52 item - I came in contact with, but most especially Drs. Lynda Wilson and Jacqueline Moss, for giving ...... 19 The Fi
OCCUPATIONAL STRESS, JOB SATISFACTION, AND JOB PERFORMANCE AMONG HOSPITAL NURSES IN KAMPALA UGANDA
by ROSE CHALO NABIRYE
KATHLEEN C. BROWN, COMMITTEE CHAIR CONNIE L. KOHLER ELIZABETH H. MAPLES NA-JIN PARK ERICA R. PRYOR
A DISSERTATION Submitted to the graduate faculty of the University of Alabama at Birmingham, in partial fulfillment of the requirements for the degree of Doctor of Philosophy BIRMINGHAM, ALABAMA 2010
Copyright by Rose Chalo Nabirye 2010
ii
OCCUPATIONAL STRESS, JOB SATISFACTION, AND JOB PERFORMANCE AMONG HOSPITAL NURSES IN KAMPALA, UGANDA
ROSE CHALO NABIRYE SCHOOL OF NURSING ABSTRACT Occupational stress, a common occurrence among various professions worldwide, is regarded as a major occupational health problem for healthcare professionals especially nurses. Occupational stress has been reported to affect job satisfaction and job performance among nurses, thus compromising nursing care and placing patients’ lives at risk. Stress is a complex phenomenon resulting from the interaction between individuals and the environment. Therefore, significant differences in occupational stress, job satisfaction and job performance among nurses may exist due to different work settings. The aims of the study were to: 1) examine the relationships between occupational stress, job satisfaction and job performance among hospital nurses in Kampala City, Uganda; 2) establish whether personal background characteristics affect the relationships between occupational stress, job satisfaction and job performance; and 3) examine whether there is a difference in levels of occupational stress, job satisfaction and job performance by type of hospital. A non-experimental correlational design was used in the study. A total of 333 nurses from four hospitals completed the Nurse Stress Index, the Job Satisfaction Survey, and the Six-Dimensional Scale of Nurse Performance scales. Study findings demonstrated that there were significant differences in levels of occupational stress, job satisfaction and job performance between the public and private not-for- profit
iii
hospitals. Nurses in the public hospital reported higher levels of occupational stress and lower levels of job satisfaction and performance. There were significant negative relationships between occupational stress and job performance and between occupational stress and job satisfaction. Nursing experience, type of hospital, and number of children had a statistically significant relationship with occupational stress, job satisfaction and job performance. Type of hospital (public versus private), ward (obstetrics/gynecology versus other ward types), and job satisfaction were significant predictors of self-rated quality of job performance. Job satisfaction was shown to mediate the relationship between occupational stress and job performance. Large scale studies were recommended to identify sources of occupational stress and factors that enhance job satisfaction among hospital nurses in Uganda. Future research is needed to examine best practices for human resource managers to improve nurse motivation, job satisfaction and nurse performance in hospitals.
Key words: occupational stress, job satisfaction, job performance, personal characteristics, work characteristics, public hospital, private not-for-profit hospitals, best practices for human resource management.
iv
DEDICATION
In loving memory of my late father Mr. Nathan Gusongoirye Waako who always encouraged me to study hard and sacrificed the little resources he had in order to provide for my education.
v
ACKNOWLEDGEMENTS First and foremost, I am most grateful to the Almighty God for giving me this opportunity to advance in my studies. I am sincerely grateful for the sponsorship/scholarship grants awarded to me for doctoral study, including the UABICER Training Grant through Dr. Eric Chamot for the initial grant which enabled me to enroll in the program. I convey my heartfelt gratitude to all the staff in the Fogarty office, especially Heather White and Alexis MacLean for the untiring patience and support accorded to me while at UAB. My gratitude to Sigma Theta Tau, the Gladys Farmer Colvin Memorial Scholarship, and to the Makerere University Staff Development and Training Division for additional grants which enabled me complete the program, and to Makerere University School of Graduate Studies for funding the study. Sincere gratitude also goes to the Good Health Program of Birmingham and Deep South Center for Occupational Health and Safety, UAB and Auburn University, for providing the pens which were distributed to the participants during data collection. I am greatly indebted and sincerely grateful to Dr. Kathleen C. Brown, my supervisor and Dissertation Committee Chair for the guidance, support and continuous advice not only on academic matters but on social issues as well. I sincerely appreciated her patience and commitment to see me through the PhD program. My sincere gratitude also goes to Dr. Erica Pryor, for the continuous advice, support and expertise in statistics and for always being there for me whenever I needed her wise counsel. I am grateful to dissertation committee members Dr. Elizabeth Maples, Dr. Connie Kohler and Dr. Na-Jin Park for their support,
vi
encouragement and guidance throughout the doctoral program and dissertation writing. I thank Dr. Isaac Okullo, my Ugandan supervisor, for his advice and encouragement throughout the PhD program. In the same way, I am thankful to Professor Wabwire-Mangen Fred of the School of Public Health, Makerere University for the professional advice and guidance, despite his busy schedule. My sincere gratitude also goes to Dr. D. K. Sekimpi for the support and encouragement throughout the doctoral program. I am so grateful to all UAB School of Nursing Faculty and other staff, whoever I came in contact with, but most especially Drs. Lynda Wilson and Jacqueline Moss, for giving me encouragement and advice which kept me “hanging in there.” To the PhD nursing students in academic years 2006/2007-2008/2009, I say thank you all for the support you gave me. Tracy, I thank you so much for providing the sisterly support academically and socially throughout my stay in Birmingham. I thank Dr. Mantana Damrongsak Brown for “showing me the ropes” and always being there for me. I am sincerely grateful to my friends Pat Yeilding and family, Sandy and Bill Myers, and family for always being there for me. The words of encouragement, spiritual and other forms of support, the love you showed me and prayers surely kept me going. I sincerely thank the Ugandan and ‘ZamUga’ communities in Birmingham and beyond, most especially Dr. Kabagambe Edmond and family, Sarah, Vincy, Jacqueline Makaaru, Jacqueline Mulundika and Margaret for the family atmosphere which made me “feel at home away from home” throughout my stay in Birmingham. I appreciate the Principal of Makerere University College of Health Sciences Professor Nelson Sewankambo for the continuous support and advice especially on funding opportunities. My sincere gratitude goes also to the Department of Nursing
vii
faculty and staff for the support and encouragement not only to me but to my family during my absence. I am grateful to Associate Professor Sara Groves from John Hopkins University, for the support and advice especially for writing the literature review and dissertation report. I thank Mr. Yovani Lubaale and Dr. Nazarius Mbona of Makerere University Institute of Statistics and School of Public Health respectively for the assistance in data analyses. Many thanks go to the Assistant Commissioner Nursing Services of Mulago Hospital, the Senior Nursing Officers and the Medical Directors of Mengo, Kibuli and Rubaga Hospitals for the support during preparation and actual data collection. To the dear nurses who participated in the study, I am so grateful for your precious time to complete the questionnaires. Many thanks also go to the research assistants including Godfrey, Scovia, Allen and Richard for the support, interest, and diligence they accorded the study. I thank my extended family members, my mum Gertrude Gusongoirye, and brothers David, Robert, Charles, Peter, Dan, Edward and sister in-law Mrs. Joyce Nankinga Gusongoirye. I thank you for your endless love, prayers and moral support. And last but not least, my beloved children, Doreen, Ellen, Pauline, Solomon and Derrick, thank you for being such wonderful children. In spite of missing motherly love and care, you gave me unconditional love, support and encouragement that actually motivated me to continue and complete my studies.
viii
TABLE OF CONTENTS Page
COPYRIGHT................................................................................................................. ii ABSTRACT..................................................................................................................iii DEDICATION ............................................................................................................... v ACKNOWLEDGEMENTS .......................................................................................... vi LIST OF TABLES ....................................................................................................... xii LIST OF FIGURES .................................................................................................... xiv LIST OF ABBREVIATIONS ...................................................................................... xv CHAPTER 1 INTRODUCTION Health Care System in Uganda ............................................................................ 3 Statement of the Problem..................................................................................... 4 Significance of the Study ..................................................................................... 6 Specific Aims of the Study .................................................................................. 7 Research Questions .............................................................................................. 7 Operational Definitions........................................................................................ 8 Conceptual Framework ........................................................................................ 9 Assumptions for the Study ................................................................................. 10 2
LITERATURE REVIEW ..................................................................................... 11 Occupational Stress............................................................................................ 11 Sources of Occupational Stress among Nurses......................................... 12 Workload .................................................................................................. 13 Organizational Pressure ............................................................................ 15 Interpersonal Relationships/Intrinsic Nature of the Work ........................ 16 Professionalism ......................................................................................... 17 Effects of Occupational Stress ........................................................................... 18 Job Satisfaction among Nurses .......................................................................... 22 Job Performance among Nurses ........................................................................ 29 Summary of Literature ....................................................................................... 32
ix
3
METHODOLOGY ............................................................................................... 33 Study Design ...................................................................................................... 33 Ethical Considerations ....................................................................................... 33 Setting ................................................................................................................ 34 Study Sample ..................................................................................................... 35 Inclusion Criteria ...................................................................................... 35 Exclusion Criteria ..................................................................................... 36 Sample Size........................................................................................................ 36 Instruments......................................................................................................... 37 The Nurse Stress Index (NSI) ................................................................... 37 The Job Satisfaction Survey (JSS) ............................................................ 38 The Six Dimension Scale of Nursing Performance (6-DSNP) ................. 39 Pilot-testing of Instruments ................................................................................ 39 Data Collection Procedures ............................................................................... 40 Data Safety and Integrity ................................................................................... 41 Data Analysis ..................................................................................................... 42 Limitations of the Study .................................................................................... 43
4
FINDINGS ............................................................................................................ 44 Demographic Characteristics ............................................................................. 44 Work Characteristics.......................................................................................... 45 Instrument Reliability ........................................................................................ 47 Descriptive Statistics for Instrument Sub-scales ............................................... 50 Nurse Stress Index .................................................................................... 50 Job Satisfaction Survey ............................................................................. 51 Nurse Performance Scale .......................................................................... 52 Occupational Stress and Demographic Characteristics ..................................... 54 Occupational Stress and Work Characteristics .................................................. 55 Job Satisfaction and Demographic Characteristics ............................................ 57 Job Satisfaction and Work Characteristics ........................................................ 58 Job Performance and Demographic Characteristics .......................................... 60 Job Performance and Work Characteristics ....................................................... 61 Findings Related to Research Questions ........................................................... 63 Research Question 1 ................................................................................. 63 Research Question 2 ................................................................................. 64 Research Question 3 ................................................................................. 65 Research Question 4 ................................................................................. 67 Research Question 5 ................................................................................. 69
5
DISCUSSION, CONCLUSIONS, IMPLICATIONS, AND RECOMMENDATIONS ............................................................................. 73 Discussion .......................................................................................................... 73 Occupational Stress................................................................................... 73 Job Satisfaction ......................................................................................... 75 Occupational Stress and Job Performance ................................................ 77 Characteristics, Stress, Job Satisfaction and Job Performance ................. 78 Mediating Role of Job Satisfaction on Stress and Job Performance ........ 79 Stress, Job Satisfaction and Job Performance by Hospital Type .............. 80 The Conceptual Framework............................................................................... 81 Conclusions ........................................................................................................ 82 x
Implications ....................................................................................................... 83 Implications for Nursing Education .......................................................... 83 Implications for Nursing Practice ............................................................. 83 Recommendations .............................................................................................. 84 REFERENCES ............................................................................................................ 86 APPENDICES ............................................................................................................. 92 Appendix A: Institutional Review Board for Human Use Approval................. 93 Appendix B: Permission to Use Research Instruments ................................... 104 Appendix C: Instruments ................................................................................. 109 Appendix D: Instrument Sub-scales and Number of Items ............................. 120
xi
LIST OF TABLES Table
Page
1 Socio-demographic Characteristics of the Sample ................................................. 45 2 Work Characteristics of the Sample ....................................................................... 46 3 Number of Items and Cronbach’s Alphas for the NSI Sub-scales ......................... 47 4 Number of Items and Cronbach’s Alphas for the JSS Sub-scales.......................... 48 5 Number of Items and Cronbach’s Alphas for the 6-DSNP Sub-scales .................. 49 6 Range of Possible Scores, Mean Scores and Standard Deviations for NSI ........... 50 7 Range of Possible Scores, Mean Scores and Standard Deviations for JSS ............ 51 8 Range of Possible Scores, Mean Scores and Standard Deviations for 6-DSNP .... 52 9 Descriptive Analyses for Job Satisfaction (JSS) .................................................... 53 10 Descriptive Analyses for Nurse Performance Scale.............................................. 54 11 Mean Scores for Occupational Stress by Demographic Characteristics ............... 55 12 Mean Scores for Occupational Stress by Work Characteristics ............................ 57 13 Mean Scores for Job Satisfaction by Demographic Characteristics ...................... 58 14 Mean Scores for Job Satisfaction by Work Characteristics .................................. 60 15 Mean Scores for Job Performance by Demographic Characteristics .................... 61 16 Mean Scores for Job Perofrmance by Work Characteristics ................................. 62 17 Correlations for Job Performance, Job Satisfaction with Occupational Stress ..... 64 18 Effect of Personal Background and Work Characteristics on the Relationships of Occupational Stress, Job Satisfaction and Job Performance Quality ............... 66 19 The Final Predictive Model for Self-Rated Job Performance ............................... 67 20 The Mediating Role of Job Satisfaction between Occupational Stress and Job Performance.......................................................................................................... 69 21 Means for Occupational Stress, Job Satisfaction and Job Performance for the Different Hospitals ............................................................................................... 70 xii
22 Influence of Type of Hospital on Job Performance, Job Satisfaction and Occupational Stress .............................................................................................. 71 23 Post Hoc Results of Differences in Means in Occupational Stress, Job Satisfaction and Job Performance by Hospital ..................................................... 72
xiii
LIST OF FIGURES Figure
Page
1 Diagram of Theoretical/Conceptual Model .............................................................. 10
xiv
LIST OF ABBREVIATIONS 6-DSNP
Six Dimensional Scale of Nurse Performance
AIDS
Acquired Immuno-Deficiency Syndrome
BSN
Bachelor of Science in Nursing
EM
Enrolled Midwife
EN
Enrolled Nurse
ENT
Ear, Nose and Throat
GDP
Gross Domestic Product
HIV
Human Immunodeficiency Virus
ICU
intensive care unit
ILO
International Labor Organization
IPR
interpersonal relations
JSS
Job Satisfaction Survey
M
Mean
MoH
Ministry of Health
NGO
Non-Governmental Organization
NHS
National Health Service
NSI
Nurse Stress Index
OSH
Occupational Safety and Health
RM
Registered Midwife
RMN
Registered Mental-health Nurse
RN
Registered Nurse
RN/M
Registered Nurse/Midwife
xv
RPN
Registered Pediatric Nurse
SD
standard deviation
SNO
Senior Nursing Officer
SPSS
Statistical Package for Social Sciences
UBOS
Uganda Bureau of Statistics
UK
United Kingdom
US
United States
WHO
World Health Organization
xvi
1
CHAPTER 1 INTRODUCTION Occupational stress can be defined as the harmful physical and emotional responses that occur when the requirements of the job do not match the resources, capabilities and needs of the worker (Alves, 2005; Bianchi, 2004; Lindholm, 2006; Nakasis & Ouzouni, 2008). The International Labor Organization (ILO) asserts that all countries, professions and all categories of workers, families and societies are affected by occupational stress (Ogon, 2001). According to Alves (2005), 40% of all American workers perceive their jobs as being extremely stressful. Similar findings are noted in the United Kingdom, where occupational stress is estimated to be the largest occupational health problem (Edwards & Burnard, 2003). Additionally, research has demonstrated that as workload and work-associated stress increase, turnover rates of workers are also noted to increase. Thus, occupational stress results in considerable costs to organizations in terms of absenteeism, loss of productivity, and health care resources (AbuAlRub, 2004; Cottrell, 2001; Gueritault-Chalvin, Kalichman, Demi, & Peterson, 2000; Nakasis & Ouzouni, 2008). Lack of productivity due to occupational stress and its related effects, including staff conflicts, recruitment and retention problems, burnout, absenteeism, litigation and rapid turn-over, and lack of job satisfaction, has been reported to cause significant monetary costs to the National Health Service [NHS] Trusts in the UK (Cottrell, 2001; Mackay, Cousins, Kelly, Lee, & McCaig, 2004). Alves (2005) reported that organizations spend as much as $75 billion a year on stress-related
2 outcomes including physical injuries at work and absenteeism; while the World Health Organization (WHO, 2002) estimates the cost of stress and stress-related problems to organizations to be in excess of $150 billion annually. Job stressors and low job control have also been shown to be risk factors for patient safety and to lead to poor job performance including reduced quality of nursing care (Sveinsdottir, Biering, & Ramel, 2006). Occupational stress has also been associated with chronic health problems like cardiovascular disease, musculoskeletal disorders, physical injuries and cancers (Alves, 2005; Bradly & Cartwright, 2002). Mental illness and serious health compromising behaviors such as increased risk for suicide, substance abuse (such as smoking and alcohol consumption), poor diet and lack of exercise are also associated with occupational stress (Adeb-Saeedi, 2002; Oginska-Bulik, 2006). The common sources and effects of occupational stress identified in the literature are from the perspective of American and European workers. Since occupational stress has been reported to affect all societies and professions (Ogon, 2001), it can be assumed that nurses working in Ugandan hospitals experience occupational stress. Research has further demonstrated that the sources of occupational stress, its levels and effects vary greatly depending on local forces such as nature of work, work setting and cultural orientation. Thus, significant differences in occupational stress among nurses in different countries may exist due to different work settings and levels of social support (Evans, 2002). At present, there are scant data about sources of occupational stress and its relationship with job satisfaction and job performance in African countries or in Uganda in particular. There is, therefore, a need to understand the predictors of occupational stress and the levels and inter-
3 relationships of occupational stress, job satisfaction and job performance among nurses working in Ugandan hospitals.
Health Care System in Uganda Uganda, which is listed among the developing countries, is located in East Africa. The country is bordered by Kenya in the east, Sudan in the north, Democratic Republic of Congo in the west, Tanzania in the south and Rwanda in the southwest (UBOS, 2005). Uganda is characterized by poor health and developmental statistics. For example, life expectancy (in years) at birth for males and females is 48 and 51 respectively and the infant mortality rate is 136 per 1000 live births; with 7.6 % of gross domestic product (GDP) expenditure on health (WHO, 2005). The health infrastructure in Uganda is composed of hospitals, health centers and aid posts which may belong to the government, non government organizations (NGOs) or to individuals (private). The hospitals are categorized as national referral, regional referral or district/rural hospitals. The health centers are graded as health center IV, III, II according to level of service, which is at the county, sub-county, and parish levels respectively. There are a number of training schools and universities which offer certificate/diploma and degree courses for healthcare professionals. Two-thirds of these belong to the government and a third to NGOs and individuals (http://www.health.go.ug/health_units.htm). The training schools for nurses are distributed in various regions/districts throughout the country. Nursing education includes various programs leading to different cadres of nurses depending on the educational level of entry to the program and the length of the program. Enrolled Nurses or Midwives (EN or EM) training lasts two and a half years and eligible students must have acquired a Uganda Certificate of Education or
4 its equivalent. This level of training has been phased out, but the cadre of nurses still exists in the hospitals. The registered level training is a three year program and eligible students must have a Uganda Advanced Level Certificate of Education or its equivalent. Graduates at this level can be general nurses (RN), midwives (RM), pediatric nurses (RPN) or mental health nurses (RMN). In the past, registered level nurses could complete further education in any other discipline of interest to get an equivalent certificate diploma after practicing for a minimum of two years. For example, an RN could apply to complete RM preparation and vice versa and train for one and a half years, creating another cadre of nurses, the Registered Nurse/Midwives (RN/M) or double trained nurses. This type of further education was referred to as horizontal training while that of an EN or EM midwife to registered level was termed vertical training. Double training has been phased out and nurses are encouraged to pursue further education in the same line, which is, general nursing, midwifery, pediatrics, etc. The Public Health Nurse (PHN) program is a two year program only offered to nurses who have done at least two disciplines at the registered level. The BSN is a four year university level program plus an additional one year internship. The BSN program prepares nurses to acquire competencies in nursing, midwifery, primary health care (community nursing), and research.
Statement of the Problem A human resource crisis exists in resource-constrained countries like Uganda due to the macroeconomic and governing factors. A shortage of nursing staff has been reported with a ratio of one nurse (nurses and midwives combined) to a population of 3,065. It has also been reported that despite employing 30,000 health
5 care workers in 2004, an extra 5,000 qualified staff were needed to address the serious staff shortage (Dieleman, et al., 2007). The situation has been compounded by the HIV/AIDS pandemic as the high number of HIV/AIDS patients leads to excessive workload and increased tasks. There are continuous complaints by Ugandan nurses about work overload, and the demand for nurses continues to grow as many drop out of work with very little intervention seen. It has been reported that there is lack of enough space in the hospitals and they are overcrowded with very sick patients. The situation is worsened by lack of facilities and shortage of nurses, which is likely to cause stress to the nurses (Ojoatre, 2008). For example, according to the government newspaper (New Vision), one of the senior staff in Mulago Hospital reported that there were only 8-12 staff members including nurses, midwives and doctors for five wards in the department of Obstetrics and Gynecology at any one time. The staff manages the patient load which is three times the load that is expected for 24 staff members on one ward. At the First Global Forum on Human Resources for Health that took place in Uganda in 2008 it was observed that there was an imbalance between the number of nurses trained in the country and those who register to practice. It was speculated that nurses have migrated to other countries, have joined other fields, or sit at home due to the poor work conditions. Nurses also complain of working all day long despite the high numbers of very seriously ill patients who require more attention. They have further complained that their work is very stressful citing the very high nurse-patient ratio which is reported to be 1:1000. The nurse-patient ratio is reported to be above the 1:2 or 1:5 recommended by the World Health Organization for fatal complications and common illnesses respectively (Natukunda, 2008).
6 Nurses have also observed that working when tired results in mistakes for which they are blamed; therefore, they would rather not go to work under the circumstances. This is in addition to the public outcry about the deteriorating nursing care in Ugandan hospitals. Since there are no established occupational health services due to limited resources and lack of occupational health professionals, there is a risk of continuous loss of nurses, either due to stress related diseases or attrition due to lack of job satisfaction (OSH WORLD, 2008; UBOS, 2005).
Significance of the Study Studies of potential sources and effects of occupational stress have been conducted among nurses in the United States and Europe. However, stress is a complex phenomenon which results from interaction between an individual and the environment in which the person exists. Thus, significant differences in occupational stress among nurses may exist due to different work settings and levels of social support (Evans, 2002). It was further asserted that occupational stress is a function of local forces, pressures and cultures that requires customized interventions (Muscroft & Hicks, 1998). Therefore, this study examined associations between occupational stress, job satisfaction and job performance among hospital nurses in Uganda. The results of this study may be used to guide policy makers and nurse managers to develop a stress prevention/management model specific to the Ugandan situation. Prevention and management of occupational stress among nurses will not only improve their health but may improve job satisfaction and nursing care, which will in turn reduce costs for the healthcare organizations as well as individuals.
7 Specific Aims of the Study The specific aims of the study were to: 1. Examine relationships between occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda. 2. Establish whether personal background characteristics influence occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda. 3. Examine whether there is a difference in levels of occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda by type of hospital; that is, government versus private not-for-profit (faithbased) hospitals.
Research Questions The research questions in this study were as follows: 1. Is there a relationship between occupational stress and job performance among hospital nurses in Kampala, Uganda? 2. Is there a relationship between occupational stress and job satisfaction among hospital nurses in Kampala, Uganda? 3. Do personal background characteristics affect the relationships among occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda? 4. Does job satisfaction mediate the relationship between occupational stress and job performance among hospital nurses in Kampala, Uganda? 5. Are there differences in levels of occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda by type of hospital?
8 Operational Definitions The following operational definitions were used in this study: Hospital nurse refers to any individual who qualified as a nurse or midwife at any level (degree, diploma, or certificate), is registered by the Uganda Nursing Council to practice nursing or midwifery, and is working in a hospital setting in Uganda. Personal background characteristics include the following demographic and work characteristics: age, gender, marital status, number of children, hospital, ward/ department, nursing education, years of nursing experience, responsibility, and hours worked on a typical day. These were measured by an investigator developed demographic questionnaire. Occupational stress refers to the harmful physical and emotional responses that occur when the requirements of the job do not match the resources, capabilities and needs of the worker (Alves, 2005). Occupational stress was measured by the Nurse Stress Index (NSI) developed by Harris (Harris, 1989). Job satisfaction refers to the level or degree to which employees like their jobs (Spector, 1997). Hospital nurses’ job satisfaction was measured by the Job Satisfaction Survey (JSS) developed by Spector in 1985 (Spector, 1997). Job performance refers to how effectively an individual carries out his/her roles and responsibilities related to his/her job (AbuAlRub, 2004). The Six Dimension Scale of Nursing Performance (Schwirian, 1978) was utilized to measure hospital nurses’ job performance or effectiveness in carrying out their roles and frequencies of responsibilities in relation to patient care.
9
Conceptual Framework The conceptual models/frameworks identified in the literature that have been used to guide the study of occupational stress and coping are Lazarus and Folkman’s cognitive theory of stress and coping (Lazarus & Folkman, 1984) and Karasek’s Demand-Control Model (Karasek, 1979). According to Lazarus and Folkman’s (1984) cognitive theory of stress and coping, stress is defined by the interaction between the individual and the environment. Demands from the environment exceeding the available resources result in stress or coping, depending on the individual’s appraisal of the environmental effects or stressors. The variables in the theory that guided this research are personal and workplace characteristics. Workplace and work characteristics act as environmental stressors while personal characteristics may facilitate the individual nurse’s ability to carry out the process of appraisal of the stressors. Occupational stress and coping will result accordingly, depending on the level of appraisal. The individual’s perception of how much control he/she has is a factor which leads to feelings of stress when the situation is perceived as uncontrollable or to feelings of positive coping if the situation is perceived as controllable. According to the Demand-Control Model (Karasek, 1979), there is interaction of objective work load demands in the environment and the employee’s decision latitude to meet the demands. Decision latitude is defined as the authority which the individual employee has to make job decisions and the opportunity to utilize and develop skills on the job (Karasek, 1979). Long term exposure to situations with increased demands but with low control leads to low productivity and health related problems. The assumption in the model is that psychological strain results from joint effects of work demands and the decision-making freedom available for the employee
10 facing the demands. In other words, jobs with high demands but with low control increase the risk of stress-related illness. The variables in this model which guided the study are job demands or job strain (workload pressures).
Work and Personal Factors Environmental Stressors (Workplace factors) - Type of hospital - Ward/unit - Responsibility - Hours worked on a typical day
Cognitive Factors
Behavioral Outcome
OCCUPATIONAL STRESS
JOB PERFORMANCE Personal Characteristics: - Age - Sex - Nursing Education - Nursing experience - Marital Status - Number of children
JJ JOB SATISFACTION
Figure 1 Diagram of Theoretical/Conceptual Model
Assumptions for the Study The study was conducted under the following assumptions: 1. Occupational stress can be measured and self reported by nurses. 2. Stressors occur in life and work environments and individuals react to these stressors. 3. Nurses work in stressful environments that each individual nurse appraises and reacts to differently.
11
CHAPTER 2 LITERATURE REVIEW This chapter provides a review of literature related to occupational stress, job satisfaction, and job performance among nurses of all specialties in or outside the hospital. The first section of the review is related to sources and effects of occupational stress. The second section includes a review of factors that affect job satisfaction and the third section reviews factors related to job performance. An abundance of research has been reported on occupational stress, job satisfaction and job performance among nurses in the United States (U.S.), Europe, and Asia, but very little in Africa. It should be noted that even the limited research in Africa was conducted only in South Africa and no research was reported on in east, west, north or central African nurses. This phenomenon was observed by Adejumo and Lekalakala-Mokgele (2009) in their study appraising African nursing scholarship where 68.3% of the publications in the past two decades were from South Africa, 5.2% from West Africa, and 3.3% from East Africa. These findings may be due to the fact that African nurses outside of South Africa have limited research skills and rarely publish. Occupational Stress Many studies have investigated occupational stress occurrence among various professions in the U.S, Europe and Asia. Researchers have examined effects of stress on employee health, job satisfaction, job performance, and coping strategies. As a
12 result of their studies, these authors, have also suggested management and prevention strategies (Bianchi, 2004; Bradley & Cartwright, 2002; Chen, Chen, Tsai, & Lo, 2007). Occupational stress is documented as a common occurrence in health professions throughout the world. The National Health Services (NHS) in the United Kingdom and in Australia reported that occupational stress occurred among health professionals at higher levels than in any other comparable profession (Adeb-Saeedi, 2002; Cottrell, 2001). This higher level of stress in health service has been attributed to the nature of the work of health professionals in which nurses, physicians and hospital administrators are involved in providing help to people experiencing life crises (Tyson & Pongruengphant, 2004). Nursing has been shown to be a strenuous profession, with nurses more exposed to stress-provoking factors than other healthcare workers. According to Evans (2002), a survey commissioned by the Sunday Times in 1997 reported that nursing was the sixth most stressful profession. This literature review will examine the sources of identified occupational stress and then the effects of stress.
Sources of Occupational Stress among Nurses Sources of stress for nurses can be divided into four areas: workload, organizational pressures, interpersonal interactions, and professionalism. In reality it is rare that only one source of stress is present. Sources of stress are frequently interrelated and synergistic effects are observed due to a variety of sources of stress. For example, interpersonal conflicts may be due to organizational and management issues. Research has demonstrated that sources of occupational stress among nurses vary between regions, countries, organizations, departments, nursing specialties and individuals. This has been attributed to the different health systems, their culture,
13 availability of resources, nature of work, different educational levels, age, employment contract, work experience and personality traits (Lee & Wang, 2002; Lindholm, 2006; Peterson & Wilson, 2002). Individual differences affect the perception of the stressful situations and the use of coping strategies. For example, in a study comparing Guatemalan and U.S. nurses’ attitudes towards nursing, Guatemalan nurses were discontent with the lack of resources to treat patients while the U.S. nurses were discontent with the work environment drawbacks (Coverston, Harmon, Keller, & Malner, 2004). Further, researchers found that younger public health nurses in the U.S., those with shorter length of current working experience, a higher level of education, and less pre-job or job continuing education perceived more occupational stress (Kirkcaldy & Martin, 2000; Lee & Wang, 2002). It can be argued that these less experienced and younger public health nurses may have lacked enough experience at their current job. This may lead to lack of confidence and or competence in their work roles, thus the higher perception of occupational stress. However, it is not clear why those with a higher level of education perceive more occupational stress, but it may be due to role ambiguity. In contrast, Kirkcaldy and Martin (2000), in their study of job stress and job satisfaction among hospital nurses in Northern Ireland, found that older nurses reported more stress while younger nurses experienced better psychological health. This could be a result of more social responsibility for the older nurses which could lead to work-home conflicts.
Workload Workload has been demonstrated to be one of the most frequent stressors (Callaghan, Shiu, & Wyatt, 2000; Khowaja, Merchant, & Hirani, 2005; Li & Lambert,
14 2008). In a study of 102 nurses in a Chinese intensive care unit, excessive workload was the most frequently cited source of workplace stress. This was a result of the nursing shortage with fewer nurses to care for more patients (Li & Lambert, 2008). Work load, shift work, overtime, and covering for absent colleagues were the most common identified stressors in other studies (Begat, Ellefsen, & Severinsson, 2005; Cottrell, 2001; Xianyu & Lambert, 2006). Lee and Wang (2002) investigated perceived occupational stress and related factors among public health nurses, and reported that personal responsibility and workloads were the major sources of occupational stress. Excessive work load was also included as a major contributor to stress among hospital based Brazilian nurses (Stacciarini & Troccoli, 2004). Heavy workload may be due to the physically arduous work of nursing jobs, as well as due to organizational pressures when there is a nursing shortage. Lack of confidence and competence in the nursing role can have a synergistic relationship with work load, creating high stress scores (Kirkcaldy & Martin, 2000). Differences in stress by work settings have been reported. For example, a study of job stress, coping and health perceptions of Hong Kong primary care nurses reported low-to-moderate frequencies of stress among primary care nurses compared to higher stress reported by those working in acute care settings (Lee & Wang, 2002). This is expected as nurses working in acute care settings deal with very sick patients, thus increased workload and emotional exhaustion. Further, Lindholm (2006) reported that nurse managers and clinical directors had a significant probability of a high level of work stress because they were exposed to high job demands. In addition, significantly greater job stress was reported among medical-surgical nurses than those working in home care (Salmond & Ropis, 2005).
15 Organizational Pressure Organizational pressure and management issues are common causes for stress (McGrath, Reid, & Boore, 2003; Tyson, Pongruengphant, & Aggarwal, 2002). Stress from a perceived lack of organizational support, lack of resources, lack of autonomy, lack of competence and confidence, lack of communication and guidance, and low salaries or absent reward systems are organizational and management issues. Lee and Wang (2002) reported personal responsibility, inadequate guidance and support, lack of consultation and communication, lack of materials or resources, inadequate manpower, and having to take risks to complete tasks as sources of institutional stress. Ongoing organizational pressure has been identified as another source of stress (Begat, Ellefsen, & Severinsson, 2005; Cottrell, 2001; Xianyu & Lambert, 2006). The work environment and institutional settings themselves have been associated with occupational stress. Sveinsdottir and colleagues observed that, in addition to stressful factors intrinsic to nursing, organizational and management attributes influenced work-related stress among nurses (Sveinsdottir, et al., 2006). Some of the organizational and management attributes identified include work environment and institutional setting. For example, in a study of occupational duties, it was reported that certified nurse anesthetists perceived that responsibilities related to patient care and anesthesia were inherently stressful (Roberts, 2005). Nurse anesthetists care for patients who are usually unconscious and who require critical nursing care and often contribute to work anxiety. Other significant stressors for the certified nurse anesthetist included lack of institutional support, lack of adequate surgical preparation, and a negative operating room environment. In a study of stress and coping among cardiovascular nurses in Brazil, Bianchi (2004) identified institutional work conditions as the major source of stress for nurses.
16 Makinen and colleagues similarly reported that occupational stress due to work social arrangements are partially determined by the organization of work (Makinen, Kivimaki, Elovainio, & Virtanen, 2003). Makinen and colleagues further reported that the patient-focused nursing modes reduced the interpersonal problems among the staff, thus decreasing work stress (2003). Evans (2002) identified six major nurse stressors, including leadership, organizational control, intrinsic job features, job image, reward systems, and human resource systems and recommended that organizations should develop policies that encourage smooth communication and socialization in the workplace.
Interpersonal Relationships/Intrinsic Nature of the Work Working with difficult patients, the nurses’ feelings about death and dying, interpersonal conflicts, managing the patients’ pain and the presence of the family also contribute to occupational stress (Adeb-Saeedi, 2002; McGrath, et al., 2003; Tyson, et al., 2002). The HIV epidemic and high mortality rates have contributed to stressful work conditions for nurses. In a study of occupational stress of nurses in South Africa, health risks posed by contact with HIV/AIDS patients, lack of recognition for the job they are doing, and insufficient staff were identified as the most common stressors for nurses (Rothmann, van der Colf, & Rothmann, 2006). These findings are consistent with literature about the effect of the HIV/AIDS pandemic on the health care workforce, with reports of increased emotional burden and stress among health workers due to anxiety and fears of occupational exposure (Dieleman, et al., 2007; WHO, 2005). Dieleman and colleagues found specifically in Uganda that 83% of the staff interviewed had increased fear of occupational exposure as a health worker, 36% had had a potential exposure injury in the past year, and the
17 only resource available for nurses was to wash the area under running water (Dieleman, et al., 2007). Eighty-six percent of the staff also reported that an increase in the amount of work increases the likelihood of injuries, contributes to a large emotional burden with burnout in caring for very sick patients who do not respond to therapy, and increases concern about being stigmatized if they do contract HIV/AIDS. The majority of the staff continued to report a significant fear of getting infected even though there was adequate protection available in the hospital. It was found that hospital administrators were haphazard in addressing this staff issue, with no written policies to prevent or mitigate the impact of HIV/AIDS, and that this affected working conditions and staff motivation (Dieleman et al, 2007).
Professionalism Professional issues have also been reported to lead to stress among nurses. For example, Evans (2002), in a Yorkshire, UK study exploring the district nurses’ perception of occupational stress, found that job image and reward systems were among the six major stress factors for the nurses. In another study of perceived occupational stress and related factors in public health nurses in Taipei City, Taiwan, Lee and Wang (2002) found that lack of recognition in the workplace was a significant stressor among nurses. Likewise, Stacciarini and Troccoli (2004) in their study of occupational stress, job satisfaction and state of health in Brazilian nurses, reported that lack of recognition, lack of status of the nursing profession, lack of autonomy, low salaries, lack of resources, and assignments outside the individual’s specialty were sources of stress for nurses. Ethical conflicts have also been identified as sources of job related stress and anxiety (Begat, et al., 2005). According to Begat and colleagues, ethical dilemmas
18 arise because of nurses’ values and desires to provide high-quality care. This is in agreement with McGrath, Reid and Boore’s (2003) findings that too little time to perform duties to one’s satisfaction and rationing of resources and services resulted in moderate to high stress. In another study of occupational responsibilities, perceived stressors and work relations of the certified nurse anesthetist, perceived occupational stress were related to patient care and anesthesia work in general (Roberts, 2004). These stressors included patient complications and unexpected patient outcomes such as death of a patient on the operating table. Lack of competence and confidence in the nursing role has also been identified as a stressor (Kirkcaldy & Martin, 2000). The nurse who lacks confidence and who is not competent in the role may be concerned about committing errors which may lead to punishment and or litigation.
Effects of Occupational Stress Occupational stress has been reported to result in a significant monetary cost for health care systems (Cottrell, 2001; Evans, 2002). This is due to lack of productivity as a result of staff conflicts, health care consumption, recruitment and retention problems, burnout, absenteeism, litigation, and rapid turnover. The World Health Organization (WHO, 2002) estimates the cost of stress and stress related problems to organizations to be in excess of $150 billion annually. According to the Health Enhancement Research Organization, a depressed employee is estimated to spend $3,189 annually on health care expenses as compared with $ 1,679 for a nondepressed employee in the UK (Cottrell, 2001). In addition, depressed workers’ accumulated short-term disability days resulted in 20 million more lost work days per year than non-depressed workers (Cottrell, 2001). Although litigation cases are not common in the developing countries like Uganda, the population is becoming
19 increasingly aware of their rights and starting to sue health institutions for health workers’ negligence of duty. Therefore, the cost of occupational stress is likely to increase in health care ministries in these countries not only in terms of medications and other supplies but also in litigation cases. It is also likely to increase individual nurses’ stress as they will be working with anxiety and fear of litigation in the event of errors as they execute their duties. Occupational stress negatively affects individuals’ health and wellbeing. Individual effort-reward imbalance has been associated with burnout, which results from prolonged intense stress. In a study of burnout among nurses in Germany, the nurses who experienced effort-reward imbalance reported higher levels on two of the three core dimensions of burnout (Bakker, Killmer, Siegrist, & Schaufeli, 2000). Bakker and colleagues found that the nurses who identified a negative imbalance between efforts spent on their job and the reward they felt from the job reported feeling more emotionally drained than those who did not. The feelings of personal accomplishment were lowest among nurses who had a mismatch between demands and rewards, and who had high intrinsic effort in their jobs. Emotional exhaustion and burnout have been recognized as occupational hazards for people-oriented professions such as nursing. Brown and colleagues examined demanding work schedules and mental health in nursing assistants working in nursing homes, and reported that working two or more double shifts per month was associated with an increased risk for all negative mental health indicators (Brown, Zijlstra, & Lyons, 2006). Furthermore, working 6-7 days per week was associated with depression and somatization. In a study of stress, coping and managerial support and work demand among nurses, consistent relationships between work stress and depression, anxiety and job satisfaction were identified (Bennett, Lowe, Matthews,
20 Dourali, & Tattersall, 2001) They suggested that lack of management support, having job overspill, making decisions under time pressure and lack of recognition by the organization were key predictors of negative effect. Chronic health problems such as cardiovascular disease, musculoskeletal disorders, physical injuries and cancers have also been associated with occupational stress (Alves, 2005). Mental illness and serious health compromising behaviors such as increased risk for suicide, substance abuse (such as smoking and alcohol consumption), poor diet, and lack of exercise weree also associated with occupational stress (Adeb-Saeedi, 2002; Oginska-Bulik, 2006). Occupational stress also contributes to many nurses leaving their jobs (Cottrell, 2001; Sveinsdottir, et al., 2006). The high turnover of nurses results in a shortage of nurses, which leads to work overload for the remaining nurses and becomes a vicious cycle. The high turnover of nurses is attributed to a lack of job satisfaction which is associated with occupational stress. For example, Flanagan and Flanagan (2002) in a study of job satisfaction and job stress reported that the NSI was the strongest explanatory variable accounting for 30.3% of job satisfaction. Tyson and colleagues, who also utilized the NSI to measure occupational stress in their study of coping with organizational stress among hospital nurses in Southern Ontario, reported a negative correlation between job stress and job satisfaction (Tyson, et al., 2002). The shortage of nurses has also been reported to affect nursing care negatively. In a study on hospital nurse staffing and patient mortality, nurse burnout and job satisfaction in Pennsylvania, it was found that surgical patients experienced a high risk-adjusted 30 day mortality and failure to rescue when the patient-to-nurse ratio was high (Aiken, Clerke, & Sloane, 2002). Patients were more at risk of dying in a 30-day period because the nurses could not rescue them when hospital units were
21 understaffed. In the same study, nurses were more likely to experience burnout and job dissatisfaction when the patient-to-nurse ratio was high. Lack of job satisfaction can lead to employees’ resentment which may be manifested in chronic absenteeism, lateness or reduced effort and increased error rate (Ackerman & Bezuidenhout, 2007). This is a manifestation of poor nursing care which places patients’ lives at risk. Sveinsdottir and others (2006) also reported that job stressors and low job control lead to poor job performance, reduced quality of nursing care, and concerns for patient safety. Because of the great shortage of nurses in Ugandan hospitals, it is a common phenomenon for nurses to work double shifts or to work for seven days or more without “off duty.” Nurses stand in for colleagues who fail to report for duty due to sick leave or other social problems. Due to poor remuneration in their primary job, nurses prolong their work schedules when they want to accumulate vacation days in order to work elsewhere to make additional money in a second job. According to Kyadondo and Whyte (2003), public sector reforms and poor remuneration have weakened workers’ positions as professionals and result in their seeking supplementary sources of income outside the health care system. Kyadondo and Whyte further observed that while other professionals in Uganda supplemented their salaries by engaging in agriculture, beer brewing or trade, health workers were found in small storefront clinics, drug shops and laboratories. This means that health workers end up overworked if they continue to work in the public sector as well as in private enterprises. Further research is needed to examine stress and related factors among Ugandan nurses.
22 Job Satisfaction among Nurses Job satisfaction is defined as the level or degree to which employees like their jobs (Spector, 1997). Numerous components of job satisfaction have been identified including including satisfaction with pay, potential for creativity, autonomy, task identity, satisfaction with organizational promotion policy and their individual promotions, satisfaction with co-workers, and available continuing education opportunities. Previous researchers have reported an inverse or negative relationship between perceived stress and job satisfaction, that is, as job satisfaction increases, stress decreases (Flanagan & Flanagan, 2002; Sveinsdottir, et al., 2006; Zangaro & Soeken, 2007). Organizational support is reported to increase nurses’ job satisfaction. The findings of Bradley and Cartwright’s study of social support, job stress, health and job satisfaction among nurses in the United Kingdom indicated that perceived organizational support was related to nurses’ health and job satisfaction (Bradley & Cartwright, 2002). Organizational support may include provision of adequate resources, good communication, training and development and adequate support supervision. In a study of staff dissatisfaction in the surgical theatre complex of a private hospital in South Africa, the results indicated that dissatisfaction was due to poor working conditions, lack of management support, unequal distribution of work, lack of resources, poor remuneration, an inflexible time system, and staff shortages (Ackerman & Bezuidenhout, 2007). Quality interprofessional collaboration between nurses and doctors and positive leadership has also been reported as important factors for nurse job satisfaction. A meta-analysis study of nurses’ job satisfaction by Zangaro and Soeken (2007) reported that nurse-physician collaboration and autonomy were strongly
23 correlated with job satisfaction and job stress among nurses. This meta-analysis, conducted in the U.S. included 31 studies published between 1991 and 2003. However, although the meta-analysis was conducted in the U.S., it is not clear whether all the reviewed research was conducted in the U.S. Therefore it is not possible to assess whether the observations could be generalized only to nurses in the U.S., or generalized more widely. In a study of factors influencing stress and job satisfaction for nurses working in psychiatric units in Greece, the findings indicated that job satisfaction was primarily influenced by the quality of clinical leadership and psychological stress (Nakasis & Ouzouni, 2008). In a study among Australian nurses, role discrepancy especially task delegation practice contributed to nurses’ intention to quit their jobs (Takase, Maude & Maniase, 2006). This report supports the above observation that clinical leadership influences job satisfaction. This may be due to poor communication and transparency which results in psychological stress and job dissatisfaction. Other researchers have reported job satisfaction among nurses in terms of intrinsic and extrinsic factors. Intrinsic factors, also known as motivators are those factors inherent to the nursing job itself, while extrinsic or hygienic factors refer to conditions of work and work environment. While some research has reported that nurses may be stressed due to the nature of their job (intrinsic factors), other studies indicate that nurses are satisfied with the intrinsic factors. For example, Lephoko and colleagues reported that nursing management and nursing staff in selected hospitals in Mpumalanga province, South Africa were content with the intrinsic factors of their jobs but were dissatisfied with the extrinsic factors of the organizational climate (Lephoko, Bezuidenhout, & Roos, 2006). On the other hand, (Salebi & Minnaar,
24 2007) in a study of nurses in a public hospital in South Africa, reported that more respondents (42%) experienced low satisfaction with motivational (intrinsic) aspects of their job as compared to only 22% who experienced low satisfaction with hygiene (extrinsic) aspects of the job. The intrinsic motivational components of this study included responsibility, opportunity for creativity and innovation, independence, and recognition while the extrinsic hygiene factors were relationships in the work place, supervisors’ decision making skills, supervision, working conditions, policies, job security and salaries. Extrinsic organizational climate in the Lephoko and colleagues (2006) study referred to management, physical environment, career development, performance management, motivation, empowerment and organizational alignment. The results of the above two studies conducted in the same country of South Africa are contradictory. This is an indication that like occupational stress, job satisfaction is influenced by many factors including differences in regions, organizations or hospitals. Furthermore, these researchers defined intrinsic and extrinsic factors very differently making study comparison difficult. A universally agreed upon definition of intrinsic and extrinsic factors is needed to better compare regions in nurse satisfaction for different regions and countries. In another study in selected hospitals in England, Lephalala and colleagues reported that the most important extrinsic factor that caused job dissatisfaction among nurses was their salary (Lephalala, Ehiers, & Oosthuizen, 2008). Salaries between private hospitals and the National Health Service varied considerably and salaries within one hospital were also not uniform. In the same study, nurses were satisfied with other extrinsic factors such as organization and administration policies, supervision and interpersonal relations. These findings are inconsistent with those of Lephoko and colleagues (2006) in South Africa in which nurses were dissatisfied with
25 extrinsic factors related to organizational climate. This result is most likely is a reflection of cultural and economic differences and the level of development of the health care systems. In the same study, Lephoko et al., (2006) reported that lack of promotion or advancement opportunities and lack of involvement in decision and policy making were the most important intrinsic factors influencing nurses’ job satisfaction among nurses in the private hospitals studied. This also contradicts other studies emphasizing that each hospital or health care environment has its own milieu that contributes to nurse satisfaction and dissatisfaction. Organizational commitment is another factor reported to have an impact on job satisfaction. In a study exploring nurses’ views and experience in mainland China, it was reported that organizational commitment had the strongest positive impact on job satisfaction (Lu, While, & Barriball, 2007). In the same study, organizational commitment was positively related to professional commitment. Although there is not much reported in the literature on spousal support, an exploratory study confirmed the hypotheses that there was a negative correlation between work-family conflict and job satisfaction and that spousal support was positively correlated with job satisfaction (Patel, Beekhan, Paruk, & Ramgoon, 2008). According to Patel and colleagues, these findings indicated that nurses who are more satisfied with their jobs were less likely to allow work to encroach on their family lives. This conclusion is in agreement with the available literature which reports that an employee who perceives himself or herself to have job control has higher job satisfaction (Chinweuba, 2007; Sveinsdottir, et al., 2006). Although this study did not specifically investigate the contribution of spousal support to job satisfaction among nurses, one of the sub-scales in the Nurse Stress Index contains items related to work/home conflict that may allude to spousal support.
26 Job satisfaction among nurses has also been studied in relation to working conditions, emotional climate, and social climate. In a study of job satisfaction of registered nurses in a community hospital in the Limpopo Province in South Africa, the majority of respondents were dissatisfied with working conditions and emotional climate of the hospital, but fairly satisfied with the social climate (Kekana, Rand, & Wya, 2007). Working conditions were defined by reflecting on the individual’s perception of working conditions as influenced by workload, salary, fringe benefits, availability of adequate resources, professional growth opportunities and the challenges of the job. Emotional working climate referred to the level of autonomy experienced by nurses, how they conceive themselves as nurses and their professional commitment. The social work climate referred to the personal interactions at work, group cohesiveness and team spirit. Under working conditions, 83% of the participants rated workload and the degree of fair remuneration as highly dissatisfying while 82% rated pressure under which they worked as the most highly dissatisfying under the emotional climate (Kekana, et al., 2007). These findings were consistent with previous reports that heavy workload is a source of occupational stress which leads to low job satisfaction. However, the factors which are related to job satisfaction may be influenced by the differences between study populations, the design and conditions under which the studies are conducted. For example, in a study of nurses in Southern Taiwan which examined the effects of job rotation and role stress on job satisfaction and organizational commitment, nurses reported that job rotation had an effect on job satisfaction (Ho, Chang, Shih, & Liang, 2009). This differs from the previous studies cited above. However, it is clear that frequent rotation on the job may not allow nurses to develop the required individual unit skills thus stress and job dissatisfaction result.
27 In a Taiwanese study of hospital nurses, the researchers found that work characteristics such as routinization had the greatest negative impact on job satisfaction followed by personality traits and job involvement (Chu, Hsu, Price, & Lee, 2003). In another study of working conditions that contribute to absenteeism among nurses in a provincial hospital in the Limpopo Province, South Africa, constraining working conditions such as inadequate group cohesion, inadequate delegation of autonomy, role ambiguity, ineffective routinization and excessive workload resulted in absenteeism in the work place (Nyathi & Jooste, 2008). As such, these findings are in agreement with Selebi and Minnar (2007) who report that routinization, role ambiguity and lack of delegation autonomy do not give opportunity for innovations and creativity by the employees, and may result in job dissatisfaction. It has been reported in the literature that there is a negative link between job satisfaction, intention to leave, and actual turnover (AbuAlRub & Al-Zaru, 2008). Work related stress has been found to increase turnover rate of workers due to less job satisfaction. Wilson and colleagues in their study of job satisfaction among a multigenerational nursing workforce also acknowledged that job satisfaction is a significant predictor for nurse retention (Wilson, Squires, Widger, Cranley, & Tourangeau, 2008). Cottrell (2001) reported that over 30,000 nurses in the UK left their profession in 1997 alone. Such a loss, coupled with the recruitment crisis, results in increasing stress on those who remain on the job. Job stress and job satisfaction have also been reported to be influenced by personal characteristics such as age and experience of nurses. For example, significant inverse correlations were reported between job satisfaction and age, and years of nursing experience with job stress (Ernst, Franco, Messmer, & Gonzalez, 2004). Other personal characteristics such as mental and physical health, marital status,
28 education level, rural/urban setting, and perceived HIV stigma were reported to have significant influence on job satisfaction. In addition, there are significant differences in job satisfaction scores among five countries in a study of HIV stigma and nurse job satisfaction in five African countries (Chirwa, et al., 2008). These results reinforce the fact that there are many factors which affect job satisfaction. In another study of the relationship between job stress and job satisfaction among nurse educators in Nigeria, educational qualification was reported to influence the relationship between job stress and job satisfaction (Chinweuba, 2007). This result was attributed to the fact that the nurse educators with a higher education had a higher chance of securing a desired and satisfying job. The education qualification has also been reported to affect the employees’ role perception, professional commitment and role conflict, thus affecting job satisfaction indirectly (Lu, et al., 2007). Chinweuba rightly observes that nurse educators with less qualifications have less chance of securing satisfying jobs, have more role conflicts role ambiguity and work overload or under load. This means that they have less control of their work days, poor promotion opportunities and low levels of salaries. The two reports highlight the importance of higher education not only to nurse educators but all nurses. It has been reported in the literature that employees who have job dissatisfaction react differently (Ackerman & Bezuidenhout, 2007). In their study of staff dissatisfaction in the theatre complex of a private hospital in South Africa, Ackerman and Bezuidenhout observed that while some employees may react by leaving the organization, others may actively and constructively attempt to improve the conditions by voicing their concerns. Further, others may wait passively for conditions to improve or worsen. Therefore, considering the nursing shortage, the need to understand job satisfaction warrants the attention of organizational leaders as
29 well as researchers. It is not known what proportion of the reported shortage of nurses in Uganda is due to lack of job satisfaction, occupational stress, or any other cause. It is therefore important to explore levels of job satisfaction among nurses in Uganda to improve the quality of nursing care in these hospitals. It is noted that studies of job satisfaction among nurses have utilized various instruments. The JSS developed by Spector (1997) was used to measure job satisfaction in the current study because its sub-scales and items include factors that may be related to the work environment for Ugandan nurses.
Job Performance among Nurses Occupational stress and low job control have been shown to be risk factors for patient safety and to lead to poor job performance (AbuAlRub, 2004). Using the SixDimensional Scale of Nurse Performance (6-DSNP) developed by Schwirian (1978), Taskase and colleagues found that the quality of job performance was reduced when job dissatisfaction was present (Takase, Maude, & Manias, 2005). A high level of occupational stress has been found to reduce nursing quality. A shortage of nursing staff due to turnover as a result of occupational stress was associated with increased patient mortality rates in an intensive-care unit (Sveinsdottir, et al., 2006). However, in a study on job stress, recognition, job performance and intention to stay at work among Jordanian hospital nurses, it was reported that recognition of nurses’ performance had a direct and buffering effect on job stress and the level of intention to stay at work (AbuAlRub & Al-Zaru, 2008). This is consistent with available literature which has reported that recognition leads to job satisfaction and nurses seeing no reason to leave their jobs (Cartledge, 2001). Higher occupational stress levels have also been significantly associated with poorer self-rated and supervisor-
30 rated job performance, more sick days, and more reported absences for mental health reasons. Supervisor support is another factor which is reported to affect nurses’ performance. In a study of primary nurses’ performance and the role of supportive management, it was found that performance increased where supervisor support was higher (Drach-Zahavy, 2004). Furthermore, nurses’ perception of the costs of seeking support had a negative impact on nurses’ performance. This means that supervisor support needs to be readily available to improve nurses’ job performance. In another study, Abu Al Rub (2004) reported that perceived social support from co-workers enhanced the level of reported job performance, and lowered the level of job stress among the nurses. In a longitudinal survey of nurses’ self-reported performance during an entry to practice program, participants reported significant increases in frequency and quality of nursing behaviors over time (Roud, Giddings, & Koziol-McLain, 2005). The researchers concluded that new graduate nurses can successfully integrate knowledge gained during training into clinical practice when provided with time and support. This is probably true for all employees because the longer one stays on the job, the more confident and competent one becomes in the skills required for the job. With confidence and competence in the job skills, performance is improved. Morale is another factor which affects nurses’ performance. Nurses’ morale can be boosted by creating a conducive environment characterized by support supervision, positive feedback, and good communication. Perceptions that they are valued, job satisfaction and organization commitment may lead to improved work place efficiency and output (Stapleton, et al., 2007). Career commitment has also been reported to have a significant positive relationship on job performance. According to
31 some researchers, career commitment is mostly attitudinal when employees become emotionally attached to the organizations and accept their goals and values (Mrayyan & Al-Faouri, 2008). With the acceptance of organizational goals and values, employees may remain in the organizations. This not only improves job performance but also reduces organizational costs due to high turnover. Many researchers have studied occupational stress, job satisfaction and job performance among nurses. However, the studies have been either in individual private or public hospitals and very few compared these variables across both public and private hospitals. In a five year follow-up study of stress among nurses in public and private hospitals in Thailand, nurses in public hospitals reported more stress than those in private hospitals. However, job satisfaction did increase over time, particularly in public hospitals (Tyson & Pongruengphant, 2004). The researchers attributed the increase in job satisfaction to maturity/age, improvement in monetary compensation, and organizational support. Religious beliefs have also been reported to affect performance and service delivery. In a study to evaluate the service delivery given by religious health care providers in Uganda, it was reported that religious not-for-profit facilities hire medical staff below the market wage but the workers provide better quality care than their government counterparts (Reinikka & Svensson, 2003). The researchers concluded that altruistic concerns of religious not-for-profit hospitals motivate the healthcare providers to provide quality care to the poor. This finding is similar to the findings of (Mrayyan & Al-Faouri, 2008) who observed that employees who are emotionally attached to the organizations and accept their goals and values are satisfied with their jobs.
32 There are indications that nursing care in hospitals in Uganda has deteriorated. This is reflected in various newspaper complaints about neglect of patients and poor nursing care in the Ugandan hospitals. It is therefore assumed that nurses’ job performance does not meet public and administrative expectations. No study has been conducted to assess hospital nurses’ job performance in Uganda. It is therefore important to conduct this study to document nurses’ perceptions of their job performance and the factors which are associated with it. This will guide policy makers and nurse managers in developing strategies for improving job performance in Ugandan hospitals.
Summary of Literature
Based on the review of the literature, occupational stress, job satisfaction, and job performance are the major factors associated with retention of nurses and quality of care. Additionally, occupational stress has been found to differ among professions and work settings. The majority of the research studies regarding sources of occupational stress and job satisfaction have been conducted on American and European nurses and their work settings. The conceptual model guiding this study indicated that cultural templates influence the appraisal of job demands, job satisfaction, and job performance (Lazarus & Folkman, 1984). Therefore, given the diversity of reported stressors for nurses, there is a need to identify the relationships among occupational stress, job satisfaction and job performance of nurses in the Ugandan hospitals.
33
CHAPTER 3 METHODOLOGY This chapter provides a description of the methods utilized in implementation of the study. The study design, data collection instruments, subjects, procedure for collection of data, and methods of data analysis are presented.
Study Design A correlational, cross-sectional study design was utilized to investigate the research questions. The relationships among occupational stress, job satisfaction, and job performance were explored. The effects of personal characteristics such as age, marital status, nursing education, nursing experience and type of unit on occupational stress, job satisfaction and job performance were also explored. Further, differences by hospital setting (government or private not-for-profit) in regard to perception of occupational stress, job satisfaction and job performance were also examined.
Ethical Considerations Protection of human subjects was evaluated by the Institutional Review Boards of the University of Alabama at Birmingham and Makerere University. Permission was sought from the hospitals’ administrators to conduct research in a large public hospital and three private not-for-profit faith-based hospitals, all situated in Kampala, Uganda. Consent was also sought from the nurses and nurses who volunteered to fill the questionnaires were assumed to have given consent. A form explaining the purpose of the study and the rights of the participants to withdraw from
34 the study anytime without penalty was attached to the questionnaires to ensure voluntary consent. Further, the questionnaires had no identifiers to ensure anonymity and confidentiality. Only aggregate data were reported, individual data were kept in strict confidence and used only for research purposes.
Setting The study was conducted on a sample of nurses working in Mulago National Referral and Teaching hospital, and three private not-for-profit (faith-based) hospitals situated in Kampala, Uganda. Mulago National Referral and Teaching Hospital is utilized for clinical and practical teaching/experience for Makerere University medical, nursing, pharmacy and other health professional students. The Directorate of Nursing in the hospital is headed by the Assistant Commissioner, Nursing Services who is assisted in her managerial duties by several Area Managers or Senior Nursing Officers (SNOs). Each Area Manager is responsible for an area which is composed of several units/wards. These areas include: Accident and Emergency, Medical, Surgical, Pediatrics, Obstetrics and Gynecology, Outpatient Clinics (Assessment Center, Private Wing), Community Health Services, Operating Theatres and Special Clinics (Ear, Nose and Throat [ENT] and Ophthalmology). Several cadres of nurses including graduate nurses (BSN), Registered Nurses (RN), Registered Midwives (RM), Registered Nurse/Midwives (RN/M), Public Health Nurses (PHN), Enrolled Nurses (EN) and Enrolled Midwives (EM) work in Ugandan hospitals. The nurses are allocated to the various units according to their availability and consideration of the expected workload on the unit. The hospital nursing staff includes 1000 nurses and the hospital has a bed capacity of 1,500
35 patients. Nursing staff on the units/wards are also assisted by nursing assistants/nursing aides and nursing students from the various nursing schools. The three non-governmental not-for-profit hospitals situated in Kampala also have different departments including Medical, Surgical, Pediatrics, Obstetrics and Gynecology, Outpatient Clinics, but no specialized clinics are available. The total number of nurses in each hospital varies between 150-200 nurses and includes Enrolled Nurses (EN), Enrolled Midwives (EM), Registered Nurses and Midwives (RN/M), Tutors, Public Health Nurses (PHN) and Diploma Trained nurses. These hospitals have bed capacity from 300 to 360 and have training schools for nurses and midwives at enrolled and registered levels.
Study Sample The target population in this study was all nurses working in the four selected hospitals namely; the national public referral and teaching hospital and three private not-for-profit faith-based hospitals at the time of the study.
Inclusion Criteria Subjects in this study had to fulfill the following conditions: (a) must have been a qualified nurse (BSN, RN, RM, RN/M, PHN, EN or EM), (b) a fulltime employee of the hospital included in the study for at least six months by the time of the study, (c) between 20 and 60 years of age, (d) willing to participate in the study, and (e) working in the general surgical, medical, pediatrics, or obstetrics and gynecology wards.
36 Exclusion Criteria The following categories of nurses were excluded from the study: (a) those nurses who were currently enrolled in advanced educational study, (b) nurses who were working on contractual terms (above the retirement age of 60 years), and (c) tutors and nurses working in Operating Theatres, specialized ward/units like intensive care units (ICU), Heart Institute or burn units. The nurses working on these units were assumed to be extremely busy and that other factors may influence nurses’ working conditions. For example, while functional or task-oriented nursing is the major mode for nursing care in the hospitals, some of the specialized wards/units have acquired different modes of nursing care. In addition, some of the specialized units operate under different projects whose working conditions are totally different from the general wards/units.
Sample Size The required sample size was calculated based on a power analysis for partial correlation and regression analyses (Cohen & Cohen, 1983). The conventional standard level of significance (.05) and power (.80) was utilized to calculate sample size (Polit & Beck, 2004). According to Polit and Beck (2004) nursing studies usually have modest effects; therefore a relatively modest effect size of .20 was utilized and the adequate sample size calculated for the study was 321. However, the questionnaires were distributed to a total of 400 nurses who met the inclusion criteria and volunteered to participate in the study in order to account for non-responses and incomplete questionnaires. A total of 333 nurses submitted completed questionnaires.
37 Instruments Three instruments were utilized to examine the relationships between occupational stress, job satisfaction and job performance. The instruments utilized included the Nurse Stress Index [NSI] (Harris, 1989), Job Satisfaction Survey [JSS] (Spector, 1997), and the Six Dimension Scale of Nursing Performance [6-DSNP] (Schwirian, 1978). The three instruments were part of a questionnaire with sections, each consisting of an instrument. The questionnaire also included a section to identify and assess personal characteristics such as age, gender, nursing education (registered, diploma, BScN, masters), nursing experience, and the department where the nurse works. This was because personal characteristics have been reported to have an influence on the perception of stress, job satisfaction and job performance (Chinweuba, 2007; Ernst, et al., 2004).
The Nurse Stress Index (NSI) The NSI (Harris, 1989) aims at identifying sources of stress among nurses in hospital and community settings. The 30 item instrument consists of six subscales which each include five items. The sub-scales include: workload pressures related to insufficient time (Managing Workload 1), items 1-5; workload pressures due to resources and conflicting priorities (Managing Workload 2), items 6-10; Organizational Support and Involvement, items 11-15; Dealing with Patients and Relatives, items 16-20; Home and Work Conflicts, items 21-25; and Confidence and Competence in Role items 26-30 (see appendix D). Respondents are asked to rate their potential stressors on a 5-point Likert scale ranging from 1= no pressure to 5= extreme pressure. The NSI is self-reported and the respondents read and circle the selected score from 1 = no pressure, 2 = very little pressure, 3 = moderate pressure, 4
38 = high pressure, and 5 = extreme pressure for each item. A total score can be computed ranging from 30-150 and means of subscales can be calculated to assess relative importance of sources of stress. The scores can also be derived from each subscale and compared directly to obtain information on perceived sources of stress (Harris, 1989). Acceptable levels of reliability and validity of the NSI were established with overall Cronbach’s alpha of 0.90 (Harris, 1989).
The Job Satisfaction Survey (JSS) The JSS aims at assessing the degree to which people like their jobs (Spector, 1997). The JSS is a self-report instrument which provides an overall job satisfaction score after assessing nine facets or sub-scales, namely; pay, promotion, supervision, fringe benefits, contingent rewards, operating conditions, coworkers, nature of work and communication. The respondents agree or disagree on a 6-point continuum for each item, with 1 representing much disagreement, 2 disagree moderately, 3 disagree slightly, 4 agree slightly, 5 agree moderately, and 6 agree very much. Each sub-scale has four items resulting in a total of 36 items; however, some items are worded positively while others are worded negatively (see appendix D). This implies therefore that agreement for a positively worded item and disagreement with a negatively worded item indicates job satisfaction and vice versa. Reverse scoring therefore is necessary for the negatively worded items making 1 represent much agreement and 6 represent very much disagreement. The numbered responses are summed after reversing the negative items to get the total satisfaction score of the 36 items ranging from 36-216. Reliability and validity of the instrument was established with an overall Cronbach’s alpha score of .91 (Spector, 1997).
39 The Six Dimension Scale of Nursing Performance (6-DSNP) The Six Dimension Scale of Nursing Performance (Schwirian, 1978) assesses nurses’ job performance. This is the person’s effectiveness in carrying out his/her roles and responsibilities in relation to patient care. The instrument is self-reported consisting of six sub-scales and 52 items. The sub-scales include: Leadership, Critical Care, Teaching/Collaboration, Planning/Evaluation, Interpersonal Relations/ Communication and Professional Development. Respondents are asked to rate the items as to how often (column A) and how well (column B) they perform the behavior/item to assess frequency and quality of performance respectively. The nurses rate the items on a 4-point scale with 1 = not expected in this job; 2 = never or seldom; 3 = occasionally and 4 = frequently for column A, while column B is rated with 1 = not very well; 2 = satisfactorily; 3 = well and 4 = very well. However, the items on the professional development sub-scale are assessed for quality only. Since the sub-scales are of different lengths ranging from 5 to 12 items, their scores are calculated according to the average of ratings on behaviors/items per sub-scale. The numbers of items in each sub-scale are as follows: Leadership = 5, Critical Care = 7, Teaching/Collaboration = 11, Planning/Evaluation = 7, Interpersonal Relations/ Communication = 12 and Professional Development = 10 (see appendix D). It is assumed that higher scores indicate better performance (Schwirian, 1978). Reliability and validity of the 6-DSNP was established and Cronbach’s alpha coefficients for each sub-scale ranged from .90 to .97 (Schwirian, 1978).
Pilot-testing of Instruments Since the data collection instruments were based on American and European cultures in previous studies, instruments for this study were pilot-tested to validate
40 their appropriateness to the Ugandan situation before actual data collection. The researcher and a small group (n=4) of nurses at the level of Senior Nursing Officers reviewed and assessed whether the questions were clear and appropriate to the Ugandan situation. It was agreed that all items were clear and understandable by the Ugandan nurses and the tools were adapted and adopted for use in the study.
Data Collection Procedures The study was conducted after getting approval from the Institutional Review Boards (IRBs) at the University of Alabama at Birmingham and Makerere University. Permission to conduct the study was also sought from the administrators of the four hospitals. Four meetings, one per hospital, were organized with the nurses through the Directors of the hospitals and the Directors of Nursing Services. The purpose of the study, the methods of data collection and time frame for the study were explained to the nurses at the meetings. It was explained to the nurses that they were free to withdraw from the study at any time without penalty. It was also made clear that there was no financial or any other form of gain from the nurses’ participation. The nurses were then invited to participate in the study and the questionnaires were distributed to those who volunteered to participate in the study. In order to maximize participation of nurses and response rate, the researcher recruited and trained four BSN prepared research assistants. The research assistants were responsible for distributing and collecting the completed questionnaires from the nurses in the various wards/units who did not attend the meetings but volunteered to participate in the study. A sealed wooden box was placed at each ward/unit in which nurses dropped the completed questionnaires. The research assistants collected the completed questionnaires from their assigned wards/units on a daily basis.
41 In order to observe confidentiality, the nurses were not required to sign a consent form. It was explained that any nurse who volunteered to complete the data collection instruments was assumed to have given informed consent. Each questionnaire had an information sheet attached to it explaining the purpose of the study, the time frame for the study and assurance that the information given was to be utilized for study purposes only and strict confidentiality was to be observed. The information sheet also included instructions to the participants not to write their names or any identifiers on the questionnaires. Further, the information sheet had instructions for the participants to place the completed questionnaires in the boxes provided on the wards/units.
Data Safety and Integrity The investigator developed and maintained a codebook for each item on the questionnaires and all questionnaires were assigned a serial number. Four research assistants were recruited and trained to distribute and collect the already completed questionnaires. The training included a review of the items in the questionnaires to ensure a common understanding of the questions and appropriate data collection techniques. These included clarifying instructions and responding to participants’ questions and ensuring confidentiality while collecting the completed questionnaires. The questionnaires were kept under lock and key in the principal investigator’s office. The principal investigator double-checked the questionnaires for completeness before data were entered in the computer programs for analysis.
42 Data Analysis Data analysis was conducted using Windows SPSS version 16.0. The data were entered in the Windows SPSS data base by two data entry clerks (double entry) to enhance the quality of data entry process and for quality control in the data entry process. The data were assessed for completeness, consistency, and missing values. A questionnaire was required to have 80% of the questions completed before it could be accepted to be entered in the computer program for analysis. No questionnaires were disqualified due to incompleteness. The few missing values of some questionnaires were imputed using the multiple imputation method. The internal consistency of the study instruments and instrument subscales was evaluated using Cronbach’s alpha. The study variables were analyzed using descriptive statistics appropriate for the scale of measurement. Bivariate relationships were evaluated with Pearson correlation coefficients for continuous variables and with analysis of variance (ANOVA) or independent t-tests to examine differences in group means for categorical variables. Multiple regression analysis was used to investigate relationships among occupational stress, job satisfaction, and job performance, controlling for personal and work characteristics. The following assumptions were evaluated for the multiple regression analyses: that the expected value of the dependent variable is a linear function of the independent variables (linearity), that the variance is the same for any fixed combination of independent variables (homoscedasticity), and that it follows a normal distribution for any fixed combinations of independent variables (normality) (Munro, 2001). The regression models were also assessed for potential problems with multicollinearity using variance inflation factor (VIF) values. The potential mediating effect of job satisfaction on the relationship between occupational stress and job performance was
43 assessed using the approach of Baron and Kenny (1986). The level of significance was set at alpha=.05 for all analyses.
Limitations of the Study 1. The study was conducted in hospitals situated in Kampala, Uganda where nurses’ work environment and organization of work may be different from other hospitals and health centers in the rural areas. Therefore, the results from this study may not be generalizable to all nurses in Ugandan hospitals. 2. The participants were volunteers; therefore the results may be biased. 3. Occupational stress measurement was based on self-report rather than by physiological biochemical analyses of blood or by physical and mental status assessments. 4. The instruments utilized in this study were based on American and European populations and may not have been culturally appropriate for the Ugandan nurses.
44
CHAPTER 4 FINDINGS This chapter presents the findings of the study. The first section presents the demographic and work characteristics including age, gender, marital status, number of children, hospital and ward/unit where participants work, nursing education, nursing experience, responsibility on the ward/unit, and hours worked on a typical day. The second section includes reliabilities of instruments used in the study. The third section presents descriptive analyses related to the study variables while section four includes the statistical analyses of data related to the study questions.
Demographic Characteristics A total of 400 eligible nurses attended the meetings and were invited to participate in the study. Although all eligible nurses agreed to participate in the study, a total of 333 nurses (response rate 83%) returned completed valid questionnaires that were included in the analyses. The age range was 20 to 60 years with a mean age of 36 years (SD = 9.1). A majority of the participants were female (95%), 61% were married while 25% had never married. More than a third (41%) had between 1-2 children and 29% had between 3-4 children. The average number of children per participant was 2.2 (SD = 1.8).
45 Table 1 Socio-demographic Characteristics of the Sample Characteristic Agea 20-29 30-39 40-49 50-60
91 135 71 36
27.33 40.54 21.32 10.81
Gender Female Male
317 16
95.20 4.80
Marital status Married Divorced/Separated Widow/Widower Never married
206 32 11 84
61.86 9.61 3.30 25.23
68 138 98 29
20.42 41.44 29.43 8.71
Number of childrenb 0 1-2 3-4 >5 a M = 36.02; SD = + 9.11 b M = 2.19; SD = + 1.84
Frequency
%
Work Characteristics Two-thirds of the participants (59%) were from the public hospital and the remaining participants were from the three private, faith-based hospitals (21%, 12%, and 8% respectively). The majority of participants (60%) were at the Registered level (RN, RM or RN/M), followed by the Enrolled level (36%), with only 4.5% at the Graduate level (BSN and above). Almost two-thirds of the participants were nurses with less than 14 years of nursing experience. Twenty percent of the participants reported 20 years or more of nursing experience. Many participants reported working longer hours than a standard shift on a typical day, with 43% working more than eight
46 hours. Two-thirds of the participants (63%) had no extra responsibility on the wards/units and 37% were ward/unit in-charges or deputies.
Table 2 Work Characteristics of the Sample Characteristic Hospital Private 1 Private 2 Private 3 Public
25 71 41 196
7.51 21.32 12.31 58.86
Nursing Education Enrolled Nurse (EN) Enrolled Midwife (EM) Registered Nurse (RN) Registered Midwife (RM) Double Trained (RN/M) BSN and above
77 42 100 49 50 15
23.12 12.61 30.03 14.71 15.02 4.50
Ward/Unit Medical Surgical Obstetrics/Gynecology Pediatrics
117 57 102 57
35.14 17.12 30.63 17.12
70 68 74 53 26 21 24
21.02 20.42 21.32 15.92 7.81 6.31 7.20
123 210
36.94 63.06
189 125 19
56.76 37.53 5.71
Nursing Experience (Years)c 30 Responsibility Ward/Unit In-charge None Hours worked on a typical dayd 12 c M = 12.59; SD = + 9.08 d M = 8.76; SD = + 8.62
Frequency
%
47
Instrument Reliability The internal consistencies of the Nurse Stress Index (NSI), Job Satisfaction Survey (JSS) and the Six-Dimension Scale of Nursing Performance (6-DSNP) were assessed using Cronbach’s alpha coefficients. The reliability estimates for all the three instruments were acceptable, ranging from .81 for the Job Satisfaction Survey to .93 for the Six-Dimension Scale of Nursing Performance, as shown in tables 3, 4 and 5. The Cronbach’s alphas for the six sub-scales of the NSI ranged from .54 for the Home and Work Conflicts scale to .80 for the Organizational Support and Involvement scale.
Table 3 Number of Items and Cronbach’s Alphas for the NSI Sub-scales Instrument Sub-scale
Number of items 5
Cronbach’s alpha
2. Managing Workload 2 (Pressures due to resources and conflicting priorities)
5
.78
3. Organizational Support and Involvement
5
.80
4. Dealing with Patients and Relatives
5
.77
5. Home and Work Conflicts
5
.54
6. Confidence and Competence in Role
5
.74
30
.92
1. Managing Workload 1 (Pressures due to insufficient time)
Total Score
.78
48 The JSS sub-scales Cronbach’s alphas ranged from -.02 for the Promotion sub-scale to .59 for the Supervision sub-scale. The Cronbach’s alpha for the total scale was acceptable at .81 as shown in table 4.
Table 4 Number of Items and Cronbach’s Alphas for the JSS Sub-scales Instrument Sub-scale
Number of items 4
Cronbach’s alpha
2. Promotion
4
-.02
3. Supervision
4
.59
4. Fringe Benefits
4
.37
5. Contingent Rewards
4
.33
6. Operating Conditions
4
.42
7. Co-workers
4
.49
8. Nature of Work
4
.54
9. Communication
4
.53
36
.81
1. Pay
Total Score
.46
49 The 6-DSNP scale examines both the self-rated frequency (Column A) and quality (Column B) of performance. The Cronbach’s alphas of the 6-DSNP Column B sub-scales ranged from .61 to .79 for the Leadership and Teaching/Collaboration subscales respectively. The total scale overall Cronbach’s alpha was .93.
Table 5 Number of Items and Cronbach’s Alphas for the 6-DSNP Sub-scales Instrument Sub-scale
Number of items 5
Cronbach’s alpha
2. Critical Care
7
.75
3. Teaching/Collaboration
11
.79
4. Planning/Evaluation
7
.74
5. Interpersonal Relations/ Communication
12
.77
6. Professional Development
10
.76
52
.93
1. Leadership
Total Score
.61
50
Descriptive Statistics for Instrument Sub-scales Nurse Stress Index The observed means for the NSI sub-scales ranged from 12.69 for the Leadership sub-scale to 14.42 for the Managing Workload 1 sub-scale. The overall mean score for the NSI was 82.18 with an SD of 21.63 as shown in table 6.
Table 6 Range of Possible Scores, Mean Scores and Standard Deviations for NSI Sub-scale 1. Managing Workload 1
Range of Possible Scores 5-25
Mean Scores for Sample 12.69
4.55
2. Managing Workload 2
5-25
14.42
4.72
3. Organizational Support and Involvement
5-25
14.14
5.21
4. Dealing with Patients and Relatives
5-25
14.23
4.34
5. Home and Work Conflicts
5-25
13.36
3.82
6. Confidence and Competence in Role
5-25
13.54
4.48
30-150
82.18
21.63
Total Score
SD
51
Job Satisfaction Survey The means for the JSS sub-scales ranged from 9.17 for the Fringe Benefits sub-scale to 18.80 for the Co-workers sub-scale. The overall mean score for the JSS was 127.65 with an SD of 19.22 as shown in table 7.
Table 7 Range of Possible Scores, Mean Scores and Standard Deviations for the JSS Sub-scale
Range of Possible Scores 4-24
Mean Scores for Sample 10.79
4.47
2. Promotion
4-24
13.43
3.55
3. Supervision
4-24
18.15
4.22
4. Fringe Benefits
4-24
9.17
3.9
5. Contingent rewards
4-24
11.25
4.11
6. Operating Conditions
4-24
12.21
3.12
7. Co-workers
4-24
18.80
3.85
8. Nature of Work
4-24
18.79
4.20
9. Communication
4-24
15.06
4.64
36-216
127.65
19.22
1. Pay
Total Score
SD
According to Spector (1997), participants can be assigned to satisfaction, ambivalent or dissatisfaction categories. For the 36-item total, where possible scores range from 36-216, the ranges 36-108 represent dissatisfaction, 108-144 ambivalence and 144-216 represent satisfaction. As shown in table 8, the majority of respondents were ambivalent (undecided) as to whether they were satisfied with their jobs or not
52 (68%) while 17.42% reported satisfaction with their job. Almost 15% reported dissatisfaction with their jobs.
Table 8 Level of Satisfaction for Job Satisfaction Survey (JSS) Level of Satisfaction
Frequency
Percentage
Dissatisfied
47
14.11
Ambivalent
228
68.47
Satisfied
58
17.42
Note: Score ranges 36-107 = Dissatisfaction, 108-143 = Ambivalent, and 144-216 = Satisfaction (Spector, 1997).
Nurse Performance Scale Two separate measures from the 6-DSNP were calculated for each subject: the total 52 item scale score for Column B (performance quality) and a mean of the sixsubscale mean scores for Column B. The measures were highly correlated (r=.997, p 5 (n = 29) 2.97 *Mean for six sub-scales of the 6-DSNP
.39 .43 .45 .36
F = 3.86, p = .010
Job Performance and Work Characteristics As shown in Table 16, the public hospital had the lowest mean score for job performance (Mean = 2.9, SD = .45) and private hospital 1 has the highest (M = 3.32,
62 SD .40). There was a significant difference in means for the different hospitals (F = 7.95, p < .001). The means for nursing education ranged from 2.84 (SD = .49) for the Registered Midwives to 3.19 (SD = .44).
Table 16 Mean Scores for Job Performance by Work Characteristics Characteristic Hospital Public (n = 196) Private 1 (n = 25) Private 2 (n = 71) Private 3(n = 41)
Mean*
SD
2.96 3.32 3.15 3.12
.45 .40 .34 .38
Nursing Education Enrolled Nurse (n = 77) Enrolled Midwife (n = 42) Registered Nurse (n = 100) Registered Midwife (n = 49) Double Trained (n = 50) BSN and above (n = 15)
3.12 3.02 3.11 2.84 3.00 3.19
.36 .37 .422 .49 .47 .44
Ward/Unit Medical (n = 117) Surgical (n = 57) Obs./Gyn. (n = 102) Pediatrics (n = 57)
3.11 2.98 2.95 3.17
.42 .43 .45 .37
ANOVA (p value F = 7.95, p < .0001
F = 3.60, p = .004
F = 4.34, p = .005
Nursing Experience (Years) 1-5 (n = 83) 6-10 (n =92) 11-15 (n = 56) 16-20 (n = 46) 21+ (n = 56)
F = 1.46, p = .214 3.13 3.02 3.06 2.95 3.03
.40 .45 .35 .47 .47
Responsibility Ward/Unit In-charge (n = 123) None (n = 210)
3.05 3.05
.44 .43
F = 0.0006, p = .980
*Mean for six sub-scales of the 6-DSNP
63 The results indicated that there were significant differences in means for nurse performance among the nursing education groups (F = 3.60, p = .004) and type of ward/unit (F = 4.34, p = .005). There were no differences in means for nurse performance between nurses who had extra responsibility (Ward/Unit in-charges or Charge nurses) and those without extra responsibility (F = .0006, p .980) or for nursing experience (F = 1.460, p = .214). The analyses presented above were on the mean subscale score for the 6DSNP. The same descriptive analyses by demographic and work characteristics were repeated on the 52-item total score. The results were congruent regarding statistically significant group differences. (Results are not shown).
Findings Related to Research Questions This section presents results of the study in relation to the research questions. Pearson correlation was utilized to answer question one, while multiple regression analyses were utilized to answer questions two and three. The fourth and fifth questions were answered using one-way Analysis of Variance (ANOVA).
Research Question 1 What is the relationship between occupational stress and job performance among hospital nurses in Kampala, Uganda? The relationship between occupational stress and job performance of the nurses was investigated using Pearson correlation coefficient. As indicated in Table 17, the results demonstrated a significant negative relationship between occupational stress (as measured by the NSI) and job performance, measured as how well the participants performed their activities (r = -.131, p = .018). This indicated that higher
64 stress levels were associated with lower levels of self-rated job performance quality. However, no significant relationship was found (r = -.018, p = .746) between occupational stress and job performance as measured by the self-rated frequency of activities (how often the participants performed the nursing activities). Research Question 2 Is there a relationship between occupational stress and job satisfaction among hospital nurses in Kampala, Uganda? The relationship between occupational stress and job satisfaction (JSS) of the nurses was also investigated using Pearson correlation coefficient, as shown in table 17. A significant inverse relationship was found between occupational stress and job satisfaction (r = -.501, p = .000). This indicated that high stress levels resulted in low job satisfaction.
Table 17 Correlations for Job Performance, Job Satisfaction with Occupational Stress r
p value
Column A (How frequent)
-.018
.746
Column B (How well)
-.131
.018
-.501
.000
Job Performance
Job Satisfaction
65 Research Question 3 Do personal background characteristics affect the relationships between occupational stress, job satisfaction and job performance among hospital nurses in Kampala, Uganda? Regression analyses were performed to investigate whether personal and work characteristics (nursing education, nursing experience, type of hospital, ward/unit, and number of children) affect the relationships between occupational stress, job satisfaction and job performance. Number of children was used a proxy indicator for family responsibility. The response variable was job performance, measured using the mean of the six sub-scale means of the 6-DSNP (Column B). The primary predictor variables were occupational stress (as measured by the NSI) and job satisfaction (as measured by the JSS). Both predictors were significant in separate simple linear regression models (R2 = .021, F = 6.87, p = .009; R2 = .033, F = 11.42, p = .001, respectively). The covariates included in the initial multiple regression model were selected based on significant bivariate relationships with job performance. Based on these analyses (see tables 15 and 16), the only personal characteristic included was number of children. The work characteristics included as covariates were type of hospital and ward and nursing education. Based on the ANOVA results, hospitals were grouped into public and private for analysis. Reference-cell coded indicator variables (Unit 1, Unit 2 and Unit 3) were created to represent the different type of wards, with the medical ward as the reference group. Nursing education was recorded into four groups: enrolled nurses/enrolled midwives, registered nurses/registered midwives, double trained, and BSN and above.
66 Preliminary examination of the data for normality, linearity, and homogeneity of variances were conducted using standard techniques and no serious violations were noted. Initial models were also evaluated for problems with multicollinearity using variance inflation factor (VIF) values, and again no problems were noted (Munro, 2001; Pallant, 2001). The set of covariates was entered in the first block, followed by occupational stress at the second step, then job satisfaction at the third step. Individual predictors and the changes in adjusted R-squared values for each step were evaluated for statistical significance (see table 18).
Table 18 Effect of Personal Background and Work Characteristics on the Relationships of Occupational Stress, Job Satisfaction and Job Performance Model Constant
Beta
t 10.856
P value .000
Number of children
-.094
-1.688
.092
Type of hospital
-.187
-2.911
.004
Unit 1(Surgical)*
-.057
-.956
.340
Unit 2 (Obs/Gyn)*
-.142
-2.322
.o21
Unit 3(Pediatrics)*
.084
1.415
.158
Nursing Educational Level
.092
1.565
.119
Occupational stress
-.019
-.302
.763
Job satisfaction
.141
2.278
.023
* Reference group – medical ward
67 Several smaller models were evaluated. As shown in Table 19, the best predictive model for job performance included type of hospital (public/private), type of ward/unit, and job satisfaction. The model including type of ward/unit and type of hospital contributed approximately 8% of the variance in job performance. When job satisfaction was added to the model with type of unit and type of hospital, the model accounted for approximately 10% of the variance in job performance.
Table 19 The Final Predictive Model for Self-Rated Job Performance Quality Model
Beta
Constant
t
P value
16.439
.000
Type of hospital
-.190
-3.425
.001
Unit 1 (Surgical)
-.060
-1.020
.308
Unit 2 (Obs/gyn)
-.155
-2.578
.010
Unit 3 (Pediatric)
.076
1.302
.194
Job satisfaction
.139
2.526
.012
All models were re-run with the job performance 52-item total score as the outcome. The predictors selected for the initial model based on bivariate analyses were the same. No differences in significant predictors in the initial and final models were noted. (Results are not shown.)
Research Question 4 Does job satisfaction mediate the relationship between occupational stress and job performance among hospital nurses in Kampala, Uganda?
68 The potential mediating role of job satisfaction between occupational stress and job performance was examined by conducting step-wise multiple regressions (Baron & Kenny, 1986) with occupational stress as the independent variable, job satisfaction as the potential mediator and job performance as the dependent variable. For this analysis, the 52 item total score for the 6-DSNP (Column B) was used. In step 1, a simple regression analysis with occupational stress (X) predicting job performance (Y) was conducted (Y = B0 + B1X + e) and it indicated a significant relationship (beta = -.124, t = -2.250, p =.025). In step 2, a simple regression analysis conducted with occupational stress predicting job satisfaction (Z) represented by Z = B0 + B1X + e also showed a significant relationship (beta = -.501, t = -10.39, p
