amongst PLWC Under this objective, distress and QoL were considered as outcomes coded as 1 “High distress and good QoL respectively as well as 0 for “Low distress and poor QoL respectively”. For this reason, logistic regression was performed to assess both crude and adjusted odd ratios of the outcomes by the independent variables. Multicollinearity was checked using variance inflation factor (VIF); however, no variable violated the rule of thumb (VIF≥10). 3.11.3 Objective 3 and 4: Mediation effect of Resilience and Social Support on the relationship between psychological distress and QoL amongst PLWC A causal path analysis was adopted to assess the mediation effect of Resilience and Social Support on the relationship between distress and QoL. This causal path analysis offers counterfactual direct, indirect, and total effects of distress on QoL (Muthén & Asparouhov, 2015). All estimations were University of Ghana http://ugspace.ug.edu.gh 31 done using the 95% confidence interval and a p-value≤0.05 was deemed significant. Initial analysis showed insignificant mediation effect of resilience and social support on the relationship between distress and QoL. 3.11.4 Objective 5: Effect Modification of Resilience on QoL among PLWC Moderation analysis involving effect modification was employed to assess association between distress and QoL by resilience and social support. The assumption was tested using the Cochran– Mantel–Haenszel test. The test of homogeneity was statistically significant (p<0.05) for resilience. The association of distress with QoL was thus found to be statistically significant depending on level of resilience. The degree to which distress influence QoL differed between low and high resilience. Adjusted logistic regression was thus performed to show association between distress and QoL modified depending on the level of resilience. 3.12 Ethical considerations Ethical approval for this study was granted by the Korle-Bu Teaching Hospital Scientific and Technical Committee Institutional Review Board (KBTH-STC/000118/2022 and KBTH- STC/IRB/000118/2022). Administrative permission was obtained from the National Radiotherapy Oncology and Nuclear Medicine Centre. Participants were informed of the study objectives both orally and in writing. Additionally, they were made aware of their right to discontinue participation in the study at any time. All respondents provided signed informed consent for this study participation. Information obtained was kept private with password protection on the computers used. Codes were used to identify respondents and were treated strictly confidential. Participants were asked to contact the KBTH-IRB on +233302667759/673034-6 from Mondays to Fridays between 8am-5pm for any issue regarding the study. University of Ghana http://ugspace.ug.edu.gh 32 CHAPTER FOUR 4.0 RESULTS 4.1 Introduction This study sought to estimate the prevalence of distress and state of QoL as well as determine factors associated with both constructs. It further sought to determine the mechanic effect of resilience and social support on the relationship between distress and QoL among PLWC. This chapter, in a logical sequence, thus summarizes and presents the findings of the study in context with the research objectives. 4.2 Sociodemographic and cancer related characteristics of study participants The study involved 469 Persons Living with Cancer (PLWC). Most respondents were aged 50-59 years (22.81%) with mean±standard deviation of 55.1±14.8 years. More than half of the respondents were females (70.8%), with a larger percentage being senior high school graduates (34.9%) and non-income earners (51.4%) who comprised the unemployed, students and pensioners. Over half of the respondents were married (59.9%) and most had between three to five children (n=49%). Majority were Christians (87.6%) and resided in the Greater Accra region (69.1%). The bulk of respondents had a valid national health insurance scheme (NHIS) card (91.3%) and a large portion fell within the low asset index category (53.9%) (Table 4.1). The most common cancers were breast, gynecological (cervical, endometrial, and ovarian) and prostate. Forty-three percent of respondents were oblivious of their cancer stage while a quarter reported the disease being advanced. The majority confirmed being diagnosed within one year of this study and more than half indicated having no family history of cancer. Three-fourth of respondents recounted experiencing at least three cancer related symptoms (pain, fatigue, weight loss, sleep disturbance and lack of appetite) and over half reported never receiving any form of psychological services prior to commencing University of Ghana http://ugspace.ug.edu.gh 33 therapeutic interventions even though over two thirds had already received between 1-2 interventions. Table 4.1: Sociodemographic and cancer related characteristics of study participants Variables Response Frequency Percent Age Groups (n=469) ≤39 73 15.57 40-49 96 20.47 50-59 107 22.81 60-69 101 21.54 70+ 92 19.62 Mean±SD 55.1±14.8 Sex (n=469) Male 137 29.21 Female 332 70.79 Education Level (n=469) No formal education 65 13.86 Basic 132 28.14 Senior high 164 34.97 Tertiary 108 23.03 Employment Status (n=469) Unemployed 130 27.72 Student 15 3.2 Employed (self) 48 10.23 Employed (non-self) 180 38.38 Pensioner 96 20.47 Asset Index (n=469) Low 253 53.94 Middle 89 18.98 High 127 27.08 Valid NHIS Card (n=649) No 41 8.74 Yes 428 91.26 Civil Status (n=469) Single 85 18.12 Married 281 59.91 Divorced/Separated 44 9.38 Widow/Widower 59 12.58 Number of children alive (n=456) None 59 12.58 1-2 116 24.73 3-5 230 49.04 6+ 64 13.65 Region of residence (n=469) Greater Accra 324 69.08 University of Ghana http://ugspace.ug.edu.gh 34 Variables Response Frequency Percent Central Region 55 11.73 Eastern Region 39 8.32 Other Regions 51 10.88 Religion (n=469) Christian 411 87.63 Muslim 58 12.37 Cancer Type (n=469) Abdominal 21 4.48 Breast 146 31.13 Gynaecological 79 16.84 Head & Neck 35 7.46 Prostate 76 16.2 Other Cancers 52 11.09 Unaware 60 12.79 Cancer Stage (n=469) Stage 1 109 23.24 Stage 2 40 8.53 Advanced 117 24.95 Unaware 203 43.28 Duration since diagnosis (n=469) <1 year 169 36.03 1 year 112 23.88 2+ years 148 31.56 Unaware 40 8.53 Family history of cancer (n=469) No 305 65.03 Yes 81 17.27 Unaware 83 17.7 Symptom(s) burden (n=469) ≤2 114 24.31 3-5 181 38.59 6+ 174 37.1 Comorbidities (n=469) No 258 55.01 Yes 211 44.99 Ever received psychological counselling (n=423) No 230 54.37 Yes 193 45.63 Therapeutic Intervention(s) (n=469) None 80 17.06 1-2 369 78.68 3+ 20 4.26 University of Ghana http://ugspace.ug.edu.gh 35 4.3 Prevalence of distress and quality of life among persons living with cancer 4.3.1 Level of psychological distress and quality of life amongst study respondents Psychological distress (distress) scores ranged from 38.71 to 78.16 with an overall mean score of 50.00 and high distress prevalence of 45.42% (95%CI=40.95-49.96). Quality of Life scores ranged from 22.14 – 68.66, with an overall mean score of 49.99 and good QoL prevalence being 52.67% (95%CI=48.12-57.16) amongst the respondents (Table 4.2). Table 4.2: Level of psychological distress and quality of life amongst study respondents Outcome Raw Scores (Min -Max) Mean ±SD Prevalence 95%CI Psychological distress 38.71 – 78.16 50.00 ±10 45.42 40.95-49.96 Quality of Life 22.14 – 68.66 49.99 ±10 52.67 48.12-57.16 4.3.2 Prevalence of high psychological distress and good quality of life by sociodemographic characteristics Differences in high distress by sociodemographic characteristics was significantly associated with education level, employment status, asset index, civil status, and number of children alive (p- value<0.05). Analysis showed that those with no formal education had a significant high proportion of high distress (58.46%; 95%CI=46.18-69.76). Among employment category, high distress was significantly observed among students (66.67%; 95%CI=40.53-85.43) and the unemployed (52.30%; 95%CI=43.71-60.76). High distress was also observed in respondents with low asset quintile (51.78%; 45.61-57.89). Patients who were single and those with no children recorded high distress prevalence of (63.52%; 95%CI=52.80-73.06) and (52.17%; 95%CI=37.91-66.08) respectively (Table 4.3). The differences in proportions of good QoL by socio-demographic showed insignificant association (p-value≥0.05) even though good QoL was observed in those with tertiary educational (63.88%; 95%CI=54.41-72.39) (Table 4.3). University of Ghana http://ugspace.ug.edu.gh 36 Table 4.3: Prevalence of high psychological distress and good quality of life by sociodemographic characteristics of participants Outcome domain Variable Psychological distress Quality of Life Prevalence[95%CI] Prevalence[95%CI] Age group ≤39 49.31 [38.05-60.65] 57.53 [45.96-68.32]] 40-49 53.12 [43.12-62.87] 43.75 [34.17-53.81] 50-59 44.85 [35.70-54.37] 49.53 [40.16-58.93] 60-69 39.60 [30.53-49.44] 58.41 [48.57-67.62] 70+ 41.30 [31.70-51.61] 55.43 [45.16-65.25] Test 4.04 (0.256) 5.54 (0.136) Sex Male 43.06 [35.01-51.49] 59.12 [50.68-67.05] Female 46.38 [41.06-51.78] 50.00 [44.62-55.37] Test 0.43 (0.510) 3.31 (0.069) Education Level No formal education 58.46 [46.18-69.76] 46.15 [34.45-58.29] Basic 55.30 [46.72-63.57] 46.21 [37.87-54.76] SHS 39.02 [31.84-46.71] 53.04 [45.37-60.57] Tertiary 35.18 [26.75-44.64] 63.88 [54.41-72.39] Test 11.31 (0.035) 1.69 (0.429) Employment Status Unemployed 52.30 [43.71-60.76] 45.38 [37.02-54.01] Student 66.67 [40.53-85.43] 66.67 [40.53-85.43] Employed (self) 31.25 [19.77-45.59] 64.58 [50.19-76.74] Employed (non-self) 47.22 [40.02-54.54] 51.67 [44.36-58.89] Pensioner 36.45 [27.44-46.53] 56.25 [46.18-65.83] Test 9.89 (0.019) 3.65 (0.301) Asset Index Low 51.78 [45.61-57.89] 48.61 [42.49-54.78] Middle 41.57 [31.80-52.05] 56.17 [45.72-66.11] High 35.43 [27.59-44.13] 58.26 [49.50-66.53] Test 10.16 (0.006) 3.75 (0.153) Valid NHIS Card No 46.34 [31.82-61.51] 51.21 [36.32-65.98] Yes 45.32 [40.65-50.08] 52.80 [48.05-57.50] Test 0.02(0.9010) 0.04 (0.846) Civil Status Divorced/Separated 43.18 [29.47-58.02] 45.45 [31.49-60.17] Single 63.52 [52.80-73.06] 44.70 [34.49-55.38] Married 40.21 [34.62-46.07] 56.93 [51.06-62.62] Widow/Widower 45.76 [33.55-58.50] 49.15 [36.68-61.72] Test 15.29(0.002) 5.49 (0.139) Number of Children alive None 52.17 [37.91-66.08] 58.69 [44.10-71.90] 1-2 34.48 [26.39-43.58] 60.34 [51.17-68.84] 3-5 47.82 [41.42-54.29] 49.13 [42.70-55.58] 6+ 48.43 [36.49-60.56] 50.00 [37.95-62.04] Test 6.50 (0.038) 4.19 (0.123) University of Ghana http://ugspace.ug.edu.gh 37 Outcome domain Variable Psychological distress Quality of Life Religion Christianity 45.01 [40.25-49.86] 52.55 [47.70-57.35] Islam 48.27[35.77-61.00] 53.44 [40.63-65.82] Test 0.22 (0.641) 0.02 (0.898) Place of Residence Greater Accra 42.59 [37.30-48.05] 53.08 [47.62-58.47] Central Region 50.91 [37.90-63.79] 45.45[32.86-58.65] Eastern Region 58.97 [43.12-73.15] 43.58 [29.06-59.30] Other Regions 47.05 [33.87-60.67] 64.70 [50.75-76.52] Test 4.74 (0.192) 5.72 (0.126) 4.3.3 Prevalence of high psychological distress and good quality of life by cancer related characteristics Differences in high distress by duration since diagnosis, symptom burden, comorbidities and receiving of psychological services prior to commencement of therapeutic intervention reflected significant association (p-value <0.05). Descriptive analysis showed the prevalence of high distress to be pronounced in respondents diagnosed less than a year of the study [50.29%;95%CI=42.78- 57.79]. In relation to symptoms burden, persons experiencing at least six symptoms recorded the largest percentage of high distress [52.87%; 95%CI=45.42-60.19]. Respondents with no comorbid conditions as well as those who reported not receiving any psychological services accounted for elevated prevalence of high distress [50.77%; 95%CI=44.67-56.84) and 53.91%; 95%CI=47.42- 60.27 respectively] (Table 4.4). Differences in proportion of good QoL by symptoms burden, comorbidities and receiving of psychological services before treatment indicated significant associations as well (p-value <0.05). Respondents reporting 1-2 cancer related symptoms experienced the highest percentage of good QoL [69.36%; 95%CI=60.17-77.24] while those with comorbid conditions reported the best good QoL (60.81%; 95%CL=53.41-66.59). Participants who reported to have received psychological services accounted for greater good QoL (60.10%; 95%CI=53.01-66.79) (Table 4.4). University of Ghana http://ugspace.ug.edu.gh 38 Table 4.4: Prevalence of high psychological distress and good quality of life by cancer related characteristics of participants Outcome domain Cancer related factors Psychological distress Quality of Life Prevalence[95%CI] Prevalence[95%CI] Cancer Type Abdominal Cancers 42.85 [23.96-64.08] 52.38 [31.78-72.19] Breast Cancer 39.72[32.09-47.89] 52.05 [43.94-60.05] Prostate Cancer 34.21 [24.43-45.54] 67.10 [55.79-76.72] Gynae Cancers 53.16 [42.15-63.87] 45.56 [34.93-56.62] Head & Neck Cancers 51.42 [35.25-67.30] 54.28 [37.86-69.82] Other Cancers 50.00[36.69-63.30] 48.07 [34.91-61.50] Test 8.51 (0.130) 9.13 (0.104) Duration since diagnosis < 1 year 50.29 [42.78-57.79] 51.47 [43.94-58.94] 1 year 46.42 [37.39-55.70] 58.03 [48.69-66.83] 2+ years 35.81[28.48-43.86] 53.37 [45.30-61.28] Test 7.26 (0.0265) 1.20 (0.549) Symptoms burden None 33.33[4.31-84.72] 1-2 24.32 [17.22-33.17] 69.36 [60.17-77.24] 3-5 51.38 [44.10-58.60] 51.93 [44.64-59.13] 6+ 52.87 [45.42-60.19] 41.95 [34.82-49.43] Test 24.14 (<0.001) 216.53 (<0.001) Comorbidities No 50.77 [44.67-56.84] 46.51 [40.49-52.63] Yes 38.86 [32.50-45.62] 60.81 [53.41-66.59] Test 6.77 (0.009) 8.91 (0.003) Ever received psychological service No 53.91[47.42-60.27] 46.52 [40.15-53.00] Yes 34.19 [27.83-41.18] 60.10 [53.01-66.79] Test 17.31 (<0.001) 7.94 (0.005) Number of interventions Yet to start 57.50 [46.45-67.84] 51.25 [40.38-61.99] 1-2 42.81 [37.84-47.93] 53.38 [48.26-58.43] 3+ 45.00 [25.27-66.43] 45.00 [25.27-66.43] Test 5.80 (0.055) 0.62 (0.734) University of Ghana http://ugspace.ug.edu.gh 39 4.4 Factors associated with high psychological distress and good quality of life amongst study participants. 4.4.1 Sociodemographic characteristics associated with high psychological distress and good quality of life amongst study participants Table 4.5 shows the crude and adjusted sociodemographic factors associated with high distress and good QoL. For high distress, univariate estimates showed significant association with educational level, employment status, asset index, civil status, and number of children alive. Multivariate analysis on the other hand confirmed significant association by education level, civil status, and number of children alive (p-value < 0.05). Respondents with no formal education were 2.19 times (95%CI=1.22-3.95) more likely to experience higher distress compared to respondents with senior high education. This was also reflected in the adjusted estimates with no formal education showing over two folds odds of high distress compared to senior high school level [aOR(95%CI=2.64(1.23-5.64). Relative to pensioners, students had higher odds of experiencing high distress [OR (95%CI=3.48 (1.96-11.03)]. No statistically significant association was identified from adjusted estimates for employment status. In relation to asset index, respondents in the low quantile showed higher odds of 1.96 compared to those in the high quantile. Though this trend appeared similar for adjusted estimated, it was not statistically significant. Compared with respondents who were married, persons who identified as single had higher odds of 2.59 and 3.93 respectively for crude and adjusted estimates. In the context of number of children, respondents with 1- 2 children had over two folds higher odds of high distress compared to those with none [OR (95%CI=32.07 (1.03-4.15)]. The odds were higher for the adjusted estimates for respondents with at least six children in comparison to the reference group [aOR (95%CI=3.00 (1.30-6.94)] (Table 4.5). In the context of QoL domain, civil status, and number of children alive were the two predictor variables with statistically significant associations. Respondents who were single experienced close University of Ghana http://ugspace.ug.edu.gh 40 to borderline significant reduction of 39% in the odds of having good QoL [aOR(95%CI=0.61(0.37- 1.00)] compared to those who were married. Relative to respondents with 1-2 children, those with 3-5 had 37% reduced odds of experiencing good QoL [OR(95%CI=0.63(0.40-1.00). The adjusted analysis showed patients with 3 to 5 children had reduced odds of 63% for good QoL compared to those s with 1-2 children [aOR(95%CI=0.37(0.22-0.63) (Table 4.5). Table 4.5: Crude and adjusted estimates showing sociodemographic characteristics associated with high psychological distress and good quality of life amongst study participants Outcomes Social-demographic factors Psychological Distress Quality of Life OR[95%CI]p-value aOR[95%CI]p-value OR[95%CI]p-value aOR[95%CI]p-value Age group 70+ ref ref ≤39 1.38 [0.74-2.56]0.305 1.09 [0.58-2.03]0.787 40-49 1.61 [0.90-2.87]0.106 0.63 [0.35-1.11]0.110 50-59 1.15 [0.66-2.03]0.614 0.79 [0.45-1.38]0.407 60-69 0.93 [0.52-1.66]0.810 1.13 [0.64-1.99]0.676 Sex Male ref ref Female 1.14 [0.76-1.70]0.512 0.69 [0.46-1.03] Education Level SHS ref ref ref No formal education 2.19 [1.22-3.95]0.008 2.64[1.23-5.64]0.013 0.76 [0.43-1.35]0.348 Basic 1.93 [1.21-3.08]0.006 2.14 [1.14-4.01]0.017 0.76 [0.48-1.20]0.243 Tertiary 0.85 [0.51-1.41]0.523 1.22 [0.61-2.45]0.572 1.56 [0.95-2.57]0.078 Employment Status Pensioner ref ref ref Unemployed 1.91[1.11-3.28]0.019 0.97 [0.45-2.07]0.940 0.64[0.37-1.10]0.107 Student 3.48 [1.96-11.03]0.034 0.83 [0.15-4.59]0.830 1.56 [0.49-4.90]0.451 Employed (self) 0.79 [0.38-1.65]0.537 0.39[0.14-1.15]0.088 1.41 [0.69-2.90]0.399 Employed (non-self) 1.56 [0.94-2.59]0.087 0.94 [0.46-1.91]0.861 0.83 [0.50-1.36]0.468 Asset Index High ref ref ref Low 1.96 [1.26-3.04]0.003 1.27 [0.69-2.33]0.438 0.68 [0.44-1.04]0.077 Middle 1.29 [0.74-2.26]0.361 0.59 [0.28-1.23]0.165 0.92 [0.53-1.59]0.760 Valid NHIS Yes ref ref No 1.04[0.55-1.97]0.901 0.94 [0.49-1.78]0.846 Civil Status Married ref ref ref ref Divorced/Separated 1.13 [0.59-2.15]0.710 1.28 [0.57-2.84]0.546 0.63 [0.33-1.19]0.157 0.45 [0.20-1.02]0.056 University of Ghana http://ugspace.ug.edu.gh 41 Outcomes Social-demographic factors Psychological Distress Quality of Life Single 2.59 [1.57-4.28]0.000 3.93 [1.64-9.45]0.002 0.61 [0.37-1.00]0.049 0.61 [0.28-1.29]0.197 Widow/Widower 1.25 [0.71-2.21]0.432 0.59 [0.28-1.29]0.190 0.73 [0.42-1.28]0.275 0.89 [ 0.47-1.68]0.725 Number of Children alive 1-2 ref ref ref ref None 2.07 [1.03-4.15]0.040 1.45[0.56-4.59]0.531 0.93[0.47-1.86]0.843 1.02 [0.41-2.55]0.964 3-5 1.74 [1.09-2.77]0.019 2.57[1.38-4.80]0.003 0.63 [0.40-1.00]0.049 0.37[0.22-0.63]0.000 6+ 1.78 [0.96-3.32]0.068 3.00 [1.30-6.94]0.010 0.66 [0.35-1.21]0.182 0.44 [0.22-0.89]0.022 Religion Christianity ref ref Islam 1.14 [0.66-1.97]0.641 1.03 [0.59-1.79]0.899 Place of Residence Greater Accra ref ref Central Region 1.39 [0.79-2.48]0.252 0.74 [0.41-1.31]0.297 Eastern Region 1.94 [0.98-3.81]0.055 0.68 [0.34-1.33]0.265 Other Regions 1.19 [0.66-2.17]0.550 1.62 [0.88-2.99]0.124 4.4.2 Cancer related characteristics associated with high psychological distress and good quality of life amongst study participants. Table 4.6 presents the crude and adjusted logistic regression for cancer related predictor variables associated with distress and QoL. While symptoms burden, comorbidities and ever received psychological services were found to be statistically significant for crude and adjusted estimations, duration since diagnosis was the significant predictor variable exclusive to crude estimation under the domain of distress (p-value <0.05). In reference to respondents whose duration since diagnosis was two or more years, the odds of experiencing high distress were 1.81 times more in persons diagnosed within less than a year of this study [OR(95%CI=1.81(1.15-2.85)]. Compared to respondents with no symptoms, the odds of high distress were 3.44 times higher in persons with six or more symptoms [OR(95%CI=3.44(2.05-5.79)]. A similar trend was observed for adjusted estimates with higher odds 5.10 [aOR(95%CI=5.10(2.59-10.01)]. While the odds of experiencing high distress was 1.62 times (95%CI=1.12-2.35) more in persons with no comorbidities, adjusted estimate showed 1.68 times higher odds compared to respondents with comorbidities. Relative to University of Ghana http://ugspace.ug.edu.gh 42 respondents who received psychological services prior to treatment, the odds of experiencing high distress were 2.25 times and 3.12 times higher for persons who did not receive such services in both crude and adjusted estimates. For QoL domain, symptoms burden, comorbid conditions and receiving psychological services were significantly associated. The odds of good QoL were significantly reduced by 69% in persons with six or more symptoms compared with those reporting no symptoms [OR(95%CI=0.31 (0.19- 0.51)]. Similar trend was observed for the adjusted model with the odds being reduced by 73% in comparison to respondents with no symptoms [aOR (95%CI=0.27(0.16-0.48)]. In relation to respondents having comorbid conditions, those who reported having no comorbidities had 43% reduced odds of good QoL [OR(95%CI=0.57(0.39-0.83)]. A similar picture was revealed in the adjusted model with the odds of good QoL being reduced by 51% in respondents with no coexisting medical conditions [aOR(9